CNS lymphoma in autoimmunity


Non-Hodgkin’s Lymphoma (a cancer involving predominantly of B cells) involving the brain, also known as primary CNS lymphoma (PCNSL) is something I see every year like clock work for the haem-oncology service at St Bartholomew’s Hospital, London. And despite the regularity of my visits, there is something extraordinary about the whole experience.

You start to either vegetate or have kittens when working in a cancer centre. Let’s say my day job treating MS by comparison comes with a sense of satisfaction that is difficult to muster when dealing with cancer. But, over time I have also found myself changing in my management of MS as a unexpected side effect of all of this.

NHL is commoner than the nice/but recalcitrant Hodgkin’s variety, and flits around the body like an immune cell on speed, eventually finding a foothold in the central nervous system (see Table below for the differences between the two).

My usual patient with PCNSL is middle aged or older, has other disorders and co-morbidities, and may have received quiet a few immunosuppressants prior to their diagnosis or are immunosuppressed, for example with HIV. Although, there is not much evidence for stopping treatment in MS, I have begun in other autoimmune conditions that I manage to withdraw immunosuppressant’s after a period of disease stability and as the person ages.

And there maybe a reason to do this also in MS. In a metanalysis 1103 publications of autoimmune disorders in PCNSL, only 39 (50 cases) could be analysed. The common disorders were SLE (n=12, 24%), MS (n=8, 16%), myasthenia gravis MG (n=7, 14%), rheumatoid arthritis RA (n=6, 12%), autoimmune uveitis (n=3, 6%) and myositis (n=3, 6%); see Figure below. Disorders that display severe disease activity and require heavy immunosuppression.

Immunosuppressants prior to diagnosis included prednisolone (n=33, 69%), mycophenolate mofetil (n=20, 42%), azathioprine (n=20, 42%), and methotrexate (n=9, 19%).

Figure: Distribution of autoimmune diseases in AD-PCNSL. Pie diagram illustrating the distribution of various autoimmune diseases (AD) in PCNSL. Systemic lupus erythematosus (24%) was the most common AD among PCNSL patients, followed by multiple sclerosis (16%), myasthenia gravis (14%), rheumatoid (rheum.) arthritis (12%), uveitis (6%) and myositis (6%)

They were mostly diffuse large B-cell lymphoma (DLBCL) n=37, 80% and Epstein-Barr-virus was reported in 36 cases. After diagnosis, immunosuppression was reduced in all 27 cases where such information was available. Treatment consisted of chemotherapy alone (methotrexate, rituximab or cytarabine), or chemotherapy and whole brain irradiation. The median overall survival (n=37) was 8, 5 and 25 months in SLE-, MS-, and MG-associated PCNSL (although not statistically significant between groups owing to the small numbers).


 Neurooncol. 2020 Jul 18. doi: 10.1007/s11060-020-03583-9. Online ahead of print.

Autoimmune disease-related primary CNS lymphoma: systematic review and meta-analysis

Leon D KaulenPhilipp KarschniaJorg DietrichJoachim M Baehring

Background: Recent studies suggest a relatively high prevalence of autoimmune disorders (AD) among primary CNS lymphoma (PCNSL) patients, however, the literature is limited to case reports. To gain a better understanding of AD-PCNSL we reviewed and analyzed all cases described in the literature.

Methods: We searched the MEDLINE database using the search terms ‘central nervous system lymphoma’ or ‘CNS lymphoma’ along with AD-related terms. We selected 39 records for qualitative synthesis of data and identified 50 AD-PCNSL. Clinical, imaging and outcome data were collected. Overall survival (OS) was analyzed with the Kaplan-Meier method. Univariate and multivariate analyses were performed using log rank test and Cox proportional hazard model.

Results: Most common AD were systemic lupus erythematosus (24%), multiple sclerosis (16%), and myasthenia gravis (14%). All patients had received immunosuppressants for their AD. Median interval from AD until PCNSL diagnosis was 108 months (range: 11-420). Male-to-female ratio was 0.42 and AD-PCNSL was diagnosed at a median age of 57 years (range: 2-88). On imaging lesions typically localized to the hemispheres (65%) and displayed peripheral enhancement (74%). Pathological evaluation revealed diffuse large-B-cell lymphoma (DLBCL) subtype (80%) and Epstein-Barr virus positivity (75%) in most AD-PCNSL. Median OS was 31 months. Age > 60 years (p = 0.014) was identified as a significant prognostic factor.

