Chemotherapy-induced weight gain in early-stage breast cancer: a prospective matched cohort study reveals associations with inflammation and gut dysbiosis

Improvements in treatment outcomes for individuals with early-stage breast cancer mark a significant achievement in cancer care. Patients with localized disease or regional nodal metastases now experience 5-year survival rates of 99% and 85%, respectively [1]. Advances in the biological understanding of the disease have improved care for these patients with the development of increasingly effective systemic therapies that reduce the risk of distant recurrence by treating subclinical residual or metastatic disease. In early studies when compared to observation, adjuvant chemotherapy reduced the risk of disease-specific mortality by nearly 30% [2] with modern regimens further reducing mortality [3].

As treatment outcomes for early-stage breast cancer improved, attention turned to reducing short-term and long-term treatment-associated morbidity. Cancer survivorship examines the field of care and research pertaining to people either living with or cured of their disease. Chemotherapy-induced syndromes such as fatigue, anxiety, depressive symptoms, and peripheral neuropathy are relatively common among survivors of breast cancer. More recently, alterations in metabolic function have also been described [4] and among breast cancer patients weight gain is a common side effect that decreases the quality of life while increasing the risk of diabetes, hypertension, and resultant cardiovascular risk [5]. Weight gain following an early-stage breast cancer diagnosis has also been associated with an increase in all-cause mortality [6, 7]. Factors associated with chemotherapy-associated weight gain include patient age, menopausal status, and reduced physical activity, while most studies do not find a simple correlative relationship between increased caloric intake and weight gain [8, 9].

The role of intestinal microbiota has also garnered recent attention as an influence on health and disease outcomes [10]. The human gut microbiome is composed of a multitude of bacteria with more than 500 characterized species classified within four dominant phyla: Firmicutes, Bacteroidetes, Actinobacteria, and Proteobacteria [11]. The gut microbiota has been shown to differ from healthy controls, a state known as dysbiosis, in a number of chronic diseases including atherosclerotic cardiovascular disease [12], type 2 diabetes [13], and obesity [14]. Dysbiosis within the gut microbiome and the relationship to inflammation is well described in the setting of inflammatory bowel disease (IBD) where reduced bacterial diversity has shown a strong association with active disease. Ott et al. demonstrated a reduction of species diversity within the gut of 30% and 50% when ulcerative colitis and Crohn’s disease patients were compared against non-inflamed controls [15]. The opposite also appears true; restoration of gut microbial diversity after fecal-microbial transplantation ameliorates disease in chronic Clostridium difficile infection [16] as well as IBD [17]. As our knowledge of the relationship between the microbiome and human health has grown, it has become increasingly apparent that dysbiosis within this complex and dynamic ecosystem may have a myriad of clinical consequences.

While the intended target of cytotoxic chemotherapy is the replicative machinery involved with cellular division in cancer, some chemotherapies are derived from antibiotics, and as such, they may also have unintended antimicrobial effects. For instance, anthracyclines act mainly through the intercalation of deoxyribonucleic acid (DNA), which in turn inhibits DNA synthesis and ribonucleic acid (RNA) production. Taxane chemotherapies, disrupt microtubule function thereby preventing mitotic depolymerization and thus inhibiting cellular division. Both of these agents are mainstays in the treatment of early-stage breast cancer, are administered systemically, and may affect the microbiota within the gut. Hepatobiliary secretion in feces is the predominant route of elimination for taxane and anthracycline chemotherapies and a recent report suggests significant bactericidal activity associated with both agents [18].

Perturbation of the gut flora in breast cancer patients receiving chemotherapy is known to occur [19], but whether weight gain reflects clinically significant changes to body composition, and the mechanisms by which this occurs are not known. We hypothesized that cytotoxic chemotherapy in women with early-stage breast cancer could affect the gut microbiota, modify host metabolism relevant to inflammation, and result in weight gain and deleterious anthropometric changes. To investigate this possibility, we undertook a prospective, controlled, matched-cohort study of women with early-stage breast cancer to characterize changes in body weight and composition in conjunction with the analysis of temporal changes in gut microbiota and inflammatory biomarkers.

