#ResearchSpeak & #NeuroSpeak: the great differentiator

Anti-drug antibodies are a major problem in clinical practice and are important for safety and efficacy reasons. #ResearchSpeak #ClinicSpeak #MSBlog

How important are anti-drug antibodies in relation to biological therapies, in particular neutralising antibodies to the monoclonal therapies used to treat MS? 

This issue became a defining issue in rheumatology amongst the monoclonal antibodies targeting TNF-alpha. Infliximab the initial anti-TNF monoclonal is so immunogenic that it has to be used in combination with another immunosuppressive, typically methotrexate or azathioprine, to try and prevent NABs. In comparison, the incidence of NABs on the newer generation anti-TNFs, for example adalimumab, is much lower. 

It will be interesting to speculate how important anti-drug antibodies (ADA), in particular neutralising antibodies (NABs), will become with the anti-CD20 monoclonals in MS. The incidence of NABs to rituximab is in the range of 6-9% compared to 0.2-0.5% for ocrelizumab and ~0% on ofatumumab (Phase 2 trial). We know that NABs are a major problem with biologicals in terms of both efficacy and safety. I would therefore predict that as in other areas, such as the anti-TNF-alphas, NABs will be a major differentiator amongst the anti-CD20s; this won’t be simply for marketing reasons, but it will be driven by efficacy and patient safety concerns. NABs blunt efficacy and are a common cause of treatment failure, and are also responsible for infusion reactions. Just ask any pwMS who has been unfortunate enough to develop anti-natalizumab antibodies. 

We also think NABs are a major problem with alemtuzumab as well (please watch this space). We have just set-up an anti-alemtuzmab antibody assay and will soon be using this assay in clinical practice and will make it available for other groups to use. We plan to extend our panel of NAB testing from anti-natalizumab and anti-alemtuzumab antibodies to include anti-rituximab and anti-ocrelizumab binding and neutralising antibodies. Knowing who has NABs affects clinical practice. 

Objective: To evaluate baseline prevalence and post-baseline incidence of anti-drug antibodies (ADAs) in ocrelizumab-treated patients from two identical Phase III, randomized, double-blind, double-dummy, IFNβ-1a-controlled trials of ocrelizumab in relapsing MS (OPERA I and OPERA II). 

Background: Ocrelizumab is a humanized monoclonal antibody (mAb) that selectively targets CD20+ B cells. Humanization is believed to potentially reduce mAb immunogenicity, a measure of which is the formation of ADAs. Theoretically, a lower immunogenicity risk is expected with humanized antibodies like ocrelizumab (85-90% human protein) compared with chimeric antibodies (typically 65-75% human protein), which could potentially improve tolerability and efficacy with long-term treatment of chronic diseases. In the treatment period of a Phase II study of ocrelizumab in relapsing-remitting MS, no treatment-emergent ADAs were detected. 

Methods: Patient serum samples were evaluated for ADAs at baseline and every 24 weeks throughout the treatment period. ADAs were detected in a validated bridging ELISA-based screening assay with appropriate sensitivity and drug tolerance. Screened positive samples were confirmed for specificity by competition with ocrelizumab and further titered. Confirmed positive samples were tested in an antibody-mediated cellular cytotoxicity-based neutralizing antibody (NAb) assay to determine the neutralizing capabilities of ADAs. 

Results: Overall, >96% of ocrelizumab-treated patients were evaluable for baseline prevalence and post-baseline incidence of ADAs in OPERA I and OPERA II. Baseline prevalence of ADAs among evaluable ocrelizumab-treated patients, but prior to ocrelizumab infusion, was 0.3% (1/396) in OPERA I and 1.0% (4/402) in OPERA II, which is expected based on assay specificity. Incidence of treatment-emergent ADAs post-baseline was 0.2% (1/402) and 0.5% (2/405) in OPERA I and OPERA II, respectively. Among these ADA-positive patients, one tested positive for NAbs.

Conclusions: There was a low incidence of immunogenicity in the ocrelizumab groups in the OPERA I and OPERA II studies. Supported by F. Hoffmann-La Roche.

CoI: multiple

About the author

Prof G

Professor of Neurology, Barts & The London. MS & Preventive Neurology thinker, blogger, runner, vegetable gardener, husband, father, cook and wine & food lover.


  • Prog G, Would you do Alemtuzumab antibodies for other centres? Would you treat with ocrelizumab after alemtuzumab?

    • Re: "Would you do Alemtuzumab antibodies for other centres?"

      Yes, once we have validated the assay and made sure it running to GLP (good laboratory practice). We will keep you posted.

    • Re: "Would you treat with ocrelizumab after alemtuzumab?"

      Yes, post-alemtuzumab and post-rituximab.

