Menopause and HRT – the state of the art in Europe

Menopause and HRT – the state of the art in Europe

Maturitas 51 (2005) 40–47 Menopause and HRT – the state of the art in Europe David H. Barlow ∗ Wolfson Medical School Building, University of Glasgow...

72KB Sizes 1 Downloads 92 Views

Maturitas 51 (2005) 40–47

Menopause and HRT – the state of the art in Europe David H. Barlow ∗ Wolfson Medical School Building, University of Glasgow, University Avenue, Glasgow G12 8QQ, UK Received 8 October 2004; received in revised form 4 February 2005; accepted 7 February 2005

Abstract The HRT field has been dramatically affected by the publication on major randomised controlled trials of the long-term effects of HRT. The publicity surrounding the publication of these data has affected public and regulatory perceptions of HRT and its role in healthcare, including the relatively short-term use of HRT for the relief of menopausal symptoms. An evidence-based appraisal of the role of HRT today is best achieved by considering the different components of the effects of HRT individually, based on the best trial evidence, and then considering these together in the context of the age of woman concerned and the specific components relevant to that woman’s health profile. This paper summarises the effects of HRT using this approach in the context of European practice today and describes the events surrounding the regulatory and scientific society position statements. © 2005 Published by Elsevier Ireland Ltd. Keywords: Menopause; Hormone therapy; Europe; State of the art; Osteoporosis

1. Introduction The basics of the menopause and hormone replacement therapy (HRT) have been established over several decades yet we have seen the field appear to be turned upside down in the past half dozen years. This has occurred as a result of the impact of landmark trials carried out in the United States which have overturned some of the established ideas about the effects of HRT in postmenopausal women, have attracted considerable media attention and have affected the regulatory framework of HRT use in Europe as well as in the US [1–4]. ∗

Tel.: +44 141 330 3362; fax: +44 141 330 3360. E-mail address: [email protected].

0378-5122/$ – see front matter © 2005 Published by Elsevier Ireland Ltd. doi:10.1016/j.maturitas.2005.02.013

This was able to happen with such an established therapy because what these recent trials were addressing were effects of HRT that were assumed from available data but had not been tested in randomised controlled trials (RCTs). Progressively over the past 15 years, the world of medical science has accepted the principle that the highest level of evidence for the effectiveness of a drug is provided by RCTs which report true disease event end-points and that other forms of evidence are less secure. Thus it has become established that RCTs with surrogate end-points, observational studies and mechanistic studies all provide evidence about the effects of drugs but represent a lower level of “proof”. However, we would generally expect that all of these forms of evidence should point in the

D.H. Barlow / Maturitas 51 (2005) 40–47

same direction. HRT has been in use for decades so how can it be, especially if you take your opinions from the media that recent RCTs can appear to have rewritten the story? It is critical to an understanding of what has been happening to remember that postmenopausal estrogen is used to address several distinct clinical areas and the quality of the evidence on each of these differs. However in the public perception and in the media attention the information about HRT is compressed into relatively simple messages which collapse the different dimensions of HRT action into headline statements that summarise HRT as “good” or “bad”. This kind of summary does no service to the place of HRT in medicine and, more importantly, does no service to the women who might benefit from the appropriate use of HRT. Another level of complexity in the discussions and arguments that have come in the wake of the publication of these major US trials is around the potentially different outcomes that might have resulted if other estrogen and progestogen products common in European practice had been used or if different, possibly lower, doses had been employed in these mostly older postmenopausal women. Similarly there have been questions around the equivalence of the US population recruited and average European populations. These points have mostly not been addressed in the popular media attention to HRT since the publication of the trials. It is likely that we shall not see future major randomised trials address these important questions which will probably remain unresolved.

