With the outbreak of the COVID-19 pandemic and the rapid development of SARS-CoV-2 vaccines worldwide, concern was raised that the immunomodulatory/immunosuppressive effects of multiple sclerosis (MS) disease modifying therapies (DMTs) might affect COVID-19 disease severity as well as the development of humoral and cellular immunity after SARS-CoV-2 exposure or vaccination (1). In Israel, a nationwide COVID-19 vaccination program was rapidly launched starting December 2020, based solely upon the BNT162b2 (Pfizer-BioNTech) mRNA vaccine, offering two doses administered 3 weeks apart, and a third vaccine booster ≥ 5 months after the 2^nd dose (2), a vaccination regime associated with a good safety profile (3). All PwMS are recommended by the National Multiple Sclerosis Society of the United States and by the MS International Federation to be vaccinated against SARS-CoV-2 (4, 5). With the variety of available DMTs, that exert their action through diverse immunological mechanisms, it is of high importance to investigate the immune response to SARS-CoV-2 vaccination in PwMS treated with various immunotherapies, in order to develop optimal and treatment-stratified guidelines. The immune response to SARS-CoV-2 consists of both a cellular and a humoral response, but measurement of SARS-CoV-2 IgG is widely used to identify persons who have recovered from COVID-19 infection or as a confirmation of a sufficient vaccine response, based on the fact that the levels of neutralizing antibodies have been shown to be highly predictive of SARS-CoV-2 immune protection (6). Initial reports have demonstrated that while most MS patients show a positive humoral immune response shortly after the 2^nd SARS-CoV-2 mRNA vaccination, a considerable proportion of patients treated with fingolimod or ocrelizumab do not develop antibodies (7–12). In light of the immunological effects of the DMTs, it is important to evaluate the effect of the patients’ immune response to the vaccine and to COVID-19 over time and the effects of the third dose of vaccine in patients treated with ocrelizumab and fingolimod. In this nation-wide multi-center prospective study we examined the levels of anti-SARS-CoV-2 antibodies in PwMS treated with most available DMTs and in healthy controls (HC), during a period of at least 6 months following the 2^nd SARS-CoV-2 vaccination and at 1-16 weeks following a 3^rd vaccine booster. In addition, we evaluated the cell-mediated immune responses against the COVID-19 spike protein in a subgroup of the participants.

To the best of our knowledge, this is the first comprehensive prospective study on the magnitude and durability of SARS-CoV-2 IgG levels combined with T cell responses in PwMS, treated with various DMTs, for over 6 months from vaccination and following a 3^rd vaccine booster. Vaccination is the major available tool to control and fight the pandemic of COVID-19. Several vaccines have been developed worldwide, using different strategies and platforms, including the novel strategy of vaccination with the SARS-CoV-2 spike protein mRNA as in the case of the BNT162b2 (Pfizer/BioNTech,Inc) and the mRNA-1273 (Moderna Tx,Inc) vaccines. These mRNA vaccines appear to be safe and have not been associated with increased incidence of MS disease activation (13–16). In general, the use of DMTs in PwMS has not been found to significantly affect COVID-19 disease course (17), although anti-CD20 therapies have been associated with higher incidence of COVID-19 infection (18) and with a possibly increased risk of severe disease course (19).

