Introduction

Despite significant advances in the field of oncology, cancer remains the second leading cause of mortality and morbidity in the United States,1 accounting for an estimated 608,570 deaths in 2021 alone.2 The incidence rate is expected to increase with a rapidly aging population. The current estimated lifetime risk of being diagnosed with cancer is 40.14% for males and 38.70% for females,3 with a projected 22 million cancer survivors in the United States by 2030.4 Furthermore, even with a marked improvement in overall survival at 5 years, cellular and organ damage from chemotherapy and/or radiation therapy frequently results in decreased quality of life for survivors, with common residual physical symptoms such as neuropathy, fatigue, cognitive problems, and pain.5 Such physical and psychosocial symptoms often persist well beyond 5 years,6 with survivors reporting unmet needs even 10 years after treatment.7 Furthermore, approximately 8% of survivors develop a second cancer, one-half of whom are likely to die from the second malignancy8, 9

Fortunately, studies of monozygotic twins suggest that hereditary factors exert a small contribution to the risk of most neoplasms and that environmental factors play a principal role in this regard.10 In particular, it has been estimated that about 42% of all cancers and 45% of cancer deaths are attributed to modifiable lifestyle risk factors, including tobacco, physical inactivity, excessive adiposity, and dietary factors such as consumption of ultra-processed food and red and processed meat and low intake of plant foods rich in dietary fiber, antioxidant vitamins, and phytochemicals.11 Overweight and obesity alone are associated with an increased risk for at least 13 different cancers,12 and excess adiposity at the time of a cancer diagnosis is associated with poorer outcomes in most cancers.13 Thus there is an opportunity for both prevention efforts and improved cancer outcomes through healthful lifestyle.

Calorie restriction (CR) without malnutrition remains the most robust intervention to date for cancer prevention in rodents and monkeys, and, in humans, it promotes anticarcinogenic adaptations such as decreased production of growth factors, inflammatory cytokines, and anabolic hormones as well as decreased oxidative stress and free-radical–induced DNA damage.14 Despite a wealth of literature on the mechanisms and impact of CR, its clinical applicability remains limited because of challenges in long-term sustainability. Intermittent fasting (IF) is becoming a popular alternative to daily CR, with IF being the most frequently cited diet pattern in 2020 among Americans aged 18 to 80 years according to the International Food Information Council survey.15 It can occur in various forms, including fasting for 24 hours on alternate days, fasting 2 days per week on nonconsecutive days, or time-restricted feeding (TRF) (Table 1). In this review, we examine the data for different forms of fasting in rodents and humans, focusing our attention on the biologic adaptations that may reduce cancer incidence and improve cancer outcomes. We also highlight new emerging scientific trends on the role of prolonged fasting and fasting-mimicking diets (FMDs) as a potential new adjunctive therapy for patients undergoing chemotherapy.

TABLE 1. Key Definitions Intermittent fasting (IF) Episodic periods of little or no calorie consumption Time-restricted feeding (TRF) A form of intermittent fasting that requires limiting the consumption of calories to a window of time, typically between 4 and 12 h daily Prolonged fasting Prolonged, periodic fasting that lasts >24 h Fasting-mimicking diet (FMD) Generalized term for low-calorie diet that is low in protein and carbohydrates but high in unsaturated fat and provides between 10% and 50% of calories of normal ad libitum intake Obesity, Adiposity, and Cancer Risk and Prognosis

Cohort studies suggest a strong link between excess body weight and multiple types of malignancies, including postmenopausal breast, endometrial, ovarian, colorectal, liver, pancreatic, gallbladder, gastric cardia, esophageal adenocarcinoma, renal cell, meningioma, thyroid, and multiple myeloma.12 Accumulating data also suggest that obesity is associated with higher rates of cancer progression, recurrence, and mortality, especially for breast, prostate, and colon cancer.13, 16-20 Furthermore, cancer survivors are at higher risk of becoming obese,21 likely because of various factors, including the use of chemotherapy, steroids, and hormonal therapy, which can accelerate weight gain.22 Although the obesity paradox refers to studies demonstrating that obesity is associated with improved overall survival, this is more likely secondary to flaws in methodological mechanisms, such as using body mass index as an obesity measure, confounding and reverse causality.23-26 However, there are studies that postulate a biologic rational for the improved survival seen in obese patients with renal cell carcinoma.13, 27, 28 In those studies, the authors found that patients with renal cell carcinoma and obesity had longer overall survival than patients without obesity. One hypothesis based on transcriptomic signature differences in the primary tumor and the peritumoral adipose tissue is that increased tumor angiogenesis and increased peritumoral inflammation in the perirenal white adipose tissue of obese patients contribute to their survival advantage. Another study found that the perinephric fat contains increased numbers of activated immune cells.28

