The difference between religion, a belief system, and science, an experimental system, is that you can falsify the latter, but not the former. So when KoL (key opinion leaders) say they believe something they don’t really mean it literally. What they are really stating is a hypothesis that needs to be interrogated.
So when I say that relapses and focal MRI activity are not MS, but are occurring in response to the cause of MS I am stating a hypothesis that needs to be falsified. Underpinning science are philosophical constructs on which to test hypotheses. One such construct, or system, is the Prentice criteria for defining a so-called surrogate endpoint in clinical medicine.
According to the Prentice criteria for a surrogate endpoint to be considered as a substitute for the disease (or is the disease), it requires that (1) the baseline measurement of the endpoint is predictive of outcome, (2) changes in the measurement of the endpoint over time is predictive of outcome and (3) changes in the measurement of endpoint in response to a therapy is also predictive of outcome (Prentice, 1989). I would add a fourth criterion that (4) the measurement of the endpoint should predict outcome independent of treatment, i.e. it should behave in the same way whether a subject is on placebo or active treatment.
When you apply these four criteria to relapses and/or focal MRI activity (new and/or enlarging T2 lesions or Gd-enhancing lesions) they don’t predict disability outcomes. This is why I state that the real MS is smouldering MS and the focal inflammatory activity is the immune system’s response to what is causing the disease. Suppressing the immune system’s response to the cause of MS may modify the course of MS, but it does not stop MS from smouldering away.
The analogy I use is one put forward way back in 1994 to Ed Thompson my supervisor when I was a PhD student. I subsequently felt confident enough to present it at a European Charcot Foundation meeting in Lausanne under the title of ‘The yin and yang of inflammation in multiple sclerosis’. I subsequently published the hypothesis as a book chapter in 2004 (see below).
It is interesting that my ideas, or hypotheses, about MS from 25 years ago haven’t changed very much. This is called cerebral stagnation. Maybe it is time to move on?
The primary neurodegenerative hypothesis (1994): Although multiple sclerosis is a clinically heterogeneous disease it can be viewed as an inflammatory neurodegenerative disease with the clinical spectrum or phenotype determined by the presence or absence of focal inflammation, similar to that which occurs in infectious diseases, e.g. leprosy. The underlying neurodegenerative component of the disease may or may not be ongoing but it is modified by superimposed focal inflammatory events. The focal inflammation may be an appropriate host response directed at an unidentified aetiological agent or an inappropriate autoimmune response. These focal inflammatory events are responsible for clinical attacks and MRI disease activity. Although damaging in themselves the focal inflammation provides the biological substrate in the form of trophic and growth factors which promote repair and clinical recovery. Inhibiting the focal inflammatory events, e.g. with generalised immunosuppression, would reduce the relapse rate and MRI activity and remove the important trophic and growth factor support provided by the inflammatory infiltrates, but it may not affect the underlying primary neurodegenerative processes. This strategy would simply convert relapsing-remitting disease into non-relapsing progressive disease. There is evidence from infectious diseases that this phenotypic variability is linked to genetic susceptibility.
Why use leprosy as an analogy? Leprosy is an infectious disease caused by a bacterium with a well-defined set of antigens that can be used to interrogate the adaptive immune systems responses to antigens. Depending on the immune response you get a different clinical picture. If you a brisk inflammatory response you get tuberculoid leprosy and it presents with very inflamed lesions. On the other side of the spectrum, you get lepromatous leprosy that is more of a smouldering disease with low-grade inflammation. Between these two extremes, you get a grey zone that is referred to as borderline subtypes. Interestingly people can convert from having lepromatous to tuberculoid leprosy if the shift their immune response to a so-called type 1 immune response that is driven by a cytokine called interferon-gamma.
