Prevention Strategies in Endometrial Carcinoma

Whilst the majority of endometrial cancer is sporadic, at least 3–5% of new diagnoses are made in women with a familial predisposition. The most common cause of this is Lynch syndrome (previously known as hereditary non-polyposis colorectal cancer, HNPCC), which is a highly penetrant autosomal dominant inherited predisposition to cancers of the endometrium, bowel, ovary and prostate, amongst others. In women with Lynch syndrome, the lifetime risk of endometrial cancer may be as high as 70%, compared with a lifetime risk of 2–3% in the general population, and often precedes a colorectal cancer diagnosis by approximately 10 years [6]. Identification of families with Lynch syndrome allows quantification of cancer risk and access to cancer surveillance programmes and could prevent cancers in other family members (Table 1) [7].

Risk-reducing surgery, in the form of hysterectomy and bilateral salpingo-oophorectomy when family complete, is well established in the primary prevention of Lynch syndrome–associated endometrial and ovarian cancers (women with Lynch syndrome have a 10–12% lifetime risk of ovarian cancer) [8, 9]. No evidence of a reduction in mortality has been demonstrated with gynaecological cancer surveillance in Lynch syndrome [10, 11]. Whilst aspirin use has demonstrated a reduction in risk for colorectal cancer, the same may not be true for endometrial cancer and more research is needed. In their primary prevention trial, Burn et al. found an unadjusted non-statistically significant reduction in endometrial cancer risk with 600 mg aspirin once daily for up to 4 years, although it was not sufficiently powered for endometrial cancer endpoints [2].

Biomarker studies have shown that combined oral contraceptives and synthetic progestin reduce endometrial proliferation, in women with Lynch syndrome. However, it is unknown whether hormonal contraceptives are as effective for chemoprevention in Lynch syndrome as they are in the general population [12].

Tamoxifen is a selective oestrogen receptor modulator (SERM) used for the prevention and adjuvant treatment of breast cancer. It has been associated with increased endometrial polyps and hyperplasia, and an increased relative risk of endometrial cancer in postmenopausal women of 4.01 (95% CI 1.7–10.9). There is no endometrial surveillance programme in place for tamoxifen users, but women and clinicians should be aware of the risks [13], and investigation of vaginal bleeding should include hysteroscopy as well as endometrial biopsy and ultrasound scan, as sensitivity and specificity of transvaginal ultrasound are low in this group. The levonorgestrel intrauterine system (LNG-IUS) has been shown to reduce the development of endometrial polyps in tamoxifen users but no convincing effect on endometrial cancer risk has yet been demonstrated [14].

PCOS affects 6–8% of women of reproductive age. Women with PCOS are overrepresented in young endometrial cancer patients, and a meta-analysis has estimated the lifetime risk of endometrial cancer in women with PCOS to be in the region of 9% (odds ratio (OR) 2.89), the increase likely driven by a toxic combination of insulin resistance and anovulatory cycles leading to oligo or amenorrhoea. For this reason, a crucial aspect of PCOS management is induction of regular withdrawal bleeds, in an attempt to reduce the risk of developing endometrial hyperplasia and cancer.

Whilst diet and weight reduction have been shown to improve the clinical manifestations of PCOS such as insulin resistance and reproductive manifestations, there is no evidence it reduces the risk of developing endometrial cancer [15]. Based on the presumption that insulin resistance is a key player in the pathogenesis of PCOS, metformin is being increasingly explored to ameliorate symptoms such as infertility and oligomenorrhoea. In vitro metformin has been shown to exert anti-tumour effects, but despite the promise of non-randomised clinical trials [16], there is no RCT evidence to support a role of metformin in endometrial cancer prevention in women with PCOS [17].

The World Cancer Research Fund has concluded that the risk of endometrial cancer is reduced by moderate physical activity and maintaining a healthy weight [18]. Whilst obesity is a well-established risk factor for multiple types of cancer, the association is seen most strongly in endometrial cancer. Excess weight has been shown to have a non-linear, dose-dependent relationship with endometrial cancer risk [19, 20]. Every 5-kg/m² increase in BMI increases the relative risk (RR) by 1.59, and at a BMI > 42 kg/m², RR 9.11 (95% CI 7.26–11.51). Risk is higher in women who have never used hormone replacement therapy (HRT) (RR 20.7; 95% CI 8.28–51.84) [21].

