Post-hoc analysis of the WHI trial report that in post-menopausal women conjugated equine estrogen plus medroxyprogesterone acetate did not increase the incidence of lung cancer but increased mortality from lung cancer.

The Lancet, Early Online Publication, 20 September 2009 doi:10.1016/S0140-6736(09)61526-9
http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(09)61526-9/abstract

 

Commentary on Lancet paper by Professor Alastair MacLennan

It is important that all long term effects of HRT are properly researched and this finding should be flagged for further study but not for a current change in policy because of the following caveats and concerns about the methodology and its mostly non-significant results in the women we treat.

These results are yet more post hoc unplanned analyses of selected WHI data.  WHI was not designed to study lung cancer.

There have literally been hundreds of unplanned post hoc analyses of the WHI data making it inevitable that there will be many chance associations at a low level of probability (p < .05).  Much stronger levels of probability are required in such multiple analyses. Despite this, the main finding of the Lancet paper is NOT significant even at the above  low cut-off  level of probability.

The non-significant increase in lung cancer overall had a HR of 1.23 (CI 0.92 - 1.63) and p= 0.16.The main group of lung cancers (non -small-cell cancers) also did NOT have a significant increase (HR 1.28, CI 0.94 - 1.73).

The actual numbers of lung cancers were relatively small from a statistical analyses point of view i.e.109 in the combined HRT group and 85 in the placebo group. Only in further unplanned post hoc sub analyses of this part of the WHI data do levels of significance appear if they are not corrected for multiple tertiary analyses i.e.HR for mortality was 1.71, CI 1.16 - 2.52 , (n = 73 versus 40).

Smokers have an enormously increased risk of lung cancer, HR of 20 - 40 in most studies, and this is the major confounding in this study. There were only 9 versus 4 deaths from lung cancer in non-smokers. These numbers are too small to make  any valid conclusions on HRT alone.

Passive smoking and many other known environmental risk factors for lung cancer have not been assessed in the analyses.

The WHI study started HRT in an atypical group of mostly asymptomatic women on average 13 years post menopause. This almost never happens in clinical practice. In practice, 99% of women start HRT near menopause for symptom control. In the Lancet paper there is NO increase in lung cancer deaths in the age group 50-59!

This sub-analysis considered only the combined oestrogen and progestogen HRT regimen used in WHI and not the oestrogen-only regimen.

The whole woman and her individual needs should, of course, be considered. Death from all causes in women starting  HRT before age 60 compared to placebo in the systematic review of 30 randomised controlled trials by Salpeter et al, 2004 showed a statistically significant reduction in overall mortality of 39% (HR 0.61, CI 0.39 - 0.95). 
Frightening women who need HRT off this therapy may well "hammer another nail in their coffin" to paraphrase the scientifically disappointing and clinically unhelpful view expressed in the accompanying Lancet editorial.

The message in this study is don't smoke and start HRT if you need it for symptom control from near menopause.

PRESS STATEMENT from the International Menopause Society

 

September 20, 2009

Menopause clinicians reply to Lancet paper on HRT and lung cancer

The Lancet has issued a press release on HRT use and lung cancer (Oestrogen plus progestin and lung cancer in postmenopausal women (Women's Health Initiative trial): a post-hoc analysis of a randomised controlled trial, by Chlebowski RT, et al.).

 

The IMS has issued the following statement, which was written by Professor Amos Pines on behalf of the IMS.

 

The new analysis by Chlebowski et al. on the incidence of lung cancer in the estrogen + progestin arm of the WHI study demonstrated a slight (non-statistically significant) increase in lung cancer cases (3 additional cases/10,000 women) in the hormone users. In addition, the paper shows an increase in lung cancer deaths in older women taking HRT, and recommends that:

"...women at high risk of lung cancer from smoking and their physicians should be made aware of this additional hazard if the initiation or continuation of combined hormone therapy is being considered."

However, the paper finds that that, in the age group 50-59 years, the numbers of deaths from lung cancers were almost identical for the HRT and placebo groups (figure 3 in the paper).

The International Menopause Society agrees with the WHI that older women who are at high risk should take care in starting HRT. However, our advice is that HRT can be prescribed without any increased risk of lung cancer death in the 50-59-year age group. This is exactly what the Chlebowski paper says, although it is unclear to us why this has not been brought out in the comment or press release, and why it is not highlighted in the paper. The relevance of age and years since menopause at the initiation of HT is recognized as a major determinant of the risks associated with HT and, therefore, it is imperative to address this issue in any WHI database analysis.

Note that the vast majority of HRT is prescribed to women in the 50-59-year age group and so, even by the terms of this paper, it is safe for the vast majority of women in this group who need to take HRT for the relief of menopausal symptoms.
The comment that follows the article, written by Dr Ganti, includes the conclusion that "These results ... seriously question whether hormone replacement therapy has any role in medicine today". Such a phrasing is misleading, biased and is not substantiated by the WHI data for women aged 50-59 years, or by results from other large-scale studies. Although care should be taken when evaluating the need for HRT in each and every woman, and the individual benefits versus potential risks should always be discussed, HRT remains the most effective therapy for vasomotor symptoms and can be prescribed without any concern in healthy, young postmenopausal women.

The advice of the International Menopause Society is that HRT is safe for healthy young women in the 50-59-year age group, but that older women should be dealt with individually and take greater care. The findings of this paper support this advice. Like all clinicians, we would urge women of any age not to smoke.

 

Reference

Chlebowski RT, Schwartz AG, Wakelee H, et al. for the Women's Health Initiative Investigators. Oestrogen plus progestin and lung cancer in postmenopausal women (Women's Health Initiative trial): a post-hoc analysis of a randomised controlled trial. Lancet 2009; September 20, 2009. E-pub ahead of print

www.imsociety.org/pdf_files/comments_and_press_statements/ims_press_statement_20_09_09.pdf

 

Content Updated 21 September 2009