Racial disparities in multiple myeloma: biological heterogeneity, treatment access, and prognostic implications

Abstract

Unlike the well-recognized two-fold higher incidence rate and an earlier age of onset of MM among African Americans (AA) compared to European Americans (EA) in the United States, findings on racial disparities in survival outcomes of MM are inconclusive, ranging from worse survival, to no difference, to better overall survival in AA patients than EA patients. The clinical outcomes for patients with MM depend on a complex interplay of factors including age, disease severity, cytogenetics, treatment modalities, biological features of the responsible clone, and non-biological factors such as treatment utilization and access to care. This review focuses on genetic and molecular heterogeneity of the disease biology and characteristics among MM patients from different populations. The review discusses possible determinants of racial and ethnic disparities in the outcomes of MM and considers potential strategies to address them with an ultimate goal of decreasing MM health disparities.

Introduction

Multiple myeloma (MM), the second most common hematologic malignancy after non-Hodgkin lymphoma [1], is a neoplastic plasma-cell disorder that is characterized by proliferation of abnormal bone marrow plasma cells, monoclonal protein in the blood or urine, and evidence of associated end-organ damage [2,3]. Globally, Northern America and Australia/New Zealand have the highest incidence rates of MM (>4 per 100,000 for both sexes combined) [4]. In the US, the incidence of MM is about two-fold higher in African Americans (AA) (16.1 per 100,000 in 2022) than Hispanics (7.5 per 100,000) and European Americans (EA) (6.9 per 100,000) with Asians having the lowest incidence (4.2 per 100,000) [1]. Despite effective contemporary multidrug combinations, especially when used with autologous stem cell transplant (ASCT), that have improved the 5-year relative survival rate to 62.4% [1,5], MM remains largely incurable for the majority of patients as they eventually relapse after first line therapy [6].

MM is a group of biologically complex and heterogeneous disease, reflected by its wide panel of genetic alterations, clinical presentation, and the individual differences in overall response and survival of patients receiving the same treatment [2,7]. Unlike the well-recognized two-fold higher incidence rate and an earlier age of onset of MM among AA than EA in the US [1], findings on racial/ethnic disparities in survival of MM are inconclusive, ranging from no difference [8–10] to better overall survival [11,12] in AA patients than EA patients, even with later access to novel therapies [13]. There are also data showing that AA patients with MM continued to have poor overall survival (OS) and progression-free survival (PFS) compared to EA patients even after accounting for ASCT and ASCT plus novel immunomodulatory therapies [14].

The clinical outcome for patients with MM depends on a complex interplay of variables including age, disease severity, cytogenetics, treatment modalities, the biological features of the responsible clone [15], and non-biological factors such as treatment utilization and access to care. This review discusses the molecular and genetic heterogeneity of the disease biology (Table 1) and characteristics among patients (Table 2) with MM from different racial groups with the aim of improving understanding of MM health disparities.

Table 1.

Association between biological prognostic factors and survival outcomes by populations.

