Mutations in K-ras codon 12 detected in plasma DNA are not an indicator of disease in patients with non-small cell lung cancer.

To the Editor: The demonstration that cell-free circulating DNA detected in the plasma of cancer patients is genetically identical to that of the primary tumor has generated substantial interest, leading to 200 publications (http:// www.ncbi.nlm.nih.gov). Recently, Wang et al. (1 ) reported in this journal that the method chosen for DNA isolation might contribute significantly to mutation detection (in their case, K-ras mutations in the plasma of patients with colorectal cancer). Briefly, they recommended the use of a modified guanidine/Promega resin method (G/R) to isolate DNA, affirming that this method enhances assay sensitivity. We used the same approach to detect K-ras mutations in the plasma of patients with nonsmall cell lung cancer (NSCLC) and compared the Qiagen vs the G/R method for isolation of circulating DNA. We purified DNA from plasma samples and cancer tissues from 12 patients. The DNA in 2 aliquots of the plasma from each patient was isolated by the Qiagen method (1, 2) and by the G/R method according to Wang et al. (1 ). Additionally, DNA was isolated by the Qiagen method from matched plasma and tissue samples (n 10 for each) and from 76 plasma-only samples (36 from cancer patients and 40 from cancer-free volunteers) according to Kopreski et al. (2 ). There was no difference in the number of K-ras mutations detected in the plasma samples collected from patients (n 15; 41.7%) and from volunteers (n 12; 30.0%; odds ratio 1.6; P 0.21). In addition, when we evaluated the presence of K-ras mutations in the matched plasma and neoplastic tissue samples, we observed no correlation. Finally, when we compared the results (K-ras status) for DNA samples isolated from plasma by the 2 different methods with the results obtained for the DNA isolated from tissue samples (12 patients), we observed K-ras codon 12 mutations in 2 different tissue samples, whereas we detected no mutations in plasma DNA isolated with the Qiagen method and 2 mutations different from those identified in the corresponding tumor tissue in the plasma DNA isolated with the G/R method. The correlation between controls and cases was not significant ( 2 0.7; P 0.5). The correlations between results obtained for tissue DNA and for plasma DNA isolated by the Qiagen or G/R method also were not significant (P 0.4 and P 1.0, respectively, nonparametric test for cases vs samples), nor was the correlation between the G/R and Qiagen isolation methods (P 0.2). These data show no relationship between K-ras mutations found in DNA from plasma and tumor tissue from NSCLC patients. Our results do not support the suggestion that K-ras mutations detected in plasma DNA are markers for tumor detection. Ramirez et al. (3 ) analyzed K-ras mutations (codon 12) in tumor and paired serum DNA of 51 NSCLC patients undergoing surgery and detected mutations in 9 tumors and 12 serum samples. As suggested by Gautschi and Ziegler (4 ), serum may contain not only DNA of tumor origin, but also a variable fraction of DNA derived from in vivoand in vitro-damaged hematopoietic cells. Indeed, lymphocytes stimulated with phytohemagglutinin or antigen may release DNA (5 ). Furthermore, human leukocytes stimulated by neutrophil-derived hydroxyl radicals may cause activation of K-ras codon 12 (6 ). In addition, inhaled particles in exposed individuals can generate reactive oxygen species that can activate K-ras (7 ). Moreover, K-ras mutations are detected in neogenetic lesions of subpleural fibrotic lesions, including ciliated bronchial epithelium and metaplastic epithelium (8 ). Because chronic inflammation may be present in patients without cancer as well as those with cancer (9 ), the analysis of K-ras mutations in plasma may be influenced by such factors. Finally, as reported recently by Keohavong et al. (10 ), K-ras mutations are frequently found in histologically normal tissues near tumors, suggesting that such mutations may represent an early event in the development of lung cancer. K-ras mutations thus may be present before clinically detectable tumors. As a final point, it is important to note that K-ras mutations have been detected in patients with ulcerative colitis, Crohn disease (11, 12), and Helicobacter pylori-associated chronic gastritis (13 ). We observed K-ras mutation in 2 volunteers diagnosed with chronic gastritis, in 1 patient with kidney failure, and in 2 who were heavy smokers. In light of these observations, plasma DNA assays for the detection of mutations in codon 12 of K-ras do not provide a reliable method to screen populations for the somatic mutations frequently found in neoplasms. Further confirmatory studies are required.