New Therapies

Multiple sclerosis (MS) is the most common demyelinating disease of the central nervous system, with recent estimates of around 1 million people living with MS in the US.¹ In many countries, MS is a leading cause of disability among young adults, second only to trauma.² Clinically, neurologic worsening (ie, disability) in MS can occur in the relapsing-remitting (RRMS) phase of disease due to incomplete recovery from neuroinflammatory relapses. However, in the 15% of patients with a progressive course from onset (PPMS), and in those with RRMS who transition to a secondary progressive phenotype (SPMS), neurologic worsening follows a slowly progressive pattern.³ A progressive disease course—either PPMS at onset or transitioning to SPMS—is the dominant factor affecting MS-related neurologic disability accumulation. In particular, epidemiologic studies have shown that, in the absence of transitioning to a progressive disease course, < 5% of individuals with MS will accumulate sufficient disability to necessitate use of a cane for ambulation.^4-6 Therefore, developing disease modifying therapies (DMTs) that are highly effective at slowing or stopping the gradual accumulation of neurologic disability in progressive MS represent a critical unmet need.

Research into the development of DMTs for progressive MS has been hindered by a number of factors. In particular, the clinical definition and diagnosis of progressive MS has been an evolving concept. Diagnostic criteria for MS, which help facilitate the enrollment of appropriate subjects into clinical trials, have only recently clarified the current consensus definition for progressive MS—steadily increasing neurologic disability independent of clinical relapses. Looking back to the Schumacher criteria in 1965 and Poser criteria in 1983, it was acknowledged that neurologic symptoms in MS may follow a relapsing-remitting or progressive pattern, but little attempt was made to define progressive MS.^7,8 The original McDonald criteria in 2001 defined diagnostic criteria for progressive MS.⁹ These criteria continued to evolve through subsequent revisions (eg, cerebrospinal fluid [CSF] specific oligoclonal bands no longer are an absolute requirement), and only in the 2017 revision was it emphasized that disability progression must occur independent of clinical relapses, concordant with similar emphasis in the 2013 revision of MS clinical course definitions.^3,10

The interpretation of prior clinical trials of DMT for progressive MS must consider this evolving clinical definition. The US Food and Drug Administration (FDA) approved mitoxantrone in 2000—making it the first DMT to carry an approved label for SPMS. While achieving significant clinical efficacy, it is clear from the details of the trial that the enrolled subjects had a high degree of inflammatory disease activity, which suggests that mitoxantrone treats neuroinflammation and not relapse-independent worsening. More recently, disparate results were seen in the anti-CD20 (rituximab, ocrelizumab) and S1P receptor modulator (fingolimod, siponimod) trials targeted at patients with primary and secondary progressive MS.^11-14 Although there are differences between these therapies, they are more similar than not within the same therapeutic class. Taken together, these trials illustrate the critical impact the narrower inclusion/exclusion criteria (namely age and extent of inflammatory activity) had on attaining positive outcomes. Other considerations, such as confounding illness, also may impact trial recruitment and generalizability of findings.

The lack of known biological targets in progressive MS, which is a complex disease with an incompletely understood and heterogeneous pathology, also hinders DMT development. Decades of research has characterized multifocal central nervous system (CNS) lesions that exhibit features of demyelination, inflammation, reactive gliosis, axonal loss, and neuronal damage. Until recently, however, much of this research focused on the relapsing phase of disease, and so the understanding of the pathologic underpinnings of progressive disease has remained limited. Current areas of investigation encompass a broad range of pathological processes, such as microglial activation, meningeal lymphoid follicles, remyelination failure, vulnerability of chronically demyelinated axons, oxidative injury, iron accumulation, mitochondrial damage, and others. There is the added complication that the pathologic processes underlying progressive MS are superimposed on the CNS aging process. In particular, the transition to progressive MS and the rate of disability accumulation during progressive MS show strong correlation with age.^6,15-17

Finally, DMT development for progressive MS also is hindered by the lack of specific surrogate and clinical outcome measures. Trials for relapsing MS have benefited greatly from the relatively straightforward assessment of clinical relapses and inflammatory disease activity on magnetic resonance imaging (MRI). With the goal of developing DMTs that are highly effective at slowing or stopping the gradual accumulation of neurologic disability in progressive MS, which by definition occurs independent of clinical relapses, these measures are not directly relevant. The longitudinal clinical disability outcome measures change at a slower rate than in early, relapsing disease. The use of standardized scales (eg, the Expanded Disability Status Scale [EDSS]), lower limb function, upper limb function, cognition, or a combination is a subject of ongoing debate. For example, the ASCEND and IMPACT trials (placebo-controlled trials for SPMS with natalizumab and interferon β-1a, respectively) showed no significant impact in EDSS progression—but in both of these trials, the 9-hole peg test (9-HPT), a performance measure for upper limb function, showed significant improvement.^10,18 Particularly in those with an EDSS of > 6.5, who are unlikely to have measurable EDSS progression, functional tests such as the 9-HPT or timed 25-foot walk may be more sensitive as measures for disability progression.¹¹ MRI measures of brain atrophy is the current gold standard surrogate outcome for clinical trials in progressive MS, but others that may warrant consideration include optical coherence tomography (OCT) or CSF markers of axonal degeneration.

DMT for Progressive MS

Current diagnostic nomenclature separates patients with active (superimposed clinical and/or radiographic relapses) from those with inactive disease.^3,12 Relapsing forms of MS include all RRMS and those with SPMS with superimposed relapses (ie, active SPMS). Following this paradigm shift, the FDA changed the indication for already approved DMT from RRMS to relapsing forms of MS. Below is a discussion of DMT that specifically use the term SPMS and PPMS in the indication, where phase 3 trial data for progressive MS is available.

In 2019, the FDA approved the first oral medication (siponimod) for active SPMS. Subsequently, updates to the labels of the older DMT expanded to include active SPMS. Until 2019, the only FDA approved medication for SPMS was mitoxantrone, and use of this medication was limited due to unfavorable adverse effects (AEs). No medications had obtained FDA approval for inactive SPMS to this point, which represented an unmet need for a considerable number of patients.

Mitoxantrone became the first DMT approved for use in SPMS in 2000 following early trials that showed reductions in EDSS worsening, change in ambulation index, reduced number of treated relapses, and prolonged time to first treated relapse. However, as with some of the other positive trials in progressive MS, it is difficult to discern the impact of suppression of relapses as opposed to direct impact on progressive pathophysiology. Within the placebo arm, for example, there were 129 relapses among the 64 subjects, which suggests that these cases had particularly active disease or were in the early stages of SPMS.¹³ Furthermore, the use of this medication was limited due to concerns of cardiotoxicity and hematologic malignancy as serious AEs.

The trials of interferon β-1b illustrate the same difficulty of isolating possible benefits in disease progression from disability accumulated from relapses. The first interferon β-1b trial for SPMS, was conducted in Europe using fingolimod and showed a delay in confirmed disability progression compared to placebo as measured with the EDSS.¹⁴ However, a North American trial that followed in 2004 was unable to replicate this finding.¹⁵ The patients in the European trial appeared to be in an earlier phase of SPMS with more active disease, and a post-hoc pooled analysis suggested that patients with active disease and those with more pronounced disability progression were more likely to benefit from treatment.¹⁶ Overall, interferons do not appear to appreciably alter disability in the long-term for patients with SPMS, though they may modify short-term, relapse-related disability.

Perhaps the most encouraging data for SPMS was found in the EXPAND trial, which investigated siponimod, an S1P receptor modulator that is more selective than fingolimod. The trial identified a 21% reduction in 3-month confirmed disability progression for SPMS patients taking siponimod compared with those taking a placebo.¹⁷ Although the patients in EXPAND did seem to have relatively less disease activity at baseline than those who participated in other SPMS trials, those who benefitted from siponimod were primarily patients who had clinical and/or radiographic relapses over the prior 2 years. Based on this, the FDA approved siponimod for active SPMS. The extent to which siponimod exerts a true neuroprotective effect beyond reducing inflammation has not been clearly established.

B-cell depleting therapies rituximab and ocrelizumab have been evaluated in both primary and secondary progressive MS populations. Early investigations of the chimeric rituximab in PPMS did not show benefits on disability (EDSS) progression; however, benefits were seen in analysis of some subgroups.¹⁸ With this in mind, the ORATORIO trial for the humanized version, ocreluzimab, included PPMS patients that were younger (aged < 55 years) and had cutoffs for disease duration (< 15 years for those with EDSS more than 5 years, < 10 years for those with EDSS less than 5 years). The study showed statistically significant changes on disability progression, which led to ocrelizumab receiving the first FDA indication for PPMS.¹¹ There are substantial pathophysiologic similarities between PPMS and SPMS in the progressive phase.¹⁹ While these medications may have similar effects across these disease processes, these benefits have not yet been demonstrated in a prospective trial for the SPMS population. Regardless, B-cell depleting therapy is a reasonable consideration for select patients with active SPMS, consistent with a relapsing form of MS.

Therapies in Development 

DMT development for progressive MS is a high priority area. Current immunomodulatory therapies for RRMS have consistently been ineffective in the inactive forms of MS, with the possible exceptions of ocrelizumab and siponimod. Therefore, instead of immunosuppression, many agents currently in phase 2 and 3 clinical trials target alternative pathophysiological processes in order to provide neuroprotection, and/or promote remyelination and neurogenesis. These targets include oxidative stress (OS), non-T cell mediated inflammation, and mitochondrial/energy failure.²⁰ Below we review a selection of clinical trials testing agents following these approaches. Many agents have more than one potential mechanism of action (MOA) that could benefit progressive MS.

Lipoic acid (LA), also known as α-lipoic acid and thiotic acid, is one such agent targeting alternative pathophysiology in SPMS. LA is an endogenous agent synthesized de novo from fatty acids and cysteine as well as obtained in small amounts from foods.²¹ It has antioxidant (AO) properties including direct radical scavenging, regeneration of glutathione, and upregulation of AO enzymes via the NrF2 pathway.²² It supports mitochondria as a key cofactor for pyruvate dehydrogenase and α-ketoglutarate dehydrogenase, and it also aids mitochondrial DNA synthesis.^21,22 Studies in experimental autoimmune encephalomyelitis, a widely used experimental mouse model of inflammatory demyelinating disease, also indicate a reduction in excessive microglial activation.²³ A phase 2 pilot randomized controlled trial (RCT) of 1200 mg LA in SPMS (n = 51) resulted in significantly less whole brain atrophy by SIENA (Structural Image Evaluation, Using Normalization, of Atrophy) at 2 years.²⁴ A follow-up multicenter trial is ongoing.

Simvastatin also targets alternative pathophysiology in SPMS. It has anti-inflammatory effects, improves vascular function, and promotes neuroprotection by reducing excitotoxicity. A phase 2 RCT demonstrated a reduction in whole brain atrophy in SPMS (n = 140), and a phase 3 trial is underway.²⁵ Ibudilast is another repurposed drug that targets alternative inflammation by inhibiting several cyclic nucleotide phosphodiesterases, macrophage migration inhibitory factor and toll-like receptor 4. A phase 2 trial (n = 225) in both SPMS and PPMS also demonstrated a reduction in brain atrophy, but participants had high rates of AEs.²⁶

Lithium and riluzole promote neuroprotection by reducing excitotoxicity. Lithium is a pharmacologic active cation used as a mood stabilizer and causes inhibition of glycogen synthase kinase-3β. Animal models also indicate that lithium may decrease inflammation and positively impact neurogenesis.²⁷ A crossover pilot trial demonstrated tolerability with trends toward stabilization of EDSS and reductions in brain atrophy.²⁸ Three neuroprotective agents, riluzole (reduces glutamate excitotoxicity), fluoxetine (stimulates glycogenolysis and improves mitochondrial energy production), and amiloride (an acid-sensing ion channel blocker that opens in response to inflammation) were tested in a phase 2b multi-arm, multi-site parallel group RCT in SPMS (n = 445). The study failed to yield differences from placebo for any agent in reduction of brain volume loss.²⁹ A prior study of lamotrigine, a sodium channel blocker, also failed to find changes in brain volume loss.³⁰ These studies highlight the large sample sizes and/or long study durations needed to test agents using brain atrophy as primary outcome. In the future, precise surrogate markers of neuroprotection will be a great need for earlier phase trials. These results also suggest that targeting > 1 MOA may be necessary to treat SPMS effectively.

Efforts to promote remyelination target one hallmark of MS damage. High dose biotin (about 10,000× usual dose) may promote myelin repair as a cofactor for fatty acid synthesis and support mitochondrial oxidative phosphorylation. While a RCT yielded a greater proportion of participants with either PPMS or SPMS with improvement in disability than placebo at 12 months, an open label trial suggested otherwise indicating a need for a more definitive trial.^31,32

Anti-LINGO-1 (opicinumab) is a monoclonal antibody that targets LINGO, a potent negative regulator of oligodendrocyte differentiation and myelination.³³ Although this agent failed in a phase 2 trial in relapsing MS, and is thus unlikely to be tested in progressive forms, the innovative approach to stimulating oligodendrocytes is ongoing. One such effort is to use thyroid hormone, crucial to myelin formation during development, as a repair agent in MS.³⁴ A dose-finding study of thyroid hormone was completed and efforts to develop a thyromimetic agent are ongoing.

Finally, efforts to promote neurogenesis remain a goal for many neurodegenerative diseases. Exercise appears to prevent age-related atrophy of the hippocampus in animals and humans and help maintain neuronal health.³⁵ In patients with RRMS, cortical thickness is impacted positively by resistance training, which suggests a neuroprotective effect.³⁶ A multi-center trial of exercise in patients with progressive MS investigating cognitive outcomes is ongoing.

Discontinuing DMT

In the early 1990s, the successful development of immune modulating therapies that reliably reduced disease activity in RRMS led to widespread initiation in patients with relapsing disease. However, guidance on when or if to discontinue DMT, even in those who have transitioned to SPMS, remains largely absent at this time. Requests to discontinue DMT may come from patients weary of taking medication (especially injections), bothered by AEs, or those who no longer perceive efficacy from their treatments. Clinicians also may question the benefit of immune modulation in patients with longstanding freedom from relapses or changes in MRI lesion burden.

To inform discussion centered on treatment discontinuation, a clinical trial is currently underway to better answer the question of when and how to withdraw MS therapy. Discontinuation of Disease Modifying Therapies in Multiple Sclerosis (DISCO-MS) is a prospective, placebo-controlled RCT and its primary endpoint is recurrence of disease activity over a 2 year follow-up period.³⁷ Eligibility requirements for the trial include age > 55 years, 5-year freedom from relapses, and 3-year freedom from new MRI lesions (criteria informed by progressive MS cohort studies).³¹ In addition to demonstrating the active disease recurrence rates in this patient population, the trial also aims to identify risk factors for recurrent disease activity among treated MS patients.³⁷ DISCO-MS builds upon a series of retrospective and observational studies that partially answered these questions, albeit in the context of biases inherent in retrospective or observational studies.

A Minneapolis MS Treatment and Research Center single-center study identified 77 SPMS patients with no acute CNS inflammatory events over 2 to 20 years and advised these patients to stop taking DMT.³² In this group, 11.7% of subjects experienced recurrent active disease. Age was the primary discriminating factor. The mean age of those who experienced disease activity was 56 years vs 61 years those who did not. A second observational study from France found that among 100 SPMS patients treated either with interferon β or glatiramer acetate for at least 6 months, 35% experienced some form of inflammatory disease upon discontinuation.³⁸ Sixteen patients relapsed and 19 developed gadolinium-enhancing MR lesions after DMT discontinuation. However, the age of the cohort was younger than the Minneapolis study (47.2 years vs 61 years), and reasons for discontinuation (eg, AEs or lack of disease activity) were not considered in the analysis.

Other studies examining the safety of DMT discontinuation have not considered MS subtype or excluded patients with progressive subtypes of MS. The largest studies to date on DMT discontinuation utilized the international MSBase global patient registry, which identified nearly 5,000 patients who discontinued interferons (73%), glatiramer acetate (18%), natalizumab (6%), or fingolimod (3%), without specifying the reasons for discontinuation.³⁹ Despite these shortcomings, data reveal trends that are helpful in predicting how MS tends to behave in patients who have discontinued therapy. Not surprisingly, disability progression was most likely among patients already characterized as having a progressive phenotype, while relapses were less likely to occur among older, progressive patients.

Although clinicians may be increasingly willing to discuss DMT discontinuation with their patients, at least 1 study exploring patient perspectives on stopping treatment suggests widespread reluctance to stop treatment. A survey conducted with participants in the North American Research Committee on Multiple Sclerosis patient-report registry found that among survey respondents, only 11.9% would discontinue their MS medication if deemed stable, while 66.3% stated they were unlikely to stop treatment.⁴⁰

These results suggest that before clinicians incorporate DMT discontinuation into the normal course of discussion with patients, they should be prepared to provide both education (on the wisdom of stopping under the right circumstances) and evidence to support when and how to make the recommendation. Based on existing evidence, criteria for recommending treatment discontinuation might include prolonged freedom from disease activity (≥ 5 years), age > 55 years or 60 years, and a progressive disease course. Thus far, no combination of factors has been shown to completely predict an event-free transition off of medicine. Since no fixed algorithm yet exists to guide DMT stoppage in MS, reasonable suggestions for monitoring patients might include surveillance MRIs, more frequent clinic visits, and possible transitional treatment for patients coming off of natalizumab or fingolimod, since these drugs have been associated with rebound disease activity when discontinued.^41,42

Clinicians wishing to maximize function and quality of life for their patients at any age or stage of disease should look to nonpharmacologic interventions to lessen disability and maximize quality of life. While beyond the scope of this discussion, preliminary evidence suggests multimodal (aerobic, resistance, balance) exercise may enhance endurance and cognitive processing speed, and that treatment of comorbid diseases affecting vascular health benefits MS. ⁴³

Conclusions

The development of numerous treatments for RRMS has established an entirely new landscape and disease course for those with MS. While this benefit has not entirely extended to those with progressive MS, those with active disease with superimposed relapses may receive limited benefit from these medications. New insights into the pathophysiology of progressive MS may lead us to new treatments through multiple alternative pathophysiologic pathways. Some early studies using this strategy show promise in slowing the progressive phase. Medication development for progressive MS faces multiple challenges due to lack of a single animal model demonstrating both pathology and clinical effects, absence of phase 1 surrogate biomarkers, and later phase trial endpoints that require large sample sizes and extended study durations. Nevertheless, the increase in number of trials and diversity of therapeutic approaches for progressive MS provides hope for effective therapy. Currently, the heterogeneity of the population with progressive MS requires an individualized treatment approach, and in some of these patients, stopping therapy may be a reasonable consideration. Symptomatic management remains critical for all patients with progressive MS as well as non-pharmacologic approaches that maximize quality of life.

Multiple sclerosis (MS) is the most common demyelinating disease of the central nervous system, with recent estimates of around 1 million people living with MS in the US.¹ In many countries, MS is a leading cause of disability among young adults, second only to trauma.² Clinically, neurologic worsening (ie, disability) in MS can occur in the relapsing-remitting (RRMS) phase of disease due to incomplete recovery from neuroinflammatory relapses. However, in the 15% of patients with a progressive course from onset (PPMS), and in those with RRMS who transition to a secondary progressive phenotype (SPMS), neurologic worsening follows a slowly progressive pattern.³ A progressive disease course—either PPMS at onset or transitioning to SPMS—is the dominant factor affecting MS-related neurologic disability accumulation. In particular, epidemiologic studies have shown that, in the absence of transitioning to a progressive disease course, < 5% of individuals with MS will accumulate sufficient disability to necessitate use of a cane for ambulation.^4-6 Therefore, developing disease modifying therapies (DMTs) that are highly effective at slowing or stopping the gradual accumulation of neurologic disability in progressive MS represent a critical unmet need.

Research into the development of DMTs for progressive MS has been hindered by a number of factors. In particular, the clinical definition and diagnosis of progressive MS has been an evolving concept. Diagnostic criteria for MS, which help facilitate the enrollment of appropriate subjects into clinical trials, have only recently clarified the current consensus definition for progressive MS—steadily increasing neurologic disability independent of clinical relapses. Looking back to the Schumacher criteria in 1965 and Poser criteria in 1983, it was acknowledged that neurologic symptoms in MS may follow a relapsing-remitting or progressive pattern, but little attempt was made to define progressive MS.^7,8 The original McDonald criteria in 2001 defined diagnostic criteria for progressive MS.⁹ These criteria continued to evolve through subsequent revisions (eg, cerebrospinal fluid [CSF] specific oligoclonal bands no longer are an absolute requirement), and only in the 2017 revision was it emphasized that disability progression must occur independent of clinical relapses, concordant with similar emphasis in the 2013 revision of MS clinical course definitions.^3,10

The interpretation of prior clinical trials of DMT for progressive MS must consider this evolving clinical definition. The US Food and Drug Administration (FDA) approved mitoxantrone in 2000—making it the first DMT to carry an approved label for SPMS. While achieving significant clinical efficacy, it is clear from the details of the trial that the enrolled subjects had a high degree of inflammatory disease activity, which suggests that mitoxantrone treats neuroinflammation and not relapse-independent worsening. More recently, disparate results were seen in the anti-CD20 (rituximab, ocrelizumab) and S1P receptor modulator (fingolimod, siponimod) trials targeted at patients with primary and secondary progressive MS.^11-14 Although there are differences between these therapies, they are more similar than not within the same therapeutic class. Taken together, these trials illustrate the critical impact the narrower inclusion/exclusion criteria (namely age and extent of inflammatory activity) had on attaining positive outcomes. Other considerations, such as confounding illness, also may impact trial recruitment and generalizability of findings.

The lack of known biological targets in progressive MS, which is a complex disease with an incompletely understood and heterogeneous pathology, also hinders DMT development. Decades of research has characterized multifocal central nervous system (CNS) lesions that exhibit features of demyelination, inflammation, reactive gliosis, axonal loss, and neuronal damage. Until recently, however, much of this research focused on the relapsing phase of disease, and so the understanding of the pathologic underpinnings of progressive disease has remained limited. Current areas of investigation encompass a broad range of pathological processes, such as microglial activation, meningeal lymphoid follicles, remyelination failure, vulnerability of chronically demyelinated axons, oxidative injury, iron accumulation, mitochondrial damage, and others. There is the added complication that the pathologic processes underlying progressive MS are superimposed on the CNS aging process. In particular, the transition to progressive MS and the rate of disability accumulation during progressive MS show strong correlation with age.^6,15-17

Finally, DMT development for progressive MS also is hindered by the lack of specific surrogate and clinical outcome measures. Trials for relapsing MS have benefited greatly from the relatively straightforward assessment of clinical relapses and inflammatory disease activity on magnetic resonance imaging (MRI). With the goal of developing DMTs that are highly effective at slowing or stopping the gradual accumulation of neurologic disability in progressive MS, which by definition occurs independent of clinical relapses, these measures are not directly relevant. The longitudinal clinical disability outcome measures change at a slower rate than in early, relapsing disease. The use of standardized scales (eg, the Expanded Disability Status Scale [EDSS]), lower limb function, upper limb function, cognition, or a combination is a subject of ongoing debate. For example, the ASCEND and IMPACT trials (placebo-controlled trials for SPMS with natalizumab and interferon β-1a, respectively) showed no significant impact in EDSS progression—but in both of these trials, the 9-hole peg test (9-HPT), a performance measure for upper limb function, showed significant improvement.^10,18 Particularly in those with an EDSS of > 6.5, who are unlikely to have measurable EDSS progression, functional tests such as the 9-HPT or timed 25-foot walk may be more sensitive as measures for disability progression.¹¹ MRI measures of brain atrophy is the current gold standard surrogate outcome for clinical trials in progressive MS, but others that may warrant consideration include optical coherence tomography (OCT) or CSF markers of axonal degeneration.

DMT for Progressive MS

Current diagnostic nomenclature separates patients with active (superimposed clinical and/or radiographic relapses) from those with inactive disease.^3,12 Relapsing forms of MS include all RRMS and those with SPMS with superimposed relapses (ie, active SPMS). Following this paradigm shift, the FDA changed the indication for already approved DMT from RRMS to relapsing forms of MS. Below is a discussion of DMT that specifically use the term SPMS and PPMS in the indication, where phase 3 trial data for progressive MS is available.

In 2019, the FDA approved the first oral medication (siponimod) for active SPMS. Subsequently, updates to the labels of the older DMT expanded to include active SPMS. Until 2019, the only FDA approved medication for SPMS was mitoxantrone, and use of this medication was limited due to unfavorable adverse effects (AEs). No medications had obtained FDA approval for inactive SPMS to this point, which represented an unmet need for a considerable number of patients.

Mitoxantrone became the first DMT approved for use in SPMS in 2000 following early trials that showed reductions in EDSS worsening, change in ambulation index, reduced number of treated relapses, and prolonged time to first treated relapse. However, as with some of the other positive trials in progressive MS, it is difficult to discern the impact of suppression of relapses as opposed to direct impact on progressive pathophysiology. Within the placebo arm, for example, there were 129 relapses among the 64 subjects, which suggests that these cases had particularly active disease or were in the early stages of SPMS.¹³ Furthermore, the use of this medication was limited due to concerns of cardiotoxicity and hematologic malignancy as serious AEs.

The trials of interferon β-1b illustrate the same difficulty of isolating possible benefits in disease progression from disability accumulated from relapses. The first interferon β-1b trial for SPMS, was conducted in Europe using fingolimod and showed a delay in confirmed disability progression compared to placebo as measured with the EDSS.¹⁴ However, a North American trial that followed in 2004 was unable to replicate this finding.¹⁵ The patients in the European trial appeared to be in an earlier phase of SPMS with more active disease, and a post-hoc pooled analysis suggested that patients with active disease and those with more pronounced disability progression were more likely to benefit from treatment.¹⁶ Overall, interferons do not appear to appreciably alter disability in the long-term for patients with SPMS, though they may modify short-term, relapse-related disability.

Perhaps the most encouraging data for SPMS was found in the EXPAND trial, which investigated siponimod, an S1P receptor modulator that is more selective than fingolimod. The trial identified a 21% reduction in 3-month confirmed disability progression for SPMS patients taking siponimod compared with those taking a placebo.¹⁷ Although the patients in EXPAND did seem to have relatively less disease activity at baseline than those who participated in other SPMS trials, those who benefitted from siponimod were primarily patients who had clinical and/or radiographic relapses over the prior 2 years. Based on this, the FDA approved siponimod for active SPMS. The extent to which siponimod exerts a true neuroprotective effect beyond reducing inflammation has not been clearly established.

B-cell depleting therapies rituximab and ocrelizumab have been evaluated in both primary and secondary progressive MS populations. Early investigations of the chimeric rituximab in PPMS did not show benefits on disability (EDSS) progression; however, benefits were seen in analysis of some subgroups.¹⁸ With this in mind, the ORATORIO trial for the humanized version, ocreluzimab, included PPMS patients that were younger (aged < 55 years) and had cutoffs for disease duration (< 15 years for those with EDSS more than 5 years, < 10 years for those with EDSS less than 5 years). The study showed statistically significant changes on disability progression, which led to ocrelizumab receiving the first FDA indication for PPMS.¹¹ There are substantial pathophysiologic similarities between PPMS and SPMS in the progressive phase.¹⁹ While these medications may have similar effects across these disease processes, these benefits have not yet been demonstrated in a prospective trial for the SPMS population. Regardless, B-cell depleting therapy is a reasonable consideration for select patients with active SPMS, consistent with a relapsing form of MS.

Therapies in Development 

DMT development for progressive MS is a high priority area. Current immunomodulatory therapies for RRMS have consistently been ineffective in the inactive forms of MS, with the possible exceptions of ocrelizumab and siponimod. Therefore, instead of immunosuppression, many agents currently in phase 2 and 3 clinical trials target alternative pathophysiological processes in order to provide neuroprotection, and/or promote remyelination and neurogenesis. These targets include oxidative stress (OS), non-T cell mediated inflammation, and mitochondrial/energy failure.²⁰ Below we review a selection of clinical trials testing agents following these approaches. Many agents have more than one potential mechanism of action (MOA) that could benefit progressive MS.

Lipoic acid (LA), also known as α-lipoic acid and thiotic acid, is one such agent targeting alternative pathophysiology in SPMS. LA is an endogenous agent synthesized de novo from fatty acids and cysteine as well as obtained in small amounts from foods.²¹ It has antioxidant (AO) properties including direct radical scavenging, regeneration of glutathione, and upregulation of AO enzymes via the NrF2 pathway.²² It supports mitochondria as a key cofactor for pyruvate dehydrogenase and α-ketoglutarate dehydrogenase, and it also aids mitochondrial DNA synthesis.^21,22 Studies in experimental autoimmune encephalomyelitis, a widely used experimental mouse model of inflammatory demyelinating disease, also indicate a reduction in excessive microglial activation.²³ A phase 2 pilot randomized controlled trial (RCT) of 1200 mg LA in SPMS (n = 51) resulted in significantly less whole brain atrophy by SIENA (Structural Image Evaluation, Using Normalization, of Atrophy) at 2 years.²⁴ A follow-up multicenter trial is ongoing.

Simvastatin also targets alternative pathophysiology in SPMS. It has anti-inflammatory effects, improves vascular function, and promotes neuroprotection by reducing excitotoxicity. A phase 2 RCT demonstrated a reduction in whole brain atrophy in SPMS (n = 140), and a phase 3 trial is underway.²⁵ Ibudilast is another repurposed drug that targets alternative inflammation by inhibiting several cyclic nucleotide phosphodiesterases, macrophage migration inhibitory factor and toll-like receptor 4. A phase 2 trial (n = 225) in both SPMS and PPMS also demonstrated a reduction in brain atrophy, but participants had high rates of AEs.²⁶

Lithium and riluzole promote neuroprotection by reducing excitotoxicity. Lithium is a pharmacologic active cation used as a mood stabilizer and causes inhibition of glycogen synthase kinase-3β. Animal models also indicate that lithium may decrease inflammation and positively impact neurogenesis.²⁷ A crossover pilot trial demonstrated tolerability with trends toward stabilization of EDSS and reductions in brain atrophy.²⁸ Three neuroprotective agents, riluzole (reduces glutamate excitotoxicity), fluoxetine (stimulates glycogenolysis and improves mitochondrial energy production), and amiloride (an acid-sensing ion channel blocker that opens in response to inflammation) were tested in a phase 2b multi-arm, multi-site parallel group RCT in SPMS (n = 445). The study failed to yield differences from placebo for any agent in reduction of brain volume loss.²⁹ A prior study of lamotrigine, a sodium channel blocker, also failed to find changes in brain volume loss.³⁰ These studies highlight the large sample sizes and/or long study durations needed to test agents using brain atrophy as primary outcome. In the future, precise surrogate markers of neuroprotection will be a great need for earlier phase trials. These results also suggest that targeting > 1 MOA may be necessary to treat SPMS effectively.

Efforts to promote remyelination target one hallmark of MS damage. High dose biotin (about 10,000× usual dose) may promote myelin repair as a cofactor for fatty acid synthesis and support mitochondrial oxidative phosphorylation. While a RCT yielded a greater proportion of participants with either PPMS or SPMS with improvement in disability than placebo at 12 months, an open label trial suggested otherwise indicating a need for a more definitive trial.^31,32

Anti-LINGO-1 (opicinumab) is a monoclonal antibody that targets LINGO, a potent negative regulator of oligodendrocyte differentiation and myelination.³³ Although this agent failed in a phase 2 trial in relapsing MS, and is thus unlikely to be tested in progressive forms, the innovative approach to stimulating oligodendrocytes is ongoing. One such effort is to use thyroid hormone, crucial to myelin formation during development, as a repair agent in MS.³⁴ A dose-finding study of thyroid hormone was completed and efforts to develop a thyromimetic agent are ongoing.

Finally, efforts to promote neurogenesis remain a goal for many neurodegenerative diseases. Exercise appears to prevent age-related atrophy of the hippocampus in animals and humans and help maintain neuronal health.³⁵ In patients with RRMS, cortical thickness is impacted positively by resistance training, which suggests a neuroprotective effect.³⁶ A multi-center trial of exercise in patients with progressive MS investigating cognitive outcomes is ongoing.

