If the patient had ever had chemotherapy for PC, regardless of whether or not that patient was LY2109761 700874-71-1 undergoing chemotherapy at the time the sample draw, the sample was classified as being post-chemotherapy. For patients in whom multiple samples were drawn on different dates, all samples were used in the data analysis unless explicitly stated in the results section. Once the predictor set was established, it was validated in a second set of randomly selected PC and healthy samples. The statistician was blinded to the identity of the samples. Applying the cut-off obtained through the Compound Covariate Prediction method, the samples were classified as either ����PC���� or ����non-PC����. The analyzer was then unblinded and the accuracy of the prediction determined by comparison with the actual diagnosis. We also applied the same equation to a subset of prechemotherapy pre-surgical PC patients to determine the ability of the predictor set to correctly classify patients into PC vs. non- PC. This is important as the influence of chemotherapy and/or surgery on the gene expression profile of PBMCs cannot be ruled out. Further, the latter group of patients represents the ideal patient population in whom the test, if validated would be applied in a clinical setting. In recent years it has been repeatedly demonstrated that genetic expression in PBMCs is altered in the context of malignancy. This observation of an altered PBMC genetic expression profile in cancer patients was first reported in diffuse large B-cell lymphoma and chronic lymphocytic leukemia and later extended beyond hematological malignancies through the analysis of PBMC expression profiling in patients with advanced renal cell carcinoma. In both hematologic malignancies and in RCC, it was reported that the variation in gene expression between patients with disease and healthy controls was much greater than the inter-sample variation observed for the healthy patients alone, suggesting that PBMCs could be useful surrogate markers with potential diagnostic and GSK2118436 Raf inhibitor prognostic applications in cancer. Further, in RCC, it was shown that an 8- gene classifier set developed from the differentially expressed genes could predict the diagnosis of malignancy with 100% accuracy. Recently, Huang et al. have reported that a differential gene expression profile does exist in PBMCs of PC patients. While this study also used microarray and Q-RT PCR validation to establish differential genetic expression in the peripheral blood of PC patients, its purpose was to establish potential biomarkers that could differentiate newly diagnosed diabetic patients with PC from diabetics without PC.