The “Walk through the Clinic Door” Effect: Differences in Survival and Toxicity Based on Site of CAR T Administration in a Default Outpatient CAR T Program

Introduction: Due to the potential of long-term PFS, CAR T-cell therapy is frequently given to patients with suboptimal performance status (PS). Differences in outcomes regarding PS of 2 or greater have been demonstrated in real world analyses of CAR T outcomes. However, it is unknown if patients who require hospitalization at the time of CAR T-cell infusion due to non-CAR T comorbidities have differential outcomes compared to patients who would otherwise be able to “walk into clinic.” To date, many patients are infused with CAR T as inpatients due to monitoring needs, not necessarily due to medical status. The Intermountain Health CAR T-cell program's default infusion setting is outpatient for all commercial CAR T-cell products, thus allowing for a more accurate evaluation of the association between true inpatient status and CAR T-cell therapy related outcomes. Methods: Patients were identified from scheduling and billing records. Patient's oncology notes for CAR-T 30-day, 100-day and 1 year follow up visits were reviewed to evaluate setting of CAR-T infusion, treatment response and toxicity. Survival was determined by any death record within the IH EMR and the patient's last known visit date. Two proportion pooled z-tests were calculated to compare OS rates between inpatient and outpatient groups. Results: There were 125 patients included in this analysis. 5 patients received CAR T for ALL, 98 patients received CAR T for B-cell lymphoma, and 22 patients received CAR T for multiple myeloma. 27/125 patients (22%) received CAR T-cell therapy in the inpatient setting, and 98/125 (78%) were infused at as outpatients. Overall survival (OS) was significantly inferior for patients who received CAR T-cell therapy as inpatients vs outpatients, with 30 -day OS of 81% vs 97% (p=0.02), 90- day OS of 56% vs 92% (p<0.01), and 1 year OS of 30% vs 77% (p<0.01). 1 year CR rates were 19% in inpatients vs 48% in outpatients. Among lymphoma patients, 7/21 (33%) of those who received inpatient CAR T were alive at 1 year, vs 59/77 (77%) of outpatients (p<0.01). For multiple myeloma patients, 1 year OS for those who received inpatient infusion was 25% (1/4 patients), vs 72% (13/18) for outpatients (p=0.02). Grade 3+ CRS in inpatients was 30% for inpatients vs 2% in outpatients (p,0.01). Grade 3+ ICANS was 26% in inpatients vs 10% in outpatients (p=0.04). Conclusion: Patients who require inpatient CAR T-cell administration due to performance status experienced significantly inferior OS and rates of CR at 1 year compared to patients who received CAR T-cell therapy in the outpatient setting. Rates of high-grade CRS and ICANS were significantly increased in patients who required inpatient CAR T administration. There appears to be a “walk through the clinic door” effect that may better capture CAR T-cell treatment risk for both outcomes and toxicity.