Impact of trial design and patient heterogeneity on the identification of clinically effective therapies for progressive MS

Abstract

Clinically effective immunomodulatory therapies have been developed for relapsing-remitting multiple sclerosis (RRMS), but they have generally not translated to a corresponding slowing of disability accumulation in progressive forms of MS. Since disability is multifaceted, progressive patients are heterogeneous, and the drivers of disease progression are still unclear, it has been difficult to identify the most informative outcome measures for progressive trials. Historically, secondary outcome measures have focused on inflammatory measures, which contributed to the recent identification of immunomodulatory therapies benefiting younger patients with more inflammatory progressive MS. Meanwhile, agents capable of treating late-stage disease have remained elusive. Consequently, measures of neurodegeneration are becoming common. Here, we review completed clinical trials testing immunomodulatory therapies in primary progressive multiple sclerosis (PPMS) or secondary progressive multiple sclerosis (SPMS) and discuss the features contributing to trial design variability in relation to trial outcomes, and how efforts toward better patient stratification and inclusion of reliable progression markers could improve outcomes.

Introduction

In contrast to the success of immunomodulatory disease modifying therapies (DMTs) in the treatment of relapsing-remitting multiple sclerosis (RRMS), most clinical trials have failed to demonstrate efficacy for these DMTs in slowing disease progression in primary progressive MS (PPMS) or secondary progressive MS (SPMS)^{1, 2}. Part of this discrepancy stems from the increased understanding of the etiology and standardized methods to measure relapses, which have remained largely elusive for progression. Consequently, the repurposing of RRMS DMTs and trial design methods toward progressive MS (PMS) may produce misleading results. Indeed, the same or highly similar DMTs have demonstrated different results in different PMS trials ². Since the progressive patient population is heterogeneous in terms of level of inflammatory activity, disease duration, and disability, patient inclusion criteria can also potentially influence trial outcomes ³. This variability then makes it difficult to determine whether the reported efficacy in positive trials is due to trial design or actual targeting of the drivers of PMS. In response to these challenges, there has been increased focus on the development and validation of biological correlates of progression in MS, and Phase 2 trials for prospective DMTs are shifting toward outcome measures of neurodegeneration ^{1, 2}. We discuss the factors contributing to trial variability, how these factors may have influenced conflicting trial outcomes, and some of the efforts toward facilitating better outcomes.

Patient characteristics

PMS is characterized by a steady accumulation of disability, however, some progressive patients also experience inflammation associated relapses evidenced by gadolinium (Gd+) enhancing lesions on magnetic resonance imaging (MRI) ⁴. Those with ‘active’ PPMS tend to be younger with shorter disease duration, while relapses in SPMS are most common during the transition phase from RRMS to SPMS. In the completed studies, the active PMS patients were the most responsive to the immunomodulatory DMTs ⁵, thus studies with a higher proportion of these patients were more likely to be positive (Table 1). This suggests that immunomodulatory therapies can be efficacious in early stages of the disease, but beyond a certain age or threshold of disability accumulation, different types of therapies may be required. To identify these therapies, future trials will need to better stratify cohorts by levels of inflammation, disability, gender, and age.

Table 1.

Baseline Demographic and Disease Characteristics for Phase 3 Progressive Multiple Sclerosis Clinical Trials.^*

