P1444: IMPACT OF CO-MORBIDITIES ON NON-RELAPSE MORTALITY AND CYTOKINE-RELEASE SYNDROME AFTER CD19 CAR-T CELL THERAPY - A RETROSPECTIVE STUDY FROM THE TCWP OF THE EBMT AND GOCART COALITION

Abstract

Background: CAR T-cell therapies have been associated with a profile of significant toxicities and previously unknown side effects. For that reason, the registration of clinical trials with commercial and non-commercial products included very strict inclusion and exclusion criteria, to minimize severe adverse events and maximize therapeutic effect. The real-life analyses revealed that a proportion of patients are treated outside of these specifications, with variable outcomes. Aims: The goal of this study was to determine the cumulative incidence of non-relapse mortality (NRM) and cytokine-release syndrome (CRS) after anti-CD19 CAR T-cell therapy, and impact of patients’ co-morbidities expressed as the hematopoietic cell transplantation-co-morbidity index (HCT-CI) scores, on the major outcome measures. Methods: This is a retrospective multicentre analysis from the Transplant Complications Working Party (TCWP) of the EBMT using the data of the GoCART coalition. Eligibility criteria for this analysis include diffuse large B-cell lymphoma (DLBCL) and acute lymphocytic leukemia (ALL) patients that underwent CD19 CAR T-cell therapy in an EBMT member centre before 31 January 2021. Exclusion criteria were lack of information on comorbidities, survival or relapse status, or age <18 years. Outcomes were measured from the date of CAR T-cell infusion for all endpoints. NRM was defined as death without relapse/progression. Cumulative incidence functions were used to estimate the incidence of complications. Multivariate analyses were performed using the Cox cause-specific model for all endpoints. The primary endpoint was NRM at 3 and at 12 months post CAR T-cell therapy, while secondary endpoints included cumulative incidence of CRS as well as NRM and CRS incidence in subgroups based on HCT-CI. Statistical analyses were performed with R 3.4.2 software packages. Results: The study identified 586 patients treated with CAR T-cells in the years 2018-2021 for DLBCL (96.4%) or ALL (3.6%). The median age was 60 (range, 18-84). 166/437 (38%) patients had a Karnofsky performance status <90%, while the HCT-CI was available for 483 (82%) patients and was intermediate (1-2) in 17.6% and high (≥3) in 13.9% of these patients. The most common co-morbidities were pulmonary (16.1%), cardiac (6.8%), obesity (6.8%), diabetes (6.4%), arrythmia (5.5%), history of solid tumor (5.3%), and active infection before treatment (4.3%). The median follow up was 12.5 months. The NRM was reported to be very low: 1.7% (95%CI 0.9-3) at 3 months and 2.9% (95%CI 1.7-4.5) at 12 months. The cumulative incidence of CRS (all grades) was 35.3% (95%CI 30.6-40.1) at 3 days and 82.7% (95%CI 78.5-86.1) at 12 days. Unexpectedly, no significant impact of HCT-CI on NRM, CRS incidence, or overall survival was observed (Fig. 1) both in univariate and multivariate analyses. Most deaths were caused by the original disease (75.4%) with concomitant infection (5.9%) or with multiorgan failure (4.3%). Image:Summary/Conclusion: Our real-life data from the GoCART registry revealed the profile and frequency of co-morbidities of patients undergoing CAR T-cell treatments at EBMT centers. While the HCT-CI score was high in a substantial proportion of patients, it did not influence any major outcome. The incidence of CRS was also high and not influenced by the HCT-CI. Our preliminary results might support application of more liberal criteria for patient qualification for CAR T-cell therapies based on co-morbidities, but this needs to be confirmed in further studies.