Discontinuing DMT

In the early 1990s, the successful development of immune modulating therapies that reliably reduced disease activity in RRMS led to widespread initiation in patients with relapsing disease. However, guidance on when or if to discontinue DMT, even in those who have transitioned to SPMS, remains largely absent at this time. Requests to discontinue DMT may come from patients weary of taking medication (especially injections), bothered by AEs, or those who no longer perceive efficacy from their treatments. Clinicians also may question the benefit of immune modulation in patients with longstanding freedom from relapses or changes in MRI lesion burden.

To inform discussion centered on treatment discontinuation, a clinical trial is currently underway to better answer the question of when and how to withdraw MS therapy. Discontinuation of Disease Modifying Therapies in Multiple Sclerosis (DISCO-MS) is a prospective, placebo-controlled RCT and its primary endpoint is recurrence of disease activity over a 2 year follow-up period.³⁷ Eligibility requirements for the trial include age > 55 years, 5-year freedom from relapses, and 3-year freedom from new MRI lesions (criteria informed by progressive MS cohort studies).³¹ In addition to demonstrating the active disease recurrence rates in this patient population, the trial also aims to identify risk factors for recurrent disease activity among treated MS patients.³⁷ DISCO-MS builds upon a series of retrospective and observational studies that partially answered these questions, albeit in the context of biases inherent in retrospective or observational studies.

A Minneapolis MS Treatment and Research Center single-center study identified 77 SPMS patients with no acute CNS inflammatory events over 2 to 20 years and advised these patients to stop taking DMT.³² In this group, 11.7% of subjects experienced recurrent active disease. Age was the primary discriminating factor. The mean age of those who experienced disease activity was 56 years vs 61 years those who did not. A second observational study from France found that among 100 SPMS patients treated either with interferon β or glatiramer acetate for at least 6 months, 35% experienced some form of inflammatory disease upon discontinuation.³⁸ Sixteen patients relapsed and 19 developed gadolinium-enhancing MR lesions after DMT discontinuation. However, the age of the cohort was younger than the Minneapolis study (47.2 years vs 61 years), and reasons for discontinuation (eg, AEs or lack of disease activity) were not considered in the analysis.

Other studies examining the safety of DMT discontinuation have not considered MS subtype or excluded patients with progressive subtypes of MS. The largest studies to date on DMT discontinuation utilized the international MSBase global patient registry, which identified nearly 5,000 patients who discontinued interferons (73%), glatiramer acetate (18%), natalizumab (6%), or fingolimod (3%), without specifying the reasons for discontinuation.³⁹ Despite these shortcomings, data reveal trends that are helpful in predicting how MS tends to behave in patients who have discontinued therapy. Not surprisingly, disability progression was most likely among patients already characterized as having a progressive phenotype, while relapses were less likely to occur among older, progressive patients.

Although clinicians may be increasingly willing to discuss DMT discontinuation with their patients, at least 1 study exploring patient perspectives on stopping treatment suggests widespread reluctance to stop treatment. A survey conducted with participants in the North American Research Committee on Multiple Sclerosis patient-report registry found that among survey respondents, only 11.9% would discontinue their MS medication if deemed stable, while 66.3% stated they were unlikely to stop treatment.⁴⁰

These results suggest that before clinicians incorporate DMT discontinuation into the normal course of discussion with patients, they should be prepared to provide both education (on the wisdom of stopping under the right circumstances) and evidence to support when and how to make the recommendation. Based on existing evidence, criteria for recommending treatment discontinuation might include prolonged freedom from disease activity (≥ 5 years), age > 55 years or 60 years, and a progressive disease course. Thus far, no combination of factors has been shown to completely predict an event-free transition off of medicine. Since no fixed algorithm yet exists to guide DMT stoppage in MS, reasonable suggestions for monitoring patients might include surveillance MRIs, more frequent clinic visits, and possible transitional treatment for patients coming off of natalizumab or fingolimod, since these drugs have been associated with rebound disease activity when discontinued.^41,42

Clinicians wishing to maximize function and quality of life for their patients at any age or stage of disease should look to nonpharmacologic interventions to lessen disability and maximize quality of life. While beyond the scope of this discussion, preliminary evidence suggests multimodal (aerobic, resistance, balance) exercise may enhance endurance and cognitive processing speed, and that treatment of comorbid diseases affecting vascular health benefits MS. ⁴³

Conclusions

The development of numerous treatments for RRMS has established an entirely new landscape and disease course for those with MS. While this benefit has not entirely extended to those with progressive MS, those with active disease with superimposed relapses may receive limited benefit from these medications. New insights into the pathophysiology of progressive MS may lead us to new treatments through multiple alternative pathophysiologic pathways. Some early studies using this strategy show promise in slowing the progressive phase. Medication development for progressive MS faces multiple challenges due to lack of a single animal model demonstrating both pathology and clinical effects, absence of phase 1 surrogate biomarkers, and later phase trial endpoints that require large sample sizes and extended study durations. Nevertheless, the increase in number of trials and diversity of therapeutic approaches for progressive MS provides hope for effective therapy. Currently, the heterogeneity of the population with progressive MS requires an individualized treatment approach, and in some of these patients, stopping therapy may be a reasonable consideration. Symptomatic management remains critical for all patients with progressive MS as well as non-pharmacologic approaches that maximize quality of life.

Multiple sclerosis (MS) is the most common demyelinating disease of the central nervous system, with recent estimates of around 1 million people living with MS in the US.¹ In many countries, MS is a leading cause of disability among young adults, second only to trauma.² Clinically, neurologic worsening (ie, disability) in MS can occur in the relapsing-remitting (RRMS) phase of disease due to incomplete recovery from neuroinflammatory relapses. However, in the 15% of patients with a progressive course from onset (PPMS), and in those with RRMS who transition to a secondary progressive phenotype (SPMS), neurologic worsening follows a slowly progressive pattern.³ A progressive disease course—either PPMS at onset or transitioning to SPMS—is the dominant factor affecting MS-related neurologic disability accumulation. In particular, epidemiologic studies have shown that, in the absence of transitioning to a progressive disease course, < 5% of individuals with MS will accumulate sufficient disability to necessitate use of a cane for ambulation.^4-6 Therefore, developing disease modifying therapies (DMTs) that are highly effective at slowing or stopping the gradual accumulation of neurologic disability in progressive MS represent a critical unmet need.

Research into the development of DMTs for progressive MS has been hindered by a number of factors. In particular, the clinical definition and diagnosis of progressive MS has been an evolving concept. Diagnostic criteria for MS, which help facilitate the enrollment of appropriate subjects into clinical trials, have only recently clarified the current consensus definition for progressive MS—steadily increasing neurologic disability independent of clinical relapses. Looking back to the Schumacher criteria in 1965 and Poser criteria in 1983, it was acknowledged that neurologic symptoms in MS may follow a relapsing-remitting or progressive pattern, but little attempt was made to define progressive MS.^7,8 The original McDonald criteria in 2001 defined diagnostic criteria for progressive MS.⁹ These criteria continued to evolve through subsequent revisions (eg, cerebrospinal fluid [CSF] specific oligoclonal bands no longer are an absolute requirement), and only in the 2017 revision was it emphasized that disability progression must occur independent of clinical relapses, concordant with similar emphasis in the 2013 revision of MS clinical course definitions.^3,10

The interpretation of prior clinical trials of DMT for progressive MS must consider this evolving clinical definition. The US Food and Drug Administration (FDA) approved mitoxantrone in 2000—making it the first DMT to carry an approved label for SPMS. While achieving significant clinical efficacy, it is clear from the details of the trial that the enrolled subjects had a high degree of inflammatory disease activity, which suggests that mitoxantrone treats neuroinflammation and not relapse-independent worsening. More recently, disparate results were seen in the anti-CD20 (rituximab, ocrelizumab) and S1P receptor modulator (fingolimod, siponimod) trials targeted at patients with primary and secondary progressive MS.^11-14 Although there are differences between these therapies, they are more similar than not within the same therapeutic class. Taken together, these trials illustrate the critical impact the narrower inclusion/exclusion criteria (namely age and extent of inflammatory activity) had on attaining positive outcomes. Other considerations, such as confounding illness, also may impact trial recruitment and generalizability of findings.

The lack of known biological targets in progressive MS, which is a complex disease with an incompletely understood and heterogeneous pathology, also hinders DMT development. Decades of research has characterized multifocal central nervous system (CNS) lesions that exhibit features of demyelination, inflammation, reactive gliosis, axonal loss, and neuronal damage. Until recently, however, much of this research focused on the relapsing phase of disease, and so the understanding of the pathologic underpinnings of progressive disease has remained limited. Current areas of investigation encompass a broad range of pathological processes, such as microglial activation, meningeal lymphoid follicles, remyelination failure, vulnerability of chronically demyelinated axons, oxidative injury, iron accumulation, mitochondrial damage, and others. There is the added complication that the pathologic processes underlying progressive MS are superimposed on the CNS aging process. In particular, the transition to progressive MS and the rate of disability accumulation during progressive MS show strong correlation with age.^6,15-17

Finally, DMT development for progressive MS also is hindered by the lack of specific surrogate and clinical outcome measures. Trials for relapsing MS have benefited greatly from the relatively straightforward assessment of clinical relapses and inflammatory disease activity on magnetic resonance imaging (MRI). With the goal of developing DMTs that are highly effective at slowing or stopping the gradual accumulation of neurologic disability in progressive MS, which by definition occurs independent of clinical relapses, these measures are not directly relevant. The longitudinal clinical disability outcome measures change at a slower rate than in early, relapsing disease. The use of standardized scales (eg, the Expanded Disability Status Scale [EDSS]), lower limb function, upper limb function, cognition, or a combination is a subject of ongoing debate. For example, the ASCEND and IMPACT trials (placebo-controlled trials for SPMS with natalizumab and interferon β-1a, respectively) showed no significant impact in EDSS progression—but in both of these trials, the 9-hole peg test (9-HPT), a performance measure for upper limb function, showed significant improvement.^10,18 Particularly in those with an EDSS of > 6.5, who are unlikely to have measurable EDSS progression, functional tests such as the 9-HPT or timed 25-foot walk may be more sensitive as measures for disability progression.¹¹ MRI measures of brain atrophy is the current gold standard surrogate outcome for clinical trials in progressive MS, but others that may warrant consideration include optical coherence tomography (OCT) or CSF markers of axonal degeneration.

DMT for Progressive MS

Current diagnostic nomenclature separates patients with active (superimposed clinical and/or radiographic relapses) from those with inactive disease.^3,12 Relapsing forms of MS include all RRMS and those with SPMS with superimposed relapses (ie, active SPMS). Following this paradigm shift, the FDA changed the indication for already approved DMT from RRMS to relapsing forms of MS. Below is a discussion of DMT that specifically use the term SPMS and PPMS in the indication, where phase 3 trial data for progressive MS is available.

In 2019, the FDA approved the first oral medication (siponimod) for active SPMS. Subsequently, updates to the labels of the older DMT expanded to include active SPMS. Until 2019, the only FDA approved medication for SPMS was mitoxantrone, and use of this medication was limited due to unfavorable adverse effects (AEs). No medications had obtained FDA approval for inactive SPMS to this point, which represented an unmet need for a considerable number of patients.

Mitoxantrone became the first DMT approved for use in SPMS in 2000 following early trials that showed reductions in EDSS worsening, change in ambulation index, reduced number of treated relapses, and prolonged time to first treated relapse. However, as with some of the other positive trials in progressive MS, it is difficult to discern the impact of suppression of relapses as opposed to direct impact on progressive pathophysiology. Within the placebo arm, for example, there were 129 relapses among the 64 subjects, which suggests that these cases had particularly active disease or were in the early stages of SPMS.¹³ Furthermore, the use of this medication was limited due to concerns of cardiotoxicity and hematologic malignancy as serious AEs.

The trials of interferon β-1b illustrate the same difficulty of isolating possible benefits in disease progression from disability accumulated from relapses. The first interferon β-1b trial for SPMS, was conducted in Europe using fingolimod and showed a delay in confirmed disability progression compared to placebo as measured with the EDSS.¹⁴ However, a North American trial that followed in 2004 was unable to replicate this finding.¹⁵ The patients in the European trial appeared to be in an earlier phase of SPMS with more active disease, and a post-hoc pooled analysis suggested that patients with active disease and those with more pronounced disability progression were more likely to benefit from treatment.¹⁶ Overall, interferons do not appear to appreciably alter disability in the long-term for patients with SPMS, though they may modify short-term, relapse-related disability.

Perhaps the most encouraging data for SPMS was found in the EXPAND trial, which investigated siponimod, an S1P receptor modulator that is more selective than fingolimod. The trial identified a 21% reduction in 3-month confirmed disability progression for SPMS patients taking siponimod compared with those taking a placebo.¹⁷ Although the patients in EXPAND did seem to have relatively less disease activity at baseline than those who participated in other SPMS trials, those who benefitted from siponimod were primarily patients who had clinical and/or radiographic relapses over the prior 2 years. Based on this, the FDA approved siponimod for active SPMS. The extent to which siponimod exerts a true neuroprotective effect beyond reducing inflammation has not been clearly established.

B-cell depleting therapies rituximab and ocrelizumab have been evaluated in both primary and secondary progressive MS populations. Early investigations of the chimeric rituximab in PPMS did not show benefits on disability (EDSS) progression; however, benefits were seen in analysis of some subgroups.¹⁸ With this in mind, the ORATORIO trial for the humanized version, ocreluzimab, included PPMS patients that were younger (aged < 55 years) and had cutoffs for disease duration (< 15 years for those with EDSS more than 5 years, < 10 years for those with EDSS less than 5 years). The study showed statistically significant changes on disability progression, which led to ocrelizumab receiving the first FDA indication for PPMS.¹¹ There are substantial pathophysiologic similarities between PPMS and SPMS in the progressive phase.¹⁹ While these medications may have similar effects across these disease processes, these benefits have not yet been demonstrated in a prospective trial for the SPMS population. Regardless, B-cell depleting therapy is a reasonable consideration for select patients with active SPMS, consistent with a relapsing form of MS.

Therapies in Development 

DMT development for progressive MS is a high priority area. Current immunomodulatory therapies for RRMS have consistently been ineffective in the inactive forms of MS, with the possible exceptions of ocrelizumab and siponimod. Therefore, instead of immunosuppression, many agents currently in phase 2 and 3 clinical trials target alternative pathophysiological processes in order to provide neuroprotection, and/or promote remyelination and neurogenesis. These targets include oxidative stress (OS), non-T cell mediated inflammation, and mitochondrial/energy failure.²⁰ Below we review a selection of clinical trials testing agents following these approaches. Many agents have more than one potential mechanism of action (MOA) that could benefit progressive MS.

Lipoic acid (LA), also known as α-lipoic acid and thiotic acid, is one such agent targeting alternative pathophysiology in SPMS. LA is an endogenous agent synthesized de novo from fatty acids and cysteine as well as obtained in small amounts from foods.²¹ It has antioxidant (AO) properties including direct radical scavenging, regeneration of glutathione, and upregulation of AO enzymes via the NrF2 pathway.²² It supports mitochondria as a key cofactor for pyruvate dehydrogenase and α-ketoglutarate dehydrogenase, and it also aids mitochondrial DNA synthesis.^21,22 Studies in experimental autoimmune encephalomyelitis, a widely used experimental mouse model of inflammatory demyelinating disease, also indicate a reduction in excessive microglial activation.²³ A phase 2 pilot randomized controlled trial (RCT) of 1200 mg LA in SPMS (n = 51) resulted in significantly less whole brain atrophy by SIENA (Structural Image Evaluation, Using Normalization, of Atrophy) at 2 years.²⁴ A follow-up multicenter trial is ongoing.

Simvastatin also targets alternative pathophysiology in SPMS. It has anti-inflammatory effects, improves vascular function, and promotes neuroprotection by reducing excitotoxicity. A phase 2 RCT demonstrated a reduction in whole brain atrophy in SPMS (n = 140), and a phase 3 trial is underway.²⁵ Ibudilast is another repurposed drug that targets alternative inflammation by inhibiting several cyclic nucleotide phosphodiesterases, macrophage migration inhibitory factor and toll-like receptor 4. A phase 2 trial (n = 225) in both SPMS and PPMS also demonstrated a reduction in brain atrophy, but participants had high rates of AEs.²⁶

Lithium and riluzole promote neuroprotection by reducing excitotoxicity. Lithium is a pharmacologic active cation used as a mood stabilizer and causes inhibition of glycogen synthase kinase-3β. Animal models also indicate that lithium may decrease inflammation and positively impact neurogenesis.²⁷ A crossover pilot trial demonstrated tolerability with trends toward stabilization of EDSS and reductions in brain atrophy.²⁸ Three neuroprotective agents, riluzole (reduces glutamate excitotoxicity), fluoxetine (stimulates glycogenolysis and improves mitochondrial energy production), and amiloride (an acid-sensing ion channel blocker that opens in response to inflammation) were tested in a phase 2b multi-arm, multi-site parallel group RCT in SPMS (n = 445). The study failed to yield differences from placebo for any agent in reduction of brain volume loss.²⁹ A prior study of lamotrigine, a sodium channel blocker, also failed to find changes in brain volume loss.³⁰ These studies highlight the large sample sizes and/or long study durations needed to test agents using brain atrophy as primary outcome. In the future, precise surrogate markers of neuroprotection will be a great need for earlier phase trials. These results also suggest that targeting > 1 MOA may be necessary to treat SPMS effectively.

Efforts to promote remyelination target one hallmark of MS damage. High dose biotin (about 10,000× usual dose) may promote myelin repair as a cofactor for fatty acid synthesis and support mitochondrial oxidative phosphorylation. While a RCT yielded a greater proportion of participants with either PPMS or SPMS with improvement in disability than placebo at 12 months, an open label trial suggested otherwise indicating a need for a more definitive trial.^31,32

Anti-LINGO-1 (opicinumab) is a monoclonal antibody that targets LINGO, a potent negative regulator of oligodendrocyte differentiation and myelination.³³ Although this agent failed in a phase 2 trial in relapsing MS, and is thus unlikely to be tested in progressive forms, the innovative approach to stimulating oligodendrocytes is ongoing. One such effort is to use thyroid hormone, crucial to myelin formation during development, as a repair agent in MS.³⁴ A dose-finding study of thyroid hormone was completed and efforts to develop a thyromimetic agent are ongoing.

Finally, efforts to promote neurogenesis remain a goal for many neurodegenerative diseases. Exercise appears to prevent age-related atrophy of the hippocampus in animals and humans and help maintain neuronal health.³⁵ In patients with RRMS, cortical thickness is impacted positively by resistance training, which suggests a neuroprotective effect.³⁶ A multi-center trial of exercise in patients with progressive MS investigating cognitive outcomes is ongoing.

Discontinuing DMT

In the early 1990s, the successful development of immune modulating therapies that reliably reduced disease activity in RRMS led to widespread initiation in patients with relapsing disease. However, guidance on when or if to discontinue DMT, even in those who have transitioned to SPMS, remains largely absent at this time. Requests to discontinue DMT may come from patients weary of taking medication (especially injections), bothered by AEs, or those who no longer perceive efficacy from their treatments. Clinicians also may question the benefit of immune modulation in patients with longstanding freedom from relapses or changes in MRI lesion burden.

To inform discussion centered on treatment discontinuation, a clinical trial is currently underway to better answer the question of when and how to withdraw MS therapy. Discontinuation of Disease Modifying Therapies in Multiple Sclerosis (DISCO-MS) is a prospective, placebo-controlled RCT and its primary endpoint is recurrence of disease activity over a 2 year follow-up period.³⁷ Eligibility requirements for the trial include age > 55 years, 5-year freedom from relapses, and 3-year freedom from new MRI lesions (criteria informed by progressive MS cohort studies).³¹ In addition to demonstrating the active disease recurrence rates in this patient population, the trial also aims to identify risk factors for recurrent disease activity among treated MS patients.³⁷ DISCO-MS builds upon a series of retrospective and observational studies that partially answered these questions, albeit in the context of biases inherent in retrospective or observational studies.

A Minneapolis MS Treatment and Research Center single-center study identified 77 SPMS patients with no acute CNS inflammatory events over 2 to 20 years and advised these patients to stop taking DMT.³² In this group, 11.7% of subjects experienced recurrent active disease. Age was the primary discriminating factor. The mean age of those who experienced disease activity was 56 years vs 61 years those who did not. A second observational study from France found that among 100 SPMS patients treated either with interferon β or glatiramer acetate for at least 6 months, 35% experienced some form of inflammatory disease upon discontinuation.³⁸ Sixteen patients relapsed and 19 developed gadolinium-enhancing MR lesions after DMT discontinuation. However, the age of the cohort was younger than the Minneapolis study (47.2 years vs 61 years), and reasons for discontinuation (eg, AEs or lack of disease activity) were not considered in the analysis.

Other studies examining the safety of DMT discontinuation have not considered MS subtype or excluded patients with progressive subtypes of MS. The largest studies to date on DMT discontinuation utilized the international MSBase global patient registry, which identified nearly 5,000 patients who discontinued interferons (73%), glatiramer acetate (18%), natalizumab (6%), or fingolimod (3%), without specifying the reasons for discontinuation.³⁹ Despite these shortcomings, data reveal trends that are helpful in predicting how MS tends to behave in patients who have discontinued therapy. Not surprisingly, disability progression was most likely among patients already characterized as having a progressive phenotype, while relapses were less likely to occur among older, progressive patients.

Although clinicians may be increasingly willing to discuss DMT discontinuation with their patients, at least 1 study exploring patient perspectives on stopping treatment suggests widespread reluctance to stop treatment. A survey conducted with participants in the North American Research Committee on Multiple Sclerosis patient-report registry found that among survey respondents, only 11.9% would discontinue their MS medication if deemed stable, while 66.3% stated they were unlikely to stop treatment.⁴⁰

These results suggest that before clinicians incorporate DMT discontinuation into the normal course of discussion with patients, they should be prepared to provide both education (on the wisdom of stopping under the right circumstances) and evidence to support when and how to make the recommendation. Based on existing evidence, criteria for recommending treatment discontinuation might include prolonged freedom from disease activity (≥ 5 years), age > 55 years or 60 years, and a progressive disease course. Thus far, no combination of factors has been shown to completely predict an event-free transition off of medicine. Since no fixed algorithm yet exists to guide DMT stoppage in MS, reasonable suggestions for monitoring patients might include surveillance MRIs, more frequent clinic visits, and possible transitional treatment for patients coming off of natalizumab or fingolimod, since these drugs have been associated with rebound disease activity when discontinued.^41,42

Clinicians wishing to maximize function and quality of life for their patients at any age or stage of disease should look to nonpharmacologic interventions to lessen disability and maximize quality of life. While beyond the scope of this discussion, preliminary evidence suggests multimodal (aerobic, resistance, balance) exercise may enhance endurance and cognitive processing speed, and that treatment of comorbid diseases affecting vascular health benefits MS. ⁴³

Conclusions

The development of numerous treatments for RRMS has established an entirely new landscape and disease course for those with MS. While this benefit has not entirely extended to those with progressive MS, those with active disease with superimposed relapses may receive limited benefit from these medications. New insights into the pathophysiology of progressive MS may lead us to new treatments through multiple alternative pathophysiologic pathways. Some early studies using this strategy show promise in slowing the progressive phase. Medication development for progressive MS faces multiple challenges due to lack of a single animal model demonstrating both pathology and clinical effects, absence of phase 1 surrogate biomarkers, and later phase trial endpoints that require large sample sizes and extended study durations. Nevertheless, the increase in number of trials and diversity of therapeutic approaches for progressive MS provides hope for effective therapy. Currently, the heterogeneity of the population with progressive MS requires an individualized treatment approach, and in some of these patients, stopping therapy may be a reasonable consideration. Symptomatic management remains critical for all patients with progressive MS as well as non-pharmacologic approaches that maximize quality of life.

Prior to the first approved disease modifying therapy (DMT) in the 1990s, treatment approaches for multiple sclerosis (MS) were not well understood. The discovery that MS was an immune mediated inflammatory disease paved the way for the treatments we know today. In 1993, interferon β‐1b became the first DMT for MS approved by the US Food and Drug Administration (FDA). Approvals for interferon β‐1a as well as glatiramer acetate (GA) soon followed. Today, we consider these the mildest immunosuppressant DMTs; however, their success verified that suppressing the immune system had a positive effect on the MS disease process.

Following these approvals, the disease process in MS is now better understood. Recently approved therapies include monoclonal antibodies, which affect other immune pathways. Today, there are 14 approved DMTs (Table 1). Although the advent of these newer DMTs has revolutionized care for patients with MS, it has been accompanied by increasing costs for the agents. Direct medical costs associated with MS management, coupled with indirect costs from lost productivity, have been estimated to be $24.2 billion annually in the US.¹ These increases have been seen across many levels of insurance coverage—private payer, Medicare, and the Veterans Health Administration (VHA).^2,3

fdp03701036s_f.png

The Figure demonstrates the cost increase that have been seen across VHA between 2004 and 2019 for the DMTs identified in Table 1. Indeed, this compound annual growth rate may be an underestimate because infusion therapies (eg, natalizumab, ocrelizumab, and alemtuzumab) are difficult to track as they may be dispensed directly via a Risk Evaluation Medication Strategy (REMS) program. According to the VHA Pharmacy Benefit Management Service (PBM), in September 2019, dimethyl fumarate (DMF) had the 13th highest total outpatient drug cost for the US Department of Veterans Affairs (VA), interferon β‐1a ranked 62nd and 83rd (prefilled pen and syringe, respectively), and GA 40 mg ranked 89th.

fdp03701036s_t1.png

Factors Impacting DMT Use

Recent changes to MS typing have impacted utilization of DMTs. Traditionally, there were 4 subtypes of MS: relapsing remitting (RRMS), secondary progressive (SPMS), progressive relapsing (PRMS), and primary progressive (PPMS). These subtypes are now viewed more broadly and grouped as either relapsing or progressive. The traditional subtypes fall under these broader definitions. Additionally, SPMS has been broken into active SPMS, characterized by continued worsening of disability unrelated to acute relapses, superimposed with activity that can be seen on magnetic resonance images (MRIs), and nonactive SPMS, which has the same disability progression as active SPMS but without MRI-visible activity.^4-6 In 2019, these supplementary designations to SPMS made their first appearance in FDA-approved indications. All existing DMTs now include this terminology in their labelling and are indicated in active SPMS. There remain no DMTs that treat nonactive SPMS.

The current landscape of DMTs is highly varied in method of administration, risks, and benefits. As efficacy of these medications often is marked by how well they can prevent the immune system from attacking myelin, an inverse relationship between safety and efficacy results. The standard treatment outcomes in MS have evolved over time. The following are the commonly used primary outcomes in clinical trials: relapse reduction; increased time between relapses; decreased severity of relapses; prevention or extend time to disability milestones as measured by the Expanded Disability Status Scale (EDSS) and other disability measures; prevention or extension of time to onset of secondary progressive disease; prevention or reduction of the number and size of new and enhancing lesions on MRI; and limitation of overall MRI lesion burden in the central nervous system (CNS).

Newer treatment outcomes employed in more recent trials include: measures of axonal damage, CNS atrophy, evidence of microscopic disease via conventional MRI and advanced imaging modalities, biomarkers associated with inflammatory disease activity and neurodegeneration in MS, and the use of no evidence of disease activity (NEDA). These outcomes also must be evaluated by the safety concerns of each agent. Short- and long-term safety are critical factors in the selection of DMTs for MS. The injectable therapies for MS (interferon β‐1a, interferon β‐1b, and GA) have established long-term safety profiles from > 20 years of continuous use. The long-term safety profiles of oral immunomodulatory agents and monoclonal antibodies for these drugs in MS have yet to be determined. Safety concerns associated with some therapies and added requirements for safety monitoring may increase the complexity of a therapeutic selection.

fdp03701036s_t2.png

Current cost minimization strategies for DMT include limiting DMT agents on formularies, tier systems that incentivize patients/prescribers to select the lowest priced agents on the formulary, negotiating arrangements with manufacturers to freeze prices or provide discounts in exchange for a priority position in the formulary, and requiring prior authorization to initiate or switch therapy. The use of generic medications and interchange to these agents from a brand name formulation can help reduce expense. Several of these strategies have been implemented in VHA.

Disease-Modifying Therapies

In 2019, 18,645 veterans with MS had either a MS-specific DMT or ≥ 1 annual encounters with a primary diagnosis of MS. Of this population, 4,720 were female and 13,357 were service connected according to VA data. About 50% of veterans with MS take a DMT. This percentage has remained stable over the past decade (Table 2). Although it appears the number of unique veterans prescribed an outpatient DMT is decreasing, this does not include the growing use of infused DMTs or DMTs obtained through the Veterans Choice Program (VCP)/Community Care (CC).

The overall outpatient pharmacy costs for veterans have remained constant despite the reduction in outpatient pharmacy prescription numbers. This may be due to increases in DMT cost to the VHA and the use of more expensive oral agents over the previously used platform injection DMTs.

Generic Conversion

GA is available in 20 mg daily and 40 mg3 times weekly subcutaneous injection dosing. The first evidence of clinical efficacy for a generic formulation for GA was evaluated by the GATE trial.⁷ This trial was a multicenter, randomized, double-blind, active- and placebo-controlled phase 3 trial. Eligible participants were randomized to receive daily SC injection for 9 months of 20 mg generic GA (n = 5,353), 20 mg brand GA (n = 5,357), or placebo (n = 584). The primary endpoint was the mean number of gadolinium (Gd1) lesions visible on MRIs during months 7, 8, and 9, which were significantly reduced in the combined GA-treated group and in each GA group individually when compared with the placebo group, confirming the study sensitivity (ie, GA was effective under the conditions of the study). Tolerability (including injection site reactions) and safety (incidence, spectrum, and severity of adverse events [AEs]) were similar in the generic and brand GA groups. These results demonstrated that generic and brand GA had equivalent efficacy, tolerability, and safety over a 9-month period.⁷

Results of a 15-month extension of the study were presented in 2015 and showed similar efficacy, safety, and tolerability in participants treated with generic GA for 2 years and patients switched from brand to generic GA.⁸ Multiple shifts for GA occurred, most notably the conversion from branded Copaxone (Teva Pharmaceutical Industries) to generic Glatopa (Sandoz). Subsequently, Sandoz released a generic 40 mg 3 times weekly formulation. Additionally, Mylan entered the generic GA market. With 3 competing manufacturers, internal data from the VHA indicated that it was able to negotiate a single source contract for this medication that provided a savings of $32,088,904.69 between September 2016 and May 2019.

The impact of generic conversions is just being realized. Soon, patents will begin to expire for oral DMTs, leading to an expected growth of generic alternatives. Already the FDA has approved 4 generic alternatives for teriflunomide, 3 for fingolimod (with 13 tentative approvals), and 15 generic alternatives for dimethyl fumarate (DMF). Implementation of therapeutic interchanges will be pursued by VHA as clinically supported by evidence.

Criteria for Use

PBM supports utilizing criteria to help guide providers on DMT options and promote safe, effective, and value-based selection of a DMT. The PBM creates monographs and criteria for use (CFU) for new medications. The monograph contains a literature evaluation of all studies available to date that concern both safety and efficacy of the new medication. Therapeutic alternatives also are presented and assessed for key elements that may determine the most safe and effective use. Additional safety areas for the new medications such as look-alike, sound-alike potential, special populations use (ie, those who are pregnant, the elderly, and those with liver or renal dysfunction), and drug-drug interactions are presented. Lastly, and possibly most importantly in an ever-growing growing world of DMTs, the monograph describes a reasonable place in therapy for the new DMT.

CFU are additional guidance for some DMTs. The development of CFU are based on several questions that arise during the monograph development for a DMT. These include, but are not limited to:

• Are there safety concerns that require the drug to receive a review to ensure safe prescribing (eg, agents with REMS programs, or safety concerns in specific populations)?
• Does the drug require a specialty provider type with knowledge and experience in those disease states to ensure appropriate and safe prescribing (eg restricted to infectious diseases)?
• Do VHA or non-VHA guidelines suggest alternative therapy be used prior to the agent?
• Is a review deemed necessary to ensure the preferred agent is used first (eg, second-line therapy)?

The CFU defines parameters of drug use consistent with high quality and evidence-based patient care. CFUs also serve as a basis for monitoring local, regional, and national patterns of pharmacologic care and help guide health care providers (HCPs) on appropriate use of medication.

