Humoral response to COVID-19 vaccination in MS patients on disease modifying therapy: Immune profiles and clinical outcomes

Abstract

Background: Patients with multiple sclerosis (MS) on some disease modifying therapies (DMTs), particularly anti-CD20 and sphingosine-1-phosphate (S1P) modulators, are at increased risk of severe Coronavirus Disease 19 (COVID-19) and death. COVID-19 vaccinations are effective in preventing infection and severe disease, but humoral response to vaccination and outcomes of COVID-19 infection after vaccination in MS patients on DMTs remain less understood.

Methods: In this retrospective single-center study, patients enrolled in the CLIMB (Comprehensive Longitudinal Investigation of Multiple Sclerosis at Brigham and Women's Hospital) study and biorepository who had been vaccinated against COVID-19 and had SARS-CoV-2 spike antibody (anti-SARS-CoV-2 S Roche-Elecsys) testing were identified and compared to healthy controls. Demographic data, serum immune profiles including lymphocyte count, B-cell count, and immunoglobulins, and clinical outcome of COVID-19 infection were collected.

Results: 254 patients (73.2% female, mean (SD) age 52.9 (11.2) years) were identified. When controlling for age, time since vaccination, and vaccine type, patients on fingolimod, ocrelizumab, rituximab, mycophenolate mofetil, natalizumab and teriflunomide had significantly lower levels of spike antibodies compared to healthy controls (n = 34). Longer duration of treatment was associated with lower spike antibody levels in patients on anti-CD20 therapy (p = 0.016) and S1P modulators (p = 0.016) compared to healthy controls. In patients on anti-CD20 therapy, higher spike antibody levels were associated with higher CD20 cell count (p<0.001), and longer time since last anti-CD20 therapy infusion (p<0.001). 92.8% (13/14) vaccine responders (spike antibody titer >100 ug/dL) on anti-CD20 therapy demonstrated B-cell reconstitution (mean CD20 3.6%). Only 1 out of 86 patients with CD20 of 0% had a measurable spike antibody response to vaccination. During follow-up (mean 270 days), five patients were diagnosed with COVID-19 after vaccination (incidence 1.9%), all of whom had spike antibody < 20 ug/dL. No patients required ICU care or died.

Conclusions: Patients on some DMTs demonstrate reduced humoral immunity after Sars-CoV-2 vaccination. Longer duration of anti-CD20 therapy and reduced CD20 cell count is associated with blunted humoral response to vaccination. CD20 reconstitution >0.1% appears necessary, but not always sufficient, for humoral response to vaccination. Breakthrough COVID-19 infection in our cohort of MS patients on DMT was higher than in population studies. We propose that adjustment of B-cell therapy administration to allow for B-cell reconstitution prior to vaccination should be considered.

Keywords: Anti-CD20; COVID-19; Disease modifying therapy; Multiple sclerosis; Vaccination.

Fig. 1

Sars-CoV-2 Spike Antibody level by Disease Modifying Therapy Type. Legend: Box plots demonstrate mean (stdev) spike antibody level ( ug/dL) for patients on each DMT. *Indicates statistically significant (p<0.05) difference compared to healthy controls. Spike antibody level at our institution is reported as a single value from <0.4 to >2500  ug/dL. DMF= Dimethyl fumarate; Fin =  Fingolimod; GA= Glatiramer acetate; IFN =  Interferon beta; MM= Mycophenolate mofetil; Nat= Natalizumab; Ocr= Ocrelizumab; Ofa= Ofatumumab; Rit=Rituximab; Ter=Teriflunomide; Unt= Untreated; HC: Healthy controls. Between group comparison statistics are shown in Table 2.

Fig. 2

Relationship of Spike Antibody titer and Time on DMT Treatment. Legend: A. Association between time on treatment and spike antibody levels ( ug/dL) in subjects treated with fingolimod (estimated Spearman's correlation coefficient -0.37 (p = 0.016). B: Association between time on treatment and spike antibody levels ( ug/dL) in subjects treated with ocrelizumab or rituximab (estimated Spearman's correlation coefficient -0.19 (p = 0.016).

Fig. 3

Antibody response by IgG level in all patients. Legend: Relationship of IgG level ( ug/dL) and spike antibody level ( ug/dL) in all patients. Estimated Spearman's correlation coefficient was 0.23 (p = 0.0017). DMF= Dimethyl fumarate; Fin =  Fingolimod; GA= Glatiramer acetate; IFN =  Interferon beta; MM= Mycophenolate mofetil; Nat= Natalizumab; Ocr= Ocrelizumab; Ofa= Ofatumumab; Rit=Rituximab; Ter=Teriflunomide; Unt= Untreated.

Fig. 4

Relationship of spike antibody level and absolute lymphocyte count. Legend: Absolute lymphocyte count vs spike antibody level. A: All patients (n = 254) estimated Spearman's correlation coefficient 0.002 (p = 0.97). B: Patients on S1P modulator therapy (fingolimod, n = 42) estimated Spearman's correlation coefficient -0.04 (p = 0.80) (Fig. 3B).

Fig. 5

SARS-CoV-2 spike antibody in patients on anti-CD20 therapy. Legend: A: Association between spike antibody level and CD20 cell count (estimated Spearman's correlation coefficient 0.59 (p<0.001). B: Association between spike antibody level and days since last B-cell infusion (estimated Spearman's correlation coefficient 0.38 (p<0.001).