Key words
obesity - gynecological malignancies - BMI
Introduction
The global obesity rate has doubled in the years from 1980 to 2008 and continues to increase. In the European Union, depending on the region, between 10 and 30% of adults are classified as clinically obese, and in the USA up to 40% of adults meet the definition of pathologically obese, a rate which does not include moderate overweight [1], [2]. Based on the 2016 WHO definition of a body mass index (BMI) of more than 30 kg/m2, obesity is associated with serious long-term health sequelae which affect almost all organ systems. In recent decades, scientific research has begun to focus on the interaction between obesity and oncogenesis. Gynecological cancers occupy a key position, as postmenopausal breast cancer and endometrial cancer – both of which make up about 33% of all new cancer diagnoses in women in Austria per year – are the most common cancers associated with obesity [3]. It is estimated that around half of all cancers could be prevented based on our current level of research; around one third of all cancers in the western world are ascribed to poor nutrition and a lack of physical activity. In the words of the WHO at the start of the 2000s, the clinical implications of new diagnoses of potentially preventable cancers represents an “unheard-of health policy challenge” [4].
Despite the association between obesity and cancer incidence reported for decades and the increasing evidence of possibly causal pathophysiological connections, scientific investigations into the role of obesity in gynecologic oncology are far from complete. Not least because, apart from the reported association as an epidemiological risk factor for endometrial and postmenopausal breast cancer, the data is limited. As the currently available data largely consists of data from retrospective cohort studies, the question whether the reported association should be interpreted as causal or correlative cannot be conclusively answered [5].
Recently published research indicates that pathologically increased BMI may be a negative predictor for recurrence and disease-specific overall survival. Current research is focusing on the role of possible influencing factors such as increased perioperative complication rates and lower response rates to cytotoxic chemotherapy regimens. Recent data have also indicated possibly clinically relevant interactions between higher BMI and the response to immunotherapies.
Given the rise in cases with disease and the new scientific data, the well-known connection between obesity and gynecological cancers is becoming more significant. The aim of this article is to review the findings of current studies on the pathophysiology and epidemiology of obesity and their clinical implications for the treatment of gynecological cancers.
Pathophysiological Background: Obesity and Oncogenesis
Pathophysiological Background: Obesity and Oncogenesis
The increase in fatty tissue due to obesity is associated with numerous pathological molecular changes which promote carcinogenesis and can have a negative impact on the course of disease in patients with gynecological cancers. In addition to the endocrine effects of the pathologically increased hormone production of adipocytes, obesity-induced insulin resistance and systemic inflammatory processes are also discussed as possible causes.
Obese fatty tissue is characterized by pathological infiltration of immunocompetent cells. While, physiologically, the majority of fatty tissue macrophages are anti-inflammatory M2 subtypes, there is an increase in both the relative and the absolute percentages of pro-inflammatory M1 macrophages in obese fatty tissue. Pathogenetically this could be interpreted as an immune response to cellular damage to the adipocytes caused by changes to cell membrane permeability by excessive absorption of saturated fatty acids [6]. Cytokines released by pro-inflammatory M1-type macrophages (including tumor necrosis factor-α [TNF-α], interleukin-6 [IL-6] and interleukin-1 β [IL-1β]) can have a direct oncogenic effect through activation of the NFκB and JNK signaling pathways and can cause oxidative DNA damage as a consequence of inflammatory processes, resulting in an increase in the cellular mutation load. Protective anti-inflammatory regulatory T-cells
(Treg) disappear in obese fatty tissue, which additionally contributes to amplification of a pro-inflammatory spiral [7]. Inflammatory processes are not limited to fatty tissue; in obese patients they lead to a measurable increase in the serum levels of IL-6 and TNF-α, which subsequently also manifest through elevated C-reactive protein (CRP) levels as markers of systemic inflammation [8], [9].
