HSV-2 encephalitis with Ocrelizumab

The effect of DMTs on lymphocytes causing lymphopenia (from E Fox et al. Neuol Clin Pract Feb 2019)

Why does this not come as a surprise to all and sundry, the report of a potentially harmful infection following immunosuppressive treatment, but this time from the new kid on the block, ocrelizumab (Ocrevus). Not to sound cliched, but if you hadn’t seen this one coming, well then the rest is history.

Risk is all about perception; what is reasonable for one person is unacceptable for another. Society, is inherently risk averse, directed by a set criteria for risks and risk mitigation. But in the land of MS, have we taken a risk too far?

Dudek et al., burst our safety bubble with the report of a case of HSV-2 encephalitis in Germany, with normal MRI head scan. The individual is 61 years old with secondary progressive MS; relapsing-remitting in 1997, secondary-progressive course in 2005 and treated with mitoxantrone until mid 2015. Ocrelizumab was started towards the end of 2018 (so this developed on ocrelizumab). Roughly 4 weeks after the second course of ocreluzimab he experienced a focal seizure involving the right arm. A subsequent CSF examination revealed an increased cell count, predominantly lymphocytes (WCC 559/ul; 546/ul mononuclear cells), and PCR positivity for HSV-2. Aciclovir was started, as well as anti-epileptics with recovery. A follow-up CSF examination revealed normalization of the cell count and drop in HSV-2 copy numbers to 1206 copies/ul. The CD19 lymphocyte level was 15.9% after the first dose of ocrelizumab and had dropped to 5% at the time of infection.

Ocrelizumab is a recombinant humanized monoclonal antibody directed against CD20-expressing B-cells; it results in antibody-dependent cellular cytolysis following cell surface binding to B lymphocytes (see figure above). In RRMS trials 58% of ocrelizumab treated individuals experienced one or more infections compared to 52% with Rebif; in PPMS trial 70% of ocrelizumab treated individuals experienced one or more infections compared to 68% on placebo (dummy tablet).

In active-controlled (RMS) clinical trials, herpes infections were reported more frequently in OCREVUS-treated patients than in REBIF-treated patients, including herpes zoster (2.1% vs. 1.0%), herpes simplex (0.7% vs. 0.1%), oral herpes (3.0% vs. 2.2%), genital herpes (0.1% vs. 0%), and herpes virus infection (0.1% vs. 0%). Infections were predominantly mild to moderate in severity. There were no reports of disseminated herpes. In the placebo-controlled (PPMS) clinical trial, oral herpes was reported more frequently in the OCREVUS-treated patients than in the patients on placebo (2.7% vs 0.8%)” (Genentech prescribing information).

As ocrelizumab is relatively new (FDA approval March 2017, EMA Jan 2018), long-term data is sadly lacking. It is important to note that unlike other DMTs, following the first infusion of ocrelizumab B cell counts are reduced within 2 weeks and remain at a low level throughout the treatment. B cells recover to baseline/lower limit of normal only 2.5 years after discontinuation in 90% of individuals.

Given that HSV-2 encephalitis is a serious outcome and that the MRI head scan is likely to be normal (possibly because of the lack of inflammatory response that results in the FLAIR abnormality on MRI), the safest scenario maybe to investigate for subclinical HSV in the CSF prior to starting ocrelizumab.

Letter to the editors
J Neurol. 2019 May 21. doi: 10.1007/s00415-019-09391-0. [Epub ahead of print]
Dudek MIR, Thies K, Kammenhuber S, Bösel J, Rösche J.

HSV-2-encephalitis in a patient with multiple sclerosis treated with ocrelizumab.