	European (1998) 43	North American (2004)44	PROMiSe (2002)83	OLYMPUS (2009)48	ORATORIO (2017)23	ASCEND: Part 1 (2016)84	INFORMS (2016)55	EXPAND (2017)56
Type of MS	SPMS	SPMS	PPMS	PPMS	PPMS	SPMS	PPMS	SPMS
DMT	Interferon beta-1b	Interferon beta-1b	Glatiramer Acetate	Rituximab	Ocrelizumab	Natalizumab	Fingolimod	Siponimod
Phase	3	3	3	2–3	3	3	3	3
Total (n) Treated (Placebo)	718 360 (358)	939 631 (308)	943 627 (316)	439 292 (147)	732 488 (244)	887 439 (448)	823 336 (487)	1651 1105 (456)
Male %**	58.1%	36.1%	47.2%	52.1%	51.4%	38.0%	51.0%	39%
Age Years (mean)	40.9±7.2	46.5±0.46	50.4±8.4	50.1±9.0	44.7±7.9	47.3±7.4	48.5±8.6	48.0±7.8
Time since onset MS** Mean (Years)	13.4±7.5	14.5±0.47	11.0±7.3	9.2±6.4	6.7±4.0	16.8±7.6	5.8±2.5	17.1±8.4
Time since diagnosis PPMS or SPMS Mean (Years)**	2.1±2.2	4.0±0.2	5.0±4.9	4.1±4.2	2.9±3.2	4.7±3.0	2.8±2.6	3.9±3.6
% Relapse-Free (Prior 2-year)	28.2%	54.3%	****	****	****	71%	****	64.0%
EDSS (mean)**	5.1±1.1	5.2±0.07	4.9±1.2	4.8±1.4	4.7±1.2	6.0 (5.0–6.5) Median	4.7±1.03	5.4±1.1
T1 Gd lesions (%)**	****	****	14.1% (Yes) 85.9% (No)	24.5% (Yes) 75.5% (No)	27.5% (Yes) 72.5% (No)	26% (Yes) 74% (No)	14% (Yes) 86% (No)	21% (Yes) 75% (No) 3% (Not assessed)
T2 Lesion Volume (mean)**	****	3831.5±222 mm2 (lesion area)	8.38±10.14 cm3	9,336.66±13,744.94 mm3	12.7±15.1 cm3	17.4±17.6 cm3	9,442.7±10,179.7 mm3	15,632±16,268 mm3
Normalized brain volume (cm3)**	****	****	****	1,202.92±120.23	1462.9±84.0	1420.9±82.8	1490.9±86.5	1,422±86
T25-FW seconds (mean)**	****	****	12.9	8.35	****	11.2 (Median)	9.05	17.4
Outcome	Positive	Negative	Negative	Negative	Positive	Negative	Negative	Positive
NCT Identifier	N/A	N/A	N/A	NCT02253264	NCT01194570	NCT01416181	NCT00731692	NCT01665144

^*Plus-minus values are means ±SD. MS multiple sclerosis, PPMS primary progressive multiple sclerosis, and SPMS secondary progressive multiple sclerosis.

^**Shown are data for patients in the treatment arm (Baseline demographic and disease characteristics for patients in the placebo arm were not included).

^****Data was not reported for this study.

Progressive trials with a higher proportion of younger patients, with shorter disease duration, with lower percentage of relapse-free patients, higher percentage of Gd+ lesions, and less disability (lower EDSS) are more likely to have positive results due to reduction of inflammatory disease activity by immunomodulatory DMTs.

Aging itself may be a critical driver of progression ^{6, 7}, particularly the aging of the immune system. The onset of PPMS is usually later than RRMS, such that after age 50 individuals are more likely to be diagnosed with PPMS ⁸. Since SPMS occurs after a period of RRMS that can last 20 years or longer, the transition to SPMS is most common as patients reach their mid-50s ⁹. This corresponds to the time when the body starts showing signs of immunosenesence and demonstrates less robust adaptive inflammatory responses ¹⁰, consistent with ‘inactive’ forms of PMS. Although the incidence of MS is higher in women, the gender disparity is greater for RRMS than PPMS, and progression is often more aggressive in men ^{11, 12}. This may be attributable to the slower rate of immunological decline in women, which has also been suggested to contribute to longevity ¹³. However, aging involves a constellation of systematic changes, including decreased myelination and repair capacity in the CNS, all of which may influence progression⁶. The age-related accumulation of CNS iron is of particular note, since iron is a feature of MS lesion formation ¹⁴, and the oxidation of Fe²⁺ to Fe³⁺ is a potent inducer of neurodegenerative oxidative stress ¹⁵. Consequently, the screening of patients for signs of immunosenescence, or iron dysregulation could facilitate the testing of more sub-population targeted DMTs.