The second carcinogenic pathomechanism currently being discussed is obesity-induced insulin resistance. High insulin serum levels have been found to exert growth-stimulating and antiapoptotic effects in vitro and in vivo [10]. In the context of the chronic increase in lipolysis associated with obesity, elevated free fatty acid concentrations have been found to suppress insulin signaling pathways [11], [12]. Physiologically, visceral fat has a higher lipolytic activity than peripheral fatty tissue, with predominantly upper-body fat distribution contributing further to the development of insulin resistance [13]. Elevated IL-6 and TNFα serum levels induced by obesity-related inflammatory processes are thought to lead to a synergistic increase in lipolytic activity [14]. Increased insulin sensitivity triggered by the
somatic overexpression of insulin and fibroblast growth factor-1 (FGF1) receptors sensitizes cells to the pro-proliferative effects of insulin, contributing to the development of endometrial hyperplasia [15].
The possible carcinogenic effect of the peptide hormones adiponectin and leptin secreted by adipocytes is also being investigated. Adiponectin increases cellular insulin sensitivity and has an anti-inflammatory effect, with physiological adiponectin concentrations having an oncoprotective effect. Endocrine dysregulation may therefore facilitate carcinogenesis [16], [17]. Leptin plays a key role in satiety regulation by the hypothalamus. Obesity is associated with leptin overexpression, with hypothalamic leptin resistance also leading to pathologically elevated concentrations. Leptin also stimulates angiogenesis processes and is likely to have a direct carcinogenic effect through the SAT3 signaling pathway. It has also been suggested that leptin may increase estrogen levels [10], [18].
Obese fatty tissue is associated with pathologically elevated estrogen levels as a consequence of the increased aromatase expression by adipocytes and the resulting increase in the aromatization of androstenedione and testosterone. By binding to estrogen receptors α or β, estrogen initiates the transcription of pro-proliferative genes, which act as a growth signal for receptor-expressing gynecological cancers. Moreover, studies have reported a direct mutagenic effect of elevated estrogen levels due to the damage caused by oxidative estrogen metabolites, which can lead to oxidative DNA damage and depurination [19]. In addition to increased estrogen synthesis, obesity is also associated with significantly decreased concentrations of the sex hormone-binding globulin (SHBG) [20]. The relative loss of sex steroid-binding capacity again results in elevated serum levels of endocrine-active estrogens. This corresponds
to the increased risk of postmenopausal breast cancer and endometrial cancer reported for patients with decreased SHBG serum levels [21], [22].
Current Clinical Data on Obesity and Gynecologic Cancers
Current Clinical Data on Obesity and Gynecologic Cancers
Endometrial cancer
The strongest correlation between obesity and the risk of developing a malignancy has been reported for endometrial cancer. Widely validated scientific evidence has highlighted the strong, positive, linear correlation between increasing BMI and the incidence of disease as well as the strong, negative, linear correlation with oncological outcomes. Currently, more than half of all newly diagnosed endometrial cancers are associated with obesity and could thus potentially be prevented. The association is strongest in postmenopausal patients with endometrioid adenocarcinoma; a cohort study of 33 436 postmenopausal patients reported that obese patients had a more than fourfold increased risk of developing endometrioid adenocarcinoma compared to normal weight patients, while the risk of developing serous, clear-cell or other high-grade cancers was more than twice as high [23]. This corresponds to a 50% risk increase per 5 BMI units in patients
with postmenopausal endometrioid adenocarcinoma [24]. A similar BMI-dependent risk constellation has also been reported for patients with premenopausal endometrial cancer, although the data for this is still limited [25]. Cancer-specific mortality is more than twice as high for patients with a BMI between 30 and 34.9, and in patients with a BMI of more than 40 it is more than six times higher [26].
Breast cancer
There is also solid evidence for an association between obesity and both the incidence and the prognosis of breast cancer, although the respective risk can be stratified according to hormone receptor status and menopausal status. Meta-analyses have found an independent, positive, linear correlation between BMI and postmenopausal, hormone receptor-positive breast cancer, with an 82% increase in the relative risk of disease. This corresponds to an increase of 10 – 32% for the relative risk per 5 kg/m2 increase in BMI. The slow but continuous increase in weight over the years in this cohort was significantly associated with disease incidence [27]. However, no association has been found between BMI and postmenopausal hormone receptor-negative breast cancer [28], [29].