Study	Data source	Study population	Overall survival, HR (95% CI)				progression-free survival, HR (95% CI)				Note
			EA	AA	H	API	EA	AA	H	A
Cytogenetics
Gasparetto [32]	Connect MM Registry	EA = 2400 AA = 211	t(11;14) + 1.0 (ref) t(11;14) − 1.16 (0.931.44)	t(11;14)+ 1.0 (ref) t(11;14) − 0.55 (0.320.94)	N.A.	N.A.	(11;14) ‘−’ vs ‘+’: 1.03 (0.871.22)	(11;14) ‘−’ vs ‘+’: 0.77 (0.521.15)	N.A.	N.A.	NDMM, voluntary registration
Badar [104]	CIBMTR	EA = 2453 AA = 1085	t(11;14) + 1.90 (1.073.38) t(11;14) − 1.55 (0.912.64)	t(11;14) + 1.0 (ref) t(11;14) − 1.35 (0.782.33)	N.A.	N.A.	t(11;14) + 1.31 (0.911.89) t(11;14) − 1.01 (0.731.39)	t(11;14) + 1.0 (ref) t(11;14) − 1.05 (0.761.46)	N.A.	N.A.	ASCT recipients registry
Derman [14]	MMRF-CoMMpass	EA = 526 AA = 113	0 HRCA 1.0 (ref) 1 HRCA 1.9 (1.2–3.0) 2+ HRCA 3.9 (2.3–6.6)	0 HRCA 1.0 (ref) 1 HRCA 0.8 (0.4–1.7) 2+ HRCA 1.4 (0.5–3.7)	N.A.	N.A.	0 HRCA 1.0 (ref) 1 HRCA 1.3 (1.0–1.7) 2+ HRCA 2.2 (1.5–3.2)	0 HRCA 1.0 (ref) 1 HRCA 1.1 (0.6–2.0) 2+ HRCA 1.5 (0.7–3.3)	N.A.	N.A.	NDMM
Wang [30]	Single institute	EA = 126 AA = 55	0 HRCA 1.0 (ref) HRCA 2.6 (1.3–5.5)	0 HRCA 1.0 (ref) HRCA 0.6 (0.1–2.9)	N.A.	N.A.	N.A.	N.A.	N.A.	N.A.	NDMM
Gene expression
Derman [14]	MMRF-CoMMpass	EA = 526 AA = 113	UAMS70 Standard risk 1.0 (ref) High risk 2.0 (1.2–3.4)	uAMS70 Standard risk 1.0 (ref) High risk 2.1 (0.7–6.2)	N.A.	N.A.	uAMS70 Standard risk 1.0 (ref) High risk 1.6 (1.1–2.4)	uAMS70 Standard risk 1.0 (ref) High risk 1.6 (0.7–3.9)	N.A.	N.A.	NDMM

EA: European Americans, AA: African Americans, H: Hispanics, API: Asian pacific islanders.

HR: Hazard ratio, Ci: confidence interval.

HRCA: high risk cytogenetic abnormalities, +: positive, −: negative, N.A.: Not available, ref: referent.

CIBMTR: Center for international Blood and Marrow Transplant Research, MMRF-CoMMpass: Multiple Myeloma Research Foundation (MMRF) CoMMpass Study.

NDMM: Newly diagnosed multiple myeloma, ASCT: autologous stem cell transplantation.

Table 2.

Association between non-biological factors and survival outcomes by populations.

