Abstract 4147: Ratio of miR-146b/miR-191 in serum predicts prognosis in surgically resected lung squamous cell carcinomas

Abstract

Abstract Background: We profiled 61 lung squamous cell carcinomas (SCC) and 10 matched adjacent normal lung tissue using the Ambion mirVana Bioarray that contains 328 human micro-RNA (miR) probes (Raponi et al. Cancer Research 2009). MiR-146b and miR-191 showed the strongest prediction accuracy for stratifying the prognostic groups. We hypothesize that these miRs can be detectable in serum and its expression in serum might be prognostic. Methods: Using the mirVana PARIS kit (Ambion), total RNA was isolated from 0.5 ml serum samples from 60 resected lung SCC patients, 30 healthy controls and 17 patients with benign lung disease. Reverse transcription reaction using Megaplex RT primers (Applied Biosystems) and preamplification using Megaplex PreAmp primers (Applied Biosystems) were performed. Expression of mature miRs, miR-146b and miR-191, were examined by real-time quantitative polymerase chain reaction using the comparative cycle threshold (Ct) method. Since the normalization of miRNA data in blood samples is still controversial, the ratio between the expression values of miR-146b and miR-191 was calculated and data log transformed. Results: miR-146b and miR-191 expression was detectable in the serum from all lung SCC patients and the serum miR-146b/miR-191 ratio was significantly higher in lung SCC patients (0.29±0.25) than in cancer-free subjects (0.16±0.11) (p-value=0.001). The miR-146b/miR-191 ratio showed a modest discriminative receiver-operator characteristic (ROC) curve, distinguishing cancer patients from cancer-free subjects with areas under the ROC curve at 0.69 (95% CI 0.59-0.79). In 20 SCC patients, the primary tumor miR-146b/miR-191 ratio (calculated using our microarray data) was compared with the miR ratio in serum and a significant correlation was observed (Pearson coefficient=0.455; p-value<0.001). Median overall survival of patients with high miR-146b/miR-191 ratio was significantly shorter (2.25 years ±0.32) than in patients with low miR ratio (5 years ±0.23, log-rank p-value=0.020). After adjusting by clinical covariates (sex, age and stage), miR-146b/miR-191 ratio remained as independent prognostic markers individually. Using the median of miR-146b/miR-191 ratio as a cutoff, the adjusted hazard ratio (HR) for death in those patients with higher miR ratio was 2.24 (95% CI 1.18-4.23). A validation of this miR-146b/miR-191 ratio as a prognostic marker in an independent cohort of resected lung SCC is warranted. Conclusions: Serum miR-146b/miR-191 ratio was significantly higher in lung SCC patients than in cancer-free subjects, but its accuracy to detect lung cancer is limited. Serum miR-146b/miR-191 ratio was significantly correlated with the primary tumor expression and the serum expression has a potential utility as independent prognostic marker in resected lung SCC. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4147. doi:1538-7445.AM2012-4147