The prognostic value of patient-reported fatigue for survival in patients with aggressive B-cell lymphomas treated with CAR T-cell therapy

Abstract

Abstract Background Chimeric antigen receptor (CAR) T-cell therapy is a valuable treatment option for patients with aggressive B-cell lymphomas. Predicting survival is critical to better identify patients at risk and some pre-infusion clinical and laboratory prognostic factors have already been identified. However, very limited evidence exists on the potential prognostic value of patient reported health status data. Objective The primary objective of this analysis was to investigate the prognostic value of pre-infusion (baseline) patient-reported fatigue for overall survival (OS). A secondary objective was to describe baseline health-related quality of life (HRQoL) profile by fatigue severity. Methods This analysis is based on data from a prospective observational study assessing HRQoL and fatigue of adult patients with aggressive B-cell lymphomas scheduled to receive CAR T-cell therapy. At baseline, HRQoL profile was assessed with the EORTC QLQ-C30 and the QLQ-NHL-HG29. Fatigue was assessed with the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F). Univariate and multivariate Cox regression analyses were performed to identify factors associated with OS. The initial univariate model included the following baseline variables: sex, age, diagnosis, ECOG PS, presence of comorbidities, number of prior lines of therapy, disease status, bulky disease, systemic B symptoms, prior stem cell transplantation, extranodal involvement, Ann Arbor stage, time from baseline HRQoL assessment to infusion, as well as FACIT-F scores. Type of CAR T-cell product was also included as covariate. A bootstrap resampling procedure was run on 5000 models to assess the prognostic importance of each variable. For this analysis patients were categorized into lower (LF) and higher fatigue (HF) groups based on the median FACIT-F score. The median follow-up (FUP) time was reported, and OS curves were estimated using the Kaplan-Meier method, stratified by fatigue groups, and compared using the log-rank test. Results Current analysis is based on 169/170 patients with a valid baseline FACIT-F questionnaire. The median age of patients was 61 years (interquartile range [IQR]: 51–68), and 31.4% were female. The median FUP time for the overall cohort was 23 months and 40 (23.7%) died during the FUP period. Most deaths (69%) were due to disease progression, followed by infections (18%), secondary malignancies (5%), and other causes (8%). In the univariate analysis, age (HR 1.04, 95% CI 1.01–1.07, p=.013), and fatigue (HR 0.83, 95% CI 0.75–0.91, p<.001) were significantly associated with OS. Ann Arbor Stage (p=.06), diagnosis using DLBCL as reference category (PMBCL: p=.07; MCL: p=.52) and disease status (p=.09) showed trends toward association with OS but did not reach statistical significance. The final multivariate model retained three baseline variables: sex, age, and FACIT-F score. In this model only age and fatigue remained significantly associated with OS with a HR of 1.03 (95% CI 1.00–1.06, p=.036) and a HR 0.86, (95% CI 0.78–0.94, p<.001), respectively. This latter data translates into a 14% decrease in the hazard of death for every 3-points increases (clinically meaningful improvement) in the FACIT-F scores. According to the bootstrap analysis, the inclusion frequencies for the variables retained in the final multivariate model were: fatigue (91%), age (59%), and sex (28%). The HF group showed a more rapid decline in survival probability (p=.018) over time compared to the LF group. The estimated OS of the overall population at 12 months was 80% (95% CI: 74–86%). The 12-month estimated OS was 87% (95% CI: 80–95%) and 73% (95% CI: 64–83%) in the LF and HF, respectively. Patients in the HF group reported a markedly worse baseline HRQoL profile compared to those with LF. To illustrate, large clinically important differences were observed with regard to key EORTC QLQ-C30 scales, that is, role functioning (Δ = 32), global quality of life (Δ = 20.5), social functioning (Δ = 19.4), and dyspnea (Δ = 17.3). Marked differences were also observed in key scales of the QLQ-NHL-HG29 questionnaire. Conclusion Our findings point to the importance of systematically assessing patient-reported fatigue before CAR T-cell infusion in the setting of aggressive B-Cell Lymphomas, as it provides independent prognostic information for survival. Assessment of fatigue may help to early identify most vulnerable individuals who are at greater risk and in need of special attention.