Vitamin D effect in the pivotal fingolimod clinical trials

Chicken or egg? Causation or association (reverse causation)? Is vitamin D a DMT or not? Does this study muddy the waters or provide some clarity? 

J Neurol. 2017 Dec 14. doi: 10.1007/s00415-017-8697-3. [Epub ahead of print]

Efficacy and safety outcomes in vitamin D supplement users in the fingolimod phase 3 trials.

Hongell K, Silva DG, Ritter S, Meier DP, Soilu-Hänninen M.

Abstract

BACKGROUND:

Low serum levels of 25-hydroxyvitamin D have been associated with worse outcomes in multiple sclerosis (MS) patients treated with interferon-beta. Association of vitamin D nutrition on the outcomes of other MS therapies has been studied less.

OBJECTIVE:

Whether patients in the phase 3 fingolimod trials using vitamin D supplements have better clinical, MRI and safety outcomes than non-users.

MATERIALS AND METHODS:

Pooled data from phase 3 FREEDOMS trials was analyzed post hoc. Vitamin D use was defined as 'non-users' (n = 562), 'casual users' (n = 157) and 'daily users' (usage 100% time in the study, n = 110).

RESULTS:

Expanded Disability Status Scale change from baseline to month 24, and annual relapse rate and proportion of patients with relapses were similar across the vitamin D user groups. Proportion of patients free of new/enlarging T2 lesions significantly favored vitamin D 'daily users' versus 'non-users'. Mean number of lesions were lower and proportion of patients free of gadolinium-enhanced T1-lesions were higher in the 'daily users'. At month 12, percent brain volume change was significantly lower in the 'daily users' versus 'non-users' and remained low at month 24 (non-significant). Incidence of depression was lower for vitamin D 'daily users' (non-significant).

CONCLUSIONS:

We observed improved MRI outcomes on percent brain volume change and proportion of patients free of new/enlarging T2 lesions, and a trend of less depression in the 'daily users' of vitamin D supplement in patients in the FREEDOMS trials.

Figure: MRI outcomes by vitamin D use in FREEDOMS and FREEDOMS II.

Sometimes it’s difficult to see the wood for the trees. You must travel through the woods again and again, and with luck you may be able to avoid the wolf. Read on and you'll understand my meaning.

In their introduction, Hongell et al. allude to an influence of vitamin D levels in the B interferon trials. Firstly, in the BEYOND trial (which compares high dose 500 ug IFN- β 1b to standard dose 250 ug), where high levels of vitamin D were associated with less disease activity on MRI, so much so that an increase of 50 nmol/L in blood levels resulted in a 31% lower rate of new lesion formation (Fitzgerald et al. JAMA Neurol 2015; 72: 1458-1465). Secondly, in the Benefit trial, where low blood levels of vitamin D was associated with worse clinical and MRI outcomes in PwMS (Ascherio et al JAMA Neurol 2014; 71(3): 306-314). They believe there to be a synergistic effect between vitamin D and the interferons and possibly even an influence on the therapeutic activity of interferon on relapses.

Armed with this knowledge, the authors (three of whom work for Novartis; the proprietor of fingolimod) have studied the effect of vitamin D use in the pivotal fingolimod Phase III clinical trials; FREEDOMS and FREEDOMS II. Just to be clear this was never a part of the original trial analysis and therefore would be considered post-hoc analyses or a data mining expedition.

They studied the effect of ‘daily’ vitamin D users with that of ‘non-users’ among those who were assigned to either receive fingolimod or placebo (dummy tablet).

What they found was that ‘daily’ users of vitamin D supplements had better MRI outcomes compared to the ‘non-users’. Although, there was no statistically significant difference in the mean numbers of new/newly enlarged T2 lesions, they did note that the proportion of PwMS free of new/newly enlarged T2 lesions significantly favoured the ‘daily’ users from month 0 – 12 and month 0 – 24 (see figure above). Surprisingly, even the percent brain volume change (or brain atrophy/volume loss) was significantly lower at month 12 in ‘daily’ users than ‘non-users’ and remained low at month 24 (although the latter was not statistically significant).

It therefore begs the question - what role did vitamin D play in the much reported favourable outcomes of fingolimod in the FREEDOMS and FREEDOMS II studies? Of course, the answer lies in the detail (or the trees metaphorically speaking) - on how much information was collected in the original data sets. The lack of blood vitamin D levels in the trials is a handicap, which makes these findings even more difficult to interpret.