Lifestyle modification can result in 4–6% weight reduction over 2 to 4 years, and anti-obesity drugs can lead to 7–10% weight reduction; however, only bariatric surgery produces significant and, crucially, durable results [22, 23]. Bariatric surgery has been shown to reduce cancer risk in women, particularly postmenopausal breast and endometrial cancers. A retrospective cohort study of more than 100,000 bariatric surgery patients in the USA estimated a 77–81% reduction in endometrial cancer risk associated with bariatric surgery (RR 0.29, 95% CI 0.26–0.32), and an even lower risk in women who attained and maintained normal weight post-bariatric surgery compared with those who remained overweight or obese (RR 0.19, 95% CI 0.17–0.22 compared with RR 0.48, 95% CI 0.43–0.55, respectively) [24].

Over 18-year median follow-up, the prospective, non-randomised SOS Study has demonstrated that bariatric surgery significantly reduces the risk of endometrial cancer (hazard ratio (HR) 0.56, 95% CI 0.35–0.89, p = 0.014) in over 2800 women who underwent surgical or conventional weight management treatment. Other cancers were also reduced after bariatric surgery and the number needed to treat (NNT) to prevent one cancer over 10 years with bariatric surgery was 31 [25•].

Studies continue to show that as well as excess body weight being a risk factor for endometrial cancer, so too is adult weight gain, weight cycling, the duration of overweight/obesity and possibly being overweight in childhood/at age 18 [26, 27•, 28‐30]. In the NOWAC study, a 5-kg weight gain was significantly associated with increased risk of postmenopausal endometrial cancer, with a dose–response relationship with increasing weight gain, emphasising the importance of maintaining a stable weight in “middle adulthood” [28]. Weight gain and weight cycling have also been shown to be significantly associated with the development of postmenopausal endometrial cancer in the Women’s Health Initiative Observational Study, of 80,943 women including 788 cases of endometrial cancer over 20 years of follow-up [27•]. Weight cycling four to six times increased risk of endometrial cancer by 38% compared with weight-stable women.

There is now convincing evidence that women with higher physical activity levels have a lower risk of endometrial cancer than women with the lowest activity levels, in particular overweight or obese women with high activity levels (BMI < 25 RR 0.97; 95% CI 0.84–1.13 cf. BMI ≥ 25 RR 0.69; 95% CI 0.52–0.91) [31, 32•]. Studies of physical activity are limited by the activity being self-reported and more difficult to ascribe a quantitative value to, and heavily intertwined with, other confounding risk factors such as adiposity.

Arthur et al. modelled a Healthy Lifestyle Index (HLI) in 3185 women and reported that for every unit increase in HLI, there was a 5% reduction in endometrial cancer risk. When they directly compared the low vs. high HLI groups, there was a 46% reduction in risk [33•].

Diabetes may increase the risk of endometrial cancer; however, meta-analyses are frequently plagued by the confounders of inactivity and obesity, as alternative or additional endometrial cancer risk factors. After multivariate adjustment, studies have estimated the increase in relative risk to be in the region of 1.42 to 4.1 [34].

An umbrella review in 2015 supported an association between type 2 diabetes and endometrial cancer. It examined 8174 cases, found no significant bias within the studies, and reported Summary Random Effects estimates of 1.97 (95% CI 1.71–2.27), and 95% prediction intervals of 1.23 to 3.16 [35]. Sacerdote et al. published the results of their meta-analysis which supported an association between endometrial cancer and diabetes (RR 1.65, 95% CI 1.5–1.81), although high heterogeneity between studies was found [36].

It has not been proven that modulation of insulin resistance is an effective mechanism for preventing EC [19, 37]. Many studies assess surrogate markers rather than clinical end points, and a definitive chemoprevention study is unlikely as it would require huge numbers and many years of follow-up. Randomised biomarker studies have so far failed to demonstrate a reduction in endometrial proliferation with 2–16 weeks of metformin treatment [38, 39]. Similarly, metformin has not been shown to reduce endometrial cancer risk (OR 1.05, 95% CI 0.82–1.35, p = 0.7), but may improve risk of recurrence or overall survival [40].

Long-term follow-up data provides convincing evidence that use of combined oral contraceptives (COC) is associated with a significant and enduring reduction in the lifetime risk of endometrial cancer. Iversen et al. demonstrate that the protective effects of COC persist for at least 30 years, in this updated follow-up study of users who were recruited between 1968 and 1969 [41•]. The faculty of Sexual and Reproductive Health advises that at a BMI > 35 kg/m², the risks of COC are likely to outweigh the benefits (UKMEC 3 recommendation), which precludes its use as chemoprevention in the most obese women [42].