CFUs are designed to ensure the HCP is safely starting a medication that has evidence for efficacy for their patient. For example, alemtuzumab is a high-risk, high-efficacy DMT. The alemtuzumab CFU acknowledges this by having exclusion criteria that prevent a veteran at high risk (ie, on another immunosuppressant) from being exposed to severe AEs (ie, severe leukopenia) that are associated with the medication. On the other hand, the inclusion criteria recognize the benefits of alemtuzumab and allows those with highly active MS who have failed other DMTs to receive the medication.

The drug monograph and CFU process is an important part of VHA efforts to optimize patient care. After a draft version is developed, HCPs can provide feedback on the exclusion/inclusion criteria and describe how they anticipate using the medication in their practice. This insight can be beneficial for MS treatment as diverse HCPs may have distinct viewpoints on how DMTs should be started. Pharmacists and physicians on a national level then discuss and decide together what to include in the final drafts of the drug monograph and CFU. Final documents are disseminated to all sites, which encourages consistent practices across the VHA.⁹ These documents are reviewed on a regular basis and updated as needed based on available literature evidence.

It is well accepted that early use of DMT correlates with lower accumulated long-term disability.¹⁰ However, discontinuation of DMT should be treated with equal importance. This benefits the patient by reducing their risk of AEs from DMTs and provides cost savings. Age and disease stability are factors to consider for DMT discontinuation. In a study with patients aged > 45 years and another with patients aged > 60 years, discontinuing DMT rarely had a negative impact and improved quality of life.^11,12 A retrospective meta-analysis of age-dependent efficacy of current DMTs predicted that DMT loses efficacy at age 53 years. In addition, higher efficacy DMT only outperforms lower efficacy DMT in patients aged < 40.5 years.¹³ Stability of disease and lack of relapses for ≥ 2 years also may be a positive predictor to safely discontinue DMT.^14,15 The growing literature to support safe discontinuation of DMT makes this a more convincing strategy to avoid unnecessary costs associated with current DMTs. With an average age of 59 years for veterans with MS, this may be one of the largest areas of cost avoidance to consider.

Off-Label Use

Other potential ways to reduce DMT costs is to consider off-label treatments. The OLYMPUS trial studied off-label use of rituximab, an anti-CD20 antibody like ocrelizumab. It did not meet statistical significance for its primary endpoint; however, in a subgroup analysis, off-label use was found to be more effective in a population aged < 51 years.¹⁶ Other case reports and smaller scale studies also describe rituximab’s efficacy in MS.^17,18 In 2018, the FDA approved the first rituximab biosimilar.¹⁹ Further competition from biosimilars likely will make rituximab an even more cost-effective choice when compared with ocrelizumab.

Alternate Dosing Regimens

Extended interval dosing of natalizumab has been studied, extending the standard infusion interval from every 4 weeks to 5- to 8-week intervals. One recent article compared these interval extensions and found that all extended intervals of up to 56 days did not increase new or enhancing lesions on MRI when compared with standard interval dosing.²⁰ Another larger randomized trial is underway to evaluate efficacy and safety of extended interval dosing of natalizumab (NCT03689972). Utilization of this dosing may reduce natalizumab annual costs by up to 50%.

Safety Monitoring

DMF is an oral DMT on the VHA formulary with CFU. Since leukopenia is a known AE, baseline and quarterly monitoring of the complete blood count (CBC) is recommended for patients taking DMF. Additionally, DMF should be held if white blood cell count (WBC) falls below 2,000/mm³.²¹ There have been recent reports of death secondary to progressive multifocal leukoencephalopathy (PML) among European patients taking DMF.^22-24 This has raised concerns about adherence to recommended CBC monitoring in veterans taking DMF. The association of DMF and leukopenia has been evident since early clinical trials.²⁵ Leukopenia in immunocompromised patients increases the risk of PML.

In the long-term extension study ENDORSE, 6% to 7% of patients continuing DMF had WBC counts of 3.0×10⁹/L compared with 7% to 10% in the new to DMF group.²⁶ In addition 6% to 8% of patients continuing DMF had lymphocyte counts of 0.5×10⁹/L, compared with 5% to 9% in the new to DMF group. The cases of PML occurred in patients who had low lymphocyte counts over an extended period with no adjustment to DMF therapy, such as holding the drug until WBC counts returned to normal levels or stopping the drug. Discussion and review within VHA resulted in the recommendation for quarterly WBC monitoring criteria.

PBM and VA Center for Medication Safety (MedSafe) conducted a medication usage evaluation (MUE) on adherence to the WBC monitoring set forth in the CFU. Data collection began in fourth quarter of fiscal year (FY) 2015 with the most recent reporting period of fourth quarter of FY 2017. The Medication Utilization Evaluation Tool tracks patients with no reported WBC in 90 days and WBC < 2,000/mm³. Over the reporting period, 20% to 23% of patients have not received appropriate quarterly monitoring. Additionally, there have been 4 cases where the WBC decreased below the threshold limit. To ensure safe and effective use of DMF, it is important to adhere to the monitoring requirements set forth in the CFU.

Impact of REMS and Special Distribution

As DMTs increase in efficacy, there are often more risks associated with them. Some of these high-risk medications, including natalizumab and alemtuzumab, have REMS programs and/or have special distribution procedures. Although REMS are imperative for patient safety, the complexity of these programs can be difficult to navigate, which can create a barrier to access. The PBM helps to assist all sites with navigating and adhering to required actions to dispense and administer these medications through a national Special Handling Drugs Microsoft SharePoint site, which provides access to REMS forms and procurement information when drugs are dispensed from specialty pharmacies. Easing this process nationwide empowers more sites to be confident they can dispense specialty medications appropriately.

Clinical Pharmacists

The VHA is unique in its utilization of pharmacists in outpatient clinic settings. Utilization of an interdisciplinary team for medication management has been highly used in VHA for areas like primary care; however, pharmacist involvement in specialty areas is on the rise and MS is no exception. Pharmacists stationed in clinics, such as neurology or spinal cord injury, can impact care for veterans with MS. Interdisciplinary teams that include a pharmacist have been shown to increase patient adherence to DMTs.²⁷ However, pharmacists often assist with medication education and monitoring, which adds an additional layer of safety to DMT treatment. At the VHA, pharmacists also can obtain a scope of practice that allows them to prescribe medications and increase access to care for veterans with MS.

Education

The VHA demonstrates how education on a disease state like MS can be distributed on a large, national scale through drug monographs, CFU, and Microsoft SharePoint sites. In addition, VHA has created the MS Centers of Excellence (MSCoE) that serve as a hub of specialized health care providers in all aspects of MS care.

A core function of the MSCoE is to provide education to both HCPs and patients. The MSCoE and its regional hubs support sites that may not have an HCP who specializes in MS by providing advice on DMT selection, how to obtain specialty medications, and monitoring that needs to be completed to ensure veterans’ safety. The MSCoE also has partnered with the National MS Society to hold a lecture series on topics in MS. This free series is available online to all HCPs who interact with patients who have MS and is a way that VA is extending its best practices and expertise beyond its own health care system. There also is a quarterly newsletter for veterans with MS that highlights new information on DMTs that can affect their care.

Conclusion

It is an exciting and challenging period in MS treatment. New DMTs are being approved and entering clinical trials at a rapid pace. These new DMT agents may offer increased efficacy, improvements in AE profiles, and the possibility of increased medication adherence—but often at a higher cost. The utilization of CFU and formulary management provides the ability to ensure the safe and appropriate use of medications by veterans, with a secondary outcome of controlling pharmacy expenditures.

The VHA had expenditures of $142,135,938 for DMT use in FY 2018. As the VHA sees the new contract prices for DMT in January 2020, we are reminded that costs will continue to rise with some pharmaceutical manufacturers implementing prices 8% to 11% higher than 2019 prices, when the consumer price index defines an increase of 1.0% for 2020 and 1.4% in 2021.²⁸ It is imperative that the VHA formulary be managed judiciously and the necessary measures be in place for VHA practitioners to enable effective, safe and value-based care to the veteran population.

Prior to the first approved disease modifying therapy (DMT) in the 1990s, treatment approaches for multiple sclerosis (MS) were not well understood. The discovery that MS was an immune mediated inflammatory disease paved the way for the treatments we know today. In 1993, interferon β‐1b became the first DMT for MS approved by the US Food and Drug Administration (FDA). Approvals for interferon β‐1a as well as glatiramer acetate (GA) soon followed. Today, we consider these the mildest immunosuppressant DMTs; however, their success verified that suppressing the immune system had a positive effect on the MS disease process.

Following these approvals, the disease process in MS is now better understood. Recently approved therapies include monoclonal antibodies, which affect other immune pathways. Today, there are 14 approved DMTs (Table 1). Although the advent of these newer DMTs has revolutionized care for patients with MS, it has been accompanied by increasing costs for the agents. Direct medical costs associated with MS management, coupled with indirect costs from lost productivity, have been estimated to be $24.2 billion annually in the US.¹ These increases have been seen across many levels of insurance coverage—private payer, Medicare, and the Veterans Health Administration (VHA).^2,3

fdp03701036s_f.png

The Figure demonstrates the cost increase that have been seen across VHA between 2004 and 2019 for the DMTs identified in Table 1. Indeed, this compound annual growth rate may be an underestimate because infusion therapies (eg, natalizumab, ocrelizumab, and alemtuzumab) are difficult to track as they may be dispensed directly via a Risk Evaluation Medication Strategy (REMS) program. According to the VHA Pharmacy Benefit Management Service (PBM), in September 2019, dimethyl fumarate (DMF) had the 13th highest total outpatient drug cost for the US Department of Veterans Affairs (VA), interferon β‐1a ranked 62nd and 83rd (prefilled pen and syringe, respectively), and GA 40 mg ranked 89th.

fdp03701036s_t1.png

Factors Impacting DMT Use

Recent changes to MS typing have impacted utilization of DMTs. Traditionally, there were 4 subtypes of MS: relapsing remitting (RRMS), secondary progressive (SPMS), progressive relapsing (PRMS), and primary progressive (PPMS). These subtypes are now viewed more broadly and grouped as either relapsing or progressive. The traditional subtypes fall under these broader definitions. Additionally, SPMS has been broken into active SPMS, characterized by continued worsening of disability unrelated to acute relapses, superimposed with activity that can be seen on magnetic resonance images (MRIs), and nonactive SPMS, which has the same disability progression as active SPMS but without MRI-visible activity.^4-6 In 2019, these supplementary designations to SPMS made their first appearance in FDA-approved indications. All existing DMTs now include this terminology in their labelling and are indicated in active SPMS. There remain no DMTs that treat nonactive SPMS.

The current landscape of DMTs is highly varied in method of administration, risks, and benefits. As efficacy of these medications often is marked by how well they can prevent the immune system from attacking myelin, an inverse relationship between safety and efficacy results. The standard treatment outcomes in MS have evolved over time. The following are the commonly used primary outcomes in clinical trials: relapse reduction; increased time between relapses; decreased severity of relapses; prevention or extend time to disability milestones as measured by the Expanded Disability Status Scale (EDSS) and other disability measures; prevention or extension of time to onset of secondary progressive disease; prevention or reduction of the number and size of new and enhancing lesions on MRI; and limitation of overall MRI lesion burden in the central nervous system (CNS).

Newer treatment outcomes employed in more recent trials include: measures of axonal damage, CNS atrophy, evidence of microscopic disease via conventional MRI and advanced imaging modalities, biomarkers associated with inflammatory disease activity and neurodegeneration in MS, and the use of no evidence of disease activity (NEDA). These outcomes also must be evaluated by the safety concerns of each agent. Short- and long-term safety are critical factors in the selection of DMTs for MS. The injectable therapies for MS (interferon β‐1a, interferon β‐1b, and GA) have established long-term safety profiles from > 20 years of continuous use. The long-term safety profiles of oral immunomodulatory agents and monoclonal antibodies for these drugs in MS have yet to be determined. Safety concerns associated with some therapies and added requirements for safety monitoring may increase the complexity of a therapeutic selection.

fdp03701036s_t2.png

Current cost minimization strategies for DMT include limiting DMT agents on formularies, tier systems that incentivize patients/prescribers to select the lowest priced agents on the formulary, negotiating arrangements with manufacturers to freeze prices or provide discounts in exchange for a priority position in the formulary, and requiring prior authorization to initiate or switch therapy. The use of generic medications and interchange to these agents from a brand name formulation can help reduce expense. Several of these strategies have been implemented in VHA.

Disease-Modifying Therapies

In 2019, 18,645 veterans with MS had either a MS-specific DMT or ≥ 1 annual encounters with a primary diagnosis of MS. Of this population, 4,720 were female and 13,357 were service connected according to VA data. About 50% of veterans with MS take a DMT. This percentage has remained stable over the past decade (Table 2). Although it appears the number of unique veterans prescribed an outpatient DMT is decreasing, this does not include the growing use of infused DMTs or DMTs obtained through the Veterans Choice Program (VCP)/Community Care (CC).

The overall outpatient pharmacy costs for veterans have remained constant despite the reduction in outpatient pharmacy prescription numbers. This may be due to increases in DMT cost to the VHA and the use of more expensive oral agents over the previously used platform injection DMTs.

Generic Conversion

GA is available in 20 mg daily and 40 mg3 times weekly subcutaneous injection dosing. The first evidence of clinical efficacy for a generic formulation for GA was evaluated by the GATE trial.⁷ This trial was a multicenter, randomized, double-blind, active- and placebo-controlled phase 3 trial. Eligible participants were randomized to receive daily SC injection for 9 months of 20 mg generic GA (n = 5,353), 20 mg brand GA (n = 5,357), or placebo (n = 584). The primary endpoint was the mean number of gadolinium (Gd1) lesions visible on MRIs during months 7, 8, and 9, which were significantly reduced in the combined GA-treated group and in each GA group individually when compared with the placebo group, confirming the study sensitivity (ie, GA was effective under the conditions of the study). Tolerability (including injection site reactions) and safety (incidence, spectrum, and severity of adverse events [AEs]) were similar in the generic and brand GA groups. These results demonstrated that generic and brand GA had equivalent efficacy, tolerability, and safety over a 9-month period.⁷

Results of a 15-month extension of the study were presented in 2015 and showed similar efficacy, safety, and tolerability in participants treated with generic GA for 2 years and patients switched from brand to generic GA.⁸ Multiple shifts for GA occurred, most notably the conversion from branded Copaxone (Teva Pharmaceutical Industries) to generic Glatopa (Sandoz). Subsequently, Sandoz released a generic 40 mg 3 times weekly formulation. Additionally, Mylan entered the generic GA market. With 3 competing manufacturers, internal data from the VHA indicated that it was able to negotiate a single source contract for this medication that provided a savings of $32,088,904.69 between September 2016 and May 2019.

The impact of generic conversions is just being realized. Soon, patents will begin to expire for oral DMTs, leading to an expected growth of generic alternatives. Already the FDA has approved 4 generic alternatives for teriflunomide, 3 for fingolimod (with 13 tentative approvals), and 15 generic alternatives for dimethyl fumarate (DMF). Implementation of therapeutic interchanges will be pursued by VHA as clinically supported by evidence.

Criteria for Use

PBM supports utilizing criteria to help guide providers on DMT options and promote safe, effective, and value-based selection of a DMT. The PBM creates monographs and criteria for use (CFU) for new medications. The monograph contains a literature evaluation of all studies available to date that concern both safety and efficacy of the new medication. Therapeutic alternatives also are presented and assessed for key elements that may determine the most safe and effective use. Additional safety areas for the new medications such as look-alike, sound-alike potential, special populations use (ie, those who are pregnant, the elderly, and those with liver or renal dysfunction), and drug-drug interactions are presented. Lastly, and possibly most importantly in an ever-growing growing world of DMTs, the monograph describes a reasonable place in therapy for the new DMT.

CFU are additional guidance for some DMTs. The development of CFU are based on several questions that arise during the monograph development for a DMT. These include, but are not limited to:

• Are there safety concerns that require the drug to receive a review to ensure safe prescribing (eg, agents with REMS programs, or safety concerns in specific populations)?
• Does the drug require a specialty provider type with knowledge and experience in those disease states to ensure appropriate and safe prescribing (eg restricted to infectious diseases)?
• Do VHA or non-VHA guidelines suggest alternative therapy be used prior to the agent?
• Is a review deemed necessary to ensure the preferred agent is used first (eg, second-line therapy)?

The CFU defines parameters of drug use consistent with high quality and evidence-based patient care. CFUs also serve as a basis for monitoring local, regional, and national patterns of pharmacologic care and help guide health care providers (HCPs) on appropriate use of medication.

CFUs are designed to ensure the HCP is safely starting a medication that has evidence for efficacy for their patient. For example, alemtuzumab is a high-risk, high-efficacy DMT. The alemtuzumab CFU acknowledges this by having exclusion criteria that prevent a veteran at high risk (ie, on another immunosuppressant) from being exposed to severe AEs (ie, severe leukopenia) that are associated with the medication. On the other hand, the inclusion criteria recognize the benefits of alemtuzumab and allows those with highly active MS who have failed other DMTs to receive the medication.

The drug monograph and CFU process is an important part of VHA efforts to optimize patient care. After a draft version is developed, HCPs can provide feedback on the exclusion/inclusion criteria and describe how they anticipate using the medication in their practice. This insight can be beneficial for MS treatment as diverse HCPs may have distinct viewpoints on how DMTs should be started. Pharmacists and physicians on a national level then discuss and decide together what to include in the final drafts of the drug monograph and CFU. Final documents are disseminated to all sites, which encourages consistent practices across the VHA.⁹ These documents are reviewed on a regular basis and updated as needed based on available literature evidence.

It is well accepted that early use of DMT correlates with lower accumulated long-term disability.¹⁰ However, discontinuation of DMT should be treated with equal importance. This benefits the patient by reducing their risk of AEs from DMTs and provides cost savings. Age and disease stability are factors to consider for DMT discontinuation. In a study with patients aged > 45 years and another with patients aged > 60 years, discontinuing DMT rarely had a negative impact and improved quality of life.^11,12 A retrospective meta-analysis of age-dependent efficacy of current DMTs predicted that DMT loses efficacy at age 53 years. In addition, higher efficacy DMT only outperforms lower efficacy DMT in patients aged < 40.5 years.¹³ Stability of disease and lack of relapses for ≥ 2 years also may be a positive predictor to safely discontinue DMT.^14,15 The growing literature to support safe discontinuation of DMT makes this a more convincing strategy to avoid unnecessary costs associated with current DMTs. With an average age of 59 years for veterans with MS, this may be one of the largest areas of cost avoidance to consider.

Off-Label Use

Other potential ways to reduce DMT costs is to consider off-label treatments. The OLYMPUS trial studied off-label use of rituximab, an anti-CD20 antibody like ocrelizumab. It did not meet statistical significance for its primary endpoint; however, in a subgroup analysis, off-label use was found to be more effective in a population aged < 51 years.¹⁶ Other case reports and smaller scale studies also describe rituximab’s efficacy in MS.^17,18 In 2018, the FDA approved the first rituximab biosimilar.¹⁹ Further competition from biosimilars likely will make rituximab an even more cost-effective choice when compared with ocrelizumab.

Alternate Dosing Regimens

Extended interval dosing of natalizumab has been studied, extending the standard infusion interval from every 4 weeks to 5- to 8-week intervals. One recent article compared these interval extensions and found that all extended intervals of up to 56 days did not increase new or enhancing lesions on MRI when compared with standard interval dosing.²⁰ Another larger randomized trial is underway to evaluate efficacy and safety of extended interval dosing of natalizumab (NCT03689972). Utilization of this dosing may reduce natalizumab annual costs by up to 50%.

Safety Monitoring

DMF is an oral DMT on the VHA formulary with CFU. Since leukopenia is a known AE, baseline and quarterly monitoring of the complete blood count (CBC) is recommended for patients taking DMF. Additionally, DMF should be held if white blood cell count (WBC) falls below 2,000/mm³.²¹ There have been recent reports of death secondary to progressive multifocal leukoencephalopathy (PML) among European patients taking DMF.^22-24 This has raised concerns about adherence to recommended CBC monitoring in veterans taking DMF. The association of DMF and leukopenia has been evident since early clinical trials.²⁵ Leukopenia in immunocompromised patients increases the risk of PML.

In the long-term extension study ENDORSE, 6% to 7% of patients continuing DMF had WBC counts of 3.0×10⁹/L compared with 7% to 10% in the new to DMF group.²⁶ In addition 6% to 8% of patients continuing DMF had lymphocyte counts of 0.5×10⁹/L, compared with 5% to 9% in the new to DMF group. The cases of PML occurred in patients who had low lymphocyte counts over an extended period with no adjustment to DMF therapy, such as holding the drug until WBC counts returned to normal levels or stopping the drug. Discussion and review within VHA resulted in the recommendation for quarterly WBC monitoring criteria.

PBM and VA Center for Medication Safety (MedSafe) conducted a medication usage evaluation (MUE) on adherence to the WBC monitoring set forth in the CFU. Data collection began in fourth quarter of fiscal year (FY) 2015 with the most recent reporting period of fourth quarter of FY 2017. The Medication Utilization Evaluation Tool tracks patients with no reported WBC in 90 days and WBC < 2,000/mm³. Over the reporting period, 20% to 23% of patients have not received appropriate quarterly monitoring. Additionally, there have been 4 cases where the WBC decreased below the threshold limit. To ensure safe and effective use of DMF, it is important to adhere to the monitoring requirements set forth in the CFU.

Impact of REMS and Special Distribution

As DMTs increase in efficacy, there are often more risks associated with them. Some of these high-risk medications, including natalizumab and alemtuzumab, have REMS programs and/or have special distribution procedures. Although REMS are imperative for patient safety, the complexity of these programs can be difficult to navigate, which can create a barrier to access. The PBM helps to assist all sites with navigating and adhering to required actions to dispense and administer these medications through a national Special Handling Drugs Microsoft SharePoint site, which provides access to REMS forms and procurement information when drugs are dispensed from specialty pharmacies. Easing this process nationwide empowers more sites to be confident they can dispense specialty medications appropriately.

Clinical Pharmacists

The VHA is unique in its utilization of pharmacists in outpatient clinic settings. Utilization of an interdisciplinary team for medication management has been highly used in VHA for areas like primary care; however, pharmacist involvement in specialty areas is on the rise and MS is no exception. Pharmacists stationed in clinics, such as neurology or spinal cord injury, can impact care for veterans with MS. Interdisciplinary teams that include a pharmacist have been shown to increase patient adherence to DMTs.²⁷ However, pharmacists often assist with medication education and monitoring, which adds an additional layer of safety to DMT treatment. At the VHA, pharmacists also can obtain a scope of practice that allows them to prescribe medications and increase access to care for veterans with MS.

Education

The VHA demonstrates how education on a disease state like MS can be distributed on a large, national scale through drug monographs, CFU, and Microsoft SharePoint sites. In addition, VHA has created the MS Centers of Excellence (MSCoE) that serve as a hub of specialized health care providers in all aspects of MS care.

A core function of the MSCoE is to provide education to both HCPs and patients. The MSCoE and its regional hubs support sites that may not have an HCP who specializes in MS by providing advice on DMT selection, how to obtain specialty medications, and monitoring that needs to be completed to ensure veterans’ safety. The MSCoE also has partnered with the National MS Society to hold a lecture series on topics in MS. This free series is available online to all HCPs who interact with patients who have MS and is a way that VA is extending its best practices and expertise beyond its own health care system. There also is a quarterly newsletter for veterans with MS that highlights new information on DMTs that can affect their care.

Conclusion

It is an exciting and challenging period in MS treatment. New DMTs are being approved and entering clinical trials at a rapid pace. These new DMT agents may offer increased efficacy, improvements in AE profiles, and the possibility of increased medication adherence—but often at a higher cost. The utilization of CFU and formulary management provides the ability to ensure the safe and appropriate use of medications by veterans, with a secondary outcome of controlling pharmacy expenditures.

The VHA had expenditures of $142,135,938 for DMT use in FY 2018. As the VHA sees the new contract prices for DMT in January 2020, we are reminded that costs will continue to rise with some pharmaceutical manufacturers implementing prices 8% to 11% higher than 2019 prices, when the consumer price index defines an increase of 1.0% for 2020 and 1.4% in 2021.²⁸ It is imperative that the VHA formulary be managed judiciously and the necessary measures be in place for VHA practitioners to enable effective, safe and value-based care to the veteran population.

Prior to the first approved disease modifying therapy (DMT) in the 1990s, treatment approaches for multiple sclerosis (MS) were not well understood. The discovery that MS was an immune mediated inflammatory disease paved the way for the treatments we know today. In 1993, interferon β‐1b became the first DMT for MS approved by the US Food and Drug Administration (FDA). Approvals for interferon β‐1a as well as glatiramer acetate (GA) soon followed. Today, we consider these the mildest immunosuppressant DMTs; however, their success verified that suppressing the immune system had a positive effect on the MS disease process.

Following these approvals, the disease process in MS is now better understood. Recently approved therapies include monoclonal antibodies, which affect other immune pathways. Today, there are 14 approved DMTs (Table 1). Although the advent of these newer DMTs has revolutionized care for patients with MS, it has been accompanied by increasing costs for the agents. Direct medical costs associated with MS management, coupled with indirect costs from lost productivity, have been estimated to be $24.2 billion annually in the US.¹ These increases have been seen across many levels of insurance coverage—private payer, Medicare, and the Veterans Health Administration (VHA).^2,3

fdp03701036s_f.png

The Figure demonstrates the cost increase that have been seen across VHA between 2004 and 2019 for the DMTs identified in Table 1. Indeed, this compound annual growth rate may be an underestimate because infusion therapies (eg, natalizumab, ocrelizumab, and alemtuzumab) are difficult to track as they may be dispensed directly via a Risk Evaluation Medication Strategy (REMS) program. According to the VHA Pharmacy Benefit Management Service (PBM), in September 2019, dimethyl fumarate (DMF) had the 13th highest total outpatient drug cost for the US Department of Veterans Affairs (VA), interferon β‐1a ranked 62nd and 83rd (prefilled pen and syringe, respectively), and GA 40 mg ranked 89th.

fdp03701036s_t1.png

Factors Impacting DMT Use

Recent changes to MS typing have impacted utilization of DMTs. Traditionally, there were 4 subtypes of MS: relapsing remitting (RRMS), secondary progressive (SPMS), progressive relapsing (PRMS), and primary progressive (PPMS). These subtypes are now viewed more broadly and grouped as either relapsing or progressive. The traditional subtypes fall under these broader definitions. Additionally, SPMS has been broken into active SPMS, characterized by continued worsening of disability unrelated to acute relapses, superimposed with activity that can be seen on magnetic resonance images (MRIs), and nonactive SPMS, which has the same disability progression as active SPMS but without MRI-visible activity.^4-6 In 2019, these supplementary designations to SPMS made their first appearance in FDA-approved indications. All existing DMTs now include this terminology in their labelling and are indicated in active SPMS. There remain no DMTs that treat nonactive SPMS.

The current landscape of DMTs is highly varied in method of administration, risks, and benefits. As efficacy of these medications often is marked by how well they can prevent the immune system from attacking myelin, an inverse relationship between safety and efficacy results. The standard treatment outcomes in MS have evolved over time. The following are the commonly used primary outcomes in clinical trials: relapse reduction; increased time between relapses; decreased severity of relapses; prevention or extend time to disability milestones as measured by the Expanded Disability Status Scale (EDSS) and other disability measures; prevention or extension of time to onset of secondary progressive disease; prevention or reduction of the number and size of new and enhancing lesions on MRI; and limitation of overall MRI lesion burden in the central nervous system (CNS).

Newer treatment outcomes employed in more recent trials include: measures of axonal damage, CNS atrophy, evidence of microscopic disease via conventional MRI and advanced imaging modalities, biomarkers associated with inflammatory disease activity and neurodegeneration in MS, and the use of no evidence of disease activity (NEDA). These outcomes also must be evaluated by the safety concerns of each agent. Short- and long-term safety are critical factors in the selection of DMTs for MS. The injectable therapies for MS (interferon β‐1a, interferon β‐1b, and GA) have established long-term safety profiles from > 20 years of continuous use. The long-term safety profiles of oral immunomodulatory agents and monoclonal antibodies for these drugs in MS have yet to be determined. Safety concerns associated with some therapies and added requirements for safety monitoring may increase the complexity of a therapeutic selection.

fdp03701036s_t2.png

Current cost minimization strategies for DMT include limiting DMT agents on formularies, tier systems that incentivize patients/prescribers to select the lowest priced agents on the formulary, negotiating arrangements with manufacturers to freeze prices or provide discounts in exchange for a priority position in the formulary, and requiring prior authorization to initiate or switch therapy. The use of generic medications and interchange to these agents from a brand name formulation can help reduce expense. Several of these strategies have been implemented in VHA.

Disease-Modifying Therapies

In 2019, 18,645 veterans with MS had either a MS-specific DMT or ≥ 1 annual encounters with a primary diagnosis of MS. Of this population, 4,720 were female and 13,357 were service connected according to VA data. About 50% of veterans with MS take a DMT. This percentage has remained stable over the past decade (Table 2). Although it appears the number of unique veterans prescribed an outpatient DMT is decreasing, this does not include the growing use of infused DMTs or DMTs obtained through the Veterans Choice Program (VCP)/Community Care (CC).

The overall outpatient pharmacy costs for veterans have remained constant despite the reduction in outpatient pharmacy prescription numbers. This may be due to increases in DMT cost to the VHA and the use of more expensive oral agents over the previously used platform injection DMTs.

Generic Conversion

GA is available in 20 mg daily and 40 mg3 times weekly subcutaneous injection dosing. The first evidence of clinical efficacy for a generic formulation for GA was evaluated by the GATE trial.⁷ This trial was a multicenter, randomized, double-blind, active- and placebo-controlled phase 3 trial. Eligible participants were randomized to receive daily SC injection for 9 months of 20 mg generic GA (n = 5,353), 20 mg brand GA (n = 5,357), or placebo (n = 584). The primary endpoint was the mean number of gadolinium (Gd1) lesions visible on MRIs during months 7, 8, and 9, which were significantly reduced in the combined GA-treated group and in each GA group individually when compared with the placebo group, confirming the study sensitivity (ie, GA was effective under the conditions of the study). Tolerability (including injection site reactions) and safety (incidence, spectrum, and severity of adverse events [AEs]) were similar in the generic and brand GA groups. These results demonstrated that generic and brand GA had equivalent efficacy, tolerability, and safety over a 9-month period.⁷

Results of a 15-month extension of the study were presented in 2015 and showed similar efficacy, safety, and tolerability in participants treated with generic GA for 2 years and patients switched from brand to generic GA.⁸ Multiple shifts for GA occurred, most notably the conversion from branded Copaxone (Teva Pharmaceutical Industries) to generic Glatopa (Sandoz). Subsequently, Sandoz released a generic 40 mg 3 times weekly formulation. Additionally, Mylan entered the generic GA market. With 3 competing manufacturers, internal data from the VHA indicated that it was able to negotiate a single source contract for this medication that provided a savings of $32,088,904.69 between September 2016 and May 2019.

The impact of generic conversions is just being realized. Soon, patents will begin to expire for oral DMTs, leading to an expected growth of generic alternatives. Already the FDA has approved 4 generic alternatives for teriflunomide, 3 for fingolimod (with 13 tentative approvals), and 15 generic alternatives for dimethyl fumarate (DMF). Implementation of therapeutic interchanges will be pursued by VHA as clinically supported by evidence.

Criteria for Use

PBM supports utilizing criteria to help guide providers on DMT options and promote safe, effective, and value-based selection of a DMT. The PBM creates monographs and criteria for use (CFU) for new medications. The monograph contains a literature evaluation of all studies available to date that concern both safety and efficacy of the new medication. Therapeutic alternatives also are presented and assessed for key elements that may determine the most safe and effective use. Additional safety areas for the new medications such as look-alike, sound-alike potential, special populations use (ie, those who are pregnant, the elderly, and those with liver or renal dysfunction), and drug-drug interactions are presented. Lastly, and possibly most importantly in an ever-growing growing world of DMTs, the monograph describes a reasonable place in therapy for the new DMT.

