The fact that Fenebrutinib is effective in non-active PPMS is encouraging, but the efficacy still remains low and the risk profile is not that good.
I’m hopeful that the next buses will be :
- IMU-838 (vidofludimus calcium) : start of phase 3 trial in PPMS in H2 2026
- Foralumab (anti-CD3) : phase 2a results for SPMS expected by this summer
- Frexalimab (anti-CD40L) : phase 3 results for SPMS expected in late 2026 or early 2027
- Remibrutinib (BTKi) : phase 3 results for RRMS expected by this summer ; highly selective and with a very good risk profile so far. A phase 3 trial for SPMS also started in late 2025.
Yes, you are right, the effect is small, but at least it tells us that we need a peripheral anti-inflammatory agent to build on. Please note that, as the MRI-inactive subgroup analyses were not part of the pre-specified analyses, they will not change the EMA or MHRA labels. In other words, ocrelizumab will remain licensed for people with active PPMS.
It would be fabulous if one of the DMTs could become available to this 74 year old female whose mobility has declined rapidly in the last 18 months. However, I’m still able to walk short distances. Improvement would be wonderful but even plateauing would be good. I fear time is running out for me.😢
RE: your comment that the O'Hand trial is a "game changer" for people with advanced MS. This comment is an example of comments made by MS professionals deep in the weeds of clinical trials, debating which results had statistical significance, which didn't, which therapies can be rammed past regulatory authorities, which can't, for people classified with one type of MS versus another. The truth is, for people with advanced disease, there are no game-changers among any of the existing therapies. Not even close.
You are welcome to dismiss an average of 8.5-9.0 years of independent hand and upper limb function as insignificant, but for the average person with advanced MS, maintaining independence is important. At the moment, we can't offer pwMS who use wheelchairs this potential treatment effect as we are not allowed to use DMTs in wheelchair users. This trial will potentially change that, which is why I call it a game-changer. In other words, what was once considered unsalvageable is now treatable. Granted, this is not a cure, but it is a start on the long journey ahead to continually improve the outcome for people with advanced MS.
My wife has advanced disease. Her hands and upper limbs sort of work and sort of don't. It varies from day to day and depending on what types of hand and upper arm movements we are discussing. So the whole construct of "independent hand and upper limb function" appears very artificial to me. And if you are asking me to believe that any existing study can establish with any reasonable degree of confidence that any existing therapy can extend "independent hand and upper limb function an average of n.5 to n+1 years", however "independent and upper limb functions" is defined, I don't believe that, especially for n as large as 8. And, BTW, my wife lost a lot of her upper limb function when on Ocrevus. Not that I think it was the Ocrevus.
Hi Dr. G. I have a general question that I have been meaning to ask for some time. In your view, why are other imaging tests such as fMRI, PET scans, SPECT scans, etc not used to assess damage, pathology, and disease activity in MS when it is well-known and well-proven by now that conventional MRI misses so much disease pathology and activity (the clinical-radiological paradox of MS)?
My suspicion (which of course I could be wrong about) is that these tests are not used not because of availability, cost, etc but because the MS drugs would not show much if any (or at the very least, not as much) efficacy against the pathology and activity detected via these imaging tests (more specific and sensitive for one thing) vs conventional MRI.
The technical and methodological barriers are real, not just convenient excuses for not doing additional imaging. PET with microglial markers (TSPO ligands such as PK11195) has shown clearly that smouldering inflammation, chronic active lesions, and cortical/meningeal pathology are detectable and clinically meaningful. But TSPO PET has problems as a trial endpoint: the TSPO genetic polymorphism stratifies patients into high/medium/low binders, kinetic modelling requires arterial input functions or validated reference regions (still debated for the cerebellum in MS), tracer half-lives are short, requiring an on-site cyclotron, inter-site standardisation is poor, and radiation exposure limits repeat scanning. Synaptic density PET, and myelin PET tracers are even earlier in development.
fMRI has its own issues — BOLD signal is an indirect haemodynamic measure, network-level changes in MS reflect a mixture of damage and compensatory reorganisation that is genuinely difficult to interpret as a "disease activity" readout, and test-retest reliability for individual-patient inference remains modest. SPECT is largely superseded. Advanced MRI techniques that do capture more of the iceberg — PRL/iron-rim lesions on susceptibility imaging, choroid plexus volume, thalamic and deep grey matter atrophy, spinal cord cross-sectional area, MTR, g-ratio mapping, sodium imaging, 7T cortical lesion imaging — exist and are increasingly used in academic cohorts, but are not standardised across vendors and sites in the way Gd-T1 and T2/FLAIR lesion counting is.
Your suspicion has more substance than the MS field is willing to admit.
Some further observations. First, the regulatory endpoint hierarchy was built around what was measurable in the 1990s — relapses and Gd-enhancing/T2 lesions — and the entire DMT pipeline has been optimised against those endpoints. There is a real selection effect: drugs that suppress focal inflammatory activity are developed, approved, and marketed; mechanisms that might address smouldering disease, chronic active lesions, compartmentalised inflammation, or neurodegeneration are harder to develop because endpoints are harder to measure and trials are longer and costlier. The PRL data make this uncomfortable — high-efficacy DMTs reduce new PRL formation but do relatively little to existing chronic active lesions, which continue to expand. If trials had been built around PRL burden or TSPO signal as primary endpoints, the apparent efficacy ranking of current DMTs would almost certainly look different, and some agents might not even have crossed the approval threshold.
Second, there is a structural disincentive. Sponsors choose endpoints that maximise the probability of a positive trial within a feasible timeframe and sample size. Regulators accept what has historical precedent. Academic investigators who run mechanistic imaging substudies often do so with sponsor cooperation that does not extend to making those endpoints primary. This is not a conspiracy — it is the ordinary operation of risk-averse drug development — but the result is that the imaging modalities most likely to expose the limits of current DMTs are the ones least likely to be deployed at trial scale.
Third, the clinico-radiological paradox is itself partly an artefact of measuring the wrong thing. When OCT-derived GCIPL thinning, spinal cord atrophy, PRL count, and serum NfL/GFAP are added, the "paradox" shrinks substantially. The field knows this. The fact that none of these has been adopted as a regulatory primary endpoint despite years of validation work is telling.
I'd push back on some of your hypotheses. I don't think it's mainly that pharma actively suppresses these modalities — it's more that no single stakeholder has a strong incentive to bear the cost of validating and standardising them to the point where regulators will accept them as primary endpoints. MAGNIMS, NAIMS, and the PRL consensus efforts are trying to close this gap. EMA and FDA have signalled openness to NfL and to PRLs as secondary/exploratory endpoints. The bottleneck is as much coordination and standardisation as it is willingness.
It's also worth noting that some highly effective therapies — anti-CD20s, natalizumab, cladribine, AHSCT — do show meaningful effects on PRL formation, the trajectory of brain volume loss, and NfL, even if they don't fully arrest smouldering disease. So the picture isn't that these modalities would reveal current DMTs to be useless; it's that they would reveal a flatter efficacy gradient and a much larger residual unmet need than the current relapse/Gd-lesion framing implies.
The under-use of advanced imaging in MS trials is overdetermined — technical, regulatory, economic, and inertial factors all contribute — but you are right that the choice of endpoints is not neutral, and that a serious shift to pathology-sensitive imaging would reshape both the apparent efficacy ranking of DMTs and the perceived size of the smouldering/PIRA problem. The honest answer is closer to "yes, partly, but not only" than to either "no, but it's purely practical" or "yes, it's mainly commercial suppression."
Hi Dr. G. Sorry for my belated response. I needed to take a while to think about this response, and I am still thinking about it.
IMHO, first we need to go back to the original “development” of the 4 alleged “sub-types” of MS, which I believe was done by Lublin et al. You yourself have said that these “sub-types” were “created” for the drug companies who developed the early MS drugs (the CRAB drugs) to be eligible for the very lucrative orphan drug designation and benefits.
