Genomic-wide cDNA microarray screening to correlate gene expression profile with chemoresistance in patients with advanced lung cancer
Fumihiro Oshita, Mizuki Ikehara, Akiko Sekiyama, Nobuyuki Hamanaka, Haruhiro Saito, Kouzo Yamada, Kazumasa Noda, Yoichi Kameda and Yohei Miyagi
We conducted a study using cDNA microarray analysis to determine whether expression levels of genes in tumors were correlated with tumor response to chemotherapy. Between September 2000 and December 2001, 47 patients were registered in the study. Eighteen patients had small cell lung cancer (SCLC), and others had non-small cell lung cancer (NSCLC). All patients except three received platinum based chemotherapy. Sixteen of the 18 patients with SCLC (89%) and 13 of the 29 patients with NSCLC (45%) responded to chemotherapy, respectively. Transbroncheal biopsy specimens of tumors were obtained before chemotherapy. The expression levels of 1176 genes in tumor specimens were analyzed using the AtlasTM Human Cancer 1.2 Array. When we analyzed the data for correlations between gene expression levels and tumor response to chemotherapy, there was a significant increase in the expression of nine genes in non-responders compared with responders to chemotherapy (p< 0.01). Multivariate regression analysis revealed that allogenic inflammatory factor, HLA-DR antigen associated invariant subunit and MHC class II HLA-DR-beta precursor were independent chemo-resistant factors (p<0.0001). When we analyzed the differences in gene expression levels between patients with SCLC and NSCLC, expression levels of one or more resistant genes were increased in comparison with the mean expression of control house keeping genes in five of 18 SCLC patients and 19 of 29 NSCLC patients, In conclusion, some chemo-resistant genes were detected in the tumor tissue of lung cancer patients using cDNA microarray analysis. A prospective study is required to confirm whether expression levels of these genes reflect chemosensitivity.