Background: The conventional tests for the diagnosis of early stage pancreatic carcinoma are not acceptable. This meta-analysis is to evaluate the accuracy of K-ras mutation for the diagnosis of pancreatic carcinoma.
Data sources: A systemic search of all relevant literature was performed in Web of Science, EMBASE, Cochrane Database, and MEDLINE (PubMed as the search engine) prior to June 1, 2011. Thirty-four studies fulfilled the inclusion criteria and data were pooled for analysis.
Results: The pooled estimates for K-ras mutation in diagnosis of pancreatic carcinoma were as follows: sensitivity 0.68 (95% CI: 0.66-0.71), specificity 0.87 (95% CI: 0.85-0.88), positive likelihood ratio 4.54 (95% CI: 3.47-5.94), negative likelihood ratio 0.37 (95% CI: 0.30-0.44) and diagnostic odds ratio 14.90 (95% CI: 10.02-22.15). Summary receiver operating characteristic analysis demonstrated that the maximum joint sensitivity and specificity was 0.79, and the overall area under the curve was 0.86.
Conclusions: Diagnostic accuracy of K-ras mutation was not superior to that of conventional tests. Therefore, K-ras mutation analysis alone is not recommended for the diagnosis of pancreatic carcinoma.