We know so much more about MS pathology and disease activity than was known in the 1990s (or at least purported to be known; maybe the neurodegeneration/PIRA component was even known back then but was minimized or hidden), but it doesn’t seem to be changing things such as MS drug clinical trial endpoints, which still remain relapses/ARR and/or conventional MRI activity, which are, IMHO, at best, surrogate endpoints.
PwMS are the ones who do not benefit here, while drug companies benefit by debuting more “me too” drugs using the lowest-hanging fruit endpoints that are reliable for them to use. The regulatory bodies are complicit in this, of course, and I have been engaged in advocacy efforts for some time to petition the FDA to move away from these endpoints.
I refuse to use the term “high efficacy” in regards to MS drugs because per my research and understanding, the MS disease activity these drugs are “highly efficacious” against is not the disease activity and pathology that is the main driver of long-term progression and disability accumulation, but this is not made clear or transparent to most PwMS, I believe. I have also always been EXTREMELY unsettled and troubled that there is no set criteria for what constitutes a “high efficacy” MS drug. My understanding is that in the UK, the ABN says that a drug that reduces relapses by 50% or more is a “high efficacy” MS drug, but we have no criteria here in the USA, I believe.
Also, IMHO, NfL and PRLs would be surrogate endpoints as well.
I’m going to use this article as an example: https://pmc.ncbi.nlm.nih.gov/articles/PMC11446943/pdf/415_2024_Article_12621.pdf. This article states: “data from randomized clinical trials indicated that PIRA events constituted roughly 70-90% of all disability accrual events over a follow-up period of 2-10 years [9, 20, 43].” Why are MS drugs called “high efficacy” when what they show this alleged “high efficacy” against is disease activity (active focal inflammation: relapses and/or new conventional MRI activity) that MINIMALLY (at best!) contributes to disability accrual/progression?! How is this right, fair, transparent, and/or ethical to PwMS to be misled like this?
This article also found that: “‘Active’ and ‘non-active’ SPMS patients had a similar risk of achieving disability milestones, suggesting that progression is primarily attributed to PIRA and only to a small extent to disease activity.” This finding then begs the critical question of WHY regulatory bodies have allowed the indication of “active SPMS” for the MS drugs when this finding shows that reducing the “active” MS activity doesn’t reduce the risk of achieving disability milestones.
You write that “So the picture isn’t that these modalities would reveal current DMTs to be useless; it’s that they would reveal a flatter efficacy gradient and a much larger residual unmet need than the current relapse/Gd-lesion framing implies.” I would push back and say that your descriptor of “a flatter efficacy gradient” is EXTREMELY generous! Per the article above, if PIRA contributes 70-90% to disability accrual events, this means that RAW is only contributing 30-10%. Therefore, if the MS drugs only show efficacy against disease activity that contributes 10-30% to disability accrual events, how can they be called “high efficacy”? I have been protesting against this for years!
Really, IMHO, the best endpoints for MS drug clinical trials would be patient-reported outcomes, assuming the PROs were validated. Otherwise, all of the other endpoints are simply surrogates that do not appear to correlate with someone’s clinical status-how they feel. This is what matters to those with MS-how they feel!!!
IMHO, the whole MS “medical-pharmaceutical-industrial-regulatory” complex needs to be restructured. It hurts, deceives, disappoints, and misleads patients at present. This complex, at present, lacks honesty, transparency, and ethics, IMHO, and this is why I am so dedicated to my advocacy work.
Thank you for such a thoughtful and candid response. I appreciate the time you took to write this and the obvious dedication you bring to your advocacy work.
You raise several critical points that are central to the current debate in MS research, and I want to address them directly.
The "Creation" of MS Sub-types: You are correct that our historical clinical phenotypes (RRMS, SPMS, PPMS) are outdated. One could argue that they reflected the limits of what neurologists could observe clinically before we had advanced imaging. However, the field is moving toward a biological understanding of MS as a single, continuous disease process driven by smouldering neuroinflammation. I am hoping the subtypes are dismantled, and MS is MS with or without superimposed focal inflammatory activity or SAW (smouldering associated worsening).
2. PIRA vs. RAW and the "High Efficacy" Label: Your assessment of ‘Progression Independent of Relapse Activity’ (PIRA) is spot on. PIRA drives the vast majority (70-90%) of long-term disability accrual, while Relapse-Associated Worsening (RAW) contributes the minority. This is particularly true in pwMS on DMTs that suppress relapses.
When neurologists and researchers use the term "high efficacy" DMTs (Disease-Modifying Therapies), we are speaking strictly about their ability to suppress focal inflammation (relapses and new Gd-enhancing lesions). To your point, this highlights a disconnect between the medical community and PwMS. To a researcher, a drug that stops 60-70% of relapses is "highly efficacious" at its specific target mechanism. To a patient, a drug isn't truly highly efficacious unless it stops all disease progression.
That said, we shouldn't dismiss the prevention of RAW entirely. While it only accounts for 10-30% of disability, preventing acute, catastrophic inflammatory events is still important to preserving neurological reserve. The problem—which you rightly point out—is that these drugs leave the smouldering, PIRA-driving pathology largely untouched. However, not all DMTs are created equal with respect to the latter. When you rank them on their ability to slow brain volume loss, alemtuzumab and AHSCT come out on top (level 4 efficacy). Why? I suspect they have a different mode of action than the other DMTs.
The Dilemma of Clinical Trial Endpoints: Your frustration with surrogate endpoints like ARR (Annualised Relapse Rate), MRI lesions, NfL, and PRLs is valid. However, regulators and researchers rely on them for time and feasibility reasons.
True disability progression in MS takes years, often decades, to fully manifest. If clinical trials relied solely on long-term disability milestones rather than surrogates, a single trial could take 10 to 15 years to complete. This would effectively paralyse the development of new treatments, and drug companies would stop investing in MS research entirely. Surrogates like NfL aren't perfect, but they give us an early signal of neuroaxonal damage, which will hopefully allow faster development and approvals.
Patient-Reported Outcomes (PROs): I agree wholeheartedly that how a patient feels is the ultimate measure of a drug's success. The challenge with using PROs as primary endpoints in clinical trials is that they are subjective and can be influenced by factors outside of MS pathology, such as sleep, mood, concurrent illnesses, or daily stress. Regulators such as the FDA require objective, quantifiable endpoints to demonstrate a drug's biological mechanism of action. The ideal future is one where trials use a composite endpoint: an objective biological marker of smouldering disease alongside validated PROs.
I do not disagree with your underlying premise: the current framework leaves a massive unmet need for PwMS, specifically regarding neuroprotection and remyelination. We need better endpoints, and we need drugs that target smouldering MS.
Please keep advocating, petitioning, and speaking up. It is pressure from passionate advocates like you that forces the "medical-pharmaceutical-regulatory" complex to sit up and listen, evolve and improve.
-Using info from the Roche publication entitled "Delay in Progression to Key Disability Milestones With Ocrelizumab in Patients With Primary Progressive Multiple Sclerosis in the ORATORIO-HAND Study" by G. Giovannoni et al
-Notice that the results for the endpoints were only statistically significant for the MRI-Inactive subgroups when the ORATORIO definition was used, NOT the ORATORIO-HAND definition. We have to ask why this is.
Here is my analysis of possibly why:
-The ORATORIO-HAND group: the MRI Inactive population becomes: Patients with no Gd+ lesions AND no active T2 lesions
-the ORATORIO group: the MRI Inactive population becomes: Patients with no Gd+ lesions (BUT, this group could still include patients with new/enlarging T2 lesions!)
-the ORATORIO-HAND group is a much "cleaner" inactive group with less inflammatory biology and less treatment-responsive disease whereas the ORATORIO group has a more MRI active population.