Here, we report the case of a 61-year-old man (K.H.) with secondary progressive multiple sclerosis, who suffered from HSV-2-encephalitis under treatment with ocrelizumab. First symptoms of a relapsing–remitting multiple sclerosis occurred in 1997. In November 2005 he developed a secondary progressive course, and hence next year the diagnosis of that type of multiple sclerosis was established. He had developed left-sided spastic hemiparesis and neurogenic bladder dysfunction. Despite treatment with mitoxantrone until May 2015 he developed spastic tetraparesis and in 2017 dysarthria. In September 2018 ocrelizumab was started with an initial dose of 300 mg and a second dose of 300 mg at the end of the month. Against spasticity he had received 4-aminopyridine for several years and triamcinolon intrathecally once a year. No other immunosuppressive or immunomodulatory drugs were applied between the termination of mitoxantrone and the start of ocrelizumab. The last treatment with triamcinolon was 3 months before the first administration of ocrelizumab. Nevertheless, he was unable to walk and his left arm was almost plegic. About 4 weeks after the second application of ocrelizumab he developed severe headache and the following day a bilateral clonic seizure with movements mainly of the right arm.

About the author

Neuro Doc Gnanapavan


  • “But in the land of MS, have we taken a risk too far?”

    “he developed spastic tetraparesis and in 2017 dysarthria.”
    “developed severe headache and the following day a bilateral clonic seizure with movements mainly of the right arm.”

    No..but what’s the point with Ocrevus here..?

    “he was unable to walk and his left arm was almost plegic”

  • I honestly have gnawing fears about this drug. Given the failed lupus and rheumatoid arthritis trials (forced to stop it because of opportunistic infections) and the extremely high cancer signal during the progressive MS trials (an eye-popping 1 in 50), there seems to be much legitimate reason for concern here. I certainly hope my fears are not borne out, but I’m afraid of what we might see in the next couple of years in regards to patients taking Ocrevus long-term…

    The one consoling factor is the robust safety profile of Rituxan, another B cell depleter, but the two drugs attack B cells in very different ways…

    • I believe there is a trial design planned that would allow you to stop Ocrelizumab i.e. de-escalation strategy.

      But this year at AAN 2019 some worrying animal data was presented:

      S26.004: Functional Characterization of Reappearing B cells After Anti-CD20 Mediated B cell Depletion in Two Animal Models of Multiple Sclerosis at Annual Meeting 2019.

      Results – In both models, a fraction of CD20+ antigen-experienced B cells persisted in splenic follicles despite extensive systemic anti-CD20 application. Upon treatment cessation, CD20+ B cells simultaneously repopulated in bone marrow and spleen before their reappearance in blood. Returning B cell population, showed a shift towards mature, differentiated phenotypes containing an increased frequency of myelin-reactive B cells with restricted B cell receptor gene diversity and enhanced capability to activate myelin-specific T cells in the model in which B cells are intrinsically activated. By contrast, in the T cell-mediated EAE model, B cells reappeared in a predominantly naïve status with a relative loss of pathogenic APC function.

      This suggests that stopping the infusions may not be a good idea!

      • First part of animal study already published https://www.ncbi.nlm.nih.gov/pubmed/30194232

        It is in my opinon one of the worse papers I have ever read, despite being written by big swingers….full of factual problems and they cannot even remember the data of their own past papers, which make a mockery of their own conclusions.

        If interested first read the paper then read below


        • You realise this is going to take ages to decipher in humans?! For a drug that is supposedly anti-B cell not surprisingly has significant anti-T cell effect. Amit Bar-Or presented findings from the OBOE study that suggests this:

          “Pretreatment CSF and serum NfL levels correlated strongly (r=0.78; p<0.001). Both serum and CSF NfL levels correlated with numbers of T1 gadolinium-enhancing lesions and new/enlarging T2 lesions on brain MRI. OCR significantly reduced serum NfL (−13.1%, −18.6% and −30.8%), CSF NfL (−24.5%, −40.0% and −54.7%) and CSF B cells (−85.5%, −84.8% and −94.0%) at Weeks 12, 24 and 52, respectively. CSF T cells were reduced by ≈60% across the same time”.

        • It was mentioned at AAN briefly in one of the platform discussions, no details as yet.

    • Wheelchair, your comment has been haunting me for a few days. Are you on O, or something else? How do you base your “legitimate reason for concern”? Where can I learn more?

