Population-specific prognostic models are needed to stratify outcomes for African-Americans with diffuse large B-cell lymphoma

Qiushi Chen, Turgay Ayer, Loretta J. Nastoupil, Jean L. Koff, Ashley D. Staton, Jagpreet Chhatwal, Christopher R. Flowers

Research output: Contribution to journalArticle

5 Scopus citations

Abstract

Diffuse large B-cell lymphoma (DLBCL) demonstrates significant racial differences in age of onset, stage, and survival. To examine whether population-specific models improve prediction of outcomes for African-American (AA) patients with DLBCL, we utilized Surveillance, Epidemiology, and End Results data and compared stratification by the international prognostic index (IPI) in general and AA populations. We also constructed and compared prognostic models for general and AA populations using multivariable logistic regression (LR) and artificial neural network approaches. While the IPI adequately stratified outcomes for the general population, it failed to separate AA DLBCL patients into distinct risk groups. Our AA LR model identified age ≥ 55 (odds ratio 0.45, [95% CI: 0.36, 0.56], male sex (0.75, [0.60, 0.93]), and stage III/IV disease (0.43, [0.34, 0.54]) as adverse predictors of 5-year survival for AA patients. In addition, general-population prognostic models were poorly calibrated for AAs with DLBCL, indicating a need for validated AA-specific prognostic models.

Original languageEnglish (US)
Pages (from-to)842-851
Number of pages10
JournalLeukemia and Lymphoma
Volume57
Issue number4
DOIs
Publication statusPublished - Apr 2 2016

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Hematology
  • Oncology
  • Cancer Research

Cite this