A review of the impact of energy balance on triple-negative breast cancer

Abstract

Cancer cells cannot proliferate without sufficient energy to generate biomass for rapid cell division, as well as to fuel their functions at baseline. For this reason, many recent observational and interventional studies have focused on increasing energy expenditure and/or reducing energy intake during and after cancer treatment. The impact of variance in diet composition and in exercise on cancer outcomes has been detailed extensively elsewhere and is not the primary focus of this review. Instead, in this translational, narrative review we examine studies of how energy balance impacts anticancer immune activation and outcomes in triple-negative breast cancer (TNBC). We discuss preclinical, clinical observational, and the few clinical interventional studies on energy balance in TNBC. We advocate for the implementation of clinical studies to examine how optimizing energy balance—through changes in diet and/or exercise—may optimize the response to immunotherapy in people with TNBC. It is our conviction that by taking a holistic approach that includes energy balance as a key factor to be considered during and after treatment, cancer care may be optimized, and the detrimental effects of cancer treatment and recovery on overall health may be minimized.

Maintenance of a healthy body weight requires exquisite matching of energy intake and output, although there is considerable debate as to whether body size is determined only by the balance of calories in and calories out. When energy output does not perfectly match intake, the surplus energy will necessarily be stored, typically as fat. However, a growing tumor can hijack the typical paradigms of storage of excess energy as fat and use some of this excess energy to generate the biomass required for rapid cell division. In addition, the insulin resistance canonically conferred by excess body weight may mediate part of the link between energy imbalance and cancer by increasing insulin-dependent glucose uptake in tumors (1,2), even without a continuous excess in energy intake. Without question, there are numerous putative mechanisms that may mediate the link between energy imbalance and cancer risk and prognosis. Many observational studies have attempted to delineate the multifactorial mechanisms and systems by which energy imbalance influences cancer risk and outcomes (3-5), and PubMed citations for cancer and energy balance, diet, and exercise have increased exponentially in recent years (Figure 1). However, energy balance in cancer has proven to be a complex field, because of the unique transcriptional programming, cellular signaling, and immunogenic features of each cancer type. This is not to argue that there are no common principles in the relationship between altered energy balance and cancer risk and prognosis that is followed by all cancer types. Some reports find a beneficial effect of implementing a neutral or negative energy balance on cancer outcomes (6-8).

Figure 1.

Interest in physical activity and cancer is exponentially increasing. Shown are the PubMed hits for “energy cancer,” “diet cancer,” and “exercise cancer” on June 16, 2022.

The relationship between energy balance and cancer (Figure 2) can be explained by the effects of obesity (resulting primarily from excess nutrient intake) and energy output (a function of basal caloric expenditure and physical activity) on tumor cells. Obesity is associated with an increased risk, development, progression, and poor prognosis of many types of cancer, including pancreatic, esophageal, colorectal, prostate, ovarian, endometrial, and breast cancer (9,10). Breast cancer has been the setting in which a plurality of observational and intervention studies of energy balance has been completed, finding higher body mass index (BMI) associated with higher risk of developing (11,12) and dying of postmenopausal breast cancer (13,14) and exercise, diet, and weight loss interventions improving tissue- and serum-based cancer biomarkers, body composition, and patient-reported outcomes in women with or at high risk for this disease (15). The relationship between energy balance and risk, outcomes, and health in survivors of all subtypes of breast cancer has been reviewed previously (16-21). However, few studies of energy balance and breast cancer have focused specifically on triple-negative breast cancer (TNBC). TNBC is characterized by the absence of estrogen receptors (ERs), progesterone receptors (PRs) and HER2, making it resistant to standard hormonal and HER2-targeted therapies. TNBC is the most lethal form of breast cancer, with a recent analysis of Surveillance, Epidemiology, and End Results program data identifying a 5-year overall survival rate of only 76.5% in patients with TNBC as compared with more than 86% in patients with hormone receptor–positive breast cancer regardless of HER2 status (22). Additionally, TNBC exhibits a clear association with obesity; case analysis and retrospective studies have found that breast cancer patients suffering from TNBC are more likely to have a higher waist-to-hip ratio (a measure of visceral adiposity) and be overweight or obese than patients with other subsets of breast cancer (23-28). As predicted by these studies, obesity confers the largest deleterious effect on patients with TNBC as compared with other breast cancer subtypes (29). The advent of immunotherapy in breast cancer treatment has improved outcomes for TNBC patients: in 2019, the US Food and Drug Administration approved immunotherapy (atezolizumab) in combination with paclitaxel for advanced-stage, programmed death-ligand 1 (PD-L1)–positive TNBC. Subsequently, this accelerated approval was withdrawn because of the drug’s failure to meet its primary endpoint of progression-free survival in a larger trial (IMpassion131) (30). However, another agent in this class, pembrolizumab, was approved for metastatic and high-risk early stage TNBC in 2020 and 2021, respectively. Given advances in adjuvant treatment for TNBC, understanding the role of energy balance in modifying and mediating treatment efficacy is of interest. This is particularly timely considering recent studies indicating that overweight may, seemingly paradoxically, improve the response to immunotherapy in lung cancer (31,32) and melanoma (33-35). However, lung cancer and melanoma are epidemiologically different tumor types from breast cancer. Although most studies have not found a correlation between excess body weight and lung cancer or melanoma incidence or prognosis after controlling for other relevant covariates, TNBC incidence and outcomes have a clear correlation with excess body weight in postmenopausal women. Considering the aggressive nature of TNBC and its poor overall survival rates, it is urgent that we elucidate modifiable factors to help improve prognosis and patient-reported outcomes in patients with TNBC (36,37) and generate evidence-based recommendations to empower patients in this area. To that end, this translational, narrative review examines observational and interventional studies on energy balance and TNBC in animals and humans.

Figure 2.

Summary of potential mechanisms that elucidate how energy balance could slow tumor growth, improve response to treatment, and improve clinical outcome. BMI = body mass index; IGF-1R = insulin-like growth factor 1 receptor; MDSC = myeloid-derived stem cell; QoL = quality of life; SGLT = sodium-glucose cotransporter; TNBC = triple-negative breast cancer.

Methods

Figure 3 shows the method used to construct this narrative review. Studies were identified via PubMed searches for “triple negative breast cancer diet,” “triple negative breast cancer exercise,” “triple negative breast cancer energy balance,” and “triple negative breast cancer fatigue” (all without quotation marks) in August 2022. The search was repeated in December 2022, and newly published studies added. Reference lists for all included studies were also cross-checked for additional relevant studies. Many human studies included patients with breast cancer, as well as other tumor types, or multiple breast cancer subtypes. Studies were included only if data from patients with TNBC were analyzed as a subgroup. Formal guidelines for systematic reviews were not followed. Throughout this narrative review, the terms physical activity and exercise are used interchangeably.

Figure 3.

Flowchart showing the strategy employed to construct this review. Figure created using Biorender.com. TNBC = triple-negative breast cancer.

Preclinical interventional studies linking energy balance to outcomes in TNBC

The 2 most commonly studied murine models of TNBC are 4T1 and E0771 cells, with PY8819 cells representing a third but substantially less studied TNBC model. The reports we reviewed uniformly demonstrated that increased energy intake and/or decreased energy expenditure accelerate tumor growth (Tables 1-3) (38-49). Relatedly, several of these published studies demonstrated that obesity increased tumor angiogenesis (46), the presence of tumor cells in lymph nodes, bone marrow (42), and lung and liver metastases (38,40,44). Caloric restriction (38) and energy wasting via promoting glycosuria with a sodium-glucose cotransporter-2 inhibitor (2,45) protected mice from the tumor-promoting effects of high-fat diet, demonstrating that it is the hypercaloric nature of an obesogenic diet, rather than the composition of the diet itself, that mediates the impact of high-fat diets to promote TNBC progression. Intriguingly, chow-fed offspring of obese mother mice also showed acceleration of tumor growth (50), possibly as a result of increased body weight in the offspring and/or an inherited factor.

Table 1.

