Development of a molecular tool for oral squamous cell carcinoma risk assessment using microrna markers

© 2015 Dr. Gareena PrasadPurpose The aim was to develop a robust methodology for analysis of microRNAs by assessing the validity of using formalin fixed paraffin embedded (FFPE) tissues, determining the most optimal cDNA conversion method and establishing a sound methodology for data analysis using PCR threshold, microRNA efficiency and appropriate normalization techniques. In addition, the aim was also to develop a microRNA biomarker panel and develop an algorithm for assessment of oral squamous cell carcinoma (OSCC) risk using microRNA abundance data. And finally, to assess the clinical utility of microRNA in oral cytological scrapings. Material and Methods To assess the utility of FFPE in microRNA analysis, 4 oral lesions were divided in half and were flash frozen and FFPE. Additionally, differing amounts of RNA were used to compare and establish the most sensitive means of cDNA conversion. Quantitative reverse transcription PCR (qRT–PCR) was conducted on all samples and comparisions were done to establish PCR efficiencies for microRNAs and establish appropriate methods for assessing PCR thresholds and normalization techniques. qRT-PCR was conducted on RNA isolated from 40 samples of formalin fixed and paraffin embedded (FFPE) biopsy samples that were histologically assessed as OSCC or histologically normal epithelium (HNE), as well as additonal oral tissue samples of oral lichen planus (OLP), mild, moderate and severe dysplasia. A panel of 11 microRNAs considered as potential biomarkers of OSCC were assessed with obtained Cq values normalized and statistically analyzed using qBase PLUS (Biogazelle) software. The microRNAs were evaluated by assessing oral cytological scrapings stored in different media at varying temperatures over differing time frames. Results Results showed that FFPE could be used for microRNA analysis along with pre- amplification in addition to Megaplex reverse transcription priming being the most sensitive method for cDNA conversion. Robust techniques were established for determining PCR thresholds and microRNA efficiencies along with global mean normalisation being deemed as the most appropriate normalization technique. Four microRNAs showed statistically significant abundance differences between the 20 OSCC and 20 HNE. Subsequently, an algorithm was developed for assessment of OSCC-risk (the “miR-OSCC-risk”) that gave sensitivity and specificity calculations for the HNE and OSCC samples tested. The algorithm output indicated high (flagged as red), indeterminate (amber) or low (green) risk, that showed a sensitivity of 95%; specificity of 70%; positive predictive value of 76%; negative predictive value of 93%; and an accuracy of 82.5%. The abundances of the microRNAs in the panel and the predictive accuracy of the algorithm were tested further using other oral lesion samples. Oral cytological scrapings of normal epithelia were in the majority of instances determined to be of low risk by the algorithm whereas the oral mucosal lichen planus and dysplastic lesions were in the majority determined to be of medium to high risk and warranting further clinical and histological investigation. Despite varying storage media, storage conditions and time frames of storage of cytological scrapings, microRNAs could be obtained and measured using qRT-PCR. Conclusions The developed methodology and miR-OSCC-risk algorithm can be used with qRT-PCR microRNA abundance data to provide an indication of OSCC risk that has potentially significant clinical utility. The diagnostic procedure developed here could be of benefit in the assessment of oral lesions of unknown pathology by the targeting of lesions requiring biopsy