Primary progressive MS is lepromatous MS and relapsing-remitting MS is tuberculoid MS and the SPMS sits in the middle. Interestingly, gamma interferon may have the same effect in MS as it does in leprosy. We know that an early trial of gamma-interferon had to be aborted as it triggered relapses. I hypothesise that if you treat PPMS with gamma-interferon you would convert it to RRMS. Unfortunately, this is an experiment that is unlikely to occur but happens naturally when you get a viral infection. Interestingly, infections are a common trigger of relapses. These clinical observations are congruent with the ‘Leprosy Hypothesis of MS‘.
Do these observations support the ‘Field Hypothesis’ and the ‘Viral Hypothesis’ of MS?
Panitch et al. Treatment of multiple sclerosis with gamma interferon: exacerbations associated with activation of the immune system. Neurology. 1987 Jul;37(7):1097-102.
We treated 18 clinically definite relapsing-remitting MS patients with recombinant gamma interferon in a pilot study designed to evaluate toxicity and dosage. Patients received low (1 microgram), intermediate (30 micrograms), or high (1,000 micrograms) doses of interferon by intravenous infusion twice a week for 4 weeks. Serum levels of gamma interferon were proportional to dose and no interferon was detected in CSF. Seven of the 18 patients had exacerbations during treatment, a significant increase compared with the prestudy exacerbation rate (p less than 0.01). Exacerbations occurred in all three dosage groups and were not precipitated by fever or other dose-dependent side effects. There were significant increases in circulating monocytes bearing class II (HLA-DR) surface antigen, in the proliferative responses of peripheral blood leukocytes, and in natural killer cell activity. These results show that systemic administration of gamma interferon has pronounced effects on cellular immunity in MS and on disease activity within the CNS, suggesting that the attacks induced during treatment were immunologically mediated. Gamma interferon is unsuitable for use as a therapeutic agent in MS. Agents that specifically inhibit gamma interferon production or counteract its effects on immune cells should be investigated as candidates for experimental therapy.
CoI: multiple
A sequence of losses
Prof G has the MS community go it wrong?
In this week’s NEJM there is an insightful perspective by Louise Aronson on ageing and driving.
Aronson. Don’t Ruin My Life — Aging and Driving in the 21st Century. N Engl J Med 2019; 380:705-707.
Louise quotes the American poet Donald Hall, who explains in Essays After Eighty how life is irrevocably and excruciatingly changed when a person must let go of their car: “For years I drove slowly and cautiously, but when I was eighty I had two accidents. I stopped driving before I killed somebody, and now when I shop or see a doctor, someone has to drive me. …Old age is a ceremony of losses.”
Although this refers to old age the same can be said for someone with MS. MS is a sequence of losses. Does it have to be this like this? I hope not, but to get to this position we need to go beyond NEDA.
I am running one of our Barts-MS teaching programmes this week in which a case was presented by one of the delegates. The lady, who is in her early thirties, has a diagnosis of relapsing MS and is NEDA, off therapy for 5 years, i.e. no relapses and no new T2 lesions. However, when you look at her sequential MRIs next to each other it is clear that she has progressive brain volume loss. She has NEDA-3, but clearly, something else is happening to her brain. I suggested to the neurologist looking after this patient to interrogate her in detail, i.e. to measure her brain volume, send her for cognitive testing, arrange for a more objective interrogation of her neurological functioning and to do a lumbar puncture to assess if she has inflammation and ongoing damage as measured by CSF neurofilament levels. In other words, don’t rely on what we have now to assess her MS disease activity.
The problem we have is that we have created a beast called NEDA and the wider MS community now think evident disease activity or EDA (relapses and focal MRI activity) is MS. EDA is obviously not MS. It is clear that EDA in untreated patients is a very poor predictor of outcome. IF EDA was MS it would predict outcome regardless of being treated or not. In other words, EDA fails one of Prentice’s criteria for being a surrogate marker of MS.
Despite writing frequently on the topic that MS is not due to relapses and/or focal MRI activity the dogma seems to stick. I have arguably helped create NEDA as a treatment target and have been responsible for some of its stickiness as a treatment target. Can I admit I am wrong? NEDA is a useful construct, but it is now becoming a barrier to treating MS properly.