Preclinical studies on the impact of energy intake and obesity on TNBC tumor growth^a

Study	Tumor model	Lifestyle intervention	Outcomes
Liu et al. (40)	4T1	Regular chow (10% fat, 70% carbohydrates, 20% protein) or a high-fat diet (45% fat, 35% carbohydrates, and 20% proteins) for 28 d	• High-fat diet increased tumor volume approximately 40% 28 d after tumor injection and increased lung metastasis.
Spielmann et al. (41)	4T1 in ovariectomized mice	Regular chow or a 50% fat high-fat diet (full nutrient composition of the diets was not reported) for 13-14 wk, after which the mice were ovariectomized. Tumor cells were injected 3 wk thereafter, and the mice were kept on the obesogenic diet for 20-21 wk.	• High-fat diet increased tumor size approximately 60% 3 wk after tumor injection.
Evangelista et al. (42)	4T1	Regular chow (10% fat) or high-fat diet (57% fat) was provided for 16 wk.	• High-fat diet increased tumor size 30 d after tumor injection. • High-fat diet increased 4T1 cells in sentinel lymph nodes and bone marrow.
Kim et al. (43)	4T1	Regular chow (10% fat) or high-fat diet (60% fat) was provided for 16 wk prior to tumor injection and an additional 4 wk after injection.	• High-fat diet increased tumor size approximately 50% 30 d after tumor injection.
Kim et al. (44)	4T1	Regular chow (10% fat) or high-fat diet (45 or 60% fat) was provided for 12 or 16 wk prior to tumor injection and an additional 25 d after injection.	• High-fat diet increased tumor size approximately 30% 25 d after tumor injection. • High-fat diet increased lung and liver metastases.
Clements et al. (48)	4T1	Low- (10% fat) or high-fat diet (60% fat) was provided for 12 wk and tumors injected after 8 wk.	• High-fat diet increased tumor size approximately sixfold 3 wk after tumor injection.
Akingbesote et al. (45)	4T1	Regular chow (18% fat) or high-fat diet (60% fat) was provided 2 wk prior to tumor injection and continued for 4 wk following injection.	• High-fat diet increased tumor size approximately 70% 4 wk after tumor injection. • Glucose wasting with dapagliflozin slowed tumor growth in lean and obese mice.
Bowers et al. (38)	E0771	Regular chow or high-fat diet (60% fat) was provided for 25 wk prior to tumor injection and an additional 3.5 wk after injection.	• High-fat diet increased tumor size approximately 70% 3.5 wk after tumor injection and increased metastasis.
Wogsland et al. (39)	E0771	Regular chow or high-fat diet (60% fat) was provided for 10 wk before tumor injection and an additional 4 wk thereafter.	• High-fat diet accelerated tumor growth.
Zhang et al. (50)	E0771	Female mice were mated and a lard-based, high-fat diet (60.3% fat) or regular chow (10.5% fat) initiated. All mice were placed on regular chow after delivery.	• Offspring of mothers fed a high-fat diet had higher body weight and faster tumor growth.
Gu et al. (46)	E0771 in ovariectomized mice	Mice received a high- (60%) or low-fat (4%) diet beginning at 60 wk of age; 8 wk later they were injected with tumor cells and maintained on the diet for an additional 4 wk.	• Breast tumor weight in obese mice was twice that of lean controls. • There was a strong positive correlation between visceral fat mass and tumor weight. • Obese mice exhibited increased tumor angiogenesis and plasma and tumor VEGF concentrations.
Pingili et al. (49)	E0771	Mice received a high- (60% fat) or low-fat diet (10% fat) for 18 wk prior to tumor cell injection then continued on the diets for an additional 3 wk.	• Tumor growth was accelerated in high-fat fed mice (approximately sixfold increase in tumor volume 3 wk after tumor cell injection).
Mentoor et al. (66)	E0771	Mice were fed a low- (10%) or high-fat (60%) diet for 8 wk prior to tumor injection and 4 wk thereafter.	• Doxorubicin was less effective to slow tumor growth in high-fat diet fed mice.
Nasiri et al. (2)	E0771	Mice were fed a low- or high-fat (60%) diet beginning the day of tumor injection and continued for 4 wk.	• Glucose wasting with dapagliflozin slowed tumor growth in obese mice.
Yelek et al. (47)	E0771, PY8819	Regular chow or high-fat diet (60% fat) was provided for 6 wk prior to tumor injection and continued until tumors reached 500 mm3.	• High-fat diet accelerated tumor growth in both models.

a

TNBC = triple-negative breast cancer; VEGF = vascular endothelial growth factor.

Table 2.

Preclinical studies on the impact of energy expenditure on TNBC tumor growth^a

Study	Tumor model	Lifestyle intervention	Outcomes
Matsumoto et al. (60)	4T1	Chow-fed mice underwent 20 min/d of vertical sine-wave vibration for 5 d/wk for 3 wk. It is not clear how soon the intervention began following tumor injection.	• Vibration exercise protected tumor-bearing mice from some but not all loss of bone mineral density.
Ahmadabadi et al. (51)	4T1	Chow-fed mice underwent high-intensity interval training (80%-95% VO2max) beginning 7-10 d after tumor injection once a tumor was visible.	• High-intensity interval training reduced tumor volume approximately 35% 4 wk after injection.
Jafari et al. (61)	4T1	Chow-fed mice were subjected to 6 wk of progressive interval training on a treadmill, beginning 2 wk after tumor cell injection.	• Interval training alone did not alter tumor growth rates, but combining exercise with vitamin D slowed tumor growth.
Ma et al. (52)	4T1	Chow-fed mice underwent low- (3 m/min), moderate- (6 m/min), or high-intensity (10 m/min) treadmill exercise, 1 h/d each day for 20 d. Subsequently, tumors were injected and allowed to grow for 20 d; exercise was not continued.	• Exercise showed a dose-dependent effect to slow tumor growth, with the high-intensity intervention reducing tumor volume approximately 20%.
Nezamdoost et al. (53)	4T1	Chow-fed mice underwent high-intensity interval training (80%-95% VO2max). The frequency and duration of the training is not clear.	• Exercise slowed tumor growth (60% reduction in tumor volume after 4 wk). • Exercise increased food intake, but there was no correlation between food intake and tumor volume.
Garritson et al. (62)	4T1	Chow-fed mice were given ad lib access to running wheels in their home cages or no running wheels. Tumor cells were injected after 6 wk, and the mice retained access to wheels (or not) until they were euthanized up to 28 d later.	• Exercise tended to slow but did not statistically significantly alter tumor growth rates.
Wang et al. (54)	4T1	Mice received low- (6 m/min), medium (10 m/min) or high-intensity treadmill exercise (15 m/min) for 20 consecutive days prior to tumor cell injection. They did not exercise after tumor injection and were euthanized 22 d later.	• Exercise reduced tumor volume by approximately 20% at the endpoint.
Vulczak et al. (55)	4T1	Chow-fed mice ran on a treadmill 18 m/min for 30 min, 5 times per week. Mice were randomized to 1 of 4 groups: sedentary throughout, exercise throughout, exercise only prior to tumor injection, or exercise only after tumor injection (after 8 wk of exercise). Mice were monitored for another 4 wk.	• Compared with mice that were sedentary throughout, mice that were exercised throughout had 40% smaller tumors at the endpoint, and mice exercised only after tumor implantation had 30% smaller tumors. • Exercise only prior to tumor injection did not affect tumor size at the endpoint.
Kim et al. (56)	4T1	Mice were fed high-fat diet and exercised on a treadmill at low- (10 m/min) or moderate-intensity (15 m/min) for a total of 13 wk; 8 wk after initiation of the diet and exercise protocol, tumor cells were injected, and the diet and exercise protocol continued thereafter.	• Low-intensity exercise delayed the detection of palpable tumor and reduced tumor volume by 50% 5 wk after injection. • Surprisingly, moderate-intensity exercise had a lesser effect, reducing tumor volume by 20% without a delay in tumor initiation.
Wang et al. (63)	4T1	Mice were given access to running wheels for 10 d before tumor cell injection and 7 d thereafter.	• Access to running wheels did not alter tumor volume, though the duration of exposure may have been insufficient to detect differences.
Bianco et al. (57)	4T1	Mice received swim training 5 times per week for 4 wk, beginning at the time of tumor injection.	• Exercise reduced tumor volume by 40% after 4 wk.
Smeda et al. (64)	4T1	Chow-fed mice were injected with tumor cells simultaneously to being assigned to ad lib running wheels. Mice were euthanized 5 wk later.	• Tumor volume was not different between sedentary and exercised mice. • Pulmonary metastasis did not differ between groups.
Molanouri Shamsi et al. (65)	4T1	Mice ran on the treadmill (10 min warm-up, and 10 two-minute intervals of running at 70% Vmax separated by 2 min of active recovery at 50% VO2max) for a total of 12 wk. Tumors were injected after 6 wk.	• Tumor volume was not different between sedentary and exercised mice.
Betof et al. (58)	4T1	Chow-fed mice were randomized to ad lib running wheels at the time of tumor injection, and terminal studies performed 18 d thereafter.	• Exercise reduced tumor volume by approximately 20%.
Goh et al. (219)	4T1	Chow-fed mice had access to running wheels for a total of 90 d (60 d before tumor cell injection).	• Precancer running distance predicted lower tumor mitotic index; however, tumor volume is not reported.
Wennerberg et al. (69)	4T1	Chow-fed mice were exercised on a treadmill (18 m/min for 30 min) beginning 8 d after tumor implantation.	• Exercise enhanced the efficacy of anti–PD-1 immunotherapy.

