Tuesday, 25 October 2016

JCV are we in safer times?

Acta Neurol Scand. 2016 Oct 20. doi: 10.1111/ane.12699. [Epub ahead of print]

JCV serology in time: 3 years of follow-up.

Cambron M, Hadhoum N, Duhin E, Lacour A, Chouraki A, Vermersch P.



Although many neurologists are reluctant to use natalizumab in MS (multiple sclerosis) given the increased risk for PML (progressive multifocal leukoencephalopathy), trust was regained with the introduction of JCV antibody titres as a potent disease-modifying therapy. Literature shows that in patients with a negative JCV serology, the risk of PML is virtually non-existent. Unfortunately, seroconversion causes concern amongst many neurologists. Furthermore, when patients seroconvert, it is still unclear what the risk is of passing the important threshold of 1.5.


JCV serology data of 161 patients were analysed, upon treatment with natalizumab at the University Hospital in Lille, France, between May 2012 and November 2014.


Of the 81 patients who tested negative for JCV antibody at baseline, 23 (28.3%) seroconverted but only seven (8.6%) passed the threshold of 1.5. Of the 80 patients testing positive for JCV antibody at baseline, eight had an initial JCV antibody titre of 0.9 or lower of which only one of eight (12.5%) patients passed the threshold of 1.5 in the following 3 years. Eight of 15 (53.3%) patients passed this threshold if the initial serology was higher than 0.9.


JCV-negative patients and JCV-positive patients with antibody levels below or equal to 0.9 both have a low risk of surpassing the 1.5 threshold.

Figure: Google trends for interest in natalizumab from 2004 - current. The peak in 2005 corresponds to suspension of marketing authorization of natalizumab (Tysabri) by the FDA.

The John Cunningham virus causes a serious demyelinating CNS disorder called progressive multifocal leukoencephalopathy (PML). Although, 33-91% of the general population is positive for the virus, only half of PwMS are positive for the antibody (the seroconverted). PML risk is at the heart of natalizumab therapy; public and professional concerns over this have not yet been convincingly allayed. In addition, there is the added concern of seroconversion on therapy; with some researchers (most recently N Schwab's group at ECTRIMS 2016) suggesting that those on natalizumab seroconvert at higher rates than expected by aging.

A lot of sweat has therefore gone into a developing an antibody test (STRATIFY) that can be used to risk stratify PML development in natalizumab users. An index threshold of 1.5 is used by many as the cut-off. In this study, Cambron et al. wanted to investigate the risk of an initially negative antibody test person passing the threshold of 1.5. They found that 28.3% seroconverted with 8.6% passing the threshold of 1.5. They found that the likelihood of seropositivity increased in those receiving a higher number of infusions. But they also then suggest that the ones with an index of =/<0.9 had a lower risk of passing the 1.5 threshold (an oxymoron possibly).

I'm not sure how this allays my fears. The authors conclude: "Patients who have a positive response to natalizumab and a positive JCV index values below 0.9 can be reassured and safely continue their natalizumab treatment, obviously with an accurate clinical and radiological follow-up". And therein lies the problem, a negative baseline test is not fool-proof or protective against PML development (see Abstract P1259 below), and where cases of diagnosed PML are concerned, their defining characteristics are being JCV positive and receiving natalizumab for >2 years (see Abstract P1111 below)!

Maybe, we need to stop worrying our heads over this one; there is only so many ways you can cut a cake.

Abstract: P1259, ECTRIMS 2016
Progressive multifocal leukoencephalopathy in a JCV seronegative patient treated by natalizumab : a new case report
Author(s):N Hadhoum ,V Neuville ,P Vermersch
Background: the prediction of the risk of progressive multifocal leukoencephalopathy (PML) is crucial to guide natalizumab (NTZ) prescription and ensure a higher safety use. The JCV index (a corollary to antibody titer) is a predictive factor of PML.

Case presentation: here we report a case of a 55-year-old female diagnosed with multiple sclerosis in 1996. Because of a continued clinical disease activity despite interferon, she was switched to NTZ. A total of 84 infusions were administered between 2011 and 2016. She never received any prior immunosuppressor and has been tested negative for anti-JCV antibodies: index at 0.22 in 06/2014 and 0.11 in 06/2015 and 10/2015. The annual MRI were strictly stable. During a routine follow-up, the patient complained of a right-sided lower limb paresis. NTZ was immediately stopped. Brain magnetic resonance imaging revealed a sub-cortical white matter lesion, hypoT1, hyperT2 in the left parietal lobe invading the U-fibers in DIR sequence. PML was suspected then confirmed by detection of JCV-DNA in the CSF (2 analyses: 11 then 68 copies/ml). The peripheral CD4+ T-cell count was 1149/mm3, CD8+ at 395/mm3. Note that the JCV index was still negative at the diagnosis and 15 days later. We investigate the lymphopenia and found a probable common variable immune deficiency (CVID) with severe hypogammaglobulinemia (3g/dL), surprisingly asymptomatic until now. A JCV seroconversion was noted at 3 months (index at 3.48).