41

• Universally licensed indication. The treatment of menopausal psychological dysfunction: • Ill-defined clinical syndrome most common below 60 years of age. • Some clinical end-point RCT evidence in favour of effectiveness. • Risks relatively low at age at which this would be relevant. • Generally not a licensed indication. The treatment of symptoms of urogenital ageing: • Multidimensional clinical syndrome which can be ongoing problem. • Effectiveness proved in clinical end-point RCTs. • Locally applied estrogen is effective and of low risk. • Widely licensed indication. The prevention of osteoporotic fracture: • Clinical problem of postmenopausal women but the incidence is greatly increased in the elderly. • Effectiveness had been dependent on observational studies and surrogate end-point trials but now proved by clinical end-point RCTs. • At younger postmenopausal ages, risks are low but risks of HRT higher in the elderly age groups where fracture risk is highest. • Has been licensed indication for many HRT preparations but regulatory revision has declared HRT not to be a first-line option. The prevention of coronary heart disease:

2. The different dimensions of HRT What are the different dimensions of HRT that should be considered separately if there is to be a fair picture of the effects of using HRT and a comprehensible portrayal of the evidence? A summary list should include the following. The treatment of menopausal vasomotor symptoms: • • • •

Effectiveness proved by clinical end-point RCTs. Mainly relevant to women below 60 years of age. Risks relatively low below 60 years. Many placebo-controlled RCTs demonstrating effectiveness.

• An important problem of postmenopausal women and a major killer. • Effectiveness in women with established disease suggested by observational and mechanistic evidence but disproved by clinical end-point RCT. • Effectiveness in women who have not experienced coronary disease events suggested by surrogate endpoint trials, observational and mechanistic evidence but not confirmed by RCTs. • Not a licensed indication for HRT. Ongoing discussion over whether HRT is effective if initiated when coronary arteries are healthy as suggested by surrogate end-point trials, observational

42

D.H. Barlow / Maturitas 51 (2005) 40–47

and mechanistic evidence but not yet well tested by clinical end-point RCTs. The prevention of Alzheimer’s disease: • A very important problem for society with predominance of elderly women. • Effectiveness in prevention suggested by observational and mechanistic evidence but disproved by clinical end-point RCTs carried out in women >65 years where the trials actually reported an increase in dementia. • Not a licensed indication for HRT. • Ongoing discussion over whether HRT is effective if initiated early after the menopause when there has been no neuronal deterioration as suggested by observational and mechanistic evidence but not yet tested by clinical end-point RCTs. In addition to this list of uses of HRT in postmenopausal women, there is the use of HRT in women who suffer premature loss of endogenous estrogen production as a result of a premature menopause or following bilateral oophorectomy. This is not a controversial use of HRT.

3. Transatlantic differences These are the different major scenarios that might be reasons for the use of HRT. The emphasis placed on these aspects of HRT has been different comparing European and North American practice. In European practice, the overwhelming predominance of use has been for symptom relief with the prevention of osteoporotic fracture as the other reason for prescription. In contrast, it is my understanding that there has been much greater interest in the prescription of HRT in the prevention of coronary heart disease in the United States than in Europe. Indeed, the increasing use of HRT in coronary heart disease prevention was a stimulus for the establishment of the Women’s Health Initiative (WHI) Trials that have been so important, subsequently, in the current reappraisal of HRT [3,4]. As a result of their origin, the trials were focussed on the question of cardiovascular effects and therefore the overall population addressed was across the span of years where cardiovascular effects are relevant. Thus the trials extended up to 79 years of age and the mean ages were in the

mid-60s. I am particularly interested in the subset of approximately one-third of the women in the trials who were aged between 50 and 59 years. This group could not be expected to have a significant number of coronary events but this group corresponds to the population of women that mainly use HRT in Europe. Much can be learned about the effects of HRT from an examination of the experience of these women in the WHI trials since even this subset of WHI represents the largest placebo-controlled RCT of HRT and its effects in the under 60s. In drawing conclusions about the effects of HRT from subsets of the WHI trials, it is critical that we understand that it is possible that for many research questions the subsets will be underpowered. Thus a failure to demonstrate a statistically significant difference within a subgroup should not be taken to indicate that there is, indeed, no difference. When this 50–59-yearold subgroup is so relevant to the real population that uses HRT in Europe it is unfortunate that often it is the overall findings in the WHI trials which are publicised. Whereas I believe it is critical when considering the quality of evidence for the effectiveness of HRT that the treatment context is known, this is not so relevant when considering risk for most women. With risk events the variable that does need to be considered is the age of the woman since, with most of the risks, there is an increase in absolute risk with increasing age. Thus risks such as breast cancer and stroke are most relevant in the oldest women.