CFU are additional guidance for some DMTs. The development of CFU are based on several questions that arise during the monograph development for a DMT. These include, but are not limited to:

• Are there safety concerns that require the drug to receive a review to ensure safe prescribing (eg, agents with REMS programs, or safety concerns in specific populations)?
• Does the drug require a specialty provider type with knowledge and experience in those disease states to ensure appropriate and safe prescribing (eg restricted to infectious diseases)?
• Do VHA or non-VHA guidelines suggest alternative therapy be used prior to the agent?
• Is a review deemed necessary to ensure the preferred agent is used first (eg, second-line therapy)?

The CFU defines parameters of drug use consistent with high quality and evidence-based patient care. CFUs also serve as a basis for monitoring local, regional, and national patterns of pharmacologic care and help guide health care providers (HCPs) on appropriate use of medication.

CFUs are designed to ensure the HCP is safely starting a medication that has evidence for efficacy for their patient. For example, alemtuzumab is a high-risk, high-efficacy DMT. The alemtuzumab CFU acknowledges this by having exclusion criteria that prevent a veteran at high risk (ie, on another immunosuppressant) from being exposed to severe AEs (ie, severe leukopenia) that are associated with the medication. On the other hand, the inclusion criteria recognize the benefits of alemtuzumab and allows those with highly active MS who have failed other DMTs to receive the medication.

The drug monograph and CFU process is an important part of VHA efforts to optimize patient care. After a draft version is developed, HCPs can provide feedback on the exclusion/inclusion criteria and describe how they anticipate using the medication in their practice. This insight can be beneficial for MS treatment as diverse HCPs may have distinct viewpoints on how DMTs should be started. Pharmacists and physicians on a national level then discuss and decide together what to include in the final drafts of the drug monograph and CFU. Final documents are disseminated to all sites, which encourages consistent practices across the VHA.⁹ These documents are reviewed on a regular basis and updated as needed based on available literature evidence.

It is well accepted that early use of DMT correlates with lower accumulated long-term disability.¹⁰ However, discontinuation of DMT should be treated with equal importance. This benefits the patient by reducing their risk of AEs from DMTs and provides cost savings. Age and disease stability are factors to consider for DMT discontinuation. In a study with patients aged > 45 years and another with patients aged > 60 years, discontinuing DMT rarely had a negative impact and improved quality of life.^11,12 A retrospective meta-analysis of age-dependent efficacy of current DMTs predicted that DMT loses efficacy at age 53 years. In addition, higher efficacy DMT only outperforms lower efficacy DMT in patients aged < 40.5 years.¹³ Stability of disease and lack of relapses for ≥ 2 years also may be a positive predictor to safely discontinue DMT.^14,15 The growing literature to support safe discontinuation of DMT makes this a more convincing strategy to avoid unnecessary costs associated with current DMTs. With an average age of 59 years for veterans with MS, this may be one of the largest areas of cost avoidance to consider.

Off-Label Use

Other potential ways to reduce DMT costs is to consider off-label treatments. The OLYMPUS trial studied off-label use of rituximab, an anti-CD20 antibody like ocrelizumab. It did not meet statistical significance for its primary endpoint; however, in a subgroup analysis, off-label use was found to be more effective in a population aged < 51 years.¹⁶ Other case reports and smaller scale studies also describe rituximab’s efficacy in MS.^17,18 In 2018, the FDA approved the first rituximab biosimilar.¹⁹ Further competition from biosimilars likely will make rituximab an even more cost-effective choice when compared with ocrelizumab.

Alternate Dosing Regimens

Extended interval dosing of natalizumab has been studied, extending the standard infusion interval from every 4 weeks to 5- to 8-week intervals. One recent article compared these interval extensions and found that all extended intervals of up to 56 days did not increase new or enhancing lesions on MRI when compared with standard interval dosing.²⁰ Another larger randomized trial is underway to evaluate efficacy and safety of extended interval dosing of natalizumab (NCT03689972). Utilization of this dosing may reduce natalizumab annual costs by up to 50%.

Safety Monitoring

DMF is an oral DMT on the VHA formulary with CFU. Since leukopenia is a known AE, baseline and quarterly monitoring of the complete blood count (CBC) is recommended for patients taking DMF. Additionally, DMF should be held if white blood cell count (WBC) falls below 2,000/mm³.²¹ There have been recent reports of death secondary to progressive multifocal leukoencephalopathy (PML) among European patients taking DMF.^22-24 This has raised concerns about adherence to recommended CBC monitoring in veterans taking DMF. The association of DMF and leukopenia has been evident since early clinical trials.²⁵ Leukopenia in immunocompromised patients increases the risk of PML.

In the long-term extension study ENDORSE, 6% to 7% of patients continuing DMF had WBC counts of 3.0×10⁹/L compared with 7% to 10% in the new to DMF group.²⁶ In addition 6% to 8% of patients continuing DMF had lymphocyte counts of 0.5×10⁹/L, compared with 5% to 9% in the new to DMF group. The cases of PML occurred in patients who had low lymphocyte counts over an extended period with no adjustment to DMF therapy, such as holding the drug until WBC counts returned to normal levels or stopping the drug. Discussion and review within VHA resulted in the recommendation for quarterly WBC monitoring criteria.

PBM and VA Center for Medication Safety (MedSafe) conducted a medication usage evaluation (MUE) on adherence to the WBC monitoring set forth in the CFU. Data collection began in fourth quarter of fiscal year (FY) 2015 with the most recent reporting period of fourth quarter of FY 2017. The Medication Utilization Evaluation Tool tracks patients with no reported WBC in 90 days and WBC < 2,000/mm³. Over the reporting period, 20% to 23% of patients have not received appropriate quarterly monitoring. Additionally, there have been 4 cases where the WBC decreased below the threshold limit. To ensure safe and effective use of DMF, it is important to adhere to the monitoring requirements set forth in the CFU.

Impact of REMS and Special Distribution

As DMTs increase in efficacy, there are often more risks associated with them. Some of these high-risk medications, including natalizumab and alemtuzumab, have REMS programs and/or have special distribution procedures. Although REMS are imperative for patient safety, the complexity of these programs can be difficult to navigate, which can create a barrier to access. The PBM helps to assist all sites with navigating and adhering to required actions to dispense and administer these medications through a national Special Handling Drugs Microsoft SharePoint site, which provides access to REMS forms and procurement information when drugs are dispensed from specialty pharmacies. Easing this process nationwide empowers more sites to be confident they can dispense specialty medications appropriately.

Clinical Pharmacists

The VHA is unique in its utilization of pharmacists in outpatient clinic settings. Utilization of an interdisciplinary team for medication management has been highly used in VHA for areas like primary care; however, pharmacist involvement in specialty areas is on the rise and MS is no exception. Pharmacists stationed in clinics, such as neurology or spinal cord injury, can impact care for veterans with MS. Interdisciplinary teams that include a pharmacist have been shown to increase patient adherence to DMTs.²⁷ However, pharmacists often assist with medication education and monitoring, which adds an additional layer of safety to DMT treatment. At the VHA, pharmacists also can obtain a scope of practice that allows them to prescribe medications and increase access to care for veterans with MS.

Education

The VHA demonstrates how education on a disease state like MS can be distributed on a large, national scale through drug monographs, CFU, and Microsoft SharePoint sites. In addition, VHA has created the MS Centers of Excellence (MSCoE) that serve as a hub of specialized health care providers in all aspects of MS care.

A core function of the MSCoE is to provide education to both HCPs and patients. The MSCoE and its regional hubs support sites that may not have an HCP who specializes in MS by providing advice on DMT selection, how to obtain specialty medications, and monitoring that needs to be completed to ensure veterans’ safety. The MSCoE also has partnered with the National MS Society to hold a lecture series on topics in MS. This free series is available online to all HCPs who interact with patients who have MS and is a way that VA is extending its best practices and expertise beyond its own health care system. There also is a quarterly newsletter for veterans with MS that highlights new information on DMTs that can affect their care.

Conclusion

It is an exciting and challenging period in MS treatment. New DMTs are being approved and entering clinical trials at a rapid pace. These new DMT agents may offer increased efficacy, improvements in AE profiles, and the possibility of increased medication adherence—but often at a higher cost. The utilization of CFU and formulary management provides the ability to ensure the safe and appropriate use of medications by veterans, with a secondary outcome of controlling pharmacy expenditures.

The VHA had expenditures of $142,135,938 for DMT use in FY 2018. As the VHA sees the new contract prices for DMT in January 2020, we are reminded that costs will continue to rise with some pharmaceutical manufacturers implementing prices 8% to 11% higher than 2019 prices, when the consumer price index defines an increase of 1.0% for 2020 and 1.4% in 2021.²⁸ It is imperative that the VHA formulary be managed judiciously and the necessary measures be in place for VHA practitioners to enable effective, safe and value-based care to the veteran population.

Parkinson disease (PD) is a progressive neurodegenerative disorder, characterized by diverse clinical symptoms. PD can present with rest tremor, bradykinesia, rigidity, falls, postural instability, and multiple nonmotor symptoms. Marras and colleagues estimated in a comprehensive meta-analysis that there were 680,000 individuals with PD in the US in 2010; this number is expected to double by 2030 based on the US Census Bureau population projections.¹ An estimated 110,000 veterans may be affected by PD; hence, understanding of PD pathology, clinical progression, and effective treatment strategies is of paramount importance to the Veterans Health Administration (VHA).²

The exact pathogenesis underlying clinical features is still being studied. Pathologic diagnosis of PD relies on loss of dopamine neurons in the substantia nigra and accumulation of the abnormal protein, α-synuclein, in the form of Lewy bodies and Lewy neurites. Lewy bodies and neurites accumulate predominantly in the substantia nigra in addition to other brain stem nuclei and cerebral cortex. Lewy bodies are intraneuronal inclusions with a hyaline core and a pale peripheral halo. Central core stains positive for α-synuclein.^3,4 Lewy neurites are widespread and are believed to play a larger role in the pathogenesis of PD compared with those of Lewy bodies.⁵

α-Synuclein

α-synuclein is a small 140 amino-acid protein with a N-terminal region that can interact with cell membranes and a highly acidic unstructured C-terminal region.⁶ α-synuclein is physiologically present in the presynaptic terminals of neurons and involved in synaptic plasticity and vesicle trafficking.⁷ There are different hypotheses about the native structure of α-synuclein. The first suggests that it exists in tetrameric form and may be broken down to monomer, which is the pathogenic form of α-synuclein. The second hypothesis suggests that it exists primarily in monomeric form, whereas other studies have shown that both forms exist and with pathologic changes, monomer accumulates in abundance and is neurotoxic.^8-11 Work by Burré and colleagues shows that native α-synuclein exists in 2 forms: a soluble, cytosolic α-synuclein, which is monomeric, and a membrane-bound multimeric form.^12,13

Alteration in aggregation properties of this protein is believed to play a central role in the pathogenesis of PD.^14,15 Pathologic α-synuclein exists in insoluble forms that can aggregate into oligomers and fibrillar structures.¹⁶ Lysosomal dysfunction may promote accumulation of insoluble α-synuclein. Prior work has shown that several degradation pathways in lysosomes, including the ubiquitin-proteasome system and autophagy-lysosomal pathway, are down regulated, thus contributing to the accumulation of abnormal α-synuclein.^17,18 Accumulation of pathologic α-synuclein leads to mitochondrial dysfunction in PD animal models, contributing further to neurotoxicity.^19,20 Aggregates of phosphorylated α-synuclein have been demonstrated in dementia with Lewy body.²¹

In addition, α-synuclein aggregates may be released into extracellular spaces to be taken up by adjacent cells, where they can cause further misfolding and aggregation of protein.²² Previous work in animal models suggested a prion proteinlike spread of α-synuclein.²³ This finding can have long-term therapeutic implications, as preventing extracellular release of abnormal form of α-synuclein will prevent the spread of pathologic protein. This can form the basis of neuroprotection in patients with PD.²⁴

It has been proposed that α-synuclein accumulation and extracellular release initiates an immune response that leads to activation of microglia. This has been shown in PD animal models, overexpressing α-synuclein. In 2008 Park and colleagues demonstrated that microglial activation is enhanced by monomeric α-synuclein, not by the aggregated variant.²⁵ Other studies have reported activated microglia around dopaminergic cells in substantia nigra.²⁶ Sulzer and colleagues showed that peptides from α-synuclein can act as antigens and trigger an autoimmune reaction via T cells.²⁷ PD may be associated with certain HLA-haplotypes.²⁸ In other words, α-synuclein can induce neurodegeneration via different mechanisms, including alteration in synaptic vesicle transmission, mitochondrial dysfunction, neuroinflammation, and induction of humoral immunity.

Immunization

Due to these observations, there had been huge interest in developing antibody-based therapies for PD. A similar approach had been tested in Alzheimer disease (AD). Intracellular tangles of tau protein and extracellular aggregates of amyloid are the pathologic substrates in AD. Clinical trials utilizing antibodies targeting amyloid showed reduction in abnormal protein accumulation but no significant improvement in cognition.²⁹ In addition, adverse events (AEs), such as vasogenic edema and intracerebral hemorrhage, were reported.³⁰ Careful analysis of the data suggested that inadequate patient selection or targeting only amyloid, may have contributed to unfavorable results.³¹ Since then, more recent clinical trials have focused on careful patient selection, use of second generation anti-amyloid antibodies and immunotherapies targeting tau.³²

Several studies have tested immunotherapies in PD animal models with the aim of targeting α-synuclein. Immunotherapies can be instituted in 2 ways: active immunization in which the immune system is stimulated to produce antibodies against α-synuclein or passive immunization in which antibodies against α-synuclein are administered directly. Once α-synuclein antibodies have crossed the blood-brain barrier, they are hypothesized to clear the existing α-synuclein. Animal studies have demonstrated the presence of these antibodies within the neurons. The mechanism of entry is unknown. Once inside the cells, the antibodies activate the lysosomal clearance, affecting intracellular accumulation of α-synuclein. Extracellularly, they can bind to receptors on scavenger cells, mainly microglia, activating them to facilitate uptake of extracellular α-synuclein. Binding of the antibodies to α-synuclein directly prevents the uptake of toxic protein by the cells, blocking the transfer and spread of PD pathology.³³

Active Immunization

Active immunization against α-synuclein was demonstrated by Masliah and colleagues almost a decade ago. They administered recombinant human α-synuclein in transgenic mice expressing α-synuclein under the control of platelet-derived growth factor β. Reduction of accumulated α-synuclein in neurons with mild microglia activation was noted. It was proposed that the antibodies produced were able to bind to abnormal α-synuclein, were recognized by the lysosomal pathways, and degraded.³⁴ Ghochikyan and colleagues developed vaccines by using α-synuclein-derived peptides. This induced formation of antibodies against α-synuclein in Lewy-bodies and neurites.³⁵ Over time, other animal studies have been able to expand on these results.³⁶

AFFiRiS, an Austrian biotechnology company, has developed 2 peptide vaccines PD01A and PD03A. Both peptides when administered to PD animal models caused antibody-based immune response against aggregated α-synuclein. Humoral autoimmune response was not observed in these studies; no neuroinflammation or neurotoxicity was noted. These peptides did not affect levels of physiologic α-synuclein, targeting only the aggregated form.³⁷ These animal models showed improved motor and cognitive function. Similar results were noted in multiple system atrophy (MSA) animal models.^38,39

The first human phase 1, randomized, parallel-group, single-center study recruited 32 subjects with early PD. Twelve subjects each were included in low- or high-dose treatment group, and 8 were included in the control group. Test subjects randomly received 4 vaccinations of low- or high-dose PD01A. Both doses were well tolerated, and no drug-related serious AEs were reported. The study confirmed the tolerability and safety of subcutaneous PD01A vaccine administration. These subjects were included in a 12-month, phase 1b follow-up extension study, AFF008E. In 2018, it was reported that administration of 6 doses of PD01A, 4 primary and 2 booster immunization, was safe. The vaccine showed a clear immune response against the peptide and cross-reactivity against α-synuclein targeted epitope. Booster doses stabilized the antibody titers. Significant increase in antibody titers against PD01A was seen over time, which was translated into a humoral immune response against α-synuclein. In addition, PD01A antibodies also were reported in cerebrospinal fluid.⁴⁰

AFFiRiS presented results of a phase 1 randomized, placebo-controlled trial in 2017, confirming the safety of PD03A in patients with PD. The study showed a clear dose-dependent immune response against the peptide and cross-reactivity against α-synuclein targeted epitope.⁴¹ AFFiRiS recently presented results of another phase 1 clinical study assessing the safety and tolerability of vaccines PD01A and PD03A in patients with early MSA. Both vaccines were well tolerated, and PD01A induced an immune response against the peptide and α-synuclein epitope.⁴² These results have provided hope for further endeavors to develop active immunization strategies for PD.

Passive Immunization

Passive immunization against α-synuclein was first reported by Masliah and colleagues in 2011. A monoclonal antibody against the C-terminus of α-synuclein, 9E4, was injected into a transgenic mouse model of PD. There was reduction in α-synuclein aggregates in the brain along with improvement in motor and cognitive impairment.⁴³ The C-terminus of α-synuclein plays a key role in the pathogenesis of PD. Changes in the C-terminus of α-synuclein induces formation of α-synuclein oligomers and subsequent neuronal spread. Antibody binds to the C-terminus and prevents structural changes that can lead to oligomerization of α-synuclein. Since the first study by Masliah, few other immunization studies utilized different antibodies against the C-terminus of α-synuclein. It was shown in a mouse model that binding of such antibodies promoted clearance of the α-synuclein by microglia.⁴⁴

Based on these animal studies, Prothena Biosciences (South San Francisco, CA) designed a phase 1, double-blind, randomized, placebo-controlled clinical trial of prasinezumab (investigational monoclonal antibody against C-terminus of α-synuclein), in subjects without PD. The results showed that it was well tolerated, and there was dose-dependent reduction in the levels of free α-synuclein in plasma.⁴⁵ A 6-month phase 1b trial to evaluate the safety, tolerability and immune system response to multiple ascending doses of prasinezumab via IV infusion once every 28 days was conducted in 64 patients with PD. The drug was found to be safe, and levels of free serum α-synuclein were reduced up to 97%.⁴⁶ Roche (Basel, Switzerland) and Prothena are conducting a multicenter, randomized, double-blind phase 2 trial in patients with early PD to evaluate the efficacy of prasinezumab vs placebo.⁴⁷

BIIB054 is another monoclonal antibody that targets the N-terminal of α-synuclein. In animal models, antibodies targeting the N-terminus reduced α-synuclein triggered cell death and reduced the number of activated microglia.⁴⁸ BIIB054, from Biogen (Cambridge, MA), was studied in 40 healthy subjects and was well tolerated with a favorable safety profile and could cross the blood-brain barrier. Like the prasinezumab study, this also was an ascending-dose study to assess safety and tolerability. In 2018, a randomized, double-blind, placebo-controlled, single-ascending dose study in patients with PD reported that BIIB054 was well tolerated, and the presence of BIIB054-synuclein complexes in the plasma were confirmed.⁴⁹ A phase 2, multicenter, randomized, double-blind, placebo-controlled study (SPARK) with an active-treatment dose-blinded period, designed to evaluate the safety, pharmacokinetics, and the pharmacodynamics of BIIB054 is currently recruiting patients with PD.

Finally, BioArctic (Stockholm, Sweden) developed antibodies that are selective for oligomeric forms of α-synuclein, which it licensed to AbbVie (North Chicago, Il).⁵⁰ These antibodies do not target the N- or C-terminus of α-synuclein. Since α-synuclein oligomers play an important role in the pathogenesis of PD, targeting them with antibodies at an early stage may prove to be an effective strategy for removal of pathogenic α-synuclein. Clinical trials are forthcoming.

Conclusions

Immunotherapy against α-synuclein has provided a new therapeutic avenue in the field of neuroprotection. Results from the first human clinical trial are promising, but despite these results, more work is needed to clarify the role of α-synuclein in the pathogenesis of PD in humans. Most of the work concerning α-synuclein aggregation and propagation has been reported in animal models. Whether similar process exists in humans is a debatable question. Similarly, more knowledge is needed about how and where in the human brain antibodies act to give neuroprotective effects. Timing of administration of immunotherapies in real time will be a crucial question.

PD is clinically evident once 80% of dopaminergic neurons in substantia nigra are lost due to neurodegeneration. Should immunotherapy be administered to symptomatic patients with PD, or if it will be beneficial only for presymptomatic, high-risk patients needs to be determined. Like AD trials, not only careful selection of patients, but determination of optimal timing for treatment will be essential. As the understanding of PD pathogenesis and therapeutics evolves, it will become clear whether immunization targeting α-synuclein will modify disease progression.

Parkinson disease (PD) is a progressive neurodegenerative disorder, characterized by diverse clinical symptoms. PD can present with rest tremor, bradykinesia, rigidity, falls, postural instability, and multiple nonmotor symptoms. Marras and colleagues estimated in a comprehensive meta-analysis that there were 680,000 individuals with PD in the US in 2010; this number is expected to double by 2030 based on the US Census Bureau population projections.¹ An estimated 110,000 veterans may be affected by PD; hence, understanding of PD pathology, clinical progression, and effective treatment strategies is of paramount importance to the Veterans Health Administration (VHA).²

The exact pathogenesis underlying clinical features is still being studied. Pathologic diagnosis of PD relies on loss of dopamine neurons in the substantia nigra and accumulation of the abnormal protein, α-synuclein, in the form of Lewy bodies and Lewy neurites. Lewy bodies and neurites accumulate predominantly in the substantia nigra in addition to other brain stem nuclei and cerebral cortex. Lewy bodies are intraneuronal inclusions with a hyaline core and a pale peripheral halo. Central core stains positive for α-synuclein.^3,4 Lewy neurites are widespread and are believed to play a larger role in the pathogenesis of PD compared with those of Lewy bodies.⁵

α-Synuclein

α-synuclein is a small 140 amino-acid protein with a N-terminal region that can interact with cell membranes and a highly acidic unstructured C-terminal region.⁶ α-synuclein is physiologically present in the presynaptic terminals of neurons and involved in synaptic plasticity and vesicle trafficking.⁷ There are different hypotheses about the native structure of α-synuclein. The first suggests that it exists in tetrameric form and may be broken down to monomer, which is the pathogenic form of α-synuclein. The second hypothesis suggests that it exists primarily in monomeric form, whereas other studies have shown that both forms exist and with pathologic changes, monomer accumulates in abundance and is neurotoxic.^8-11 Work by Burré and colleagues shows that native α-synuclein exists in 2 forms: a soluble, cytosolic α-synuclein, which is monomeric, and a membrane-bound multimeric form.^12,13

Alteration in aggregation properties of this protein is believed to play a central role in the pathogenesis of PD.^14,15 Pathologic α-synuclein exists in insoluble forms that can aggregate into oligomers and fibrillar structures.¹⁶ Lysosomal dysfunction may promote accumulation of insoluble α-synuclein. Prior work has shown that several degradation pathways in lysosomes, including the ubiquitin-proteasome system and autophagy-lysosomal pathway, are down regulated, thus contributing to the accumulation of abnormal α-synuclein.^17,18 Accumulation of pathologic α-synuclein leads to mitochondrial dysfunction in PD animal models, contributing further to neurotoxicity.^19,20 Aggregates of phosphorylated α-synuclein have been demonstrated in dementia with Lewy body.²¹

In addition, α-synuclein aggregates may be released into extracellular spaces to be taken up by adjacent cells, where they can cause further misfolding and aggregation of protein.²² Previous work in animal models suggested a prion proteinlike spread of α-synuclein.²³ This finding can have long-term therapeutic implications, as preventing extracellular release of abnormal form of α-synuclein will prevent the spread of pathologic protein. This can form the basis of neuroprotection in patients with PD.²⁴

It has been proposed that α-synuclein accumulation and extracellular release initiates an immune response that leads to activation of microglia. This has been shown in PD animal models, overexpressing α-synuclein. In 2008 Park and colleagues demonstrated that microglial activation is enhanced by monomeric α-synuclein, not by the aggregated variant.²⁵ Other studies have reported activated microglia around dopaminergic cells in substantia nigra.²⁶ Sulzer and colleagues showed that peptides from α-synuclein can act as antigens and trigger an autoimmune reaction via T cells.²⁷ PD may be associated with certain HLA-haplotypes.²⁸ In other words, α-synuclein can induce neurodegeneration via different mechanisms, including alteration in synaptic vesicle transmission, mitochondrial dysfunction, neuroinflammation, and induction of humoral immunity.

Immunization

Due to these observations, there had been huge interest in developing antibody-based therapies for PD. A similar approach had been tested in Alzheimer disease (AD). Intracellular tangles of tau protein and extracellular aggregates of amyloid are the pathologic substrates in AD. Clinical trials utilizing antibodies targeting amyloid showed reduction in abnormal protein accumulation but no significant improvement in cognition.²⁹ In addition, adverse events (AEs), such as vasogenic edema and intracerebral hemorrhage, were reported.³⁰ Careful analysis of the data suggested that inadequate patient selection or targeting only amyloid, may have contributed to unfavorable results.³¹ Since then, more recent clinical trials have focused on careful patient selection, use of second generation anti-amyloid antibodies and immunotherapies targeting tau.³²

Several studies have tested immunotherapies in PD animal models with the aim of targeting α-synuclein. Immunotherapies can be instituted in 2 ways: active immunization in which the immune system is stimulated to produce antibodies against α-synuclein or passive immunization in which antibodies against α-synuclein are administered directly. Once α-synuclein antibodies have crossed the blood-brain barrier, they are hypothesized to clear the existing α-synuclein. Animal studies have demonstrated the presence of these antibodies within the neurons. The mechanism of entry is unknown. Once inside the cells, the antibodies activate the lysosomal clearance, affecting intracellular accumulation of α-synuclein. Extracellularly, they can bind to receptors on scavenger cells, mainly microglia, activating them to facilitate uptake of extracellular α-synuclein. Binding of the antibodies to α-synuclein directly prevents the uptake of toxic protein by the cells, blocking the transfer and spread of PD pathology.³³

Active Immunization

Active immunization against α-synuclein was demonstrated by Masliah and colleagues almost a decade ago. They administered recombinant human α-synuclein in transgenic mice expressing α-synuclein under the control of platelet-derived growth factor β. Reduction of accumulated α-synuclein in neurons with mild microglia activation was noted. It was proposed that the antibodies produced were able to bind to abnormal α-synuclein, were recognized by the lysosomal pathways, and degraded.³⁴ Ghochikyan and colleagues developed vaccines by using α-synuclein-derived peptides. This induced formation of antibodies against α-synuclein in Lewy-bodies and neurites.³⁵ Over time, other animal studies have been able to expand on these results.³⁶

AFFiRiS, an Austrian biotechnology company, has developed 2 peptide vaccines PD01A and PD03A. Both peptides when administered to PD animal models caused antibody-based immune response against aggregated α-synuclein. Humoral autoimmune response was not observed in these studies; no neuroinflammation or neurotoxicity was noted. These peptides did not affect levels of physiologic α-synuclein, targeting only the aggregated form.³⁷ These animal models showed improved motor and cognitive function. Similar results were noted in multiple system atrophy (MSA) animal models.^38,39

The first human phase 1, randomized, parallel-group, single-center study recruited 32 subjects with early PD. Twelve subjects each were included in low- or high-dose treatment group, and 8 were included in the control group. Test subjects randomly received 4 vaccinations of low- or high-dose PD01A. Both doses were well tolerated, and no drug-related serious AEs were reported. The study confirmed the tolerability and safety of subcutaneous PD01A vaccine administration. These subjects were included in a 12-month, phase 1b follow-up extension study, AFF008E. In 2018, it was reported that administration of 6 doses of PD01A, 4 primary and 2 booster immunization, was safe. The vaccine showed a clear immune response against the peptide and cross-reactivity against α-synuclein targeted epitope. Booster doses stabilized the antibody titers. Significant increase in antibody titers against PD01A was seen over time, which was translated into a humoral immune response against α-synuclein. In addition, PD01A antibodies also were reported in cerebrospinal fluid.⁴⁰

AFFiRiS presented results of a phase 1 randomized, placebo-controlled trial in 2017, confirming the safety of PD03A in patients with PD. The study showed a clear dose-dependent immune response against the peptide and cross-reactivity against α-synuclein targeted epitope.⁴¹ AFFiRiS recently presented results of another phase 1 clinical study assessing the safety and tolerability of vaccines PD01A and PD03A in patients with early MSA. Both vaccines were well tolerated, and PD01A induced an immune response against the peptide and α-synuclein epitope.⁴² These results have provided hope for further endeavors to develop active immunization strategies for PD.

Passive Immunization

Passive immunization against α-synuclein was first reported by Masliah and colleagues in 2011. A monoclonal antibody against the C-terminus of α-synuclein, 9E4, was injected into a transgenic mouse model of PD. There was reduction in α-synuclein aggregates in the brain along with improvement in motor and cognitive impairment.⁴³ The C-terminus of α-synuclein plays a key role in the pathogenesis of PD. Changes in the C-terminus of α-synuclein induces formation of α-synuclein oligomers and subsequent neuronal spread. Antibody binds to the C-terminus and prevents structural changes that can lead to oligomerization of α-synuclein. Since the first study by Masliah, few other immunization studies utilized different antibodies against the C-terminus of α-synuclein. It was shown in a mouse model that binding of such antibodies promoted clearance of the α-synuclein by microglia.⁴⁴

Based on these animal studies, Prothena Biosciences (South San Francisco, CA) designed a phase 1, double-blind, randomized, placebo-controlled clinical trial of prasinezumab (investigational monoclonal antibody against C-terminus of α-synuclein), in subjects without PD. The results showed that it was well tolerated, and there was dose-dependent reduction in the levels of free α-synuclein in plasma.⁴⁵ A 6-month phase 1b trial to evaluate the safety, tolerability and immune system response to multiple ascending doses of prasinezumab via IV infusion once every 28 days was conducted in 64 patients with PD. The drug was found to be safe, and levels of free serum α-synuclein were reduced up to 97%.⁴⁶ Roche (Basel, Switzerland) and Prothena are conducting a multicenter, randomized, double-blind phase 2 trial in patients with early PD to evaluate the efficacy of prasinezumab vs placebo.⁴⁷

BIIB054 is another monoclonal antibody that targets the N-terminal of α-synuclein. In animal models, antibodies targeting the N-terminus reduced α-synuclein triggered cell death and reduced the number of activated microglia.⁴⁸ BIIB054, from Biogen (Cambridge, MA), was studied in 40 healthy subjects and was well tolerated with a favorable safety profile and could cross the blood-brain barrier. Like the prasinezumab study, this also was an ascending-dose study to assess safety and tolerability. In 2018, a randomized, double-blind, placebo-controlled, single-ascending dose study in patients with PD reported that BIIB054 was well tolerated, and the presence of BIIB054-synuclein complexes in the plasma were confirmed.⁴⁹ A phase 2, multicenter, randomized, double-blind, placebo-controlled study (SPARK) with an active-treatment dose-blinded period, designed to evaluate the safety, pharmacokinetics, and the pharmacodynamics of BIIB054 is currently recruiting patients with PD.

Finally, BioArctic (Stockholm, Sweden) developed antibodies that are selective for oligomeric forms of α-synuclein, which it licensed to AbbVie (North Chicago, Il).⁵⁰ These antibodies do not target the N- or C-terminus of α-synuclein. Since α-synuclein oligomers play an important role in the pathogenesis of PD, targeting them with antibodies at an early stage may prove to be an effective strategy for removal of pathogenic α-synuclein. Clinical trials are forthcoming.

Conclusions

Immunotherapy against α-synuclein has provided a new therapeutic avenue in the field of neuroprotection. Results from the first human clinical trial are promising, but despite these results, more work is needed to clarify the role of α-synuclein in the pathogenesis of PD in humans. Most of the work concerning α-synuclein aggregation and propagation has been reported in animal models. Whether similar process exists in humans is a debatable question. Similarly, more knowledge is needed about how and where in the human brain antibodies act to give neuroprotective effects. Timing of administration of immunotherapies in real time will be a crucial question.

PD is clinically evident once 80% of dopaminergic neurons in substantia nigra are lost due to neurodegeneration. Should immunotherapy be administered to symptomatic patients with PD, or if it will be beneficial only for presymptomatic, high-risk patients needs to be determined. Like AD trials, not only careful selection of patients, but determination of optimal timing for treatment will be essential. As the understanding of PD pathogenesis and therapeutics evolves, it will become clear whether immunization targeting α-synuclein will modify disease progression.