Discussion: it is well-known that PML risk is influenced by NTZ duration, prior use of immunosuppressors and seropositivity for the JCV. The JCV seronegativity status does not allow to exclude any risk of PML. Indeed, de novo infection with seroconversion is theoretically possible and underestimate rate of infection appears likely (in case of low viremia). One recent publication reports a NTZ-related PML diagnosed 2 weeks after negative anti-JCV antibody assay and 2 others PML due to NTZ were tested negative (prior to diagnostic) among the 541 reported cases. In our case, we can hypothesize that the JCV index were false negative results as the asymptomatic CVID imply a fail in humoral responses.

Conclusion: this case must not undermine the central role of the JCV index in predicting the risk of PML. However, it seems wise to regularly control the blood lymphocyte phenotype and immunoglobulins. CVID is the one of the most common causes of adult immunodeficiency.

Abstract: P1111, ECTRIMS 2015
JCV (John Cunningham virus) index: follow-up of a French cohort
Author(s):S. Mathais ,A. Gayou ,X. Moisset ,J.C. Ouallet ,B. Pereira ,A. Ruet ,F. Taithe ,K.-K. Kounkou ,E. Dumont ,B. Brochet ,P. Clavelou

Objectives: Patients receiving natalizumab (NTZ) are at risk of PML (Progressive Multifocal Leukoencephalopathy) due to JCV reactivation. In order to predict that risk we currently use the JCV index. Recently, cut-off values for the JCV index have been published. The risk of PML seems to be very low under 0.9, low between 0.9 and 1.5 and high above 1.5.

We wanted to better characterize the JCV index and its evolution and check whether the suggested thresholds were applicable to our cohort.

Methods: This study was conducted in two French University Hospitals (Bordeaux and Clermont-Ferrand) until the 31/12/2014. We included 352 patients over 18 years with a relapsing-remitting multiple sclerosis and at least one JCV index. We got 908 samples. Three subjects developed a PML. We evaluated the seropositivity for JCV, the potential modifying factors of JCV index (before and under NTZ), its evolution under NTZ and particularly in case of PML.

Results: 63.2% of NTZ-naïve patients were positive for JCV. At this stage the JCV index is statistically related with the age but independent from sex, disease duration or anterior exposure to immunosuppressive agents. Among the patients finally treated with NTZ 39.7% were positive for JCV at baseline and 55% at the last JCV index. On NTZ only the treatment duration was statistically related to the evolution of the JCV index. 84.5% of subjects remained stable during the follow-up specially if they were initially positive for JCV (45.4%). 6.7% seroconverted and 1.5% seroreverted. Our three cases of PML presented two risk factors (over 2 years of NTZ and seropositivity for JCV). Their indexes were above 1.5 for at least 3 months prior PML.

Conclusions: JCV index seems to remain fairly constant over time and the published threshold for high PML risk (1.5) was confirmed in our cohort.


  1. Thanks NDG.

    I really enjoy your posts and I like this topic in particular; am coming up on the two year mark and have made my 4 weekly trip part of my schedule and don't see any need to change so long as the numbers continue to look good.

  2. I seroconverted 2 years ago (age 48) and after a drug holiday went on extended dosing. We were hoping extended dosing would bring the titer down. It did not. My titers are way beyond 1.5. I wonder about the confidence interval when the titers are beyond 1.5. I have not seen any data discussing this.

    Thanks for any insight,


    1. So if you look at the original publication from Plavina et al the confidence interval ranges from just under 2 to just under 3 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4282070/) in those not receiving prior immunosuppression. This is because the case numbers are small relative to the non-case numbers. But you can also see on Fig 4 that 50/51 patients who developed PML infact had an index above 0.9!

  3. I'm on tecfidera. Living in London. Can I request to be tested for jcv? I was told Biogen isn't keen for patients to be tested!?

    1. I dont know, maybe one of the neuros can answer...but the question is what if you are positive? One would expect there is a 50:50 chance you are

    2. Yes you can get tested. It's all about transparency these days!


Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.