a

PD-1 = programmed death 1; TNBC = triple-negative breast cancer; VO₂max = maximal oxygen consumption.

Table 3.

Preclinical studies on the impact of energy balance on TNBC tumor growth (interventions targeting energy intake and expenditure are shown)^a

Study	Tumor model	Lifestyle intervention	Outcomes
Turbitt et al. (71)	4T1.2	Regular chow-fed mice were randomized to sedentary or ad lib wheel cages and/or energy restriction (90% of ad lib dietary intake) for 13 wk. Tumor cells were injected after 8 wk on the intervention.	• Physical activity alone had no impact on tumor volume, but the combination of physical activity and energy restriction reduced tumor volume by approximately 40% at the endpoint.
Gomes-Santos et al. (59)	E0771	Mice were randomized to high-fat (60%) diet beginning 12 wk before tumor implantation and/or incremental treadmill exercise training at the time of tumor implantation.	• Exercise reduced tumor size by 30%. • Surprisingly, the impact of diet per se on tumor growth and the impact of combined dietary and exercise interventions were not reported.

a

TNBC = triple-negative breast cancer.

We acknowledge that publication bias may reduce negative studies appearing in the literature, however, we are encouraged by the uniformly positive effect of healthy body weight on tumor growth rates in preclinical observational studies that we identified in this narrative review. The results of preclinical studies testing interventions increasing energy expenditure are less homogeneous. Whereas some exercise studies have showed slowed tumor growth (51-59) and protection against loss of bone mineral density (60), in other studies, exercise did not slow tumor growth in mouse models of TNBC (61-65). These mixed results may be at least in part attributable to the models studied. Most of the published reports on exercise and TNBC progression have used lean mice, and it is possible that there could be a greater effect of exercise to improve systemic metabolism under conditions of overnutrition. The timing of exercise may also be important. Although exercise before and after tumor implantation slowed tumor growth, as did exercise only after tumor implantation, Vulczak et al. (55) observed no tumor-suppressive effect of exercise performed only prior to tumor cell injection. To our knowledge, no published preclinical studies have used more than 1 type of exercise, but several have examined a potential dose response to exercise. Ma et al. (52) found a dose-dependent effect of increasing treadmill running speed to slow 4T1 tumor volume. On the contrary, surprisingly, Kim et al. (56) found a greater effect of low-intensity exercise to slow tumor growth than moderate-intensity exercise. Finally, increased food intake in ad lib–fed, exercised mice may counteract a tendency for exercise to slow tumor growth (53). Consistent with this possibility, Nezamdoost et al. (53) demonstrated that exercise alone did not slow TNBC tumor growth, however, the combination of exercise and energy restriction counteracting the tendency of the exercised mice to increase food intake did slow tumor growth. Several preclinical studies have also examined the impact of approaches targeting energy balance on the effectiveness of cancer therapy. In E0771 TNBC, doxorubicin chemotherapy was less effective in obese as compared with lean mice (66), and in 4T1 breast cancer, glucose wasting with dapagliflozin enhanced the efficacy of paclitaxel chemotherapy in overweight and lean mice (45). Additional studies will be required to determine whether these beneficial effects are due to negative energy balance per se and/or to mechanisms downstream of negative energy balance, including reductions in whole-body adiposity, improved insulin sensitivity, and associated effects.

Preclinical impact of altering energy balance on immune function in TNBC

Our review identified only 3 published preclinical studies that have examined the impact of energy balance on the effectiveness of immunotherapy: 2 targeting diet or food intake and/or absorption and 1 targeting exercise. Gibson and colleagues (67) found that obese mice exhibited a defective response to immunotherapy, but depleting immune-suppressive myeloid-derived stem cells (MDSCs) improved the response to immunotherapy. In a similar vein, the efficacy of immunotherapy against E0771 breast cancer was improved by bariatric surgery resulting in weight loss (68). Targeting energy balance with a different intervention, Wennerberg et al. (69) found that treadmill exercise increased the efficacy of anti–PD-1 treatment. Although Pingili et al. (49) did not compare the efficacy of immunotherapy in lean vs obese mice, this group did document that anti–PD-1 therapy was effective at reversing some of the high-fat diet–induced markers of immune suppression in obese animals with E0771 breast cancer. In addition, several preclinical studies have examined the effect of changes in energy intake or expenditure (Tables 4-6) on markers of immune activation and function. Tumor-bearing mice fed a high-fat diet exhibited increased expression of the activating natural killer (NK) cell receptor NKG2D ligand (41) and PD-L1–positive dendritic cells and immunosuppressive granulocyte-like MDSCs (39,49,67) and decreased CD8 T cells (39) and monocytes (70). Depleting MDSCs slowed 4T1 tumor growth in obese mice, although growth rates were still more rapid than were observed in lean mice (48). Taken together, these results suggest that overweight may impair the antitumor immune response. Conversely, increasing energy expenditure with exercise reduced the accumulation of immature myeloid cells and immunosuppressive MDSCs (62,69), reduced M2 macrophage polarization (56), and reduced regulatory T cells in spleen (57), and it increased tumor CD8 T-cell infiltration (59). The combination of exercise and calorie restriction had a greater effect than either intervention alone to increase CD4 T-cell proliferation (71).

Table 4.

Preclinical studies on the impact of interventions targeting energy intake and obesity on immune cell infiltration and function in TNBC^a

