Expression data from non-small cell lung carcinoma (NSCLC)

The analytical validation of a 15 gene prognostic signature for early-stage, completely resected, non-small-cell lung carcinoma that distinguishes between patients with good and poor prognoses. Matched fresh frozen (FF) and RNALater (RNAL)-treated NSCLC specimens from 43 distinct patients were employed in this study.  Each specimen was split into 2 biological replicates, equaling 172 samples (86 FF, 86 RNAL) in total which were analyzed via microarray.  Microarray output was used to calculate the risk score (risk of recurrence) using a previously-described signature. Concordance (in risk category – high, low) and correlation (in risk score) between the FF and RNAL specimens was calculated. After concordance and correlation between FF and RNAL specimens was established, RNAL specimens only were used to evaluate the analytical performance of the signature. Accuracy was evaluated using 34 samples, comparing signature concordance and correlation between the clinical laboratory and a reference laboratory. Precision was evaluated using 8 samples (4 high risk, 4 low risk), accounting for both intra-run and inter-run concordance. Analytical sensitivity was evaluated using 1 sample, assessed at 3 different RNA input amounts. Analytical specificity was evaluated using 2 samples (1 high risk, 1 low risk). Contributor: Precision Therpeutics, Inc.