Chimeric antigen receptor (CAR) T-cell therapy has revolutionized the treatment landscape of relapsed/refractory B-cell lymphomas. Despite its efficacy, it induces profound immune suppression, predisposing patients to opportunistic infections. Cytomegalovirus (CMV) reactivation is a known complication in immunocompromised individuals, yet its specific prognostic impact following CAR T-cell therapy in lymphoma remains insufficiently characterized.

To investigate the role of CMV reactivation in this setting, we analyzed 1124 lymphoma patients treated with commercial CAR T-cell products and enrolled in the Italian prospective observational CART-SIE study (NCT06339255). All CMV reactivations within 3 months of CAR T infusion were taken into account as well as their association to treatment response, overall survival (OS), progression-free survival (PFS), cytokine release syndrome (CRS), and immune effector cell-associated neurotoxicity syndrome (ICANS), infections, immune effector cell-associated hematotoxicity (ICAHT) and use of immunosuppressive therapies. A 3-month landmark analysis was performed for OS and PFS. Univariate and multivariate Cox proportional hazards models were used to identify independent risk factors. Median follow-up was 16.5 months (95% CI: 13.2–17.8).

CMV reactivation occurred in 122 patients (10.9%), while 1002 did not. Notably, no cases of CMV disease were recorded in the cohort.

Baseline characteristics were well balanced between patients with and without CMV reactivation, including age >70 years (14.8% vs. 12.5%, p=0.472), male gender (66.4% vs. 63.3%, p=0.550), DLBCL diagnosis (64.8% vs. 63.3%, p=0.232), prior autologous transplant (30.6% vs. 29.4%, p=0.833), CAR-T product (axi-cel: 55.7% vs. 53.2%, p=0.340), median number of previous therapies (2 vs. 2, p=0.117) and refractory disease status at study entry (55.9% vs. 54.0%, p=0.207), respectively.

Complete response (CR) rates at day +30 and +90 were similar in patients with vs. without CMV reactivation (55.3% vs. 56.3%, p=0.484 and 58.6% vs. 63.6%, p=0.410, respectively).

However, in the 3-month landmark analysis, CMV reactivation was associated with significantly worse outcomes: 1-year OS was 70.0% in the CMV-positive group vs. 78.6% in the CMV-negative group, and 2-year OS was 52.4% vs. 67.1% (p=0.023). Similarly, 3-month landmark PFS were 61.9% vs. 71.2% at 1 year and 46.2% vs. 60.8% at 2 years (p=0.026), respectively.

Rates of CRS (90.7% vs. 87.3%, p=0.305) and ICANS (33.1% vs. 26.2%, p=0.128) were comparable between patients with and without CMV reactivation. Severity grading of both CRS and ICANS also did not differ significantly (p=0.401 and p=0.321, respectively). Similarly, the incidence of infections (13.1% vs. 11.1%, p=0.544) and ICAHT (29.5% vs. 23.3%, p=0.144) did not differ significantly between patients with and without CMV reactivation, respectively. However, patients who experienced CMV reactivation more frequently received immunosuppressive treatments, including tocilizumab (79.5% vs. 70.7%, p=0.043), anakinra (12.3% vs. 6.7%, p=0.040), and corticosteroids (59.8% vs. 45.5%, p=0.003).

In multivariate 3-month landmark analysis, after adjusting for age, gender, prior autologous transplant, number of prior treatment lines, bridging therapy, diagnosis, CAR-T product, and steroid use, CMV reactivation remained independently associated with inferior OS (HR 1.47, 95% CI: 1.03–2.11, p=0.034) and PFS (HR 1.46, 95% CI: 1.01–2.11, p=0.045).In conclusion, CMV reactivation is a relatively frequent early event following CAR T-cell therapy for lymphoma and is correlated with increased use of immunomodulatory agents. While it was not linked to differences in response rates, overt CMV disease, or an increased incidence of infectious or CAR T-related complications, CMV reactivation emerged as an independent predictor of inferior survival. These findings underscore the importance of active CMV surveillance and highlight the need for further studies investigating prophylactic or preemptive antiviral strategies to improve outcomes in this vulnerable patient population.

This content is only available as a PDF.
2025
Sign in via your Institution