Study	Tumor model	Lifestyle intervention	Outcomes
Spielmann et al. (41)	4T1 in ovar-iectomized mice	Mice were randomly assigned to receive regular chow or a 50% fat high-fat diet (full nutrient composition of the diets was not reported) for 13-14 wk, after which they were ovariectomized. Tumor cells were injected 3 wk thereafter, and the mice were kept on the obesogenic diet for 20-21 wk.	• High-fat diet fed mice exhibited reduced NKp46 expression on circulating NK cells. • High-fat diet fed mice exhibited increased expression of the activating NK cell receptor NKG2D ligand.
Clements et al. (48)	4T1	Low- (10% fat) or high-fat diet (60% fat) was provided for 12 wk, and tumors injected after 8 wk.	• High-fat diet increased circulating MDSCs. • Depleting MDSCs slowed tumor growth in high-fat diet fed mice; however, high-fat fed mice still had more rapid tumor growth than low-fat fed mice.
Núñez-Ruiz et al. (70)	E0771	Regular chow (10% fat) or high-fat diet (60% fat) was provided for 25 wk prior to tumor injection and an additional 3.5 wk after injection.	• High-fat diet reduced circulating regulatory T cells and decreased monocytes. • High-fat diet did not alter tumor-infiltrating regulatory T cells.
Wogsland et al. (39)	E0771	Regular chow or high-fat diet (60% fat) was provided for 10 wk before tumor injection and an additional 4 wk thereafter.	• High-fat diet increased PD-L1–positive dendritic cells and granulocyte-like myeloid-derived stem cells and decreased CD8 T cells.
Zhang et al. (50)	E0771	Female mice were mated and a lard-based, high-fat diet (60.3% fat) or regular chow (10.5% fat) initiated. All mice were placed on regular chow after delivery.	• Offspring of mothers fed a high-fat diet exhibited lower granzyme B and strongly tended to have lower interferon-gamma but did not show differences in the frequency of CD4, CD8, or Treg cells or CD8 T-cell exhaustion.
Mentoor et al. (66)	E0771	Mice were fed a low- (10%) or high-fat (60%) diet for 8 wk prior to tumor injection and 4 wk thereafter.	• Inflammatory cytokine concentrations (TNFα, IL-6, MCP-1) were not different. It is not clear whether these were measured in plasma or in the tumor.
Gibson et al. (67)	E0771	Mice were fed a low- (14%) or high-fat (60%) diet for 20 wk. Tumor cells were injected after 16 wk on the diet. Immunotherapy (adenovirus encoding murine TNF-related apoptosis-inducing ligand) was administered to a subset of mice.	• Obesity increased CXCL2-mediated chemotaxis and accumulation of granulocytic MDSCs in the tumor microenvironment. • Obesity caused hyperactivation of CD8 T cells but also increased CD8 T-cell apoptosis. • Obesity-induced resistance to immunotherapy but disrupting tumor CXCL2a antagonism improved the response to immunotherapy.
Pingili et al. (49)	E0771	Mice received a high- (60% fat) or low-fat (10% fat) diet for 18 wk prior to tumor cell injection then continued on the diets for an additional 3 wk.	• Anti–PD-1 reversed obesity-associated immunosuppression (ie, increased tumor dendritic cell infiltration and the ratio of M1/M2 macrophages). • Obesity increased MDSC and M2 macrophage content, but these immunosuppressive changes were reversed with anti–PD-1 treatment.

a

CXCL2a = chemokine (C-X-C motif) ligand 2a; IL-6 = interleukin-6; MCP-1 = monocyte chemotactic protein-1; MDSC = myeloid-derived stem cell; NK = natural killer; NK2GD = natural killer group 2D; PD-1 = programmed death 1; TNBC = triple-negative breast cancer; TNF = tumor necrosis factor; Treg = regulatory T cell.

Table 5.

Preclinical studies on the impact of interventions targeting energy expenditure on immune cell infiltration and function in TNBC^a

Study	Tumor model	Lifestyle intervention	Outcomes
Garritson et al. (62)	4T1	Chow-fed mice were given ad lib access to running wheels in their home cages or no running wheels. Tumor cells were injected after 6 wk, and the mice retained access to wheels (or not) until they were euthanized up to 28 d later.	• Exercise delayed the accumulation of immature myeloid cells and immunosuppressive MDSCs in tumors.
Kim et al. (56)	4T1	Mice were fed high-fat diet and exercised on a treadmill at low- (10 m/min) or moderate-intensity (15 m/min) for 13 wk; 8 wk after initiation of the diet and exercise protocol, tumor cells were injected, and the diet and exercise protocol continued thereafter.	• Low- and moderate-intensity exercise decreased the number of M2 but did not change M1 macrophages. • Incubation in the myokine myostatin reduced M2 polarization in U937 cells.
Bianco et al. (57)	4T1	Mice received swim training 5 times per week for 4 wk, beginning at the time of tumor injection.	• Exercise reduced splenic Treg and CD4+ IL-10+ cells. • Exercise reduced tumor dendritic cells.
Wennerberg et al. (69)	4T1	Chow-fed mice were exercised on a treadmill (18 m/min for 30 min) beginning 8 d after tumor implantation.	• Exercise reduced PD-1+ CD8+ T cells and NK cells. • Exercise reduced MDSCs.

a

IL-10 = interleukin 10; MDSC = myeloid-derived stem cells; NK = natural killer; PD-1 = programmed death 1; TNBC = triple-negative breast cancer; Treg = regulatory T cell.

Table 6.

Preclinical studies on the impact of interventions targeting energy intake and expenditure on immune cell infiltration and function in TNBC^a

Study	Tumor model	Lifestyle intervention	Outcomes
Turbitt et al. (71)	4T1.2	Regular chow-fed mice were randomized to sedentary or ad lib wheel cages and/or energy restriction (90% of ad lib dietary intake) for 13 wk. Tumor cells were injected after 8 wk on the intervention.	• Physical activity alone had no impact on CD4+ T-cell proliferation, but energy restriction increased CD4+ T-cell proliferation, and the combination of physical activity and energy restriction had a greater effect to increase CD4+ T-cell proliferation.
Gomes-Santos et al. (59)	E0771	Mice were randomized to high-fat (60%) diet beginning 12 wk before tumor implantation and/or incremental treadmill exercise training at the time of tumor implantation.	• Exercise increased tumor CD8+ T-cell infiltration and effector function.

a

TNBC = triple-negative breast cancer.

Clinical observational studies of energy balance and TNBC

Body weight

A BMI higher than 30 kg/m² at diagnosis correlates with the presence of larger tumors with higher T-stage and increased tumor grades (23,26,27,72) in patients with breast cancer. This increased aggressiveness correlates to at least a 30% increased risk of breast cancer recurrence or death (73,74). Numerous potential mechanistic hypotheses may be devised from the correlation between higher BMI and worsened breast cancer risk and outcomes. Obesity is a condition of chronic inflammation, which is characterized by increased levels of circulating inflammatory cytokines that can promote local inflammation in adipose tissues (23,75,76). Adipose tissue inflammation is linked to metabolic syndrome, which includes disorders such as insulin resistance, dyslipidemia, and visceral obesity, all of which may advance tumor growth through dysregulated adipokine, hormone, and insulin signaling. Given the multiple ways in which energy imbalance may promote TNBC tumor growth (37), it is important to gain a deeper understanding of the interrelationship between energy balance and TNBC risk and outcomes.

Exercise

Exercise has long been recognized to be associated with a lower risk of cancer, cancer recurrence, and cancer-specific mortality (20,77-81), with a recent review concluding that physical activity had the greatest effect of any of the lifestyle factors examined to reduce the risk of breast cancer recurrence and death (82). This relationship between physical activity and modest protection from breast cancer incidence can be observed even without any differences in body composition: a recent meta-analysis by Pizot and colleagues (83) found that regular physical activity reduced the risk of all breast cancer for women of any age and body weight by 12% and reduced the incidence of ER- and PR-negative breast cancer by 20%.

The American Society for Clinical Oncology (84) and American College of Sports Medicine (85) have recently published guidelines for diet, exercise, and weight maintenance during breast cancer treatment and in survivors. These and most other cancer guidelines recommend breast cancer patients who are overweight to lose weight and those with a healthy BMI to maintain a stable body weight. This can be particularly challenging for breast cancer patients: one-third of patients gain more than 5% of their baseline body weight within 3 years of diagnosis, particularly in patients within the healthy weight range (86). Unfortunately, the relationship between overweight and obesity and TNBC outcomes is not completely clear; although Mowad et al. (27) showed that TNBC patients with overweight or obesity exhibited a trend toward worse survival as compared with patients with a healthy weight, this difference was not statistically significant. However, increased adiposity tended to have a greater effect on negative vs positive hormone receptor–driven tumors (87). A multisite cohort study revealed that the risk of postmenopausal breast cancer—not necessarily TNBC—was 42% lower in patients who had undergone bariatric surgery prior to diagnosis as compared with matched participants who had not undergone bariatric surgery (88). Taken together, these data suggest that optimizing energy balance may be of particular importance in human breast cancer patients treated with immunotherapy. Future prospective trials will be of great importance.

