The impact of Duffy genotype on progression-free survival (PFS) with lenalidomide, Bortezomib, and dexamethasone (RVd) alone or RVd plus autologous stem cell transplantation (ASCT) and continuous R maintenance in patients (pts) with newly diagnosed multiple myeloma (NDMM): Updated subgroup analysis of the phase 3 DETERMINATION trial

Abstract Background: A single nucleotide polymorphism (SNP) in ACKR1/DARC results in erythrocyte Duffy null phenotype in ~66% of African American (AA) and <1% of White pts. This is associated with lower absolute neutrophil count (ANC) and has a key role in cytokine homeostasis, which may influence MM pathobiology, response to inflammatory stressors, and treatment (tx) outcomes. DETERMINATION found a PFS benefit with RVd+ASCT (RVd-alone vs RVd+ASCT: hazard ratio [HR] 1.53; 95% confidence interval (CI) 1.23–1.91), but subgroup analysis showed differential PFS effect by race (AA pts: HR 1.07, 95% CI 0.61–1.89; White pts: HR 1.67, 95% CI 1.29–2.15). As race is a social construct, and with equitable access to care and cost-free therapy provision in DETERMINATION, we sought a pathobiological explanation. Given the prevalence of Duffy null and its known impact on the inflammasome, we hypothesized that Duffy status may affect magnitude of PFS benefit. We report updated PFS analyses in all pts in DETERMINATION with Duffy status available. Methods: Pts received RVd-alone or RVd+ASCT followed by R maintenance until progression in both groups. Peripheral blood samples underwent genomics analysis for the SNP rs2814778 and were classified as C/C (Duffy null) or non-C/C (Duffy non-null). Impact of Duffy status on PFS was evaluated with Cox proportional hazards regression in univariate models. Heterogeneity of tx effect was assessed by a test for interaction. Results: Overall, 592 enrolled pts were evaluable for Duffy status; consistent with US population data, 63.3% (n=62/98) of AA pts and 1.1% (n=5/465) of White/other pts were Duffy null (n=4/29 pts with missing race). 493 randomized pts had Duffy status evaluated (68.3% of intent-to-treat [ITT] population, N=722), with 238 assigned to RVd-alone and 255 to RVd+ASCT. Pts were broadly representative of the ITT population. PFS with RVd-alone vs RVd+ASCT in the analysis cohort, overall and by race, was consistent with the ITT population. Of the 493 pts, 59 (12.0%) pts were Duffy null (53 [89.8%] AA, 5 [8.5%] white/other race, 1 missing), and 434 (88.0%) were Duffy non-null (403 [92.9%] white/other race, 28 [6.5%] AA, 3 missing). Median (interquartile range) baseline ANC was 2.8 (2.1–4.3) x 109/L in Duffy null pts vs 3.4 (2.6–4.6) x 109/L in Duffy non-null pts. In Duffy null vs Duffy non-null pts, median duration of tx from randomization (35.1 vs 33.8 months [mos]) and of R maintenance (42.1 vs 36.4 mos) were similar in the RVd-alone arm but numerically shorter (30.3 vs 40.4 mos; 32.5 vs 42.4 mos) in the RVd+ASCT arm. Rates of grade ≥3 neutropenia in the RVd-alone vs RVd+ASCT arms (for all tx) were 48.3% vs 80.0% in Duffy null pts and 40.7% vs 88.9% in Duffy non-null pts; rates of grade ≥3 febrile neutropenia (FN) were 0% vs 3.3% in Duffy null pts and 7.2% vs 9.3% in Duffy non-null pts. Overall pooled PFS was similar in Duffy null vs Duffy non-null pts (median 62.5 vs 56.7 mos; HR 0.94, 95% CI 0.62–1.42). Duffy non-null pts had PFS findings consistent with the ITT analysis (RVd-alone vs RVd+ASCT: 120/209 vs 85/225 events/pts; median 46.7 vs 67.5 mos; HR 1.76, 95% CI 1.33–2.34). In contrast, Duffy null pts had longer PFS with RVd-alone vs RVd+ASCT (9/29 vs 16/30 events/pts; median NR vs 44.0 mos; HR 0.64, 95% CI 0.27–1.50) (interaction p-value 0.005). When these analyses were restricted to AA pts, the same PFS pattern was seen with RVd-alone vs RVd+ASCT among Duffy null (8/25 vs 15/28 events/pts; median NR vs 45.4 mos; HR 0.66, 95% CI 0.27–1.60) compared to Duffy non-null pts (6/13 vs 3/15 events/pts; median 64.4 mos vs NR; HR 5.29, 95% CI 1.20–23.4). On univariate analysis by Duffy status (null vs non-null), PFS HR was 0.51 (95% CI 0.26–1.00) with RVd-alone and 1.63 (95% CI 0.95–2.78) with RVd+ASCT. Conclusions: These exploratory analyses of DETERMINATION suggest that Duffy status drives a difference in tx effect that is more pronounced than for PFS by race. With RVd-alone vs RVd+ASCT, PFS appeared better in Duffy null pts and poorer in Duffy non-null pts. In Duffy null vs non-null pts, PFS appeared better with RVd-alone and poorer with RVd+ASCT. Similar trends were seen in analyses restricted to AA pts, indicating Duffy status may provide a biological rationale for observed differential tx effects, rather than race. Further studies are warranted to assess impact of Duffy status on clinical outcomes such as treatment response, overall survival, and effects on the inflammasome, as well as ANC and FN.