Conclusions: AD requiring immunosuppression appear over-represented in the population of PCNSL patients. Aggressive polychemotherapy can accomplish long term OS in AD-PCNSL comparable to immunocompetent patients. Age > 60 may serve as a prognostic factor.

About the author

Neuro Doc Gnanapavan


Leave a Reply to Tony F Cancel reply

  • My usual patient with PCNSL is middle aged or older, has other disorders and co-morbidities, and may have received quiet a few immunosuppressants prior to their diagnosis or are immunosuppressed, for example with HIV

    Do you know that its always the case with organ transplanted patients ,i am thinking about heart ,.liver, kidney

    They recive daily immunosuppressors are they at risk also?


    • Yes, unfortunately. The transplant patients unfortunately are another group at risk – mycophenolate is a big culprit. Although this has reduced as tacrolimus, sirolimus are now favoured for organ transplants to prevent rejection.

  • A chilling post!
    It is easy to assume that Neuros who don’t buy into the treat early and treat hard (well proven) dogma are dinosaurs.
    But some that I have spoken to are very cautious about the C word, and feel that it is being discounted somehow when MS treatment advice is given.
    We compare ARRs at 2-5-10ys, BVL, hand dexterity, EDSS, new T2s, etc…. but how many studies out there compare cancer risk (excluding HSCT)? Where is the efficacy-to-cancer-risk chart on this blog?

    If the NIH study is correct, that all MS treatments stop working after the age of 54, then what is the incremental benefit on being immunosupressed after that milestone with a maintenance therapy?

  • A good argument in favour of IRTs (immune reconstitution therapies) in which longterm immunosuppression is avoided. There is a clear signal of CNS lymphoma with natalizumab (reduced CNS immune surveillance) and a systemic lymphoma signal with fingolimod.

    The question I ask “Is life -long immunosuppression a tenable solution for treating MS?”.

    • At which point is it too late to switch to IRT for a long term JCV- natalizumaber?

      I am guessing there is a cut-off point of some sort beyond which the switch makes less sense?

        • Is Dr.Angry’s test ready to tell me if I will respond as well to Lemtrada as I do to Tysabri?

          • It is. But if you’ve never had Lemtrada before this shouldn’t be an issue. It’s the third course where you need to consider this question.

    • The question I ask “Is life -long immunosuppression a tenable solution for treating MS?

      Nice 🙂

      Do you ask that question to pharma stake holders?

      What do they respond?


      • Pharma or stakeholders 😉 If the latter
        …the parable of the blind men and an elephant comes to mind.

        • Thanks NDG,

          3 quick questions:
          1) How can I know in advance if I will be one of the lucky ones who will develop thyroid autoimmunity? (they do better apparently)
          2) If time is brain, and I have had smoldering MS for 15 years since CIS (NEDA-3), is mybrain still young enough to benefit from IRT?
          3) Am I legally allowed to have Lemtrada if I have not relapsed on Tysabri? is brain volume loss sufficient to make a case for Lemtrada in the UK?

          • 1) As far as I know there isn’t a test for this.
            2) IRT strategy is heavy immunosuppression at start, but front loading of side effects. Technically can be done at any time. For example outside of MS in vasculitis (inflammation of blood vessels) first line is normally regular azathioprine tablets, but if there is disease break through you use pulsed cyclophosphamide which is a block of infusions.
            3) When a person is switched from Natalizumab they need to be stepped onto another equivalent therapy to avoid rebound relapses, this includes IRT strategies.

        • Stake holder: Stakeholder Management, stakeholder mapping
          stake·hold·er (stāk′hōl′dər)
          1. One who has a share or an interest, as in an enterprise.

          I think they are basically the same


          Most researchers in academia are stake holders and vice versa

          Some members of your team are or were stakeholders in some biotech company project (not so big pharma)


    • Reading about PCNSL, I can see a number of case studies online. Some case are of people who were diagnosed after only 3 doses of natalizumab. Is it possibly to query the Biogen Tysabri risk monitoring database to have some high level data of how prominent this risk is and possibly which groups (age or other biological indicators) is at most risk?



Recent Posts

Recent Comments