Drivers of disease progression

The inflammatory drivers of relapses are relatively well-understood, which has led to the development of several DMTs highly effective at preventing relapses ¹⁶. In contrast to relapses, progression is generally characterized by diffuse neurodegeneration not obviously associated with specific inflammatory activity, and the underlying mechanisms are less clear ¹⁷. Demyelination itself can cause axons to lose trophic support, and thus induce neurodegeneration¹⁸, but the lack of translation by highly effective RRMS DMTs suggests that the axonal pathology in PMS may be driven by distinct mechanisms from relapses. Cortical pathology appears to play a prominent role¹⁹, and is associated with the presence of meningeal inflammation and ectopic lymphoid follicles^{20, 21}. While RRMS has been described as a T cell driven disease, these observations suggest an enhanced role for B cells in progression²². This may explain the recent success of the B cell depleting therapy, ocrelizumab, in a trial for PPMS²³, however, it remains unclear how well it actually targets the CNS resident B cell follicles. Progression may also involve a shift from the adaptive to the innate immune system, as microglial activation is associated with diffuse neuropathology in PMS²⁴. Non-inflammatory mediated mechanisms of axonal pathology are also possible, as synaptic alterations, mitochondrial dysfunction, and defects in axon transport have been reported in PMS ²⁵. Irrespective of the cause, treatments for advanced PMS are likely to require neuroprotective activity.

Measures of disease progression

Unlike annualized relapse rate, which is an easily quantifiable efficacy metric for RRMS, the assessment of disability progression is more subjective and time intensive. A change in the Expanded Disability Status Scale (EDSS) score is the accepted standard primary outcome measure in Phase 3 progressive trials ²⁶. However, disability is multifaceted including both physical and cognitive impairment, but EDSS is heavily weighted on physical disability, and becomes increasingly insensitive to small changes in progression as disability increases ²⁷. Therefore, newer trials are also incorporating the more quantifiable features of the MSFC ²⁶, such as the timed 25’ walk test (T25FW) and 9 hole peg test (9-HPT), as well as the symbol digit modality test, which has been demonstrated to be a highly sensitive and easily performed test of cognitive function in MS patients ²⁸ (Table 2). Since these measures are not yet accepted or standardized by regulatory agencies, it can be difficult to prioritize or reconcile disparate results across the measures (Table 2), however, new combinatorial measures, such as CombiWISE ²⁹, have the potential to overcome the biases in any particular test. Furthermore, outcome measures may also need to be tailored to best correspond to the disease course in specific sub-populations of progressive patients.

Table 2.