An effective, long-term, memory immune response is driven by the adaptive immune system, consisting of both a humoral response mediated by B cells, producing antigen-specific antibodies, and a T cell-mediated response, causing destruction of virus-infected cells and necessary for the development of plasma cells and memory B cells. Thus, measurement of both the humoral and the cell-mediated responses is required to precisely estimate the immune response to SARS-CoV-2 vaccine. However, the relatively easily accessible measurement of anti SARS-CoV-2 antibody levels in the serum is the common and mostly used method to identify individuals who recovered from COVID-19 infection or to confirm a sufficient vaccine response, especially since SARS-CoV-2 neutralizing antibodies were shown to be highly predictive of SARS-CoV-2 immune protection (6). Evidence for positive humoral immune responses to the SARS-CoV-2 vaccine in MS patients is accumulating (9, 10, 12, 20, 21). Our current study provides longitudinal accumulating data regarding the IgG serum levels following >6 months follow-up and also after a 3^rd vaccine booster. Our findings show that IgG levels to the SARS-CoV-2 spike protein in MS patients were slightly lower than in healthy individuals, declined by >80% within 6 months from the 2^nd vaccination, and were significantly increased following a 3^rd vaccine booster. A similar decline in IgG levels at 6 months from vaccination and a beneficial effect of the 3^rd vaccine booster was also demonstrated in the healthy cohort. Interestingly, we did not detect a negative correlation between IgG levels after COVID-19 infection and time, suggesting a more robust and long-lasting immune response to the SARS-CoV-2 virus following natural infection vs vaccination. However, the median time between COVID-19-infection and blood collection in our cohort was 27 weeks, thus may not being able to capture the major, initial decline in IgG levels. Antibody levels following COVID-19 infection were shown to be associated with COVID-19 disease severity (a parameter that was not registered in our study), and were reported to start declining after 60 days, but still to be detectable for at least 120 days (22). For the COVID-19-recovered patients in our study, the follow up data were limited, as the majority of these patients received the recommended post–infection vaccination. Interestingly, we found higher IgG levels in vaccinated female than in male patients, not reported in previous reports (20, 23). While we did not detect a similar difference in IgG levels between genders in HC, probably due to the relatively small cohort, both levels of IgGs and of neutralizing antibodies have been shown to be higher in females than in males receiving the BNT162b2 Covid-19 Vaccine, especially in association with age (24), a difference that at least in part is thought to be hormone-mediated (25). The negative correlation between IgG levels and age in vaccinated PwMS or healthy participants presented in our study has also been suggested elsewhere (11, 26), and confirms the potential elevated risk of COVID-19 infection in elderly people.

Our data show, similarly to previous reports, that MS patients treated with most DMTs develop a positive humoral and cell-mediated immune response to the mRNA vaccine, which in general does not differ significantly from that observed in untreated patients or healthy individuals. The reduced humoral response and low frequency of seropositive patients that was observed in PwMS treated with fingolimod and ocrelizumab are in line with several recent studies (7, 8, 10–12, 20, 21, 27–30). In addition, we could show that a similar humoral response and low frequency of seropositive patients is also observed after the 3^rd vaccine booster in the patients treated with fingolimod or ocrelizumab, and that in those patients (in contrast to healthy individuals and PwMS in general), the 3^rd vaccination did not boost the IgG levels further than the first 2 vaccine doses. One recent study reported similarly that only 1 out of 16 ocrelizumab-treated patients was seropositive after the 3^rd BNT162b2 vaccine booster (29). It seems thus, that while some PwMS treated with S1PR-modulators or anti-CD20 therapies do benefit from the additional vaccine dose, an optional vaccine strategy for these patients should be considered. On the other hand, we found that the vast majority of ocrelizumab-treated patients developed a positive anti-spike protein IFNγ-T-cell immune response (both after the 2^nd and the 3^rd vaccinations) and thus, despite the lack of a sufficient humoral response, they may carry a relative protection against COVID-19 infection or severe disease. Similar results were recently reported by other investigators: Gadani et al. found that 96.9% of ocrelizumab patients developed a T-cell response (30), Tortorella et al. reported a 92% positive T-cell response rate (21) and Brill et al. found a response rate of 89.7% (11). Aposolidis et al. reported positive CD-4 and CD-8 T-cell responses to vaccination in all anti CD-20-treated patients, but these responses were somewhat skewed, with reduced follicular helper T (T_FH) cell development and elevated CD8 T-cell responses, while Th1 responses were preserved, especially in patients who were seronegative (31). Although these findings of relatively preserved T-cell-mediated anti-COVID-19 responses are encouraging, studies with larger patient cohorts, including follow-up on the risk and outcome of COVID-19 infection are needed to accurately estimate the vaccine-induced protection in anti-CD20- treated patients. Interestingly, we observed both positive humoral and cellular immune responses to the vaccine in patients treated with another anti-CD20 mAB - ofatumumab. We are not aware of published data on the response to vaccination in these patients, however, one study found that COVID-19- recovered ofatumumab-treated, B-cell depleted patients were seronegative for anti SARS-Cov-2 antibodies, but developed a positive cellular response (32), while another study reported a positive anti-SARS-CoV-2 antibody response after COVID-19 in a B-cell depleted ofatumumab-treated patient (33). Confirmation of this finding in a larger group of ofatumumab-treated patients will be of high interest.

