Racial and Ethnic Differences in Clonal Hematopoiesis, Tumor Markers, and Clinical Outcomes of Patients with Multiple Myeloma

Background: Multiple myeloma (MM) is twice as common in Blacks compared to Non-Hispanic (NH) Whites and Hispanics. While treatment and mortality differences have been reported for Black patients with MM compared to NH White patients, there is limited data on Hispanic populations. Furthermore, the factors driving observed differences in MM presentation and treatment responses by race and ethnicity are largely unknown. We investigated demographic, clinical, and molecular features, including tumor mutations and clonal hematopoiesis (CH), in a diverse population of patients with MM to elucidate mechanisms driving clinical disparities.

Methods: Patients diagnosed with MM who consented to our institutional biorepository protocol were eligible for inclusion. Demographic and clinical data were obtained from cancer registry and abstracted from electronic medical records. MM tumor cells were purified from bone marrow aspirates by CD138 affinity chromatography. DNA was isolated from tumor cells and whole blood for each patient, and whole exome sequencing (WES) data was generated. Tumor somatic mutations were characterized using paired tumor-normal (blood) WES. CH was classified based on blood-derived somatic mutations, using paired tumors and reference populations as germline comparators. Outcomes included overall survival (OS; date of diagnosis to death/last contact) and progression-free survival (PFS; 1 ^st-line treatment start to 1 ^st disease progression/death).

Results: A diverse group of MM patients (n=496) were included: NH White (80%), NH Black (10%) and Hispanic (9%). NH Black and Hispanic MM patients had a younger median age at diagnosis (57 and 53 yrs, respectively) compared to NH Whites (63 yrs, p = 0.0001; Fig A). There was no statistical difference in treatment categories received by race/ethnicity. NH Black patients had a longer time to hematopoietic cell transplant (HCT; 376 days) than NH White or Hispanic patients (248 and 270 days, respectively, p = 0.011). There was an improvement in OS for NH Black (HR 0.49, 95% CI 0.30-0.81) and Hispanic (HR 0.66, 95% CI 0.37-1.18) patients compared to NH White patients, but the association was not statistically significant in Hispanics. In univariable analysis, OS was also associated with age at diagnosis, International Staging System (ISS), treatment with HCT, and treatment regimen category. In multivariable analysis, after adjusting for age, ISS, HCT, and treatment category there was no longer a statistically significant association between OS and race/ethnicity. Although a worse PFS was present among Hispanic patients (adjusted HR 1.45, 95% CI 0.99-2.13), there was no statistically significant difference in PFS by race/ethnicity.

The most mutated genes in MM tumors were KRAS (24%), NRAS (17%), TP53 (11%), DIS3 (9%), and BRAF (9%) (Fig B). Genes with significantly higher tumor mutation rates in Black compared to NH White patients were SP140 (12% v 4%, p = 0.026), AUTS2 (8% v 2%, p = 0.04), and SETD2 (6% v 1%, p = 0.037). IRF4 was most commonly mutated in Hispanics (11% v 3% in NH White and 0% in Black, p = 0.019). We identified CH using WES in 60 (12%) patients. The most CH mutations were in ASXL1, DNMT3A, and TET2. There was no difference in the prevalence of CH by race/ethnicity (p=0.8). There was a statistically significant difference in OS by race/ethnicity and CH status (Fig C). For NH Black patients, CH (HR 4.36, 95% CI 1.36-14.0) and age at diagnosis (HR 1.08, 95% CI 1.03-1.14) were associated with inferior OS (Fig C). After adjusting for age in multivariable analysis, the positive association with CH status among Black patients was no longer statistically significant (HR 2.72, 95% CI 0.48-15.4). A positive, but not statistically significant, association for PFS in NH White patients with CH was also noted (adjusted HR 1.38, 95% CI 0.95-2.0).

Conclusions: This is the first study to examine differences in tumor mutation profiles, CH, and treatment among different racial and ethnic groups of patients diagnosed with MM. Our data suggest that age at diagnosis, tumor mutations, and CH may all contribute to clinical disparities observed in patients with MM. Efforts to expand our cohort and incorporate additional molecular biomarkers, epidemiologic characteristics, and clinical parameters are ongoing with the ultimate goal of elucidating targetable biological mechanisms to personalize management and optimize outcomes for diverse patients diagnosed with MM.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal