Circulating Mir-130a in Multiple Myeloma and Extramedullary Myeloma Patients

Background

MicroRNAs (miRNAs) are a class of short, non-coding, single stranded RNAs regulating a broad spectrum of processes. Circulating miRNAs are an important emerging biomarker in cancer as well as a possible non-invasive diagnostic solution for a wide range of clinical disorders due to their high stability and association with disease state, although their source still remains uncertain. In multiple myeloma (MM), a plasma cell malignancy, circulating miRNAs have been reported to have a diagnostic and prognostic potential. It is therefore plausible to assume that they are involved in pathogenesis of this disease and thus could be used as diagnostic tool not only for MM, its extramedullary (EM) form but also for monitoring the clinical course of the disease. Therefore, in this study, we aimed to identify such miRNAs.

Methods

Screening analysis of 667 miRNAs was performed on 5 EM serum samples, 5 newly diagnosed MM samples and 6 healthy donors (HD) serum samples using TaqMan Low Density Arrays (TLDA) from Life Technologies. QPCR was performed for miR-130a on 118 serum samples obtained in Brno from newly diagnosed MM patients (pts) (35 pts), primary and secondary EM (35 pts), relapsed MM (18 pts) and HD (30). Further, 45 serum samples (12 diagnostic and 33 follow-up) of pts reaching VGPR/better response, enrolled in Italian CRD/MEL-200 and EMN-02 studies were used for circulating miRNA estimation. Receiver Operating Characteristic (ROC) analysis was used to calculate specificity and sensitivity of the miRNA as a biomarker. Biochemical characteristics were also available for EM and MM pts from Brno. P values <0.05 were considered as significant.

Results

TLDA profiling revealed 14 deregulated miRNAs (all p<0.05, adjusted p<0.41) between MM pts and EM pts, and 20 miRNAs were on the top of the list of deregulated miRNAs between EM and HD serum samples (all p<0.05, adjusted p<0.40). MiR-130a was chosen for further verification by qPCR as it was on the top of the list of deregulated miRNAs between the groups. qPCR revealed that level of miR-130a was significantly decreased in MM and EM samples when compared with HD (all p<0.005); moreover, level of miR-130a was decreased also in EM when compared with MM sera (p<0.06). To discriminate EM pts from other groups, ROC curve was calculated. It revealed that miR-130a is potent to distinguish EM pts from HD with area under the curve (AUC) = 0.805, specificity of 86.7% and sensitivity of 65.7% using cut-off value = 3377 copies/1ng of miRNA/RNA. Most importantly, miR-130a was able to distinguish EM pts from newly diagnosed MM pts with AUC = 0.628, specificity of 94.3% and sensitivity of 28.6% using cut-off value = 1438 copies/1ng of miRNA/RNA, and EM pts from MM pts in relapse with AUC = 0.702, specificity of 94.4% and sensitivity of 28.6% using cut-off value = 1438 copies/1ng of miRNA/RNA. In the cohort of EM pts, miR-130a significantly correlated with most of clinically relevant parameters; there was a positive correlation with level of hemoglobin and thrombocytes count (r_s=0.397 and 0.439, all p<0.05) and a negative correlation with levels of monoclonal immunoglobulin, β₂-microglobulin and C-reactive protein (r_s=-0.398, -0.427 and -0.488, all p<0.05) and it was also associated with higher ISS stage (p=0.017).

Further, in the analysis of miR-130a dynamics in follow-up samples from Italy, we observed increase of miR-130a levels in 8/12 MM pts during the follow-up sampling (p<0.06) in comparison with diagnostic samples, whereas in remaining 4 MM pts it remained stable or decreased.

Conclusions

In this study, miR-130a was decreased in serum samples of pts developing EM disease and distinguished EM pts from newly diagnosed MM pts and relapsed/progressed MM pts with specificity over 90%. Further, we observed increased level of miR-130a in the follow-up samples of MM pts. It suggests that miR-130a could be associated with EM disease; however, underlying biology and origin of miR-130a still remains to be explored.

Work was supported by grants IGA NT 12130, NT 14575 and NT 13190.