      • George, if you go to the Wheelchair Kamikaze website and search for “ocrelizumab” you’ll find several articles I wrote when the drug was first approved that offer in-depth analysis of the drug and its history.

        These articles don’t go specifically into my reasons for concern, but do hit on all of the pertinent points. In a nutshell, it’s important to remember the history of the drug. It was initially trialed for the treatment of lupus, rheumatoid arthritis, and multiple sclerosis. The lupus and rheumatoid arthritis trials were forced to stop because of serious opportunistic infections in trial subjects. These stoppages represented a near disaster for the drugs developer, Genentech/Roche. The multiple sclerosis trials were ultimately allowed to continue because of the perception that MS patients have a higher tolerance for risk than other patient populations. Hardly comforting logic.

        Furthermore, the cancer signal in the progressive MS trials was extremely high. In fact, more than 1 in 50 progressive MS trial subjects developed cancer within three years of starting ocrelizumab. Now, as many researchers and neurologist assert, this number could very well be an outlier and not indicative of the true cancer risk of ocrelizumab. Indeed, this metric was not seen in the relapsing MS ocrelizumab trials, although there was a much less significant cancer signal detected. Furthermore, follow-up studies have not confirmed this number, and, thankfully, we have not seen an uptick in cancers in patients put on ocrelizumab since the drug was FDA approved. However, most patients have not been on the drug long enough to make any determination one way or another. I doubt that the true risk is anywhere near one in 50, however, even double that would be clearly unacceptable. If the progressive MS cancer numbers do turn out to be a statistical anomaly, then, hallelujah! Still, my feelings are that some degree of caution is warranted.

        I have spoken privately to some MS neurologists who have voiced concern about the drug due to the above mentioned reasons. It must be said that these same neurologists are prescribing the drug despite their reservations. Genentech has provided many US doctors with robust incentives for prescribing the drug. I won’t go further than that statement here, but I suggest all US patients look up there doctors on the “Dollars for Docs” website. Just Google it. The results may be enlightening.

        As I stated in my earlier comment, ocrelizumab’s older cousin, Rituximab, has a quite robust safety record. However, the two drugs work in very different ways to deplete B cells, and ocrelizumab’s mechanism of action may make it a more potent and more effective B cell destroyer, but at the same time might make it a more dangerous agent. Again, I emphasize “might”.

        Only time will tell whether my concerns are justified, and I do hope I am proven absolutely wrong. I am not currently on Ocrevus, nor do I plan to be so. I have taken Rituxan in the past, to no obvious benefit. I’ve had progressive MS (or some other progressive neurologic condition – my diagnosis has never been able to be definitively confirmed) for at least 16 years.

        Hope this clarifies my stance. Remember, I am just an educated patient, and in no way wish to portray myself as some sort of trained medical professional. I am just a guy with some strong opinions.

        • Lets not forget the risk of under treatment. I’m about to go on Ocrevus and it will be drug number 3 for me. I personally am desperate not to lose my mind. So far what I see is far greater risk of letting my MS progress. The risk looks lower than a than Tysabri or Lemtrada with a relatively high level of effectiveness.

          Can we also please just stop comparing it to Rebif. Most of us are not going to consider going onto Rebif if we are looking at Ocrelizumab. There is a meta-analysis comparing it to other drugs that is much more useful from a patient perspective: https://www.msard-journal.com/article/S2211-0348(18)30580-7/fulltext

          I’m not saying we should ignore the risk but it needs to be put in relative terms.

      • If you read the article summarizing the phase III trials with rheumatoid arthritis using ocrevus-

        PLoS One. 2014; 9(2): e87379.

        a) many RA patients were on methotrexate in addition to ocr
        b) there was no indication of an increased benefit for using ocr beyond rituximab or other existing therapies, which is why it wasn’t moved forward
        c) infection rates were much higher in asian studies than in the rest of the world (some of the trials were done there)

        “In the meta-analysis, a statistically significant difference from placebo +MTX in incidence of SIEs per 100 patient-years of 2.4 (95% CI, 0.3–4.5) was observed with OCR500+MTX, but not with OCR200+MTX (0.6; 95% CI, −1.3 to 2.4).”