-As a result, the ORATORIO-HAND MRI Inactive group becomes a more difficult population for the drug to show benefit in
-As you remove inflammatory patients from the MRI Inactive group, the treatment effect weakens significantly, suggesting that much of OCR's alleged "benefit" for those with PPMS is benefit against active inflammation
We seem to be going around in circles. Ocrelizumab is effective in people with active PPMS, now confirmed in three independent studies, and its effect in inactive PPMS is small.
You say here that “its effect in inactive PPMS is very small,” but in other posts and comments, you seem to have espoused the view that the effect is more significant.
Your view would be that the “very small effect in inactive PPMS” still justifies the use of OCR for those with inactive PPMS, and that the risks would outweigh the possible or alleged “benefit(s)”?
In the inactive group, it delays worsening by 25% (relative reduction), which most consider a small but significant effect. 25% is better than the treatment effect we saw with interferon-1a (Avonex) in relapsing MS, which was licensed to delay disability worsening in relapsing MS.
Please note this is an average effect, with some doing much better than others and some doing worse. However, it is something to build on, i.e. to have at the base of the pyramid when we move in the future to combination therapies.
2nd comment re: fenebrutinib for PPMS and the FENtrepid trial:
-Per Roche's own published info, in the "Baseline T1 Gd+ lesions" sub-group, there was a significant difference between the response to OCR and/or FEN depending on if the subgroup had Baseline T1 Gd+ lesions or not, with the Gd+ positive group favoring OCR and the Gd- group favoring FEN. Interestingly enough, no other sub-groups had the result that this sub-group did, where one group favored OCR and the other favored FEN. This tells us that active inflammation as indicated by Gd+ lesions at baseline is a significant factor which determines treatment response.
You have to wonder if what drove the "success" of this trial's results was the effect of the Gd+ subgroup.
Beyond this aspect, you look at the upper bound of the CI for the 9HPT endpoint, which was 0.98-not meaningful, as the true effect could be as small as a 2% reduction.
For the other endpoints (cCDP12, CDP12-EDSS, and T25FW), the HRs are below 1 (barely), but the upper bounds of the CIs cross 1, indicating uncertainty. The HRs here of 0.88, 0.84, and esp 0.93 are not strong and may even be statistical "noise."
Even the allegedly "strongest" effect "found" in this information is not very impressive: This would be the post hoc analysis based on composite endpoint used in ORATORIO-HAND of CDP12-EDSS + 9HPT, with results of: HR (95% CI), 0.78 (0.64-095). This is NOT an impressive HR by any means, and the upper bound of the CI of 95 is NOT impressive in the least either!
Yes, you are correct. As I said, these results will not change the label or marketing authorisation for PPMS in Europe and the UK. All they do is confirm the original ORATORIO trial results. In other words, it was not a false positive result or a type 1 error. The EMA licensed ocrelizumab for active PPMS on the condition that Roche did a second study to confirm the result; this is called a conditional approval. I am surprised the FDA did not do the same, but that is now water under the bridge, which has long since reached the ocean.
You know my feelings about the FDA's approval of Ocrevus for PPMS: absolutely egregious and non-evidence-based in my view and opinion. However, the way FDA approval works here, we don't have conditional approval. We have accelerated approval but that doesn't seem to have been applicable to Ocrevus for PPMS. The fact that Ocrevus for PPMS got full approval because the FDA allowed the results from the two RMS trials to serve as confirmatory evidence is just absolutely appalling to me and always will be, but when you look at the head of the FDA Neuroscience division who signed off on the Ocrevus for PPMS approval, the decision is far less shocking and surprising, because he was also involved in and/or signed off on other very controversial and questionable approvals of neuroscience drugs: Dr. William Dunn. One of my life goals and dreams is to write a book about all the damage that Dr. Dunn has done.
Well, the good news is that it's now history and time to move on. The ORATORIO trial has been reproduced, and hence ocrelizumab does work in PPMS. It should be viewed as a good news story, as people with PPMS have had access to an effective therapy for 9 years. If it were not approved, many would have missed out on an effective treatment. Maybe the PPMS community should celebrate this fact rather than ruminate on what could or should have happened.
Another way of thinking about this is in terms of days of mobility or upper-limb function saved by the FDA's earlier licensing of ocrelizumab for PPMS, rather than waiting for the O'HAND trial data.
Please note that, as the MRI-inactive subgroup analyses were not part of the pre-specified analyses, they will not change the EMA or MHRA labels. In other words, ocrelizumab will remain licensed for people with active PPMS.
Please bear with me as I compose 1 comment at a time. This 1st comment will regard the HERCULES vs. the PERSEUS tolebrutinib trials in progressive MS:
1) Differences in trial inclusion criteria:
a) HERCULES
-ages 18 to 60
-dx of nrSPMS according to the 2017 McD criteria
-EDSS: 3 to 6.5
-documented evidence of disability progression observed during the 12 months before screening
-absence of clinical relapses for at least 24 months
-primary endpoint: time to onset of 6-month CDP as assessed by EDSS
b) PERSEUS:
-ages 18 to 55
-dx of PPMS per 2017 McD criteria
-EDSS: 2 to 6.5
-positive CSF OCBs and/or elevated IgG index either during screening history or documented previous history
-primary endpoint: 6-month cCDP
The FDA's CRL for tolebrutinib from last Dec (2025) stated the following:
-"Enrolled subjects could have either non-active SPMS...or active SPMS (i.e., no clinical relapses but evidence of inflammatory MRI activity..."
-"Because of the lack of historical MRI data collection in the study, we were unable to retrospectively determine whether the subjects enrolled in Study...[HERCULES] had active or non-active SPMS..."
-"Subgroup analyses of Study...[HERCULES] indicate that the observed treatment effect was greater in subjects who had T1 gadolinium-enhancing (GdE) lesions at baseline (i.e., active SPMS), which compromised 13% of the enrolled population...This subgroup analysis suggests that the treatment effect is largely driven by the small group of subjects with active SPMS based on their baseline MRI scan..."
-"There were much larger treatment effects in the population with baseline GdE lesions than in those without baseline GdE lesions"
-"...the diminished treatment effects in the subgroup analysis of subjects without baseline GdE lesions"
Commentary:
-The "success" of HERCULES vs the failure of PERSEUS could be due, at least in some part, to differences in the trial populations; I would really like to know what percentage of the PERSEUS population had GdE+ lesions at baseline (i.e., active PPMS) AND/OR the fact that these 2 trials had different primary endpoints
Sanofi are not seeking a PPMS license for tolebrutinib. However, they have just received EMA approval for non-relapsing SPMS, regardless of MRI activity (Gd-lesions or new T2 lesions).
It's also interesting to see the differences between the CHMP's positive opinion about tolebrutinib for nrSPMS regardless of MRI activity vs the FDA's CRL from last Dec. Per the FDA CRL, it seems that there may be little to no benefit from tolebrutinib for those with nrSPMS without MRI activity.
o quote Richard Peto one of the giants of British medicine: “The appropriate interpretation of apparently different results in different subgroups of trial results is still one of the most difficult matters of judgement in the interpretation of randomised evidence; at present, many clinicians and regulatory agencies pay far too much attention to irregularities between the apparent effects in different subgroups, to the potential detriment of the care of individual patients” (Peto, 2011).
I think the FDA is making a mistake by moving the goalposts. The trial was powered to answer whether tolebrutinib works in nrSPMS, not in subgroups.
It makes no difference as the inclusion criteria are based on the clinical presentation, i.e. no relapses. The MRI is done after the subjects enter the trial as part of the baseline and is not read locally. This would align with clinical practice, so people with and without Gd enhancement would be classified as nr-SPMS. What I find galling is that the FDA vetted and accepted the protocol, and now they want to potentially limit the treatment to those with Gd-enhancement. This is trying to retrofit the subjects.