Dietary composition

Caloric intake, rather than energy expenditure, can explain the preponderance of variance in body weight (89). Although many observational and fewer interventional studies on diet and breast cancer incidence and outcomes have been completed, surprisingly few have focused on TNBC. The published clinical studies on diet and breast cancer, including TNBC and stratifying by subtype, are shown in Table 7. Two observational studies in which participants were interviewed about their dietary habits failed to elucidate any link between dietary fat, meat, fish, or vegetables and TNBC risk (90,91), although high fruit intake was associated with a lower occurrence of TNBC in a case-control study of Vietnamese women (91).

Table 7.

Studies of diet and breast cancer outcomes^a

Study	Study design	Methods	Subtype(s)	Outcomes
Holm et al. (90)	Retrospective cohort	• 240 women aged 50-65 y with early stage breast cancer were interviewed about their dietary habits at diagnosis.	Breast cancer, stratified by subtype	• Lower dietary fat intake tended to correlate with improved survival, but only in ER-positive tumors not ER-negative tumors.
Nguyen et al. (91)	Case-control	• 476 Vietnamese women with breast cancer (45 with TNBC) and 454 age-matched participants without cancer were interviewed about their dietary habits over the past 5 y.	Breast cancer, stratified by subtype	• High fruit intake was associated with an 80% lower risk of TNBC. • There was no association between TNBC risk and intake of meat, fish, or vegetables.
Chlebowski et al. (93)	Randomized controlled trial	• 362 women with early stage ER- and PR-negative breast cancer • Participants were randomly assigned in the Women’s Intervention Nutrition Study to counseling to reduce dietary fat intake to 15% while maintaining adequate nutrition. • Participants were not asked to reduce total caloric intake.	Breast cancer, stratified by subtype	• The dietary intervention tended to exert a greater effect to reduce breast cancer recurrence in women with ER- and PR-negative vs receptor-positive tumors, but the interaction between hormone receptor status and dietary intervention was not statistically significant.
Pierce et al. (94)	Randomized controlled trial	• 619 women with ER- and PR-negative breast cancer • Survivors of early stage breast cancer were randomly assigned in the Women’s Healthy Eating and Living trial to counseling to implement a diet high in vegetables, fruit, and fiber and low in fat or to maintain their usual diet. • Participants were not asked to reduce total caloric intake.	ER- and PR-negative breast cancer	• The dietary intervention did not alter the risk of breast cancer recurrence or mortality.

a

ER = estrogen receptor; PR = progesterone receptor; TNBC = triple-negative breast cancer.

Clinical interventional studies of energy balance and TNBC

Energy composition and intake

Numerous confounders complicate the relationship between reported dietary nutrient intake and breast cancer incidence and outcomes. Among adult cancer survivors, adhering to a healthy diet was associated with higher levels of education, higher family income, non-Hispanic White race, and older age (92), all of which are associated with better outcomes in TNBC. Therefore, interventional studies of the relationship between energy balance and TNBC outcomes are of great interest. Although interventions reducing dietary fat without reducing total calorie intake failed to improve outcomes (93,94), the lone study randomizing participants to an intervention aiming to generate an energy deficit with primary outcomes changes in body fat, physical fitness, and quality of life did improve quality of life associated with a reduction in body fat (95) (Table 8). Unfortunately, this study did not capture survival outcomes and would have been underpowered to detect differences in survival, but nevertheless it highlights the opportunity to examine the impact of energy balance interventions on quality of life, as we will discuss later in this narrative review.

Table 8.

Studies correlating physical activity to TNBC risk and outcomes

Study	Study design	Methods	Subtype(s)	Outcomes
Trivers et al. (24)	Case-control	• Data from a multicenter population-based case-control study in younger women (aged 20-54 y), 135 of whom were diagnosed with TNBC • Self-reported physical activity data were included.	Breast cancer, stratified by subtype	• Above the median level of physical activity the year before the interview was associated with decreased risk of TNBC.
Ellingjord-Dale et al. (108)	Case-control	• Data from participants in the Cancer Registry of Norway were analyzed, 361 of whom had TNBC. • Participants without a cancer diagnosis were asked to complete a questionnaire reporting their current hours of weekly physical activity when they came for a screening mammogram.	Breast cancer, stratified by subtype	• There was no effect of physical activity on the risk of TNBC, but high levels of physical activity did reduce the risk for luminal A–like ER+PR+HER2− breast cancer.
Ma et al. (97)	Prospective cohort	• Patients were asked to complete questionnaires reporting physical activity long-term (from high school through age 54 y or age at cohort entry, whichever was younger) and baseline (during 3 y prior to baseline). • Of the 108,907 women in the analysis, 348 developed TNBC.	Breast cancer, stratified by subtype	• Women in the highest group of self-reported strenuous physical activity had a lower risk of TNBC. • Baseline physical activity and long-term physical activity tended to lower the risk of TNBC, but this did not reach statistical significance.
Bigman et al. (98)	Case-control	• Nigerian women with breast cancer reported their leisure-time physical activity, which was converted to MET-h. • 123 women with TNBC were included.	Breast cancer, stratified by subtype	• Women in the highest quartile of physical activity had a 49% lower risk of TNBC.
Beasley et al. (99)	Prospective pooled cohort of breast cancer patients	• Physical activity between 18 and 48 mo postdiagnosis was reported and converted into METs. • Meeting physical activity guidelines was defined as completing ≥10 MET-h per wk.	Breast cancer, stratified by hormone receptor status (HER2 not assessed)	• Risks of all-cause mortality and breast cancer mortality were reduced by 28% and 36%, respectively, in ER- and PR-negative breast cancer patients who met physical activity guidelines.
Chen et al. (100)	Prospective cohort of breast cancer patientsa	• 1494 women with ER- and PR-negative breast cancer were asked to report their physical activity 6, 18, and 36 mo postdiagnosis.	Breast cancer, stratified by hormone receptor status (HER2 not assessed)	• ER- and PR-negative patients at the highest level of physical activity exhibited 48% and 55% reductions in all-cause and disease-specific mortality, respectively • ER- and PR-negative patients were included, but TNBC patients were not studied.
Dixon-Suen et al. (220)	Case-control	• 4964 women with TNBC were included. • Mendelian randomization analysis of case-control data from the Breast Cancer Association Consortium, associating SNPs correlated with accelerometer data, and the full instrument (5 SNPs) correlated with self-reported vigorous physical activity (≥3 vs 0 d/wk).	Breast cancer, stratified by subtype	• Greater genetically predicted sedentary time was associated with a statistically non-significant twofold risk for TNBC with the single SNP. • No association with risk for TNBC was observed with the full instrument.
Bao et al. (101)	Prospective cohorta	• 518 women with metastatic TNBC were asked to report their physical activity 6, 18, 36, and 60 mo postdiagnosis.	TNBC	• Both duration and MET-h per week of physical activity at 6, 18, 36, and 60 mo were inversely correlated with overall and disease-free survival.
Phipps et al. (103)	Prospective cohort study combined with Women’s Health Initiative randomized clinical trials	• Data from postmenopausal women in the observational study and the randomized clinical trials within the Women’s Health Initiative (307 women with TNBC) were asked to report physical activity at the time of enrollment. • Incident cases of breast cancer were identified in the mean 7.9 y follow-up. • Participants completed questionnaires on the frequency and duration of physical activity, and these data were converted to MET-h.	Breast cancer, stratified by subtype	• There was a modest, non-statistically significant inverse relationship between weekly physical activity and TNBC risk.
Ma et al. (104)	Retrospective case-control	• Data from the population-based Women’s Contraceptive and Reproductive Experiences case-control study, including 390 women with TNBC • Self-reported physical activity (type, age at which the woman started and stopped the activity, months of participation per year, and average hours per week) was converted to MET-h at each year of age.	Breast cancer, stratified by subtype	• There was a modest, non-statistically significant inverse relationship between lifetime physical activity and TNBC risk.
Schmidt et al. (105)	Retrospective chart analysis	• Data from 197 women with TNBC were analyzed. • Participants were stratified based on whether or not they reported regular physical activity (≥2 times per week).	TNBC	• Regular physical activity did not affect overall or disease-free survival. • This study is limited by its relatively restricted assessment of physical activity.
Delrieu et al. (107)	Prospective cohort study	• 100 women with metastatic TNBC were asked to report their physical activity at the time of enrollment, and the data were converted to MET per min and per week.	TNBC	• Physical activity was not associated with survival in patients with metastatic TNBC.
Holmes et al. (106)	Prospective cohort study of breast cancer patients	• Leisure-time physical activity data collected longitudinally from women with stage I-III breast cancer in the Nurses’ Health Study were analyzed. • 421 women with metastatic ER- and PR-negative breast cancer were included.	Breast cancer, stratified by hormone receptor status (HER2 not assessed)	• Breast cancer mortality did not differ based on activity (<9 MET-h per week vs >9 MET h per week) in patients with ER- and PR-negative tumors. • Patients with ER- and PR-negative tumors were included, however, TNBC patients were not studied specifically.
Courneya et al. (102)	Randomized clinical trial	• Women with stage I-IIIA breast cancer, including 62 women with TNBC, were randomly assigned to receive aerobic exercise training, resistance training, or usual care for the duration of their chemotherapy. • Fitness was considered in the aerobic-trained group: participants trained at 60% to 80% VO2 peak.	Breast cancer, stratified by subtype	• No differences in disease-free survival or recurrence-free survival were observed in patients with TNBC.
Swisher et al. (95)	Randomized controlled trial	• 28 survivors of TNBC were randomly assigned to complete moderate-intensity (150 min/wk, targeted to achieve 50%-65% maximal heart rate for 12 wk) aerobic exercise and diet counseling (goal to decrease dietary fat caloric intake by 200 kcal/wk).	TNBC	• Participants in the intervention group lost more body fat (2.4% vs 0.9%). • The intervention improved quality of life, as assessed on the FACT-B questionnaire.

