The epidemiology of cytomegalovirus (CMV) after chimeric antigen receptor-modified T-cell immunotherapy (CARTx) is poorly understood owing to a lack of routine surveillance.
We prospectively enrolled 72 adult CMV-seropositive CD19-, CD20-, or BCMA-targeted CARTx recipients and tested plasma samples for CMV before and weekly up to 12 weeks after CARTx. We assessed CMV-specific cell-mediated immunity (CMV-CMI) before and 2 and 4 weeks after CARTx, using an interferon γ release assay to quantify T-cell responses to IE-1 and pp65. We tested pre-CARTx samples to calculate a risk score for cytopenias and infection (CAR-HEMATOTOX). We used Cox regression to evaluate CMV risk factors and evaluated the predictive performance of CMV-CMI for CMV reactivation in receiver operator characteristic curves.
CMV was detected in 1 patient (1.4%) before and in 18 (25%) after CARTx, for a cumulative incidence of 27% (95% confidence interval, 16.8-38.2). The median CMV viral load (interquartile range) was 127 (interquartile range, 61-276) IU/mL, with no end-organ disease observed; 5 patients received preemptive therapy based on clinical results. CMV-CMI values reached a nadir 2 weeks after infusion and recovered to baseline levels by week 4. In adjusted models, BCMA-CARTx (vs CD19/CD20) and corticosteroid use for >3 days were significantly associated with CMV reactivation, and possible associations were detected for lower week 2 CMV-CMI and more prior antitumor regimens. The cumulative incidence of CMV reactivation almost doubled when stratified by BCMA-CARTx target and use of corticosteroids for >3 days (46% and 49%, respectively).
CMV testing could be considered between 2 and 6 weeks in high-risk CARTx recipients.