Primary and Secondary Outcome Measures for Progressive MS Trials

STUDY (YEAR)	MS TYPE	Primary Outcome	Outcome met?	Secondary Outcomes	Outcome met?	Comments
Interferon-β EUROPEAN43 (1998)	SPMS	Time to confirmed progression in disability (CDP) as measured by a 1.0 point increase on the Expanded Disability Status Scale (EDSS), sustained for at least 3 months, or a 0.5 point increase if the baseline EDSS was 6.0 or 6.5 (36 months)	YES	Time to becoming wheelchair bound	YES	Benefits primarily due to reduction in inflammatory relapse related activity.
				Proportion of patients becoming wheelchair bound	YES
				Annual relapse rate	YES
				Percent change in annual T2 lesion volume	YES
				Number newly active lesions	YES
				Proportion of patients with CDP	YES
				Change in EDSS from baseline	YES
				EDSS at endpoint	NO
				Time to first relapse	YES
				Proportion of patients with moderate/severe relapse	YES
Interferon -β NORTH AMERICAN44 (2004)	SPMS	Time to CDP by increase of ≥1.0 point from the baseline EDSS score (≥0.5 point if the baseline EDSS score was 6.0 to 6.5) confirmed at 2 consecutive schedule exams spanning ≥ 6 months from onset of progression (36 months)	NO	Change in EDSS score from baseline	NO	Effective for reducing inflammatory relapse activity, but not progression.
				Annual relapse rate	YES
				Change in T2 lesion area	YES
				Active (new) lesion rate	YES
				Change in composite neuropsychological function score (Rao Brief Repeatable Battery)	NO
Glatiramer Acetate (PROMiSe)83 (2002)	PPMS	Time to CDP. An EDSS change of ≥1 points sustained for 3 months in patients with an EDSS score at baseline of 3–5, or ≥0.5 EDSS point sustained for 3 months in patients with a baseline EDSS score of 5.5–6.5 over the 36-month study.	NO	Proportion of patients without progression	NO	Minor reduction in inflammation, otherwise ineffective for progressive MS across all measures.
				Change in EDSS from baseline	NO
				Change in MSFC from baseline	NO
				Volume of T2 lesions	YES (partial)
				Number Gd+ T1 lesions	YES (partial)
				Number hypointense T1 lesions	NO
				Change in brain volume	NO
Rituximab (OLYMPUS)48 (2009)	PPMS	Time to CDP. Sustained EDSS increase of ≥1.0 point from baseline EDSS if the baseline EDSS was between 2.0–5.5 points or an EDSS increase of ≥0.5 points if the baseline EDSS was >5.5 points sustained for ≥12 weeks. (96 weeks)	NO	Change in T2 lesion volume from baseline	YES	Effective for reducing inflammatory relapse activity, but not progression. Possible benefit for younger patients with active PPMS
				Change in brain volume from baseline	NO
				Change in EDSS from baseline	NO
				Change in Multiple Sclerosis Functional Composite score (MSFC) from baseline	NO
				Change in Timed 25’ Walk (T25FW) from baseline	YES
Ocrelizumab (ORATORIO)23 (2017)	PPMS	An increase in the EDSS of at least 1.0 point from baseline that was sustained on subsequent visits for at least 12 weeks if the baseline score was 5.5 or less or an increase of at least 0.5 points that was sustained for at least 12 weeks if the baseline was more than 5.5. (120 weeks)	YES	CDP ≥ 24 weeks	YES	Primary effect on reducing inflammatory activity with minor effect on progression. Approved for PPMS 2017. Most effective for ‘active’ PPMS, benefit for ‘inactive’ PPMS requires further study.
				Change in T25FW from baseline	YES
				Change in T2 lesion volume from baseline	YES
				Percent change in brain volume (week 24120)	YES
				Change in Physical Component Summary core from baseline	NO
Natalizumab (ASCEND)84 (2016)	SPMS	Percentage of participants with confirmed progression of disability in 1+ of the EDSS, T25FW, or 9-HPT. (96 weeks)	NO	T25FW responders	NO	Used more comprehensive composite primary endpoint. Effective for reducing inflammatory relapse activity, but not progression.
				Multiple Sclerosis Walking Scale (MSWS-12) Change	NO
				ABILHAND Change	NO
				Multiple Sclerosis Impact Scale 29 (MSIS-29) Change	NO
				Change in brain volume	NO
				Confirmed progression by EDSS functional subscore	NO
				Annual relapse rate	YES
				Number of Gd+ lesions	YES
				Number new/enlarging T2 lesions	YES
				Volume T2 lesions	YES
				Confirmed progressors Symbol digit modalities test (SDMT)	NO
