The study purpose was to develop a patient's prognostic index (PI) reflecting the genetic information in cDNA microarray-based CGH experiment data for estimating a gastric cancer patient's survival time. The developed methodology was fit to and validated using data from the Cancer Metastasis Research Center at Yonsei University; 30 pairs of gastric tumors and normal gastric tissues were used in the cDNA microarray-based CGH. The cDNA microarrays containing 17,000 sequence-verified human gene probes were directly compared. Genetic alteration score (GAS) was constructed based on the genes that had a high frequency of alteration among all the genes displaying small variations across the arrays. GAS was determined using a technique that finds linear combinations of the original variables that best account for the variability in the data. When classifying cancer patients with the PI predicted by the model incorporating GAS, the correct classification rate for recurrence was 83.33%. In conclusion, GAS allowed for providing an independent patient's PI that reflects the genetic information for prognosis on hazard rate of recurrence, which was capable of distinguishing a patient's recurrence status, survival status and cancer stage status. The predicted PI also provided each patient's estimated disease-free survival rate. In this study, 82 genes were selected for analysis based on a high frequency of alteration and small variations across the arrays. In addition, 13 genes displaying a possible relationship with disease-free survival time were identified. GAS was found to be associated with the recurrence status and survival status.

Related Articles