In this prospective and longitudinal study of patients with surgically resected early-stage breast cancer, we observed alterations within the gut microbiome following treatment with chemotherapy, as well as significant weight gain and an increase in body fat. Furthermore, the perturbation of the microbiome was accompanied by the development of an inflammatory response that may link chemotherapy-dependent alterations within the gut microbiome with our reported anthropometric changes.

Obesity has been implicated in the development of breast cancer [27] and weight gain after treatment may impact the risk of disease recurrence and death [28]. Furthermore, the impact of weight gain during survivorship is a significant public health concern given the high prevalence of breast cancer. Quite notably, prior studies have revealed that the proportion of overweight or obese individuals increases from less than 50% to 67% when pre- and post-cancer diagnosis BMI are compared [29].

Whereas much of the available evidence to suggest an association between treatment with chemotherapy and weight gain is retrospective or anecdotal, we present prospectively collected data and furthermore we conducted DXA scans, the gold standard for body composition analysis both pre- and post-treatment. We also collected and analyzed stool and blood samples in conjunction with DXA studies, which allowed us to test the hypothesis that chemotherapy-dependent alterations in gut microbial ecology impacts host metabolism. Our study confirms earlier reports of weight gain after treatment with chemotherapy for early-stage breast cancer, but in addition, our study was designed to include appropriate control groups to permit a more extensive interpretation of our data. In particular, we included collection of data from patients treated only with endocrine therapy as adjuvant treatment following surgery, a group of patients we believe comprise the optimal comparator group. Strikingly, within this comparator group we observed none of the findings we report within our chemotherapy-treated cohort of patients.

The use of DXA for body composition analyses within our study was important for the added information it provides beyond weight gain or loss. Android fat deposition is a sensitive predictor of subsequent metabolic disease [30] and more recently, an inverse association between intestinal microbial diversity and android fat distribution was reported [31]. With our results we demonstrate the use of adjuvant cytotoxic chemotherapy, but not adjuvant endocrine therapy, is associated with significant weight gain and increased adiposity in patients with early-stage breast cancer. More specifically, our results demonstrate a significant increase in the ratio of android to gynoid fat deposition in chemotherapy-treated patients that was not seen in patients treated with only endocrine therapy. Furthermore, we demonstrate stability of lean body mass across our entire patient cohort, including those patients treated with and without chemotherapy, suggesting chemotherapy-associated body composition changes may be isolated to increased adiposity. It has been suggested that chemotherapy-associated weight gain during treatment for early-stage breast cancer may be transient [5], but if in fact body composition (i.e., an increase in body-fat percentage) and not just weight is altered during treatment with chemotherapy, the long-term health consequences may be significant even if an individual returns to their pre-chemotherapy weight. We did not record change in body composition or weight beyond approximately one year following enrolment to study, but we believe our end-of-study findings of sustained patient weight gain, increased body-fat proportion, and increased android to gynoid body fat distribution imply a significant risk for the development of long-term health consequences.

With respect to the impact of gut microbes on obesity, a few central themes appear consistent within the literature. First, a reduction with respect to bacterial diversity may contribute to the development of obesity as well as a host of additional diseases, perhaps most notably IBD [15]. And second, apart from inter-species variation, the compositional shift of the microbiota at higher hierarchical levels demonstrates an association with disease in human studies and animal models [32, 33]. We observed a significant treatment effect on microbial composition within the gut following treatment with chemotherapy, but not in women who received only endocrine therapy as adjuvant treatment. In our analysis, treatment with chemotherapy predominantly resulted in a reduction in microbial abundance, whereas treatment with endocrine therapy had a dissimilar effect. Phylum representation appeared more diverse in the pre-chemotherapy samples, whereas in the post-chemotherapy sample there appeared to be a relative abundance of Bacteroidetes. This observation suggests that the effect of chemotherapy on the gut microbiome may be to reduce the diversity of the population as opposed to a simple, pan-organism bactericidal effect.

The prospective nature of our study allowed for a comparison of intra-patient variation with respect to microbiota, as we were able to analyze pre- and post-treatment samples from the patients within our study. This is important as the composition of the gut microbiota is known to be dynamic [34], even within an individual, and without pre- and post-treatment sampling it would be difficult to determine whether differences between patients groups (e.g., chemotherapy and endocrine therapy-treated) truly relates to a treatment effect.