Fingolimod (INFORMS)55 (2016)	PPMS	Time to CDP by increase from baseline EDSS score by 1 point (≥0.5 point if the baseline EDSS is score is ≥5.5), increase of ≥ 20% from baseline in T25FW, or ≥ 20% from baseline in 9 hole peg test (9-HPT). Had to be confirmed for the same component at least 3 months later. (36 months)	NO	Percentage change in brain volume	NO	Used more comprehensive composite primary endpoint. Effective for reducing inflammatory relapse activity, but not progression.
				Number of new/newly enlarging T2 lesions	YES
				Number Gd+ T1 lesions	YES
				Number patients without Gd+ T1 lesions	YES
				Number new hypointense T1 lesions	YES
				Number of patients without hypointense T1 lesions	YES
Siponimod (EXPAND)56 (2017)	SPMS	Time to CDP. An Increase of score of 1 point in patients with baseline score of 3.0 to 5.0 and 0.5 point increase with baseline score of 5.5 to 6.5. (24 months)	YES	Time to confirmed worsening ≥20% from baseline T25FW	NO	Primary effect on inflammatory activity with modest effect on progression. Expected to be approved for SPMS. Possible neuroprotective activity requires further study.
				Time to 6 months CDP	YES
				Annual relapse rate	YES
				Time to first relapse	YES
				Change in score on MSWS	NO
				Change in T2 lesion volume from baseline	YES
				Number new/enlarging T2 lesions	YES
				Number of Gd+ T1 lesions	YES
				Percent change in brain volume from baseline	YES
MD1003 (MS-SPI)68 (2016)	PPMS -OR- SPMS	Proportion of patients with improvement of MS-related disability at month 9, confirmed at month 12. A decrease of ≥0.5 point or ≥1 point in EDSS (if baseline score was 6–7 or 4.5–5.5, respectively) or a ≥20% decrease in T25FW time, compared with the best recorded EDSS or T25FW value. (12 months)	YES	Change in EDSS from baseline	NO	Possible benefit for progression, but short study and inconsistencies in outcome measures. Expected to be resolved by larger Phase 3 study (MS-SPI2).
				Change in Clinical Global Impression scale	NO
				Change in MSWS	NO
				Short-form 36 Health Survey subscores	NO
				Change in Modified Fatigue Impact scale	NO
				Change in 9-HPT	NO
				Change in T25FW	NO
				Change in EDSS functional subscores	NO
Ibudilast (SPRINT-MS)67, 85 (2018)	PPMS -OR- SPMS	Change in whole brain atrophy as measured by Brain Parenchymal Fraction (BPF) over 96 weeks.	YES	Change in magnetization transfer ratio (MTR) for normal brain tissue	YES	Possible benefit for progression, but limited to imaging measures so more comprehensive clinical outcome measures assessment needed.
				Change in MTR for normal gray matter	YES
				Change in white matter by diffusion tensor imaging (DTI)	NO
				Retinal fiber thickness (RLFL) measured by Optical coherence tomography (OCT)	N/A
				Change in cortical atrophy	N/A
Lipoic Acid (2017)76	SPMS	Annualized rate of whole brain atrophy as measured by percent change in brain volume.	YES	Change in total deep grey matter volume	NO	Possible benefit for progression/ neuroprotection. Inconsistencies in outcome measures, but lacked power to detect clinical outcomes. Larger study cohort needed (planned).
				Change in cortical thickness	NO
				Change in T2 lesion volume	NO
				Change in C1 cross sectional area	NO
				Change in cervical spinal cord lesion occupancy	NO
				Change in average RNFL thickness	NO
				Change in average retinal ganglion cell inner plexiform layer (GCIPL) thickness	NO
				Change in SDMT measure, number correct	NO
				Change in T25FW	NO (trend)
				Change in EDSS	NO
				Change in MSWS-12	NO
				Change in RAND 36-Item Short Form Health Survey	NO
				Change in Activities of Balance (ABC)	NO
Simvastatin (MS-STAT) (2014)81	SPMS	Rate of whole brain atrophy per year as measured by percent change in brain volume at 12 and 24 months.	YES	Change in EDSS	YES	Possible benefit for progression/ Neuroprotection. May be better for inactive than active SPMS, but inconsistencies in outcomes measures suggest possible small sample size bias, to be addressed by Phase 3 study (MS-STAT2).
				Change in MSFC: total, physical, and psychological scores	YES (total and physical)
				Change in MSIS-29: z-score, walk, peg-test, and PASAT scores	NO
				Rate of new and enlarging T2 lesions	NO
				Rate of relapses over 24 months	NO