4. The state of the art in Europe Against this background what has happened in European practice is that the balance of effectiveness has been revised as a result of the WHI trials and the HERS trial [1–4]. The revision of the place of HRT in clinical practice has been necessary because of the negation of the previously assumed cardiovascular benefit and the evidence concerning risk. The revision has been demanded by the position expressed by the European Medicines Evaluation Agency (EMEA) on 3 December 2003 [5]. In the development of that position a further influence, in addition to HERS and WHI, has been the observational study, the Million Women Study (MWS), which is focussed on breast cancer risk in HRT users [6]. Since the MWS is an observational study, and not an RCT, its evidence should be considered alongside

D.H. Barlow / Maturitas 51 (2005) 40–47

43

the many other observational studies which have been best summarised in the major reanalysis of the observational study data published by the Collaborative Group on Hormones and Breast Cancer [7]. The particular influence of the MWS appears to be that it has suggested higher levels of breast cancer risk associate with HRT use than had been found in previous studies. The influence of the MWS on EMEA thinking is clearly indicated in the EMEA statement since it specifically refers to MWS. With the publication of the EMEA statement, the individual European regulatory agencies that are linked to EMEA issued guidance on the same day. In the UK, this is the role of the Medicines and Healthcare Products Regulatory Agency (MHRA) and guidance for British practice was issued on the same day [8]. The press headlines were very negative but when the actual recommendations are examined the position is much less controversial that implied by the press comment. Since there has been what I perceive to be a mismatch between the actual regulatory position and the popular perception of the regulatory position I would like to review the context for HRT use in Europe at present and give my personal view, based on European recommendations, of how we can address the different HRT indications. As stated above following the publication of the WHI estrogen–progestogen trial, terminated early because the global health index was unfavourable, there was considerable media and professional attention but it was not until after the publication of the MWS that major regulatory statements were made. The EMEA statement [5] indicated the EMEA had convened an “Ad-Hoc Expert Group” to review the HRT data and to advise whether the data “raise a public health concern in relation to the safe and effective use of HRT”. This group was convened by EMEA at the request of some member States. The EMEA statement indicated that the “review was not a full scientific benefit/risk assessment as in the context of Marketing Authorisation Applications”. It was presumably a consensus process. It is valuable to list the major conclusions of the review:

• HRT prevents fractures during therapy, in women with and without risk factors for osteoporosis. • HRT prevents fractures in women with low bone density and/or established osteoporosis (based on limited data). • The HRT-mediated fracture prevention disappears within a few years after discontinuation. • There is no benefit of combined HRT on the risk of heart disease. • There is some evidence of increased risk of acute myocardial infarction during the first year of combined HRT. • There is some evidence of a link with stroke. • There is no evidence of a beneficial effect of combined HRT on cognitive function and some evidence of an increased risk of dementia after use of combined therapy. • HRT is associated with increased risk of breast cancer and endometrial cancer, especially with increased duration of use and with all types of HRT. • Progestogens in HRT reduce the endometrial cancer risk and augment the breast cancer risk. • HRT is associated with an increased risk of blood clotting (VTE) especially in the first year of use.

• HRT provides effective relief of climacteric (vasomotor) symptoms. • HRT prevents bone loss or increases bone density.

Despite the adverse headlines that were produced following this announcement, the actual statements made are reasonable although there are several pro-

The group concluded that “there is a public health concern with regards to the safe and effective us of HRT” and this outcome had been communicated to the member States for further evaluation and regulatory action. The summary advice on practice issued by EMEA itself can be presented in the following three EMEA statements: • The benefit/risk balance of HRT in the treatment of climacteric symptoms adversely affecting quality of life is favourable; the minimum effective dose and shortest duration of HRT treatment should be used. • The benefit/risk balance of HRT in healthy women without climacteric symptoms is generally unfavourable. • For the prevention of osteoporosis or osteoporotic fractures in women with risk factors or established osteoporosis, the benefit/risk balance of HRT is not favourable as first line treatment for this indication.