One thing I do wonder about, and please weigh in on, is the FDA's definitions of active vs non-active SPMS per their tolebrutinib CRL, which are as follows:
a) non-active SPMS: no clinical relapses and no inflammatory MRI activity
b) active SPMS: no clinical relapses but evidence of inflammatory MRI activity
These definitions don't make sense in that if non-active is NO relapses and NO inflammatory MRI activity, then why is active NO relapses but YES, evidence of inflammatory MRI activity.
This further doesn't make sense when the FDA writes the following about active SPMS: "active SPMS...which is considered a relapsing form of MS (RMS) and for which there are approved therapies." (FDA tolebrutinib CRL, page 3).
How can the FDA say the or their definition of active SPMS is one in which there are NO relapses but then say active SPMS "is considered a relapsing form of MS (RMS)"?
Maybe this is correct and makes sense, and thus, I do not understand correctly, but I have been very puzzled and stumped by this. I would appreciate your thoughts and insight here.
Yes. Sanofi has no grounds to seek a PPMS license for tolebrutinib IMHO in light of the failed PERSEUS trial.
My understanding of the tolebrutinib for nrSPMS status with the EMA at this time is that the CHMP issued a positive opinion but that the EMA decision has not been made or at least announced yet. My understanding here could be wrong, though.
EMA decision is a rubber-stamping exercise, i.e. part of the EMA bureaucracy. Once the CHMP gives the green light, it is almost certain to be approved. I don't think it is a mistake. We must not let underpowered subgroup analyses mislead us. To quote Richard Peto one of the giants of British medicine: “The appropriate interpretation of apparently different results in different subgroups of trial results is still one of the most difficult matters of judgement in the interpretation of randomised evidence; at present, many clinicians and regulatory agencies pay far too much attention to irregularities between the apparent effects in different subgroups, to the potential detriment of the care of individual patients” (Peto, 2011).
“It is for this reason why I think higher doses of ocrelizumab are better than lower doses. I think this is due to deeper B-cell depletion and possibly greater CNS effects with higher doses”
Does this suggest high dose ocrelizumab could be beneficial for those with smouldering MS?
Only in those with active MS. We need CNS penetrant therapies. Either using a shuttle to get more anti-CD20 monoclonal antibodies into the CNS, BTKis, CAR T-cells etc.
There is a P1 study by Roche to use a brain shuttle with an anti CD20 (probably ocrelizumab?). The treatment is called RO7121932. Results expected next year.
“Is this because Gd-enhancing lesions are an indicator or message from the immune system that whatever is causing MS is active and modifiable by anti-CD20 therapy? Could this be related to EBV?”
We seem to be far way from any treatments that effectively tackle smouldering MS / progression. MS researchers (Nero) seem to be fiddling around with anti-CD20s whilst the CNSs (Rome) of MS patients burns. Where are we with therapies to purge the CNS of EBV? What’s happening to the trials of anti-virals targeting EBV?
Why aren’t MS patients (150k) seeing the same advances as in cancer?
There are three MS antiviral trials that started in 2026.
There will be a readout next year from the trials, and 2 are phase 3 so this means faster approval. 2 MS trials are testing Tenofovir Alafenamide for MS-EBV.
A 3rd trial is testing spironolactone and famciclovir.
My understanding and impression as to why MS patients are not seeing the same advances as cancer patients is because tackling neurodegenerative we see the same challenge playing out in other neurodegenerative diseases that lack meaningful treatment options. The fact that MS is framed, per the conventional narrative, as a primary autoimmune disease is very misleading and deceitful in my opinion, it intentionally downplays the neurodegeneration that occurs in MS, which per my understanding of the research, is the primary driver and cause of long-term progression and disability accumulation.
I totally agree. The money was in anti-relapse drugs and pharma chased the money. Unfortunately, the MS researchers were also happy to follow the money. No one is willing to rock the boat / bite the hand that feeds them. Two things you’ll never see in this life: a neuro living in social housing or a neuro who can point to a patient who has got better / stabilised (MS, MND, dementia…). Michael J Fox has raised over $1 billion for Parkinson’s research yet I don’t see any cure on the horizon.
The "medical-pharmaceutical-industrial" complex has to keep the conventional/convenient MS narrative going-that MS is a primary disease consisting of relapses and/or new conventional MRI activity-the 2 aspects that Big Pharma's drugs show efficacy against.
What they and their narrative doesn't and won't tell PwMS:
-relapses aren't the driver of long-term progression and disability accumulation
-the clinical-radiological paradox of MS
-the differences between NMOSD and MOGAD, where progression is driven by relapses, which is not the case for MS
-the neurodegeneration in MS (PIRA) and that their drugs aren't effective against it
-that MS may likely not be a primary autoimmune disease-one example: no target antigen(s) have ever been found despite decades of searching
-the Numbers Needed to Treat for their drugs for MS (some are as high as in the 20s; some may be even higher), meaning that not everyone who takes an MS drug receives the alleged "benefits"
-that MS is a progressive neurodegenerative disease
-that there is SO MUCH pathology and disease activity in MS that can't be seen on conventional MRI
I could go on and on with this list, but I will stop here.
The current MS drugs are over-hyped, and their "benefits" are exaggerated, distorted, and misrepresented. It's wrong, unethical, and harmful, IMHO
I agree, but pretty much any info one reads or finds about MS describes it as an autoimmune disease, and the MS drugs are all anti-inflammatory treatments based on MS as an autoimmune disease is my understanding, which is why they don't work (IMHO and per my research; happy to provide evidence upon request) against the PIRA and neurodegeneration in MS, only the active focal inflammation disease activity.
Have you heard of the inside-out vs outside-in theories of MS? Maybe you would be in the inside-out camp, then. Have you heard of Dr. Peter K Stys and his research alleging the inside-out theory of MS? Highly recommend reading.
I believe the "autoimmune" activity is a reaction to a primary neurodegenerative pathology in MS personally.
I downloaded your slides from ECTRIMS - "Efficacy and Safety of Ocrelizumab vs Placebo in Primary Progressive MS: Results of the Phase IIIb ORATORIO-HAND Study" and I read through them. Very interesting, thank you. Please share a link related to your data about wheelchair users or impacts data from your oratorio study on non-active PPMS patients.
This data is directly applicable to me: I am in the US, 55 yo male, progressive MS since 2001. Ocrevus for the last six years. Clear progression despite Ocrevus use in terms of balance and walking and timed 25 ft walk test. No active lesions on imaging the past four years. I have been on standard dosing of every six months, and I am looking to moving to reduced dosing schedule of once a year if my next MRI shows no new lesions and evaluating at a later time discontinuing Ocrevus.
I know given past comments in your writing and your papers that you are against extended interval dosing and with these Oratorio results discontinuation of drug (at least if have active disease as defined by lesions on imaging). A fair statement or am I mischaracterizing?
Of course I don't know what my disease presentation would look like if I had never been on Ocrevus the last six years, so it's hard to compare A to B and this hypothetical. I'm confident it has had a positive benefit within the limitations of the drug, but perhaps because it is not penetrant to the CNS, it's had minimal effect as compared to a drug that might be more penetrating?
if I had additional data to review related to impacts on non-active PPMS patients, it might tell me something about EID or discontinuation especially as it relates to hand function. I think the results are promising for peope with more active disease but I am still unclear on numbers regarding non active PPMS. Thank you
I'm still hopeful that CN|M Au8 will have a positive impact on my PPMS. Its such a shame that Tolebrutinib will never be available to pwPPMS. Never feels like too negative an outlook but is there any way back Tolebrutinib as a treatment for PPMS?
The fact that Fenebrutinib is effective in non-active PPMS is encouraging, but the efficacy still remains low and the risk profile is not that good.
I’m hopeful that the next buses will be :
- IMU-838 (vidofludimus calcium) : start of phase 3 trial in PPMS in H2 2026
- Foralumab (anti-CD3) : phase 2a results for SPMS expected by this summer
- Frexalimab (anti-CD40L) : phase 3 results for SPMS expected in late 2026 or early 2027
- Remibrutinib (BTKi) : phase 3 results for RRMS expected by this summer ; highly selective and with a very good risk profile so far. A phase 3 trial for SPMS also started in late 2025.
