Protein Expression Profiling of Irinotecan Pathway in Lung Cancer
W Zhang, WS Zhang, EL Lee, HL McLeod. Howard University Hospital, Washington, DC; The Second Hospital of Nanjing, Nanjing, Jiangsu, China; UNC School of Pharmacy, Chapel Hill, NC
Background: The exact mechanism for variation of response to chemotherapy remains unclear. Individualized therapy strategies for cancer will require a more thorough understanding of the pathways influencing drug fate, including expression of cellular target enzymes, metabolism enzymes and cellular transporters. Irinotecan is an excellent example of an anticancer drug in need of individualization. We profile expression of proteins in an irinotecan pathway in lung small cell carcinoma and non-small cell carcinoma tissues and construct a new pharmacologic pathway to help advance individualization in the selection of cancer therapy.
Design: Paraffin embedded tumor tissues from 15 patients with lung small cell carcinoma, 49 patients with stage IV non-small cell carcinoma (21 squamous cell carcinoma, 18 adenocarcinoma and 10 undifferentiated large cell carcinoma) were used to construct a tissue microarray. Fifteen markers, including CES2, NFkappaB, XRCC1, CDC45, TDP1, TOP1, PARP1, ABCC1, ABCC2, ABCC3, ABCB1, ABCG2, p53, ERCC2, UGT1A1, that are related to the irinotecan pathway were immunohistochemically stained on the tissue microarray. Protein expression was assessed to derive an overall expression level. Hierarchical clustering was used with unsupervised algorithms to identify patterns of protein expression that produced distinct clusters of patients.
Results: The relative expression levels across the 15 pathway proteins and the interpatient variability were considerable. Using hierarchical clustering, 4 protein clusters and 3 patient groups had highly similar indices based on the protein expressions. Protein expression of this drug pathway is not associated with lung carcinoma histological type. Cluster analysis identified a variety of histological types with the same pharmacological profile. The 3 patient groups had no unique clinical pathologic features but could be differentiated by the statistically significant differences in the protein expression levels of 5 proteins.
Conclusions: Gene expression profiling could be valuable for predicting tumor response to chemotherapy and for tailoring therapy to individual cancer patients. Establishing the ability of pathway analysis to discern patient populations who have a high likelihood of benefit from a given chemotherapy agent or those who will receive no benefit will provide the basis for selection of therapy for individual patients.
Category: Pan-genomic/Pan-proteomic Approaches to Diseases
Monday, March 22, 2010 1:00 PM
Poster Session II # 240, Monday Afternoon