Study	Data source	Study population	Overall survival, HR (95% CI)				Progression-free survival, HR (95% CI)				Note
			EA	AA	H	A	EA	AA	H	A
Treatment Utilization – novel agents
Ailawadhi [64]	SEER-Medicare	EA = 3574 AA = 945 H = 531 API = 288	1.00 (ref)	1.11 (1.01–1.22)	1.08 (0.97–1.22)	1.06 (0.91–1.23)	N.A.	N.A.	N.A.	N.A.
Ailawadhi [10]	MM Cooperative Group trials	EA = 2373 AA = 392 H = 76 Non-H = 55	1.00 (0.75–1.33)	0.95 (0.70–1.29)	1.00 (ref)	1.13 (0.73–1.75)	1.20 (0.92–1.55)	1.14 (0.86–1.51)	1.0 (ref)	1.32 (0.88–1.96)	Phase 3 trials = 7 Phase 2 trials = 2
Pulte [8]	Meta-analysis of 5 clinical trials	EA = 2400 API = 167 Other = 250	1.0 (ref)	N.A.	N.A.	1.03 (0.81–1.32)	N.A.	N.A.	N.A.	N.A.	NDMM, not transplant eligible
Kaur [67]	MMCCR & Connect MM registry	MMCCR: EA = 169 AA = 489 H = 281 Connect: EA = 2327 AA = 404 H = 173	MMCCR: 1.0 (ref) Connect MM: 1.51 (1.15–2.00)	1.2. (0.57–2.4) 1.21 (0.88–1.66)	1.3 (0.97–1.01) 1.0 (ref)	N.A.	MMCCR: N.A. Connect MM: 1.34 (1.09–1.65)	N.A. 1.24 (0.98–1.57)	N.A. 1.0 (ref)	N.A.	NDMM, MMCCR: Montefiore Medical Center Cancer Registry
Dong [65]	Matched population-based, SEER-Medicare	AA = 3319 EA = 3319	Treatment matched: 1.17 (1.09–1.26) Demographics matched 1.09 (1.01–1.17) SES matched: 1.03 (0.96–1.11)	1.0 (ref)	N.A.	N.A.	N.A.	N.A.	N.A.	N.A.	3319 EA matched sequentially to 3319 AAs on demographics, SES, & treatment
Joseph [69]	single institute	EA = 800 AA = 499 API = 27	1.00 (ref)	0.84 (0.68–1.04)	N.A.	1.15 (0.57–2.32)	1.00 (ref)	0.97 (0.82–1.96)	N.A.	0.84 (0.43–1.63)	Rvd and D-RVd treated NDMM
Wang [105]	SEER-Medicare	EA = 11983 AA = 2094	Tx ≤1 yr dx 1.00 (ref) Tx >1 yr dx 1.00 (ref)	0.98 (0.09–1.06) 0.84 (0.77–0.93)	N.A.	N.A.	N.A.	N.A.	N.A.	N.A.	NDMM
Treatment Utilization – ASCT
Schriber [75]	CIBMTR and SEER	EA = 18046 AA = 4123 H = 1933	1.07 (0.97–1.18)	0.99 (0.89–1.11)	1.0 (ref)	N.A.	N.A.	N.A.	N.A.	N.A.	NDMM, ASCT recipient
Fiala [81]	SEER-Medicare	EA = 17574 AA = 3342	1.00 (ref)	Adjusting for demographics: 1.12 (1.05–1.19) Demographics & treatment 1.05 (0.99–1.12) Demographics, treatment, & accessibility 0.91 (0.85–0.97)	N.A.	N.A.	N.A.	N.A.	N.A.	N.A.
Derman [14]	CoMMpass	EA = 526 AA = 113	No ASCT 1.0 (ref) ASCT 0.7 (0.3–1.8)	No ASCT 1.0 (ref) ASCT 0.8 (0.2–3.6)	N.A.	N.A.	No ASCT 1.0 (ref) ASCT 0.7 (0.4–1.3)	No ASCT 1.0 (ref) ASCT 0.1 (0.02–0.5)	N.A.	N.A.
Ailawadhi [70]	Connect MM Registry	SCT: EA = 878 AA = 126 NSCT: EA = 1562 AA = 271	SCT 1.0 (ref) NSCT 1.0 (ref)	0.56 (0.35–0.89) 0.86 (0.70–1.06)	N.A. N.A.	N.A. N.A.	SCT 1.0 (ref) NSCT 1.0 (ref)	0.86 (0.65–1.15) 0.96 (0.82–1.14)	N.A. N.A.	N.A. N.A.	NDMM receiving first-line stem cell transplantation (SCT) or not receiving SCT (NSCT)
Treatment Utilization – CAR-T
Peres [66]	US MM immunotherapy consortium	EA = 149 AA = 36 H = 22	1.0 (ref)	1.13 (0.54–2.38)	1.39 (0.54–3.63)	N.A.	1.0 (ref)	1.33 (0.81–2.21)	1.45 (0.77–2.73)		Refractory/relapsed MM treated with CAR-T
Access to care – SES
Castaneda-Avila [103]	SEER-18	EA = 35314 AA = 10023 H = 6663 API = 3330	High SES: 1.0 (ref) Medium SES 1.15 (1.11–1.19) Low SES 1.17 (1.12–1.22)	High SES 1.0 (ref) Medium SES 1.08 (0.98–1.19) Low SES 1.14 (1.04–1.23)	High SES 1.0 (ref) Medium SES 1.07 (0.96–1.19) Low SES 1.09 (0.98–1.20)	High SES 1.0 (ref) Medium SES 1.18 (1.04–1.33) Low SES 1.35 (1.16–1.57)
Buradagunta [102]	SEER-Medicare	EA = 1591 AA = 1591	High SES 1.0 (ref) Low SES 1.0 (ref)	Not reported, p>0.1 1.18 (1.02–1.37)

EA: European Americans, AA: African Americans, H: Hispanics, API: Asian Pacific Islanders.

HR: Elazard ratio, CI: confidence interval, N.A.: Not available, ref: referent.

CIBMTR: Center for International Blood and Marrow Transplant Research, MMRF-CoMMpass: Multiple Myeloma Research Foundation (MMRF) CoMMpass Study.

NDMM: Newly diagnosed multiple myeloma, ASCT: autologous stem cell transplantation.