A higher premenopausal BMI is associated with a decreased relative risk of hormone receptor-positive breast cancer, with every BMI increase of 5 kg/m2 associated with a 10% lower incidence. The underlying pathomechanisms are not clear [28]. Another meta-analysis observed a 43% increased risk for triple-negative breast cancer in populations of premenopausal women but no significant association between obesity and triple-negative cancer in postmenopausal women [30], [31].
After adjusting for tumor subtypes and menopausal status, obesity was found to increase the cancer-specific mortality risk and risk of recurrence by about 35 – 40% as well as the probability of developing distant metastasis and late recurrence. The available data do not show whether this is due to late diagnosis or more aggressive tumor biology [32].
Epithelial ovarian cancer
The association between BMI and the risk of disease is only significant for specific histological subtypes and not for all epithelial ovarian cancers. This is according to the current state of research which considers epithelial ovarian cancers as a heterogeneous group of different malignancies. A meta-analysis of 13 448 patients found no association between BMI and the incidence of high-grade serous or high-grade endometrioid adenocarcinoma. However, the relative risk of low-grade serous or low-grade endometrioid ovarian cancer was increased by 13% and 20%, respectively, with a linear positive association with increasing BMI. There was also a weak association with an increase in the relative disease risk for mucinous and clear-cell ovarian cancers of 17% and 19%, respectively. The strongest association was for serous borderline tumors, which showed a relative risk increase of 20 – 25% per 5 kg/m2 BMI, while the increase was weaker for mucinous borderline
tumors at 9 – 11% per 5 kg/m2 BMI. The pathogenetic reasons have not been determined. The effect of BMI on disease risk was found to be stable when compared to BMI data for early adulthood (age 18 – 20), which points to obesity as an independent risk factor [33]. A retrospective multicenter study with 370 patients found that BMI had no effect on the interval to recurrence, although the study was not sufficiently powered to be able to show differences between the different histotypes [34]. As regards the known stronger effect on hormone receptor-positive breast cancer and endometrial cancer, a subgroup analysis of high-grade serous cancer stratified according to hormone receptor status could provide clinically relevant findings. However, no solid data are currently available.
Vulvar cancer
Data on obesity for patients with vulvar cancer is limited because of the low incidence of disease. AGO-CaRE-1, currently the largest study with a total of 1618 patients with vulvar cancer, analysed the BMI values of 849 patients with squamous cell carcinomas and found that the rate of local recurrence was twice as high in the cohort of patients with a BMI of more than 30, although there were no differences in the rates of distant recurrence. Similarly, disease-free overall survival rates were 36% lower for this cohort. Multivariate analysis showed this to be an independent factor, although there were no differences in R0 resection rates, chosen therapies and response rates between groups [35]. Further analysis of the same cohort confirmed a higher rate of local recurrence and shorter progression-free survival rates for obese patients [36].
Cervical cancer
The currently available data on the association between obesity, disease risk and disease-specific overall survival for cervical cancer is inconsistent. While one meta-analysis of 128 233 patients reported a relative increase in disease risk of 8 – 40% for obese patients, the possible pathomechanisms remain unclear [37]. The disease-specific mortality rate was higher for obese American women of Caucasian origin, but this could not be verified for obese Afro-American patients. It has been suggested that screening compliance of obese patients may be lower, and lower screening rates can lead to higher cancer-specific mortality rates [38]. According to a number of small case-control studies, part of this effect could be due to the relative increase in the risk for adenocarcinoma but not squamous cell carcinoma [39], [40]. A retrospective cohort
analysis of 944 227 American screening patients which included a total of 490 cases with cervical carcinoma highlights the increased risk of disease for obese patients, while no differences were reported with regard to HPV status or histology. The lower rate of diagnosed precancerous cervical lesions in this patient population indicates that screening quality is lower in cohorts with higher BMI [41].
