e19063 Background: Primary cutaneous B-cell lymphomas (PCBCL), primary cutaneous diffuse large B-cell lymphoma (PCDLBCL), follicle center lymphoma (PCFCL), and marginal zone lymphoma (PCMZL)—demonstrate distinct prognoses, yet population-level data on social determinants of health (SDOH) remain limited. This study evaluated the influence of demographic, socioeconomic, and geographic factors on survival across subtypes. Methods: We conducted a retrospective cohort study using the SEER-17 database to identify patients with PCDLBCL (ICD-O-3 9680/3), PCFCL (9597/3), and PCMZL (9699/3) restricted to primary skin sites. Variables included age, sex, race/ethnicity, median household income, and geography. Overall survival (OS) and cancer-specific survival (CSS) were evaluated using Kaplan-Meier methods and Cox regression (univariate and multivariate), with patients diagnosed through 2017 for ≥5-year follow-up. Analyses were performed in Python (v3.12) using lifelines and pandas, reporting hazard ratios (HRs) with 95% confidence intervals (CIs); p <0.05 was considered statistically significant. Results: A total of 4,525 patients were identified: PCDLBCL (n=1,744), PCFCL (n=751), and PCMZL (n=2,030). Age was the strongest predictor of survival across all subtypes; in PCDLBCL, patients aged ≥80 years had worse OS (univariate: median 24 vs. 65 months, HR 3.98; multivariate: HR 7.71; p0.001) and CSS (univariate: HR 2.26; multivariate: HR 4.57; p0.001) compared to patients aged 45-64 years, while in PCFCL, age was the only significant predictor (≥80 years: multivariate OS HR 20.60, CSS HR 12.57; p0.001). Male sex was independently associated with worse OS in PCDLBCL (HR 1.13, p=0.046) and PCMZL (HR 1.23, p=0.036). Racial disparities were significant only in PCDLBCL, with Asian patients demonstrating worse OS (univariate: median 40 vs. 65 months; multivariate: HR 1.52, p=0.007) and CSS (HR 1.34, p=0.001) compared to White patients, and American Indian patients showing worse OS (HR 2.49, p=0.032). Higher income (Q3 vs. Q1) was protective for OS in PCDLBCL (univariate: median 88 vs. 50 months; multivariate: HR 0.73, p=0.008) and for both OS (HR 0.62, p=0.028) and CSS (HR 0.38, p=0.021) in PCMZL. Geographic region significantly influenced OS in PCDLBCL (p=0.029) and PCMZL (p=0.001). No significant SDOH-related differences were observed in PCFCL. Conclusions: This population-based study demonstrates that SDOH differentially impact PCBCL subtypes. Age is the strongest independent predictor across all subtypes. PCDLBCL demonstrates the most pronounced disparities, with independent effects of race, income, sex, and geography. PCMZL shows socioeconomic and sex-based influences. PCFCL appears resistant to SDOH. These findings underscore the need for targeted interventions to address modifiable barriers in aggressive PCBCL subtypes.
Yang et al. (Thu,) studied this question.