Reply to: "COVID-19 vaccination in IMID patients receiving rituximab: A personalized regimen should be formulated".

To the Editor: We read with interest the letter submitted by Yang et al in response to our article. The authors propose a personalized vaccination schedule to address COVID-19 susceptibility among patients with immune-mediated inflammatory diseases treated with rituximab, which may blunt the humoral immune response to the COVID-19 vaccine. In their article, the authors cite mounting evidence of a higher risk of infection among rituximab-treated patients and raise concerns that our study was underpowered to detect this risk and did not account for behavior modifications among these vulnerable patients.

To address these concerns, we emphasize that our study was a unique matched study comparing the risk of infection and overall mortality within a large cohort of patients (n = 7361) treated with biologics during the initial wave of the pandemic in the state of Massachusetts before any availability of vaccines. We report the risk of infection by medication class in the supplemental materials (Supplementary Table VII, available via Mendeley at https://data.mendeley.com/datasets/w4478kftkk/1). There was a small suggested increased risk of infection with rituximab, although it was not statistically significant (odds ratio, 1.16; 95% CI, 0.73-1.83; P = .53). Based on the 1.4% rate of COVID-19 in our control group of 75,961 patients, assuming 80% power and 90% CI, a statistically significant difference would require a relative risk of infection of at least 1.61 for rituximab-treated patients. As the authors point out, impaired immune response to the vaccine among rituximab-treated patients, particularly those vaccinated within a few months of infusion, is now well recognized. The authors reference a 2.28-fold increase in the infection risk among rituximab-treated patients but do not provide a citation; thus, it is unclear whether vaccinated or unvaccinated patients were compared and whether the risk was adjusted for established infection risk factors. We recognize that our results are not generalizable to vaccinated patients. However, the validity of our results is supported by a retrospective cohort study of 1895 rituximab-treated patients with multiple sclerosis and 4.81 million controls, which did not find higher infection or mortality rates in the study group. Similarly to our cohort, this cohort from California was assessed during a period when vaccines were not available.

Regarding the second concern raised by the authors, we fully acknowledge the potential for behavior modifications to have affected our results. We urged readers to consider our results in the context of the real-world data that they were based upon, with particular attention to the risk reduction strategies reported in surveys of patients with inflammatory bowel and rheumatic diseases.

Both humoral and cellular immunity play a role in protection against COVID-19. Interestingly, CD8+ T-cell–mediated immunity has been found in the absence of circulating B cells.3, 4, 5 Further research into the benefits of cellular immunity in the absence of humoral immunity is needed to guide vaccination among patients for whom vaccination before rituximab treatment is not possible.

We thank the authors for their interest in our study. Their recommendations for a tailored approach for COVID-19 vaccination in patients treated with rituximab are evidence-based and should be considered in the management of rituximab-treated patients.

Conflicts of interest

None disclosed.