Biological determinants

Cytogenetic abnormalities

Cytogenetic abnormalities are a hallmark of MM and carry significant prognostic implications. MM can be classified into two distinct karyotypic groups: a hyperdiploid phenotype, characterized by trisomies of odd-numbered chromosomes that are present in 50–60% of tumors, and non-hyperdiploid, primarily characterized by translocations of 14q32, but also gains of 1q and monosomy 13 [16–19]. Hyperdiploidy and t(11;14) are associated with a favorable prognosis, whereas despite advances in therapeutic options, abnormalities such as t(4;14), t(14;16), t(14;20), deletion 17p, and gain or amplification of 1q are associated with inferior survival [18–21]. As a result, the presence of any of these abnormalities have been labeled as high-risk cytogenetic abnormalities (HRCA) [19,22,23]. These HRCAs have also been used to determine the intensity of frontline therapy. For example, patients with t(4;14) have more favorable outcomes when they are treated with proteasome inhibitors (PI) [24], whereas patients with the high-risk cytogenetics t(14;16), t(14;20), and/or del(17p) are better candidates for triplet combination therapy (e.g. bortezomib-lenalidomide-dexamethasone) compared with patients with intermediate-risk or standard-risk disease [19,25,26].

Early evidence suggests that racial disparities in prognosis may be due to different frequencies and number of HRCAs. Compared with EA patients, AA patients have a slightly lower prevalence of high-risk del 17/TP53 mutation [27] and are less likely to harbor t(4;14), t(11;14), monosomy 13, and monosomy 17 [28]. However, not all studies have found these differential frequencies to exist. For example, recent studies [14,29,30] found no difference in the frequency of HRCA between AA and EA patients. These conflicting findings arise partly because of the challenge to directly compare results across studies due to the heterogeneity in laboratory practices related to profiling cytogenetic abnormalities [31], including the heterogeneity of fluorescent in situ hybridization (FISH) probes and the lack of uniform CD138+ selection for FISH analysis which could lead to false negatives and under-reported cytogenetic abnormalities [14].

In addition to frequency differences, HRCAs may be associated with survival outcomes differently between populations. For example, t(11;14) [32] and t(4;14) [30] were associated with worse OS in AA but not EA, while gain/amplification of 1q was associated with inferior OS in EA but not AA [30]. These population-specific associations were not attenuated with additional adjustment for established clinical prognostic factors or concomitant presence of other HRCAs [30]. Additionally, Derman et al. [14] and Wang et al. [30] found that the International Myeloma Working Group (IMWG) definition of HRCAs were associated with worse OS only in EA patients with MM, and not in AA patients. These findings suggest that whereas conventional HRCA has been associated with inferior outcomes, this needs to be separately considered for AA patients to optimize risk prognostication [14].

Genomics

Aside from identifying high-risk cytogenetic abnormalities by FISH, advanced molecular diagnostics such as whole-genome sequencing [33] and gene expression profiling (GEP) [34–36] have also been applied to identify high risk MM and predict disease behavior. GEP classifiers such as EMC92 [35,36], UAMS70 [14,37], and SKY92 [38,39], identify subgroups of MM associated with adverse survival outcomes. Differences in gene expression and mutation may also be a factor behind the prognostic differences across populations. In an analysis of the Multiple Myeloma Research Foundation (MMRF) CoMMpass data, Manojlovic, et al. [40] found racial differences in the frequency of gene mutations with AA patients having a higher frequency of BCL7A, BRWD3, and AUTS2 mutations, and a lower frequency of TP53 and IRF4 mutations compared to EA patients. Using the same MMRF database, Derman et al. [14] found that high-risk gene expression profile by UAMS70 was associated with worse OS and PFS in EA patients with MM after controlling for stage, number of HRCAs, and triplet therapy with ASCT, but to a lesser extent, in AA patients with wide confidence interval. Another study by Kazandjian et al. [27] found a higher rate of BRAF^V600E mutation in AA (14%) compared with EA patients (2%) with MM. Albeit a rare driver mutation in MM, BRAF^V600E is an important druggable target as BRAF mutations has been found to occur in ~5% of MM patients with tumors that respond to tyrosine kinase inhibition [41]. Overall, these findings suggest that similar to HRCA, defining high risk MM by GEP needs to also separately consider AA patients or other racial groups to improve understanding unique disease biology and genes that may be associated with prognosis.