As with prior studies, we observed greater post-chemotherapy weight gain among younger patients [5, 35], although given the small number of patients within this subgroup specific caution should be exercised when interpreting our data. We stratified our patient cohort both by age and menopausal status and observed increased weight gain among patients younger than 60 years of age, and also in the group of patients who were pre-menopausal. The relationship between estrogen, the gut microbiome, and health and disease is an emerging area of study, and the estrobolome, an aggregate of enteric bacterial genes whose products are capable of metabolizing estrogens, has been described [36]. Biliary excretion of estrogen and its metabolites is a well-known phenomenon [37] and indeed, estrogen metabolites in conjugated form may be recovered from feces [38, 39]. Furthermore, antibiotic exposure in pre-menopausal women has been shown to increase estrogen excretion [40], possibly through perturbation of the gut microbiome and a corresponding reduction in the deconjugation of non-absorbable metabolites of the hormone. The effect of estrogens on metabolism is well known, with estrogen deficiency contributing to the development of obesity and the metabolic syndrome [41, 42]. As is the case with antimicrobial therapy, perturbation of the microbiome associated with chemotherapy may result in a reduction in the re-absorption of estrogen metabolites from the gastrointestinal tract. Thus, the relative increase in weight gain observed in younger patients may be due in part to an anti-estrogenic effect. Conversely, the pleiotropic effects of estrogen are reduced with aging through multiple mechanisms, including a reduction in the level of circulating hormone but also through a reduction in estrogen receptor expression via transcriptional and epigenetic mechanisms [43, 44], which may explain why older, postmenopausal patients exhibit less weight gain versus their younger, pre-menopausal counterparts.

At present, we do not fully understand the complex mechanisms by which perturbation of gut microbiota may impact human health and disease, but others have postulated these changes may be mediated through systemic inflammation (recently reviewed by Cox et al [45]). After recognizing that treatment with chemotherapy appears to reduce microbial diversity and abundance within the gut in a manner similar to that seen in active IBD, we measured circulating cytokines and chemokines among our patient cohort, as well as fecal calprotectin levels. For patients who received treatment with chemotherapy an increase in fecal calprotectin levels as well as circulating pro-inflammatory mediators was observed, but a similar response was not seen among patients who received treatment with endocrine therapy after breast cancer surgery; in fact, the opposite appeared true. We observed a robust inflammatory response in chemotherapy-treated patients characterized by more than a 2.5-fold increase in fecal calprotectin levels, as well as increased levels of interleukin-6, interleukin-8, interleukin-17, and interleukin-18. We also discovered an increase in circulating chemokine levels following treatment with chemotherapy, and while a component of this response may indicate a homeostatic response to chemotherapy-induced cytopenias, when considered as a whole the picture is more consistent with a pro-inflammatory response similar to that which may be seen in the setting of IBD. The relationship between gut microbiota and inflammation seen with inflammatory bowel disease is a complex one and evidence suggests that perturbation of the microbiome may lead to the development of aberrant inflammation rather than inflammation leading to perturbation of the microbiome. Further study will be required to determine whether the same process occurs following treatment with chemotherapy, however, we believe such a scenario may provide a mechanistic link between treatment with chemotherapy, resultant intestinal microbial dysbiosis, and the subsequent occurrence of weight gain and metabolic disease seen within these patients.

Within our patient cohort, weight gain after treatment with chemotherapy was associated with increased FGF-21 and MCP1 levels. Within the gut FGF-21, a hepatokine secreted by the liver functions as a potent activator of glucose uptake in adipocytes [46]. FGF-21 is elevated in obese subjects and is independently associated with the development of type 2 diabetes [47]. MCP1 is known to confer resistance to insulin signaling [48], and we think that the increase after treatment with chemotherapy seen with both cytokines may be of particular significance within the context of weight gain and body compositional changes. Interestingly, the increase of circulating MCP-1 within our chemotherapy-treated patients is of additional interest when considered within the broader literature as MCP-1 is known to play a key pathogenic role in the development of immune-mediated illness including colitis [49, 50].

