Get ready to throw out your Premarin (again): Thoughts on hormones and lung cancer

June 23, 2009 - 10:00 am email to a friend

Print This Post view / write comments

(1 votes, average: 5.00 out of 5)

Loading ...

Dr. Pinder

The role of hormones in the development and progression of lung cancer in women has generated much interest. Unfortunately, a lot of the data to date has been observational, which doesn’t establish a “cause and effect” relationship. The Nurses Health Study (more on this below) is a good example: a large cohort of women was observed over time. The women completed questionaires on all sorts of exposures (diet, hormone replacement therapy, tobacco, etc), and they were followed over time. Then, investigators tried to sort out whether there were differences in exposures between women who got a given disease and those who didn’t. These types of studies can be “hypothesis-generating” but rarely yield clear results. It can be very difficult to isolate one exposure amongst many other confounding variables.

At ASCO this year, the Women’s Health Initiative (WHI) Study investigators presented important new findings from this landmark study of hormone replacement therapy (HRT). Most women are familiar with the earlier reported findings of this study because those findings led physicians to STOP recommending HRT for post-menopausal women (previously considered beneficial). Most may not be familiar with the details of the study design so here is a refresher. This study enrolled healthy post-menopausal women with no history of breast cancer. Over 16,000 women were enrolled: half were randomized to HRT with estrogen plus progesterone, the other half received placebo. The investigators were primarily interested in the risk of breast cancer and cardiovascular disease. Other cancers specifically studied included endometrial cancer and colorectal cancer. The study was stopped early, because the investigators found significantly more risks than benefits associated with HRT compared to placebo. Specifically, women who received HRT were 30% more likely to develop cardiovascular disease, 26% more likely to be diagnosed with breast cancer, and had a 40% increased risk of stroke.

Though the study stopped early, the women continued to be followed. In a recent update, the investigators noted an increase risk of malignancy (other than the three previously studied cancers) in women who took HRT. They decided to analyze the risk of lung cancer in the two cohorts of women.

The women who received HRT did not differ from those who received placebo in smoking status; about 50% were never-smokers, 40% former smokers and 10% were smoking when they entered the study. Women who received HRT did not appear to be at increased risk of small cell lung cancer. However, the risk of non-small cell lung cancer was 28% higher for those who received HRT (though the p-value was not significant at 0.12). Furthermore, women who received HRT and developed NSCLC were more likely to die of their lung cancer than those who were on placebo. It also appeared that women on HRT were more likely to have metastatic disease and poorly differentiated tumors. When they looked at smoking status, it appeared that the greatest risk was for women who took HRT and were current smokers. The authors estimated that 1/100 smoking women who received HRT died unnecessarily from lung cancer. While the risk was greatest for smokers, never smokers and former smokers also appeared to be at increased risk of NSCLC if they were assigned to HRT.

Results presented from the Nurses Health Study were a little less clear-cut. This study included a large number of women (over 107,000) who reported on their exposure to tobacco smoke (both first and secondhand), reproductive factors, oral contraceptive (OCP) and HRT use, diet, etcetera. They were followed over time and ~1700 were eventually diagnosed with lung cancer. Younger age at menopause, older age at first birth, longer OCP use and fewer number of children were associated with an increased risk of lung cancer. When broken down by smoking status, however, it appeared that these findings applied mainly to current smokers. In never smokers, lung cancer risk decreased with increasing number of children. Overall, no significant association between HRT use and lung cancer risk but when broken down by histology, there was a trend for increased risk of adenocarcinoma in HRT users and decreased risk of small cell lung cancer in HRT users.

How can we interpret and use the findings from these two studies? Well, I think the WHI data confirm that long-term HRT is harmful for post-menopausal women. We can add lung cancer to the list of negative outcomes with HRT. It is helpful to have a randomized trial since observational studies of HRT in lung cancer have yielded mixed results - some suggested it was helpful, others harmful. Given that we also had mixed results for other outcomes prior to definitive data from the WHI study, I am inclined to go with the results from WHI. For lung cancer, the increased risk does not become apparent (the curves do not separate) until after 3 year of HRT. The same was true for breast cancer in the WHI study. Therefore, many physicians try to utilize the shortest duration of HRT possible to help women through the most intense period of menopausal symptoms (six months is often accepted as a reasonable timeframe). It’s difficult to interpret whether these findings apply to young women who go through early menopause (from chemotherapy or hysterectomy) as the women in this study were all over 50.

The Nurses Health Study highlights the complexity of the relationship between hormonal factors, smoking and lung cancer and I don’t know how much the findings help us. For example, the investigators found an apparent “protective effect” of HRT on the risk of small cell lung cancer whereas the investigators in the WHI study saw no impact of HRT on small cell lung cancer. Most of the trends seen in the Nurses Health Study were not statistically significant. I think I’d be more inclined to give weight to a finding from an observational study if the signal were really strong, but these results were not.

With both of these studies, it is difficult to know whether estrogen is to blame or whether other hormones or molecular pathways stimulated/inhibited by estrogen may be at work. And while estrogen is certainly an attractive therapeutic target (given great successes in breast cancer), I think that it is probably going to prove a much more minor player in lung cancer than in breast cancer.

Bookmark to: