Barts-MS rose-tinted-odometer: ★★
In response to a question over the weekend about what has happened to the DODO and ADIOS studies. Both are alive and kicking. The more insightful question would be ‘how can I support both the DODO (double-dose ocrelizumab study) and the ADIOS (adaptive dosing ocrelizumab study) studies?’.
Surely, the DODO and ADIOS studies are incompatible with each other scientifically? How can I, on the one hand, support a higher dose of ocrelizumab and on the other hand suggest reducing the dose in the longterm. The hypothesis is all about timing and how you use anti–CD20 therapies.
You need higher doses of anti-CD20 therapy initially as an induction strategy to purge the various B-cell compartments of memory B-cells, which house latent EBV and the highly autoreactive population of B-cells that drive and maintain the MS-state. This population of cells may reside in the deep tissues and/or the central nervous system, which is why we are also testing CNS penetrant anti-B-cell strategies, simultaneously. Time is short so we need to run trials in parallel.
However, once you have purged these compartments say after 2 years of treatment you don’t need to maintain such high-doses of anti-CD20 therapies that are then suppressing normal B-cell biology and immune responses, which result in longterm complications. This is why we want to use ocrelizumab as an immune reconstitution therapy, i.e. high-dose upfront followed by no treatment and wait to see if MS remains in remission or disease-activity returns requiring additional courses. The latter is one of the arms of our proposed ADIOS study.
In reality, if we could convince a national funding agency, a pharma company or a wealthy philanthropist I would use anti-CD20 therapy as part of an induction-maintenance protocol. After two years of induction therapy with high-dose ocrelizumab, I would test different maintenance strategies in parallel. My agents of choice would be teriflunomide, leflunomide, IMU-838 (vidofludimus) or ASLAN003 (selective second-generation DHODH inhibitors), HAART (highly active antiretrovirals), famciclovir or another anti-EBV viral agent. The hypothesis is to allow B-cell reconstitution after anti-CD20 therapy in the presence of an anti-viral agent to prevent EBV reactivation and reinfection of new memory B cells. By doing this you will also be derisking the long-term immunosuppression associated with anti-CD20 therapies and prevent the development of hypogammaglobulinaemia. In addition, you will be allowing patients to respond to vaccines.
The problem with this trial proposal is the outcome measure; the power calculations are not trivial and the study would have to be very long. I also have reservations about whether or not the regulators will accept the induction maintenance strategy. Maybe we can sell it to them on safety, i.e. to prevent the development of hypogammaglobulinaemia and infections rather than on efficacy? If we go this route then there is only one agent we can use and that is teriflunomide, which is licensed to treat MS. As teriflunomide is coming off patent there is a chance the NHS may be interesting in funding such a trial; i.e. it would save them money. This is something I am exploring as a proof-of-concept trial.
The good news is that Roche-Genentech is testing the principles of the DODO study and announced at MSVirtual2020 two high-dose ocrelizumab trials (see below). These trials up the stakes in the anti-CD20 wars and I am confident that we need higher doses upfront to purge deep tissue and possibly CNS pools of B-cells. Please note that you don’t need higher doses of anti-CD20 therapy to suppress relapses and focal MRI activity you can do that with current or lower doses. I am confident both these studies will show that higher-dose ocrelizumab is superior to standard dose ocrelizumab on disability progression or smouldering MS, but not on focal inflammatory events. In relation to the latter, we have hit the ceiling already.
You need higher doses up-front to target the drivers of smouldering MS; i.e. disease progression independent of relapses, accelerated brain volume loss, slowly expanding lesions (SELs) and the subpial cortical lesions. If these higher-dose studies are positive it will put clear daylight between ocrelizumab and the other anti-CD20 therapies and it would mean the ofatumumab and rituximab are currently being underdosed, at least initially in the first two years. But don’t we have a hint of this already? Ofatumumab was not better than teriflunomide at slowing down brain volume loss in year two of the ASCLEPIOS I and II clinical trials (NCT02792218 and NCT02792231) despite being superior to teriflunomide on relapses and MRI activity. The latter is more proof that focal inflammatory disease (relapses and MRI activity) is not MS but in response to what is causing the disease. The real MS is what causes smouldering pathology and end-organ damage.
DODO vs. ADIOS vs. iTeri: which one would I prioritise? Almost certainly iTeri; the iTeri trial makes the most sense in terms of our current understanding of the pathogenesis of MS, mode of action of anti-CD20 therapies and the long-term risks of chronic B-cell depletion.
Hauser et al. Ofatumumab versus Teriflunomide in Multiple Sclerosis. N Engl J Med. 2020 Aug 6;383(6):546-557.
Background: Ofatumumab, a subcutaneous anti-CD20 monoclonal antibody, selectively depletes B cells. Teriflunomide, an oral inhibitor of pyrimidine synthesis, reduces T-cell and B-cell activation. The relative effects of these two drugs in patients with multiple sclerosis are not known.
Methods: In two double-blind, double-dummy, phase 3 trials, we randomly assigned patients with relapsing multiple sclerosis to receive subcutaneous ofatumumab (20 mg every 4 weeks after 20-mg loading doses at days 1, 7, and 14) or oral teriflunomide (14 mg daily) for up to 30 months. The primary end point was the annualized relapse rate. Secondary end points included disability worsening confirmed at 3 months or 6 months, disability improvement confirmed at 6 months, the number of gadolinium-enhancing lesions per T1-weighted magnetic resonance imaging (MRI) scan, the annualized rate of new or enlarging lesions on T2-weighted MRI, serum neurofilament light chain levels at month 3, and change in brain volume.
Results: Overall, 946 patients were assigned to receive ofatumumab and 936 to receive teriflunomide; the median follow-up was 1.6 years. The annualized relapse rates in the ofatumumab and teriflunomide groups were 0.11 and 0.22, respectively, in trial 1 (difference, -0.11; 95% confidence interval [CI], -0.16 to -0.06; P<0.001) and 0.10 and 0.25 in trial 2 (difference, -0.15; 95% CI, -0.20 to -0.09; P<0.001). In the pooled trials, the percentage of patients with disability worsening confirmed at 3 months was 10.9% with ofatumumab and 15.0% with teriflunomide (hazard ratio, 0.66; P = 0.002); the percentage with disability worsening confirmed at 6 months was 8.1% and 12.0%, respectively (hazard ratio, 0.68; P = 0.01); and the percentage with disability improvement confirmed at 6 months was 11.0% and 8.1% (hazard ratio, 1.35; P = 0.09). The number of gadolinium-enhancing lesions per T1-weighted MRI scan, the annualized rate of lesions on T2-weighted MRI, and serum neurofilament light chain levels, but not the change in brain volume, were in the same direction as the primary end point. Injection-related reactions occurred in 20.2% in the ofatumumab group and in 15.0% in the teriflunomide group (placebo injections). Serious infections occurred in 2.5% and 1.8% of the patients in the respective groups.
Conclusions: Among patients with multiple sclerosis, ofatumumab was associated with lower annualized relapse rates than teriflunomide. (Funded by Novartis; ASCLEPIOS I and II ClinicalTrials.gov numbers, NCT02792218 and NCT02792231.).