Abstract B064: Genetic similarity, Duffy-null genotype, and autologous transplant in multiple myeloma: Disparities in depth of response but not survival
TL;DR
African genetic similarity/Duffy-null genotype was associated with a lower likelihood of achieving deep response but not with PFS or OS, and the immortal-time-corrected ASCT effect was uniform across ancestry and Duffy groups.
Abstract
Patients of African ancestry have higher multiple myeloma (MM) prevalence and historically worse outcomes, often attributed to unequal access to care. Whether genetically inferred ancestry or the Duffy-null genotype (rs2814778-CC; common in individuals of African-genetic similarity and associated with lower baseline neutrophil counts) influences treatment response and/or survival is poorly defined. Previous work has suggested that Duffy status may negatively influence progression-free survival (PFS) in MM patients undergoing transplant1. We examined associations of genetic similarity and Duffy-null status with treatment response, PFS, overall survival (OS), and the benefit of autologous stem-cell transplant (ASCT). Single-institution MM cohort (n=279) with genetic ancestry assigned by principal-component analysis compared with 1000 Genomes reference populations: African n=115, European n=137, Asian n=22, Other/Admixed n=5. Duffy status was defined by rs2814778 (CC=Duffy-null; TT/TC=Duffy-positive). Regression and survival analyses used the three major ancestry groups (n=274; reference=African). Depth-of-response endpoints (≥VGPR and CR/sCR) used logistic regression; PFS and OS used Cox models. Because ASCT timing introduces immortal-time bias, ASCT was modeled as a time-dependent covariate, and ancestry/Duffy effect modification was tested by likelihood-ratio test. Median follow-up was 94 months (IQR 50-138) in patients seen from 1993-2018. Duffy-null genotype (92/279, 33%; 91 in African-genetic similarity patients) was associated with lower odds of achieving ≥VGPR (unadjusted OR 0.58, 95% CI 0.34–0.97, p=0.037; adjusted for ancestry, age, treatment era, and ASCT: OR 0.34, 0.11–0.96, p=0.041). ASCT was the strongest predictor of ≥VGPR (adjusted OR 4.04, 2.14–7.86, p<0.001). Genetic similarity was not independently associated with ≥VGPR (p=0.93), CR/sCR (p=0.90), PFS (p=0.77), or OS (p=0.57); Duffy-null was not associated with PFS (HR 1.06, 0.77–1.45) or OS (HR 0.98, 0.69–1.40). In time-dependent models, ASCT conferred no PFS (HR 1.17, 0.85–1.60, p=0.33) or OS (HR 1.25, 0.88–1.78, p=0.21) benefit, and its effect did not differ by genetic similarity (interaction p=0.92 PFS, 0.68 OS) or Duffy (p=0.56 PFS, 0.74 OS). Because Duffy-null was ∼99% restricted to African ancestry, ancestry and Duffy were highly collinear. In this diverse MM cohort, across treatment eras, African genetic similarity/Duffy-null genotype was associated with a lower likelihood of achieving deep response (≥VGPR) but not with PFS or OS, and the immortal-time-corrected ASCT effect was uniform across ancestry and Duffy groups. Observed differences were in depth of initial response rather than survival. Strong collinearity between Duffy-null and African genetic similarity limits causal attribution; larger multi-institutional cohorts are needed to disentangle ancestral genetic effects from access- and treatment-related factors. Ayo S. Falade, Hongwei Tang, Michael Howe, Mahek Bai, Suganti Shivaram, Stacey Lehman, Huihuang Yan, Yan Asmann, Shaji Kumar, Vincent Rajkumar, Rahma Warsame, Megan Weivoda, Sikander Ailawahdhi, Linda Baughn, Joselle M. Cook. Genetic similarity, Duffy-null genotype, and autologous transplant in multiple myeloma: Disparities in depth of response but not survival [abstract]. In: Proceedings of the 19th AACR Conference on the Science of Cancer Health Disparities; 2026 Oct 18-21; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2026;35(10 Suppl):Abstract nr B064.