Abstract
“The findings of an adverse effect of estrogen on prognosis are supported by a robust body of recent preclinical data demonstrating that estrogen can stimulate NSCLC growth and that blocking estrogen can inhibit tumor growth in xenograft models.”
Reflecting the increasing rate of tobacco use in women after World War II, the age-adjusted rate of lung cancer incidence in women has climbed from less than 20 per 100,000 in 1973 to approximately 40 per 100,000 by 1999 [1]. Currently, approximately 20% of women in the USA smoke, with little evidence that the frequency will decline due to antismoking campaigns [2]. The smoking prevalence is highest in younger women, especially those from economically and educationally deprived backgrounds. Although the rate of lung cancer deaths in men has started to decline, it only appears to have slowed its rate of increase in women, who now account for more than 40% of lung cancer deaths [3]. In the USA, lung cancer-related deaths in women now exceed those from breast or colon cancer combined [4].
Women may be more susceptible to the carcinogenic effects of tobacco smoke than men. This has been demonstrated epidemiologically, as well as in laboratory and clinical studies. In a study in 800 Canadian women, the association of smoking and nonsmall-cell lung cancer (NSCLC) was significantly stronger for females than for males [5]. In subjects with a history of 40 pack-years, compared with lifelong nonsmoking, the odds ratio for women to develop lung cancer was 27.9 (95% confidence interval [CI]: 14.9–52.0) and for men was 9.60 (95% CI: 5.64–16.3). Higher odds ratios for females were also seen within each of the major histological groupings. Thus, the elevated risk of lung cancer currently observed in other studies for female ever-smokers compared with male ever-smokers may be due to higher susceptibility among females. A case–control study of 4000 patients and control subjects was conducted in the USA [6]. The authors found that the “dose–response overall risks over cumulative exposure to cigarette smoking were 1.2–1.7-fold higher in women than in men for the three major histologic types”. They concluded that “these results confirm that overall risks for major lung cancer types are consistently higher for women than for men at every level of exposure to cigarette smoke. Furthermore, this gender difference cannot be explained by differences in baseline exposure, smoking history or body size, but it is likely due to the higher susceptibility to tobacco carcinogens in women.” Henschke and Miettinene reviewed a large series of at-risk individuals screened by computerized tomography and found that for individuals with at least 10 pack-years of exposure, the relative risk for women was 2.7 [7]. However, other epidemiological studies failed to confirm the increased risk [8,9].
In addition to potential sex differences in relative risk for developing lung cancer, there are major biological differences in the disease between men and women. Female smokers and nonsmokers are more likely to develop adenocarcinomas than squamous cell carcinomas, which are more common in men. Women develop cancer at a younger age than men [3,10]. Among never-smokers who develop lung cancer, women are over-represented by a factor of 2.5 [11]. This is seen most dramatically in Asian women in whom never-smokers account for 70% of female lung cancer. Several molecular and genetic differences between men and women have been suggested as explanations for these differences [12].
One such hypothesis suggests that sex differences in susceptibility to NSCLC may be due to the effect of estrogens on pulmonary physiology and carcinogen metabolism [13,14]. Estrogens, through their interactions with estrogen receptors, play an important role in early lung development in both sexes [15,16]. Cells derived from normal lung tissue and nonsmall-cell lung tumors express both estrogen receptor-α and -β and show biological responses to estrogen [17]. An epidemiological link between duration of estrogen exposure and incidence of NSCLC has been suggested from large cohort studies in Japanese nonsmoking women. Women who had early age of menarche or late menopause had a twofold risk of developing NSCLC [18]. Use of hormone replacement therapy (HRT) in women with artificial menopause compared with hormone replacement alone or natural menopause conferred a relative risk of 2.4 for developing lung cancer and 2.7 for adenocarcinoma [19]. In a recent retrospective analysis of data from a community hospital in the USA of HRT and prognosis of women with NSCLC, those who used HRT had a substantially shorter survival than those who did not (hazard ratio: 1.97; 95% CI: 1.14–3.39) [20].
Survival for stage-matched women with NSCLC appears somewhat better than for men overall [21–23]. However, in retrospective analyses of cooperative group data, younger, presumably premenopausal women had shorter survival than older women. In an analysis of patients with advanced NSCLC enrolled in Southwest Oncology Group trials, women aged over 70 years had a 34% 1-year survival compared with 11% for those aged under 45 years [21]. A subanalysis of women enrolled in two recent large Phase III trials (Selective Targeting for Efficacy in Lung cancer, Lower Adverse Reactions [STELLAR]− 3 and −4) in frontline therapy for patients with advanced NSCLC who were performance status 2, demonstrated a highly significant decrease in survival for women in the control arms who received standard therapy and were aged under 55 years. This was also found in women who had pretreatment estradiol levels of greater than 30 pg/ml versus those with lower levels [24]. In these trials the median survival was 181 days in premenopausal women compared with 263 days in postmenopausal women. In a recent analysis of Surveillance Epidemiology and End Results data involving all stages of NSCLC, premenopausal women, particularly those with squamous cell cancer and bronchoalveolar cancer, were found to have a worse survival than postmenopausal women [25]. Overall, these data suggest that advanced NSCLC is more aggressive in premenopausal women (or in women receiving HRT) than in older women [26].
The findings of an adverse effect of estrogen on prognosis are supported by a robust body of recent preclinical data demonstrating that estrogen can stimulate NSCLC growth and that blocking estrogen can inhibit tumor growth in xenograft models [17,27]. Given the dismal results with conventional therapy in premenopausal women from retrospective analyses of cooperative study group data and the short survival observed in the control arms of STELLAR-3 and −4 for premenopausal women, an urgent need for new and effective therapies exists in this patient population.
Owing to differences in the biology of NSCLC between men and women, it is recommended that all large trials stratify by sex in addition to the standard factors. Moreover, attention should be given to defining the hormonal status of newly diagnosed patients with NSCLC, perhaps by measuring pretreatment estradiol levels and other relevant parameters. Given the relatively high levels of estradiol seen in some men, estradiol may be of clinical importance in both sexes. Gender-specific prospective studies should also be considered as certain drugs such as paclitaxel poliglumex (Xyotax™) may have increased efficacy in women, and particularly in premenopausal women. If, as current retrospective data suggest, higher levels of estrogen are highly associated with an adverse outcome in advanced NSCLC, it will be of great interest to evaluate anti-estrogenic therapies in conjunction with standard and targeted therapeutics in both men and women.