The increase in measured chemokines including chemokine ligand 23 (CCL23) may also represent a mechanism by which chemotherapy results in systemic inflammation. CCL23 is a chemokine with potent chemoattractant properties for resting T-lymphocytes [51]. CCL23-dependent signaling has been implicated in several inflammatory disease states including rheumatoid arthritis [52] and IBD [53]. Similarly, our observed upregulation of interleukin-18 (IL-18), a member of the IL-1 family of cytokines, may translate treatment with chemotherapy to inflammation. IL-18 plays a major role in the induction of interferon-gamma signaling, and an increase in circulating IL-18 has been demonstrated in numerous inflammatory disease states including metabolic syndromes, psoriasis, and IBD [54]. Although we are limited by a relatively small sample size, our study informs the possibility of the potential development of gut inflammation in subsets of chemotherapy-treated patients.

Previous studies established that chemotherapy is associated with harmful anthropometric change, including weight gain, which is a known risk factor for poor patient outcomes in survivorship after treatment for early-stage breast cancer. In this study, we link cytotoxic chemotherapy with perturbation of the gut microbiome and development of a robust inflammatory response with resultant weight and body compositional changes.

We acknowledge the extensive recent work performed by Terrisse et al. [19], in which the authors also demonstrate patient weight gain following treatment with chemotherapy. However, the alterations within the gut microbiome within the study by Terrisse would appear opposite to those that we observed within our prospective study. Why treatment with cytotoxic therapy would increase the diversity of bacterial populations within the gut is unclear, but it seems likely that any change within the gut microbiome after treatment with chemotherapy would be transient, and thus the difference between the results of our two studies may reflect a difference in the timing of sample collection. Our analysis of the effect of chemotherapy on the composition of the gut microbiome was made on samples collected just before or immediately following a patient’s final cycle of cytotoxic therapy — whereas if samples were collected at too distant a time following cessation of treatment the true chemotherapy effect may have been missed.

We also note the recent work of Uzan-Yulzari et al., in which evidence is supplied to indicate baseline microbiome characteristics may predict weight gain and metabolic derangement after treatment with chemotherapy [55]. We find this hypothesis intriguing and believe it merits further investigation. This work is complementary to our own; however, our study design allows us to demonstrate that treatment with chemotherapy also alters the composition of the gut microbiome (in addition to any baseline changes which may be relevant) in association with these anthropometric changes.

Strengths of our study include a priori data and biosample collection. Our matched-cohort study design allowed for the enrolment of 40 individuals with early-stage breast cancer, divided between those patients who received chemotherapy after curative-intent surgery, and those who would only receive endocrine therapy following surgery, thus including within our analysis a population of patients whom we believe to represent the optimal control group for studying chemotherapy-dependent microbiome and body compositional change. Our study represents a comparatively large analysis with respect to characterization of the microbiota in the setting of chemotherapy, and indeed cancer in general, but nonetheless, given the complexity of the gut microbiome our sample size still limits the interpretation and generalizability of our data. In addition, prospective collection of biomarkers specific for metabolic disease would have added value to the study, and future projects will include prospective data collection over a longer duration, as well as prospective evaluation of glycaemic control, blood pressure, and lipid indices following treatment with chemotherapy.

Biological mechanisms by which post-chemotherapy alterations within the gut microbiota could lead to weight gain are yet to be elucidated. However, it has been postulated that a systemic inflammatory response may occur due to a microbial shift, which could lead to the development of obesity or disease. We sought to connect alterations within the gut microbiota with the development of systemic inflammation as a consequence of cytotoxic chemotherapy. Notably, we show that treatment with chemotherapy, but not endocrine therapy, results in a transient rise in circulating pro-inflammatory cytokines and chemokines, as well as a significant increase in fecal calprotectin levels, which indeed indicates a connection between post-chemotherapy gut microbiota change and inflammation. Moreover, this illuminates a potential mechanism by which chemotherapy could result in weight gain in breast cancer survivorship, and a possible target of future treatments designed to improve long-term patient outcomes.