Epigenetics

Aberrant epigenetic regulation of gene expression plays an important role in the pathogenesis and progression of MM [7]. Greater epigenetic heterogeneity is linked with more aggressive subtypes, cytogenetic subgroups, and disease progression [42,43]. Hypermethylation of GITR [44], MIR34B/C [45], DNMT3A [46], or the combined inactivation of tumor suppressor genes GPX3, RBP1, SPARC, and TGFBI [47] have also been associated with poor prognosis, inferior survival, and progression of MM. In addition, hypermethylation of RASD1 has been correlated with resistance to dexamethasone [48]. DNA methylation score also predicts the efficacy of decitabine, a DNA methyltransferase inhibitor, in MM [49]. In addition to more-studied 5-methylcytosines (5mC), emerging evidence supports pathobiological roles of 5-hydroxymethylcytosines (5hmC), a cytosine modification with a distinct genomic distribution and regulatory role from 5mC [50], in the development and progression of MM. Compared to normal plasma cells, the global 5hmC levels is significantly reduced in MM cells [51], and a lower 5hmC global level, but not 5mC level, was associated with worse OS of MM [52]. MM-specific hydroxymethylome has also been associated with MM cell proliferation and prognosis [53].

Cytosine modifications vary substantially among natural populations, indicating that epigenetic variation is an intrinsic feature in humans [54]. However, unlike solid tumors (e.g. endometrial cancer [55], prostate cancer [56], and breast cancer [57] where population-specific patterns of DNA methylation have been associated with molecular subtypes and survival, the association between cytosine modifications and clinical outcomes of MM in different populations are largely unknown. Recently, emerging evidence from our work showed that 5hmC-modified genes in cell-free DNA at the time of diagnosis were associated with OS and PFS in MM patients [58] and the association differed between EA and AA patients [59]. Further research is warranted to identify population-specific cytosine modifications and their epigenetic modifiers in MM aggressiveness. In addition, the association between cytosine modification changes and cytogenetic abnormalities (e.g. HRCA) by populations remain to be characterized.

Clinical prognostic indices

Emerging evidence suggests that racial differences may exist in the distribution of some clinical prognostic indices that are part of the staging system. For example, in a pooled analysis of nine clinical trials, Ailawadhi et al. [10] found AA patients to be more likely to present with elevated lactate dehydrogenase (LDH) level, an indicator of disease burden and tumor proliferation, and inferior survival. Interestingly, Derman et al. [60] found that AA patients exhibited a greater increase in estimated glomerular filtration rate (eGFR) following treatment compared with EA patients, suggesting a greater improvement of renal function. Despite these interesting findings, how distribution differences impact survival outcomes across populations remains largely unknown. One reason is the limited representation of AA patients and other minorities in studies that informed the development of International Staging System (ISS) or Revised-ISS: only 10% of IMWG-cited studies reported the race/ethnicity of participants, and among the 38,050 U.S. clinical trial participants, only 2,493 of them were AA [61]. Future research using administrative databases with results of laboratory tests (e.g. National Cancer Database [NCDB]) and more diverse patient populations is warranted to better understand the performance of ISS/R-ISS in prognostication across populations.

Non-biological determinants

Treatment utilization

The advent of novel agents (e.g. proteasome inhibitors [PI], immunomodulatory drugs [IMiDs], monoclonal antibodies) and contemporary multidrug combinations (e.g. triplet and quadruplet induction), especially when used with ASCT, have significantly improved survival outcomes of MM. However, the steady improvement in survival outcomes has disproportionally been found among non-Hispanic White patients [11,62,63]. In a population-based study using SEER-9 data (1973–2005), Waxman et al. [11] found that, after the introduction of novel therapies, AA patients had half the observed survival improvements of their EA counterparts. Another analysis of SEER-Medicare data (2007–2013) found that although uptake of novel therapies increased across all populations, the rate of adoption was slower among AA patients compared to EA and Hispanic patients [13]. Importantly, there is evidence that with equitable access to care and receipt of novel therapies, AA [8,10, 64–66] and Hispanic [67] patients appear to have similar or better OS compared to EA patients. Taken together, these findings suggest there may be a yet undescribed interplay of biological differences in treatment response or nonmedical barriers (e.g. unequal access to care) to racial disparities in MM outcomes.