a

Reports are from the same study. ER = estrogen receptor; FACT-B = Functional Assessment of Cancer Therapy—Breast; MET = metabolic equivalent of task; PR = progesterone receptor; SNP = single nucleotide polymorphism; TNBC = triple negative breast cancer; VO₂peak = peak oxygen consumption.

Clinical interventional studies of energy balance and TNBC

Energy expenditure

To date, there is considerable heterogeneity in the published observational studies of physical activity or exercise and TNBC and incidence of or outcomes in patients with ER- and PR-negative breast cancer (Table 8). Some of the published studies on this topic that we are aware of showed benefit (24,96-101), whereas others did not (102-108). The heterogeneity may be partially attributed to limited numbers of patients with TNBC: only approximately 10% of breast cancers are TNBC, so studies that do not specifically aim to recruit a critical mass of patients with TNBC are likely underpowered to capture differences in outcomes in patients with this breast cancer subtype. Though the mechanisms by which exercise affects cancer risk and prognosis remain elusive and are likely multifactorial, the ability of exercise to affect energy balance indicates that it is capable of changing body composition by decreasing body fat mass and increasing fat-free mass, which, in turn, will increase energy expenditure. The effects of exercise on energy balance are generally more profound in individuals with obesity as compared with normal weight individuals (109,110). It is also important to note that, to our knowledge, no study has measured total daily energy expenditure in breast cancer patients or survivors, as was recently noted (111); for this reason, although it stands to reason that increasing energy expenditure using recreational physical activity would improve outcomes, we lack any foundational data in this area. However, given exercise’s ability to directly alter energy balance by altering body composition and to potentially improve the above-mentioned relationship between TNBC and obesity, it is important that we pursue rationally designed studies to understand the interrelationships between these factors in patients with TNBC.

Many breast cancer patients undergoing therapy reduce their daily energy expenditure, which is associated with loss of muscle mass (112). Breast cancer or breast cancer treatment reduces cardiorespiratory fitness. In one study, the majority (22 of 30) of breast cancer survivors had a peak oxygen consumption (VO₂peak) beneath the 20th percentile (113), with the average VO₂peak 22% lower in breast cancer survivors than their age-matched, sedentary, healthy counterparts (114). Importantly, these studies did not stratify by breast cancer subtypes. Intriguingly, hormone receptor status correlated with reported intensive physical activity (ie, patients with ER- and PR-negative breast cancer were more likely to complete intensive physical activity) (115). However, no relationship was observed between HER2 status and intensive physical activity.

In parallel, breast cancer survivors are at a higher risk of cardiovascular mortality than their healthy peers (116). In particular, African American breast cancer survivors are at a fourfold higher risk of cardiovascular mortality as compared with similar patients without African ancestry (117). Although this study did not specifically examine TNBC survivors, African American women have a threefold higher risk of developing TNBC as compared with White women (118) and have approximately 40% higher breast cancer mortality rates (119). Taken together, these data suggest that breast cancer patients and survivors with low cardiorespiratory fitness would be at greater mortality risk, and indeed, a prospective study of patients with breast cancer—not exclusively TNBC—did demonstrate an increased risk of death in those with low fitness (120).

These data imply that interventions to improve cardiorespiratory fitness by optimizing whole-body energy balance may yield survival benefits in breast cancer survivors and, perhaps, patients. Before examining the impact of physical activity interventions on survival, the field must first be confident that physical activity interventions can improve fitness in this high-risk patient population. Several recent studies have, indeed, demonstrated an improvement in VO₂peak resulting from an exercise intervention in breast cancer patients and survivors (121-127). Importantly, breast cancer patients who were exercising and currently undergoing treatment could maintain and even increase their VO₂peak throughout the intervention (123). It should be noted that more than 80% of cancer patients are unable to complete the VO₂max or VO₂peak test (128), so surrogate tests are of great importance. The 6- and 12-minute walk tests (6MWT and 12MWT, respectively) are the most commonly used tests that indirectly measure aerobic power (129), particularly on a population scale. Five interventional studies reported the impact of a physical fitness intervention on breast cancer patients’ performance in the 6MWT or 12MWT. Each of these found a beneficial effect of the exercise intervention on performance (127,130-133), whereas none found a non-statistically significant or negative relationship between exercise interventions and performance in the 6MWT or 12MWT. These data demonstrate that the low-cost, highly scalable option of a 6MWT or 12MWT is sufficiently sensitive to assess the impact of a physical fitness intervention in breast cancer patients. As with many other topics discussed in this narrative review, to our knowledge there have been no trials specifically examining cardiorespiratory fitness in TNBC patients treated with immunotherapy. Such trials will be of great importance moving forward, as virtually all patients with advanced PD-L1–positive TNBC are now treated with immunotherapy in combination with chemotherapy as preferred first line of therapy, and such therapy is also the standard of care for high-risk, early stage TNBC regardless of PD-L1 status.

We identified 2 interventional studies that have been completed using specific exercise prescriptions (aerobic and strength training) in TNBC patients and survivors in which primary outcomes included muscle strength and cardiovascular fitness. Most of the outcomes in these studies were improved in intervention arms compared with the control participants. Swisher and colleagues’ study (95) was the only reviewed trial focused solely on TNBC patients. Although this study did not evaluate cancer outcomes, the authors observed a reduction in body fat in the intervention arm and addition of body fat in the control arm and an improvement in quality of life in the intervention arm. Peak exercise capacity, however, did not change. Travier et al. (134) studied breast cancer patients with multiple tumor subtypes, but twice as many patients with TNBC were randomly assigned to the intervention arm rather than the control arm. Although the results were not stratified by breast cancer subtype, overall the intervention increased fitness and leg strength and reduced reported fatigue. Additionally, Nock et al. (133) examined the effectiveness of community-based exercise programs in African American breast cancer survivors and found that exercise programs were well embraced, with 70% adherence to the exercise regimen (133) and a reduction in self-reported depression symptoms in the intervention arm (135). This interventional study is important as it isolates a specific group of individuals at high risk for developing aggressive breast cancer subtypes and assesses how exercise affects their health outcomes. These studies demonstrate the value of physical activity in breast cancer rehabilitation. However, neither focused on TNBC patients treated with immunotherapy. There is, therefore, an urgent need and promising opportunity to add substantially to what is known about exercise interventions for TNBC patients taking checkpoint inhibitors and to broaden the studied interventions to include work on modulating energy balance in TNBC patients.

