Trends in Real-World Characteristics and Outcomes of Patients with Diffuse Large B-Cell Lymphoma Treated in the US Community Oncology Setting Pre- and Post-CAR-T Approvals

Background: Diffuse large B-cell lymphoma (DLBCL) is the most common of the aggressive non-Hodgkin’s lymphomas. However, the entry of CAR-T therapy in October 2017 marked a significant shift by introducing a potentially curative therapy for relapsed/refractory (R/R) DLBCL. The purpose of this study was to describe the real-world trends in characteristics and outcomes of DLBCL patients in the eras pre- and post-CAR-T approval in a US community oncology setting.

Methods: This retrospective observational study used electronic health records data from within The US Oncology Network and selected non-Network practices, covering a nationally representative network of over 2,700 community-based oncology providers seeing more than 1 million patients annually. All adult patients with an initial diagnosis of or first visit for DLBCL from 10/1/2014 to 10/1/2023 were included. This was considered their index date. Patients were grouped into a pre- and post-CAR-T era based on index on or after 10/1/2017, respectively, with three years of data capture in each cohort. Patient characteristics as well as the proportion of patients surviving 1 and 2 years following index were descriptively summarized overall, by index year, and by pre- vs. post-CAR-T era. No statistical testing was done to compare the two cohorts.

Results: A total of 15,910 patients with DLBCL were identified. One-third (32%) of patients were observed in the pre-CAR-T era, compared to two-thirds in the post-CAR-T era (68%). Mean (standard deviation [SD]) age at diagnosis was 67 (14) years. More than half of patients were male (55%); most patients were White (70%). The most common documented stage at diagnosis was Stage IV (40%), followed by Stage III (24%) and Stage II (19%). Information on germinal center B-cell–like (GCB) vs. non-GCB subtypes was not available. The mean (standard deviation [SD]) age at diagnosis remained stable over time (66 [14] years in the pre-CART-T era vs. 67 [14] years in the post-CAR-T era), and there was no difference in the proportion of males vs. females over time. Among patients with documented race, there was a decrease in the proportion of White patients (89% in the pre-cohort vs. 84% in the post-cohort), while the proportion of Black patients and patients with other races increased slightly (5% vs. 6% and 6% vs. 10%, respectively). The proportion of patients diagnosed with documented advanced disease (Stage III or IV) remained stable over time, as did the geographic distribution of patients (38% West, 32% South, 24% Midwest and 6% Northeast). The proportion of patients who died within 1 year of diagnosis did not differ over time (14% and 15%, pre and post, respectively) nor did the proportion of patients who died within 2 years of diagnosis (20% in each). Year-level analyses were roughly in line with era-level analyses.

Conclusions: In this retrospective descriptive study of DLBCL patients in the US community setting in the pre and post CAR-T eras, there were no substantial changes in patient characteristics, except an increase in non-White races, which likely reflect underlying population demographic changes. There were more patients in the post CAR-T era, which likely reflects growth of participating practices more than an increased incidence of new patients. We did not see any difference in shorter term 1- and 2-year survival outcomes pre- and post-CAR-T introduction. This may reflect limited use of CAR-T in this real-world community setting up to this point, with greater heterogeneity regarding eligibility and challenges in access. Further insight will require updated time analyses, longer patient follow-up and more detailed data on CAR-T patients in the spectrum of evolving treatment options in DLBCL, which now also includes bispecific antibody therapies.