Parkinson disease (PD) is a progressive neurodegenerative disorder, characterized by diverse clinical symptoms. PD can present with rest tremor, bradykinesia, rigidity, falls, postural instability, and multiple nonmotor symptoms. Marras and colleagues estimated in a comprehensive meta-analysis that there were 680,000 individuals with PD in the US in 2010; this number is expected to double by 2030 based on the US Census Bureau population projections.¹ An estimated 110,000 veterans may be affected by PD; hence, understanding of PD pathology, clinical progression, and effective treatment strategies is of paramount importance to the Veterans Health Administration (VHA).²

The exact pathogenesis underlying clinical features is still being studied. Pathologic diagnosis of PD relies on loss of dopamine neurons in the substantia nigra and accumulation of the abnormal protein, α-synuclein, in the form of Lewy bodies and Lewy neurites. Lewy bodies and neurites accumulate predominantly in the substantia nigra in addition to other brain stem nuclei and cerebral cortex. Lewy bodies are intraneuronal inclusions with a hyaline core and a pale peripheral halo. Central core stains positive for α-synuclein.^3,4 Lewy neurites are widespread and are believed to play a larger role in the pathogenesis of PD compared with those of Lewy bodies.⁵

α-Synuclein

α-synuclein is a small 140 amino-acid protein with a N-terminal region that can interact with cell membranes and a highly acidic unstructured C-terminal region.⁶ α-synuclein is physiologically present in the presynaptic terminals of neurons and involved in synaptic plasticity and vesicle trafficking.⁷ There are different hypotheses about the native structure of α-synuclein. The first suggests that it exists in tetrameric form and may be broken down to monomer, which is the pathogenic form of α-synuclein. The second hypothesis suggests that it exists primarily in monomeric form, whereas other studies have shown that both forms exist and with pathologic changes, monomer accumulates in abundance and is neurotoxic.^8-11 Work by Burré and colleagues shows that native α-synuclein exists in 2 forms: a soluble, cytosolic α-synuclein, which is monomeric, and a membrane-bound multimeric form.^12,13

Alteration in aggregation properties of this protein is believed to play a central role in the pathogenesis of PD.^14,15 Pathologic α-synuclein exists in insoluble forms that can aggregate into oligomers and fibrillar structures.¹⁶ Lysosomal dysfunction may promote accumulation of insoluble α-synuclein. Prior work has shown that several degradation pathways in lysosomes, including the ubiquitin-proteasome system and autophagy-lysosomal pathway, are down regulated, thus contributing to the accumulation of abnormal α-synuclein.^17,18 Accumulation of pathologic α-synuclein leads to mitochondrial dysfunction in PD animal models, contributing further to neurotoxicity.^19,20 Aggregates of phosphorylated α-synuclein have been demonstrated in dementia with Lewy body.²¹

In addition, α-synuclein aggregates may be released into extracellular spaces to be taken up by adjacent cells, where they can cause further misfolding and aggregation of protein.²² Previous work in animal models suggested a prion proteinlike spread of α-synuclein.²³ This finding can have long-term therapeutic implications, as preventing extracellular release of abnormal form of α-synuclein will prevent the spread of pathologic protein. This can form the basis of neuroprotection in patients with PD.²⁴

It has been proposed that α-synuclein accumulation and extracellular release initiates an immune response that leads to activation of microglia. This has been shown in PD animal models, overexpressing α-synuclein. In 2008 Park and colleagues demonstrated that microglial activation is enhanced by monomeric α-synuclein, not by the aggregated variant.²⁵ Other studies have reported activated microglia around dopaminergic cells in substantia nigra.²⁶ Sulzer and colleagues showed that peptides from α-synuclein can act as antigens and trigger an autoimmune reaction via T cells.²⁷ PD may be associated with certain HLA-haplotypes.²⁸ In other words, α-synuclein can induce neurodegeneration via different mechanisms, including alteration in synaptic vesicle transmission, mitochondrial dysfunction, neuroinflammation, and induction of humoral immunity.

Immunization

Due to these observations, there had been huge interest in developing antibody-based therapies for PD. A similar approach had been tested in Alzheimer disease (AD). Intracellular tangles of tau protein and extracellular aggregates of amyloid are the pathologic substrates in AD. Clinical trials utilizing antibodies targeting amyloid showed reduction in abnormal protein accumulation but no significant improvement in cognition.²⁹ In addition, adverse events (AEs), such as vasogenic edema and intracerebral hemorrhage, were reported.³⁰ Careful analysis of the data suggested that inadequate patient selection or targeting only amyloid, may have contributed to unfavorable results.³¹ Since then, more recent clinical trials have focused on careful patient selection, use of second generation anti-amyloid antibodies and immunotherapies targeting tau.³²

Several studies have tested immunotherapies in PD animal models with the aim of targeting α-synuclein. Immunotherapies can be instituted in 2 ways: active immunization in which the immune system is stimulated to produce antibodies against α-synuclein or passive immunization in which antibodies against α-synuclein are administered directly. Once α-synuclein antibodies have crossed the blood-brain barrier, they are hypothesized to clear the existing α-synuclein. Animal studies have demonstrated the presence of these antibodies within the neurons. The mechanism of entry is unknown. Once inside the cells, the antibodies activate the lysosomal clearance, affecting intracellular accumulation of α-synuclein. Extracellularly, they can bind to receptors on scavenger cells, mainly microglia, activating them to facilitate uptake of extracellular α-synuclein. Binding of the antibodies to α-synuclein directly prevents the uptake of toxic protein by the cells, blocking the transfer and spread of PD pathology.³³

Active Immunization

Active immunization against α-synuclein was demonstrated by Masliah and colleagues almost a decade ago. They administered recombinant human α-synuclein in transgenic mice expressing α-synuclein under the control of platelet-derived growth factor β. Reduction of accumulated α-synuclein in neurons with mild microglia activation was noted. It was proposed that the antibodies produced were able to bind to abnormal α-synuclein, were recognized by the lysosomal pathways, and degraded.³⁴ Ghochikyan and colleagues developed vaccines by using α-synuclein-derived peptides. This induced formation of antibodies against α-synuclein in Lewy-bodies and neurites.³⁵ Over time, other animal studies have been able to expand on these results.³⁶

AFFiRiS, an Austrian biotechnology company, has developed 2 peptide vaccines PD01A and PD03A. Both peptides when administered to PD animal models caused antibody-based immune response against aggregated α-synuclein. Humoral autoimmune response was not observed in these studies; no neuroinflammation or neurotoxicity was noted. These peptides did not affect levels of physiologic α-synuclein, targeting only the aggregated form.³⁷ These animal models showed improved motor and cognitive function. Similar results were noted in multiple system atrophy (MSA) animal models.^38,39

The first human phase 1, randomized, parallel-group, single-center study recruited 32 subjects with early PD. Twelve subjects each were included in low- or high-dose treatment group, and 8 were included in the control group. Test subjects randomly received 4 vaccinations of low- or high-dose PD01A. Both doses were well tolerated, and no drug-related serious AEs were reported. The study confirmed the tolerability and safety of subcutaneous PD01A vaccine administration. These subjects were included in a 12-month, phase 1b follow-up extension study, AFF008E. In 2018, it was reported that administration of 6 doses of PD01A, 4 primary and 2 booster immunization, was safe. The vaccine showed a clear immune response against the peptide and cross-reactivity against α-synuclein targeted epitope. Booster doses stabilized the antibody titers. Significant increase in antibody titers against PD01A was seen over time, which was translated into a humoral immune response against α-synuclein. In addition, PD01A antibodies also were reported in cerebrospinal fluid.⁴⁰

AFFiRiS presented results of a phase 1 randomized, placebo-controlled trial in 2017, confirming the safety of PD03A in patients with PD. The study showed a clear dose-dependent immune response against the peptide and cross-reactivity against α-synuclein targeted epitope.⁴¹ AFFiRiS recently presented results of another phase 1 clinical study assessing the safety and tolerability of vaccines PD01A and PD03A in patients with early MSA. Both vaccines were well tolerated, and PD01A induced an immune response against the peptide and α-synuclein epitope.⁴² These results have provided hope for further endeavors to develop active immunization strategies for PD.

Passive Immunization

Passive immunization against α-synuclein was first reported by Masliah and colleagues in 2011. A monoclonal antibody against the C-terminus of α-synuclein, 9E4, was injected into a transgenic mouse model of PD. There was reduction in α-synuclein aggregates in the brain along with improvement in motor and cognitive impairment.⁴³ The C-terminus of α-synuclein plays a key role in the pathogenesis of PD. Changes in the C-terminus of α-synuclein induces formation of α-synuclein oligomers and subsequent neuronal spread. Antibody binds to the C-terminus and prevents structural changes that can lead to oligomerization of α-synuclein. Since the first study by Masliah, few other immunization studies utilized different antibodies against the C-terminus of α-synuclein. It was shown in a mouse model that binding of such antibodies promoted clearance of the α-synuclein by microglia.⁴⁴

Based on these animal studies, Prothena Biosciences (South San Francisco, CA) designed a phase 1, double-blind, randomized, placebo-controlled clinical trial of prasinezumab (investigational monoclonal antibody against C-terminus of α-synuclein), in subjects without PD. The results showed that it was well tolerated, and there was dose-dependent reduction in the levels of free α-synuclein in plasma.⁴⁵ A 6-month phase 1b trial to evaluate the safety, tolerability and immune system response to multiple ascending doses of prasinezumab via IV infusion once every 28 days was conducted in 64 patients with PD. The drug was found to be safe, and levels of free serum α-synuclein were reduced up to 97%.⁴⁶ Roche (Basel, Switzerland) and Prothena are conducting a multicenter, randomized, double-blind phase 2 trial in patients with early PD to evaluate the efficacy of prasinezumab vs placebo.⁴⁷

BIIB054 is another monoclonal antibody that targets the N-terminal of α-synuclein. In animal models, antibodies targeting the N-terminus reduced α-synuclein triggered cell death and reduced the number of activated microglia.⁴⁸ BIIB054, from Biogen (Cambridge, MA), was studied in 40 healthy subjects and was well tolerated with a favorable safety profile and could cross the blood-brain barrier. Like the prasinezumab study, this also was an ascending-dose study to assess safety and tolerability. In 2018, a randomized, double-blind, placebo-controlled, single-ascending dose study in patients with PD reported that BIIB054 was well tolerated, and the presence of BIIB054-synuclein complexes in the plasma were confirmed.⁴⁹ A phase 2, multicenter, randomized, double-blind, placebo-controlled study (SPARK) with an active-treatment dose-blinded period, designed to evaluate the safety, pharmacokinetics, and the pharmacodynamics of BIIB054 is currently recruiting patients with PD.

Finally, BioArctic (Stockholm, Sweden) developed antibodies that are selective for oligomeric forms of α-synuclein, which it licensed to AbbVie (North Chicago, Il).⁵⁰ These antibodies do not target the N- or C-terminus of α-synuclein. Since α-synuclein oligomers play an important role in the pathogenesis of PD, targeting them with antibodies at an early stage may prove to be an effective strategy for removal of pathogenic α-synuclein. Clinical trials are forthcoming.

Conclusions

Immunotherapy against α-synuclein has provided a new therapeutic avenue in the field of neuroprotection. Results from the first human clinical trial are promising, but despite these results, more work is needed to clarify the role of α-synuclein in the pathogenesis of PD in humans. Most of the work concerning α-synuclein aggregation and propagation has been reported in animal models. Whether similar process exists in humans is a debatable question. Similarly, more knowledge is needed about how and where in the human brain antibodies act to give neuroprotective effects. Timing of administration of immunotherapies in real time will be a crucial question.

PD is clinically evident once 80% of dopaminergic neurons in substantia nigra are lost due to neurodegeneration. Should immunotherapy be administered to symptomatic patients with PD, or if it will be beneficial only for presymptomatic, high-risk patients needs to be determined. Like AD trials, not only careful selection of patients, but determination of optimal timing for treatment will be essential. As the understanding of PD pathogenesis and therapeutics evolves, it will become clear whether immunization targeting α-synuclein will modify disease progression.

The development of biologic therapies has led to some of the most significant advances in the treatment of cancer, but these drugs are also very expensive. As patents for the biologics begin to expire, the development of biosimilars has the potential to dramatically cut therapy costs thereby making the therapies more readily accessible to patients. Here, we discuss biosimilar development and the challenges that need to be overcome to create a robust market.

Biosimilar, not generic

Biologic therapies are derived from living organisms and include the targeted monoclonal antibodies (mAbs) and cell-based therapies that have revolutionized the treatment of certain cancer types. Yet, their greater complexity makes them more difficult to manufacture, store, and administer, making them a costly therapeutic option that ultimately drives up health care costs. According to a 2011 drug expenditure analysis, biologic therapies accounted for more than half of the total expenditure on anticancer drugs in the US health care system.^1,2

Generally, when drug patents expire, other companies can develop their own identical generic versions to increase competition in the marketplace and drive down costs. However, the paradigm for generic development cannot be applied to biologic therapies because the way in which they are manufactured makes it impossible to generate an identical copy.

Instead, the Biologics Price Competition and Innovation Act, a provision of the Patient Protection and Affordable Care Act, has allowed for submission of an application for “licensure of a biologic product based on its similarity to a licensed biologic product”.³

These “biosimilars” have been positioned as game-changers in oncology, with the potential to reduce costs and improve access to biologic therapies. With the patents on several blockbuster cancer biologics already expired or due to expire by 2020, an increasing number of biosimilars are being developed.⁴

Totality of evidence

Biosimilars require more rigorous testing than generics, but they don’t require the same type of scientific data that the original biologic products, termed “reference products,” did. Therefore, they are governed by legislation unique to them and approved by different regulatory pathways. The US Food and Drug Administration (FDA) has established a unique shortened regulatory pathway for their approval, known as the 351(k) pathway. So whereas the pathway for reference products is geared toward demonstrating patient benefit, biosimilars are required instead to show equivalence to the reference product.⁵

Biosimilars are produced through reverse engineering the reference product. Then, through a stepwise process, to generate what the FDA calls a “totality of evidence,” biosimilar manufacturers must demonstrate structural and functional similarities (through comparative quality studies) and comparable pharmacokinetics and pharmacodynamics (through comparative nonclinical and clinical studies) to the reference product. Final approval is based on 1 or more comparative clinical studies performed in the most sensitive patient population(s) (Figure 1).⁶

618_de Lartigue bios_F1_web.PNG

A flurry of approvals

The first biosimilar approvals in oncology in the United States came in the supportive care niche (Table 1). Filgrastim-sndz (Zarxio), approved in March 2015, is a biosimilar of the granulocyte-macrophage colony stimulating factor (G-CSF) analog filgrastim (Neupogen). Owing to its mechanism of action in stimulating the production of neutrophils in the bone marrow, filgrastim is used to help reduce the risk or severity of neutropenia in patients undergoing myelosuppressive chemotherapy regimens.

Filgrastim-sndz was approved for use across all 5 indications for which the reference product is approved, based on the totality of evidence, which included results from the key phase 3 PIONEER study.⁹ Market entry was initially delayed by lawsuits filed by Amgen, the maker of the reference product, but the biosimilar was subsequently cleared by the US Court of Appeals for the Federal Circuit. The wholesale acquisition cost (WAC) for a 300µg syringe is $324.30 for filgrastim and $275.66 figrastim-sndz, representing a 15% reduction on the reference product.¹⁰

618_de Lartigue bios_T1_web.PNG

9,19-22

618_de Lartigue bios_T2_web.PNG

Biosimilars in development

While numerous other biosimilars of filgrastim and pegfilgrastim are in development, the major focus has been on the development of more biosimilars to treat cancer (Table 3). BLAs have been submitted for 4 biosimilars of trastuzumab and 1 bevacizumab biosimilar. Approval for several of the trastuzumab biosimilars has been delayed by CRLs from the FDA, mostly regarding issues with the manufacturing process or facility. Several other trastuzumab and bevacizumab biosimilars are in late-stage clinical trials.

618_de Lartigue bios_T3_web.PNG

The results of several phase 3 comparative clinical trials were recently published or reported at annual conferences. Pfizer’s PF-05280014 was compared with the European Union (EU)–approved trastuzumab, both in combination with paclitaxel, in patients with previously untreated HER2-positive metastatic breast cancer. Data reported at the European Society for Medical Oncology congress in 2017 demonstrated equivalence between the reference product and biosimilar in overall response rate (ORR).²³

Another recently published trial compared this biosimilar to EU-trastuzumab, both in combination with carboplatin and docetaxel, as neoadjuvant treatment for patients with resectable HER2-positive breast cancer. Among 226 patients randomized to receive 8 mg/kg in cycle 1 and 6 mg/kg thereafter of the biosimilar or reference product, every 3 weeks for 6 cycles, the pathologic complete response (pCR) rates were 47% and 50%, respectively.²⁴

The results of a phase 3 study comparing Samsung Bioepis/Merck’s joint offering SB3 were recently published. A total of 875 patients were randomized 1:1 to receive SB3 or reference trastuzumab in combination with chemotherapy (4 cycles docetaxel followed by 4 cycles 5-fluorouracil/epirubicin/cyclophosphamide) prior to surgery, followed by 10 cycles of adjuvant SB3 or trastuzumab reference. Rates of event-free survival (EFS) were comparable between the 2 groups at 12 months (93.7% vs 96.1%, respectively).²⁵

Amgen’s ABP980 was evaluated in the phase 3 LILAC trial, which measured the effect of the biosimilar on pCR in women with HER2-positive early breast cancer compared with reference trastuzumab. After 4 cycles of run-in anthracycline-based chemotherapy, ABP980 or reference trastuzumab were administered in combination with paclitaxel. This was followed by surgery and then ABP980 or reference trastuzumab in the adjuvant setting for up to 1 year, with the option to continue on the same drug as the neoadjuvant setting or to switch to the other. Among 696 assessable patients, the pCR rates were 48% and 42%, respectively.²⁶

Most advanced in clinical testing among the upcoming bevacizumab biosimilars is Pfizer’s PF-06439535, for which the results of a phase 3 comparative trial were presented at the 2018 annual meeting of the American Society for Clinical Oncology. PF-06439535 was compared with the EU-approved bevacizumab, both in combination with paclitaxel and carboplatin, as first-line therapy for patients with advanced non-squamous NSCLC. Among 719 patients, the primary endpoint of ORR was 45.3% and 44.6%, respectively.²⁷

Biosimilars of a third blockbuster cancer drug, the CD20-targeting mAb rituximab (Rituxan) are also in development and FDA approval is pending for 2. The patent for Rituxan expired in 2016, so these drugs could hit the market as soon as they are approved.

In a race to the finish for the first US-approved rituximab biosimilar, Celltrion-Teva’s CT-P10 (Truxima) seems most likely to come first; the Oncologic Drugs Advisory Committee voted unanimously in October 2018 to recommend its approval. Phase 3 comparative data were recently published; patients with newly diagnosed advanced-stage follicular lymphoma were randomized to receive intravenous infusions of 375 mg/m² CT-P10 or reference rituximab, both in combination with cyclophosphamide, vincristine, and prednisone, on day 1 of 8 21-day cycles. The ORRs were identical (92.6%) for both drugs, pharmacokinetics data also suggested bioequivalence, and the incidence of AEs was also comparable (83% vs 80%).²⁸

Biosimilars of the epidermal growth factor receptor (EGFR)-targeting mAb cetuximab are also listed in the pipeline for several biosimilar developers, but there is no indication of their developmental status as yet and no clinical trials are ongoing in the US.

Sorrento is developing STI-001, a cetuximab biosimilar, and reported that a phase 3 trial had been completed. Instead of a comparison with the reference product, however, the trial compared STI-001 in combination with irinotecan with irinotecan alone. They reported significantly higher ORR, PFS, and OS with the biosimilar compared with irinotecan alone, and a significant increase over historical data with the reference product, as well as fewer side effects and immunogenicity, which they attribute to its manufacture in a different cell line. However, no data has been published and no trials are ongoing in the United States, so the status of its development remains unclear.²⁹
 

Challenges to a robust market

It is an exciting time for biosimilars, with many approvals and drugs being brought to market in the US in the past several years and more poised to follow suit as patents expire. Yet many challenges remain around the growth of a robust biosimilars market.

Several surveys conducted in recent years have demonstrated suboptimal knowledge of all aspects of biosimilars and highlighted the need for evidence-based education across specialties.^30,31 In response, the FDA recently announced that it was launching an educational campaign to further understanding of biosimilars, including naming conventions (Figure 2).^32,33 Numerous other medical professional societies have produced or are in the process of producing biosimilar guidelines.

618_de Lartigue bios_F2_web.PNG

The development of biologic therapies has led to some of the most significant advances in the treatment of cancer, but these drugs are also very expensive. As patents for the biologics begin to expire, the development of biosimilars has the potential to dramatically cut therapy costs thereby making the therapies more readily accessible to patients. Here, we discuss biosimilar development and the challenges that need to be overcome to create a robust market.

Biosimilar, not generic

Biologic therapies are derived from living organisms and include the targeted monoclonal antibodies (mAbs) and cell-based therapies that have revolutionized the treatment of certain cancer types. Yet, their greater complexity makes them more difficult to manufacture, store, and administer, making them a costly therapeutic option that ultimately drives up health care costs. According to a 2011 drug expenditure analysis, biologic therapies accounted for more than half of the total expenditure on anticancer drugs in the US health care system.^1,2

Generally, when drug patents expire, other companies can develop their own identical generic versions to increase competition in the marketplace and drive down costs. However, the paradigm for generic development cannot be applied to biologic therapies because the way in which they are manufactured makes it impossible to generate an identical copy.

Instead, the Biologics Price Competition and Innovation Act, a provision of the Patient Protection and Affordable Care Act, has allowed for submission of an application for “licensure of a biologic product based on its similarity to a licensed biologic product”.³

These “biosimilars” have been positioned as game-changers in oncology, with the potential to reduce costs and improve access to biologic therapies. With the patents on several blockbuster cancer biologics already expired or due to expire by 2020, an increasing number of biosimilars are being developed.⁴

Totality of evidence

Biosimilars require more rigorous testing than generics, but they don’t require the same type of scientific data that the original biologic products, termed “reference products,” did. Therefore, they are governed by legislation unique to them and approved by different regulatory pathways. The US Food and Drug Administration (FDA) has established a unique shortened regulatory pathway for their approval, known as the 351(k) pathway. So whereas the pathway for reference products is geared toward demonstrating patient benefit, biosimilars are required instead to show equivalence to the reference product.⁵

Biosimilars are produced through reverse engineering the reference product. Then, through a stepwise process, to generate what the FDA calls a “totality of evidence,” biosimilar manufacturers must demonstrate structural and functional similarities (through comparative quality studies) and comparable pharmacokinetics and pharmacodynamics (through comparative nonclinical and clinical studies) to the reference product. Final approval is based on 1 or more comparative clinical studies performed in the most sensitive patient population(s) (Figure 1).⁶

618_de Lartigue bios_F1_web.PNG

A flurry of approvals

The first biosimilar approvals in oncology in the United States came in the supportive care niche (Table 1). Filgrastim-sndz (Zarxio), approved in March 2015, is a biosimilar of the granulocyte-macrophage colony stimulating factor (G-CSF) analog filgrastim (Neupogen). Owing to its mechanism of action in stimulating the production of neutrophils in the bone marrow, filgrastim is used to help reduce the risk or severity of neutropenia in patients undergoing myelosuppressive chemotherapy regimens.

Filgrastim-sndz was approved for use across all 5 indications for which the reference product is approved, based on the totality of evidence, which included results from the key phase 3 PIONEER study.⁹ Market entry was initially delayed by lawsuits filed by Amgen, the maker of the reference product, but the biosimilar was subsequently cleared by the US Court of Appeals for the Federal Circuit. The wholesale acquisition cost (WAC) for a 300µg syringe is $324.30 for filgrastim and $275.66 figrastim-sndz, representing a 15% reduction on the reference product.¹⁰

618_de Lartigue bios_T1_web.PNG

9,19-22

618_de Lartigue bios_T2_web.PNG

Biosimilars in development

While numerous other biosimilars of filgrastim and pegfilgrastim are in development, the major focus has been on the development of more biosimilars to treat cancer (Table 3). BLAs have been submitted for 4 biosimilars of trastuzumab and 1 bevacizumab biosimilar. Approval for several of the trastuzumab biosimilars has been delayed by CRLs from the FDA, mostly regarding issues with the manufacturing process or facility. Several other trastuzumab and bevacizumab biosimilars are in late-stage clinical trials.

618_de Lartigue bios_T3_web.PNG

The results of several phase 3 comparative clinical trials were recently published or reported at annual conferences. Pfizer’s PF-05280014 was compared with the European Union (EU)–approved trastuzumab, both in combination with paclitaxel, in patients with previously untreated HER2-positive metastatic breast cancer. Data reported at the European Society for Medical Oncology congress in 2017 demonstrated equivalence between the reference product and biosimilar in overall response rate (ORR).²³

Another recently published trial compared this biosimilar to EU-trastuzumab, both in combination with carboplatin and docetaxel, as neoadjuvant treatment for patients with resectable HER2-positive breast cancer. Among 226 patients randomized to receive 8 mg/kg in cycle 1 and 6 mg/kg thereafter of the biosimilar or reference product, every 3 weeks for 6 cycles, the pathologic complete response (pCR) rates were 47% and 50%, respectively.²⁴

The results of a phase 3 study comparing Samsung Bioepis/Merck’s joint offering SB3 were recently published. A total of 875 patients were randomized 1:1 to receive SB3 or reference trastuzumab in combination with chemotherapy (4 cycles docetaxel followed by 4 cycles 5-fluorouracil/epirubicin/cyclophosphamide) prior to surgery, followed by 10 cycles of adjuvant SB3 or trastuzumab reference. Rates of event-free survival (EFS) were comparable between the 2 groups at 12 months (93.7% vs 96.1%, respectively).²⁵

Amgen’s ABP980 was evaluated in the phase 3 LILAC trial, which measured the effect of the biosimilar on pCR in women with HER2-positive early breast cancer compared with reference trastuzumab. After 4 cycles of run-in anthracycline-based chemotherapy, ABP980 or reference trastuzumab were administered in combination with paclitaxel. This was followed by surgery and then ABP980 or reference trastuzumab in the adjuvant setting for up to 1 year, with the option to continue on the same drug as the neoadjuvant setting or to switch to the other. Among 696 assessable patients, the pCR rates were 48% and 42%, respectively.²⁶

Most advanced in clinical testing among the upcoming bevacizumab biosimilars is Pfizer’s PF-06439535, for which the results of a phase 3 comparative trial were presented at the 2018 annual meeting of the American Society for Clinical Oncology. PF-06439535 was compared with the EU-approved bevacizumab, both in combination with paclitaxel and carboplatin, as first-line therapy for patients with advanced non-squamous NSCLC. Among 719 patients, the primary endpoint of ORR was 45.3% and 44.6%, respectively.²⁷

Biosimilars of a third blockbuster cancer drug, the CD20-targeting mAb rituximab (Rituxan) are also in development and FDA approval is pending for 2. The patent for Rituxan expired in 2016, so these drugs could hit the market as soon as they are approved.

In a race to the finish for the first US-approved rituximab biosimilar, Celltrion-Teva’s CT-P10 (Truxima) seems most likely to come first; the Oncologic Drugs Advisory Committee voted unanimously in October 2018 to recommend its approval. Phase 3 comparative data were recently published; patients with newly diagnosed advanced-stage follicular lymphoma were randomized to receive intravenous infusions of 375 mg/m² CT-P10 or reference rituximab, both in combination with cyclophosphamide, vincristine, and prednisone, on day 1 of 8 21-day cycles. The ORRs were identical (92.6%) for both drugs, pharmacokinetics data also suggested bioequivalence, and the incidence of AEs was also comparable (83% vs 80%).²⁸

Biosimilars of the epidermal growth factor receptor (EGFR)-targeting mAb cetuximab are also listed in the pipeline for several biosimilar developers, but there is no indication of their developmental status as yet and no clinical trials are ongoing in the US.

Sorrento is developing STI-001, a cetuximab biosimilar, and reported that a phase 3 trial had been completed. Instead of a comparison with the reference product, however, the trial compared STI-001 in combination with irinotecan with irinotecan alone. They reported significantly higher ORR, PFS, and OS with the biosimilar compared with irinotecan alone, and a significant increase over historical data with the reference product, as well as fewer side effects and immunogenicity, which they attribute to its manufacture in a different cell line. However, no data has been published and no trials are ongoing in the United States, so the status of its development remains unclear.²⁹
 

Challenges to a robust market

It is an exciting time for biosimilars, with many approvals and drugs being brought to market in the US in the past several years and more poised to follow suit as patents expire. Yet many challenges remain around the growth of a robust biosimilars market.

Several surveys conducted in recent years have demonstrated suboptimal knowledge of all aspects of biosimilars and highlighted the need for evidence-based education across specialties.^30,31 In response, the FDA recently announced that it was launching an educational campaign to further understanding of biosimilars, including naming conventions (Figure 2).^32,33 Numerous other medical professional societies have produced or are in the process of producing biosimilar guidelines.

618_de Lartigue bios_F2_web.PNG

The development of biologic therapies has led to some of the most significant advances in the treatment of cancer, but these drugs are also very expensive. As patents for the biologics begin to expire, the development of biosimilars has the potential to dramatically cut therapy costs thereby making the therapies more readily accessible to patients. Here, we discuss biosimilar development and the challenges that need to be overcome to create a robust market.

Biosimilar, not generic

Biologic therapies are derived from living organisms and include the targeted monoclonal antibodies (mAbs) and cell-based therapies that have revolutionized the treatment of certain cancer types. Yet, their greater complexity makes them more difficult to manufacture, store, and administer, making them a costly therapeutic option that ultimately drives up health care costs. According to a 2011 drug expenditure analysis, biologic therapies accounted for more than half of the total expenditure on anticancer drugs in the US health care system.^1,2

Generally, when drug patents expire, other companies can develop their own identical generic versions to increase competition in the marketplace and drive down costs. However, the paradigm for generic development cannot be applied to biologic therapies because the way in which they are manufactured makes it impossible to generate an identical copy.

Instead, the Biologics Price Competition and Innovation Act, a provision of the Patient Protection and Affordable Care Act, has allowed for submission of an application for “licensure of a biologic product based on its similarity to a licensed biologic product”.³

These “biosimilars” have been positioned as game-changers in oncology, with the potential to reduce costs and improve access to biologic therapies. With the patents on several blockbuster cancer biologics already expired or due to expire by 2020, an increasing number of biosimilars are being developed.⁴

Totality of evidence

Biosimilars require more rigorous testing than generics, but they don’t require the same type of scientific data that the original biologic products, termed “reference products,” did. Therefore, they are governed by legislation unique to them and approved by different regulatory pathways. The US Food and Drug Administration (FDA) has established a unique shortened regulatory pathway for their approval, known as the 351(k) pathway. So whereas the pathway for reference products is geared toward demonstrating patient benefit, biosimilars are required instead to show equivalence to the reference product.⁵

Biosimilars are produced through reverse engineering the reference product. Then, through a stepwise process, to generate what the FDA calls a “totality of evidence,” biosimilar manufacturers must demonstrate structural and functional similarities (through comparative quality studies) and comparable pharmacokinetics and pharmacodynamics (through comparative nonclinical and clinical studies) to the reference product. Final approval is based on 1 or more comparative clinical studies performed in the most sensitive patient population(s) (Figure 1).⁶

618_de Lartigue bios_F1_web.PNG

A flurry of approvals

The first biosimilar approvals in oncology in the United States came in the supportive care niche (Table 1). Filgrastim-sndz (Zarxio), approved in March 2015, is a biosimilar of the granulocyte-macrophage colony stimulating factor (G-CSF) analog filgrastim (Neupogen). Owing to its mechanism of action in stimulating the production of neutrophils in the bone marrow, filgrastim is used to help reduce the risk or severity of neutropenia in patients undergoing myelosuppressive chemotherapy regimens.