Oral, injectable and intrauterine progestin use has been shown to reduce the risk of endometrial cancer. Data from the Finnish Cancer Registry showed a standardised incidence ratio of 0.46 for endometrial cancer in users of the LNG-IUS. The NOWAC study was a Norwegian population-based prospective cohort study. Median follow-up was 12.5 years; 9% of the cohort reported LNG-IUS use during or prior to 1998–2007. After adjusting for BMI, activity, age at start of follow-up, combined oral contraceptive use, menopausal status and parity ever-users of LNG-IUS had RR of endometrial cancer of 0.34 (95% CI 0.18–0.65) compared with never-users [43•].

Almost 25 years ago, the harmful effects of unopposed oestrogen HRT in women with a uterus were apparent. Sequential HRT has also been shown to increase the risk of endometrial cancer, with risk being inversely proportional to the number of days progestin is given for. Continuous combination has not been shown to increase endometrial cancer risk and may even reduce it, presumably because of the protective effects of progesterone on the endometrium [44].

Recent years have seen the emergence of the first oestrogen-based, progestin-free oral menopausal HRT for non-hysterectomised women. Conjugated oestrogens with the SERM bazedoxifene (CE/BZA) minimise estrogenic effects on endometrium and breast whilst effectively addressing menopausal symptoms and protecting against osteoporosis. A similar randomised, double-blind study of 17β-estradiol/raloxifene did not provide adequate endometrial protection. CE/BZA has been studied in five RCTs involving more than 7500 women and no increase in endometrial hyperplasia was found. It may be an option for women who poorly tolerate the side effects of progestins [45•].

Preclinical and animal model studies have shown that bisphosphonates have anti-tumour effects, in part through induction of apoptosis and inhibition of proliferation and angiogenesis. They have been shown to affect the growth and metastasis of gynaecological cancers in cell lines and animal models. A meta-analysis of cohort and case control studies demonstrated a statistically significant endometrial cancer risk reduction with 1–3 years of bisphosphonate use in 6499 endometrial cancer cases from 226,560 participants. With 3 or more years of bisphosphonate use, a 56% reduction in endometrial cancer risk was seen (pooled RR 0.44; 95% CI 0.28–0.7) [46•].

In their 2005 randomised controlled primary prevention trial of 39,876 women in the Women’s Health Study, Cook et al. found no association between low-dose aspirin use (100 mg/alternate days cf. placebo) and endometrial cancer risk (RR 1.22, 95% CI 0.94–1.58, p = 0.14) [47]. A 2016 meta-analysis of cohort and case control studies did suggest a small-to-modest protective effect with aspirin use, and a weak effect with regular NSAID use, from statistically non-significant pooled risk estimates [48]. The results should be interpreted with caution; only cohort and case control studies were included, and doses and duration of aspirin use varied widely between the studies included. A 2018 meta-analysis by Qiao and Yang et al. also found an inverse association between aspirin use and endometrial cancer risk (RR 0.92, 95% CI 0.85–0.99) from 8410 endometrial cancer cases in six case control studies, and 3127 endometrial cancer cases in eight cohort studies. Again, it is unclear what dose of aspirin participants were taking [49].

Continued efforts to promote breastfeeding may help to reduce endometrial cancer risk in the general population. In their meta-analysis, Jordan et al. demonstrated an 11% reduction in endometrial cancer risk in women who had breastfed (pooled OR 0.89; 95% CI 0.81–0.98) comparing 8981 endometrial cancer cases with 17,241 controls from cohort and case control studies [50].

Felix et al. presented a pooled analysis of individual level data from 18 epidemiological studies, which was the largest investigation of IUD use and EC risk to date (8801 cases and 15,357 controls). Pooled OR for ever-use of IUD was 0.81 (95% CI 0.74 to 0.9) compared with never-use with the inverse association being strongest in users of inert IUDs, and weaker in users of copper, hormonal or a combination of types of IUDs. Nulliparous IUD users gained more benefit than parous IUD users. Proposed mechanisms of action by which IUDs reduce endometrial cancer risk are increased decidual loss, alterations in hormone receptor expression and stimulation of an inflammatory microenvironment in the uterus [51].