    • So the question arises is the failure of the fingolimod to lemtrada approach real as ocrelizumab depletes ina similar way to lemtrada. However advantage is you get a second does of ocrelizumab 6 months after the first. The split month first dose used in early trials could have been an advantage here.

      However, it highlights an issue and this is when you plan a switch to a new drug do you have a plan how to transission off to another.

  • The antibody story is an interesting one… Because for the first 10+ years of TNF-alpha inhibitors we did not have an option of testing for anti-drug antibodies in the clinic. The effect of the drug could be less with years, we would switch to another drug and so on…
    Then the competition kicked in and the companies had to show that their product was better than all the others.
    Now we have an option of routine testing so we do it in all patients… what do we find? Yes, in some patients there are antibodies that are responsible for the the drug being less effective(not common). The most common reason for the drug level reading 0 is… non-compliance. Life gets in the way, people do not take thier drugs as prescribed.
    Sorry for the digression, back to rituximab… so say 6% of rituximab patients get the antibodies per year of treatment and have to switch them to ocrelizumab… how many more patients could we treat with rituximab to NEDA for the cost of one on ocrelizumab? ( would Roche give us some quantity-rebate on its new drug please?)

    • Re: "how many more patients could we treat with rituximab to NEDA for the cost of one on ocrelizumab?"

      In the UK it will be zero as the NHS will not allow us to use rituximab in MS. We tried to get rituximab into our guidelines, but our business case was rejected.

    • So you know if you are taking alemtuzumab, is it safe to say you will probably will be Nab positive at some stage (about 80%. Its humanised whats going on), but for most it won't have impact in terms of depletion and severity infusion (which occur in about 90% of people) reactions in the initial treatment cycle. You will start with no Nabs when you start first cycle and few will still have then before second infusion, but if you need third infusion or more then I suspect influences to occur and as been said they can be "problematic"..why no publication..it's safety we are talking about.

      As for Babs no wonder people get infusion reactions their frequency is so high and they appear rapidly within a month of first infusion in the majority of people. Not sure why it was not very clear in the European Label, in contrast to the American label which spells it out…maybe a difference between first and third line or maybe American Companies don't care about us Europeans and we are not that important.

      Not clear why one has to dig to find this important information.

      If you are not depleting, it is likely to be a Nabs issue. Ask for details after your treatment cycles

    • MD, Lemtrada seems to be the worst in class. (maybe because it is the oldest). We never needed huge doses of SoluMedrol with infliximab or rituximab (with rituximab 125 mg i.v., with Lemtrada 1000 mg i.v.)
      Could a smaller dose do? I don't think so, I was getting a rash on 500mg the last 2 days. Then again it was probably some cytokine realease reaction and not the antibodies.

    • Worst in class, tell that to the people who are doing very well and past their four years monitoring I suspect they are happy getting on with their lives.

    • As to rashes it is the nature of the beast. Rituximab depletes a fraction of the cells that lemtrada depletes. Bcells are about 10-20 percent so lemtrada kills 5 times more cells causing them to liberate their contents. It also destroys monocytes which are known to have toxic contents so whilst rituximab will cause infusion reactions they will be in a different league to lemtrada. The cytokine storm from lemtrada will reactivate old lesions. This is why people get a cocktail of anti histamines steroid paracetamol etc to reduce these but still 90 percent get infusion reactions. It is amazing that Cambridge group carried on in view of this and autoimmunities.

      On first infusion there will be no antibodies and few will have antibodies for second infusion. It will be more possible on third infusion but reactions do occur so. These will be limited by the drug cocktail including steroids but hospitals are trained and can deal with anaphylactoid responses and have adrenaline to hand should it be needed. Thankfully these issues are rare.

  • How do you identify those who developed NABs; can you do with a blood test?
    What is the percentage of those on alemtuzumab and natalizumab who developed NABs?

    • To test for NABs you need a blood test. For natalizumab ~5-6% develop persistent NABs and everyone who gets to 12 months on treatment should have themselves checked for NABs. Regarding alemtuzumab we don't know how many people develop persistent NABs, but we suspect it is higher than 5%.

    • We don't really know because Genzyme did not properly publish the data they hold on this. Why not?

      You can find the partial answer in the FDA reference, above MD 12:10:00 pm


      Its alot of people that have them, nearly 80%. They are boosted with each cycle of injection. They rocket after the infusion and then wain over the year.

      The binding antibodies are much more frequent a titre of 6 million in some people (This means alot..I found this information elsewhere).

      The argument for ignoring I guess, is that they dont stop the drug working.

      So have an ECTRIMS astract in abbout 2012 and we can forget it.
      Is this OK?
      Why was this information not published with the pivotal trial data..What were the authors and reveiwers doing, it is common fayre for any antibody tiral data these days.

By Prof G



Recent Posts

Recent Comments