© Oxford University Press 2006.
IN THIS ISSUE
Practice Change After a Meeting PresentationPublication of results from clinical trials can lead to rapid and dramatic changes in treatment patterns. Giordano et al. (p. 382) show that presentation of results of a clinical trial at a major medical meeting can produce similar changes, well before full publication of the study. This conclusion emerged from their analysis of changes in adjuvant chemotherapy use in women in the SEERMedicare linked database who had been diagnosed with stage I-III breast cancer during 19941999. In May 1998, the initial results of Cancer and Leukemia Group B Study 9344, which showed that paclitaxel improves survival of women with lymph nodepositive breast cancer, were presented at a meeting of the American Society of Clinical Oncology. The use of taxanes as part of chemotherapy for lymph nodepositive primary breast cancer rose dramatically shortly after the meeting, long before the study was published (in 2003) and even before FDA approved the regimen (in 1999). The authors note that a process for rapid dissemination of detailed data from practice-changing trials is essential.
In an editorial, Woloshin and Schwartz (p. 372) note that the CALGB 9344 investigators and patients who received the new regimen were lucky in the sense that the initial results held up over time. The authors suggest that physicians approach preliminary research cautiously because early reports may have undergone only limited peer review and because an initially promising regimen may turn out to be ineffective or to have toxic effects that are apparent only over time.
HPV Infection and Incidence of Skin Cancers
Because the relationship between human papillomavirus (HPV) infection and keratinocyte cancers, such as squamous cell carcinomas (SCCs) and basal cell carcinomas (BCCs), remains unclear, Karagas et al. (p. 389) carried out a population-based casecontrol study of 252 SCC case patients, 525 BCC case patients, and 461 control subjects. They detected HPV antibodies more frequently in plasma from SCC patients than in plasma from control subjects but found no difference in HPV seropositivity between BCC patients and control subjects. Seropositivity for HPV types in the genus beta, particularly HPV 5, was associated with SCC risk. The highest SCC risk was associated with seropositivity for multiple HPV types and a tendency to sunburn, but the associations had limited statistical precision. The authors conclude that their findings support a role for HPV types from the genus beta in the pathogenesis of SCC.
C/EBP
Activity and Lung Cancer
Protein expression of the tumor suppressor CCAAT/enhancer-binding protein
(C/EBP
) is low in several tumor types. To determine whether DNA methylation and histone acetylation of the C/EBP
promoter regulate expression of this gene in lung cancer, Tada et al. (p. 396) examined C/EBP
expression and the C/EBP
gene promoter in 15 human lung cancer cell lines and 122 human primary lung tumors. The authors observed low or no C/EBP
expression in 12 of the cell lines and in 81 of the tumors, which was associated with hypermethylation and histone acetylation in the upstream C/EBP
promoter. C/EBP
promoter activity was restored in cells treated with the histone deacetylase inhibitor trichostatin A and the demethylating agent 5-aza-2'-deoxycytidine. The authors conclude that DNA hypermethylation of the upstream C/EBP
gene promoter but not the core promoter is critical to the regulation of C/EBP
expression in lung cancer.
Folate Intake and Pancreatic Cancer Risk
Experimental and epidemiologic evidence have pointed to a possible role of folate in carcinogenesis, but epidemiologic studies of the association between folate intake and pancreatic cancer have yielded conflicting results. In this issue, Larsson et al. (p. 407) report the results of a prospective cohort study of more than 80,000 Swedish men and women in which they analyzed the association between folate intake and pancreatic cancer risk. Individuals in the highest category of dietary folate intake had one-fourth the pancreatic cancer risk of those in the lowest category of intake. High total folate intake (i.e., dietary plus supplemental) was also associated with reduced risk of pancreatic cancer. However, folic acid from supplements did not show an association with pancreatic cancer.
Mesothelioma Mortality and Fibrous Zeolite Exposure
Mesothelioma incidence is high in certain villages on the Turkish Anatolian plateau where erionite, a form of zeolite fibers, occurs in the local volcanic tuff. Baris and Grandjean (p. 414) report results of a prospective study of residents of two exposed villages and one nearby control village. A total of 891 men and women aged 20 years or older were included, 230 of them residing in the control village. After a 23-year follow-up, there were 372 deaths, 119 of them from mesothelioma. Mesothelioma, in fact, was the cause of 44.5% of all deaths in the exposed villages. When standardized to world population, the incidence of pleural mesothelioma was approximately 700 and 200 cases per 100,000 people annually in the two exposed villages and approximately 10 cases per 100,000 people in the control village. The authors conclude that their results emphasize the severity of the mesothelioma endemic in erionite-exposed areas of Turkey.
Influenza Epidemics and Childhood Leukemia
Childhood leukemias may vary in etiology. Some forms have been suggested to be associated with infection or lack of immunity. To investigate this relationship, Kroll et al. (p. 417) examined time trends in leukemia incidence between 1974 and 2000 using data from a registry of childhood tumors for all of Britain. The incidence of acute lymphoblastic leukemia (ALL), particularly the precursor B-cell subtype, gradually increased whereas that of other forms of leukemia did not. In addition, the authors observed small peaks in incidence of ALL in 1976 and 1990. Both peaks occurred during the year immediately following a winter influenza epidemic, as indicated by general practitioner consultation rates. The authors note that this coincidence could be due to chance or other factors but that if there were a real association it would be consistent with hypotheses that some cases of childhood leukemia are triggered by infection, especially in conditions of low herd immunity.
![]()
CiteULike
Connotea
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||