Finally, after a week or more of thinking and contemplation my opinion about the ofatumumab vs. teriflunomide trial data (ASCLEPIOS I and II); another of my ECTRIMS highlights.
The result of the ASCLEPIOS I and II are not unexpected and in line with the treatment effects of anti-CD20 therapies with some caveats.
- Both ASCLEPIOS I and II studies met their primary endpoints in patients with relapsing forms of MS (RMS); overall ofatumumab (OMB157), a subcutaneous, potent, fully-human antibody targeting CD20 positive B-cells, delivered efficacy with a favorable safety profile
- RMS patients on ofatumumab had a reduction in annualized relapse rate (ARR) by 50.5% (0.11 vs. 0.22) and 58.5% (0.10 vs. 0.25) compared to Aubagio®* (teriflunomide) (both studies p<0.001) in ASCLEPIOS I and II studies respectively
- Ofatumumab showed highly significant suppression of gadolinium (Gd) T1 lesions when compared to Aubagio®, demonstrating a profound suppression of new inflammatory activity
- Ofatumumab showed a relative risk reduction of 34.4% in 3-month confirmed disability progression (CDP†) (p=0.002) and 32.5% in 6-month CDP (p=0.012) versus Aubagio® in pre-specified pooled analyses
- Ofatumumab, if approved, will potentially become a treatment for a broad RMS population and the first B-cell therapy
My interpretation:
Inflammation: relapse rate, focal MRI activity (Gd-enhancing & new T2 lesions) and neurofilament data.
I have made the point that these three markers measure focal inflammation, driven by adaptive immunity, and there is little doubt that ofatumumab is superior in suppressing inflammation compared to teriflunomide. Does this make ofatumumab superior to other very high efficacy DMTs, such as natalizumab, rituximab, ocrelizumab, alemtuzumab and HSCT? I suspect not. To prove this we would need head-2-head studies. I also think there are floor effects on these outcomes, i.e. you can only reduce relapse rates to around 0.1 to 0.2 and no lower. Why? I suspect some relapses are pseudo-relapses and are due to intermittent symptoms in relation to infections, fatigue and possibly hidden symptoms.
Please note that I don’t consider peripheral blood neurofilament levels (pbNFL) to be a neurodegenerative marker in the context of MS. All the data I have seen to date indicates that it is linked to focal inflammatory activity. Clearly more needs to be done in progressive MS with pbNFL to understand what it means in inactive or smouldering MS.
End-organ damage: disability progression and brain volume data
I was disappointed with how ofatumumab did against teriflunomide in delaying disability progression and reducing the relative loss of brain volume. This will be ofatumumab’s Achille’s heel. Why? It is clear that MS the disease is not focal inflammation; I have made the point that based on the Prentice criteria, both relapse and focal MRI activity don’t predict disability outcomes in natural history studies and placebo arms of clinical trials. If focal inflammation was MS then relapses and focal MRI activity would predict outcome whether or not you are on a DMT. The point I making here may be a philosophical one, but it a very important one.
In comparison, sustained or confirmed disability progression has to be MS and is based on the pathological correlates that define MS (demyelination, neuroaxonal loss and gliosis).
Why did ofatumumab do so poorly on these metrics relative to teriflunomide? It could be that teriflunomide is the outlier and this opinion is based on several observations.
- Teriflunomide is the only DMT to have a consistent effect on disability progression; i.e. both teriflunomide phase 3 placebo-controlled trials were positive on this outcome. In addition, the treatment effect or impact of teriflunomide on disability progression has always been greater than what you would expect from its impact on relapses. For the tuned-on readers, you would have noticed the same disconnect between relapses and disease progression was observed in the ponesimod vs. teriflunomide trial.
- Teriflunomide also has a significant effect on brain volume loss compared to placebo, which again is out of proportion to its impact on relapses (see picture below).
- Teriflunomide is more effective when used 2nd and 3rd line. Teri is the only DMT to show the latter and this observation was seen in both phase 3 studies, which makes it likely to be a real, and a very important, finding.
- Teriflunomide is a broad-spectrum antiviral agent, which may be part of its mode of action in MS. Could teriflunomide be targeting the viral cause of MS independent of its effects on the immune system’s response to that virus? This needs more study, but teriflunomide is the outlier, or exception, that disproves the dogma.
Is ofatumumab being underdosed?