Toxicity associated with breast cancer treatment

Checkpoint inhibitors have yielded promising results in metastatic breast cancer, however, they are associated with clinically significant toxicities. In fact, adverse effects occur in more than 90% of breast cancer patients treated with immunotherapy (136-138). Although most are low grade, these effects can substantially reduce quality of life and could lead to poor treatment adherence. To our knowledge, no studies thus far have specifically examined the impact of energy balance on quality of life in patients receiving immunotherapy. However, the literature on dietary modifications and increased physical activity, which will be discussed in the following sections, would strongly suggest that optimizing energy balance would improve quality of life and mitigate common cancer-related symptom toxicities in patients treated with immunotherapy.

Cancer-related fatigue (CRF) is almost universal in patients during cancer treatment; 99% of breast cancer patients reported CRF during treatment, and more than 60% rated their CRF as moderate to severe (139). Immune checkpoint inhibitors are associated with CRF in 30%-50% of patients with multiple tumor types (140). In addition to its effects to worsen psychological health and quality of life in patients (141-143), CRF can lead to a reduced likelihood that patients will benefit from potentially curative therapies. Cancer fatigue worsens adherence to lifestyle adaptations, which may improve outcomes, and standard-of-care cancer treatment (144,145); 10% of cancer patients reported that their treatments were delayed, changed, or stopped because of CRF (146).

Although our review did not identify any studies specifically examining the impact of physical activity interventions on CRF in TNBC patients treated with immunotherapy, there is considerable literature suggesting that increasing energy expenditure has a beneficial effect. Numerous studies of pedometer-based exercise programs reported the intervention’s effect on patients’ level of CRF (147-152). Studies have confirmed a beneficial effect of physical activity to minimize CRF and improve psychosocial health (149-151), and another found that Functional Assessment of Cancer Therapy—Breast scores improved in the intervention arm (147). It is possible, although not definite, that CRF accounted for part of the improvements observed, as CRF is one component of the Functional Assessment of Cancer Therapy—Breast assessment. In contrast, in a small randomized control trial, Gokal et al. (148) found no statistically significant change in CRF between those in the intervention arm, who started the trial walking for 10 minutes at any one time and steadily increased that time to 30 minutes 5 times a week, and those in the control arm, who participated in usual care. We are not aware of any studies finding negative correlations between daily physical activity and level of CRF during treatment. Additionally, dietary interventions may help treat CRF. Kleckner et al. (153) recently demonstrated that a Mediterranean diet intervention improved symptoms of CRF in breast cancer patients, although this study was not restricted to TNBC patients. Considering the substantially worse prognosis in TNBC as compared with other breast cancer subtypes, as well as the shifting therapeutic landscape considering the recent approval of checkpoint inhibitors for metastatic TNBC, it will be important to design trials that specifically examine the impact of interventions to optimize energy balance in patients treated with immunotherapy for TNBC.

CRF does not only affect breast cancer patients during their course of treatment; 15%-35% of adult cancer survivors experience chronic CRF or fatigue that lasts months or years after cancer diagnosis and treatment (154). A randomized controlled trial conducted by Zick et al. (155) found that in breast cancer survivors, a 3-month diet rich in fruit, vegetables, whole grains, and omega-3 fatty acid–rich foods improved fatigue symptoms by 44%. Alfano et al. (156) conducted a cross-sectional survey of long-term breast cancer survivors to assess changes in self-reported exercise behavior and current symptoms (women were on average 12 years since diagnosis). Those who reported an increase in exercise since diagnosis reported lower fatigue levels. This group also reported higher social support compared with those who did not increase exercise since diagnosis. However, it is unknown whether increased physical activity led to higher social support or if higher social support led to increases in physical activity as the data were cross-sectional. It is also unclear whether survivors who increased their exercise experienced a decrease in fatigue levels or whether those with high fatigue levels found it impossible to increase exercise. Additionally, trials parsing how modulating energy balance, not just diet and not just physical activity, are needed. Motivating to Exercise and Diet, and Educating to healthy behaviors After breast cancer is a multicenter, randomized controlled trial examining the effect of a combined diet and exercise intervention on CRF in overweight and obese patients with breast cancer (157) and will provide important data in this vein, but there remains the opportunity to pursue interventions to test the impact of negative energy balance on CRF in patients treated with immunotherapy specifically.

Additionally, data on the impact of diet and/or physical activity on other toxicities related to checkpoint inhibitor treatment (rash, myalgia, pneumonitis, autoimmune) are lacking in patients with TNBC. In examining colitis as a common side effect of checkpoint inhibitors, we must consider a potential bidirectional relationship between microbial shifts and the response and tolerability of immunotherapy; checkpoint inhibitors are known to cause microbial shifts (158), and recent studies demonstrated effects of differences in the gut microbiome, which are typically diet driven, on the efficacy of immunotherapy against melanoma (159-166), lung (167,168), and kidney cancer (167). Further, microbial signatures predict the occurrence of adverse effects of checkpoint inhibitor treatment in metastatic melanoma (164,165,169). Data on the impact of excess body weight on the incidence of adverse effects of checkpoint inhibition are mixed; whereas some studies have detected increases in the occurrence of adverse effects in overweight patients (170-172), others have not observed a difference (31,173-176). However, it is important to note that these studies have not focused on TNBC and, in most cases, have not included TNBC patients. As TNBC biology and host tumor–immune cell interactions are likely quite different from those observed in other immunogenic cancers, there is an important opportunity to examine how alterations in energy balance may affect the incidence of adverse effects of immunotherapy treatment beyond fatigue.

Exercise heterogeneity in TNBC patients

The goal of exercise therapy during TNBC treatment is typically to reduce cancer proliferation and cancer morbidity and mortality rather than create athletes with high recreational physical activities. The aim is to utilize doable interventions within the patient’s cardiorespiratory capacity to improve systemic energy balance and reduce the risk of advanced disease through low-cost, low-toxicity, and short-term lifestyle (physical activity) interventions. Effective-intensity training is feasible, safe, and effective for cancer patients, and a shorter time of training is likely sufficient to obtain benefits, which should be considered in the implementation of exercise strategies (177). Systematic reviews and meta-analyses also tend to aggregate exercise programs into general categories and rarely investigate the specific features of exercise programs that may make them more or less effective. Despite mixed results from different reviews and studies, the preponderance of evidence supports the need for adequately powered studies to investigate associations between high-intensity physical activity and TNBC outcomes. Also, clinical trials on the impact of exercise on tumor biology should continue to collect biomarker data to provide further insight into the mechanistic basis of exercise and interactions (178). More research is needed to determine the dose-response relationship, type, and frequency between exercise and survival during TNBC therapy.

Potential mechanisms for a link between exercise and improved prognosis in TNBC

Exercise is a safe and effective modifiable factor for adjuvant therapy provided that the energy expenditure is in line with or exceeds the energy intake of patients. Exercise can influence adiposity, reduce inflammation and other hormones, and have an overall positive effect on body composition. These effects could provide strategies to improve breast cancer prognosis, ideally using precision medicine metabolic approaches, as we have discussed previously (45). TNBC etiology is multifactorial, and there has been no conclusive evidence concerning the specific physiological mechanisms in which physical activity reduces its risk and improves its outcomes. However, it is postulated that physical activity beneficially affects adiposity, insulin resistance, and the production of adipokines and inflammatory markers. Each of these is a potential mechanism that could explain the beneficial effect of exercise on TNBC progression. The roles of adiposity (179), insulin and insulin-like growth factors (IGF) (180-182), adipokines (183-185), and inflammatory cytokines (186,187) have been reviewed previously and will be discussed further in the following subsections. A study by De Santi and his team (188) found that the proliferation of TNBC cells decreased when they were cultured and incubated in serum induced by exercise. These data suggest that a secreted protein(s) found in the serum of exercised participants may have antitumor properties, but this exercise factor (or factors) has not yet been identified.

