Project description:This series represents expression profiles of 34 non-seminoma germ cell tumors (NSGCTs) from patients who received cisplatin based chemotherarpy for treatment of their disease for whom full clinical follow-up information was available. These specimens were used as a validation set to test outcome prediction models using a subset of previously profiled GCT specimens (see GEO accession #GSE3218). Keywords: Disease state analysis (good vs. poor patient outcome)
Project description:This series represents expression profiles of 34 non-seminoma germ cell tumors (NSGCTs) from patients who received cisplatin based chemotherarpy for treatment of their disease for whom full clinical follow-up information was available. These specimens were used as a validation set to test outcome prediction models using a subset of previously profiled GCT specimens (see GEO accession #GSE3218). Experiment Overall Design: Tumor tissues were collected under IRB-approved protocols at the Memorial Sloan-Kettering Cancer Center, New York, between 1987 and 2003. The tumor samples consist of 34 NSGCT specimens. RNA was isolated, labeled, and hybridized to Affymetrix U133A and U133B microarrays using standard protocols. Data were processed by the RMA method.
Project description:Expression profiling of a panel of 101 adult male germ cell tumors and 5 normal testis specimens was performed on Affymetrix U133A and U133B microarrays. This data has been used to: 1) generate a gene classifier that predicts histology (see PMID 15870693). 2) identify candidate target genes on 12p, a region that is gained in almost 100% of germ cell tumors (see PMID 16424014) 3) identify pluripotency associated genes through comparison of pluripotent embryonal carcinoma vs. undifferentiated seminoma. ONGOING: 4) Identification of genes associated with patient outcome and cisplatin resistance. Keywords: germ cell tumor, histologic subtypes
Project description:Microarray analysis was used to determine the expression of 12,000 genes in a set of 50 gliomas, 28 glioblastomas and 22 anaplastic oligodendrogliomas. Supervised learning approaches were used to build a two-class prediction model based on a subset of 14 glioblastomas and 7 anaplastic oligodendrogliomas with classic histology. A 20-feature k-nearest neighbor model correctly classified 18 of the 21 classic cases in leave-one-out cross-validation when compared with pathological diagnoses. This model was then used to predict the classification of clinically common, histologically nonclassic samples. When tumors were classified according to pathology, the survival of patients with nonclassic glioblastoma and nonclassic anaplastic oligodendroglioma was not significantly different (P = 0.19). However, class distinctions according to the model were significantly associated with survival outcome (P = 0.05). This class prediction model was capable of classifying high-grade, nonclassic glial tumors objectively and reproducibly. Moreover, the model provided a more accurate predictor of prognosis in these nonclassic lesions than did pathological classification. These data suggest that class prediction models, based on defined molecular profiles, classify diagnostically challenging malignant gliomas in a manner that better correlates with clinical outcome than does standard pathology.