Association between soluble immune biomarkers and monoclonal gammopathy of undetermined significance (MGUS) in three large cancer epidemiology cohorts with participants from underrepresented groups.

Abstract

Abstract Introduction: MGUS is a necessary (but not sufficient) precursor of multiple myeloma (MM), thus, examining determinants of MGUS may provide clues to the etiology of MM. We evaluated the role of soluble immune biomarkers that may influence B-cell differentiation to plasma cells and plasma cell growth and microbial translocation in MGUS. Methods: We obtained previously collected serum samples from MM-free participants in 3 prospective cohorts (Black Women's Health Study [BWHS], Multiethnic Cohort [MEC] and the Southern Community Cohort Study [SCCS]) including those from underrepresented groups and tested them for MGUS using standard laboratory methods. We confirmed 898 MGUS cases and 952 MGUS/MM-free controls. We measured CXCL10, CCL/MCP2, gp130, sIL6Rα, angiopoietin2, hepatic growth factor (HGF), IL10, CCL/MIP1α, and IL6 from the same samples using Luminex multiplex assays (R&D). We also measured sCD14 and lipopolysaccharide binding protein (LBP) levels to reflect microbial translocation. Biomarker distributions were normal with log transformation and analyzed as continuous measures, except for MP1α and LBP which were analyzed as tertiles. Demographic and anthropometric information was obtained from study questionnaires. The association between biomarker levels and MGUS was assessed using unconditional logistic regression, adjusted for race/ethnicity, sex, age at blood draw and body mass index (BMI) at baseline . Additional analyses were stratified by sex, race/ethnicity, age (<60 years, ≥60 years), and BMI (<30 kg/m2, ≥30 kg/m2). Statistical significance was assessed with Wald p-values corrected for multiple comparisons using FDR. Results: The racial/ethnic distribution of the participants was 60% Black, 16.5% White, 10.6% Japanese American, 9.3% Latino, and 3.6% Native Hawaiian. Females comprised 64.4% (Black females comprised roughly half of the total participants) and the mean age at blood draw was 66.3 years (SD=9.02). Thirty-three percent and 28% of cases and controls were obese, respectively. Most biomarkers levels were mildly correlated. Among controls, there were significant differences in some biomarker levels by race/ethnicity, sex, and obesity, with IL6 significantly different by all demographic categories (highest in Black controls). All biomarker levels significantly increased with age except MIP1α and LBP (no association). Increasing HGF levels were significantly associated with MGUS overall and within many subgroups. The overall odds ratio [OR] was 1.66 per log pg/mL (95% confidence interval (CI): 1.32, 2.09, p=0.0002). Significant ORs ranged from 2.40 (p= 0.008) for those with IgA to 1.52 in Black participants (p=0.04). Increasing sIL6Rα levels were significantly associated with MGUS overall (OR= 1.70, 95% CI= 1.17,2.47, p=0.017), and among women (p=0.046), non-obese participants (p=0.011) and IgGk (p=0.021). Increasing gp130 levels were also significantly associated with MGUS among women <60 years (p=0.042) non-obese (p=0.015) and Black participants (p=0.042). Increasing angiopoietin levels were associated with MGUS only among younger women (p=0.042). Neither sCD14 nor LBP levels were associated with MGUS in any subgroup, although the effect estimates were inversely associated with MGUS. Conclusions: HGF was the biomarker most strongly and consistently associated with MGUS, across multiple subgroups. HGF is produced by plasma cells and myeloma cells, and promotes migration of plasma cells to bone marrow, angiogenesis and plasma cell growth, and is associated with progression from MGUS to myeloma and myeloma prognosis. The robust association observed here reflects its central role in myelomagenesis. IL6Ra was also associated with MGUS overall and in multiple groups and is another biomarker that is expressed on plasma cells. sIL6Rα and gp130 are receptors for IL6, a known plasma cell growth factor. There were no strong associations between immune biomarkers and MGUS among obese patients (at higher risk for MGUS, MM and progression), suggesting a different mechanism of risk. Finally, there was no evidence that microbial translocation was associated with MGUS. A limitation was the cross-sectional study design, with MGUS identification and biomarkers measured from the same sample, which precludes assignment of causal direction for association with MGUS. Nevertheless, the evidence supporting the association between HGF and IL6Ra and MGUS in Black participants and in female vs. male patients is noteworthy.