BRAF Mutations and Inflammatory Gene Expression in Myeloma Cells from Patients with Renal Dysfunction

Introduction: Multiple Myeloma (MM), the second most common blood cancer, is a clonal disease of long-lived plasma cells (PC) that is currently incurable. Symptomatic myeloma-defining events include hyper calcemia, renal dysfunction, anemia, and bone disease (CRAB criteria). Renal dysfunction is primarily due to free light chain (FLC) precipitation with uromodulin in the distal tubules resulting in light chain cast nephropathy (LCCN). Renal dysfunction may have a negative impact on patient outcomes as it can influence the ability to optimally treat patients, and in some may result in hemodialysis dependence. Therefore, understanding the factors that can contribute to renal dysfunction could allow for early intervention to prevent associated morbidity.

Methods: To determine whether certain proteins or genetic mutations in MM patients are associated with renal failure, we analyzed data from the Multiple Myeloma Research Foundation's CoMMpass Study (NCT01454297, Interim Analysis 16) using creatinine levels as a surrogate for renal dysfunction (>176 μmol/L indicates renal failure). We correlated creatinine with other measures of renal dysfunction (BUN), serum protein levels (total protein, M-protein, IgG, IgA, IgM, IgL-kappa, IgL-lambda), as well as structural changes including t(11;14), t(12;14), t(14;16), t(14;20), t(6;14), t(8;14), hyperdiploidly, 1q21 gains and 17p13 loss as determined by whole genome sequencing (WGS). We used whole exosome sequencing (WES) data to assess if common mutations (KRAS, NRAS, BRAF, DIS3, FAM46C) were associated with renal dysfunction. The numbers in each comparison differed based on the availability of WGS (structural events: 621 vs. 96) or WES (SNVs: 1003 vs. 119). To study gene expression we focused on the patients with the highest and lowest creatinine levels (70/group). Since there was a sex bias in these groups we included a covariate for sex to determine gene expression changes independent of sex. Finally, we analyzed RNAseq analysis for expression of light chain variable regions to determine if there was an association of light chain usage with renal dysfunction.

Results: We initially correlated BUN, serum protein levels and structural events with creatinine levels, but only found a significant correlation with BUN levels (Pearson R=0.7275, P=<0.0001, N=816). We next performed a contingency analysis comparing genetic events in patients with creatinine >176 μmol/L versus those <176 μmol/L. While there were no differences in structural events between those with normal vs. elevated creatinine, we did observe a significant increase in BRAF mutations in patients with high creatinine (Fisher's exact test, P=0.0093). We also observed a trend towards significance with KRAS mutations (Fisher's exact test P=0.0930). We examined light chain variable region usage but found no association with high creatinine levels. We expanded this analysis to all genes, and found 110 genes differentially expressed with renal dysfunction, which were enriched for genes involved in inflammatory responses (AIF1, LY96, LYVE1, MPEG1, SELL1, STAB1, TNFRSF12A) and metabolism (COX7A2, FOLR2, MT-ND1, MT-ND3).

Conclusions: Our data are consistent with previous studies with respect to the lack of association of renal dysfunction with serum proteins and translocations. However, we observed interesting associations with BRAF mutations and a trend with KRAS mutations, suggesting a role in disease pathogenesis beyond driving tumorigenesis. Additionally, we detected expression changes with several genes involved in inflammatory responses. We cannot conclude if this is due to inflammation related to acute kidney disease or if the myeloma plasma cells expressing these genes are more likely to produce FLC that result in LCCN. Finally, we did not observe a variable chain usage bias, however our study is limited to expression and therefore points to the likelihood that somatic hypermutation plays an important role in whether a light chain can lead to renal disease. Taken together, these data suggest the presence of genetic and biological differences in myeloma cells associated with renal dysfunction and provide insights into identifying patients that could develop renal disease.

ACKNOWLEDGEMENTS

This work was supported through the STEP-UP HS program from the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health, R25DK113659