Genomic and Systemic Metabolism Differences Associated with Racial Disparities in Multiple Myeloma


Background: Racial disparities exist in the prevalence of multiple myeloma (MM) and monoclonal gammopathy of unknown significance (MGUS), with significantly higher prevalence in African American/ black (AA) compared to non-AA individuals. AA patients are also younger than non-AA at the time of diagnosis. We aimed to determine if differences in systemic metabolism and genetics might contribute to these racial disparities in MM. There are currently no approved anti-myeloma therapeutics that simultaneously reverse the metabolic drivers for MM tumor growth.

Methods: Studies were approved by the IRB and IACUC. We used a large health system electronic medical record database to characterize the metabolic phenotype of AA and white individuals with MM. Body mass index (BMI) was calculated from height and weight measures. BMI≥30kg/m2 defined obesity. HbA1c ≥6.5% defined diabetes. Genomic analysis was performed by extracting mutational signatures from tumor samples of MM patients as previously described (PMID: 23945592), and comparing their prevalence between AA and white patients. We used an immunodeficient non-obese mouse model of type 2 diabetes (Rag1-/-/MKR) and controls (Rag1-/-) for in vivo MM1.S xenograft studies. ON123300, a dual ARK5/CDK4 kinase inhibitor was used for preclinical studies.

Results: Of our population of 3170 people, 2128 individuals were AA (21.7%) or white (78.3%). Females comprised 53.2% (n=246) of the AA population and 40.2% (n=669) of the white population. The highest prevalence of obesity was in AA women (46%) > white men (41%) > AA men (36%) > white women (35%). The prevalence of diabetes was greater in AA men (37%) and women (34%) than white men (24%) and women (19%). BMI was a poor predictor of diabetes in the AA population with MM, where diabetes affected 1 in 5 AA individuals with normal BMI, but only 1 in 12 white individuals.

Genomic characterization of AA patients, revealed a significant enrichment for the COSMIC SBS1 clock-like mutational signature in AA patients compared to white (p < 0.05), which was confirmed after adjusting for age. The SBS1 signature correlates with the age of individuals and may represent a cell division clock.

The male Rag1-/-/MKR mice had hyperglycemia and insulin resistance, and were not obese relative to control mice when fed a regular chow diet. The MM1.S tumor xenografts grew much more rapidly in the Rag1-/-/MKR mice compared with controls; 60 days post injection, tumors were approximately 5 times larger than controls and ex vivo analysis of protein lysates revealed increased activation of the insulin receptor (IR) / insulin-like growth factor 1 receptor (IGF-1R) and mTOR signaling pathway. Our previous studies (Perumal et al, CCR 2016) have characterized a role for ARK5(AMPK-related protein kinase 5), which is a major regulator of glutamine uptake in tumor cells that co-express ARK5 and MYC. Preclinical testing of MM cell lines and xenograft mouse models show a robust reduction of MM tumor growth as well as inhibition of IGF-1R-downstream signaling pathways, which contribute to tumor proliferation.

Conclusions: In our large diverse cohort of individuals with MM, we found significant racial disparities in the prevalence of obesity and type 2 diabetes, in addition to enrichment of a clock-like mutational signature in MM from AA individuals. Our findings suggest a signature of accelerated aging in MM from AA patients, which is concordant with younger age at diagnosis. In our mouse model, MM xenografts grew more rapidly in the diabetic mice and had more activation of mTOR signaling, a pathway known to regulate clock and aging gene signatures. ON123300 was able to reverse the metabolic drive as well as directly reduce tumor growth in vivo xenograft models.