If I was a behavioural psychologist I would be referring to NEDA as the new cognitive bias. We need to shift our worldview of MS away from an MRI worldview. What we should be doing is creating a biological worldview of MS and asking what is happening in the ‘field‘ or the brains of people with MS. We have to explain why end-organ damage is ongoing despite switching off focal inflammatory activity. What is driving SELs (slowly expanding lesions), the subpial cortical lesion, grey matter atrophy and the accelerated brain volume loss? If we don’t then MS will remain a sequence of losses.
To B or not to B
Is targeting the B-cell sufficient to get on top of MS or do we need something extra?
I spoke at the MS Nurses’ MS@TheLimits2019 meeting at the Royal College of Physicians yesterday. My brief was to cover the role of B-cells in the pathogenesis of MS and to review the converging evidence that supports B-cells being the central player in the pathogenesis of MS.
It is clear that depleting B-cell therapies are very effective in controlling relapses and MRI activity. With a very favourable safety profile and relatively low treatment and monitoring burden, B cell therapies are likely to become one of the most widely prescribed classes of DMT. However, B-cell therapies don’t match HSCT, alemtuzumab and natalizumab when it comes to downstream end-organ damage markers, in particular, brain volume loss. Why? I wish I knew. But if I knew the answer to this question I would have a pretty good idea about the cause of MS.
A clue may be in the ‘Field Hypothesis‘. It is clear to me that relapses and focal MRI activity are not the primary events in MS. Focal inflammation is not MS. Focal inflammation is in response to what is causing MS and the cause is likely to be something in the CNS. Focal changes occur in the white matter weeks to months before you get a Gd-enhancing lesion. When you stop natalizumab and allow re-trafficking of lymphocytes you get rebound disease activity way and above what one would expect from pre-treatment baseline levels of disease activity. What is happening in the brain, or field, of these patients to trigger such a vigorous inflammatory response? Could it be a virus? Importantly, B-cells appear to be needed for the rebound response. Rituximab, and I suspect ocrelizumab, are very effective in preventing rebound. However, as both these agents target a small subset of T-cells you can’t claim categorically that the rebound is only driven by B-cells.
The difference between HSCT, alemtuzumab and natalizumab and the anti-B cell therapies (rituximab, ocrelizumab and possibly cladribine) is the former take out or inhibit trafficking of both T & B cells. As HSCT and Alemtuzumab have the best data in relation to long-term remission, or potential cures, you have to conclude that you need to target both B cell and T cells (substantial peripheral depletion) if this is your treatment aim.
Please note that I classify cladribine as a B-cell depelter and not a dual B and T cell depelter. The level of T-cell depletion with cladribine is modest at the licensed dose (~50%) which is not sufficient to put it into the same class as alemtuzumab and HSCT. This is one of the reasons why I refer to cladribine as being a SIRT (selective immune reconstitution therapy) and the others as NIRTs (non-selective immune reconstitution therapies).
I have always made the point that to treat MS you need much more than an anti-inflammatory and that you also have to have neuroprotective therapies and potentially remyelinating agents on board as well. If you have disabilities we need to be thinking about neurorestorative therapies and finally you need to target lifestyle and wellness to tackle the issue of comorbidities and ageing.
So in short, targeting B-cells is important, but not sufficient to get on top of the shredder.
You will see that a large part of my talk was covering the link between EBV, B–cells and MS. The B-cell hypothesis at least strengthens the case for EBV being the cause of MS and the need for an EBV vaccine for MS prevention trials. Please don’t forget that EBV lives inside memory B cells and hijacks the B cell’s biology in many ways that have potential relevance to MS and other autoimmune diseases.
My talk will be available online in a few weeks to help you interpret my presentation. In the interim you can download my talk from my slide sharing site.