The etiology of multiple myeloma remains unknown, but some heritable contribution is suspected. Very few genetic risk markers have been reported in more than one study population. A nested analysis (72 cases, 144 controls) in the Nurses’ Health Study (NHS) and Health Professionals Follow-up Study (HPFS) cohorts suggested a role in multiple myeloma susceptibility for single nucleotide polymorphisms (SNP) in the genes for the insulin receptor substrate type 1 (IRS1; rs17208470, rs1801278) and the interleukin (IL)-6 receptor (IL6R; rs6684439, rs7529229, rs2228145). The NHS and HPFS cohort members are predominantly of European origin; the present analysis further explored the IRS1 and IL6R loci in persons of European descent and in African Americans in a large pooled population comprised of five U.S. population-based studies.
We used unconditional logistic regression to calculate odds ratios (OR) and 95% confidence intervals (CI) assuming a log-additive inheritance model and adjusting for age, sex, and study site. We performed all analyses separately in 824 participants of European origin (259 cases, 565 controls) and 298 African Americans (114 cases, 186 controls). We also calculated a risk allele score by summing all variant alleles across the five SNPs in the ∼95% of participants with no missing genotype data, and we used additional logistic regression models to estimate a race-specific multivariable OR and 95% CI for multiple myeloma risk per variant allele carried. All statistical tests were two-tailed and assumed an α-error of 0.05.
In persons of European descent, at the locus rs17208470 (IRS1) we observed a statistically significant 42% increase in multiple myeloma risk per copy of the variant allele (OR, 95% CI: 1.42, 1.02-1.98; p=0.037). In IL6R, we observed a statistically significant association with multiple myeloma at rs6684439 (OR, 95% CI: 1.27, 1.01-1.58; p=0.037) and marginally significant associations with multiple myeloma at two loci: rs7529229 (1.24, 0.99-1.55, p=0.057) and rs2228145 (1.24, 0.99-1.55, p=0.056). The risk allele score analysis (240 cases, 532 controls of European origin with no missing genotype data) showed that the risk of multiple myeloma increased significantly by 11% per variant allele carried (OR, 95% CI per risk allele: 1.11, 1.03-1.19; p=0.008). In African Americans, two loci in IL6R, rs6684439 and rs7529229, showed suggestive non-significant associations with multiple myeloma risk (rs6684439, OR, 95% CI, p-value per copy of the minor allele: 1.37, 0.94-2.00, p=0.11; rs7529229: 1.39, 0.95-2.04, p=0.089). Neither SNP in IRS1 was associated with multiple myeloma in this group. The risk allele score analysis in African Americans (109 cases, 176 controls) suggested a 17% increase in multiple myeloma risk per variant allele carried (1.17, 0.99-1.38; p=0.067).
Based on a priori hypotheses, we examined the association of five previously reported SNPs in IRS1 and IL6R with multiple myeloma in a larger series of persons of European descent than previously published, and for the first time, in African Americans. Both genes belong to biologic signaling pathways with important roles in multiple myeloma pathogenesis. In participants of European origin, SNPs in IRS1 and IL6R were at least marginally associated with multiple myeloma, and the risk of multiple myeloma increased significantly per risk allele carried. We observed slightly stronger ORs for the IL6R SNPs in African Americans, but no associations were statistically significant in that group, probably due to a lack of statistical power because of small numbers for this subgroup and/or because the true causal SNPs in these genes have differential linkage disequilibrium across racial groups. The nominally significant and suggestive associations that we observed are plausible and should be further evaluated in other large multi-ethnic studies of multiple myeloma susceptibility.
Anderson:celgene: Consultancy; onyx: Consultancy; gilead: Consultancy; sanofi aventis: Consultancy; oncopep: Equity Ownership; acetylon: Equity Ownership.
Asterisk with author names denotes non-ASH members.