Abstract Intro Angioimmunoblastic T cell lymphoma is a unique subtype of peripheral T cell lymphoma (PTCL), often associated with poor prognosis. It accounts for ∼2% of all non-Hodgkin lymphomas and ∼15% to 20% of PTCL (Blood, PMID: 34292324). Given its rarity, there is minimal data that analyzes the difference in demographics, treatment patterns, and outcomes of angioimmunoblastic T cell lymphoma among different types of cancer centers. In this study we outline the key differentiating factors seen between Academic Cancer Programs (ACPs) and Community Cancer Programs (CCPs). Methods We performed a retrospective analysis of AITL cases diagnosed in the United States between the years 2004-2022 using the National Cancer Database (NCDB) between the years of 2004-2022. Patient's demographics, clinical characteristics and outcomes were compared between ACPs and CCPs. ACPs included academic and research programs, including NCI-designated comprehensive cancer centers. CCPs comprised community, comprehensive community, and integrated network cancer programs. Kaplan-Meier and Cox proportional hazards models were used to compare overall survival (OS), adjusting for age, race/ethnicity, insurance status, comorbidity score (Charlson-Deyo), and distance from treating facility. Results A total of 6,663 patients were identified, with 3,967 (60%) treated at ACPs and 38% treated at CCPs. CCP-treated patients were older, with a median age 71 vs. 68 years, and a greater proportion of patients age 75 or older (40% vs. 29%, p0.001). ACPs cared for more patients under 60 years of age (25 vs. 20%). More patients at ACPs were diagnosed at stage III or IV vs CCPs (83% vs 78.4%). Race/ethnicity distribution showed that patients affected by AITL were mostly white (82% in ACPs and 87% in CCPs). ACPs saw slightly more Black patients compared to CCPs (10% vs. 7% respectively). There were no significant differences between income and education levels between the two facility types. Patients treated at ACPs lived further from treatment facility compared to patients being treated at CCPs (12 miles vs. 8 miles). Insurance status demonstrated significant differences, with CCPs seeing more Medicare patients (62% vs. 53%), whereas ACPs having more private insurance rates (34% vs. 29%) (p 0.001. Additionally, ACPs saw slightly more Medicaid-covered patients compared to CCPs (7% vs. 4%, respectively). There were significant differences in treatment patterns, with ACPs being more likely to provide treatment compared to CCPs (72% vs. 62%, p0.001). The median time-to-treatment initiation was 22 days for both facility types. Survival analysis revealed an overall survival at 2, 5, and 10 years consistently favored ACPs, with the most pronounced differences observed in the earlier years. At 2 years, survival was 51% for ACPs versus 46% for CCPs; at 5 years, 36% versus 33%; and at 10 years, 24% versus 22%, respectively. In a multivariable Cox model—adjusting for age, race/ethnicity, insurance status, great-circle distance to care, and Charlson–Deyo comorbidity score—receiving care at an ACP remained independently associated with improved OS (2.14 years vs. 1.52 years, p=0.005). Conclusion In this large, retrospective analysis of 6,663 patients diagnosed with AITL across the United States, treatment at academic center programs (ACPs) was associated with improved survival compared to community cancer programs (CCPs), even after adjusting for key demographic and clinical factors. Patients treated at ACPs were younger, more likely to present with advanced stage disease, and lived further away from their treatment facility. Despite comparable time-to-treatment initiation, ACPs demonstrated higher rates of treatment initiation. Moreover, ACPs led to higher 2-, 5-, and 10-year survival rates, with the most notable differences seen in early survival. These findings suggest that early referral to academic centers may confer a survival advantage in this rare lymphoma, potentially reflecting key differences in access to specialized care, clinical trial availability and supportive infrastructure.
Mendiola et al. (Mon,) studied this question.