Insulin and other hormones

Insulin promotes cellular development and growth and also exhibits anti-apoptotic and mitogenic effects in breast cancer (189-192). There are several putative mechanisms by which insulin may drive breast cancer, including TNBC. Hyperinsulinemia promotes the synthesis and action of IGF-1 (190). IGF-1 is a well-known mitogen that encourages cell growth, differentiation, and transformation and inhibits apoptosis, therefore promoting tumor cell division. It does this by modulating the availability of free IGF-1 present through its binding proteins (IGFBP1-6). Obesity-related chronic hyperinsulinemia lowers IGFBP levels, which in turn increases levels of free IGF-1 (193). The phosphatidylinositol-3-kinase, extracellular signal–regulated kinase (ERK), RAC(Rho family)-alpha serine/threonine-protein kinase (AKT), and mitogen-activated protein kinase (MAPK) pathways are activated when insulin binds to the insulin receptor. These pathways can be activated to cause growth, invasion, angiogenesis, and a reduction in apoptosis (193). Increased levels of insulin and IGF-1 may also increase the risk of TNBC via modification of circulating estrogen levels. For instance, insulin can inhibit the hepatic secretion of sex hormone–binding globulin and stimulate the activity of estrogen synthase resulting in an excess of free bioavailable estrogen (189,192). Insulin increases ER alpha-mediated transcription in breast cancer cells, and ER stimulates the insulin signaling pathway by increasing MAPK activation. Insulin and estrogen can work together to drive cell-cycle progression and growth in breast cancer cells (189). Davison et al. (194) found that 7 TNBC cell lines expressed IGF receptors. These intriguing data suggest that insulin-IGF cross-signaling can enhance proliferation and survival of breast cancer cells, including in TNBC models. Acute and chronic exercise are generally accepted to improve insulin sensitivity (195,196), including in breast cancer patients and survivors with multiple tumor subtypes (190,191,193). Numerous studies (193,197-201) have shown a statistically significant decrease in insulin levels and insulin resistance, thereby decreasing fasting glucose, insulin, and IGF-1 and increasing IGFBP-1 and IGFBP-3 following a high- or moderate-intensity exercise program. Improved insulin sensitivity leads to reductions in fasting and particularly postprandial insulin. Studies on exercise and weight loss in breast cancer have shown the suppression of IGF-1–related signaling pathways, such as protein kinase B anti-apoptosis, and rat sarcoma MAPK proliferation leads to cell cycle inactivation and cancer suppression (202).

High- or moderate-intensity physical activity may also increase glucose disposal and insulin sensitivity by a number of mechanisms including increased glucose transporter protein and mRNA, increased muscle glucose delivery because of increased muscle capillary density, increased activity of glycogen synthase and hexokinase, and increased postreceptor insulin signaling (203). The inhibitory effect of exercise on cancer cell proliferation in TNBC as well as other breast cancer subtypes may also be mediated by mammalian target of rapamycin targets (204,205). It is believed that exercise affects the activation of mammalian target of rapamycin signaling, which enhances apoptosis; modulates systemic signals such as IGF-1, insulin, metabolic, and inflammatory pathways; and reduces angiogenesis. Leptin is a peptide hormone produced by white adipose tissue. It regulates appetite and energy expenditure and has been suggested to have an impact on carcinogenesis, angiogenesis, immune responses, cytokine production, and other biological processes. In TNBC metastases, a decrease in the leptin-to-adiponectin ratio resulting from caloric restriction in turn decreases proliferation, increases apoptosis, and downregulates the IGF1-1R pathway (6,204). Likely as a result of the effect to reduce body fat, chronic increases in physical activity reduce circulating leptin levels (206-208), despite a seemingly paradoxical increase in plasma leptin observed acutely after a single bout of exercise in volunteers with high body fat percentages (209). Importantly, it is likely that there is not a single mechanism by which exercise slows tumor growth, much less a single hormonal mechanism. Just as cancer is broadly multifactorial, mechanisms that may mediate the effects of any intervention on breast cancer, including TNBC, likely are multifactorial as well.

Impact of exercise on inflammation and immune function

Studies have proposed that chronic inflammation may be one of the major factors that contribute to breast cancer development and progression. Tumor necrosis factor–α, interleukin-1 (IL-1), IL-6, and C-reactive protein (CRP) produced by T lymphocytes are biomarkers of inflammation that are elevated in patients with breast cancer (210,211). Pro-inflammatory cytokines like IL-1, IL-6, and tumor necrosis factor–α, which are increased in individuals with obesity, enhance the synthesis of CRP. Elevated CRP has been associated with worsened survival in breast cancer patients (212), although interventional studies to test a potential direct impact of CRP to promote tumor growth are lacking. In addition, NK cells are crucial in the prevention of early and metastatic breast cancer, at least in part by releasing cytolytic granules containing perforin, granzymes, and granulysin (213). Unfortunately, NK cell toxicity is limited in advanced breast cancer (214), contributing to immune evasion.

Physical exercise and weight loss resulting from caloric restriction can shift the immune milieu to a more favorable landscape in TNBC. Regular exercise may have an anti-inflammatory impact by decreasing visceral fat mass and increasing the number and activity of anti-inflammatory regulatory T cells while reducing the number and activity of pro-inflammatory cells (69,215). These improvements can translate to a reduction in fatigue and improved quality of life (216-218). In addition, myokines such as IL-6 that are released from active skeletal muscle during exercise, in turn, cause an increase in cortisol and adrenaline release. Taken together, these data indicate that exercise can stimulate the immune-cell mobilization response in the cancer microenvironment, including the infiltration of immune NK cells, the immune anti-inflammatory effect of cytokines, and the proliferation of T cells, thus inhibiting the growth of TNBC cells. With immunotherapy recently having become standard-of-care treatment for metastatic TNBC expressing PD-L1, it will be of great interest to understand whether exercise can enhance the response to checkpoint inhibitors in patients with breast cancer.

Summary and conclusions

Physical activity improves overall health through multiple mechanisms, and it has become increasingly clear that it deserves attention as an ant-cancer adjuvant therapy as well. The first PubMed citation for exercise prescription was published in 1969, but despite this, evidence-based guidelines for an exercise prescription in patients with TNBC are still lacking. It will be important to design interventional trials that will allow clinicians to partner with patients with TNBC to design specific exercise goals (type, duration, intensity) that are feasible and effective for the patient seeking to lower cancer risk or improve treatment efficacy and/or tolerability. Additionally, maintenance of a healthy body weight has clearly been shown to improve outcomes in breast cancer including in TNBC, however, more nuanced studies of specific dietary interventions will be necessary to better understand both mechanisms and potential therapeutic strategies for TNBC patients. This would position diet and exercise as a clinical and investigational intervention deserving of continued attention in patients with this disease.

Future directions

Remarkably few prospective, interventional studies have been conducted to determine the impact of changes in energy balance on outcomes in patients treated with immunotherapy for TNBC. It will be important to better understand how optimizing energy balance—intake and output—can affect outcomes in patients with this disease. In addition, future trials will be needed to determine whether prospective studies to optimize energy balance can reduce checkpoint inhibitor treatment toxicity and improve overall quality of life in cancer patients and survivors. In understanding these mechanisms, it is conceivable that new targets to improve survival and quality of life in TNBC patients will be uncovered.

Data availability

No new data were generated or analysed in support of this research.

Author contributions

Ngozi D. Akingbesote, BS (Investigation; Writing—original draft), Dennis Owusu, BS (Investigation; Writing—original draft), Ryan Liu (Investigation; Writing—original draft), Brenda Cartmel, PhD (Writing—review & editing), Leah M. Ferrucci, PhD, MPH (Funding acquisition; Writing—review & editing), Michelle Zupa, BS (Writing—review & editing), Maryam B. Lustberg, MD, MPH (Writing—review & editing), Tara Sanft, MD (Writing—review & editing), Kim R. M. Blenman, PhD (Visualization; Writing—review & editing), Melinda L. Irwin, PhD, MPH (Funding acquisition; Writing—review & editing), and Rachel J. Perry, PhD (Funding acquisition; Investigation; Writing—original draft)

Funding

This work was supported in part by a Pilot Grant (Team Challenge Award) from the Yale Cancer Center, by the Lion Heart Foundation, and by the Breast Cancer Research Foundation.

Conflicts of interest

The authors have no conflicts of interest to disclose.

Acknowledgements

Role of the funder: The funders had no role in the writing, editing, or decision to publish this review.