Anti-CD20 therapies such as ocrelizumab reduce the number of circulating B cells, thus preventing a sufficient B cell response to antigens and the development of antibodies, an effect which is likely to persist until sufficient B cell reconstitution occurs. It would therefore be of interest to determine the optimal time interval between the infusion and vaccination that would allow for sufficient reconstitution of B cells to enable an effective humoral immune response. This issue has been studied by a few groups but remains debatable; while in some reports there was a significant correlation between time from last treatment to vaccination and the IgG levels (11, 20, 30), others could not confirm this finding (21). In our study, IgG levels correlated only very weakly with the time between infusion and vaccination, both after the 2^nd and the 3^rd vaccine administration. One study found that 143 days following ocrelizumab infusion is the time-point when IgG starts to increase (20). We found that a time interval of more than 5 months between ocrelizumab infusion and vaccination allows for higher IgG levels than a time interval of <5 months, and a similar difference was also observed in patients vaccinated ≥6 months vs. those vaccinated <6 months after ocrelizumab infusion. Our data also showed that a higher proportion of ocrelizumab-treated patients were seropositive if they were at least 5 months from last infusion at time of vaccine, and in patients with ≥ 6 months time interval, sero-positivity increased to 75% post the 3^rd vaccination. Based on our data, and since the humoral response to the vaccine is effective about 14 days after vaccination (34), we recommend that immunization should be timed to the window of 5 months after the last ocrelizumab dose and two weeks before the next dose of ocrelizumab. It was recently suggested that measuring the association between IgG response and B cell reconstitution, rather than the time interval since infusion, is more useful for determining the optimal timing of vaccination administration in ocrelizumab-treated patients (31). A recent study found that the mRNA-1273 vaccine, containing 100 micrograms of the spike protein mRNA, induced higher IgG levels in MS patients compared to the BNT162b2 mRNA vaccine which contains only 30 micrograms of mRNA (20). Based on this finding, the investigators suggested that the mRNA-1273 vaccine would be preferable as vaccination booster in PwMS on anti-CD20 therapy or fingolimod, who did not develop efficient humoral responses following BNT162b2 vaccination.

Our findings indicate that the vast majority of patients treated with fingolimod fail to mount an IFNγ-T-cell immune response. Tortorella et al. similarly found a T-cell response in only 14.3% of fingolimod-treated patients (21). With both reduced or insufficient humoral and cell-mediated immune responses, these patients may be at increased risk of COVID-19 infection and severe disease. Interestingly, it has been suggested that the immunosuppressive effects of fingolimod could be beneficial for prevention of acute respiratory distress syndrome in patients with severe COVID-19, by reducing the cytokine storm (35), and in a recent study fingolimod-treated PwMS were not found to be at increased risk of severe COVID-19 (19). Our data did not confirm a previous report of correlation between IgG response and lymphocyte count in fingolimod-treated patients (20).

Our study has certain limitations. Due to the rapid vaccination program in Israel, the serum samples after both the 2^nd and the 3^rd vaccines were collected during a variable and rather long time period (2-16 weeks, median ~7 weeks) when patients visited the outpatient clinics, and thus the antibody measurement does not represent the peak of post-vaccination humoral response in every patient. Since there is a natural reduction in IgG levels with time, adjustment for time between vaccination and blood collection was integrated in the statistical analysis. The number of patients treated with other than fingolimod or ocrelizumab S1PR-modulators or anti-CD20 therapies was small, thus limiting any conclusion-making on their effects on the vaccine response. The number of patients assessed for T-cell mediated responses (assessed only in a sub-group of patients) was also low and not sufficiently representative for each DMT, but still our findings add to information emerging from other sites. Data on COVID-19 infection before and during the follow-up period were based only on reports by the participants attending the clinics, and was not intended for interpretation regarding the efficacy of vaccination and the risk of infection in our study population. Since our samples were collected before the outbreak of the latest Omicron variant in Israel, which seems to have the capability to escape immune-surveillance, the impact of the vaccination program on Omicron and future variants is yet to be determined. However, the information on the immune responses of PwMS after the first 3 mRNA vaccines is likely to be relevant for future vaccine development.

The merit of this study is the longitudinal follow-up of IgG levels and cell-mediated responses after 2 vaccinations and the response to a 3^rd vaccine booster in PwMS treated with a wide variety of DMTs. While for most DMTs, the humoral and cell-mediated responses appear to be similar to those of untreated PwMS, the finding of sero-negativity in >50% of patients treated with the S1PR-modulator fingolimod and the anti-CD20 therapy ocrelizumab even after a 3 vaccine booster, along with the lack of cell-mediated immune response in the vast majority of fingolimod-treated patients, may suggest that the strategy of boosting the immune system with additional vaccine doses is not effective enough for these patients and that other vaccination strategies should be considered. Such considerations could include optimization of the timing between vaccine administrations, specifically in the context of MS immunotherapies, use of combinational vaccines based on different development platforms and targets, optimization of the vaccine dose and choice of appropriate vaccine in relation to its ability to induce higher antibody levels. In general, efforts should be focused on development of optimal vaccine strategies aiming at improving immunogenicity and long-lasting immunity, tailored for each PwMS under treatment with a specific immunotherapy. Based on the cumulative data until today, updated recommendations about the type and timing of SARS-CoV-2 vaccinations of MS patients are needed.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Helsinki Committee, Barzilai Medical Center; Helsinki Committee, Carmel Medical Center; Helsinki Committee, Hadassah Medical Center; Helsinki Committee, Tel Aviv Sourasky Medical Center; Helsinki Committee, Rabin Medical Center;. The patients/participants provided their written informed consent to participate in this study.

The Israeli Neuroimmunology Study Group on COVID-19 vaccination in Multiple Sclerosis:

Lea Glass-Marmor, Carmel Medical Center; Anat Volkovitz, Carmel Medical Center; Sara Dishon, Carmel Medical Center; Zeev Dishon, Carmel Medical Center; Netta Kugelman, Carmel Medical Center; Zeev Nitsan, Barzilai Medical Center; Marwan Alkrenawi, Barzilai Medical Center; Nurit Hovel, Barzilai Medical Center; Nir Michal, Barzilai Medical Center; Vered Loew-Shavit, Barzilai Medical Center; Lital Mizrahi, Barzilai Medical Center; Marsel Zafrani, Barzilai Medical Center; Panayiota Petrou, Hadassah Medical Center; Nour Eddine-Yaghmour, Hadassah Medical Center; Ariel Ginzberg, Hadassah Medical Center; Ibrahim Kassis, Hadassah Medical Center; Michelle Halimi, Hadassah Medical Center; Keren Regev, Tel Aviv Sourasky Medical Center; Hadar Kolb, Tel Aviv Sourasky Medical Center; Ifat Vigiser, Tel Aviv Sourasky Medical Center; Yoav Piora, Tel Aviv Sourasky Medical Center; Irina Komarov, Tel Aviv Sourasky Medical Center; Avigail Hindi, Tel Aviv Sourasky Medical Center; Adi Wilf-Yarkoni, Rabin Medical Center; Itay Lotan, Rabin Medical Center; Elia Uri, Israel Institute for Biological Research; Shaher Rotem, Israel Institute for Biological Research; Hila Cohen, Israel Institute for Biological Research.

AM conceptualized and designed the study, contributed to acquisition and interpretation of the data and revised the manuscript for intellectual content. ES-R designed the study, analyzed and interpreted the data, drafted the manuscript. RM, DK and AK contributed to the design of the study, acquisition and interpretation of the data, and revised the manuscript for intellectual content. MH contributed to acquisition and interpretation of the data and revised the manuscript for intellectual content. EB-H contributed to the acquisition and analysis of the data. The Israeli Neuroimmunology Study Group on COVID-19 vaccination in Multiple Sclerosis contributed to acquisition of the data. All authors contributed to the article and approved the submitted version.

This study was coordinated and funded by the Israel Neuroimmunological Society through grants received by Merck, Roche and Novartis companies. The funding sources were not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.