P1204: TREATMENT PATTERNS AND REAL-WORLD OUTCOMES IN PATIENTS (PT) WITH LARGE B-CELL LYMPHOMA (LBCL) WHO RECEIVED SECOND-LINE (2L) THERAPY

Abstract

Background: Pts with LBCL who fail first-line (1L) treatment may be candidates for high-dose chemotherapy (HDCT)/hematopoietic stem cell transplantation (HSCT) if they respond to salvage chemotherapy. However, pts unable to undergo HDCT/HSCT have limited treatment options and often have poor outcomes. Aims: To characterize the demographics, clinical characteristics, treatment patterns, and outcomes in pts with LBCL receiving 2L therapy with or without HDCT/HSCT in the real-world setting. Methods: A retrospective observational study was performed in adult pts with relapsed or refractory LBCL in the United States using electronic health record data from the Flatiron Health database from 01/01/2011 to 09/30/2021. The index date was defined as the 2L therapy start date. Pts were followed from index date until the end of study or death, whichever occurred first. Eligibility criteria included a diagnosis of LBCL, 1L treatment with an anti-CD20 therapy and anthracycline, 2L therapy with or without HDCT/HSCT, and ≥ 12 mo of follow-up. Pt demographics, clinical characteristics, treatment patterns, and overall survival (OS) were examined. Survival probabilities were estimated by Kaplan-Meier (KM) methods. Risk of death between groups was assessed by Cox proportional hazards regression models. Results: Out of 881 pts who met the eligibility criteria, 142 (16.1%) received HDCT/HSCT and 739 (83.9%) did not. Median follow-up time was 17.3 mo. Of all pts, median (range) age was 67.0 yr (19–85), 58.6% were male, 61.9% had stage III or IV disease, and 84.7% received treatment in a community setting. Of pts with/without HDCT/HSCT, median (range) age was 60.0 yr (19–77)/69.0 yr (23–85), 65.5%/57.2% were male, and 69.8%/60.3% had stage III or IV disease, respectively. Among pts with HDCT/HSCT, the most common 2L treatments were rituximab plus ifosfamide, carboplatin, and etoposide (R-ICE)-based (72.5%) and rituximab plus gemcitabine, dexamethasone, and cisplatin (6.3%); the most common third-line (3L) treatments were bendamustine plus rituximab (BR; 12.2%) and lenalidomide plus rituximab (12.2%). Among pts without HDCT/HSCT, the most common 2L treatments were R-ICE (25.3%) and BR (12.2%); the most common 3L treatments were rituximab plus gemcitabine and oxaliplatin (R-GemOX, 11.9%) and R-ICE (8.2%). Pts who received R-ICE might have been intended to proceed to HDCT/HSCT but did not due to lack of response to R-ICE. The median (95% confidence interval [CI]) KM estimates of OS were 22.1 mo (17.9–29.0), 100.0 mo (71.3–not reached), and 15.4 mo (12.9–19.7) in all pts and pts with or without HDCT/HSCT, respectively (P < 0.0001); OS probabilities at 12/24 mo were 60.2%/48.4%, 85.9%/77.3%, and 55.2%/42.8%. Of the 881 pts, 645 (73.2%) and 236 (26.8%) relapsed in ≤ 12 mo and > 12 mo from 1L, respectively. The median (95% CI) KM estimate of OS was 14.3 mo (12.1–18.7) and 47.8 mo (35.8–not reached) in pts who relapsed in ≤ 12 mo versus > 12 mo from 1L, respectively (P < 0.0001; Figure); OS probabilities at 12/24 mo were 54.0%/42.4% and 77.1%/64.9%. The risk of death was significantly greater among pts who relapsed in ≤ 12 mo versus > 12 mo from 1L therapy (hazard ratio, 1.81; 95% CI, 1.45–2.26; P < 0.0001). Image:Summary/Conclusion: Pts with LBCL who did not receive HDCT/HSCT in 2L (including those who were intended but failed to proceed to HSCT) or relapsed in ≤ 12 mo of 1L therapy initiation have a poor prognosis. Significant unmet need exists in pts who were unable to derive the benefit from HDCT/HSCT.