© 2002 by Oxford University Press
Journal of the National Cancer Institute, Vol. 94, No. 18, 1343,
September 18, 2002
© 2002 Oxford University Press
IN THIS ISSUE |
Germline BRCA1 mutations confer a substantial lifetime risk for breast and ovarian cancer, but whether the mutation increases the risk of other cancers is not known. Thompson and Easton (p. 1358) evaluated the risk for cancer at other sites in a large cohort of individuals from families in Europe and North America that carry a BRCA1 mutation. They found that BRCA1 mutation carriers had an increased risk for several cancers, including pancreatic cancer and cancer of the uterine body and cervix. Men with BRCA1 mutations who were younger than 65 years old had an increased risk for prostate cancer but carriers 65 years old or older did not. Overall, the authors found that increases in the risk for cancer at sites other than the breast or ovary were small and were primarily evident in women.
In another study, Brose et al. (p. 1365) determined cancer risk in a population of BRCA1 mutation carriers (some of whom were included in the study by Thompson and Easton) who were identified because they had been referred to cancer risk evaluation clinics on the basis of a family history of breast and/or ovarian cancer. The risk of breast and ovarian cancer among BRCA1 mutation carriers was higher than population-based estimates but lower than that in families ascertained for linkage studies. Mutation carriers also were at increased risk of colon, pancreatic, gastric, and fallopian tube cancers. The authors note that their data may provide the most relevant cancer risk estimates to use in advising women who undergo BRCA1 mutation testing through cancer risk evaluation clinics.
In an editorial, Gruber and Petersen (p. 1344) note that family studies tend to overestimate risks but can provide insights into associations between BRCA1 mutations and cancers other than breast or ovary that can be investigated with the use of other study designs. In the meantime, they conclude that the useful message of these studies is that the increased risks of cancers other than breast, ovarian, and fallopian tube cancers are likely to be small.
Variability in Mammogram Reading
The use of "test sets" of films have shown that radiologists’ interpretations of mammograms vary widely. However, little is known about the extent to which such variability exists in the real-world setting. To investigate this question, Elmore et al. (p. 1373) analyzed medical records from more than 2100 women whose nearly 9000 mammograms had been read by 24 radiologists. The radiologists varied widely in their interpretations, with false-positive rates ranging from 2.6% to 15.9%. Adjusting for patient, testing, and radiologist characteristics narrowed this range to 3.5% to 7.9%. Younger radiologists and those who had graduated more recently from medical school had higher false-positive rates. The authors note that better understanding of the sources of variability in mammogram interpretations should ultimately result in reduced rates of false-positive mammograms but continued high sensitivity and accuracy.
In an editorial, Kessler et al. (p. 1346) note that the source of the variability in false-positive rates that Elmore et al. observed could not be determined: The radiologists may be equal in quality, with different thresholds for suspicion, or they may vary in quality. In any event, they suggest that double reading of mammograms, as is done in a number of countries (but not the United States), would improve both sensitivity and specificity of mammography.
CD40-Mediated Apoptosis in Human Urothelial Cells
In epithelial cells, binding of the cell-surface receptor CD40 to its ligand, CD40L, can cause different effects, which include growth inhibition, pro-inflammatory cytokine secretion, or apoptosis. Bugajska et al. (p. 1381) examined the effects of ligation of CD40 and other tumor necrosis factor superfamily members on the survival of normal and malignant human urothelial cells that express CD40. They found that the apoptotic responses of the cells to CD40 ligation depended on how CD40L was presented to the cells. For example, soluble trimeric CD40L did not induce apoptosis in normal or malignant urothelial cells, whereas CD40L that was presented on the surfaces of co-cultured mouse fibroblasts induced apoptosis in malignant but not in normal urothelial cells. The authors conclude that susceptibility to CD40 ligation-induced apoptosis may be a useful therapeutic approach for eliminating neoplastically transformed urothelial cells and that loss of CD40 expression by urothelial cells may be an important adaptive mechanism that facilitates the development and progression of bladder cancer.
BRCA2 Alterations in Ovarian Cancer
The BRCA2 gene is well known to be involved in hereditary ovarian cancer, but its role in sporadic ovarian cancer has been unclear. To investigate the role of BRCA2 mutations in ovarian cancer and the mechanisms of inactivation of this gene, Hilton et al. (p. 1396) used a protein truncation test to identify inactivating mutations in the BRCA2 gene in ovarian tumors from 92 unrelated women with ovarian cancer whose BRCA1 mutation status was already known. Five tumors had a germline mutation in BRCA2, and four tumors had a somatic mutation. Twelve additional tumors lacked BRCA2 mRNA, but BRCA2 promoter hypermethylation was identified in only one of them. Another 23 tumors demonstrated loss of heterozygosity. Overall, 82% of the tumors contained an alteration in BRCA1, BRCA2, or both genes. The frequency of BRCA1 or BRCA2 dysfunction was independent of family history, but the mechanism of dysfunction was not.
Human Papillomavirus Cofactors for Cervical Cancer
Although infection of the cervical epithelium by one of 13 oncogenic human papillomavirus (HPV) types is the primary cause of cervical cancer and its immediate precursor, cervical intraepithelial neoplasia grade 3 (CIN3), most HPV infections do not progress to CIN3. Case–control studies have suggested that HPV cofactors, such as smoking, parity, and oral contraceptive use, influence the progression of an HPV infection to cervical cancer. Castle et al. (p. 1406) conducted a 10-year prospective study among 1812 women who were positive for oncogenic HPV DNA at enrollment to examine associations between these potential cofactors and the risk of CIN3 or cervical cancer. They found that current and past smoking, but not current oral contraceptive use at enrollment or the number of past live births, were associated with the risk of CIN3 or cervical cancer. The authors conclude that smoking is a risk factor for CIN3 and cervical cancer among HPV-infected women.
| ||||||||||||||||||||||||||||||||||||