Ofatumumab is being given at a dose of 20mg subcutaneously monthly. This dose was chosen to keep B-cells depleted, but not severely depleted, so as to allow rapid repopulation of peripheral B-cells numbers if ofatumumab is stopped. In other words, B-cell depletion is relatively mild compared to ocrelizumab 600mg every 6 months. With ocrelizumab, it takes 6 months or longer to start to see B-cell reconstitution.
I don’t buy this argument. The repopulation kinetics with ofatumumab are based on relatively short-term dosing studies in which deep tissue and in bone marrow B-cell depletion is likely to be relatively modest. I suspect with long-term dosing with ofatumumab deep tissue and bone marrow B-cell depletion is more likely and hence the B-cell repopulation kinetics will mimic that of rituximab and ocrelizumab.
I also think rapid B-cell repopulation is likely not to be relevant as the new B-cells will almost certainly be bone marrow-derived naive B-cells and not memory B-cells.
The question I have is the 20mg per month of ofatumumab sufficient to penetrate the CNS and clear the intrathecal of CNS resident B-cell follicles?
At the AAN this year Stephen Hauser presented data indicating that when it comes to disability progression, not relapse rate or MRI activity, the extent of exposure to ocrelizumab is very important.
The greater the ocrelizumab exposure the more effective it was at delaying disability progression. This could be related to deep tissue (peripheral) and end-organ (central) B-cell depletion. There is mounting evidence that the B-cells and plasma cells within the brain and spinal cord of MSers are driving some of the slow-burn we see clinically and on MRI (smouldering MS). What I am saying is that ocrelizumab could be superior to ofatumumab when it comes to scrubbing the brain clean of pathogenic B-cell follicles. Therefore it more important than ever to test this hypothesis in a head-2-head study of ocrelizumab vs. ofatumumab (OVO study) or the DODO study comparing double-dose (1200 mg) vs standard-dose (600 mg) ocrelizumab (DODO study) to see if the higher dose of ocrelizumab has a bigger impact on the intrathecal B cell response than the standard dose.
I would suggest these studies include next-generation MRI and other biomarkers to test the CNS penetration hypothesis. If these studies are positive, i.e. ocrelizumab is superior to ofatumumab and double-dose ocrelizumab is superior to single-dose ocrelizumab, it will not only tell us a lot about how anti-CD20 therapies work in MS, but it may answer the question of whether or not we need to target the intrathecal or CNS B-cell response in MS. The latter hypothesis is being tested by our group in two studies at present. We would love to add a third and fourth study to the portfolio. If you work for Novartis or Roche please tell the powers that we are really, really, interested in doing both the OVO and the DODO studies.
What about teriflunomide?
Don’t forget that the implications from the ponesimod vs. teriflunomide and ofatumumab vs. teriflunomide trials are quite profound. Teriflunomide is quite a remarkable DMT and we need to explore its antiviral effects in MS in more detail and understand what it is doing in MS independent of its rather weak anti-inflammatory effects. This is why I have proposed using teriflunomide as a maintenance therapy post-induction. In my ECTRIMS hot topic presentation, I called the trial the iTeri study (see slide show above).
If you work for Genzyme-Sanofi please tell the powers that be that we are really, really, interested in an induction-maintenance trial with both teriflunomide (iTeri study) and a second with your BTK inhibitor (iBruT study).
CoI: multiple
Prof G with the DoDo study aren’t you concerned with safety? Recently you seem to be trashing anti-CD20 as a class based on safety concerns.
I have just posted the following response to a similar query:
Anti-CD20 therapies are a game-changer in that many more people will be treated early with highly effective therapies. Anti-CD20 therapies are also relatively safe in the short to intermediate-term. All I am concerned about is the long-term safety of prolonged and profound B cell depletion. It is best to acknowledge this risk and deal with it before it becomes a major problem. One way of addressing is this is via our proposed ADIOS trial, optimising vaccinations before starting anti-CD20 therapy and monitoring and managing the hypogammaglobulinaemia when it emerges. All DMTs have risks and most of these risks can be derisked, which is a good thing.
What I am trying to tell the companies that market anti-CD20 therapies in MS that they need to avoid the Ostrich Syndrome (head-in-the-sand syndrome) and proactively manage and study the risks of prolonged B-cell depletion, otherwise the problem will come back and bite them (known-unknowns are vicious).
My longterm safety concerns are not going to change my use of anti-CD20 therapies only how I monitor and deal with the potential long-term complications and how I counsel my patients.