The season of MS: Glandular fever may not be linked to time of year

Vitamin D (as measured by time of year) and glandular fever are not linked. Should we be surprised?

I guess it boils down to how you think it is all working. 

If the month of birth theory holds any water then the major risk period is when you're in the womb or shortly after birth. That the vitamin D hypothesis is also evident in type I diabetes suggests the risk factor is early in life. The biology suggests that vitamin D can shape your immune repertoire (range of different things your immune system responses to). So your die is cast as to whether you are at risk of autoimmunity. But this does not give you autoimmunity. Your other genes also put you at risk.

Whether vitamin D exhibits a major influence once the condition appears, remains to be seen. The trials are being done. 

Migration studies from low to high incidence countries suggest you acquire the trigger before you are 15 years for risk of MS. One suggestion is that this trigger is Epstein Barr Virus. This shapes your B cell (antibody making cells) repertoire. Which in turn may determine whether you get MS or other autoimmune conditions associated with EBV. 

About half the western population become infected with EBV in infancy. Are these the diabetes-prone people?  I guess infection goes unnoticed (undiagnosed) as one of the many episodes of fever that a child has. 

In adolescent life the naive/mature B cell gets infected by EBV because the virus enters the B cell via CD21 and an action via HLA-DR. Is this where the number one autoimmune genetic MS risk factor occurs? Is HLA-DRB1*1501 a gene that is good at getting B cells infected by EBV?

The mature B cell is then triggered to proliferate by the virus, and their killing by anti-EBV CD8, cytotoxic T cells cause a cytokine storm and sickness behaviour...that is glandular fever. 

The virus causes the B cells to mature into memory B cells and not antibody-producing plasma cells by a protein called EBNA3. 

The virus then hides in the memory B cells out of sight of the immune system, but switching off virus production. And as memory B cells are like buses that go round the body, the virus hitches a ride all round the body. 

However, the virus does other things, including making the memory B cell independent of the requirement of T cell help. So they do not need to get stimulated by a molecule called CD40. This means, for example, if a memory B cell enters the brain and sees its target it will become activated and so does not need a T cell of the same specificity to be present.

As more damage occurs, more antigens capable of stimulating B cells are liberated. You don't have to postulate that it is one antigen causing the problem and these can change over time. However there is specificity because the B cell would have to see a brain derived signal to trigger release of cytokines to start the lesion.

This would be a great immune-evolution because it makes us more able to fight infection quickly, creating a survival advantage for the human population.

On the down-side, autoimmunity could occur. However, EBV has been with us for thousands of years (co-evolving with us). As humans used to have children earlier in life - and died earlier than we do now - autoimmunity would arguably not be selected against. Why? Because most autoimmunity occurs later in life (old-age and after child birth for our ancestors). 

So, evolution-wise, this would not be selected against. Furthermore, even if it did occur in our ancestors, it is at such a low frequency that it does not affect the overall survival of the population.  

What do you think of the idea?

Please post refuting (& supporting) work

Downham C, Visser E, Vickers M, Counsell C. Season of infectious mononucleosis as a risk factor for multiple sclerosis: A UK primary care case-control study. Mult Scler Relat Disord. 2017 Oct;17:103-106.

BACKGROUND:Infectious mononucleosis (IM) and vitamin D deficiency are both risk factors for multiple sclerosis (MS).
OBJECTIVE:We wished to establish if IM in the winter months when vitamin D levels are low may be a greater risk factor for MS than IM in the summer months.
METHODS:We identified all patients with MS diagnosed aged 16-60 in a large primary care database in the United Kingdom and matched each by age, sex, general practice and observation period with up to six controls. We identified a coded diagnosis of IM prior to the index date (date of diagnosis). Logistic regression was used to calculate the odds ratio for prior IM exposure in cases versus controls and for winter versus summer exposure in cases and controls with prior IM exposure.
RESULTS:Based on 9247 cases and 55,033 matched controls (246 and 846 with prior IM respectively), IM was associated with the development of MS (OR 1.77, 95%CI 1.53-2.05) but there was no evidence that IM in the winter as opposed to summer was associated with developing MS (OR 1.09, 95%CI 0.72-1.66).
CONCLUSION: We found no evidence that the season of IM influences the risk of subsequent MS.