AUTHORS
Jeet Patel, DO, Matthew P Dandan, DO2, Muhammad Kashif Amin, MD, MSc, FACP3
AFFILIATIONS
1 The University of Kansas Cancer Center, Kansas City, Kansas, United States, The University of Kansas Medical Center, Olathe, Kansas, United States
2 Medical College of Wisconsin, Milwaukee, Wisconsin, United States
3 The University of Kansas Cancer Center, Kansas City, Kansas, United States
ABSTRACT
Background: Idecabtagene vicleucel (ide-cel) is a standard B-cell maturation antigen (BCMA)–directed chimeric antigen receptor (CAR)-T cell therapy for relapsed/refractory multiple myeloma (RRMM). These heavily pretreated, frequently hypogammaglobulinemic patients are at substantial infection risk, yet real-world predictors of infection after commercial ide-cel have not been defined at registry scale, with prior data largely limited to single-center studies.
Objective: To characterize the incidence and predictors of clinically significant infection after ide-cel for RRMM, and its association with progression-free survival (PFS) and overall survival (OS) using the Center for International Blood and Marrow Transplant Research (CIBMTR) registry.
Methods: In this retrospective multicenter study using the publicly available CIBMTR dataset (P-6000), we analyzed 821 patients receiving commercial ide-cel for RRMM (infused 2021–2023). The outcome was clinically significant infection (first-year reporting). Baseline comparisons used chi-square and Mann-Whitney-U tests. Pre-infusion predictors were evaluated by univariable logistic regression; variables with p<0.05 on univariable analysis were considered for multivariable modeling. ECOG was excluded due to collinearity with the frailty index, and marrow plasma-cell burden was excluded due to high missingness (38%). Missing covariates were handled by multiple imputation (chained equations, M=25, Rubin’s rules), with complete-case sensitivity analysis. OS and PFS were assessed by Kaplan-Meier estimation and Cox regression; cause-specific mortality and relapse versus treatment-related mortality (TRM) were estimated by competing-risks (Aalen-Johansen) methods. Multivariable Cox models evaluated infection alongside these predictors for OS/PFS, with a landmark sensitivity analysis excluding early deaths. Cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS) were treated as peri-infusion factors. Analyses used Python.
Results: Median age was 66 years (31% were ≥70); 42% were frail; median follow-up was 11.6 months. Clinically significant infection occurred in 367/821 patients (44.7%; 95% CI 41.3–48.1). On univariable analysis, infection was associated with frailty (OR 1.42), ECOG ≥2 (OR 2.89), comorbidity (OR 1.60), hemoglobin <8 g/dL (OR 1.98), platelets <50×10?/L (OR 2.14), marrow plasma cells >50% (OR 1.81), and elevated LDH (OR 1.61), but not age ≥70 (OR 1.00); peri-infusion CRS (OR 1.85) and ICANS (OR 1.70) were also associated. After multivariable adjustment and imputation, only thrombocytopenia (aOR 1.65; 95% CI 1.00–2.72; p=0.048) and elevated LDH (aOR 1.50; 95% CI 1.08–2.08; p=0.015) remained independent; frailty’s association attenuated and was no longer significant (aOR 1.21; p=0.26). Overall response rate was 73.1% and complete response rate was 25.2%. Median OS was 15.6 months, and median PFS was 8.8 months. Infection accounted for only 7.9% of deaths (vs. 75.8% due to disease progression); 12-month cumulative incidence of infection-related death was ~2.5% versus ~30% for other-cause death. Relapse, not TRM, dominated PFS events (12-month cumulative incidence 55.7% vs. 4.8%). Infection was associated with inferior OS and PFS on univariable analysis (HR 1.47 and 1.38, respectively); however, in multivariable models that included the same burden markers, this association attenuated toward the null (aHR 1.20, non-significant), while thrombocytopenia and elevated LDH independently predicted both OS and PFS in their own right (aHR 1.3–1.8, p=0.001–0.01).
Conclusion: Clinically significant infection affects nearly half of commercial ide-cel recipients but accounts for a small minority of deaths, which are chiefly driven by disease progression and relapse. Frailty was associated with infection on univariable analysis, but this effect attenuated after adjustment; thrombocytopenia and elevated LDH were the only independent predictors, and independently predicted OS/PFS as well. These routinely available pre-infusion markers can flag high-risk patients for intensified monitoring and prophylaxis. Infection's own association with survival was largely explained by this shared burden rather than reflecting an independent effect. Findings support burden-based, risk-adapted infection prevention after ide-cel; validation incorporating infection timing and pathogen-level detail is warranted.
DISCLOSURES
Jeet Patel, DO: No relevant conflicts of interest to declare.