The field is working toward the use of fluid based biomarkers to monitor progression, which is expected to be more time and cost effective than current methods, and to improve patient stratification ³⁰. The elevation of certain factors in CSF and/or serum, such as neurofilament light ³¹ and GFAP ³², has been associated with neurodegeneration. However, these factors are typically identified through cross-sectional studies, and no single factor has been found to be a reliable indicator of progression at the individual level. In the meantime, MRI measures of neurodegeneration, such as whole brain atrophy, are increasingly incorporated into progressive trials ³³ (Table 2). But since these methods of measuring brain volume detect very small units of change, they are subject to confounders if not carefully controlled for age, gender, and level of inflammation. Neuroprotective activity could potentially be obscured by anti-inflammation associated ‘pseudoatrophy’ ³⁴, and while cortical thickness is considered more sensitive ³⁵, this measure can be confounded by its relation to age ³⁶ and intracranial volume ³⁷. Optical coherence tomography measures of optic nerve thickness are beginning to be used in trials, as they are reliable correlates of brain degeneration in MS ³⁸, and easier to perform and standardize than MRI measures ³⁹.

Ideally, it would be most useful to have measures related to drivers of progression which could be detected early in the disease course and predict future progression. Recent MRI advances have allowed detection of the differential distribution of brain iron in RRMS compared to PMS ⁴⁰. Furthermore, the monitoring of iron levels could potentially be a tool for predicting progression, as decreased thalamic iron levels has been associated with faster progression^{41, 42}.

Trial comparisons

As demonstrated by the interferon-beta trials^{43, 44}, trial design differences, such as patient selection, can affect trial success³. Consequently, the different results in trials for similar drugs in the anti-CD20 and S1P modulator families raises the concern whether this stems from differences in biological activity or trial design.

Anti-CD20 therapy: rituximab and ocrelizumab

Anti-CD20 therapy primarily targets B cells and is very effective for reducing relapse rate in RRMS ⁴⁵. The depletion of B cells is thought to impact the activation status and function of the remaining T cells ^{46, 47}. The chimeric anti-CD20 monoclonal antibody rituximab failed to decrease 3 month confirmed disability progression (CDP) assessed by a change in EDSS in the PPMS OLYMPUS trial, but post-hoc analysis revealed that there was a small benefit to a subpopulation of active early stage PPMS patients that were younger and had more Gd+ enhancing lesions ⁴⁸. Ocrelizumab, a second generation humanized anti-CD20 therapy, was engineered to be slightly more effective at activating cell death in CD20 expressing cells. While the patient cohorts in the two trials were comparable in terms of EDSS and Gd+ enhancing lesions, the ORATORIO trial for ocrelizumab had a higher percentage of the younger PPMS patients with shorter disease duration ²³. Ocrelizumab was found to decrease 3 and 6 month CDP, slow the performance decline in the T25FW, and decrease brain volume loss (BVL) compared to placebo ²³, which led to its recent approval for PPMS. However, based on the OLYMPUS trial, it appears that rituximab may have similar efficacy for active early stage PPMS, and it remains unclear if anti-CD20 therapy actually slows progression for older and late stage PPMS patients, or if its efficacy is more limited to a particular sub-population which was overrepresented in the ORATORIO trial.

S1P modulators: fingolimod and siponimod

The functional antagonism of S1PR1 reduces the level of circulating pro-inflammatory lymphocytes⁴⁹, which is believed to contribute to the efficacy for S1P modulators in RRMS⁵⁰. Fingolimod was shown to decrease brain volume loss in RRMS ^{51, 52} and in vitro studies indicated neuroprotective potential ^{53, 54}, suggesting it may also benefit PMS. The INFORMS trial tested the efficacy of fingolimod in PPMS, but failed to meet its primary endpoint of a change in 3 month CDP compared to baseline or lead to a reduction in BVL ⁵⁵. Notably, this trial used composite assessment of disability progression by also incorporating the T25FW and the 9HPT. Similar to ocrelizumab, fingolimod treatment reduced Gd+ enhancing lesions, but the percentage of patients (14%) with these inflammatory lesions at baseline was half of that in the ORATORIO trial ^55,23, and did not translate into a corresponding reduction in the rate of disease progression.