Novel agents use for induction therapy

Over the last decade, combination induction therapy regimens, such as triplet and recently, quadruplet combination therapies, have emerged as the standard of care for newly diagnosed MM patients [68,69]. However, findings regarding racial and ethnic disparities in the receipt of novel induction regimens are mixed. In a comprehensive review of disparities in MM treatment patterns in the US [62], no differences in the utilization of upfront novel agents between race and ethnicity groups were reported in some studies [70–73]. For example, an analysis of The Connect MM Registry (2009–2016) data showed similar proportions of AA and EA patients who received triplet treatment regimen irrespective of ASCT [70]. The types and duration of induction therapy were also similar between AA and EA patients [70]. Of note, enrollment in the community-based Connect MM registry is voluntary, potentially enriched with higher educated and higher SES patients. In contrast, studies have also reported that, compared with EA MM patients, AA and Hispanic patients were less likely to receive novel therapies [10,13, 64,74]. The MMRF CoMMpass study [14] showed that AA patients were less likely to receive triplet induction therapy with PI and IMiDs, compared to EA patients. A recent analysis of the NCDB (2004–2020) [75] found AA patients to have significantly lower odds of receiving triplet or quadruplet therapy compared to EA patients. Interestingly, Dennis et al. [76] found that disparities in access to novel therapies differ by age group. Among MM patients <65 years, AA patients were less likely to receive a triplet regimen compared with EA and Hispanic patients even after controlling for age, year of diagnosis, and comorbidities, but there were no significant racial/ethnic differences in the use of triplet therapy among patients aged ≥65 years [76]. Because racial differences in treatment efficacy have not been demonstrated, it is critical to understand why gaps in treatment utilization exist. As the treatment landscape for newly diagnosed MM patients changes rapidly, there is potential for physician referral bias, type of treatment center, and insurance coverage to all play a role in the underutilization of novel agents by minority MM patients.

ASCT utilization

ASCT is the standard of care for transplant-eligible patients, providing a larger benefit when used earlier in the treatment paradigm of MM [77]. With ASCT being adopted into the standard of care, the rate of ASCT within one year of diagnosis has increased among EA and AA patients, but have stayed the same for Hispanic patients as found in the analysis of the SEER-Medicare database (2007–2013) [13]. However, despite increases in utilization overtime, disparities in ASCT utilization across racial and ethnic groups continue to exist [14,62,74,78–85]. For example, in an analysis of the Nationwide Inpatient Sample of Healthcare Cost and Utilization Project database, Al Hadidi et al. [79] found that AA patients had lower receipt of ASCT compared to EA patients. Similarly, using the NCDB (1998 to 2010), Al-Hamadani et al. [80] reported that compared with EA MM patients, AA, Hispanic, and Asian patients were less likely to receive ASCT [13]. ASCT utilization is known to have a significant impact on outcomes of MM patients. Differential utilization across racial groups may partly explain disparities in survival. Using the MMRF CoMMpass database, Derman et al. [14] found that lack of ASCT was significantly associated with worse OS and PFS in EA and AA patients. Interestingly, they also found racial differences in receiving ASCT or triplet induction across different HRCA groups. Compared to EA patients, AA patients with 0 or 1 HRCA were less likely to receive ASCT or ASCT and triplet therapy, where AA patients with 2+ HRCA were more likely to receive ASCT and ASCT+triplet therapy [14].

Reasons for disparities in ASCT utilization or ASCT+novel agents are likely complex and multifactorial. It may be attributed to implicit bias among physicians against ASCT in AA patients with MM [14]. Alternatively, ASCT is primarily performed at academic centers and this facility-related difference in care delivery can have a direct impact on prognosis. Academic centers typically have facilities and infrastructure to perform molecular testing techniques [86], and MM specialists in large academic centers are thus more likely to rely on risk stratification and the latest clinical evidence to guide risk-adapted treatment decision-making than community oncologist [87,88]. Unfortunately, most population-based studies or those using administrative data (e.g. SEER-Medicare linked data) lack data on both treatment facility and prognostic indices (e.g. disease severity or cytogenetic risk stratification) which limited a comprehensive evaluation of relative contribution of these determinants to disparities in treatment approaches and outcomes.

