Häfner et al., pp. 1842–1846 Even in countries with a high standard of health care, large numbers of cervical cancer patients succumb to the disease. To improve the detection of metastases, Hafner et al. investigated RNA expression of two molecular markers in lymph nodes. They found the tumor-associated gene HPV16-E6 to be a specific and sensitive marker for the detection of intact tumor cells in lymph nodes, with the potential to aid detection of metastases too small to be recognized using standard techniques. Despite advances in early diagnosis and improved treatments, 40% of cervical cancer patients die. Even when no cancer can be detected in the lymph nodes, the disease often recurs. In lymph nodes that appear disease-free by routine diagnostic evaluation, exhaustive serial sectioning can turn up prognostically significant micrometastases. These detection methods, however, are time-consuming and costly. The authors used a PCR-based assay, more sensitive and less susceptible to the distribution of tumor cells within the lymph node compared with immunohistochemical assays. They compared expression of the viral oncogene HPV16-E6, a hallmark of cervical cancer cells, and that of cytokeratin 19 (CK-19), often used as a tissue-specific marker for metastatic cells. About a third of the histologically negative lymph nodes showed HPV16-E6 expression, compared with 80% that contained CK-19. In addition, 8 of 10 lymph nodes from noncancer patients expressed CK-19 but not HPV 16-E6. Another study, currently in progress, will evaluate how well HPV oncogene expression can be used to predict metastases, a tool that could potentially reduce cervical cancer's death toll. Kaz et al., pp. 1922–1929 Abnormal gene methylation influences the initiation of colorectal cancer differently in sporadic and familial colon cancers, and methylation patterns may be useful in early detection, a new study finds. Methylated genes frequently appear in colon cancers, but whether they cause the disease or simply occur as part of the disease progression has not been determined. This study examined the frequency and pattern of methylation in certain tumor suppressor genes in adenomas from individuals with hereditary nonpolyposis colon cancer (HNPCC), or the Lynch syndrome. Recent studies have shown that abnormal methylation affects many genes in various cancers, including colon cancer. These epigenetic alterations could lead to cancer by inactivating tumor suppressor genes. Though most colorectal cancers occur sporadically, some familial forms exist; individuals with the Lynch syndrome carry mutations in genes responsible for DNA mismatch repair. Tumors then arise in these individuals when a genetic alteration inactivates the second allele. Epigenetic changes such as methylation might also disable the DNA mismatch repair system and provide the “second hit” required for tumor formation. Using methylation-specific PCR assays, the researchers assessed the frequency of aberrant methylation in four genes previously shown to be methylated in premalignant neoplasms in the colon. One gene involved in mismatch repair, MLH1, was methylated in more than half the adenomas from Lynch syndrome patients, but fewer than 5% of sporadic adenomas. Another gene, p14ARF, also showed more frequent methylation in familial adenomas than in sporadic adenomas, while two other genes, MGMT and CDKN2A/p16, presented the opposite pattern: more likely to be methylated in sporadic adenomas. These results suggest possible pathways by which formation of sporadic and Lynch syndrome adenomas are controlled and indicate that gene methylation could serve as an early detection marker for Lynch syndrome adenomas. Larsson et al., pp. 1993–1998 The risk of developing pancreatic cancer increases along with body mass index (BMI), according to a new systematic review of published studies. This meta-analysis of existing data reveals a 12% increase in pancreatic cancer risk associated with a 5 kg/m2 increase in BMI. Pancreatic cancer poses a particular threat because it is aggressive and usually advances to an incurable stage before it is diagnosed. Fewer than 5% of patients survive 5 years. Therefore, any modifiable risk factors that can be identified could help focus prevention strategies. Epidemiological studies have associated overweight with various cancer types. Because obesity has been related to glucose intolerance, insulin resistance, hyperinsulinemia and type 2 diabetes, it is biologically plausible that BMI could influence pancreatic cancer. Previous work has shown an increased risk of pancreatic cancer accompanying long-term diabetes, and plentiful evidence suggests that insulin resistance might also contribute to its onset. In 2003, a systematic review of case-control and prospective studies evaluating BMI and pancreatic cancer risk identified a positive association: 2% increase in risk per 1 kg/m2 increase in BMI. Since then, several new studies have been published, and Larsson et al. have now updated this meta-analysis. The authors analyzed 21 prospective studies and determined a relative risk of 1.12 per 5 kg/m2 increase in BMI, with no study reporting results dramatically off that number. No evidence of publication bias was found. Future work, therefore, should provide insight into how insulin resistance might play a role in pancreatic cancer.