| There is a well-described, race-based outcome disparity in patients diagnosed with the most common form of non-Hodgkin lymphoma, diffuse large B-cell lymphoma. Despite significant speculation regarding the sources of this disparity, few firm conclusions have been drawn. Using data from the United States Veterans Health Administration, a cohort was constructed of 3,227 patients with a new diagnosis of diffuse large B-cell lymphoma made between October 1, 1998 and September 30, 2008. Consistent with previous studies, black patients had poorer overall survival, were younger, and more likely to have advanced stage disease compared to white patients. New findings from this study were a higher prevalence of HIV in black patients compared to white patients (21% versus 4%, p <.0001), and higher frequency elevated lactate dehydrogenase (64% versus 54%, respectively, p=.003), a serum tumor marker. Cox analysis controlling for differences in baseline characteristics noted between the black and white patients demonstrated that patients who were HIV positive had a markedly increased risk of death (hazard ratio=1.67, 95% confidence interval 1.27--2.2). Treatment variables associated with a reduced risk of death included: doxorubicin (hazard ratio=.57, 95% confidence interval 0.49--0.67), rituximab (hazard ratio=.6, 95% confidence interval 0.52--0.69), and myeloid growth factors (hazard ratio=.74, 95% confidence interval 0.56--0.98). Logistic regression analyses demonstrated no race-based differences in the use of doxorubicin, rituximab, or myeloid growth factors. While univariate analyses suggested decreased rituximab use in black patients, multivariable analysis demonstrated that this was driven by the differences in HIV prevalence. Patients who were HIV positive were less likely to receive rituximab (odds ratio=.1, 95% confidence interval 0.06--0.18), consistent with standard of care practice during the study time frame. Taken together, these results demonstrate that a higher prevalence of HIV in black patients is an important factor in racial outcome disparities in diffuse large B-cell lymphoma. There was no evidence of systematic bias in the application of curative intent therapy after controlling for HIV status. Measures to prevent or control HIV may be the best way to reduce racial disparities in diffuse large B-cell lymphoma. |
