Genome-Wide Copy Number Analyses Correlated with Outcomes in Untreated Multiple Myeloma Patients

We reviewed the “Multiple Myeloma Research Consortium's (MMRC) copy number” data set and corresponding clinical annotation available from the Multiple Myeloma Research Foundation (MMRF) Genomics Portal in attempt to identify clinically relevant mutations. Agilent 244k aCGH on DNA from CD138-selected plasma cells were performed on 254 MM patients. We identified 105 untreated patients from this data set for further analysis that had at least partial clinical annotation available.

Median age of the population at diagnosis was 63-years-old (range 40–89), 84% were Caucasian, and 63% were male. Fifty-eight percent had IgG isotype, 18% had IgA, 10% had no heavy chain, and 2% had IgD. Eighty-five patients had albumin and beta-2 available to calculate ISS stage. Fifty- three percent were ISS stage I, 26% were stage II, and 21% were stage III. Median M-Spike was 3.0 g/dL, 78% had elevated free light chains, and 56% had lytic bone disease. Patients received a variety of therapies for MM.

A univariate analysis of the clinical annotation found several previously discovered high-risk groups that were prognostic for survival including: age > 65 years at diagnosis (HR 3.424; p = 0.012), serum creatinine > 2.0 mg/dL (HR 3.197; p = 0.028), and ISS stage 3 compared to 1 and 2 (HR 2.701; p = 0.077). Genome-wide copy number analysis yielded several genomic aberrations that were significantly more common in these high risk sub-groups. Patients 65-years-old or older at diagnosis were more likely to have deletions at chromosome 1p12, 8p21, 10p12, 13q34, 14q24, 16p13, 22q13 and amplifications at 5q35 and 15q15. Deletion of chromosome 10p12 (HR 3.618; p < 0.01) involving genes ANKRD26, MEG4 and amplification of 5q35 involving GRK6, DBN1, DOK3, DDX41, ABS, PDLIM7, F12 and SLC34A1 (HR 2.358; p = 0.07) both correlated with survival. Patients with serum creatinine > 2g/dL were more likely to have deletions of chromosome 1p21, 2p11, and amplifications at 8q24, 14q32.2 and 16p11 (p< 0.05). Interestingly gain of chromosome 14q32.2 (EVL and RNB6) correlated with survival (HR 5.539; p < 0.001).

Our analysis revealed that patients with ISS stage III had higher percent of deletion on chromosome 12q23, 13q32, 16q and amplification of 1q (p < 0.01), although none of these genomic aberrations correlated with survival. Interestingly, the isotype of MM (IgA vs. IgG) did not correlate with survival in this data set, however, deletion of chromosome 8p21 correlated with survival (HR 2.760; p = 0.02) and occurred more frequently in IgA patients (p < 0.01).We also analyzed previously identified high-risk groups: LDH > 300 units/L, CRP > 6.0 mg/dL and patients with lytic bone disease. Several genetic aberrations were more frequent in these groups (p < 0.01), but neither the genomic changes nor the clinical characteristics correlated with survival.

Several high-risk groups have been previously identified using clinical characteristics or genetic data, but are rarely analyzed together. In this data set, we found several factors that correlated with survival including: Age >65, serum creatinine >2.0mg/dL, ISS stage 3, amplification of chromosomes 5q35, 14q32, deletions of chromosomes 8p21, and 10p12. Additional multivariate analysis would determine if these genetic aberrations or the clinical characteristics are independently significant for survival. The heterogeneity of treatments limits the ability of retrospective studies to draw firm conclusions; however, this study is illustrative of the power of combing clinical and genomic data to narrow the multitude of genomic changes to those of clinical relevance.

No relevant conflicts of interest to declare.