44

D.H. Barlow / Maturitas 51 (2005) 40–47

visos that might apply to the guidance on the role of HRT in relation to osteoporosis. The further advice to health professionals in the UK was provided through a statement from the UK regulatory agency, the MHRA, through its Committee of Safety of Medicines [8] and two accompanying documents of advice for patients [9] and professionals [10]. Its advice to professionals [10] published on the Internet emphasised the EMEA advice and gave further detail on the advice relating to osteoporosis. The MHRA wording describing the EMEA review statement was that “despite its effectiveness in preventing osteoporosis, the review concluded that for long-term use, the balance of risks and benefits is such that HRT should no longer be considered a first-line therapy for preventing osteoporosis.” Where this statement appears to clarify the EMEA statement is the use of the phrase “long-term use”. It is well recognised that osteoporosis prevention has to be a long-term intervention which will only be fully effective if it carries the woman through to the later phase of postmenopausal life where fracture is common. The statement thus appears to be considering the concept of long-term use of HRT over a long span of years into the 60s and beyond into ages where the absolute risk associated with HRT is increasing. The philosophy appears to be that the extended use of HRT, necessary for osteoporosis prevention, is associated with levels of absolute risk not seen with alternative interventions, primarily bisphosphonates, and as a result HRT in long-term use should not be the first-line approach in osteoporosis prevention. This is further clarified in the advice given in the MHRA “Advice to Professionals” Document [10]. The advice is phrased as a relatively negative statement but it is important to examine what is being excluded and what appears to be permitted. First, the statement indicates that it is referring to women over 50 years of age. This leaves the place of HRT in women under 50 years who might need osteoporosis prevention outside the negative statement. Furthermore, there is a subsequent statement devoted to women with a premature menopause (natural or surgical) where the use of HRT is accepted as appropriate. Despite this we have many women who are requiring HRT well below 50 years of age who have been very concerned about the adverse publicity. In relation to the long-term use of HRT in osteoporosis prevention, the MHRA advice is that this should

only be as a second-line approach where there are problems with bisphosphonate treatment. These problems are given as follows: • Intolerance of bisphosphonates – this is a relatively uncommon problem now that once weekly oral preparations are available. • Where bisphosphonates are contraindicated – again this is a small group. • Where there is evidence of a lack of response – most women will respond to bisphosphonates and there are no clear criteria for the definition of poor response. Certainly it would be inappropriate to use the occurrence of a fracture as indicative of poor response since this could occur on effective therapy. It is likely that this advice will be taken to mean a loss of bone density on bisphosphonates therapy but this can only be fairly estimated after an interval of a couple of years. Bone metabolic marker measurement might in future be used to estimate lack of response but its use remains outside mainstream clinical practice in the UK. The MHRA statement permits the long-term use of HRT in osteoporosis prevention where any of these situations apply and following an assessment of the individual’s risk:benefit balance. In practical terms, I would not expect many women to fall into these categories. In my opinion the main role of HRT in relation to osteoporosis prevention is in three scenarios. (1) First, in women with premature menopause. This is consistent with the MHRA advice. (2) Second, in women who are requiring HRT for the management of menopausal symptoms. These women can be reassured that whilst taking HRT for symptom relief they will be maintaining bone mass. The MHRA statement makes it clear that the benefit:risk balance is favourable for such women. Some of these women will require HRT for symptom relief over quite a few years but generally below 60 years. Whilst reassurance that their bone mass has been maintained whiles they have had symptom relief might be helpful this will not be a sufficient strategy, by itself, if the woman is at high risk of osteoporotic fracture because the intervention would need to be continued for many years beyond the time of symptom relief. However, the bone protection during HRT-mediated symp-

D.H. Barlow / Maturitas 51 (2005) 40–47

45

Table 1 Clinical events per 10,000 women per year estimated from the estrogen plus progestogen vs. placebo component of the Women’s Health Initiative Study Women 50–59 years old E+P/placebo CHD Stroke Hip fracture All fractures Colon cancer Breast cancer

Women 70–79 years old Difference

22/17 14/10 1/3 111/141 4/5 31/26

E+P/placebo

+5 +4 −2 −30 −1 +5

78/55 61/48 33/48 224/285 14/28 54/41

Difference +23 +13 −15 −61 −14 +13

A comparison of event rates in the 50–59-year and the 70–79-year subgroups [3,13–17]. Table 2 Clinical events per 10,000 women per year estimated from the estrogen-only vs. placebo component of the Women’s Health Initiative Study Women 50–59 years old E/placebo CHD Stroke Hip fracture VTE Colon cancer Breast cancer

