Symptomatic COVID-19 course and outcomes after three mRNA vaccine doses in multiple sclerosis patients treated with high-efficacy DMTs

Prior to widespread vaccination, pwMS with higher disability and those on anti-CD20 therapies had increased risks of severe COVID-19 outcomes (hospitalization, ICU, death). With mRNA vaccines, most DMTs, including some high-efficacy agents such as natalizumab (NTZ), do not substantially blunt vaccine responses; however, anti-CD20 therapies and S1P-receptor modulators (e.g., fingolimod, FNG) reduce humoral responses, and S1P modulators may also impair cellular responses. After two-dose vaccination, breakthrough infections and more severe courses appeared more frequent in pwMS on OCR and S1P modulators. Given the booster’s potential to revive immune responses, this study investigated whether a third mRNA vaccine dose mitigates COVID-19 severity in pwMS on OCR and FNG, compared with NTZ. Primary outcome: hospitalization, ICU admission, death. Secondary outcomes: symptom type/duration and need for symptomatic or COVID-19-specific therapies.

Existing literature shows: (1) anti-CD20 and S1P modulators dampen humoral (and for S1P modulators, cellular) responses to SARS-CoV-2 mRNA vaccines; (2) pwMS on most other DMTs, including NTZ, mount adequate responses; (3) breakthrough infections after two doses are more frequent in pwMS on OCR/S1P modulators, with OCR linked to more severe outcomes. Preliminary studies suggest boosters can revive immune responses in these groups, but real-world effectiveness against severe COVID-19 after a third dose remained insufficiently characterized prior to this study.

Design: Multicentre longitudinal observational study at 17 Italian MS centers (RIREMS). Period: Data collection September 2021–July 2022. Inclusion criteria: pwMS ≥18 years; treated with NTZ, OCR, or FNG; no DMT change from first vaccine dose through study end; laboratory-confirmed COVID-19 (molecular or antigen swab) ≥14 days after third mRNA vaccine dose. Exclusion: missing COVID-19 duration/outcome data; steroid treatment within 1 month prior to any vaccine dose or COVID-19; COVID-19 before the booster. Data collected: demographics and MS characteristics (age, sex, onset, type, EDSS, disease/treatment duration; for OCR, last infusion date); comorbidities (e.g., cardiovascular disease, diabetes, obesity, CKD, liver disease, COPD, cancer, autoimmune thyroiditis); vaccination details (dates and type: BNT162b2 or mRNA-1273); COVID-19 details (diagnosis date/test type, symptoms, medications including NSAIDs, oxygen, steroids, antibiotics, anti-SARS-CoV-2 monoclonal antibodies—regdanvimab, sotrovimab, casirivimab-imdevimab—care setting, outcomes including symptom duration, recovery status, death). Statistical analysis: Descriptive statistics; ANOVA with Scheffé procedure for continuous variables; Fisher’s exact test for categorical variables. Linear and logistic regression models assessed associations between DMT group and COVID-19 course/outcomes (dependent variables), adjusting for age, sex, comorbidities, EDSS, DMT duration, and interval between primary vaccination cycle and booster. Significance set at p=0.05; Benjamini–Hochberg correction for multiple comparisons. Sensitivity analyses excluded (1) progressive MS phenotypes and (2) pwMS treated with anti-SARS-CoV-2 monoclonal antibodies.

- Cohort: 290 pwMS with COVID-19 after booster (79 NTZ; 126 OCR; 85 FNG) from 17 centers. Most received the same vaccine type for primary and booster (BNT162b2 n=236; mRNA-1273 n=18); heterologous boosters occurred in 36 cases. - Group characteristics: NTZ patients were younger; OCR patients had higher EDSS and more progressive phenotypes; FNG patients had longer disease and treatment durations. - Booster timing: Mean days from primary series to booster differed (FNG 194±88 vs NTZ 227±78, p=0.028; OCR 207±69). Breakthrough infection occurred later after booster in FNG (mean 96.5±67.9 days) vs NTZ (64.1±50.3, p=0.001) or OCR (73.0±48.3, p=0.01). - Symptoms: Overall asymptomatic rate 8.3% (no significant group difference). Symptom duration longer in OCR vs NTZ (p=0.008). NTZ group reported lower frequencies of sore throat, myalgia, loss of taste/smell, fever >38°C, cough, and dyspnea than OCR/FNG. - Treatments during COVID-19: Higher use of NSAIDs, steroids in OCR/FNG vs NTZ; antibiotics and oxygen more frequent in OCR vs NTZ; anti-SARS-CoV-2 monoclonal antibodies used more in OCR (11.1%) and FNG (9.4%) vs NTZ (0%). - Outcomes: Hospitalization rates: OCR 11.1%, FNG 7.1%, NTZ 1.3% (overall p=0.02; NTZ vs OCR p=0.006), but differences did not persist after multiple-comparison correction. ICU admission occurred in 1 patient (OCR); no deaths. - Adjusted analyses: After controlling for age, sex, comorbidities, EDSS, DMT duration, and vaccination interval, many observed differences (including hospitalization and symptom duration) were associated with demographic/clinical factors (age, EDSS) rather than DMT type. DMT-related effects persisted for certain variables (time from booster to infection, loss of taste/smell, cough, and prescription of monoclonal antibodies). - Sensitivity analyses: Excluding progressive MS or patients receiving COVID-19 monoclonal antibodies did not materially change results.

Although crude comparisons suggested somewhat worse symptom burden and higher resource use among pwMS on OCR and FNG vs NTZ, adjusted analyses indicated that key outcomes—especially hospitalization and symptom duration—were primarily driven by age and disability (EDSS), not DMT class. This contrasts with pre-vaccination or two-dose era reports where anti-CD20 therapy associated with more severe COVID-19 courses. The likely contributors include the protective effect of a third booster dose reviving immune responses even in OCR/FNG-treated pwMS and the predominance of the Omicron variant during the study period, which is generally associated with milder disease. Persisting DMT-linked differences for cough, loss of taste/smell, and monoclonal antibody use may reflect residual differences in immune response profiles and clinical practice patterns. Overall, findings support the real-world effectiveness of booster vaccination in reducing severe COVID-19 among pwMS on OCR and FNG.

In the largest cohort to date assessing breakthrough COVID-19 after three mRNA vaccine doses in pwMS on OCR, FNG, and NTZ, severe outcomes were uncommon and no deaths occurred. After adjusting for clinical and demographic factors, hospitalization risk and symptom duration did not differ by DMT class, suggesting that a third booster dose confers meaningful protection against severe disease in pwMS on OCR and FNG. These results reinforce booster vaccination strategies for pwMS receiving high-efficacy DMTs known to blunt vaccine immunogenicity. Future work should integrate immunologic correlates (humoral and cellular) and compare outcomes after two vs three doses.

- Potential selection bias: participants were identified after COVID-19 diagnosis, likely underrepresenting asymptomatic infections and precluding estimation of true breakthrough rates by DMT. - Possible unmeasured confounding from prior asymptomatic infections before booster, which could affect immune responses. - No comparison group with COVID-19 after only two vaccine doses; unable to directly quantify incremental benefit of the booster. - Lack of humoral and cellular immune response measurements to correlate with clinical outcomes.