Evaluating the discriminatory capacity of miR-371A-3P in the context of pure seminomatous testicular cancer metastases.

Abstract

424 Background: Conventional serum tumor markers (STMs) are noted to have poor specificity in detecting seminomatous germ cell tumors (SGCTs). Recent publications have suggested that miR-371a-3p may offer improved performance for these cases. However, miR-371a-3p’s performance in detecting pure seminoma at retroperitoneal lymph node dissection (RPLND) has yet to be elucidated in any publicly available study. This lack of data is in part due to the rare occurrence of such cases, and it represents a distinct path for exploring miR-371a-3p’s clinical utility. Methods: Pre-surgical serum samples were collected prospectively from patients prior to RPLND. Within the 15-patient cohort, 6 were assigned as ‘Control’ and 9 were assigned as ‘Seminoma’ based on pathological confirmation of viable GCT. Out of all 15 patients, 10 were chemotherapy naïve prior to RPLND. Post-chemotherapy patients made up 5/6 ‘Control’ patients. MiR-371a-3p levels were quantified by RT-qPCR. Using 2−∆∆Cq method, miR-371a-3p expression was first normalized to the miR-30b-5p reference gene and then calculated relative to the mean target expression across a collection of RNA extracts from healthy male donors’ serum. The Cq values were statistically evaluated to obtain performance characteristics (sensitivity and specificity). Results: Although assay results revealed no significant difference in miR-371a-3p expression between groups, median relative expression in the seminoma group trended higher than the control group (Rq = 3705 and 241 respectively, p = 0.2844). Out of all chemotherapy naïve patients, 9/10 had viable GCT at RPLND whereas 7/10 showed elevated miR-371a-3p. Among the five post-chemotherapy patients, 0/5 had viable GCT at RPLND and 2/5 had elevated miR-371a-3p. The assay provided 7/9 true positive designations in the seminoma group and 4/6 true negatives in the control group. The two false positive results were from post-chemotherapy patients. Of the two optimal thresholds calculated by Youden’s J statistic, the lower threshold of 28.62 (78% sensitivity ad 67% specificity) was selected because it was more in line with previously published data. ROC analysis provided an AUC of 0.704 (95% confidence interval: 0.43-0.98, p = 0.1949). STMs performance was poorer at 67% sensitivity and 17% specificity. Conclusions: Performance metrics for miR-371a-3p exceed those of STMs but were substantially lower than previous reports that evaluated performance in pre-orchiectomy and primary RPLND settings. A possible explanation for this disparity is that miR-371a-3p’s performance is hindered in the post-chemotherapy RPLND setting. However, any strong conclusions from these results are limited by the small sample size which is partly due to the rarity of this clinical scenario. These results suggest that pure seminoma at RPLND may be a key clinical context wherein the miRNA assay may require further refinement.