Women 70–79 years old Difference

14/24 16/16 4/1 15/13 7/12 21/29

E/placebo

−10 0 +3 +2 −5 −8

88/84 71/57 32/52 40/28 32/15 32/34

Difference +4 +14 −20 +12 +17 −2

A comparison of event rates in the 50–59-year and the 70–79-year subgroups [4].

tom relief could be built into a longer-term strategy for osteoporosis prevention in high-risk individuals with other treatment options, principally bisphosphonates or SERMs, subsequently being used. If this approach is to be viewed as cost-effective it would certainly only apply to individuals at highrisk of osteoporotic fracture since osteoporosis interventions will only have the possibility of being cost-effective in women in their 50s if the women are at high-risk of fracture. (3) Third, there will be women in their 50s or early 60s who are at high-risk of osteoporosis and who might, at that phase of their life, wish to use HRT as an effective intervention for bone protection. The broad sweep of the MHRA statement indicates that it views HRT as having a negative benefit:risk balance if the woman has no need for symptom relief. This indicates that the MHRA view is that as a long-term strategy the use of HRT aimed at osteoporosis prevention is associated with such risk that the delivery of the intended skeletal benefit through HRT is not justifiable. Since the majority of the burden of absolute risk occurs in the lat-

ter decades of treatment I believe that a case can be made that an osteoporosis strategy might commence in the first postmenopausal decade as involving HRT, with other agents used subsequently, but I accept that the risk aspect of the balance of benefit:risk becomes more accentuated in the 60s and especially in the 70s as indicated by the WHO results. Thus I would argue that the use of HRT in this way in asymptomatic women below 60 years of age, who are at high-risk of fracture, as part of an osteoporosis strategy where other agents will then be used in later years is an acceptable approach. I do not consider that this is ruled out by the MHRA statement since I am not proposing that the woman will use HRT long-term as her anti-osteoporosis agent into ages where the absolute risk associated with HRT begins to rise. The gradient of difference between the absolute risk associated with HRT use in the first postmenopausal decade (approximately 50–59 years of age) and later years can be illustrated by comparing the absolute risk estimates published from the WHI studies for women

46

D.H. Barlow / Maturitas 51 (2005) 40–47

50–59 years old and women 70–79 years at study entry and who then had a mean of 5.2 years HRT or placebo in the combined HRT study involving more than 16,000 women (Table 1) [2] or a mean of 6.8 years HRT or placebo in the estrogen-only HRT study involving approximately 11,000 (Table 2) [4]. We are unlikely to have better risk estimates from an RCT source in the foreseeable future. This comparison indicates the relatively low absolute increase in risk due to HRT in the younger decade of the study, especially where the women had undergone hysterectomy, as is the case for about twenty percent of British women.

5. Conclusion In conclusion, I am of the opinion that following the publication of the major trials, particularly the two WHI trials there has been a deep re-appraisal of the place of HRT in postmenopausal health. An important pressure for the re-appraisal has come through the positions adopted by EMEA and in the national regulatory authorities which for my country is the MHRA. I believe that the extent of the adverse press and public concern which followed these statements has lead to health professional interpreting these statements as being more severe than is the case when the texts are examined. Indeed, the position statements published by EMAS [11], is, in fact, reasonably consistent with the statements although the perspectives of these latter statements have a more positive flavour. For example, the EMAS statement, in discussing the role of HRT in the prevention of osteoporosis emphasises that “while alternatives to HRT use are available for the prevention and treatment of osteoporosis, estrogen may still remain the best option particularly in young and/or menopausally symptomatic women.” [11]. In contrast the statement by the International Menopause Society is critical of the EMAS position stating that “IMS considers that the EMEA recommendations are unjustified and potentially harmful for the health of postmenopausal women”. The IMS view on osteoporosis being that “HRT is a most effective therapy for the prevention of osteoporosis and related fractures; indeed, the Women’s Health Initiative (WHI) confirmed the fracture reduction at both the spine and hip. No other intervention has been shown to be so cost-effective in reducing hip and spine fractures.” [12].