Complication Rates After Surgical Interventions and Obesity
Complication Rates After Surgical Interventions and Obesity
Operating times are longer and intraoperative blood loss is higher during abdominal oncologic surgery carried out in obese patients, and the rate of short-term postoperative complications is also higher in this cohort. However, obesity had no impact on cancer-specific long-term survival [42]. Although a retrospective cohort study of 514 patients with endometrial cancer showed an increased rate of mild postoperative complications, particularly infection and wound healing complications after laparotomy, a finding that was borne out by a retrospective systematic analysis of 8453 cases, no correlation was found with intraoperative complications or postoperative long-term morbidity [43]. This was also confirmed for patients with ovarian cancer who had cytoreductive surgery [44] and patients with cervical cancer who underwent radical abdominal hysterectomy with pelvic lymph node
dissection [45]. The oncologic prognosis of obese patients who had breast cancer surgery was not poorer than that of non-obese patients; however, cosmetic outcomes were significantly poorer, which also reduced postoperative quality of life [46]. Based on the evidence obtained from large, retrospective cohort analyses, a high BMI alone is not a contraindication for radical surgery in gynecologic oncology patients. However, this does not apply to patients with pre-existing obesity-related comorbidities [47]. Given the fact that no indications of an independent correlation between BMI and serious surgical complications have been found to date, carrying out an individual risk-benefit analysis of the perioperative risk for obese patients is essential.
Adjuvant Treatment Strategies
Adjuvant Treatment Strategies
A non-systematic review of 66 studies found an association between obesity and poorer oncologic outcomes following cytotoxic chemotherapy in patients with ovarian or endometrial cancer. This effect can be ascribed, in the first instance, to the lack of evidence about the pharmacokinetics of cytotoxic therapies in patients with a high BMI or calculated body surface. As regards correlations between dose intensity and response rates, toxicity and survival rates, all of the above suggests that the dose intensity should be maintained, irrespective of the level of obesity [48], [49], [50]. The guidelines of the American Society of Clinical Oncology argue in favour of fully weight-adapted dosages of cytotoxic therapies and against any reduction of doses, even though the data, particular for morbid obesity, is extremely thin [51]. Even though obese
patients have higher toxicity rates, it was not possible to deduce a consequent negative impact on survival from this. Although analysis of the prospective phase-III GAIN trial on adjuvant dose-dense therapy strategies in 3023 breast cancer patients found significantly higher toxicities compared to non-obese patients, the survival rates were the same, which points to suboptimal dose regimens in this patient population [52]. There are currently no comparable data for vulvar or cervical cancer.
It is possible that obese patients with ovarian cancer also benefit less from adjuvant bevacizumab therapy than normal weight patients; however, the data on whether this effect is due to pharmacokinetics or tumor biology is limited and unclear [53]. Comparable observations have also been reported for non-gynecological malignancies [54].
Early clinical data obtained in the context of experimental immunotherapies investigating the therapy response to checkpoint inhibitors (CPI) in patients with non-gynecological malignancies could offer a counterpoint to the findings on cytotoxic therapies. After adjusting for possible cofounders, an Italian retrospective multicenter study with 876 patients, the majority of whom had had malignant melanoma, kidney cell cancer or non-small cell lung cancer, found significantly better response and survival rates for both patients with BMI ≥ 25 and patients in the cohort with BMI ≥ 30. The reason for this was not clear; irrespective of possible dosage bias, tumors encouraged by pathologically increased obese tissue might respond better to CPI therapies. The same effect has also been previously described for smaller independent cohorts, particularly for malignant melanomas [55], [56]. But because similar data on
gynecological malignancies is lacking, there are currently no clinical implications for gynecologic oncology [57].
Outlook and Future Issues
Outlook and Future Issues
While BMI is an easy target because it is simple to use and cost-effective in clinical practice, the validity of using BMI has repeatedly been criticized, particularly with regard to extreme values and the lack of discriminatory data on muscle percentages or body fat distribution. Clinical alternatives such as waist circumference or waist-to-hip ratio could serve as prognostic factors, as they reflect the oncogenic potential of visceral fatty tissue better. However, the data on this is currently still limited. Although waist-to-hip ratio was identified as an independent prognostic factor in a study of 4062 Chinese patients with breast cancer, there are currently no direct comparisons of the prognostic value of waist-to-hip ratio compared to BMI [58].