In contrast, siponimod was recently shown to have clinical efficacy for SPMS in slowing 3 month CDP based only on EDSS, decreasing in T1 Gd+ enhancing lesions, and decreasing BVL relative to baseline in the EXPAND trial⁵⁶. Although modest, the success of this trial is notable due to the high percentage of patients (64.5%) well into the progressive phase of the disease, defined as not experiencing a relapse for at least 2 years prior to the study⁵⁶. The patients in EXPAND were older, had longer disease duration and higher EDSS scores than in other positive trials ⁵⁶ (Table 1), suggesting efficacy for non-inflammatory associated progression.

The INFORMS and EXPAND trials included patients with different forms of PMS, which may have different etiologies, thus the results may not be directly translatable⁵⁷. The differential S1P receptor specificity for these two therapeutic agents may also play a role. Fingolimod acts on S1P receptors 1, 3, 4, and 5, while siponimod is more narrowly targeted to S1P receptors 1 and 5 ⁵⁰ S1PR1 selective agents have shown benefits in animal models of neurodegeneration ^58,59, and S1PR5 is expressed by oligodendrocytes ⁶⁰, suggesting that these are the dominant receptors responsible for mediating CNS effects. However, human astrocytes can mediate S1P signaling through either S1PR1 or S1PR3 ⁶¹, thus the difference in selectivity for S1PR3 may underlie the differential activity in PMS. Ultimately, further studies are needed to tease out whether the efficacy of siponimod in SPMS is dependent upon intrinsic neuroprotective activity.

Moving Forward

In order to find DMTs effective for inactive PMS, the focus has shifted to neuroprotective agents with neurodegeneration focused outcome measures ³³. There have been a few recent promising Phase 2 trials (Table 3), however, as demonstrated by the recent failure of laquinimod⁶² to slow MS disease progression⁶³, these results based on imaging outcomes may not hold up when using more comprehensive clinical outcomes in Phase 3 trials.

Table 3.

Baseline Demographic and Disease Characteristics for Phase 2 Progressive Multiple Sclerosis Clinical Trials.^*

	MS-SPI (2016)68	SPRINT-MS (2018)67, 85	Lipoic Acid (2017)76	MS-STAT (2014)81
Type of MS	PPMS & SPMS	PPMS & SPMS	SPMS	SPMS
DMT	MD1003 Biotin	Ibudilast	Lipoic Acid	Simvastatin
Phase	2	2	2	2
Total (n) Treated (Placebo)	154 103 (51)	255 (1:1 radomization)	51 27 (24)	140 70 (70)
Male %**	47.0	46.3%	39%	49%
Age Years (mean)	51.8±9.1	55.6±7.3	58.5±5.9	51.5±7
Time since onset MS Mean (Years)**	14.8±8.9	11.88±9.05	29.6±9.5	22.1± 8.3
Time since diagnosis SPMS Mean (Years)**	****	****	****	7.3±5.6
% Relapse-Free (Prior 2-year)	****	****	****	89%
EDSS (mean)**	5.98±0.75	5.14±1.15	6.0 (median)	5.76 ±0.84
T2 Lesion Volume (mean)**	****	10,350±11,140 mm3	10.0±7.3 cm3	****
Normalized brain volume (cm3)**	****	0.8042±0.0295 BPF	1,451±63	1,104±118
T25-FW seconds (mean)**	21.8	16.55	11.7±11.3	1.67±0.91 ft/s (speed)
Outcome	Positive	Positive	Positive	Positive
NCT Identifier	NCT02220933	NCT01982942	NCT01188811	NCT00647348

^*Plus-minus values are means ±SD. MS multiple sclerosis, PPMS primary progressive multiple sclerosis, and SPMS secondary progressive multiple sclerosis.