Where disparity especially emerges between the considerations of the menopause societies and the regulatory statements is in relation to the other potential areas where HRT has been discussed. Since no licences exist for coronary heart disease prevention or Alzheimer’s disease prevention it is not surprising that the regulatory statements do not refer to these topics since these are uses not endorsed by the regulators. In the society statements the references to these considerations are considerably more measured than would have been the case in the past but they leave open the possibility that early use of HRT may deliver such benefits; clearly areas for future research and debate. In the context of the risks that have been identified in association with HRT use, especially in older women I believe that very substantial evidence of benefit for these unrecognised indications will need to emerge before the regulatory position is revised. What I would hope might happen however is that the regulatory position will better recognise the apparently different levels of absolute risk associated with estrogen-only HRT in hysterectomised women and combined HRT used by other women. In my view this difference as revealed by the huge RCTs should by represented in a differential view within the regulatory statements.

References [1] Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. J Am Med Assoc 1998;280:605–13. [2] Grady D, Herrington D, Bittner V, et al. HERS Research Group. Cardiovascular disease outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin Replacement Study follow-up (HERS II). J Am Med Assoc 2002;288:49–57. [3] Rossouw JE, Anderson GL, Prentice RL, et al. Writing Group for the Women’s Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. J Am Med Assoc 2002;288:321–3. [4] The Women’s Health Initiative Steering Committee. Effects of conjugated equine estrogen in postmenopausal women with hysterectomy. J Am Med Assoc 2004;291:1701–12. [5] EMEA public statement on recent publications regarding hormone replacement therapy. Post-authorisation evaluation of medicines for human use. EMEA/33065/03. London: The European Agency for the Evaluation of Medicinal Products; 2003. http://www.emea.eu.int.

D.H. Barlow / Maturitas 51 (2005) 40–47 [6] Beral V, Million Women Study Collaborators. Breast cancer and hormone-replacement therapy in the Million Women Study. Lancet 2003;362:419–27. [7] Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and hormone replacement therapy: collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer. Lancet 1997;350:1047–59. [8] Duff G, Committee on Safety of Medicines. Use of hormone replacement therapy in the prevention of osteoporosis: important new information; 2003. http://medicines.mhra.gov.uk. [9] MHRA and CSM. Hormone replacement therapy (HRT): Latest safety update; 2003. http://medicines.mhra.gov.uk/ourwork/ monitorsafequalmed/safetymessages/hrtsafetyupdatedecember 2003.pdf. [10] Duff G, Committee on Safety of Medicines. Further advice on safety of HRT: Risk:benefit unfavourable for first-line use in prevention of osteoporosis. CEM/CMO/2003/19; 2003. http://medicines.mhra.gov.uk/ourwork/monitorsafequalmed/ safetymessages/hrtepinet 31203.pdf. [11] European Menopause and Andropause Society. Position statement on postmenopausal hormone therapy. http://emas. obgyn.net/content/PDF/PosPaper.pdf.

47

[12] International Menopause Society. Statement from the International Menopause Society. http://emas.obgyn.net/content/ IMS/statement IMS.htm. [13] Manson JE, Hsia J, Johnson KC, et al. Women’s Health Initiative Investigators. Estrogen plus progestin and the risk of coronary heart disease. N Engl J Med 2003;349:523–34. [14] Wassertheil-Smoller S, Hendrix SL, Limacher M, et al. WHI Investigators. Effect of estrogen plus progestin on stroke in postmenopausal women: The Women’s Health Initiative: a randomized trial. J Am Med Assoc 2003;289:2673–84. [15] Cauley JA, Robbins J, Chen Z, et al. Women’s Health Initiative Investigators. Effects of estrogen plus progestin on risk of fracture and bone mineral density: The Women’s Health Initiative randomized trial. J Am Med Assoc 2003;290:1729– 38. [16] Chlebowski RT, Hendrix SL, Langer RD, et al. WHI Investigators. Influence of estrogen plus progestin on breast cancer and mammography in healthy postmenopausal women: The Women’s Health Initiative Randomized Trial. J Am Med Assoc 2003;289:3243–53. [17] Cheblowski RT, Wactawski-Wnede J, Ritenbaugh C, et al. Estrogen plus progestin and colorectal cancer in postmenopausal women. N Eng J Med 2004;350:991–1004.