Chimeric antigen receptor (CAR) T-cell therapy utilization

Compared to alternate treatment regimens, CAR-T cell therapy has shown improved survival outcomes for patients with relapsed and refractory multiple myeloma, resulting in it becoming part of the standard of care [89]. However, due to the low enrollment of minority patients in clinical trials for CAR-T cell therapy, survival outcomes from CAR-T cell therapy across racial groups are less known. A study using the US Multiple Myeloma Immunotherapy Consortium found that, although best overall response rate was lower among Hispanic patients than AA and EA patients, there are no racial differences in PFS or OS [66].

Despite having equal survival outcomes across racial groups, minority MM patients have been found to not have equal access to CAR-T cell therapy in real-world situations. In the Vizient Clinical Database, Ahmed et al. [90] found that AA and Hispanic patients comprised only 1% and 5.4% of the CAR-T cell recipients, respectively, despite representing 17% and 7.4% of the overall MM population identified in the database, a finding consistent with previous disparities published from the National Inpatient Sample (NIS) [91]. The low rates of CAR-T cell utilization among AA and Hispanic patients could be related to the disparities observed across insurance types and CAR-T cell utilization. Patients with private insurance are significantly overrepresented in the CAR-T cell patient population, compared to Medicare, Medicaid, and patients without insurance [90]. AA and Hispanic patients are less likely to have private insurance [92,93], which may have a direct correlation to the underutilization seen in CAR-T therapy. In addition, the barriers to ASCT utilization (e.g. facility, geographic locations, insurance, etc.) remain for CART-cell therapy utilization.

Access to care

In the treatment of patients with MM, access to and rapid initiation of effective therapy is critical. Studies have reported racial disparities in treatment access or clinical outcomes [64,78,94]. Inferior access to care could be resulted from insurance type and lower socioeconomic status.

Insurance type

Insurance type plays an important role in determining access to treatment, the timeliness of interventions, and consequently, treatment outcomes of MM. Generally, patients with private insurance have better access to advanced treatment, including novel agents and ASCT, which have been associated with improved survival outcomes [95–97]. AA patients are more likely to be uninsured or covered by Medicaid compared to EA patients, who tend to have higher rates of private insurance [92,93,98]. Fiala et al. [81] found that compared to traditional Medicare coverage, patients with Medicaid/Medicare double coverage were less likely to receive bortezomib, and ASCT. However, other studies found that insurance status was not significantly associated with receipt of ASCT in EA patients [85] or delayed time to initial treatment [99]. The discrepancies maybe due partly to the challenges of identifying insurance type from electronic healthcare record or lack of prognostic information in population-based administrative database (e.g. SEER-Medicare linked data).

Socioeconomic status (SES)

SES has been shown as an independent predictor of survival in MM [92,100,101]. Both AA and Hispanics are more likely to be in low census-tract SES compared to EA [100,102]. AA and EA patients with MM living in low SES neighborhoods have been found to have worse prognosis compared to their respective counterparts living in high SES neighborhoods [103]. The association between SES and survival of MM in Hispanic patients is less clear. Prior to matching for SES, Hispanics had worse survival outcomes compared to EA patients [102]. When high SES Hispanics were matched with high SES EA patients, there were no differences in survival outcomes. However, when low SES Hispanics were matched to low SES EA patients, low SES Hispanics continue to have worse survival outcomes compared with low SES EA patients [102]. How SES differences actually influence MM treatment and survival outcomes remain largely unknown, mainly because census-tract SES is typically the only method of measuring SES in studies using SEER-Medicare linked data or administrative health records. Census-track SES may not reflect the actual household income, oversimplifies the complexity of SES and financial burden from treatment, and prone to misclassification [90].

Summary and future direction

Although survival outcomes in MM patients resulting from advances in the MM treatment paradigm have improved substantially in the last 20 years, disparities in outcomes among different racial and ethnic groups remain a significant clinical and health equity concern. The underlying reasons for the disparity are likely multifactorial. A complex interplay of population-specific disease biology, treatment utilization, and access to care may all influence treatment decision and contribute to the observed racial and ethnic disparities in outcomes.