Filgrastim-sndz was approved for use across all 5 indications for which the reference product is approved, based on the totality of evidence, which included results from the key phase 3 PIONEER study.⁹ Market entry was initially delayed by lawsuits filed by Amgen, the maker of the reference product, but the biosimilar was subsequently cleared by the US Court of Appeals for the Federal Circuit. The wholesale acquisition cost (WAC) for a 300µg syringe is $324.30 for filgrastim and $275.66 figrastim-sndz, representing a 15% reduction on the reference product.¹⁰

618_de Lartigue bios_T1_web.PNG

9,19-22

618_de Lartigue bios_T2_web.PNG

Biosimilars in development

While numerous other biosimilars of filgrastim and pegfilgrastim are in development, the major focus has been on the development of more biosimilars to treat cancer (Table 3). BLAs have been submitted for 4 biosimilars of trastuzumab and 1 bevacizumab biosimilar. Approval for several of the trastuzumab biosimilars has been delayed by CRLs from the FDA, mostly regarding issues with the manufacturing process or facility. Several other trastuzumab and bevacizumab biosimilars are in late-stage clinical trials.

618_de Lartigue bios_T3_web.PNG

The results of several phase 3 comparative clinical trials were recently published or reported at annual conferences. Pfizer’s PF-05280014 was compared with the European Union (EU)–approved trastuzumab, both in combination with paclitaxel, in patients with previously untreated HER2-positive metastatic breast cancer. Data reported at the European Society for Medical Oncology congress in 2017 demonstrated equivalence between the reference product and biosimilar in overall response rate (ORR).²³

Another recently published trial compared this biosimilar to EU-trastuzumab, both in combination with carboplatin and docetaxel, as neoadjuvant treatment for patients with resectable HER2-positive breast cancer. Among 226 patients randomized to receive 8 mg/kg in cycle 1 and 6 mg/kg thereafter of the biosimilar or reference product, every 3 weeks for 6 cycles, the pathologic complete response (pCR) rates were 47% and 50%, respectively.²⁴

The results of a phase 3 study comparing Samsung Bioepis/Merck’s joint offering SB3 were recently published. A total of 875 patients were randomized 1:1 to receive SB3 or reference trastuzumab in combination with chemotherapy (4 cycles docetaxel followed by 4 cycles 5-fluorouracil/epirubicin/cyclophosphamide) prior to surgery, followed by 10 cycles of adjuvant SB3 or trastuzumab reference. Rates of event-free survival (EFS) were comparable between the 2 groups at 12 months (93.7% vs 96.1%, respectively).²⁵

Amgen’s ABP980 was evaluated in the phase 3 LILAC trial, which measured the effect of the biosimilar on pCR in women with HER2-positive early breast cancer compared with reference trastuzumab. After 4 cycles of run-in anthracycline-based chemotherapy, ABP980 or reference trastuzumab were administered in combination with paclitaxel. This was followed by surgery and then ABP980 or reference trastuzumab in the adjuvant setting for up to 1 year, with the option to continue on the same drug as the neoadjuvant setting or to switch to the other. Among 696 assessable patients, the pCR rates were 48% and 42%, respectively.²⁶

Most advanced in clinical testing among the upcoming bevacizumab biosimilars is Pfizer’s PF-06439535, for which the results of a phase 3 comparative trial were presented at the 2018 annual meeting of the American Society for Clinical Oncology. PF-06439535 was compared with the EU-approved bevacizumab, both in combination with paclitaxel and carboplatin, as first-line therapy for patients with advanced non-squamous NSCLC. Among 719 patients, the primary endpoint of ORR was 45.3% and 44.6%, respectively.²⁷

Biosimilars of a third blockbuster cancer drug, the CD20-targeting mAb rituximab (Rituxan) are also in development and FDA approval is pending for 2. The patent for Rituxan expired in 2016, so these drugs could hit the market as soon as they are approved.

In a race to the finish for the first US-approved rituximab biosimilar, Celltrion-Teva’s CT-P10 (Truxima) seems most likely to come first; the Oncologic Drugs Advisory Committee voted unanimously in October 2018 to recommend its approval. Phase 3 comparative data were recently published; patients with newly diagnosed advanced-stage follicular lymphoma were randomized to receive intravenous infusions of 375 mg/m² CT-P10 or reference rituximab, both in combination with cyclophosphamide, vincristine, and prednisone, on day 1 of 8 21-day cycles. The ORRs were identical (92.6%) for both drugs, pharmacokinetics data also suggested bioequivalence, and the incidence of AEs was also comparable (83% vs 80%).²⁸

Biosimilars of the epidermal growth factor receptor (EGFR)-targeting mAb cetuximab are also listed in the pipeline for several biosimilar developers, but there is no indication of their developmental status as yet and no clinical trials are ongoing in the US.

Sorrento is developing STI-001, a cetuximab biosimilar, and reported that a phase 3 trial had been completed. Instead of a comparison with the reference product, however, the trial compared STI-001 in combination with irinotecan with irinotecan alone. They reported significantly higher ORR, PFS, and OS with the biosimilar compared with irinotecan alone, and a significant increase over historical data with the reference product, as well as fewer side effects and immunogenicity, which they attribute to its manufacture in a different cell line. However, no data has been published and no trials are ongoing in the United States, so the status of its development remains unclear.²⁹
 

Challenges to a robust market

It is an exciting time for biosimilars, with many approvals and drugs being brought to market in the US in the past several years and more poised to follow suit as patents expire. Yet many challenges remain around the growth of a robust biosimilars market.

Several surveys conducted in recent years have demonstrated suboptimal knowledge of all aspects of biosimilars and highlighted the need for evidence-based education across specialties.^30,31 In response, the FDA recently announced that it was launching an educational campaign to further understanding of biosimilars, including naming conventions (Figure 2).^32,33 Numerous other medical professional societies have produced or are in the process of producing biosimilar guidelines.

618_de Lartigue bios_F2_web.PNG

Infection with certain viruses has been causally linked to the development of cancer. In recent years, an improved understanding of the unique pathology and molecular underpinnings of these virally associated cancers has prompted the development of more personalized treatment strategies, with a particular focus on immunotherapy. Here, we describe some of the latest developments.

The link between viruses and cancer

Suspicions about a possible role of viral infections in the development of cancer were first aroused in the early 1900s. The seminal discovery is traced back to Peyton Rous, who showed that a malignant tumor growing in a chicken could be transferred to a healthy bird by injecting it with tumor extracts that contained no actual tumor cells.¹

The infectious etiology of human cancer, however, remained controversial until many years later when the first cancer-causing virus, Epstein-Barr virus (EBV), was identified in cell cultures from patients with Burkitt lymphoma. Shortly afterward, the Rous sarcoma virus was unveiled as the oncogenic agent behind Rous’ observations.²Seven viruses have now been linked to the development of cancers and are thought to be responsible for around 12% of all cancer cases worldwide. The burden is likely to increase as technological advancements make it easier to establish a causal link between viruses and cancer development.³

In addition to making these links, researchers have also made significant headway in understanding how viruses cause cancer. Cancerous transformation of host cells occurs in only a minority of those who are infected with oncogenic viruses and often occurs in the setting of chronic infection.

Viruses can mediate carcinogenesis by direct and/or indirect mechanisms (Figure 1). Many of the hallmarks of cancer, the key attributes that drive the transformation from a normal cell to a malignant one, are compatible with the virus’s needs, such as needing to avoid cell death, increasing cell proliferation, and avoiding detection by the immune system.

218_de Lartigue_NT_F_web.png

Vaccines lead the charge in HPV-driven cancers

German virologist Harald zur Hausen received the Nobel Prize in 2008 for his discovery of the oncogenic role of human papillomaviruses (HPVs), a large family of more than 100 DNA viruses that infect the epithelial cells of the skin and mucous membranes. They are responsible for the largest number of virally associated cancer cases globally – around 5% (Table 1).

218_de Lartigue_NT_T1_web.png

A number of different cancer types are linked to HPV infection, but it is best known as the cause of cervical cancer. The development of diagnostic blood tests and prophylactic vaccines for prevention and early intervention in HPV infection has helped to reduce the incidence of cervical cancer. Conversely, another type of HPV-associated cancer, head and neck squamous cell carcinoma (HNSCC), has seen increased incidence in recent years.

HPVs are categorized according to their oncogenic potential as high, intermediate, or low risk. The high-risk HPV16 and HPV18 strains are most commonly associated with cancer. They are thought to cause cancer predominantly through integration into the host genome. The HPV genome is composed of 8 genes encoding proteins that regulate viral replication and assembly. The E6 and E7 genes are the most highly oncogenic; as the HPV DNA is inserted into the host genome, the transcriptional regulator of E6/E7 is lost, leading to their increased expression. These genes have significant oncogenic potential because of their interaction with 2 tumor suppressor proteins, p53 and pRb.^6,7

The largest investment in therapeutic development for HPV-positive cancers has been in the realm of immunotherapy in an effort to boost the anti-tumor immune response. In particular, there has been a focus on the development of therapeutic vaccines, designed to prime the anti-tumor immune response to recognize viral antigens. A variety of different types of vaccines are being developed, including live, attenuated and inactivated vaccines that are protein, DNA, or peptide based. Most developed to date target the E6/E7 proteins from the HPV16/18 strains (Table 2).^8,9

218_de Lartigue_NT_T2_web.png

Hepatocellular carcinoma: a tale of two viruses

The hepatitis viruses are a group of 5 unrelated viruses that causes inflammation of the liver. Hepatitis B (HBV), a DNA virus, and hepatitis C (HCV), an RNA virus, are also oncoviruses; HBV in particular is one of the main causes of hepatocellular carcinoma (HCC), the most common type of liver cancer.

The highly inflammatory environment fostered by HBV and HCV infection causes liver damage that often leads to cirrhosis. Continued infection can drive permanent damage to the hepatocytes, leading to genetic and epigenetic damage and driving oncogenesis. As an RNA virus, HCV doesn’t integrate into the genome and no confirmed viral oncoproteins have been identified to date, therefore it mostly drives cancer through these indirect mechanisms, which is also reflected in the fact that HCV-associated HCC predominantly occurs against a backdrop of liver cirrhosis.

HBV does integrate into the host genome. Genome sequencing studies revealed hundreds of integration sites, but most commonly they disrupted host genes involved in telomere stability and cell cycle regulation, providing some insight into the mechanisms by which HBV-associated HCC develops. In addition, HBV produces several oncoproteins, including HBx, which disrupts gene transcription, cell signaling pathways, cell cycle progress, apoptosis and other cellular processes.^15,16

Multitargeted tyrosine kinase inhibitors (TKIs) have been the focal point of therapeutic development in HCC. However, following the approval of sorafenib in 2008, there was a dearth of effective new treatment options despite substantial efforts and numerous phase 3 trials. More recently, immunotherapy has also come to the forefront, especially immune checkpoint inhibitors.

Last year marked the first new drug approvals in nearly a decade – the TKI regorafenib (Stivarga) and immune checkpoint inhibitor nivolumab (Opdivo), both in the second-line setting after failure of sorafenib. Treatment options in this setting may continue to expand, with the TKIs cabozantinib and lenvatinib and the immune checkpoint inhibitor pembrolizumab and the combination of durvalumab and tremelimumab hot on their heels.^17-20 Many of these drugs are also being evaluated in the front-line setting in comparison with sorafenib (Table 3).

218_de Lartigue_NT_T3_web.png

Adoptive cell therapy promising in EBV-positive cancers

More than 90% of the global population is infected with EBV, making it one of the most common human viruses. It is a member of the herpesvirus family that is probably best known as the cause of infectious mononucleosis. On rare occasions, however, EBV can cause tumor development, though our understanding of its exact pathogenic role in cancer is still incomplete.

EBV is a DNA virus that doesn’t tend to integrate into the host genome, but instead remains in the nucleus in the form of episomes and produces several oncoproteins, including latent membrane protein-1. It is associated with a range of different cancer types, including Burkitt lymphoma and other B-cell malignancies. It also infects epithelial cells and can cause nasopharyngeal carcinoma and gastric cancer, however, much less is known about the molecular underpinnings of these EBV-positive cancer types.^26,27Gastric cancers actually comprise the largest group of EBV-associated tumors because of the global incidence of this cancer type. The Cancer Genome Atlas Research Network recently characterized gastric cancer on a molecular level and identified an EBV-positive subgroup as a distinct clinical entity with unique molecular characteristics.²⁹

The focus of therapeutic development has again been on immunotherapy, however in this case the idea of collecting the patients T cells, engineering them to recognize EBV, and then reinfusing them into the patient – adoptive cell therapy – has gained the most traction (Table 4).

218_de Lartigue_NT_T4_web.png

Newest oncovirus on the block

The most recently discovered cancer-associated virus is Merkel cell polyomavirus (MCV), a DNA virus that was identified in 2008. Like EBV, virtually the whole global adult population is infected with MCV. It is linked to the development of a highly aggressive and lethal, though rare, form of skin cancer – Merkel cell carcinoma.

MCV is found in around 80% of MCC cases and in fewer than 10% of melanomas and other skin cancers. Thus far, several direct mechanisms of oncogenesis have been described, including integration of MCV into the host genome and the production of viral oncogenes, though their precise function is as yet unclear.^32-34

The American Cancer Society estimates that only 1500 cases of MCC are diagnosed each year in the United States.³⁵ Its rarity makes it difficult to conduct clinical trials with sufficient power, yet some headway has still been made.

Around half of MCCs express the programmed cell death ligand 1 (PD-L1) on their surface, making them a logical candidate for immune checkpoint inhibition. In 2017, avelumab became the first FDA-approved drug for the treatment of MCC. Approval was based on the JAVELIN Merkel 200 study in which 88 patients received avelumab. After 1 year of follow-up the ORR was 31.8%, with a CR rate of 9%.³⁶

Genome sequencing studies suggest that the mutational profile of MCV-positive tumors is quite different to those that are MCV-negative, which could have therapeutic implications. To date, these implications have not been delineated, given the challenge of small patient numbers, however an ongoing phase 1/2 trial is evaluating the combination of avelumab and radiation therapy or recombinant interferon beta, with or without MCV-specific cytotoxic T cells in patients with MCC and MCV infection.

The 2 other known cancer-causing viruses are human T-lymphotropic virus 1 (HTLV-1), a retrovirus associated with adult T-cell leukemia/lymphoma (ATL) and Kaposi sarcoma herpesvirus (KSHV). The latter is the causative agent of Kaposi sarcoma, often in combination with human immunodeficiency virus (HIV), a rare skin tumor that became renowned in the 1980s as an AIDS-defining illness.

The incidence of HTLV-1- and KSHV-positive tumors is substantially lower than the other virally associated cancers and, like MCC, this makes studying them and conducting clinical trials of novel therapeutic options a challenge. Nonetheless, several trials of targeted therapies and immunotherapies are underway.

Infection with certain viruses has been causally linked to the development of cancer. In recent years, an improved understanding of the unique pathology and molecular underpinnings of these virally associated cancers has prompted the development of more personalized treatment strategies, with a particular focus on immunotherapy. Here, we describe some of the latest developments.

The link between viruses and cancer

Suspicions about a possible role of viral infections in the development of cancer were first aroused in the early 1900s. The seminal discovery is traced back to Peyton Rous, who showed that a malignant tumor growing in a chicken could be transferred to a healthy bird by injecting it with tumor extracts that contained no actual tumor cells.¹

The infectious etiology of human cancer, however, remained controversial until many years later when the first cancer-causing virus, Epstein-Barr virus (EBV), was identified in cell cultures from patients with Burkitt lymphoma. Shortly afterward, the Rous sarcoma virus was unveiled as the oncogenic agent behind Rous’ observations.²Seven viruses have now been linked to the development of cancers and are thought to be responsible for around 12% of all cancer cases worldwide. The burden is likely to increase as technological advancements make it easier to establish a causal link between viruses and cancer development.³

In addition to making these links, researchers have also made significant headway in understanding how viruses cause cancer. Cancerous transformation of host cells occurs in only a minority of those who are infected with oncogenic viruses and often occurs in the setting of chronic infection.

Viruses can mediate carcinogenesis by direct and/or indirect mechanisms (Figure 1). Many of the hallmarks of cancer, the key attributes that drive the transformation from a normal cell to a malignant one, are compatible with the virus’s needs, such as needing to avoid cell death, increasing cell proliferation, and avoiding detection by the immune system.

218_de Lartigue_NT_F_web.png

Vaccines lead the charge in HPV-driven cancers

German virologist Harald zur Hausen received the Nobel Prize in 2008 for his discovery of the oncogenic role of human papillomaviruses (HPVs), a large family of more than 100 DNA viruses that infect the epithelial cells of the skin and mucous membranes. They are responsible for the largest number of virally associated cancer cases globally – around 5% (Table 1).

218_de Lartigue_NT_T1_web.png

A number of different cancer types are linked to HPV infection, but it is best known as the cause of cervical cancer. The development of diagnostic blood tests and prophylactic vaccines for prevention and early intervention in HPV infection has helped to reduce the incidence of cervical cancer. Conversely, another type of HPV-associated cancer, head and neck squamous cell carcinoma (HNSCC), has seen increased incidence in recent years.

HPVs are categorized according to their oncogenic potential as high, intermediate, or low risk. The high-risk HPV16 and HPV18 strains are most commonly associated with cancer. They are thought to cause cancer predominantly through integration into the host genome. The HPV genome is composed of 8 genes encoding proteins that regulate viral replication and assembly. The E6 and E7 genes are the most highly oncogenic; as the HPV DNA is inserted into the host genome, the transcriptional regulator of E6/E7 is lost, leading to their increased expression. These genes have significant oncogenic potential because of their interaction with 2 tumor suppressor proteins, p53 and pRb.^6,7

The largest investment in therapeutic development for HPV-positive cancers has been in the realm of immunotherapy in an effort to boost the anti-tumor immune response. In particular, there has been a focus on the development of therapeutic vaccines, designed to prime the anti-tumor immune response to recognize viral antigens. A variety of different types of vaccines are being developed, including live, attenuated and inactivated vaccines that are protein, DNA, or peptide based. Most developed to date target the E6/E7 proteins from the HPV16/18 strains (Table 2).^8,9

218_de Lartigue_NT_T2_web.png

Hepatocellular carcinoma: a tale of two viruses

The hepatitis viruses are a group of 5 unrelated viruses that causes inflammation of the liver. Hepatitis B (HBV), a DNA virus, and hepatitis C (HCV), an RNA virus, are also oncoviruses; HBV in particular is one of the main causes of hepatocellular carcinoma (HCC), the most common type of liver cancer.

The highly inflammatory environment fostered by HBV and HCV infection causes liver damage that often leads to cirrhosis. Continued infection can drive permanent damage to the hepatocytes, leading to genetic and epigenetic damage and driving oncogenesis. As an RNA virus, HCV doesn’t integrate into the genome and no confirmed viral oncoproteins have been identified to date, therefore it mostly drives cancer through these indirect mechanisms, which is also reflected in the fact that HCV-associated HCC predominantly occurs against a backdrop of liver cirrhosis.

HBV does integrate into the host genome. Genome sequencing studies revealed hundreds of integration sites, but most commonly they disrupted host genes involved in telomere stability and cell cycle regulation, providing some insight into the mechanisms by which HBV-associated HCC develops. In addition, HBV produces several oncoproteins, including HBx, which disrupts gene transcription, cell signaling pathways, cell cycle progress, apoptosis and other cellular processes.^15,16

Multitargeted tyrosine kinase inhibitors (TKIs) have been the focal point of therapeutic development in HCC. However, following the approval of sorafenib in 2008, there was a dearth of effective new treatment options despite substantial efforts and numerous phase 3 trials. More recently, immunotherapy has also come to the forefront, especially immune checkpoint inhibitors.

Last year marked the first new drug approvals in nearly a decade – the TKI regorafenib (Stivarga) and immune checkpoint inhibitor nivolumab (Opdivo), both in the second-line setting after failure of sorafenib. Treatment options in this setting may continue to expand, with the TKIs cabozantinib and lenvatinib and the immune checkpoint inhibitor pembrolizumab and the combination of durvalumab and tremelimumab hot on their heels.^17-20 Many of these drugs are also being evaluated in the front-line setting in comparison with sorafenib (Table 3).

218_de Lartigue_NT_T3_web.png

Adoptive cell therapy promising in EBV-positive cancers

More than 90% of the global population is infected with EBV, making it one of the most common human viruses. It is a member of the herpesvirus family that is probably best known as the cause of infectious mononucleosis. On rare occasions, however, EBV can cause tumor development, though our understanding of its exact pathogenic role in cancer is still incomplete.

EBV is a DNA virus that doesn’t tend to integrate into the host genome, but instead remains in the nucleus in the form of episomes and produces several oncoproteins, including latent membrane protein-1. It is associated with a range of different cancer types, including Burkitt lymphoma and other B-cell malignancies. It also infects epithelial cells and can cause nasopharyngeal carcinoma and gastric cancer, however, much less is known about the molecular underpinnings of these EBV-positive cancer types.^26,27Gastric cancers actually comprise the largest group of EBV-associated tumors because of the global incidence of this cancer type. The Cancer Genome Atlas Research Network recently characterized gastric cancer on a molecular level and identified an EBV-positive subgroup as a distinct clinical entity with unique molecular characteristics.²⁹

The focus of therapeutic development has again been on immunotherapy, however in this case the idea of collecting the patients T cells, engineering them to recognize EBV, and then reinfusing them into the patient – adoptive cell therapy – has gained the most traction (Table 4).

218_de Lartigue_NT_T4_web.png

Newest oncovirus on the block

The most recently discovered cancer-associated virus is Merkel cell polyomavirus (MCV), a DNA virus that was identified in 2008. Like EBV, virtually the whole global adult population is infected with MCV. It is linked to the development of a highly aggressive and lethal, though rare, form of skin cancer – Merkel cell carcinoma.

MCV is found in around 80% of MCC cases and in fewer than 10% of melanomas and other skin cancers. Thus far, several direct mechanisms of oncogenesis have been described, including integration of MCV into the host genome and the production of viral oncogenes, though their precise function is as yet unclear.^32-34

The American Cancer Society estimates that only 1500 cases of MCC are diagnosed each year in the United States.³⁵ Its rarity makes it difficult to conduct clinical trials with sufficient power, yet some headway has still been made.

Around half of MCCs express the programmed cell death ligand 1 (PD-L1) on their surface, making them a logical candidate for immune checkpoint inhibition. In 2017, avelumab became the first FDA-approved drug for the treatment of MCC. Approval was based on the JAVELIN Merkel 200 study in which 88 patients received avelumab. After 1 year of follow-up the ORR was 31.8%, with a CR rate of 9%.³⁶

Genome sequencing studies suggest that the mutational profile of MCV-positive tumors is quite different to those that are MCV-negative, which could have therapeutic implications. To date, these implications have not been delineated, given the challenge of small patient numbers, however an ongoing phase 1/2 trial is evaluating the combination of avelumab and radiation therapy or recombinant interferon beta, with or without MCV-specific cytotoxic T cells in patients with MCC and MCV infection.

The 2 other known cancer-causing viruses are human T-lymphotropic virus 1 (HTLV-1), a retrovirus associated with adult T-cell leukemia/lymphoma (ATL) and Kaposi sarcoma herpesvirus (KSHV). The latter is the causative agent of Kaposi sarcoma, often in combination with human immunodeficiency virus (HIV), a rare skin tumor that became renowned in the 1980s as an AIDS-defining illness.

The incidence of HTLV-1- and KSHV-positive tumors is substantially lower than the other virally associated cancers and, like MCC, this makes studying them and conducting clinical trials of novel therapeutic options a challenge. Nonetheless, several trials of targeted therapies and immunotherapies are underway.

Infection with certain viruses has been causally linked to the development of cancer. In recent years, an improved understanding of the unique pathology and molecular underpinnings of these virally associated cancers has prompted the development of more personalized treatment strategies, with a particular focus on immunotherapy. Here, we describe some of the latest developments.

The link between viruses and cancer

Suspicions about a possible role of viral infections in the development of cancer were first aroused in the early 1900s. The seminal discovery is traced back to Peyton Rous, who showed that a malignant tumor growing in a chicken could be transferred to a healthy bird by injecting it with tumor extracts that contained no actual tumor cells.¹

The infectious etiology of human cancer, however, remained controversial until many years later when the first cancer-causing virus, Epstein-Barr virus (EBV), was identified in cell cultures from patients with Burkitt lymphoma. Shortly afterward, the Rous sarcoma virus was unveiled as the oncogenic agent behind Rous’ observations.²Seven viruses have now been linked to the development of cancers and are thought to be responsible for around 12% of all cancer cases worldwide. The burden is likely to increase as technological advancements make it easier to establish a causal link between viruses and cancer development.³

In addition to making these links, researchers have also made significant headway in understanding how viruses cause cancer. Cancerous transformation of host cells occurs in only a minority of those who are infected with oncogenic viruses and often occurs in the setting of chronic infection.

Viruses can mediate carcinogenesis by direct and/or indirect mechanisms (Figure 1). Many of the hallmarks of cancer, the key attributes that drive the transformation from a normal cell to a malignant one, are compatible with the virus’s needs, such as needing to avoid cell death, increasing cell proliferation, and avoiding detection by the immune system.

218_de Lartigue_NT_F_web.png

Vaccines lead the charge in HPV-driven cancers

German virologist Harald zur Hausen received the Nobel Prize in 2008 for his discovery of the oncogenic role of human papillomaviruses (HPVs), a large family of more than 100 DNA viruses that infect the epithelial cells of the skin and mucous membranes. They are responsible for the largest number of virally associated cancer cases globally – around 5% (Table 1).

218_de Lartigue_NT_T1_web.png

A number of different cancer types are linked to HPV infection, but it is best known as the cause of cervical cancer. The development of diagnostic blood tests and prophylactic vaccines for prevention and early intervention in HPV infection has helped to reduce the incidence of cervical cancer. Conversely, another type of HPV-associated cancer, head and neck squamous cell carcinoma (HNSCC), has seen increased incidence in recent years.

HPVs are categorized according to their oncogenic potential as high, intermediate, or low risk. The high-risk HPV16 and HPV18 strains are most commonly associated with cancer. They are thought to cause cancer predominantly through integration into the host genome. The HPV genome is composed of 8 genes encoding proteins that regulate viral replication and assembly. The E6 and E7 genes are the most highly oncogenic; as the HPV DNA is inserted into the host genome, the transcriptional regulator of E6/E7 is lost, leading to their increased expression. These genes have significant oncogenic potential because of their interaction with 2 tumor suppressor proteins, p53 and pRb.^6,7

The largest investment in therapeutic development for HPV-positive cancers has been in the realm of immunotherapy in an effort to boost the anti-tumor immune response. In particular, there has been a focus on the development of therapeutic vaccines, designed to prime the anti-tumor immune response to recognize viral antigens. A variety of different types of vaccines are being developed, including live, attenuated and inactivated vaccines that are protein, DNA, or peptide based. Most developed to date target the E6/E7 proteins from the HPV16/18 strains (Table 2).^8,9

218_de Lartigue_NT_T2_web.png

Hepatocellular carcinoma: a tale of two viruses

The hepatitis viruses are a group of 5 unrelated viruses that causes inflammation of the liver. Hepatitis B (HBV), a DNA virus, and hepatitis C (HCV), an RNA virus, are also oncoviruses; HBV in particular is one of the main causes of hepatocellular carcinoma (HCC), the most common type of liver cancer.

The highly inflammatory environment fostered by HBV and HCV infection causes liver damage that often leads to cirrhosis. Continued infection can drive permanent damage to the hepatocytes, leading to genetic and epigenetic damage and driving oncogenesis. As an RNA virus, HCV doesn’t integrate into the genome and no confirmed viral oncoproteins have been identified to date, therefore it mostly drives cancer through these indirect mechanisms, which is also reflected in the fact that HCV-associated HCC predominantly occurs against a backdrop of liver cirrhosis.

HBV does integrate into the host genome. Genome sequencing studies revealed hundreds of integration sites, but most commonly they disrupted host genes involved in telomere stability and cell cycle regulation, providing some insight into the mechanisms by which HBV-associated HCC develops. In addition, HBV produces several oncoproteins, including HBx, which disrupts gene transcription, cell signaling pathways, cell cycle progress, apoptosis and other cellular processes.^15,16

Multitargeted tyrosine kinase inhibitors (TKIs) have been the focal point of therapeutic development in HCC. However, following the approval of sorafenib in 2008, there was a dearth of effective new treatment options despite substantial efforts and numerous phase 3 trials. More recently, immunotherapy has also come to the forefront, especially immune checkpoint inhibitors.

Last year marked the first new drug approvals in nearly a decade – the TKI regorafenib (Stivarga) and immune checkpoint inhibitor nivolumab (Opdivo), both in the second-line setting after failure of sorafenib. Treatment options in this setting may continue to expand, with the TKIs cabozantinib and lenvatinib and the immune checkpoint inhibitor pembrolizumab and the combination of durvalumab and tremelimumab hot on their heels.^17-20 Many of these drugs are also being evaluated in the front-line setting in comparison with sorafenib (Table 3).

218_de Lartigue_NT_T3_web.png

Adoptive cell therapy promising in EBV-positive cancers

More than 90% of the global population is infected with EBV, making it one of the most common human viruses. It is a member of the herpesvirus family that is probably best known as the cause of infectious mononucleosis. On rare occasions, however, EBV can cause tumor development, though our understanding of its exact pathogenic role in cancer is still incomplete.

EBV is a DNA virus that doesn’t tend to integrate into the host genome, but instead remains in the nucleus in the form of episomes and produces several oncoproteins, including latent membrane protein-1. It is associated with a range of different cancer types, including Burkitt lymphoma and other B-cell malignancies. It also infects epithelial cells and can cause nasopharyngeal carcinoma and gastric cancer, however, much less is known about the molecular underpinnings of these EBV-positive cancer types.^26,27Gastric cancers actually comprise the largest group of EBV-associated tumors because of the global incidence of this cancer type. The Cancer Genome Atlas Research Network recently characterized gastric cancer on a molecular level and identified an EBV-positive subgroup as a distinct clinical entity with unique molecular characteristics.²⁹

The focus of therapeutic development has again been on immunotherapy, however in this case the idea of collecting the patients T cells, engineering them to recognize EBV, and then reinfusing them into the patient – adoptive cell therapy – has gained the most traction (Table 4).

218_de Lartigue_NT_T4_web.png

Newest oncovirus on the block

The most recently discovered cancer-associated virus is Merkel cell polyomavirus (MCV), a DNA virus that was identified in 2008. Like EBV, virtually the whole global adult population is infected with MCV. It is linked to the development of a highly aggressive and lethal, though rare, form of skin cancer – Merkel cell carcinoma.

MCV is found in around 80% of MCC cases and in fewer than 10% of melanomas and other skin cancers. Thus far, several direct mechanisms of oncogenesis have been described, including integration of MCV into the host genome and the production of viral oncogenes, though their precise function is as yet unclear.^32-34

The American Cancer Society estimates that only 1500 cases of MCC are diagnosed each year in the United States.³⁵ Its rarity makes it difficult to conduct clinical trials with sufficient power, yet some headway has still been made.

Around half of MCCs express the programmed cell death ligand 1 (PD-L1) on their surface, making them a logical candidate for immune checkpoint inhibition. In 2017, avelumab became the first FDA-approved drug for the treatment of MCC. Approval was based on the JAVELIN Merkel 200 study in which 88 patients received avelumab. After 1 year of follow-up the ORR was 31.8%, with a CR rate of 9%.³⁶

Genome sequencing studies suggest that the mutational profile of MCV-positive tumors is quite different to those that are MCV-negative, which could have therapeutic implications. To date, these implications have not been delineated, given the challenge of small patient numbers, however an ongoing phase 1/2 trial is evaluating the combination of avelumab and radiation therapy or recombinant interferon beta, with or without MCV-specific cytotoxic T cells in patients with MCC and MCV infection.

The 2 other known cancer-causing viruses are human T-lymphotropic virus 1 (HTLV-1), a retrovirus associated with adult T-cell leukemia/lymphoma (ATL) and Kaposi sarcoma herpesvirus (KSHV). The latter is the causative agent of Kaposi sarcoma, often in combination with human immunodeficiency virus (HIV), a rare skin tumor that became renowned in the 1980s as an AIDS-defining illness.

The incidence of HTLV-1- and KSHV-positive tumors is substantially lower than the other virally associated cancers and, like MCC, this makes studying them and conducting clinical trials of novel therapeutic options a challenge. Nonetheless, several trials of targeted therapies and immunotherapies are underway.

Although there have been significant improvements in patient outcomes for some forms of pediatric cancer, progress has been painfully slow for others. An increasing understanding of pediatric cancers is highlighting the unique molecular drivers and challenging the assumption that drugs developed in adults can be applied to children and young adults. Here, we discuss game-changing therapeutic advances and a shifting view of childhood cancers.

Unique genomic background

Although pediatric cancers are rare, representing just 1% of all new cancers diagnosed annually in the United States, they are the second leading cause of death in children aged 1 to 14 years. There are many different histological tumor types under the umbrella of childhood cancers, of which the most common are leukemias, central nervous system tumors, and lymphomas (Figure 1).^1,2

518_deL NT ped ca_F1_web.PNG

A ‘magic bullet’?

Chromosomal rearrangements are common in pediatric cancers. This type of molecular abnormality can result in a fusion of 2 different genes when the chromosome breaks apart and the pieces join back together in a muddled order. If the genetic code fuses in a manner that is “readable” by the cell, then it can drive aberrant activation of one or both genes.⁷ Gene fusions often involve kinase enzymes that are essential players in cell signaling pathways regulating hallmark cancer processes, such as unchecked cell proliferation. The fusion drives the constitutive activation of the kinase and, thus, these downstream signaling pathways.

One of the first chromosomal rearrangements linked to cancer, BCR-ABL1 – more commonly known as the Philadelphia chromosome – results in aberrant activation of the ABL1 kinase. It is present in nearly all cases of chronic myeloid leukemia (CML) and 3% to 5% of patients with ALL, and thus became the central focus of targeted drug development. Imatinib was initially approved by the US Food and Drug Administration (FDA) in 2001 for the treatment of adult patients with CML and had such a significant impact on the treatment landscape that it made the cover of Time magazine as a “magic bullet” in the war on cancer.⁸

Approval was expanded into pediatric patients in 2006 and for pediatric patients with ALL in 2013. However, as with the use of most kinase inhibitors, tumors can evolve under the selective pressure of treatment, developing additional molecular abnormalities that drive resistance.⁹

Next-generation multikinase inhibitors that more potently inhibit the BCR-ABL1 fusion protein have been developed to provide additional treatment options for patients who become resistant to imatinib. Dasatinib and nilotinib are among several drugs that have recently been approved for pediatric cancer therapy (Table 1). Both therapies were approved to treat children with Philadelphia chromosome-positive CML in the chronic phase in either the front- or second-line setting after failure of imatinib.

518_deL NT ped ca_T1_web.PNG

Other prominent gene fusions

Gene fusions involving the anaplastic lymphoma kinase (ALK) occur in patients with non–small-cell lung cancer and ALK inhibitors have provided an effective new treatment option for patients whose tumors display this abnormality.

ALK fusions are also a prominent feature of several kinds of pediatric cancers and ALK inhibitors offer promise in this setting.^7,12 An NPM-ALK fusion is found in 90% of pediatric anaplastic large cell lymphoma (ALCL) cases,¹³ whereas a variety of ALK fusions are found in up to half of patients with inflammatory myofibroblastic tumor (IMT), a rare form of soft tissue sarcoma.¹⁴ ALK inhibitors are being tested in a variety of clinical trials in pediatric patients (Table 2).

518_deL NT ped ca_T2_web.PNG

518_deL NT ped ca_F2_web.PNG

CAR T cells transformative in ALL

A variety of different types of immunotherapy have been tested in patients with pediatric cancers. In general, immunotherapy has proved less effective than in adult cancers, possibly because of the lower tumor mutation burden in pediatric cancers, which means there are likely fewer cancer antigens to provoke an anti-tumor immune response.

There are notable exceptions among the disappointments, however, and most exciting is the development of chimeric antigen receptor (CAR) T cells. CAR T cells fall into a category of immunotherapy known as adoptive cell therapy (ACT), in which immune cells are harvested from a patient and grown outside the body to increase their numbers before being reinfused into the patient.

In the case of CAR T-cell therapy, the cells are genetically engineered to express a CAR that endows them with tumor-targeting capabilities. To date, the development of CAR T cells has focused on the use of the CD19 antigen as a target, which is highly expressed on a variety of B-cell malignancies, including several of the most common forms of pediatric cancer. ASCO shined the spotlight on CAR T-cell therapy this year, naming it the Advance of the Year for 2018, saying that the treatment is “poised to transform childhood ALL.”¹⁹

Two CD19-targeted CAR T-cell therapies – tisagenlecleucel and axicabtagene ciloleucel – were brought to market in 2017. Only tisagenlecleucel is approved in the pediatric ALL population, however, having been awarded approval for the treatment of patients aged up to 25 years whose disease is refractory to or relapsed after receiving at least 2 prior therapies. In the pivotal trial, complete responses were observed in more than 60% of patients.²⁰ Clinical trials of both CAR T-cell therapies in pediatric ALL and non-Hodgkin lymphoma are ongoing (Table 3).

518_deL NT ped ca_T3_web.PNG

More immunotherapy approvals

The immune checkpoint inhibitors, which work by blocking inhibitory receptors on the surface of T cells, have also had recent approvals in pediatric patient populations. Pembrolizumab and nivolumab, inhibitors of the programmed cell death receptor 1 (PD-1) protein, have both been approved for use in adult and pediatric patients (older than 12 years) with relapsed/refractory metastatic colorectal cancer (and other solid tumors in the case of pembrolizumab) that display defects in the mismatch repair pathway that fixes damaged DNA or in patients that have high levels of microsatellite instability. Both deficient mismatch repair and microsatellite instability–high can indicate a high mutation burden in a tumor, which may predict increased sensitivity to immunotherapy.²⁹

The approval in pediatric patients in both of those instances, however, was not based on data in pediatric patient populations but extrapolated from adult patients. Pembrolizumab is also approved for the treatment of adults and pediatric patients with classical Hodgkin lymphoma (cHL) after 3 or more previous treatments, but once again efficacy in the pediatric population was inferred from clinical trials performed in adults. Most recently, pembrolizumab was approved for the treatment of adult and pediatric patients with relapsed or refractory primary mediastinal large B-cell lymphoma.³⁰Ipilimumab, which targets a different T cell receptor – cytotoxic T lymphocyte antigen-4 (CTLA-4) – has been approved for the treatment of pediatric patients aged 12 years and older with metastatic melanoma. This expanded indication, following on from its approval in adult patients in 2011, was based on data from 2 trials in which objective responses were observed in 2 out of 17 patients, including 1 partial response that lasted 16 months.³¹Finally, antibody-drug conjugates (ADC), in which tumor antigen-targeting monoclonal antibodies are conjugated to cytotoxic payloads to combine the specificity of an antibody with the cell-killing potency of chemotherapy, have also generated some recent successes in pediatric cancers.

Gemtuzumab ozogamicin is an ADC that targets the CD33 protein, which is highly expressed on 85%-90% of cases of acute myeloid leukemia (AML). In 2000, it was the first ADC to be brought to market in the United States, but it was subsequently voluntarily withdrawn by the manufacturer in 2010 after confirmatory trials failed to show a survival benefit.

Recently, a meta-analysis of gemtuzumab ozogamicin trials suggested that the drug likely does improve long-term overall survival (OS) and reduce the risk of relapse and researchers developed an intermittent dosing schedule to help mitigate toxicity.³² This new dosing regimen received FDA approval in 2017 for the treatment of pediatric patients aged 2 years and older on the basis of 2 clinical trials.

In the MyloFrance-1 trial, 57 patients were administered 3 mg/m² gemtuzumab ozogamicin on days 1, 4, and 7 followed by cytarabine consolidation therapy and demonstrated a 26% CR rate and median recurrence-free survival of 11.6 months. In the phase 3 AML-19 trial, 237 patients received gemtuzumab ozogamicin at a dose of 6 mg/m² on day 1 and 3 mg/m² on day 8 or best supportive care. Gemtuzumab ozogamicin improved OS from 3.6 to 4.9 months.^33,34

Inotuzumab ozogamicin is a CD22-targeting ADC that has been FDA approved for the treatment of adult patients with relapsed/refractory B-cell precursor ALL since last year. The therapy has been available to pediatric patients through a compassionate access program, but it has not been extensively evaluated in this population. The results of a retrospective analysis of pediatric patients who received at least 1 dose of inotuzumab ozogamicin were presented at ASCO in 2017. Among 29 patients with heavily pretreated disease the CR rate was 62%, 72% of whom achieved MRD negativity.³⁵

Although there have been significant improvements in patient outcomes for some forms of pediatric cancer, progress has been painfully slow for others. An increasing understanding of pediatric cancers is highlighting the unique molecular drivers and challenging the assumption that drugs developed in adults can be applied to children and young adults. Here, we discuss game-changing therapeutic advances and a shifting view of childhood cancers.

Unique genomic background

Although pediatric cancers are rare, representing just 1% of all new cancers diagnosed annually in the United States, they are the second leading cause of death in children aged 1 to 14 years. There are many different histological tumor types under the umbrella of childhood cancers, of which the most common are leukemias, central nervous system tumors, and lymphomas (Figure 1).^1,2

518_deL NT ped ca_F1_web.PNG

A ‘magic bullet’?

Chromosomal rearrangements are common in pediatric cancers. This type of molecular abnormality can result in a fusion of 2 different genes when the chromosome breaks apart and the pieces join back together in a muddled order. If the genetic code fuses in a manner that is “readable” by the cell, then it can drive aberrant activation of one or both genes.⁷ Gene fusions often involve kinase enzymes that are essential players in cell signaling pathways regulating hallmark cancer processes, such as unchecked cell proliferation. The fusion drives the constitutive activation of the kinase and, thus, these downstream signaling pathways.

One of the first chromosomal rearrangements linked to cancer, BCR-ABL1 – more commonly known as the Philadelphia chromosome – results in aberrant activation of the ABL1 kinase. It is present in nearly all cases of chronic myeloid leukemia (CML) and 3% to 5% of patients with ALL, and thus became the central focus of targeted drug development. Imatinib was initially approved by the US Food and Drug Administration (FDA) in 2001 for the treatment of adult patients with CML and had such a significant impact on the treatment landscape that it made the cover of Time magazine as a “magic bullet” in the war on cancer.⁸

Approval was expanded into pediatric patients in 2006 and for pediatric patients with ALL in 2013. However, as with the use of most kinase inhibitors, tumors can evolve under the selective pressure of treatment, developing additional molecular abnormalities that drive resistance.⁹

Next-generation multikinase inhibitors that more potently inhibit the BCR-ABL1 fusion protein have been developed to provide additional treatment options for patients who become resistant to imatinib. Dasatinib and nilotinib are among several drugs that have recently been approved for pediatric cancer therapy (Table 1). Both therapies were approved to treat children with Philadelphia chromosome-positive CML in the chronic phase in either the front- or second-line setting after failure of imatinib.

518_deL NT ped ca_T1_web.PNG

Other prominent gene fusions

Gene fusions involving the anaplastic lymphoma kinase (ALK) occur in patients with non–small-cell lung cancer and ALK inhibitors have provided an effective new treatment option for patients whose tumors display this abnormality.

ALK fusions are also a prominent feature of several kinds of pediatric cancers and ALK inhibitors offer promise in this setting.^7,12 An NPM-ALK fusion is found in 90% of pediatric anaplastic large cell lymphoma (ALCL) cases,¹³ whereas a variety of ALK fusions are found in up to half of patients with inflammatory myofibroblastic tumor (IMT), a rare form of soft tissue sarcoma.¹⁴ ALK inhibitors are being tested in a variety of clinical trials in pediatric patients (Table 2).

518_deL NT ped ca_T2_web.PNG

518_deL NT ped ca_F2_web.PNG

CAR T cells transformative in ALL

A variety of different types of immunotherapy have been tested in patients with pediatric cancers. In general, immunotherapy has proved less effective than in adult cancers, possibly because of the lower tumor mutation burden in pediatric cancers, which means there are likely fewer cancer antigens to provoke an anti-tumor immune response.

There are notable exceptions among the disappointments, however, and most exciting is the development of chimeric antigen receptor (CAR) T cells. CAR T cells fall into a category of immunotherapy known as adoptive cell therapy (ACT), in which immune cells are harvested from a patient and grown outside the body to increase their numbers before being reinfused into the patient.

In the case of CAR T-cell therapy, the cells are genetically engineered to express a CAR that endows them with tumor-targeting capabilities. To date, the development of CAR T cells has focused on the use of the CD19 antigen as a target, which is highly expressed on a variety of B-cell malignancies, including several of the most common forms of pediatric cancer. ASCO shined the spotlight on CAR T-cell therapy this year, naming it the Advance of the Year for 2018, saying that the treatment is “poised to transform childhood ALL.”¹⁹

Two CD19-targeted CAR T-cell therapies – tisagenlecleucel and axicabtagene ciloleucel – were brought to market in 2017. Only tisagenlecleucel is approved in the pediatric ALL population, however, having been awarded approval for the treatment of patients aged up to 25 years whose disease is refractory to or relapsed after receiving at least 2 prior therapies. In the pivotal trial, complete responses were observed in more than 60% of patients.²⁰ Clinical trials of both CAR T-cell therapies in pediatric ALL and non-Hodgkin lymphoma are ongoing (Table 3).

518_deL NT ped ca_T3_web.PNG

More immunotherapy approvals

The immune checkpoint inhibitors, which work by blocking inhibitory receptors on the surface of T cells, have also had recent approvals in pediatric patient populations. Pembrolizumab and nivolumab, inhibitors of the programmed cell death receptor 1 (PD-1) protein, have both been approved for use in adult and pediatric patients (older than 12 years) with relapsed/refractory metastatic colorectal cancer (and other solid tumors in the case of pembrolizumab) that display defects in the mismatch repair pathway that fixes damaged DNA or in patients that have high levels of microsatellite instability. Both deficient mismatch repair and microsatellite instability–high can indicate a high mutation burden in a tumor, which may predict increased sensitivity to immunotherapy.²⁹

The approval in pediatric patients in both of those instances, however, was not based on data in pediatric patient populations but extrapolated from adult patients. Pembrolizumab is also approved for the treatment of adults and pediatric patients with classical Hodgkin lymphoma (cHL) after 3 or more previous treatments, but once again efficacy in the pediatric population was inferred from clinical trials performed in adults. Most recently, pembrolizumab was approved for the treatment of adult and pediatric patients with relapsed or refractory primary mediastinal large B-cell lymphoma.³⁰Ipilimumab, which targets a different T cell receptor – cytotoxic T lymphocyte antigen-4 (CTLA-4) – has been approved for the treatment of pediatric patients aged 12 years and older with metastatic melanoma. This expanded indication, following on from its approval in adult patients in 2011, was based on data from 2 trials in which objective responses were observed in 2 out of 17 patients, including 1 partial response that lasted 16 months.³¹Finally, antibody-drug conjugates (ADC), in which tumor antigen-targeting monoclonal antibodies are conjugated to cytotoxic payloads to combine the specificity of an antibody with the cell-killing potency of chemotherapy, have also generated some recent successes in pediatric cancers.

Gemtuzumab ozogamicin is an ADC that targets the CD33 protein, which is highly expressed on 85%-90% of cases of acute myeloid leukemia (AML). In 2000, it was the first ADC to be brought to market in the United States, but it was subsequently voluntarily withdrawn by the manufacturer in 2010 after confirmatory trials failed to show a survival benefit.

Recently, a meta-analysis of gemtuzumab ozogamicin trials suggested that the drug likely does improve long-term overall survival (OS) and reduce the risk of relapse and researchers developed an intermittent dosing schedule to help mitigate toxicity.³² This new dosing regimen received FDA approval in 2017 for the treatment of pediatric patients aged 2 years and older on the basis of 2 clinical trials.

In the MyloFrance-1 trial, 57 patients were administered 3 mg/m² gemtuzumab ozogamicin on days 1, 4, and 7 followed by cytarabine consolidation therapy and demonstrated a 26% CR rate and median recurrence-free survival of 11.6 months. In the phase 3 AML-19 trial, 237 patients received gemtuzumab ozogamicin at a dose of 6 mg/m² on day 1 and 3 mg/m² on day 8 or best supportive care. Gemtuzumab ozogamicin improved OS from 3.6 to 4.9 months.^33,34

Inotuzumab ozogamicin is a CD22-targeting ADC that has been FDA approved for the treatment of adult patients with relapsed/refractory B-cell precursor ALL since last year. The therapy has been available to pediatric patients through a compassionate access program, but it has not been extensively evaluated in this population. The results of a retrospective analysis of pediatric patients who received at least 1 dose of inotuzumab ozogamicin were presented at ASCO in 2017. Among 29 patients with heavily pretreated disease the CR rate was 62%, 72% of whom achieved MRD negativity.³⁵

Although there have been significant improvements in patient outcomes for some forms of pediatric cancer, progress has been painfully slow for others. An increasing understanding of pediatric cancers is highlighting the unique molecular drivers and challenging the assumption that drugs developed in adults can be applied to children and young adults. Here, we discuss game-changing therapeutic advances and a shifting view of childhood cancers.

Unique genomic background

Although pediatric cancers are rare, representing just 1% of all new cancers diagnosed annually in the United States, they are the second leading cause of death in children aged 1 to 14 years. There are many different histological tumor types under the umbrella of childhood cancers, of which the most common are leukemias, central nervous system tumors, and lymphomas (Figure 1).^1,2

518_deL NT ped ca_F1_web.PNG

A ‘magic bullet’?

Chromosomal rearrangements are common in pediatric cancers. This type of molecular abnormality can result in a fusion of 2 different genes when the chromosome breaks apart and the pieces join back together in a muddled order. If the genetic code fuses in a manner that is “readable” by the cell, then it can drive aberrant activation of one or both genes.⁷ Gene fusions often involve kinase enzymes that are essential players in cell signaling pathways regulating hallmark cancer processes, such as unchecked cell proliferation. The fusion drives the constitutive activation of the kinase and, thus, these downstream signaling pathways.

One of the first chromosomal rearrangements linked to cancer, BCR-ABL1 – more commonly known as the Philadelphia chromosome – results in aberrant activation of the ABL1 kinase. It is present in nearly all cases of chronic myeloid leukemia (CML) and 3% to 5% of patients with ALL, and thus became the central focus of targeted drug development. Imatinib was initially approved by the US Food and Drug Administration (FDA) in 2001 for the treatment of adult patients with CML and had such a significant impact on the treatment landscape that it made the cover of Time magazine as a “magic bullet” in the war on cancer.⁸

Approval was expanded into pediatric patients in 2006 and for pediatric patients with ALL in 2013. However, as with the use of most kinase inhibitors, tumors can evolve under the selective pressure of treatment, developing additional molecular abnormalities that drive resistance.⁹

Next-generation multikinase inhibitors that more potently inhibit the BCR-ABL1 fusion protein have been developed to provide additional treatment options for patients who become resistant to imatinib. Dasatinib and nilotinib are among several drugs that have recently been approved for pediatric cancer therapy (Table 1). Both therapies were approved to treat children with Philadelphia chromosome-positive CML in the chronic phase in either the front- or second-line setting after failure of imatinib.

518_deL NT ped ca_T1_web.PNG

Other prominent gene fusions

Gene fusions involving the anaplastic lymphoma kinase (ALK) occur in patients with non–small-cell lung cancer and ALK inhibitors have provided an effective new treatment option for patients whose tumors display this abnormality.

ALK fusions are also a prominent feature of several kinds of pediatric cancers and ALK inhibitors offer promise in this setting.^7,12 An NPM-ALK fusion is found in 90% of pediatric anaplastic large cell lymphoma (ALCL) cases,¹³ whereas a variety of ALK fusions are found in up to half of patients with inflammatory myofibroblastic tumor (IMT), a rare form of soft tissue sarcoma.¹⁴ ALK inhibitors are being tested in a variety of clinical trials in pediatric patients (Table 2).

518_deL NT ped ca_T2_web.PNG

518_deL NT ped ca_F2_web.PNG

CAR T cells transformative in ALL

A variety of different types of immunotherapy have been tested in patients with pediatric cancers. In general, immunotherapy has proved less effective than in adult cancers, possibly because of the lower tumor mutation burden in pediatric cancers, which means there are likely fewer cancer antigens to provoke an anti-tumor immune response.

There are notable exceptions among the disappointments, however, and most exciting is the development of chimeric antigen receptor (CAR) T cells. CAR T cells fall into a category of immunotherapy known as adoptive cell therapy (ACT), in which immune cells are harvested from a patient and grown outside the body to increase their numbers before being reinfused into the patient.

In the case of CAR T-cell therapy, the cells are genetically engineered to express a CAR that endows them with tumor-targeting capabilities. To date, the development of CAR T cells has focused on the use of the CD19 antigen as a target, which is highly expressed on a variety of B-cell malignancies, including several of the most common forms of pediatric cancer. ASCO shined the spotlight on CAR T-cell therapy this year, naming it the Advance of the Year for 2018, saying that the treatment is “poised to transform childhood ALL.”¹⁹

Two CD19-targeted CAR T-cell therapies – tisagenlecleucel and axicabtagene ciloleucel – were brought to market in 2017. Only tisagenlecleucel is approved in the pediatric ALL population, however, having been awarded approval for the treatment of patients aged up to 25 years whose disease is refractory to or relapsed after receiving at least 2 prior therapies. In the pivotal trial, complete responses were observed in more than 60% of patients.²⁰ Clinical trials of both CAR T-cell therapies in pediatric ALL and non-Hodgkin lymphoma are ongoing (Table 3).

518_deL NT ped ca_T3_web.PNG

More immunotherapy approvals

The immune checkpoint inhibitors, which work by blocking inhibitory receptors on the surface of T cells, have also had recent approvals in pediatric patient populations. Pembrolizumab and nivolumab, inhibitors of the programmed cell death receptor 1 (PD-1) protein, have both been approved for use in adult and pediatric patients (older than 12 years) with relapsed/refractory metastatic colorectal cancer (and other solid tumors in the case of pembrolizumab) that display defects in the mismatch repair pathway that fixes damaged DNA or in patients that have high levels of microsatellite instability. Both deficient mismatch repair and microsatellite instability–high can indicate a high mutation burden in a tumor, which may predict increased sensitivity to immunotherapy.²⁹

The approval in pediatric patients in both of those instances, however, was not based on data in pediatric patient populations but extrapolated from adult patients. Pembrolizumab is also approved for the treatment of adults and pediatric patients with classical Hodgkin lymphoma (cHL) after 3 or more previous treatments, but once again efficacy in the pediatric population was inferred from clinical trials performed in adults. Most recently, pembrolizumab was approved for the treatment of adult and pediatric patients with relapsed or refractory primary mediastinal large B-cell lymphoma.³⁰Ipilimumab, which targets a different T cell receptor – cytotoxic T lymphocyte antigen-4 (CTLA-4) – has been approved for the treatment of pediatric patients aged 12 years and older with metastatic melanoma. This expanded indication, following on from its approval in adult patients in 2011, was based on data from 2 trials in which objective responses were observed in 2 out of 17 patients, including 1 partial response that lasted 16 months.³¹Finally, antibody-drug conjugates (ADC), in which tumor antigen-targeting monoclonal antibodies are conjugated to cytotoxic payloads to combine the specificity of an antibody with the cell-killing potency of chemotherapy, have also generated some recent successes in pediatric cancers.

Gemtuzumab ozogamicin is an ADC that targets the CD33 protein, which is highly expressed on 85%-90% of cases of acute myeloid leukemia (AML). In 2000, it was the first ADC to be brought to market in the United States, but it was subsequently voluntarily withdrawn by the manufacturer in 2010 after confirmatory trials failed to show a survival benefit.

Recently, a meta-analysis of gemtuzumab ozogamicin trials suggested that the drug likely does improve long-term overall survival (OS) and reduce the risk of relapse and researchers developed an intermittent dosing schedule to help mitigate toxicity.³² This new dosing regimen received FDA approval in 2017 for the treatment of pediatric patients aged 2 years and older on the basis of 2 clinical trials.

In the MyloFrance-1 trial, 57 patients were administered 3 mg/m² gemtuzumab ozogamicin on days 1, 4, and 7 followed by cytarabine consolidation therapy and demonstrated a 26% CR rate and median recurrence-free survival of 11.6 months. In the phase 3 AML-19 trial, 237 patients received gemtuzumab ozogamicin at a dose of 6 mg/m² on day 1 and 3 mg/m² on day 8 or best supportive care. Gemtuzumab ozogamicin improved OS from 3.6 to 4.9 months.^33,34

Inotuzumab ozogamicin is a CD22-targeting ADC that has been FDA approved for the treatment of adult patients with relapsed/refractory B-cell precursor ALL since last year. The therapy has been available to pediatric patients through a compassionate access program, but it has not been extensively evaluated in this population. The results of a retrospective analysis of pediatric patients who received at least 1 dose of inotuzumab ozogamicin were presented at ASCO in 2017. Among 29 patients with heavily pretreated disease the CR rate was 62%, 72% of whom achieved MRD negativity.³⁵

The rarity and complexities of bone and soft tissue sarcomas pose a major challenge to effective treatment. Historically, there has been a blanket approach to treatment, but more recently that has begun to change thanks to genome profiling studies and novel clinical trial strategies. Here, we discuss the resulting enrichment of the therapeutic armamentarium with molecularly targeted and immune therapies.

A challenging tumor type

Sarcomas are a large group of histologically diverse cancers that arise in the mesenchymal cells. They can be broadly divided into bone and soft tissue sarcomas (STS) but are further subdivided according to the type of cell from which they derive; osteosarcomas in the bone, rhabdomyosarcomas in the skeletal muscle, liposarcomas in the fat tissues, leiomyosarcomas in the smooth muscle, and chondrosarcomas in the cartilaginous tissue, for example.

Each sarcoma subtype itself encompasses a range of different cancers with unique biology. Under the umbrella of liposarcoma, for example, are well/dedifferentiated liposarcomas and myxoid liposarcomas, which have very different pathologies and clinical courses.

As a whole, sarcomas are extremely rare tumors, accounting for less than 1% of all adult cancers, although they disproportionately affect children and young adults, with a prevalence closer to 15%.^1,2 Certain sarcoma subtypes are exceptionally rare, with only a few cases diagnosed worldwide each year, whereas liposarcomas are at the other end of the spectrum, comprising the most common form of STS (Figure 1).³

418_de Lartigue_NT sarc_F1_web.PNG

Magic bullet for GIST

Despite their clear heterogeneity and complexity, sarcomas have tended to be treated as a single entity. Chemotherapy has played a central role in the treatment of advanced sarcomas and continues to do so, with 2 newer drugs approved by the United States Food and Drug Administration (FDA) in the past several years.^6,7

The development of targeted therapy, on the other hand, for the most part proved unsuccessful. In general, studies examining the somatic mutation landscape in sarcomas found very few that were highly recurrent. The exception was gastrointestinal stromal tumors (GIST), which represent around 8% of STS.⁸ Frequent mutations in several highly targetable tyrosine kinases, notably KIT, which is mutated in around 85% of cases,⁹ and platelet-derived growth factor receptor alpha (PDGFRα) were identified in these tumors.¹⁰This prompted the development of tyrosine kinase inhibitors (TKIs), targeting these and other kinases, for the treatment of patients with GIST, and culminated in the approval of imatinib for this indication in 2002. This revolutionized the treatment of GIST, which had a poor prognosis and were resistant to chemotherapy, extending median overall survival in patients with metastatic disease almost to 5 years.^11-13

Imatinib was also shown to benefit patients with surgically resectable disease and was subsequently approved in the adjuvant setting in 2008. A recent trial demonstrated that 3-year continuation of adjuvant imatinib resulted in a significantly longer progression-free survival (PFS) compared with 1 year of adjuvant imatinib, and even longer time periods are now being evaluated.^14,15 The TKIs sunitinib and regorafenib have also been approved for the treatment of patients who become resistant to imatinib.^16,17 Avapritinib, a newer, more specific inhibitor of KIT is also being evaluated in patients with GIST (Table).

418_de Lartigue_NT sarc_T_web.PNG

Long-sought success for STS

Sunitinib and regorafenib include PDGFRα and the vascular endothelial growth factor receptors (VEGFRs) among their targets, receptors that play crucial roles in the formation of new blood vessels (angiogenesis). Many types of non-GIST sarcomas have been shown to be highly vascularized and express high levels of both of those receptors and other angiogenic proteins, which sparked interest in the development of multitargeted TKIs and other anti-angiogenic drugs in patients with STS.¹⁸

In 2012, pazopanib became the first FDA-approved molecularly targeted therapy for the treatment of non-GIST sarcomas. Approval in the second-line setting was based on the demonstration of a 3-month improvement in PFS compared with placebo.¹⁹ Four years later, the monoclonal antibody olaratumab, a more specific inhibitor of PDGFRα, was approved in combination with doxorubicin, marking the first front-line approval for more than 4 decades.²⁰Numerous other anti-angiogenic drugs continue to be evaluated for the treatment of advanced STS. Among them, anlotinib is being tested in phase 3 clinical trials, and results from the ALTER0203 trial were presented at the 2018 annual meeting of the American Society of Clinical Oncology (ASCO).²¹ After failure of chemotherapy, 223 patients were randomly assigned to receive either anlotinib or placebo. Anlotinib significantly improved median PFS across all patients, compared with placebo (6.27 vs 1.4 months, respectively; hazard ratio [HR], 0.33; P < .0001), but was especially effective in patients with alveolar soft part sarcoma (ASPS; mPFS: 18.2 vs 3 months) and was well tolerated.²¹

Sarcoma secrets revealed

Advancements in genome sequencing technologies have made it possible to interrogate the molecular underpinnings of sarcomas in greater detail. However, their rarity presents a significant technical challenge, with a dearth of samples available for genomic testing. Large-scale worldwide collaborative efforts have facilitated the collection of sufficiently large patient populations to provide statistically robust data in many cases. The Cancer Genome Atlas has established a rare tumor characterization project to facilitate the genomic sequencing of rare cancer types like sarcomas.

Genome sequencing studies have revealed 2 types of sarcomas: those with relatively stable genomes and few molecular alterations, exemplified by Ewing sarcoma, which has a mutational load of 0.15 mutations/Megabase (Mb); and those that are much more complex with frequent somatic mutations, the prime example being leiomyosarcoma. The latter are characterized by mutations in the TP53 gene, dubbed the “guardian of the genome” for its essential role in genome stability.

The 2 types are likely to require very different therapeutic strategies. Although genomically complex tumors offer up lots of potential targets for therapy, they also display significant heterogeneity and it can be challenging to find a shared target across different tumor samples. The p53 protein would make a logical target but, to date, tumor suppressor proteins are not readily druggable.

The most common type of molecular alterations in sarcomas are chromosomal translocations, where part of a chromosome breaks off and becomes reattached to another chromosome. This can result in the formation of a gene fusion when parts of 2 different genes are brought together in a way in which the genetic code can still be read, leading to the formation of a fusion protein with altered activity.^22-25

In sarcomas, these chromosomal translocations predominantly involve genes encoding transcription factors and the gene fusion results in their aberrant expression and activation of the transcriptional programs that they regulate.

Ewing sarcoma is a prime example of a sarcoma that is defined by chromosomal translocations. Most often, the resulting gene fusions occur between members of theten-eleven translocation (TET) family of RNA-binding proteins and the E26 transformation-specific (ETS) family of transcription factors. The most common fusion is between the EWSR1 and FLI1 genes, observed in between 85% and 90% of cases.

