Interrupting methotrexate to improve immunity after COVID-19 booster vaccination: is it really worth it?

Abhishek and colleagues conducted an important clinical trial (the VROOM study), examining the effect of a 2-week interruption of methotrexate on the antibody response to the COVID-19 booster vaccination. The rigorous study design yielded insightful findings regarding the effects of this intervention on vaccine immunogenicity. The antibody concentrations of receptor-binding domain (RBD) at 4 weeks in the suspended methotrexate group were more than twice as high as in the continued methotrexate group (geometric mean ratio 2·2 [95% CI 1·6–3·0]). Despite this clear statistical difference, antibody concentrations showed inherent variability, which was reflected by considerable standard deviations (mean 34 556 U/mL [SD 38 323] in the suspended methotrexate group vs 17 682 U/mL [20 872] in the continued methotrexate group). The authors postulated a greater vaccine efficacy following the strategy of suspending methotrexate. However, antibody concentrations are ultimately a surrogate parameter of uncertain clinical significance. The Article refers to the dose–response association between RBD antibody titres following vaccination and protection against COVID-19,2, 3 but this non-linear correlation was shown for RBD antibody concentrations up to 6000 binding antibody units per mL, which is far below the magnitudes observed in the VROOM study. Presumably, the correlation curves plateau at higher antibody titres. Furthermore, these historical correlations have been observed for the alpha (B.1.1.7) SARS-CoV-2 variant. Although higher concentrations might be needed to neutralise new variants of concern, such as the omicron (B.1.1.529) variant, the correlation between clinical protection and binding antibody units per mL is even less understood.

By contrast, concerning self-reported disease flares in the first 4 weeks, the number of flares needed to harm for discontinuing methotrexate was four (56% in the suspended methotrexate group vs 31% in the continued methotrexate group; absolute risk increase 25%). Due to the open-label study design, self-reported flares might overestimate the true effect of suspending methotrexate. However, such a subjective sense of less controlled disease is also to be expected in real-world clinical practice. In the present study, slightly more patients in the suspended methotrexate group sought medical assistance because of flares than did individuals in the continued methotrexate group (six [4·7%] of 127 vs three [2·4%] of 127); however, in real life, this could mean an additional burden on the health-care system, which would be a counterproductive effect during the pandemic.

Clinical endpoints regarding increased protection against COVID-19 were not reported in the suspended methotrexate group. Thus, we have to balance the unknown clinical benefit of a biomarker with a real—albeit only temporary—clinical harm for 25% of patients (absolute risk increase). Therefore, we believe the conclusions that endorse the discontinuation of methotrexate before COVID-19 booster vaccination in this patient population are too optimistic. The relevance of the findings of the VROOM study and similar studies relying on antibody concentrations needs to be shown in long-term studies.

We declare no competing interests.