Cellular responses to SARS-CoV-2 vaccination after B-cell depletion: conflicting results from studies – Authors' reply

We thank Pavan Bhargava for pointing out an error in the Methods section of our Article. For the interferon γ (IFNγ) release assay, cells were indeed stimulated for 16 h, not 1 h. The original Article¹ has now been formally corrected. Bhargava then raises several points that we find difficult to agree with.

First, Bhargava appears to question the merit of studying a mixed population of patients that have different disease aetiologies and that are predominantly under co-treatment with additional immunosuppressives. We believe we have addressed this point, however. In the appendix of our Article, we present subanalyses for the different disease subpopulations and co-treatments.

Next, our analysis included only four patients with multiple sclerosis that were treated with ocrelizumab. The study by Gadani and colleagues²—one of the key studies Bhargava cited to support their argument—exclusively focused on patients with multiple sclerosis, the majority of whom were treated with ocrelizumab monotherapy, and hence the results of these two studies cannot be compared. As stated in the appendix of our Article (appendix p 7), ten (56%) of 18 participants on anti-CD20 monotherapy mounted a positive IFNγ release (table). Since our Article was published, we have completed analyses in more patients; the full dataset shows a positive IFNγ release in 41 (44%) of 93 patients.

Furthermore, the results of different measures of cell-mediated immunity are not interchangeable and methodological differences might yield different results—as perhaps Bhargava was also alluding to. For example, previous literature on tuberculosis showed differences between the whole-blood-based quantitative IFNγ release assays and the more sensitive, but only semiquantitative, ELISpot analyses from peripheral blood mononuclear cells.³ Most of the published studies on SARS-CoV-2-specific T-cell responses so far have used ELISpot assays. This highly sensitive, yet only semiquantitative, assay measures the number of T cells that recognise a given antigen. In our study, however, we measured the total amount of IFNγ released after stimulation. This approach integrates the number of reactive T cells with the amount of IFNγ produced per cell. Increasingly, data show that anti-B-cell monotherapy also affects T cells.⁴ A recent study by Apostolidis and colleagues⁵ reported a detectable, but altered, CD4+ T-cell response in patients on anti-CD20 therapies but also found a higher CD8+ T-cell response. Thus, it seems possible that, overall, B-cell depleting therapies impair the functional quality of T cells rather than their quantity. This finding is poorly understood for now, but points towards a key role of B cells in the promotion of cellular immunity.

Finally, as age is an important influencing factor regarding the strength of immune responses, it is worth considering that our anti-CD20-depleted patient population had a median age of 67 years, whereas the anti-CD20 treated population in the study from Gadani and colleagues was considerably younger (median age of 48 years). In addition, Gadani and colleagues’ regression analysis was based on 30 patients and 12 controls and included four independent parameters. A model with such few participants and relatively many parameters may be at risk of overfitting, and thus might fail to correctly adjust for the age difference. Moreover, of the 38 patients, none of the 30 in whom cellular responses were analysed had been subjected to a serological assessment of previous SARS-CoV-2 exposure, as far as reported in Gadani and colleagues’ paper.

We declare no competing interests.