Because of the negative impact of obesity on the probability of recurrence and on cancer-specific overall survival, it has been suggested that obesely increased fatty tissue could be a potential oncological therapy target in preclinical models [59]. Although widely validated in non-oncological settings, clinical data on the availability of anti-obesity drugs in oncological therapy are lacking. Glucagon-like peptide 1 (GLP-1) receptor antagonists such as liraglutide which inhibits insulin synthesis and decelerates gastric emptying result in a clinical weight loss of 5 – 10% of body weight within 20 weeks, with stable long-term effects if treatment is maintained. The effect is comparable to that of other licensed preparations such as the lipase inhibitor orlistat or the serotonin antagonist lorcaserin and is not inferior to bariatric surgery [60], [61]. A retrospective cohort
analysis with a 5-year follow-up after surgically enforced weight loss reported a reduction in the risk of developing endometrial cancer of up to 40%. There is also broad evidence for a decrease in serum concentrations of sex steroids and insulin [62], [63], [64]. A forced reduction of obese fatty tissue by the administration of drugs could therefore be a promising future therapeutic approach.
Conclusion: Relevance for Clinical Practice
Conclusion: Relevance for Clinical Practice
Obesity is an independent risk factor which considerably increases the probability of developing endometrial and postmenopausal breast cancer compared with normal weight patients. The epidemiological associations for cervical cancer, vulvar cancer, epithelial ovarian cancer and premenopausal breast cancer are plausible, but the evidence is less solid and the reported effects less pronounced, especially as HPV screening for the prevention of vulvar and cervical cancer plays a far greater clinical role and no connection has been found for the most common high-grade serous subtypes of epithelial ovarian cancer ([Table 1]). The data on clinical recommendations for action is thin. Based on current evidence, patients with an increased BMI but no associated comorbidities have no increased risk of serious complications during surgical primary therapy nor do they have higher perioperative mortality rates, which is why radical surgical strategies are
not contraindicated. Similarly, dose-reduced cytotoxic chemotherapy regimens for ovarian and endometrial cancer are associated with poorer oncologic outcomes in obese patients, and therefore the recommendation is that obese patients are treated using fully weight-adapted dosages despite the potentially higher toxicity. Despite the known negative correlation between BMI, the probability of recurrence and oncologic outcomes, because of the lack of evidence, forced reduction of obese fatty tissue, either through bariatric surgery or drug therapy, must be limited to clinical trials.
Table 1 Key messages on the clinical relevance and management of obesity and its association with gynecological cancers, based on tumor location. Obesity is defined as a body mass index (BMI) ≥ 30 kg/m2.
|
Cancer location
|
Key messages: relevance for clinical practice
|
|
1 The relative increase in the risk of developing cervical cancer which has been reported in some studies is probably due to the poorer quality of screening [41].
2 A relative increase in the risk of developing low-grade, mucinous and clear-cell cancers has been reported, but because of the rarity of these histological subtypes and the limited effect, the clinical relevance is limited [33].
|
|
Endometrium
|
pre- und postmenopausal: positive linear correlation with incidence and poorer oncological outcome [23]
|
|
postmenopausal endometrioid adenocarcinoma: 50% increase in relative risk per 5 BMI units [24]
|
|
Breast (hormone receptor-positive)
|
postmenopausal: positive linear correlation with incidence and poorer oncological outcome [32]
|
|
postmenopausal cancer: 10 – 32% increase in relative risk per 5 BMI units [27]
|
|
premenopausal cancer: 10% increase in relative risk per 5 BMI units [28]
|
|
Breast (hormone receptor-negative)
|
postmenopausal cancer: relevance for disease risk and risk of cancer-specific mortality is clinically negligible [31]
|
|
premenopausal cancer: associated with development of triple-negative cancer [30]
|
|
Cervix
|
relevance for disease risk and risk of cancer-specific mortality is clinically negligible1,2
|
|
Ovary (epithelial)
|
|
Vulva
|
relevance for disease risk is clinically negligible [35], [36]
|
|
significantly shorter progression-free and disease-specific survival rates [35], [36]
|
|
Therapy strategies
|
Surgical: obesity without associated morbidity is no contraindication for radical surgical treatment strategies [43], [44], [45]
|
|
Cytotoxic chemotherapy: fully weight-adapted doses associated with better oncological outcomes despite higher toxicity [48], [49], [50], [51]
|
