Association Between Patient-Reported Outcomes Prior to Chimeric Antigen Receptor (CAR) T-Cell Therapy and Clinical...

Chimeric antigen receptor T-cell therapy (CAR-T) is a transformative therapy for relapsed/refractory hematologic malignancies but often results in significant toxicities. While patient-reported outcomes (PROs) are associated with outcomes in patients receiving other oncologic therapies, the relationship between pre-CAR-T PROs and CAR-T outcomes remains unknown.

We aimed to analyze the association between pre-CAR-T infusion PROs (Quality of life [QOL], physical symptoms, anxiety, depression, post-traumatic stress disorder [PTSD] symptoms, physical symptoms) and clinical outcomes in CAR-T, including cytokine release syndrome (CRS), neurotoxicity, hospital length of stay (LOS), and overall survival (OS).

We conducted a secondary analysis of a longitudinal study of 100 adults 18+ years with relapsed/refractory hematologic malignancies receiving CAR-T at a single academic center. We assessed QOL (Functional Assessment of Cancer Therapy-General), mood (Hospital Anxiety and Depression Scale), and physical symptoms (Edmonton Symptom Assessment Scale-revised) prior to CAR-T infusion. We assessed the association of pre-CAR-T PROs with CRS (grade 2+), neurotoxicity (yes or no), LOS, and OS using univariate models and multivariable models adjusting for patient-, disease-, and treatment-factors. We assessed all PROs continuously.

The median age was 66 (range 23-90) years, 37% of patients were female, and the majority were White (87%) and married/partnered (77%). The most common diagnosis was non-Hodgkin lymphoma (70%) followed by multiple myeloma (28%). The most frequently used CAR-T products were tisagenlecleucel (34%), followed by lisocabtagene maraleucel (16%), axicabtagene ciloleucel (13%), and idecabtagene vicleucel (12%). CRS occurred in 76% (26% grade 2+) and neurotoxicity occurred in 33%. The median LOS for CAR-T was 14.5 days. In multivariable analyses, greater pre-CAR-T QOL was associated with lower risk of CRS (odds ratio [OR] 0.97, P = .03) and neurotoxicity (OR = 0.97, P = .03); while greater depression symptoms pre-CAR-T (OR = 1.15, P = .02) were associated with higher risk of neurotoxicity. In multivariable analyses, worse physical symptoms pre-CAR-T (HR 1.03, P = .04) were associated with worse OS. There was no significant association between QOL, depression or anxiety symptoms with OS. Pre-CAR-T PROs were not associated with LOS.

Pre-CAR-T PROs are associated with OS post-CAR-T and risk of CRS and neurotoxicity in CAR-T recipients. These findings underscore the potential utility of pre-CAR-T PROs as important prognostic factors for CAR-T outcomes.