        “A total of 10 opportunistic infections were recorded during the DBPC period. One occurred in the PBO+MTX group (Mycobacterium abscessus on the thigh in a patient from Thailand), 5 in the OCR200+MTX group (de novo pulmonary tuberculosis [n = 2; Mexico], hepatitis B reactivation [Japan], Mycobacterium kansasii infection [Germany] and histoplasmosis [United States]) and 4 in the OCR500+MTX group (Pneumocystis jiroveci [Japan], esophageal candidiasis [France], Varicella pneumonia [South Korea] and systemic Candida infection [South Korea]).”

        d) pooled data showed no increase in malignancies

        “Pooled long-term follow-up data (of up to 5 years) from the four studies showed that the rate of malignancies per 100 patient-years was 1.18 (95% CI, 0.61–2.06) in the PBO+MTX group (n = 865; 1018 patient-years), 1.51 (95% CI, 0.99–2.19) in the OCR200+MTX group (n = 1121; 1792 patient-years) and 1.41 (95% CI, 1.07–1.83) in the OCR 500+MTX group (n = 1849; 4034 patient-years). The rate of all active treatments combined (n = 2434; 5826 patient-years) was 1.44 (95% CI, 1.15–1.78). In summary, these pooled analyses revealed no differences in the rate of malignancies between treatment groups.”

        The papers are there- people need to look for them.

        You can’t compare RA patients using mtx + ocr to ms patients not using mtx. Mtx inhibits T cell activation- you stack a T cell inhibitor on top of a B cell depletor, yes you get increased infections. That is basic immunology and not unexpected.

        Also the increased risk of cancer means that 1 in 50 taking the drug has an increased risk of cancer. 1 in 50 isnt eye popping. 1 in 5 is eye popping.

        • Points will made about the use of Methotrexate in the rheumatoid arthritis ocrelizumab trial populations. Still, doesn’t explain the lupus trial failures.

          As Neuro Doc points out in his previous comment, it appears that Ocrevus reduces CSF T cells by 60%, so it’s not only affecting T cell populations.

          Also, Rituxan has been used safely in RA patient’s for over a decade, some of whom use it concurrently with Methotrexate. So there’s that…

          As for the cancer rates, 1 in 50 progressive MS ocrelizumab trial subjects developed cancer, they didn’t develop a risk for cancer. They actually developed malignancies. Real-world use of the drug has not (thankfully) given us the same numbers, but they do warrant attention and at least give some reason for concern. As I stated previously, the cancer rates in the RRMS population were much, much lower. Still, there have been drugs halted from coming to market for numbers significantly lower than those seen in the PMS trials…

          • As the question when are they looking for the CSF T cells, if it is say after 6 months then the drug had had an effect and disease is under control and there are no cells in the CNS, including T cells, this is very different from saying drug has an effect on T cells.

          • Ah, yes, OK got it. Was wondering by what mechanism O might also eliminate T cells. Thanks for the reply.

          • When I posed the same question to Amit, his hypothesis was that the feedback that often exists between the B-cells and T-cells is taken away by Ocrevus and may explain the drop in T cells. I did inquire similarly about the effect on the innate immune system, but he felt that since the study was done across many centers the quality of the samples could not be guaranteed to accurately estimate a population that is hard to measure in the CSF in the first place.

      • Ocrelizumab was trialled in rheumatoid arthritis but the program was terminated early due to an increase in serious infections. See – https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3911947/

        In the phase III MS studies, breast cancer occurred more frequently in patients treated with ocrelizumab than those treated with rebif. Only time will tell whether malignancy is a risk with this drug.

        The review completed by the Australian regulator can be accessed here – https://www.tga.gov.au/sites/default/files/auspar-ocrelizumab-180808.pdf

        It gives you an idea of the available data at the time the drug was registered (it also gives you an insight into how regulators assess risk)



Recent Posts

Recent Comments