With regard to the DODO study; double-dose may only be required for a short period of time to scrub clean the CNS of B-cells. Once the latter has been achieved the dose could be lowered and or the dosing interval extended. It is clear we have a lot to study.
Prof G,
“It is clear that MS the disease is not focal inflammation”. You are pushing this hard at the moment. It’s quite discouraging for MSers who opted for a highly effective therapy to suppress relapses and activity on MRI to now be told we are deluding ourselves as these treatments are not targeting MS just the side effects of the disease. More worrying is that underlying disability progression will eventually show its face.
It doesn’t surprise me at all that highly effective anti inflammatory therapies aren’t targeting the real disease. When Campath (Alemtuzumab) was tried on SPMS patients in the early 1990s MRI activity was suppressed but MSers still experienced “unrelenting cerebral atrophy”.
Please can you highlight what, if any, trials are underway to target real MS. My overall impression is that the MS research world targeted the easy pickings (relapses, MRI activity) and pharma were quite happy to develop expensive drugs to target these aspects of the disease. 25 years after the first MS therapies were licensed, we now know that these treatments are piss poor at stopping the real disease (Hauser’s comment about patients on Ocrelizumab still becoming SP should have given a clue). Am I right to angry that once again thousands of MSers are likely to have a bleak outlook because of lazy researchers and greedy pharma?
Paul, I don’t think you should be too hard on Pharma and the MS community. Many of these insights have only recently emerged and are still emerging. I suspect many of my colleagues will disagree with me. I am just one opinion in a sea of opinions. It is clear that we need more than just anti-inflammatories to treat MS. The combination therapy paradigm is finally gaining traction. I realise you are angry and disappointed, but funders and science move very slowly. As I write this I am at home working on a grant application to tackle the ‘real MS’ or as I call it ‘smouldering MS’.
Prof G,
Thanks for your response. My target isn’t you, but your blog provides an outlet for the frustration I feel. Having the disease and researching the disease are very different things. I’m 18 years into this disease and am guessing you are 30 years into researching it. A real breakthrough would both allow us to go to bed without thinking about this wretched dishes. Good luck with the application.
Very interesting about Teri, the anti-viral how would one go about investigating if this is the moa. EVB-viral load? Will you be posting something on the updated mavenclad data? If 75% was free of progression 7 years after the first dose the effect on BVL should be significant no?
EBV shedding is one way. Unfortunately, this does not get to whether or not it has effects within the CNS.
Why not add Aubagio to a first line like Tysabri??? Especially for those of us who were started early and have been stable for a long time. This is my worry like the reader above. I too am 14 years into this journey and am fortunate to have been placed on Tysabri from the get go. I worry daily about the smoldering MS and if my neuroradiologist will catch this before I show any real symptoms.
Thank you as always for your insights into this disease.
About the ocrelizumab-time-disability progression: is this disability progression independent of what is caused by relapses and new lesions? Otherwise it could just be that it takes time for ocrelizumab to work. It is known that relapses and new MRI lesions occur in the first few months when starting ocrelizumab treatment.
Also, is this a meta-analysis for dosing? Are groups similar in baseline characteristics?
I think that measuring BVL over the short term 12 – 24 mo. is fraught with hazards.
The measurement methods and equipment vary between testing facilities.
The measurement methods are somewhat crude and vary with repeats and hydration levels of subjects, etc.
This leads to very noisy data.
The rate of BVL is not linear vs age, within the healthy cohort, much less the MS group.
Comparing BVL between DMTs with all these variables is a noble cause, but should be looked at with the proper perspective.
So while I think this is one of the best metrics, it’s real value is the long term data over several to many years. It should be verified independently with clinical metrics over a larger user population and should become a more generally accepted annual measure along with MRI for most patients.
Unfortunately people with MS don’t want to wait too long while their brains are turning to mush.
I guess your maven friend Antonio Scalfari friend thinks otherwise
“In the multivariate analysis, the only variables significantly predicting faster cortical thinning were the accumulation of new CLs (-0.26886; p< 0.001) and new relapses"
"Accelerated cortical thinning distinguishes patients with more severe disease progression and is associated with more prominent focal cortical pathology and frequent relapses"
Disease activity, brain atrophy and long term clinical outcome of Multiple Sclerosis
http://www.professionalabstracts.com/ectrims2019/iplanner/#/presentation/398
I suspect the majority of these patients are on DMTs. Yes, relapses and MRI are predictive when you are on a DMT, but not when you are untreated (natural history or placebo). A very important distinction.