The fourth and last chapter in a series of articles that address methodological issues related to major studies on breast cancer risk in HRT users was published 2 weeks ago. The chapter, written by Shapiro, Farmer, Stevenson, Burger and Mueck [1], focuses on the Million Women Study, while the previous chapters analyzed the Women’s Health Initiative (WHI) estrogen + progestogen and estrogen-alone arms and the Collaborative Analysis. Here is the abstract of the manuscript:

Background: Based principally on findings in three studies, the collaborative reanalysis (CR), the Women’s Health Initiative (WHI) and the Million Women Study (MWS), it is claimed that hormone replacement therapy (HRT) with estrogen plus progestogen (E+P) is now an established cause of breast cancer; the CR and MWS investigators claim that unopposed estrogen therapy (ET) also increases the risk, but to a lesser degree than does E+P. The authors have previously reviewed the findings in the CR and WHI (Parts 1–3). 
Objective: To evaluate the evidence for causality in the MWS. 
Methods: Using generally accepted causal criteria, in this article (Part 4) the authors evaluate the findings in the MWS for E+P and for ET. 
Results: Despite the massive size of the MWS, the findings for E+P and for ET did not adequately satisfy the criteria of time order, information bias, detection bias, confounding, statistical stability and strength of association, duration-response, internal consistency, external consistency or biological plausibility. Had detection bias resulted in the identification in women aged 50–55 years of 0.3 additional cases of breast cancer in ET users per 1000 per year, or 1.2 in E+P users, it would have nullified the apparent risks reported. 
Conclusion: HRT may or may not increase the risk of breast cancer, but the MWS did not establish that it does.

The International Menopause Society issued a press statement concomitantly with the lifting of embargo from this article. 

President-Elect of the International Menopause Society, Professor Rod Baber (Sydney) said: ‘This review of the Million Women Study (MWS) highlights the problems associated with any observational study regardless of size. Potential biases inherent in any observational study make it difficult for such studies to accurately comment on relative risks of less than 2 and this paper identifies them in a clear concise manner. In the MWS, identified biases including time order, information and detection bias, confounding, internal and external consistency may all have contributed to the excess risk of breast cancer with HRT use found in this study. HRT may contribute to increased breast cancer risk but the evidence from this study is flawed and does not assist in the decision-making process. Every woman has individual risk factors for breast cancer, a number of which are modifiable and she should discuss these with her physician when considering whether or not to use hormone replacement therapy so that an accurate assessment of risks and benefits may be made.’

International Menopause Society Board member, Professor Anne Gompel (Paris) said: ‘This paper is a meaningful analysis of the MWS by an expert in epidemiology, and puts forward the fact that the numbers of patients included in a study does not guarantee the accuracy of results. I would advise everyone to read this paper closely, since it highlights some of the important requirements of a good-quality epidemiological study. For example, Professor Shapiro shows that the real relative risk is actually difficult to evaluate when it is less than 2. There are other studies which do show risks and benefits for HRT, so I believe that the take-home message is that, for a given woman, the important evaluation is her own risk. Recent papers on breast density, alcohol, obesity and exercise suggest that taking all these factors into account should help to identify which patients are at risk. HRT does not carry the same risk and benefit for each woman; some women will have increased risks, some will have only benefits, and this also applies to breast cancer. ‘

Tom Parkhill, IMS Press Officer, informed us that the press release was well taken by the media, and that the article itself received a lot of attention and was brought as an item in hundreds of websites in various languages. The typical headings were ‘Research linking HRT therapy to cancer is riddled with flaws’, ‘Study faults research linking hormone therapy to cancer’, ‘Can a study of a million women be wrong? Hormone replacement therapy and breast cancer link questioned.’ Links to some examples of the coverage are given below.

Click here for a popular, but obviously influential, press article in the NY Daily News. 

A good review from the NHS can be found by clicking here.

A further one from Brazil can be read here.

Reference
1. Shapiro S, Farmer RDT, Stevenson JC, Burger HG, Mueck AO. Does hormone replacement therapy cause breast cancer? An application of causal principles to three studies. Part 4. The Million Women Study. J Fam Plann Reprod Health Care 2011. doi: 1136/jfprhc-2011-100229