The mutator phenotype and genomic instability are key steps in the origin of many human cancers. In hereditary nonpolyposis colorectal carcinoma, defects in the DNA mismatch repair proteins are associated with microsatellite instability. In this issue, Catto and colleagues (pages 484–490) examine the expression levels of two DNA mismatch repair proteins and the presence of microsatellite instability in 111 cases of transitional cell carcinoma of the bladder (TCC). They report that a significant proportion of TCCs exhibit reduced expression of hMLH1 or hMSH2, two DNA mismatch repair proteins. This occurred in younger patients and was related to both tumor stage and grade. The least differentiated and most invasive tumors were more likely to exhibit reduced DNA mismatch repair protein expression. Unexpectedly, reduced expression of either protein appeared to predict subsequent outcome, with fewer recurrences and relapses in all tumors and in superficial tumors; however, 5-year survival rates did not reach statistical significance with respect to protein expression. Microsatellite instability was observed in less than 10% of all tumors but, unexpectedly, was not correlated with the loss of DNA mismatch repair proteins. These observations suggest that mismatch repair proteins expression levels could serve as prognostic indicators for TCC. Despite improved diagnostic and treatment options, colorectal cancer remains one of the most common and fatal cancers. Nakayama and colleagues previously reported that p16 promoter methylation is present in the tumors of colorectal cancer patients and can be detected in the serum of these patients using methylation-specific PCR. In this article (pages 491–493) the same group has tested a group of 45 colorectal cancer patients for the presence of p16 promoter methylation in serum. They report that methylated tumor DNA was detected in the serum samples of 31 out of 45 (69%) patients with recurrent colorectal cancer. In sharp contrast, none of the 50 patients whose tumor DNA had no methylation in p16 promoter were positive. These observations suggest that analysis of the state of methylation of the p16 promoter in serum DNA could become a sensitive and selective method to detect recurrent colorectal cancer. DNA methylation controls gene expression and is often dysregulated in cancer cells. Aberrant DNA methylation not only can silence gene transcription when it occurs near promoter regions, it can also lead to enhanced gene mutations as 5-methylcytosine is deaminated to thymidine or can promote carcinogenesis through enhanced allelic loss. As a measure of DNA methyltransferase activity in certain tumors, levels of DNA methyltransferase I (DNMT1) mRNA have been studied in the past. DNMT1 is the best characterized of the three methyltransferase enzymes known so far. In this issue, Saito and colleagues focus on the measurement of DNMT1 protein levels in hepatocellular carcinoma (pages 527–532). They generated a specific antibody directed against human DNMT1 and measured protein levels by immunohistochemistry. Normal liver tissue or precancerous conditions such as cirrhosis or chronic hepatitis did not stain positive. However, nearly half of the samples obtained from individuals with hepatocellular carcinoma were positive for DNMT1. DNMT1-positive tumors were poorly differentiated and showed portal vein involvement and, most importantly, correlated significantly with low survival. These results indicate that increased DNMT1 protein expression might play a role in the malignant progression of hepatocellular carcinoma. Thymomas develop in the thymic epithelium, which normally directs T-lymphocyte maturation. These tumors manifest asymptomatically as mediastinal masses, cause local chest pain or present concurrently with neurological symptoms of myasthenia gravis. This issue of IJC features the first population-based study describing the demographic pattern of thymoma incidence in the United States. Engels and colleagues (pages 546–551) focused exclusively on malignant thymomas, which differ from the more common benign thymomas based on evidence for capsular invasion. Their data confirmed that malignant thymoma is an extremely rare disease. Thymoma incidence was higher in men than in women and peaked late in adulthood (72 years). Interestingly, incidence rates were nearly doubled in individuals from Hawaii, indicating a genetic influence of the disease. To examine whether malignant thymomas arise as a component of an inherited cancer syndrome, special attention was placed on subsequent malignancies. Interestingly, the risk of developing a second malignancy was not broadly elevated although the risk for B-cell non-Hodgkin lymphoma was enhanced in surviving thymoma patients. Whether non-Hodgkin lymphomas were caused by abnormal B-cell proliferation due to dysfunctional T cells or presented as a consequence of radiotherapy remains to be determined. The authors further pointed out that 5 patients developed 2 subsequent malignancies, and 4 of the 5 patients developed lung tumors. They will further examine this novel association in future studies. Kaplan-Meier survival curves for individuals with hepatocellular carcinoma grouped by their DNMT1 status. (A) Recurrence-free survival is shown, while overall survival is plotted in (B).