An incentive to start hormone replacement: the effect of postmenopausal hormone replacement therapy on the risk of colorectal cancer.

Abstract

PURPOSE: To analyze the evidence available on risk reduction of colorectal cancer with postmenopausal hormone replacement therapy (HRT). BACKGROUND: Colorectal cancer is a major public health problem in the United States. Approximately 135,000 new cases will be diagnosed in 2001, and 56,000 will die of this disease.1 It accounts for 15% of all cancers in men and women and, after lung cancer, is the second leading cause of cancer deaths overall and the first in women aged 75 and older.1 Men and women have a 6% lifetime chance of developing a colorectal cancer; the lifetime risk of dying is 2.6%.2 The incidence continues to rise throughout life, approximately doubling with each decade over the age of 50. The incidence is four to five times higher in people aged 65 and older; for women this ratio is higher for colorectal cancers than it is for cancer in general. More than two-thirds of colorectal cancer occurs in people aged 65 and older.1 Overall 5-year survival is 57% and is somewhat worse in older people, even after adjusting for increases in other causes of death in older people.2 There are many well-identified risk factors cited in the causation of colon cancer, and age is a dominant one for sporadic cases.3 Many protective factors, such as diet high in fruits and vegetables,4–6 folic acid supplements,7 calcium supplements,8 regular use of aspirin or nonsteroidal antiinflammatory drugs,9,10 physical activity,11 hydroxy methyl glutaryl coenzyme A reductase inhibitors, and hormone therapy, are being identified. There is now accumulating epidemiological and biological evidence12–18 suggesting an inverse relationship between hormone use and incidence of colorectal cancers. A careful appraisal of a meta-analysis of the data available from many trials investigating this relationship is presented. DATA SOURCES: All English-language articles with quantitative data regarding the effect of postmenopausal HRT on the risk of colorectal cancer in women were sought by a search of Medline databases (from January 1966 to September 1998). Articles were identified using medical-subject-heading terms “colorectal, colon, rectal cancer, estrogen, and estrogen replacement therapy” or the text word terms “hormone replacement therapy, post menopausal hormones, and noncontraceptive hormones.” STUDY SELECTION CRITERIA: Two unblinded reviewers initially reviewed 28 articles to determine eligibility. For the primary analysis, 10 studies were excluded for the following reasons: no clear distinction between oral contraceptive pills (OCP) and postmenopausal therapy in two, definition of use was not clear in one, preliminary analyses were superseded by recent publications with same study samples in six, and no quantitative data were available in one. For the secondary analysis, five studies were excluded because premenopausal women were included in reference category (hormone status of premenopausal is same as postmenopausal on HRT), and comparison would obscure the beneficial effect. DATA EXTRACTION: Case control and prospective study designs were used. Relative risks (RRs), odds ratios, and 95% confidence intervals (CIs) were abstracted from all the articles reviewed and adjusted for confounding factors when available. RRs were weighted according to the precision of study, thereby giving larger studies with more precise estimates and narrower confidence intervals greater weights. Summary RR estimates and 95% CIs were calculated using the general variance method based on fixed effects model. Separate analyses were performed within categories of hormone use (ever use, current use, and duration of current use) and cancer type (colon, rectal, and colorectal).