- TYK2 inhibitors (SUDO-550, A-005, etc.) entering phase 2 soon
- And of course anti-BCMA CAR-T therapies (CT103A, AZD0120)
Yes, you are right, the effect is small, but at least it tells us that we need a peripheral anti-inflammatory agent to build on. Please note that, as the MRI-inactive subgroup analyses were not part of the pre-specified analyses, they will not change the EMA or MHRA labels. In other words, ocrelizumab will remain licensed for people with active PPMS.
It would be fabulous if one of the DMTs could become available to this 74 year old female whose mobility has declined rapidly in the last 18 months. However, I’m still able to walk short distances. Improvement would be wonderful but even plateauing would be good. I fear time is running out for me.😢
Moira, you are a champion - MS and still walking at 74! I hope you get access to a DMT.
RE: your comment that the O'Hand trial is a "game changer" for people with advanced MS. This comment is an example of comments made by MS professionals deep in the weeds of clinical trials, debating which results had statistical significance, which didn't, which therapies can be rammed past regulatory authorities, which can't, for people classified with one type of MS versus another. The truth is, for people with advanced disease, there are no game-changers among any of the existing therapies. Not even close.
You are welcome to dismiss an average of 8.5-9.0 years of independent hand and upper limb function as insignificant, but for the average person with advanced MS, maintaining independence is important. At the moment, we can't offer pwMS who use wheelchairs this potential treatment effect as we are not allowed to use DMTs in wheelchair users. This trial will potentially change that, which is why I call it a game-changer. In other words, what was once considered unsalvageable is now treatable. Granted, this is not a cure, but it is a start on the long journey ahead to continually improve the outcome for people with advanced MS.
My wife has advanced disease. Her hands and upper limbs sort of work and sort of don't. It varies from day to day and depending on what types of hand and upper arm movements we are discussing. So the whole construct of "independent hand and upper limb function" appears very artificial to me. And if you are asking me to believe that any existing study can establish with any reasonable degree of confidence that any existing therapy can extend "independent hand and upper limb function an average of n.5 to n+1 years", however "independent and upper limb functions" is defined, I don't believe that, especially for n as large as 8. And, BTW, my wife lost a lot of her upper limb function when on Ocrevus. Not that I think it was the Ocrevus.
Hi Dr. G. I have a general question that I have been meaning to ask for some time. In your view, why are other imaging tests such as fMRI, PET scans, SPECT scans, etc not used to assess damage, pathology, and disease activity in MS when it is well-known and well-proven by now that conventional MRI misses so much disease pathology and activity (the clinical-radiological paradox of MS)?
My suspicion (which of course I could be wrong about) is that these tests are not used not because of availability, cost, etc but because the MS drugs would not show much if any (or at the very least, not as much) efficacy against the pathology and activity detected via these imaging tests (more specific and sensitive for one thing) vs conventional MRI.
Thank you.
The technical and methodological barriers are real, not just convenient excuses for not doing additional imaging. PET with microglial markers (TSPO ligands such as PK11195) has shown clearly that smouldering inflammation, chronic active lesions, and cortical/meningeal pathology are detectable and clinically meaningful. But TSPO PET has problems as a trial endpoint: the TSPO genetic polymorphism stratifies patients into high/medium/low binders, kinetic modelling requires arterial input functions or validated reference regions (still debated for the cerebellum in MS), tracer half-lives are short, requiring an on-site cyclotron, inter-site standardisation is poor, and radiation exposure limits repeat scanning. Synaptic density PET, and myelin PET tracers are even earlier in development.
fMRI has its own issues — BOLD signal is an indirect haemodynamic measure, network-level changes in MS reflect a mixture of damage and compensatory reorganisation that is genuinely difficult to interpret as a "disease activity" readout, and test-retest reliability for individual-patient inference remains modest. SPECT is largely superseded. Advanced MRI techniques that do capture more of the iceberg — PRL/iron-rim lesions on susceptibility imaging, choroid plexus volume, thalamic and deep grey matter atrophy, spinal cord cross-sectional area, MTR, g-ratio mapping, sodium imaging, 7T cortical lesion imaging — exist and are increasingly used in academic cohorts, but are not standardised across vendors and sites in the way Gd-T1 and T2/FLAIR lesion counting is.
Your suspicion has more substance than the MS field is willing to admit.
Some further observations. First, the regulatory endpoint hierarchy was built around what was measurable in the 1990s — relapses and Gd-enhancing/T2 lesions — and the entire DMT pipeline has been optimised against those endpoints. There is a real selection effect: drugs that suppress focal inflammatory activity are developed, approved, and marketed; mechanisms that might address smouldering disease, chronic active lesions, compartmentalised inflammation, or neurodegeneration are harder to develop because endpoints are harder to measure and trials are longer and costlier. The PRL data make this uncomfortable — high-efficacy DMTs reduce new PRL formation but do relatively little to existing chronic active lesions, which continue to expand. If trials had been built around PRL burden or TSPO signal as primary endpoints, the apparent efficacy ranking of current DMTs would almost certainly look different, and some agents might not even have crossed the approval threshold.
Second, there is a structural disincentive. Sponsors choose endpoints that maximise the probability of a positive trial within a feasible timeframe and sample size. Regulators accept what has historical precedent. Academic investigators who run mechanistic imaging substudies often do so with sponsor cooperation that does not extend to making those endpoints primary. This is not a conspiracy — it is the ordinary operation of risk-averse drug development — but the result is that the imaging modalities most likely to expose the limits of current DMTs are the ones least likely to be deployed at trial scale.
Third, the clinico-radiological paradox is itself partly an artefact of measuring the wrong thing. When OCT-derived GCIPL thinning, spinal cord atrophy, PRL count, and serum NfL/GFAP are added, the "paradox" shrinks substantially. The field knows this. The fact that none of these has been adopted as a regulatory primary endpoint despite years of validation work is telling.
I'd push back on some of your hypotheses. I don't think it's mainly that pharma actively suppresses these modalities — it's more that no single stakeholder has a strong incentive to bear the cost of validating and standardising them to the point where regulators will accept them as primary endpoints. MAGNIMS, NAIMS, and the PRL consensus efforts are trying to close this gap. EMA and FDA have signalled openness to NfL and to PRLs as secondary/exploratory endpoints. The bottleneck is as much coordination and standardisation as it is willingness.
It's also worth noting that some highly effective therapies — anti-CD20s, natalizumab, cladribine, AHSCT — do show meaningful effects on PRL formation, the trajectory of brain volume loss, and NfL, even if they don't fully arrest smouldering disease. So the picture isn't that these modalities would reveal current DMTs to be useless; it's that they would reveal a flatter efficacy gradient and a much larger residual unmet need than the current relapse/Gd-lesion framing implies.
The under-use of advanced imaging in MS trials is overdetermined — technical, regulatory, economic, and inertial factors all contribute — but you are right that the choice of endpoints is not neutral, and that a serious shift to pathology-sensitive imaging would reshape both the apparent efficacy ranking of DMTs and the perceived size of the smouldering/PIRA problem. The honest answer is closer to "yes, partly, but not only" than to either "no, but it's purely practical" or "yes, it's mainly commercial suppression."
Would you weigh it differently?
Hi Dr. G. Sorry for my belated response. I needed to take a while to think about this response, and I am still thinking about it.
IMHO, first we need to go back to the original “development” of the 4 alleged “sub-types” of MS, which I believe was done by Lublin et al. You yourself have said that these “sub-types” were “created” for the drug companies who developed the early MS drugs (the CRAB drugs) to be eligible for the very lucrative orphan drug designation and benefits.
