Microarray versus conventional prediction of lymph node metastasis in colorectal carcinoma

Background: The authors investigated whether microarray-based gene expression analysis of primary tumor biopsy material could be used to predict lymph node status in patients with colorectal carcinoma (CRC). Lymphatic metastasis strongly determines treatment algorithms in CRC. Currently, postoperative histology results are needed to determine lymph node status. Reliable preoperative information would be useful to advance treatment strategies. Methods: In specimens from 66 patients with CRC from the Erlangen Registry of Colorectal Cancer, 41 shock-frozen samples of International Union Against Cancer (UICC) Stage I?II CRC and 25 samples of UICC Stage III CRC were microdissected manually, RNA was isolated, and gene chips (HG-U133A; Affymetrix) were hybridized. Prediction rates for lymphatic metastasis were calculated using conventional clinicopathologic parameters, gene expression data, and a combination of both. Prediction error, specificity, and sensitivity were analyzed using six different statistical classifiers. Results: Analysis of conventional parameters produced a positive prediction rate that ranged between 53% and 61%, sensitivity of 42%, and specificity of 72%. Microarray prediction rates were between 62% and 67% for lymphatic metastasis. Specificity was between 76% and 83%, and sensitivity was between 38% and 48%, depending on the statistical procedure. The conventional estimates were improved by 9?12% when array data were added. Conclusions: Current data show that the prediction of lymphatic metastasis can be improved by gene expression profiling of the primary tumor biopsy, alone or in combination with conventional parameters. Gene expression profiling may become valuable increasingly in planning treatment for patients with CRC.