Diagnostic potential of microRNAs (miRs) in malignant pleural mesothelioma (MPM)

Introduction: MPM is histologically difficult to distinguish from reactive mesothelial proliferations (RMPs), partly because proposed MPM-markers are not specific, reproducible or validated enough (Zimling ZG, Jørgensen A & Santoni-Rugiu E, Histopathology 2012; 60(6B):E96-105). MiRs are small non-coding RNA-strands (∼22 nt) that post-transcriptionally regulate gene-expression, vital cellular processes and oncogenesis. However it is uncertain whether currently published miR-data may provide candidate diagnostic biomarkers for differentiating MPM from RMP. To pursue this goal, we performed a screening of miR-expression in FFPE diagnostic biopsies, surgically resected MPM-specimens and corresponding surrounding non-neoplastic pleura (NP).

Materials and Methods: We performed a RT-qPCR-based (Exiqon®) screening of 742 human miRs’ expression in preoperative biopsies, epithelioid MPM-, and NP-specimens from 5 patients treated with extra-pleural pneumonectomy as part of trimodal protocol. The relevant identified differentially expressed miRs were validated on tissue samples from 41 independent MPM-patients by RT-qPCR TaqMan® MicroRNA Assays (Applied Biosystems). Quantification-cycles (Cq) were determined by the related 7500 RT-PCR system software, comparatively analyzing PCR-data with an internal reference-gene (RNU49). Significant (p < 0.05) differences between groups were detected by 1-way ANOVA with Bonferroni post-tests. Validated miR-targets were assessed by immunohistochemistry (IHC).

Results: We identified significant difference in expression of 7 cancer-relevant miRs (down-regulation of miR-17-5p, -126, -143, -145 and -652 and up-regulation of miR-221 and -193a-3p) that correctly differentiated MPM from RMPs and was not influenced by chemotherapy (comparable miR-expression-levels in surgical samples and diagnostic biopsies). We furthermore tested if the 7 miRs could fulfill the recommendations of the International Mesothelioma Interest Group (IMIG) for a suitable MPM-marker (sensitivity/specificity > 80%) by generating ROC-curves using the RT-qPCR-data. The IMIG-criteria were not met for miR-17-5p, -221 and -143, while miR-126, -145, -193a-3p and -652 showed high sensitivity and specificity. Preliminary IHC-data of the validated miR-126 target L-type amino acid transporter 1 (LAT1) showed good correlation between miR-126 down-regulation and LAT1 up-regulation in MPM.

Conclusion: The cancer-relevant miR-126, -145, -193a-3p and -652 have great potential as biomarkers for distinguishing MPM from RMPs, while miR-17-5p, -143 and -221, previously reported to be specific for MPM's histological subtypes and to play a pathogenic role in MPM, are not entirely suitable for this differential diagnostic purpose. Further studies by in situ-hybridization techniques will additionally corroborate the possible use of these miRs in MPM diagnostics.