We know so much more about MS pathology and disease activity than was known in the 1990s (or at least purported to be known; maybe the neurodegeneration/PIRA component was even known back then but was minimized or hidden), but it doesn’t seem to be changing things such as MS drug clinical trial endpoints, which still remain relapses/ARR and/or conventional MRI activity, which are, IMHO, at best, surrogate endpoints.
PwMS are the ones who do not benefit here, while drug companies benefit by debuting more “me too” drugs using the lowest-hanging fruit endpoints that are reliable for them to use. The regulatory bodies are complicit in this, of course, and I have been engaged in advocacy efforts for some time to petition the FDA to move away from these endpoints.
I refuse to use the term “high efficacy” in regards to MS drugs because per my research and understanding, the MS disease activity these drugs are “highly efficacious” against is not the disease activity and pathology that is the main driver of long-term progression and disability accumulation, but this is not made clear or transparent to most PwMS, I believe. I have also always been EXTREMELY unsettled and troubled that there is no set criteria for what constitutes a “high efficacy” MS drug. My understanding is that in the UK, the ABN says that a drug that reduces relapses by 50% or more is a “high efficacy” MS drug, but we have no criteria here in the USA, I believe.
Also, IMHO, NfL and PRLs would be surrogate endpoints as well.
I’m going to use this article as an example: https://pmc.ncbi.nlm.nih.gov/articles/PMC11446943/pdf/415_2024_Article_12621.pdf. This article states: “data from randomized clinical trials indicated that PIRA events constituted roughly 70-90% of all disability accrual events over a follow-up period of 2-10 years [9, 20, 43].” Why are MS drugs called “high efficacy” when what they show this alleged “high efficacy” against is disease activity (active focal inflammation: relapses and/or new conventional MRI activity) that MINIMALLY (at best!) contributes to disability accrual/progression?! How is this right, fair, transparent, and/or ethical to PwMS to be misled like this?
This article also found that: “‘Active’ and ‘non-active’ SPMS patients had a similar risk of achieving disability milestones, suggesting that progression is primarily attributed to PIRA and only to a small extent to disease activity.” This finding then begs the critical question of WHY regulatory bodies have allowed the indication of “active SPMS” for the MS drugs when this finding shows that reducing the “active” MS activity doesn’t reduce the risk of achieving disability milestones.
You write that “So the picture isn’t that these modalities would reveal current DMTs to be useless; it’s that they would reveal a flatter efficacy gradient and a much larger residual unmet need than the current relapse/Gd-lesion framing implies.” I would push back and say that your descriptor of “a flatter efficacy gradient” is EXTREMELY generous! Per the article above, if PIRA contributes 70-90% to disability accrual events, this means that RAW is only contributing 30-10%. Therefore, if the MS drugs only show efficacy against disease activity that contributes 10-30% to disability accrual events, how can they be called “high efficacy”? I have been protesting against this for years!
Really, IMHO, the best endpoints for MS drug clinical trials would be patient-reported outcomes, assuming the PROs were validated. Otherwise, all of the other endpoints are simply surrogates that do not appear to correlate with someone’s clinical status-how they feel. This is what matters to those with MS-how they feel!!!
IMHO, the whole MS “medical-pharmaceutical-industrial-regulatory” complex needs to be restructured. It hurts, deceives, disappoints, and misleads patients at present. This complex, at present, lacks honesty, transparency, and ethics, IMHO, and this is why I am so dedicated to my advocacy work.
Thank you for such a thoughtful and candid response. I appreciate the time you took to write this and the obvious dedication you bring to your advocacy work.
You raise several critical points that are central to the current debate in MS research, and I want to address them directly.
The "Creation" of MS Sub-types: You are correct that our historical clinical phenotypes (RRMS, SPMS, PPMS) are outdated. One could argue that they reflected the limits of what neurologists could observe clinically before we had advanced imaging. However, the field is moving toward a biological understanding of MS as a single, continuous disease process driven by smouldering neuroinflammation. I am hoping the subtypes are dismantled, and MS is MS with or without superimposed focal inflammatory activity or SAW (smouldering associated worsening).
2. PIRA vs. RAW and the "High Efficacy" Label: Your assessment of ‘Progression Independent of Relapse Activity’ (PIRA) is spot on. PIRA drives the vast majority (70-90%) of long-term disability accrual, while Relapse-Associated Worsening (RAW) contributes the minority. This is particularly true in pwMS on DMTs that suppress relapses.
When neurologists and researchers use the term "high efficacy" DMTs (Disease-Modifying Therapies), we are speaking strictly about their ability to suppress focal inflammation (relapses and new Gd-enhancing lesions). To your point, this highlights a disconnect between the medical community and PwMS. To a researcher, a drug that stops 60-70% of relapses is "highly efficacious" at its specific target mechanism. To a patient, a drug isn't truly highly efficacious unless it stops all disease progression.
That said, we shouldn't dismiss the prevention of RAW entirely. While it only accounts for 10-30% of disability, preventing acute, catastrophic inflammatory events is still important to preserving neurological reserve. The problem—which you rightly point out—is that these drugs leave the smouldering, PIRA-driving pathology largely untouched. However, not all DMTs are created equal with respect to the latter. When you rank them on their ability to slow brain volume loss, alemtuzumab and AHSCT come out on top (level 4 efficacy). Why? I suspect they have a different mode of action than the other DMTs.
The Dilemma of Clinical Trial Endpoints: Your frustration with surrogate endpoints like ARR (Annualised Relapse Rate), MRI lesions, NfL, and PRLs is valid. However, regulators and researchers rely on them for time and feasibility reasons.
True disability progression in MS takes years, often decades, to fully manifest. If clinical trials relied solely on long-term disability milestones rather than surrogates, a single trial could take 10 to 15 years to complete. This would effectively paralyse the development of new treatments, and drug companies would stop investing in MS research entirely. Surrogates like NfL aren't perfect, but they give us an early signal of neuroaxonal damage, which will hopefully allow faster development and approvals.
Patient-Reported Outcomes (PROs): I agree wholeheartedly that how a patient feels is the ultimate measure of a drug's success. The challenge with using PROs as primary endpoints in clinical trials is that they are subjective and can be influenced by factors outside of MS pathology, such as sleep, mood, concurrent illnesses, or daily stress. Regulators such as the FDA require objective, quantifiable endpoints to demonstrate a drug's biological mechanism of action. The ideal future is one where trials use a composite endpoint: an objective biological marker of smouldering disease alongside validated PROs.
I do not disagree with your underlying premise: the current framework leaves a massive unmet need for PwMS, specifically regarding neuroprotection and remyelination. We need better endpoints, and we need drugs that target smouldering MS.
Please keep advocating, petitioning, and speaking up. It is pressure from passionate advocates like you that forces the "medical-pharmaceutical-regulatory" complex to sit up and listen, evolve and improve.
“it is the ordinary operation of risk-averse drug development.” Thank you for telling it like it is, even if it’s depressing.
3rd comment: Getting to O'HAND:
-Using info from the Roche publication entitled "Delay in Progression to Key Disability Milestones With Ocrelizumab in Patients With Primary Progressive Multiple Sclerosis in the ORATORIO-HAND Study" by G. Giovannoni et al
-Notice that the results for the endpoints were only statistically significant for the MRI-Inactive subgroups when the ORATORIO definition was used, NOT the ORATORIO-HAND definition. We have to ask why this is.
Here is my analysis of possibly why:
-The ORATORIO-HAND group: the MRI Inactive population becomes: Patients with no Gd+ lesions AND no active T2 lesions
-the ORATORIO group: the MRI Inactive population becomes: Patients with no Gd+ lesions (BUT, this group could still include patients with new/enlarging T2 lesions!)
-the ORATORIO-HAND group is a much "cleaner" inactive group with less inflammatory biology and less treatment-responsive disease whereas the ORATORIO group has a more MRI active population.
