Black/White Differences in the Treatment and Outcomes of Diffuse Large B Cell Lymphoma: A Matched Cohort Analysis.

While African-Americans have a lower incidence of diffuse large B-cell lymphoma (DLBCL) than Whites (W) in the United States, there appear to be racial differences in clinical features, the use of chemoimmunotherapy(Flowers, AACR 2008), and treatment outcomes. However, previous analyses have been hampered by limited numbers of Black (B) patients (pts); lack of direct clinical information on pathology, prognostic factors, treatment, and outcomes; limited follow-up; or all three.

To examine B/W differences in the use of CHOP-based chemotherapy vs. rituximab CHOP (RCHOP), we performed a retrospective, matched cohort analysis comparing the impact of race, on presenting features, treatment, and outcomes in a clinical setting with detailed data ascertainment from pathology, clinical, and pharmacy records. Patients diagnosed with DLBCL from 1981 to 2009 were identified from Emory pathology and medical records using previously published methods (Graiser, Cancer Informatics 2007). Baseline demographic data, components of the International Prognostic Index (IPI: age, performance status, lactate dehydrogenase [LDH], number of extranodal sites involved, stage), insurance status, employment status, treatment and survival data were extracted. Differences by race in baseline features and treatment category (CHOP vs. RCHOP) were analyzed using Chi-squared statistical analysis. Univariate and multivariate logistic regression analyses were performed for the entire population and cohorts matched by IPI, age, and year of diagnosis, to determine predictors of RCHOP use, 2-year and 5-year overall survival (OS). Cox regression was used to analyze the predictors of OS.

A total of 531 (348 W and 107 B) pts with a confirmed diagnosis of DLBCL were identified. The median age of diagnosis for B was significantly less than that for W pts (median age 46 (range: 18–87) vs. 56 (range: 18–86), respectively; p<0.001). There were no differences in stage, performance status or extranodal disease, but B pts more frequently presented with elevated LDH (50% vs. 37%, p=0.04). There were no racial differences in RCHOP use. Pts who had elevated LDH were more likely to receive RCHOP (p<0.001). W race was a predictor of improved 2yr OS. W race, early stage, IPI <2 were predictors of improved 5yr OS. However, there were no B/W differences in OS, in the CHOP cohort or the RCHOP cohort. After matching for revised IPI group, age at diagnosis, and diagnosis year B (n=104) and W (n=104) pts had similar rates of RCHOP use (41% vs. 46%), 2-yr OS, and 5-yr OS.

These data corroborate findings that B pts present with DLBCL at a younger age. In the setting of a single institution referral center with detailed ascertainment of treatment, there do not appear to be racial differences in RCHOP use or outcomes. Since IPI, age, and year of treatment may influence treatment selection and outcomes, matching cohorts on these factors is necessary when examining B/W differences.

Flowers: Amos Medical Faculty Development Program grant from the American Society of Hematology/Robert Wood Johnson Foundation: Research Funding.