The presence of HRCA has a well-established impact on survival outcomes in MM and is being used to guide risk-adapted treatment strategies. However, current HRCA classifications do not adequately account for racial variabilities in their clinical implications. Two recent studies [14,30] that have evaluated the racial differences in HRCA and survival outcomes found no difference in the frequency of HRCA between EA and AA MM patients and yet, HRCA was associated with worse OS in EA patients, but not in AA patients - even after controlling for access to optimal frontline therapy. These findings highlight a critical need for population-specific risk stratification that better reflect the biological diversity across racial groups and ultimately, to ensure equal access to risk-adapted treatment across populations. In addition, genetic and epigenetic alterations not only vary among populations but may also associated with OS and PFS differently between EA and AA patients. Investigating the interplay between HRCA, gene expression, and epigenetic modifications in racially diverse cohorts is essential to elucidate the biological underpinnings of MM health disparities. Despite this need, data on these biological prognostic parameters remain limited for Hispanic and Asian patients in single- and multi-institutional studies. Given the rarity of MM, future collaborative efforts – such as the MMRF CoMMpass study – could ensure sufficient representation of minority patients and implement standardized, comprehensive collection of molecular and clinical data to enable equitable advancements in precision medicine.

With the rapidly evolving therapeutic landscape of MM, differences in the uptake of novel agents across racial groups have become increasingly evident. Evidence suggests that when access to care – particularly novel induction regimens, ASCT, and CAR-T cell therapy, is equal across populations, racial differences in survival outcomes are largely attenuated. This underscores the need to critically examine the multifactorial barriers that contribute to treatment underutilization. These include disease biology and severity, structural barriers to care (e.g. lack of referral pathways or provider networks), provider-related factors (e.g. implicit bias in recommending novel therapy for in minority patients, patient-level decision-making influenced by limited health literacy, and broader socioeconomic constraints, including insurance types. Geographic disparities also play a critical role in treatment utilization, as community hospitals—where minority patients are often overrepresented—frequently lack the infrastructure or resources to administer advanced therapies, including ASCT [80] and/or CAR-T, and report lower rates of triplet and quadruplet regimen utilization [75]. The recognition and dismantling of these systemic barriers are essential to address racial disparities in MM outcomes, particularly as novel therapies continue to redefine the standard of care and prolong survival in MM.

A promising avenue to investigate the influence of these barriers on treatment outcomes across racial groups lies in the use of population-based, real-world datasets such as the NCDB, SEER-Medicare, and other large administrative registries. These datasets offer substantial sample sizes and racial diversity, making them valuable for MM health disparities research. However, a critical limitation of these sources is the absence of granular biological prognostic data, including cytogenetic risk stratification, molecular profiles, and disease burden at diagnosis—all of which are essential to interpreting treatment patterns and survival outcomes. This limitation complicates the interpretation of observed racial disparities, as it remains unclear whether differences in outcomes are driven primarily by inequities in treatment access or by unmeasured differences in disease biology across populations. Additionally, these datasets often rely on census-track SES as a surrogate for SES, which may not accurately capture individual-level financial burden or social determinants of health. Future research using real-world data must prioritize the integration of biological disease characteristics, individual-level SES, and financial burden to elucidate the complex and often interdependent factors contributing to racial disparities in MM outcomes. Furthermore, these analyses must be contextualized within the rapidly evolving therapeutic landscape, accounting for temporal trends in the adoption of novel therapies across populations, to fully understand how disparities emerge and persist over time.

Conclusion

The interplay between molecular and genetic heterogeneity of MM biology and disparities in treatment access and utilization presents a significant challenge in evaluating and addressing prognostic and survival differences across diverse populations. Racial and ethnic disparities in MM outcomes are multifactorial, stemming in part from unequal access to novel therapies, including triplet/quadruplet regimens, ASCT, and emerging CAR-T cell therapies. However, growing evidence also points to population-specific biological differences in treatment response and disease progression – differences that remain poorly understand due to limited inclusion of diverse populations in studies. Addressing these disparities requires an integrated clinical and research agenda that combines equitable access to cutting-edge therapies with rigorous investigation into the biological, genomic, and epigenetic drivers of differential outcomes. Future research must prioritize the inclusion of underrepresented populations in studies, ensure standardized data collection on molecular, cytogenetics, clinical, and sociodemographic data to disentangle the relative impact of biology versus access. Moreover, temporal analyses capturing evolving treatment patterns will be critical to understanding how disparities shift with the introduction of new therapies. Advancing this research agenda is vital to inform equitable treatment delivery and optimal MM therapy for all patients.