Utility of Immunohistochemistry Tissue Array in the Prediction of Epithelial Ovarian Carcinoma Response to Therapy

Tyler O. Kirby,Warner K. Huh,Robert S. Seitz,James E. Kendrick,Y. Wang,Rachel Beck,S.G. DeFoe,BZ Ring,D. T. Ross
DOI: https://doi.org/10.1200/jco.2005.23.16_suppl.5033
IF: 45.3
2005-01-01
Journal of Clinical Oncology
Abstract:5033 Background: Ovarian carcinoma is generally associated with poor prognosis, yet there exists significant variability in outcome amongst patients with similar pathological diagnoses. Robust markers to predict response to therapy are lacking. We sought to define prognostic categories for epithelial ovarian cancer using immunohistochemistry tissue microarray (TMA) technology. Methods: All patients from 1995- 2003 with epithelial ovarian or primary peritoneal carcinoma undergoing initial debulking at a single institution were included. TMA’s were constructed using 0.6mm cores obtained from paraffin embedded blocks and according to standardized protocols using an ordinal scoring system. Antibodies were selected based on targets suggested by large-scale gene expression studies. Univariate analyses were used to identify IHC biomarkers associated with increased disease free survival (DFS), platinum sensitivity (PS), and optimal debulking status. Significant or near-significant variables (p<0.10) were then used in an unsupervised two-step cluster analysis to identify patient groups distinguished by patterns of IHC staining. These cluster groups were analyzed for association with DFS and PS using Cox and logistic regression analysis controlling for stage, grade, histology, and debulking status. Results: 161 (127 St III/IV, 23 St II, 9 St I, 2 unk) patients were evenly distributed with respect to debulking status (75 optimal, 72 suboptimal, 14 unknown). Of the 22 proteins analyzed by univariate analysis, 5 were found to be significantly associated with DFS and 2 with debulking status. Two-step cluster analysis distinguished 3 patient clusters of which membership in cluster 1 (n=82, 50.9%) was associated with significantly longer DFS, optimal debulking and PS. Cox regression analysis demonstrated that the association with DFS was independent of stage, grade, histology, and debulking status. Conclusions: TMA based technology represents a cost-effective method for discovery of prognostic IHC classifiers for epithelial ovarian cancer. Univariate analysis combined with cluster analysis of TMA data has the ability to discriminate classifiers with clinical prognostic significance. Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Applied Genomics
What problem does this paper attempt to address?