Authors' objectives

To assess the association between hormone replacement therapy (HRT) and the risk of developing or dying from colorectal cancer.

Searching

English language literature was sought from the following databases: MEDLINE (1966 to April 1998); Current Contents (July 1995 to April 1998); Cancerlit (1983 to April 1998); and HEALTHSTAR (1975 to April 1998). Reference lists of published studies were searched. Details of the search strategy were given. Unpublished studies were sought by consulting experts and manufacturers of HRT products.

Study selection

Assessment of study quality

Validity was assessed by examining selection bias, misclassification bias, and detection bias. Studies were reviewed independently by two of the authors.

Data extraction

The following data were extracted by two authors using a pretested data extraction sheet with disagreements being resolved by consensus: study design; study size; tumour site; multivariable adjusted odds ratio (OR) or relative risk (RR) with 95% confidence intervals; years of HRT use; and factors used to estimate adjusted RR. Relative risk and 95%CI were calculated by recency of HRT use, by duration of use of HRT, and by colon cancer subsite. Women who had never used HRT were used as the reference group. The OR from case control studies was used as an approximation to the RR. Where necessary, additional information was sought from the original authors.

Methods of synthesis

Results of the review

Twenty-five studies (including one case series with 735 users, one RCT with 168 women, 12 case control studies with over 14,760 women; and 11 cohort studies with at least 1,030,905 person-years of use) were included.

Studies were excluded from summary calculations if they had fewer than five colorectal cancer cases or use of an inappropriate control group.

All but one of the studies was observational. It was not possible to assess the effects of dose, type of oestrogen preparation, and use of combined HRT on colorectal cancer due to insufficient published data.

No difference was found in risks of developing colorectal cancer for ever users of HRT compared to never users (17 studies): RR = 0.89 (95%CI: 0.76, 1.05). No difference was found in risks of developing colon cancer for ever users of HRT compared to never users (15 studies): RR = 0.92 (95%CI: 0.79, 1.08). Significant heterogeneity was found. After omitting the one study accounting for most of the heterogeneity, the remaining studies were homogeneous (P = 0.11) and RR for colon cancer and ever use of HRT = 0.88 (95%CI: 0.80, 0.97). Rectal cancer use was not affected by HRT use (11 studies): RR = 0.97 (95%CI: 0.85, 1.11).

Recency of HRT use and colon cancer: The risk of developing colon cancer was less in recent HRT users than in nonusers (7 studies): RR = 0.67 (95%CI: 0.59, 0.77). No difference was found in risks of developing colon cancer for former users of HRT compared to never users (6 studies): RR = 0.93 (95%CI: 0.82, 1.06).

Duration of use of HRT and colon cancer (4 studies): No difference was found in RR of colon cancer in women who had used HRT for less than 5 years: RR = 1.03 (95%CI: 0.74, 1.44) and those who had used it for five or more years: RR = 0.89 (95%CI% 0.70, 1.14), compared to never users.

Colon cancer subsite: No difference was found in risk of right colon cancer in ever users of HRT with RR = 0.85 (95%CI: 0.65, 1.11) or in left colon cancer with RR = 0.90 (95%CI% 0.75, 1.08), compared to never users.

Fatal colon cancer was less common in ever users of HRT compared to never users (3 cohort studies): RR = 0.72 (95%CI: 0.64, 0.81).

Authors' conclusions

The risk of colon cancer may be decreased amongst recent post menopausal HRT users. Although the data is limited, the risk of fatal colon cancer may be lower in HRT users.

CRD commentary

The aims and inclusion criteria were stated. Full details of the literature search were given. Published and unpublished data were sought. Some relevant details of the included studies and results were clearly presented. Validity and statistical heterogeneity were assessed. Methods used to assess validity and extract data were described. Sensitivity analysis was conducted and reasons for heterogeneity postulated. The discussion included consideration of the following: limitation of the review due to evidence being largely based on observational studies; potential sources of bias including publication bias, misclassification bias, recall bias, selection bias, and detection bias.

By limiting the literature search to English language studies some other relevant studies may have been omitted. Details of methods used to select control for case control studies and details of methods used to select subjects for cohort studies would have been welcome as would a more comprehensive description of patient characteristics and the type of HRT used.

The authors cautious conclusion is supported by the evidence, though it must be noted that it was not possible to assess the association of combined HRT with colorectal cancer due to insufficient published data and this should perhaps have been mentioned in the conclusion.

Implications of the review for practice and research

Practice: The authors suggest that the findings of the review indicate that protection from colon cancer could be included in information discussed when counselling women about HRT.

Research: The authors mention an ongoing randomised controlled trial of HRT being conducted by the Women's Health Initiative which will examine colon cancer as a primary outcome. Results from this study are expected in 2006.

Bibliographic details

Nanda K, Bastian L A, Hasselblad V, Simel D L. Hormone replacement therapy and the risk of colorectal cancer: a meta-analysis. Obstetrics and Gynecology 1999; 93(5 Part 2 Supplement S): 880-888. [PubMed: 10912438]

Indexing Status

Subject indexing assigned by NLM

MeSH

Colorectal Neoplasms /mortality /prevention & control; Female; Hormone Replacement Therapy; Humans; Incidence; Risk Factors; United States /epidemiology; Women's Health

AccessionNumber

11999001021

Database entry date

30/09/2000

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.