-As a result, the ORATORIO-HAND MRI Inactive group becomes a more difficult population for the drug to show benefit in
-As you remove inflammatory patients from the MRI Inactive group, the treatment effect weakens significantly, suggesting that much of OCR's alleged "benefit" for those with PPMS is benefit against active inflammation
We seem to be going around in circles. Ocrelizumab is effective in people with active PPMS, now confirmed in three independent studies, and its effect in inactive PPMS is small.
You say here that “its effect in inactive PPMS is very small,” but in other posts and comments, you seem to have espoused the view that the effect is more significant.
Your view would be that the “very small effect in inactive PPMS” still justifies the use of OCR for those with inactive PPMS, and that the risks would outweigh the possible or alleged “benefit(s)”?
In the inactive group, it delays worsening by 25% (relative reduction), which most consider a small but significant effect. 25% is better than the treatment effect we saw with interferon-1a (Avonex) in relapsing MS, which was licensed to delay disability worsening in relapsing MS.
Please note this is an average effect, with some doing much better than others and some doing worse. However, it is something to build on, i.e. to have at the base of the pyramid when we move in the future to combination therapies.
2nd comment re: fenebrutinib for PPMS and the FENtrepid trial:
-Per Roche's own published info, in the "Baseline T1 Gd+ lesions" sub-group, there was a significant difference between the response to OCR and/or FEN depending on if the subgroup had Baseline T1 Gd+ lesions or not, with the Gd+ positive group favoring OCR and the Gd- group favoring FEN. Interestingly enough, no other sub-groups had the result that this sub-group did, where one group favored OCR and the other favored FEN. This tells us that active inflammation as indicated by Gd+ lesions at baseline is a significant factor which determines treatment response.
You have to wonder if what drove the "success" of this trial's results was the effect of the Gd+ subgroup.
Beyond this aspect, you look at the upper bound of the CI for the 9HPT endpoint, which was 0.98-not meaningful, as the true effect could be as small as a 2% reduction.
For the other endpoints (cCDP12, CDP12-EDSS, and T25FW), the HRs are below 1 (barely), but the upper bounds of the CIs cross 1, indicating uncertainty. The HRs here of 0.88, 0.84, and esp 0.93 are not strong and may even be statistical "noise."
Even the allegedly "strongest" effect "found" in this information is not very impressive: This would be the post hoc analysis based on composite endpoint used in ORATORIO-HAND of CDP12-EDSS + 9HPT, with results of: HR (95% CI), 0.78 (0.64-095). This is NOT an impressive HR by any means, and the upper bound of the CI of 95 is NOT impressive in the least either!
Yes, you are correct. As I said, these results will not change the label or marketing authorisation for PPMS in Europe and the UK. All they do is confirm the original ORATORIO trial results. In other words, it was not a false positive result or a type 1 error. The EMA licensed ocrelizumab for active PPMS on the condition that Roche did a second study to confirm the result; this is called a conditional approval. I am surprised the FDA did not do the same, but that is now water under the bridge, which has long since reached the ocean.
You know my feelings about the FDA's approval of Ocrevus for PPMS: absolutely egregious and non-evidence-based in my view and opinion. However, the way FDA approval works here, we don't have conditional approval. We have accelerated approval but that doesn't seem to have been applicable to Ocrevus for PPMS. The fact that Ocrevus for PPMS got full approval because the FDA allowed the results from the two RMS trials to serve as confirmatory evidence is just absolutely appalling to me and always will be, but when you look at the head of the FDA Neuroscience division who signed off on the Ocrevus for PPMS approval, the decision is far less shocking and surprising, because he was also involved in and/or signed off on other very controversial and questionable approvals of neuroscience drugs: Dr. William Dunn. One of my life goals and dreams is to write a book about all the damage that Dr. Dunn has done.
Well, the good news is that it's now history and time to move on. The ORATORIO trial has been reproduced, and hence ocrelizumab does work in PPMS. It should be viewed as a good news story, as people with PPMS have had access to an effective therapy for 9 years. If it were not approved, many would have missed out on an effective treatment. Maybe the PPMS community should celebrate this fact rather than ruminate on what could or should have happened.
Another way of thinking about this is in terms of days of mobility or upper-limb function saved by the FDA's earlier licensing of ocrelizumab for PPMS, rather than waiting for the O'HAND trial data.
Please note that, as the MRI-inactive subgroup analyses were not part of the pre-specified analyses, they will not change the EMA or MHRA labels. In other words, ocrelizumab will remain licensed for people with active PPMS.
Please bear with me as I compose 1 comment at a time. This 1st comment will regard the HERCULES vs. the PERSEUS tolebrutinib trials in progressive MS:
1) Differences in trial inclusion criteria:
a) HERCULES
-ages 18 to 60
-dx of nrSPMS according to the 2017 McD criteria
-EDSS: 3 to 6.5
-documented evidence of disability progression observed during the 12 months before screening
-absence of clinical relapses for at least 24 months
-primary endpoint: time to onset of 6-month CDP as assessed by EDSS
b) PERSEUS:
-ages 18 to 55
-dx of PPMS per 2017 McD criteria
-EDSS: 2 to 6.5
-positive CSF OCBs and/or elevated IgG index either during screening history or documented previous history
-primary endpoint: 6-month cCDP
The FDA's CRL for tolebrutinib from last Dec (2025) stated the following:
-"Enrolled subjects could have either non-active SPMS...or active SPMS (i.e., no clinical relapses but evidence of inflammatory MRI activity..."
-"Because of the lack of historical MRI data collection in the study, we were unable to retrospectively determine whether the subjects enrolled in Study...[HERCULES] had active or non-active SPMS..."
-"Subgroup analyses of Study...[HERCULES] indicate that the observed treatment effect was greater in subjects who had T1 gadolinium-enhancing (GdE) lesions at baseline (i.e., active SPMS), which compromised 13% of the enrolled population...This subgroup analysis suggests that the treatment effect is largely driven by the small group of subjects with active SPMS based on their baseline MRI scan..."
-"There were much larger treatment effects in the population with baseline GdE lesions than in those without baseline GdE lesions"
-"...the diminished treatment effects in the subgroup analysis of subjects without baseline GdE lesions"
Commentary:
-The "success" of HERCULES vs the failure of PERSEUS could be due, at least in some part, to differences in the trial populations; I would really like to know what percentage of the PERSEUS population had GdE+ lesions at baseline (i.e., active PPMS) AND/OR the fact that these 2 trials had different primary endpoints
-
In the Perseus trial, 14.1% of subjects in the placebo arm and 10% in the tolebrutinib arm had Gd-enhancing lesions at baseline.
Sanofi are not seeking a PPMS license for tolebrutinib. However, they have just received EMA approval for non-relapsing SPMS, regardless of MRI activity (Gd-lesions or new T2 lesions).
It's also interesting to see the differences between the CHMP's positive opinion about tolebrutinib for nrSPMS regardless of MRI activity vs the FDA's CRL from last Dec. Per the FDA CRL, it seems that there may be little to no benefit from tolebrutinib for those with nrSPMS without MRI activity.
o quote Richard Peto one of the giants of British medicine: “The appropriate interpretation of apparently different results in different subgroups of trial results is still one of the most difficult matters of judgement in the interpretation of randomised evidence; at present, many clinicians and regulatory agencies pay far too much attention to irregularities between the apparent effects in different subgroups, to the potential detriment of the care of individual patients” (Peto, 2011).
I think the FDA is making a mistake by moving the goalposts. The trial was powered to answer whether tolebrutinib works in nrSPMS, not in subgroups.
One question is whether "nrSPMS" includes or should or should not include those with MRI activity, I believe.
It makes no difference as the inclusion criteria are based on the clinical presentation, i.e. no relapses. The MRI is done after the subjects enter the trial as part of the baseline and is not read locally. This would align with clinical practice, so people with and without Gd enhancement would be classified as nr-SPMS. What I find galling is that the FDA vetted and accepted the protocol, and now they want to potentially limit the treatment to those with Gd-enhancement. This is trying to retrofit the subjects.