Significant efforts have been made to target EWSR1-FLI1. Since direct targeting of transcription factors is challenging, those efforts focused on targeting the aberrant transcriptional programs that they initiate. A major downstream target is the insulin-like growth factor receptor 1 (IGF1R) and numerous IGF1R inhibitors were developed and tested in patients with Ewing sarcoma, but unfortunately success was limited. Attention turned to the mammalian target of rapamycin (mTOR) as a potential mechanism of resistance to IGF1R inhibitors and explanation for the limited responses. Clinical trials combining mTOR and IGF1R inhibitors also proved unsuccessful.²⁶

Although overall these trials were deemed failures, they were notable for the dramatic responses that were seen in 1 or 2 patients. Researchers are probing these “exceptional responses” using novel N-of-1 clinical trial designs that focus on a single patient (Figure 2).^27-30 More recently, the first drug to specifically target the EWSR1-FLI1 fusion protein was developed. TK216 binds to the fusion protein and prevents it from binding to RNA helicase A, thereby blocking its function.³¹

Another type of gene fusion, involving the neurotrophic tropomyosin receptor kinase (NTRK) genes, has recently come into the spotlight for the treatment of lung cancer. According to a recent study, NTRK fusions may also be common in sarcomas. They were observed in 8% of patients with breast sarcomas, 5% with fibrosarcomas, and 5% with stomach or small intestine sarcomas.³²

The NTRK genes encode TRK proteins and several small molecule inhibitors of TRK have been developed to treat patients with NTRK fusion-positive cancers. Another novel clinical trial design – the basket trial – is being used to test these inhibitors. This type of trial uses a tumor-agnostic approach, recruiting patients with all different histological subtypes of cancer that are unified by the shared presence of a specific molecular alteration.³³

418_de Lartigue_NT sarc_F2_web.PNG

Repurposing gynecologic cancer drugs

More recently, a third group of sarcomas was categorized, with intermediate genomic complexity. These tumors, including well/dedifferentiated liposarcomas, were characterized by amplifications of chromosome 12, involving genes such as cyclin-dependent kinase 4 (CDK4). In fact, more than 90% of patients with well/dedifferentiated sarcomas display CDK4 amplification, making it a logical therapeutic target.³⁶

CDK4 encodes CDK4 protein, a cell cycle-associated protein that regulates the transition from G1-S phase, known as the restriction point, beyond which the cell commits to undergoing mitosis. Aberrant expression of CDK4 in cancer drives the hallmark process of unchecked cellular proliferation.

Some small molecule CDK4/6 inhibitors have been developed and have shown significant promise in the treatment of breast cancer. They are also being evaluatedin patients with sarcoma whose tumors display CDK4 overexpression. In a recently published phase 2 trial of palbociclib in 60 patients with well/dedifferentiated liposarcomas, there was 1 CR.³⁷

Another group of drugs that has advanced the treatment of gynecologic cancers comprises the poly (ADP-ribose) polymerase (PARP) inhibitors. In this context, PARP inhibitors are used in patients with mutations in the breast cancer susceptibility genes, BRCA1/2. The BRCA and PARP proteins are both involved in DNA repair pathways and the inhibition of PARP in patients who already have a defective BRCA pathway renders a lethal double blow to the cancer cell. According to the Broad Institute Cancer Cell Line Encyclopedia, Ewing sarcomas express high levels of the PARP1 enzyme, which could render them sensitive to PARP inhibition. Preclinical studies seemed to confirm that sensitivity, however, so far this has yet to translate into success in clinical trials, with no objective responses observed as yet.³⁸
 

Expanding the field

Other treatment strategies being tested in patients with sarcoma are moving the field beyond conventional targeted therapies. There has been substantial focus in recent years on epigenetic alterations and their potential role in the development of cancer. Epigenetics is the secondary layer of regulation that acts on the genome and directs the spatial and temporal expression of genes.

Both DNA and the histone proteins they are packaged up with to form chromatin in nondividing cells can be modified by the attachment of chemical groups, such as acetyl and methyl groups, which can alter access to the DNA for transcription.

EZH2 is an enzyme that participates in histone methylation and thereby regulates transcriptional repression. Some types of sarcoma are characterized by a loss of expression of the INI1 gene, also known as SMARCB1. The INI1 protein is part of a chromatin remodeling complex that relieves transcriptional repression and when INI1 is lost, cells become dependent upon EZH2.³⁹Clinical trials of the EZH2 inhibitor tazemetostat are ongoing in several types of sarcoma. Results from a phase 2 study in adults with INI1-negative tumors were presented at ASCO in 2017. Among 31 patients treated with 800 mg tazemetostat in continuous 28-day cycles, mPFS was 5.7 months, disease control rate was 10%, and confirmed overall response rate was 13%. The FDA has granted tazemetostat orphan drug designation in this indication.⁴⁰A pediatric basket trial of tazemetostat is also ongoing, but the FDA recently placed it under a clinical hold as a result of a safety update from the trial in which a pediatric patient with advanced poorly differentiated chordoma developed a secondary T-cell lymphoma.⁴¹

Targeting the unique metabolism of sarcomas may offer a promising therapeutic strategy, although this is in the preliminary stages of evaluation. A recent study showed that the expression of the argininosuccinate synthase 1 enzyme, which is involved in the generation of arginine through the urea cycle, was lost in up to 90% of STS. A pegylated arginine deaminase (ADI-PEG20), is being evaluated in a phase 2 clinical trial.⁴²

Finally, the concept of using immunotherapy to boost the anti-tumor immune response is also being examined in sarcomas. A significant number of cases of STS, osteosarcoma and GIST have been shown to express programmed cell death protein-ligand 1, therefore the use of immune checkpoint inhibitors that block this ligand or its receptor and help to reactive tumor-infiltrating T cells, could be a beneficial strategy.

Limited activity has been observed in studies conducted to date, however combination therapies, especially with inhibitors of the indoleamine 2,3-dioxygenase (IDO) enzyme, which plays a key role in immunosuppression, could help to harness the power of these drugs. Studies have suggested that sarcomas may be infiltrated by immunosuppressive macrophages that express IDO.⁴³

It is generally believed that immunotherapy is most effective in tumors that are highly mutated because that allows a large number of cancer antigens to provoke an anti-tumor immune response. However, a single highly expressed antigen can also be strongly immunogenic. Synovial sarcomas have a relatively low mutational burden but they do express high levels of the cancer testis antigen NY-ESO-1.

NY-ESO-1 has provided a useful target for the development of adoptive cell therapies and vaccines for the treatment of sarcomas. CMB305 is an NY-ESO-1 vaccine that also incorporates a toll-like receptor 4 agonist. It is being evaluated in the phase 3 Synovate study as maintenance monotherapy in patients with locally advanced, unresectable or metastatic synovial sarcoma. In a phase 1 study, at a median follow-up of just under 18 months, the median OS for all 25 patients was 23.7 months.⁴⁴

The rarity and complexities of bone and soft tissue sarcomas pose a major challenge to effective treatment. Historically, there has been a blanket approach to treatment, but more recently that has begun to change thanks to genome profiling studies and novel clinical trial strategies. Here, we discuss the resulting enrichment of the therapeutic armamentarium with molecularly targeted and immune therapies.

A challenging tumor type

Sarcomas are a large group of histologically diverse cancers that arise in the mesenchymal cells. They can be broadly divided into bone and soft tissue sarcomas (STS) but are further subdivided according to the type of cell from which they derive; osteosarcomas in the bone, rhabdomyosarcomas in the skeletal muscle, liposarcomas in the fat tissues, leiomyosarcomas in the smooth muscle, and chondrosarcomas in the cartilaginous tissue, for example.

Each sarcoma subtype itself encompasses a range of different cancers with unique biology. Under the umbrella of liposarcoma, for example, are well/dedifferentiated liposarcomas and myxoid liposarcomas, which have very different pathologies and clinical courses.

As a whole, sarcomas are extremely rare tumors, accounting for less than 1% of all adult cancers, although they disproportionately affect children and young adults, with a prevalence closer to 15%.^1,2 Certain sarcoma subtypes are exceptionally rare, with only a few cases diagnosed worldwide each year, whereas liposarcomas are at the other end of the spectrum, comprising the most common form of STS (Figure 1).³

418_de Lartigue_NT sarc_F1_web.PNG

Magic bullet for GIST

Despite their clear heterogeneity and complexity, sarcomas have tended to be treated as a single entity. Chemotherapy has played a central role in the treatment of advanced sarcomas and continues to do so, with 2 newer drugs approved by the United States Food and Drug Administration (FDA) in the past several years.^6,7

The development of targeted therapy, on the other hand, for the most part proved unsuccessful. In general, studies examining the somatic mutation landscape in sarcomas found very few that were highly recurrent. The exception was gastrointestinal stromal tumors (GIST), which represent around 8% of STS.⁸ Frequent mutations in several highly targetable tyrosine kinases, notably KIT, which is mutated in around 85% of cases,⁹ and platelet-derived growth factor receptor alpha (PDGFRα) were identified in these tumors.¹⁰This prompted the development of tyrosine kinase inhibitors (TKIs), targeting these and other kinases, for the treatment of patients with GIST, and culminated in the approval of imatinib for this indication in 2002. This revolutionized the treatment of GIST, which had a poor prognosis and were resistant to chemotherapy, extending median overall survival in patients with metastatic disease almost to 5 years.^11-13

Imatinib was also shown to benefit patients with surgically resectable disease and was subsequently approved in the adjuvant setting in 2008. A recent trial demonstrated that 3-year continuation of adjuvant imatinib resulted in a significantly longer progression-free survival (PFS) compared with 1 year of adjuvant imatinib, and even longer time periods are now being evaluated.^14,15 The TKIs sunitinib and regorafenib have also been approved for the treatment of patients who become resistant to imatinib.^16,17 Avapritinib, a newer, more specific inhibitor of KIT is also being evaluated in patients with GIST (Table).

418_de Lartigue_NT sarc_T_web.PNG

Long-sought success for STS

Sunitinib and regorafenib include PDGFRα and the vascular endothelial growth factor receptors (VEGFRs) among their targets, receptors that play crucial roles in the formation of new blood vessels (angiogenesis). Many types of non-GIST sarcomas have been shown to be highly vascularized and express high levels of both of those receptors and other angiogenic proteins, which sparked interest in the development of multitargeted TKIs and other anti-angiogenic drugs in patients with STS.¹⁸

In 2012, pazopanib became the first FDA-approved molecularly targeted therapy for the treatment of non-GIST sarcomas. Approval in the second-line setting was based on the demonstration of a 3-month improvement in PFS compared with placebo.¹⁹ Four years later, the monoclonal antibody olaratumab, a more specific inhibitor of PDGFRα, was approved in combination with doxorubicin, marking the first front-line approval for more than 4 decades.²⁰Numerous other anti-angiogenic drugs continue to be evaluated for the treatment of advanced STS. Among them, anlotinib is being tested in phase 3 clinical trials, and results from the ALTER0203 trial were presented at the 2018 annual meeting of the American Society of Clinical Oncology (ASCO).²¹ After failure of chemotherapy, 223 patients were randomly assigned to receive either anlotinib or placebo. Anlotinib significantly improved median PFS across all patients, compared with placebo (6.27 vs 1.4 months, respectively; hazard ratio [HR], 0.33; P < .0001), but was especially effective in patients with alveolar soft part sarcoma (ASPS; mPFS: 18.2 vs 3 months) and was well tolerated.²¹

Sarcoma secrets revealed

Advancements in genome sequencing technologies have made it possible to interrogate the molecular underpinnings of sarcomas in greater detail. However, their rarity presents a significant technical challenge, with a dearth of samples available for genomic testing. Large-scale worldwide collaborative efforts have facilitated the collection of sufficiently large patient populations to provide statistically robust data in many cases. The Cancer Genome Atlas has established a rare tumor characterization project to facilitate the genomic sequencing of rare cancer types like sarcomas.

Genome sequencing studies have revealed 2 types of sarcomas: those with relatively stable genomes and few molecular alterations, exemplified by Ewing sarcoma, which has a mutational load of 0.15 mutations/Megabase (Mb); and those that are much more complex with frequent somatic mutations, the prime example being leiomyosarcoma. The latter are characterized by mutations in the TP53 gene, dubbed the “guardian of the genome” for its essential role in genome stability.

The 2 types are likely to require very different therapeutic strategies. Although genomically complex tumors offer up lots of potential targets for therapy, they also display significant heterogeneity and it can be challenging to find a shared target across different tumor samples. The p53 protein would make a logical target but, to date, tumor suppressor proteins are not readily druggable.

The most common type of molecular alterations in sarcomas are chromosomal translocations, where part of a chromosome breaks off and becomes reattached to another chromosome. This can result in the formation of a gene fusion when parts of 2 different genes are brought together in a way in which the genetic code can still be read, leading to the formation of a fusion protein with altered activity.^22-25

In sarcomas, these chromosomal translocations predominantly involve genes encoding transcription factors and the gene fusion results in their aberrant expression and activation of the transcriptional programs that they regulate.

Ewing sarcoma is a prime example of a sarcoma that is defined by chromosomal translocations. Most often, the resulting gene fusions occur between members of theten-eleven translocation (TET) family of RNA-binding proteins and the E26 transformation-specific (ETS) family of transcription factors. The most common fusion is between the EWSR1 and FLI1 genes, observed in between 85% and 90% of cases.

Significant efforts have been made to target EWSR1-FLI1. Since direct targeting of transcription factors is challenging, those efforts focused on targeting the aberrant transcriptional programs that they initiate. A major downstream target is the insulin-like growth factor receptor 1 (IGF1R) and numerous IGF1R inhibitors were developed and tested in patients with Ewing sarcoma, but unfortunately success was limited. Attention turned to the mammalian target of rapamycin (mTOR) as a potential mechanism of resistance to IGF1R inhibitors and explanation for the limited responses. Clinical trials combining mTOR and IGF1R inhibitors also proved unsuccessful.²⁶

Although overall these trials were deemed failures, they were notable for the dramatic responses that were seen in 1 or 2 patients. Researchers are probing these “exceptional responses” using novel N-of-1 clinical trial designs that focus on a single patient (Figure 2).^27-30 More recently, the first drug to specifically target the EWSR1-FLI1 fusion protein was developed. TK216 binds to the fusion protein and prevents it from binding to RNA helicase A, thereby blocking its function.³¹

Another type of gene fusion, involving the neurotrophic tropomyosin receptor kinase (NTRK) genes, has recently come into the spotlight for the treatment of lung cancer. According to a recent study, NTRK fusions may also be common in sarcomas. They were observed in 8% of patients with breast sarcomas, 5% with fibrosarcomas, and 5% with stomach or small intestine sarcomas.³²

The NTRK genes encode TRK proteins and several small molecule inhibitors of TRK have been developed to treat patients with NTRK fusion-positive cancers. Another novel clinical trial design – the basket trial – is being used to test these inhibitors. This type of trial uses a tumor-agnostic approach, recruiting patients with all different histological subtypes of cancer that are unified by the shared presence of a specific molecular alteration.³³

418_de Lartigue_NT sarc_F2_web.PNG

Repurposing gynecologic cancer drugs

More recently, a third group of sarcomas was categorized, with intermediate genomic complexity. These tumors, including well/dedifferentiated liposarcomas, were characterized by amplifications of chromosome 12, involving genes such as cyclin-dependent kinase 4 (CDK4). In fact, more than 90% of patients with well/dedifferentiated sarcomas display CDK4 amplification, making it a logical therapeutic target.³⁶

CDK4 encodes CDK4 protein, a cell cycle-associated protein that regulates the transition from G1-S phase, known as the restriction point, beyond which the cell commits to undergoing mitosis. Aberrant expression of CDK4 in cancer drives the hallmark process of unchecked cellular proliferation.

Some small molecule CDK4/6 inhibitors have been developed and have shown significant promise in the treatment of breast cancer. They are also being evaluatedin patients with sarcoma whose tumors display CDK4 overexpression. In a recently published phase 2 trial of palbociclib in 60 patients with well/dedifferentiated liposarcomas, there was 1 CR.³⁷

Another group of drugs that has advanced the treatment of gynecologic cancers comprises the poly (ADP-ribose) polymerase (PARP) inhibitors. In this context, PARP inhibitors are used in patients with mutations in the breast cancer susceptibility genes, BRCA1/2. The BRCA and PARP proteins are both involved in DNA repair pathways and the inhibition of PARP in patients who already have a defective BRCA pathway renders a lethal double blow to the cancer cell. According to the Broad Institute Cancer Cell Line Encyclopedia, Ewing sarcomas express high levels of the PARP1 enzyme, which could render them sensitive to PARP inhibition. Preclinical studies seemed to confirm that sensitivity, however, so far this has yet to translate into success in clinical trials, with no objective responses observed as yet.³⁸
 

Expanding the field

Other treatment strategies being tested in patients with sarcoma are moving the field beyond conventional targeted therapies. There has been substantial focus in recent years on epigenetic alterations and their potential role in the development of cancer. Epigenetics is the secondary layer of regulation that acts on the genome and directs the spatial and temporal expression of genes.

Both DNA and the histone proteins they are packaged up with to form chromatin in nondividing cells can be modified by the attachment of chemical groups, such as acetyl and methyl groups, which can alter access to the DNA for transcription.

EZH2 is an enzyme that participates in histone methylation and thereby regulates transcriptional repression. Some types of sarcoma are characterized by a loss of expression of the INI1 gene, also known as SMARCB1. The INI1 protein is part of a chromatin remodeling complex that relieves transcriptional repression and when INI1 is lost, cells become dependent upon EZH2.³⁹Clinical trials of the EZH2 inhibitor tazemetostat are ongoing in several types of sarcoma. Results from a phase 2 study in adults with INI1-negative tumors were presented at ASCO in 2017. Among 31 patients treated with 800 mg tazemetostat in continuous 28-day cycles, mPFS was 5.7 months, disease control rate was 10%, and confirmed overall response rate was 13%. The FDA has granted tazemetostat orphan drug designation in this indication.⁴⁰A pediatric basket trial of tazemetostat is also ongoing, but the FDA recently placed it under a clinical hold as a result of a safety update from the trial in which a pediatric patient with advanced poorly differentiated chordoma developed a secondary T-cell lymphoma.⁴¹

Targeting the unique metabolism of sarcomas may offer a promising therapeutic strategy, although this is in the preliminary stages of evaluation. A recent study showed that the expression of the argininosuccinate synthase 1 enzyme, which is involved in the generation of arginine through the urea cycle, was lost in up to 90% of STS. A pegylated arginine deaminase (ADI-PEG20), is being evaluated in a phase 2 clinical trial.⁴²

Finally, the concept of using immunotherapy to boost the anti-tumor immune response is also being examined in sarcomas. A significant number of cases of STS, osteosarcoma and GIST have been shown to express programmed cell death protein-ligand 1, therefore the use of immune checkpoint inhibitors that block this ligand or its receptor and help to reactive tumor-infiltrating T cells, could be a beneficial strategy.

Limited activity has been observed in studies conducted to date, however combination therapies, especially with inhibitors of the indoleamine 2,3-dioxygenase (IDO) enzyme, which plays a key role in immunosuppression, could help to harness the power of these drugs. Studies have suggested that sarcomas may be infiltrated by immunosuppressive macrophages that express IDO.⁴³

It is generally believed that immunotherapy is most effective in tumors that are highly mutated because that allows a large number of cancer antigens to provoke an anti-tumor immune response. However, a single highly expressed antigen can also be strongly immunogenic. Synovial sarcomas have a relatively low mutational burden but they do express high levels of the cancer testis antigen NY-ESO-1.

NY-ESO-1 has provided a useful target for the development of adoptive cell therapies and vaccines for the treatment of sarcomas. CMB305 is an NY-ESO-1 vaccine that also incorporates a toll-like receptor 4 agonist. It is being evaluated in the phase 3 Synovate study as maintenance monotherapy in patients with locally advanced, unresectable or metastatic synovial sarcoma. In a phase 1 study, at a median follow-up of just under 18 months, the median OS for all 25 patients was 23.7 months.⁴⁴

The rarity and complexities of bone and soft tissue sarcomas pose a major challenge to effective treatment. Historically, there has been a blanket approach to treatment, but more recently that has begun to change thanks to genome profiling studies and novel clinical trial strategies. Here, we discuss the resulting enrichment of the therapeutic armamentarium with molecularly targeted and immune therapies.

A challenging tumor type

Sarcomas are a large group of histologically diverse cancers that arise in the mesenchymal cells. They can be broadly divided into bone and soft tissue sarcomas (STS) but are further subdivided according to the type of cell from which they derive; osteosarcomas in the bone, rhabdomyosarcomas in the skeletal muscle, liposarcomas in the fat tissues, leiomyosarcomas in the smooth muscle, and chondrosarcomas in the cartilaginous tissue, for example.

Each sarcoma subtype itself encompasses a range of different cancers with unique biology. Under the umbrella of liposarcoma, for example, are well/dedifferentiated liposarcomas and myxoid liposarcomas, which have very different pathologies and clinical courses.

As a whole, sarcomas are extremely rare tumors, accounting for less than 1% of all adult cancers, although they disproportionately affect children and young adults, with a prevalence closer to 15%.^1,2 Certain sarcoma subtypes are exceptionally rare, with only a few cases diagnosed worldwide each year, whereas liposarcomas are at the other end of the spectrum, comprising the most common form of STS (Figure 1).³

418_de Lartigue_NT sarc_F1_web.PNG

Magic bullet for GIST

Despite their clear heterogeneity and complexity, sarcomas have tended to be treated as a single entity. Chemotherapy has played a central role in the treatment of advanced sarcomas and continues to do so, with 2 newer drugs approved by the United States Food and Drug Administration (FDA) in the past several years.^6,7

The development of targeted therapy, on the other hand, for the most part proved unsuccessful. In general, studies examining the somatic mutation landscape in sarcomas found very few that were highly recurrent. The exception was gastrointestinal stromal tumors (GIST), which represent around 8% of STS.⁸ Frequent mutations in several highly targetable tyrosine kinases, notably KIT, which is mutated in around 85% of cases,⁹ and platelet-derived growth factor receptor alpha (PDGFRα) were identified in these tumors.¹⁰This prompted the development of tyrosine kinase inhibitors (TKIs), targeting these and other kinases, for the treatment of patients with GIST, and culminated in the approval of imatinib for this indication in 2002. This revolutionized the treatment of GIST, which had a poor prognosis and were resistant to chemotherapy, extending median overall survival in patients with metastatic disease almost to 5 years.^11-13

Imatinib was also shown to benefit patients with surgically resectable disease and was subsequently approved in the adjuvant setting in 2008. A recent trial demonstrated that 3-year continuation of adjuvant imatinib resulted in a significantly longer progression-free survival (PFS) compared with 1 year of adjuvant imatinib, and even longer time periods are now being evaluated.^14,15 The TKIs sunitinib and regorafenib have also been approved for the treatment of patients who become resistant to imatinib.^16,17 Avapritinib, a newer, more specific inhibitor of KIT is also being evaluated in patients with GIST (Table).

418_de Lartigue_NT sarc_T_web.PNG

Long-sought success for STS

Sunitinib and regorafenib include PDGFRα and the vascular endothelial growth factor receptors (VEGFRs) among their targets, receptors that play crucial roles in the formation of new blood vessels (angiogenesis). Many types of non-GIST sarcomas have been shown to be highly vascularized and express high levels of both of those receptors and other angiogenic proteins, which sparked interest in the development of multitargeted TKIs and other anti-angiogenic drugs in patients with STS.¹⁸

In 2012, pazopanib became the first FDA-approved molecularly targeted therapy for the treatment of non-GIST sarcomas. Approval in the second-line setting was based on the demonstration of a 3-month improvement in PFS compared with placebo.¹⁹ Four years later, the monoclonal antibody olaratumab, a more specific inhibitor of PDGFRα, was approved in combination with doxorubicin, marking the first front-line approval for more than 4 decades.²⁰Numerous other anti-angiogenic drugs continue to be evaluated for the treatment of advanced STS. Among them, anlotinib is being tested in phase 3 clinical trials, and results from the ALTER0203 trial were presented at the 2018 annual meeting of the American Society of Clinical Oncology (ASCO).²¹ After failure of chemotherapy, 223 patients were randomly assigned to receive either anlotinib or placebo. Anlotinib significantly improved median PFS across all patients, compared with placebo (6.27 vs 1.4 months, respectively; hazard ratio [HR], 0.33; P < .0001), but was especially effective in patients with alveolar soft part sarcoma (ASPS; mPFS: 18.2 vs 3 months) and was well tolerated.²¹

Sarcoma secrets revealed

Advancements in genome sequencing technologies have made it possible to interrogate the molecular underpinnings of sarcomas in greater detail. However, their rarity presents a significant technical challenge, with a dearth of samples available for genomic testing. Large-scale worldwide collaborative efforts have facilitated the collection of sufficiently large patient populations to provide statistically robust data in many cases. The Cancer Genome Atlas has established a rare tumor characterization project to facilitate the genomic sequencing of rare cancer types like sarcomas.

Genome sequencing studies have revealed 2 types of sarcomas: those with relatively stable genomes and few molecular alterations, exemplified by Ewing sarcoma, which has a mutational load of 0.15 mutations/Megabase (Mb); and those that are much more complex with frequent somatic mutations, the prime example being leiomyosarcoma. The latter are characterized by mutations in the TP53 gene, dubbed the “guardian of the genome” for its essential role in genome stability.

The 2 types are likely to require very different therapeutic strategies. Although genomically complex tumors offer up lots of potential targets for therapy, they also display significant heterogeneity and it can be challenging to find a shared target across different tumor samples. The p53 protein would make a logical target but, to date, tumor suppressor proteins are not readily druggable.

The most common type of molecular alterations in sarcomas are chromosomal translocations, where part of a chromosome breaks off and becomes reattached to another chromosome. This can result in the formation of a gene fusion when parts of 2 different genes are brought together in a way in which the genetic code can still be read, leading to the formation of a fusion protein with altered activity.^22-25

In sarcomas, these chromosomal translocations predominantly involve genes encoding transcription factors and the gene fusion results in their aberrant expression and activation of the transcriptional programs that they regulate.

Ewing sarcoma is a prime example of a sarcoma that is defined by chromosomal translocations. Most often, the resulting gene fusions occur between members of theten-eleven translocation (TET) family of RNA-binding proteins and the E26 transformation-specific (ETS) family of transcription factors. The most common fusion is between the EWSR1 and FLI1 genes, observed in between 85% and 90% of cases.

Significant efforts have been made to target EWSR1-FLI1. Since direct targeting of transcription factors is challenging, those efforts focused on targeting the aberrant transcriptional programs that they initiate. A major downstream target is the insulin-like growth factor receptor 1 (IGF1R) and numerous IGF1R inhibitors were developed and tested in patients with Ewing sarcoma, but unfortunately success was limited. Attention turned to the mammalian target of rapamycin (mTOR) as a potential mechanism of resistance to IGF1R inhibitors and explanation for the limited responses. Clinical trials combining mTOR and IGF1R inhibitors also proved unsuccessful.²⁶

Although overall these trials were deemed failures, they were notable for the dramatic responses that were seen in 1 or 2 patients. Researchers are probing these “exceptional responses” using novel N-of-1 clinical trial designs that focus on a single patient (Figure 2).^27-30 More recently, the first drug to specifically target the EWSR1-FLI1 fusion protein was developed. TK216 binds to the fusion protein and prevents it from binding to RNA helicase A, thereby blocking its function.³¹

Another type of gene fusion, involving the neurotrophic tropomyosin receptor kinase (NTRK) genes, has recently come into the spotlight for the treatment of lung cancer. According to a recent study, NTRK fusions may also be common in sarcomas. They were observed in 8% of patients with breast sarcomas, 5% with fibrosarcomas, and 5% with stomach or small intestine sarcomas.³²

The NTRK genes encode TRK proteins and several small molecule inhibitors of TRK have been developed to treat patients with NTRK fusion-positive cancers. Another novel clinical trial design – the basket trial – is being used to test these inhibitors. This type of trial uses a tumor-agnostic approach, recruiting patients with all different histological subtypes of cancer that are unified by the shared presence of a specific molecular alteration.³³

418_de Lartigue_NT sarc_F2_web.PNG

Repurposing gynecologic cancer drugs

More recently, a third group of sarcomas was categorized, with intermediate genomic complexity. These tumors, including well/dedifferentiated liposarcomas, were characterized by amplifications of chromosome 12, involving genes such as cyclin-dependent kinase 4 (CDK4). In fact, more than 90% of patients with well/dedifferentiated sarcomas display CDK4 amplification, making it a logical therapeutic target.³⁶

CDK4 encodes CDK4 protein, a cell cycle-associated protein that regulates the transition from G1-S phase, known as the restriction point, beyond which the cell commits to undergoing mitosis. Aberrant expression of CDK4 in cancer drives the hallmark process of unchecked cellular proliferation.

Some small molecule CDK4/6 inhibitors have been developed and have shown significant promise in the treatment of breast cancer. They are also being evaluatedin patients with sarcoma whose tumors display CDK4 overexpression. In a recently published phase 2 trial of palbociclib in 60 patients with well/dedifferentiated liposarcomas, there was 1 CR.³⁷

Another group of drugs that has advanced the treatment of gynecologic cancers comprises the poly (ADP-ribose) polymerase (PARP) inhibitors. In this context, PARP inhibitors are used in patients with mutations in the breast cancer susceptibility genes, BRCA1/2. The BRCA and PARP proteins are both involved in DNA repair pathways and the inhibition of PARP in patients who already have a defective BRCA pathway renders a lethal double blow to the cancer cell. According to the Broad Institute Cancer Cell Line Encyclopedia, Ewing sarcomas express high levels of the PARP1 enzyme, which could render them sensitive to PARP inhibition. Preclinical studies seemed to confirm that sensitivity, however, so far this has yet to translate into success in clinical trials, with no objective responses observed as yet.³⁸
 

Expanding the field

Other treatment strategies being tested in patients with sarcoma are moving the field beyond conventional targeted therapies. There has been substantial focus in recent years on epigenetic alterations and their potential role in the development of cancer. Epigenetics is the secondary layer of regulation that acts on the genome and directs the spatial and temporal expression of genes.

Both DNA and the histone proteins they are packaged up with to form chromatin in nondividing cells can be modified by the attachment of chemical groups, such as acetyl and methyl groups, which can alter access to the DNA for transcription.

EZH2 is an enzyme that participates in histone methylation and thereby regulates transcriptional repression. Some types of sarcoma are characterized by a loss of expression of the INI1 gene, also known as SMARCB1. The INI1 protein is part of a chromatin remodeling complex that relieves transcriptional repression and when INI1 is lost, cells become dependent upon EZH2.³⁹Clinical trials of the EZH2 inhibitor tazemetostat are ongoing in several types of sarcoma. Results from a phase 2 study in adults with INI1-negative tumors were presented at ASCO in 2017. Among 31 patients treated with 800 mg tazemetostat in continuous 28-day cycles, mPFS was 5.7 months, disease control rate was 10%, and confirmed overall response rate was 13%. The FDA has granted tazemetostat orphan drug designation in this indication.⁴⁰A pediatric basket trial of tazemetostat is also ongoing, but the FDA recently placed it under a clinical hold as a result of a safety update from the trial in which a pediatric patient with advanced poorly differentiated chordoma developed a secondary T-cell lymphoma.⁴¹

Targeting the unique metabolism of sarcomas may offer a promising therapeutic strategy, although this is in the preliminary stages of evaluation. A recent study showed that the expression of the argininosuccinate synthase 1 enzyme, which is involved in the generation of arginine through the urea cycle, was lost in up to 90% of STS. A pegylated arginine deaminase (ADI-PEG20), is being evaluated in a phase 2 clinical trial.⁴²

Finally, the concept of using immunotherapy to boost the anti-tumor immune response is also being examined in sarcomas. A significant number of cases of STS, osteosarcoma and GIST have been shown to express programmed cell death protein-ligand 1, therefore the use of immune checkpoint inhibitors that block this ligand or its receptor and help to reactive tumor-infiltrating T cells, could be a beneficial strategy.

Limited activity has been observed in studies conducted to date, however combination therapies, especially with inhibitors of the indoleamine 2,3-dioxygenase (IDO) enzyme, which plays a key role in immunosuppression, could help to harness the power of these drugs. Studies have suggested that sarcomas may be infiltrated by immunosuppressive macrophages that express IDO.⁴³

It is generally believed that immunotherapy is most effective in tumors that are highly mutated because that allows a large number of cancer antigens to provoke an anti-tumor immune response. However, a single highly expressed antigen can also be strongly immunogenic. Synovial sarcomas have a relatively low mutational burden but they do express high levels of the cancer testis antigen NY-ESO-1.

NY-ESO-1 has provided a useful target for the development of adoptive cell therapies and vaccines for the treatment of sarcomas. CMB305 is an NY-ESO-1 vaccine that also incorporates a toll-like receptor 4 agonist. It is being evaluated in the phase 3 Synovate study as maintenance monotherapy in patients with locally advanced, unresectable or metastatic synovial sarcoma. In a phase 1 study, at a median follow-up of just under 18 months, the median OS for all 25 patients was 23.7 months.⁴⁴