AUTHORS
Tara Rakiewicz, MD, Zachary Fusfeld2, Joshua Banks, MD3, Beatrice Razzo, MD4, Theresa Hyslop5, Adam Binder, MD6
AFFILIATIONS
1 Thomas Jefferson University Sidney Kimmel Comprehensive Cancer Center, Philadelphia, Pennsylvania, United States
2 Thomas Jefferson University Sidney Kimmel Comprehensive Cancer Center, Philadelphia, Pennsylvania, United States
3 Thomas Jefferson University Sidney Kimmel Comprehensive Cancer Center, Philadelphia, Pennsylvania, United States
4 Thomas Jefferson University Sidney Kimmel Comprehensive Cancer Center, Philadelphia, Pennsylvania, United States
5 Thomas Jefferson University Sidney Kimmel Comprehensive Cancer Center, Philadelphia, Pennsylvania, United States
6 Thomas Jefferson University Sidney Kimmel Comprehensive Cancer Center, Philadelphia, Pennsylvania, United States
ABSTRACT
Introduction: There is a well-established risk of progression along the spectrum of plasma cell dyscrasias, from the premalignant Monoclonal Gammopathy of Undetermined Significance (MGUS) to Smoldering Myeloma (SM) and eventually to Multiple Myeloma (MM). What isn’t well established is the etiology for the disproportionate incidence of these diseases amongst races/ethnicities. The Surveillance, Epidemiology, and End Results (SEER) data from 2019-2023 revealed the incidence of myeloma in Non-Hispanic Black (NHB) men was 17.9 per 100,000 persons, more than double the incidence of myeloma in Non-Hispanic White (NHW) men at 8.4 per 100,000 persons. We do not have a strong understanding of why NHB individuals have a higher risk of developing multiple myeloma. Studies show evidence of germline predisposition for plasma cell dyscrasias, but these have been performed largely in limited populations and individuals of western European ancestry. We aimed to evaluate patients with plasma cell dyscrasias and use geocoding to evaluate how various publicly available social determinants of health (SDOH) measures impact development of disease.
Methods: A cohort of patients was created by query of Epic’s database. Patients were included if they had an ICD-10 diagnosis of a plasma cell dyscrasia between 2008-2025, reside in Pennsylvania (PA), New Jersey (NJ) or Delaware (DE), and seen in the Jefferson Health System at least one time. We use an Integrated Nested LaPlace Approximation approach to account for spatial autocorrelation, using a modified Besag-York-Mollie areal model to assess associations between two binary latent class determinants of neighborhood SES and diagnostic stage of plasma cell dyscrasias (MGUS, SM, and MM) using a Proportional Odds Model. The two latent classes used are derived from American Community Survey covariates: Advantage class uses proportion of professional class males and females, and residents with below a high school education; these classes are separated to High Advantage (HA) or Low Advantage (LA). Disadvantage class uses various inputs such as proportion of single-headed, female-headed, crowded, and renter households, as well as those with no vehicle or phones. It also uses the proportion of residents who are NHB, below the poverty line, on public assistance, and unemployed. This class is separated to High Disadvantage (HD) or Low Disadvantage (LD). Models are additionally adjusted for categorical age at diagnosis, birthyear, sex, and NHB racial identity.
Results: A total of 6832 patients were evaluated. 4476 patients with MGUS, 117 with SM, and 2239 with MM. 1980 patients identified as NHB, while 3823 identified as NHW; 50.89% of the cohort were female. Most patients (n=5254) reside in PA, with less in NJ (n=1469) or DE (n=118). The breakdown of all existing tracts in PA, NJ, and DE were LD n=4035 (75%) and HD n=1379 (25%), as opposed to HA n=3409 (63%) and LA n=2005 (37%). Our total cohort’s distribution was as follows: LD n=4871 (70%), HD n=2079 (30%), as opposed to HA n=4662 (67%) and LA n=2288 (33%). For the patients with MGUS, tract distribution was LD n=3162 (70.7%) and HD n=1312 (29.3%), as well as HA n=3065 (68.5%), LA n=1409 (31.5%). For SM, tract distribution was LD n=83 (70.9%) and HD n=34 (29.1%), as well as HA n=79 (67.5%), LA n=38 (32.5%). For MM, tract distribution was LD n=1542 (68.9%) and HD n=697 (31.1%), as well as HA n=1440 (64.3%), LA n=799 (35.7%). Patients of HD showed an OR of 0.79 (CI 0.63-0.99), whereas patients of LA had an OR of 1.43 (CI 1.16-1.76).
Conclusion: Analysis of the plasma cell dyscrasia population seen at Jefferson Health shows that patients who are of LA have a significantly increased OR of having a diagnosis of MM as opposed to MGUS or SM. Compared with individuals in high advantage tracts, those in low disadvantage tracts had a 43% higher odds of SM or MM vs MGUS, as well as a 43% higher odds of MM vs MGUS or SM. This highlights SDOH as a potential risk factor for pathogenesis of disease. Interestingly, there was a protective effect seen of HD; a possible nuanced aspect of the models used, as high levels of public assistance and urban dwelling may impact results. Notably, our population was not consistent with SEER data, as national data shows the NHB population greatly outnumbered NHW population in all groups. We plan to further evaluate our patient population in the future with chemical and pollutant exposures to create a more complex map.
DISCLOSURES
Tara Rakiewicz, MD: No relevant conflicts of interest to declare.