One thing I do wonder about, and please weigh in on, is the FDA's definitions of active vs non-active SPMS per their tolebrutinib CRL, which are as follows:
a) non-active SPMS: no clinical relapses and no inflammatory MRI activity
b) active SPMS: no clinical relapses but evidence of inflammatory MRI activity
These definitions don't make sense in that if non-active is NO relapses and NO inflammatory MRI activity, then why is active NO relapses but YES, evidence of inflammatory MRI activity.
This further doesn't make sense when the FDA writes the following about active SPMS: "active SPMS...which is considered a relapsing form of MS (RMS) and for which there are approved therapies." (FDA tolebrutinib CRL, page 3).
How can the FDA say the or their definition of active SPMS is one in which there are NO relapses but then say active SPMS "is considered a relapsing form of MS (RMS)"?
Maybe this is correct and makes sense, and thus, I do not understand correctly, but I have been very puzzled and stumped by this. I would appreciate your thoughts and insight here.
Thank you so much!
Yes. Sanofi has no grounds to seek a PPMS license for tolebrutinib IMHO in light of the failed PERSEUS trial.
My understanding of the tolebrutinib for nrSPMS status with the EMA at this time is that the CHMP issued a positive opinion but that the EMA decision has not been made or at least announced yet. My understanding here could be wrong, though.
EMA decision is a rubber-stamping exercise, i.e. part of the EMA bureaucracy. Once the CHMP gives the green light, it is almost certain to be approved. I don't think it is a mistake. We must not let underpowered subgroup analyses mislead us. To quote Richard Peto one of the giants of British medicine: “The appropriate interpretation of apparently different results in different subgroups of trial results is still one of the most difficult matters of judgement in the interpretation of randomised evidence; at present, many clinicians and regulatory agencies pay far too much attention to irregularities between the apparent effects in different subgroups, to the potential detriment of the care of individual patients” (Peto, 2011).
“It is for this reason why I think higher doses of ocrelizumab are better than lower doses. I think this is due to deeper B-cell depletion and possibly greater CNS effects with higher doses”
Does this suggest high dose ocrelizumab could be beneficial for those with smouldering MS?
Only in those with active MS. We need CNS penetrant therapies. Either using a shuttle to get more anti-CD20 monoclonal antibodies into the CNS, BTKis, CAR T-cells etc.
There is a P1 study by Roche to use a brain shuttle with an anti CD20 (probably ocrelizumab?). The treatment is called RO7121932. Results expected next year.
“Is this because Gd-enhancing lesions are an indicator or message from the immune system that whatever is causing MS is active and modifiable by anti-CD20 therapy? Could this be related to EBV?”
We seem to be far way from any treatments that effectively tackle smouldering MS / progression. MS researchers (Nero) seem to be fiddling around with anti-CD20s whilst the CNSs (Rome) of MS patients burns. Where are we with therapies to purge the CNS of EBV? What’s happening to the trials of anti-virals targeting EBV?
Why aren’t MS patients (150k) seeing the same advances as in cancer?
https://www.bbc.co.uk/news/articles/cx214vld41ko
There are three MS antiviral trials that started in 2026.
There will be a readout next year from the trials, and 2 are phase 3 so this means faster approval. 2 MS trials are testing Tenofovir Alafenamide for MS-EBV.
A 3rd trial is testing spironolactone and famciclovir.
The trials are in Australia and Norway.
My understanding and impression as to why MS patients are not seeing the same advances as cancer patients is because tackling neurodegenerative we see the same challenge playing out in other neurodegenerative diseases that lack meaningful treatment options. The fact that MS is framed, per the conventional narrative, as a primary autoimmune disease is very misleading and deceitful in my opinion, it intentionally downplays the neurodegeneration that occurs in MS, which per my understanding of the research, is the primary driver and cause of long-term progression and disability accumulation.
I totally agree. The money was in anti-relapse drugs and pharma chased the money. Unfortunately, the MS researchers were also happy to follow the money. No one is willing to rock the boat / bite the hand that feeds them. Two things you’ll never see in this life: a neuro living in social housing or a neuro who can point to a patient who has got better / stabilised (MS, MND, dementia…). Michael J Fox has raised over $1 billion for Parkinson’s research yet I don’t see any cure on the horizon.
The "medical-pharmaceutical-industrial" complex has to keep the conventional/convenient MS narrative going-that MS is a primary disease consisting of relapses and/or new conventional MRI activity-the 2 aspects that Big Pharma's drugs show efficacy against.
What they and their narrative doesn't and won't tell PwMS:
-relapses aren't the driver of long-term progression and disability accumulation
-the clinical-radiological paradox of MS
-the differences between NMOSD and MOGAD, where progression is driven by relapses, which is not the case for MS
-the neurodegeneration in MS (PIRA) and that their drugs aren't effective against it
-that MS may likely not be a primary autoimmune disease-one example: no target antigen(s) have ever been found despite decades of searching
-the Numbers Needed to Treat for their drugs for MS (some are as high as in the 20s; some may be even higher), meaning that not everyone who takes an MS drug receives the alleged "benefits"
-that MS is a progressive neurodegenerative disease
-that there is SO MUCH pathology and disease activity in MS that can't be seen on conventional MRI
I could go on and on with this list, but I will stop here.
The current MS drugs are over-hyped, and their "benefits" are exaggerated, distorted, and misrepresented. It's wrong, unethical, and harmful, IMHO
Multiple sclerosis is not autoimmune. The immune system tries to do its job to clean up the mess that's left by whatever is causing the pathology.
I agree, but pretty much any info one reads or finds about MS describes it as an autoimmune disease, and the MS drugs are all anti-inflammatory treatments based on MS as an autoimmune disease is my understanding, which is why they don't work (IMHO and per my research; happy to provide evidence upon request) against the PIRA and neurodegeneration in MS, only the active focal inflammation disease activity.
Have you heard of the inside-out vs outside-in theories of MS? Maybe you would be in the inside-out camp, then. Have you heard of Dr. Peter K Stys and his research alleging the inside-out theory of MS? Highly recommend reading.
I believe the "autoimmune" activity is a reaction to a primary neurodegenerative pathology in MS personally.
Pathology... like an infections agent, EBV?
I downloaded your slides from ECTRIMS - "Efficacy and Safety of Ocrelizumab vs Placebo in Primary Progressive MS: Results of the Phase IIIb ORATORIO-HAND Study" and I read through them. Very interesting, thank you. Please share a link related to your data about wheelchair users or impacts data from your oratorio study on non-active PPMS patients.
This data is directly applicable to me: I am in the US, 55 yo male, progressive MS since 2001. Ocrevus for the last six years. Clear progression despite Ocrevus use in terms of balance and walking and timed 25 ft walk test. No active lesions on imaging the past four years. I have been on standard dosing of every six months, and I am looking to moving to reduced dosing schedule of once a year if my next MRI shows no new lesions and evaluating at a later time discontinuing Ocrevus.
I know given past comments in your writing and your papers that you are against extended interval dosing and with these Oratorio results discontinuation of drug (at least if have active disease as defined by lesions on imaging). A fair statement or am I mischaracterizing?
Of course I don't know what my disease presentation would look like if I had never been on Ocrevus the last six years, so it's hard to compare A to B and this hypothetical. I'm confident it has had a positive benefit within the limitations of the drug, but perhaps because it is not penetrant to the CNS, it's had minimal effect as compared to a drug that might be more penetrating?
if I had additional data to review related to impacts on non-active PPMS patients, it might tell me something about EID or discontinuation especially as it relates to hand function. I think the results are promising for peope with more active disease but I am still unclear on numbers regarding non active PPMS. Thank you
I'm still hopeful that CN|M Au8 will have a positive impact on my PPMS. Its such a shame that Tolebrutinib will never be available to pwPPMS. Never feels like too negative an outlook but is there any way back Tolebrutinib as a treatment for PPMS?