^**Shown are data for patients in the treatment arm (Baseline demographic and disease characteristics for patients in the placebo arm were not included).

^****Data was not reported for this study.

Ibudilast is a phosphodiesterase inhibitor (4, 10), MIF inhibitor, and TLR4 functional antagonist used as an anti-asthmatic treatment in Japan ⁶⁴. Ibudiliast can readily cross the BBB and has been shown to protect against microglial induced neurotoxicity through the inhibition of inflammation and induction of neurotrophic factors⁶⁵. It had no effect on relapse rate or inflammatory lesion load in RRMS, calling into question its therapeutic mechanism of action, yet it did slow BVLs, suggesting it may benefit progressive patients ⁶⁶. The completed SPRINT-MS trial in PPMS and SPMS found a significant reduction in their primary endpoint of brain atrophy at 96 weeks comparing ibudilast to placebo ⁶⁷, which paves the way for its use in a larger Phase 3 trial.

Biotin has come to be sought after by MS patients following the positive results of two small studies examining the effect of high dose pharmaceutical grade oral biotin (300 mg) on disability progression in PMS^{68, 69}. As an essential coenzyme involved in brain energy metabolism and myelin synthesis, biotin’s protective effect is thought to stem from enhanced remyelination⁷⁰. The MS-SPI trial used disability improvement as measured by a change EDSS or the T25FW, and found that approximately 13% of progressive patients treated with biotin experienced disability improvement⁶⁸. However, a more recent study with 43 progressive patients found that none of the patients improved and EDSS measured disability actually worsened in one-third of biotin treated patients⁷¹. Another report found that high dose biotin exacerbated inflammatory activity in a subset of MS patients⁷². It has also been suggested that the levels of other essential nutrients, such as iron, could impact the efficacy of biotin supplementation⁷³. An ongoing Phase 3 trial (NCT02936037) is expected to settle these conflicting results.

Lipoic Acid is an antioxidant suggested to have neuroprotective properties⁷⁴. It is involved in the reduction of Fe³⁺ to Fe²⁺, and may help prevent iron overload associated toxicity⁷⁵. Treatment with lipoic acid decreased brain atrophy as measured by the percent change in brain volume in SPMS patients⁷⁶. However, all other CNS imaging associated and clinical measures failed to show significant benefits, but may stem from lack of power due to the small sample size⁷⁶. A larger follow-up cohort is needed to adequately interpret the results of this study.

Simvastatin is used to prevent atherosclerosis related complications, and was tested in MS based on its immunomodulatory and neuroprotective activity⁷⁷. Though best known for their effects on triglycerides, statins also reduce ferritin levels and deplete iron stores^{78, 79}. Though not effective for RRMS⁸⁰, it was found to reduce the percent change in BVL as well as reduce the rate of change in EDSS and MSFC scores in patients with SPMS⁸¹. Additionally, a sub-study of this trial revealed cognitive and quality of life improvements with simvastatin treatment⁸². The clinical efficacy of simvastatin in SPMS will be determined in an upcoming Phase 3 trial (NCT03387670).

Conclusion

The greatest impediment to developing effective treatments for PMS continues to be the lack of clear biological mechanisms underlying disease progression. However, the recent success of ocrelizumab and siponimod for at least some subsets of PMS patients provides the unprecedented opportunity to try to tease apart these mechanisms through comparisons of clinical responders and non-responders. These studies could also facilitate the development of new biomarkers directly related to the physiology of progression. Since iron regulation is a proposed activity of several of the prospective PMS DMTs, the monitoring of iron levels may be useful to determine patients most likely to benefit and measure efficacy. However, due to the multifaceted nature of progression, combinatorial metrics are likely going to be needed. The mixed results of past trials highlight the need for better patient stratification and correspondingly tailored outcome measures. Due to the heterogeneous nature of the PMS patient population, there needs to be a shift from a ‘one size fits all’ to a more targeted approach.