The mechanisms by which excessive adiposity affects cancer risk and prognosis are complex and continue to be elucidated. Chronic inflammation, insulin resistance, and altered sex hormone metabolism appear to be key factors.29 Weight gain results in adipose tissue hypertrophy and immune infiltration, increased production and decreased clearance of free fatty acids,30 and changes in proinflammatory cytokines and adipokine signaling, which lead to systemic insulin resistance.31, 32 Multiple mechanisms linking insulin dysregulation and cancer have been proposed. For example, compensatory hyperinsulinemia promotes cell proliferation and protects mutated cells from apoptosis through activation of the PI3K/AKT pathway and has been associated with increased risk of cancer recurrence and death.33-35 Insulin resistance also increases levels of bioavailable sex hormones and insulin-like growth factor 1 (IGF-1) by reducing liver production of SHBG,36, 37 IGF-binding protein 1 (IGFBP-1), and IGFBP-2.38 In addition, excessive adiposity raises circulating estrogens through increased aromatase expression.39 Altered plasma concentrations of estrogen-related and androgen-related hormones and IGF-1 are linked to breast, endometrial, prostate, and colon cancer risk in humans.40-43 In preclinical models, elevated estrogen, testosterone, and IGF-1 promote tumorigenesis by inducing genetic instability, free radical–mediated DNA damage, and an impaired DNA repair response.44, 45 In addition, systemic inflammation can promote cancer development46 and limit antitumor responses by means of immune dysregulation of natural killer cells and stromal tumor-infiltrating lymphocytes.47, 48

Obese cancer survivors are not only at risk for poorer cancer prognosis, but they also have increased risk of diabetes and cardiovascular, liver, and kidney disease, among many other adiposity-related clinical conditions.49-51 Therefore, there is an urgent need to improve cancer care beyond novel therapeutics by elucidating the effects of diet, exercise, and weight management in cancer prevention and treatment.

Diet and Cancer Prognosis: Randomized Clinical Trials

Whether weight loss, without a significant change in diet composition, has a casual role in reducing cancer risk and improving prognosis remains an important but unanswered question. Much of the data regarding diet and cancer survival are in the breast cancer population, although there are data in colorectal and other cancers as well. The Women's Health Initiative Dietary Modification Trial randomized 48,835 postmenopausal women to a control arm (usual American diet) or an intervention arm of a low-fat diet rich in fruit and vegetables. There was a small but sustained 3% weight loss in the intervention arm.52 Although these women did not experience a reduction in the risk of breast or colorectal cancer, post-hoc analysis suggests that death as a result of breast cancer was significantly lower in women who developed breast cancer after randomization to the dietary intervention compared with controls (hazard ratio, 0.78; 95% CI, 0.65-0.94; P = .01), with no change in clinical outcomes after adjusting for weight loss.52 Similarly, in the Women's Intervention Nutrition Study of early stage breast cancer survivors, a low-fat dietary pattern resulted in a small but significant 2.7-kg weight loss and was associated with a 24% higher relapse-free survival rate.53, 54 In contrast, results of the Women's Healthy Eating and Living Study showed that an isocaloric high-vegetable, low-fat diet did not result in any difference in body weight or breast cancer outcomes in patients with early stage breast cancer.55 The findings of these large, randomized trials led to the hypothesis that a negative energy balance is necessary for improving breast cancer outcomes.

The ongoing Breast Cancer Weight Loss Study (ClinicalTrials.gov identifier NCT02750826) will help to determine whether weight loss after breast cancer treatment can improve prognosis. However, additional questions will remain, including how such changes may impact treatment efficacy if implemented earlier and whether macronutrient and/or micronutrient dietary modifications can potentiate the effects of weight loss on cancer prognosis, not only for breast cancer but for many other common cancers.

Calorie Restriction and Cancer Prevention

Chronic daily CR without malnutrition has a powerful effect in preventing spontaneous and chemically induced tumors in animal models.56 This cancer-protective effect was first discovered in 1942 by Tannenbaum, who demonstrated that CR could markedly reduce the development of mammary tumors in rodents.57 As reviewed elsewhere,14 this finding has been consistently replicated in hundreds of CR studies on various tumors, including lymphomas, breast cancers, and skin cancers.58-61 Tumor xenografts in mice are also sensitive to dietary restriction, with the exception of tumors that have PI3K pathway activations.62 More recently, CR has been evaluated in rhesus monkeys, animals that share a strikingly similar genome to that of humans.63 Both the University of Wisconsin and National Institute on Aging CR primate trials have shown a 50% decrease in the incidence of spontaneous cancer, most commonly gastrointestinal adenocarcinoma, in monkeys consuming a 30% CR diet compared with ad libitum-fed animals.64, 65

Data on the effects of daily CR in humans are slowly accumulating and suggest a beneficial effect, even when started in older adults with obesity who most likely already harbor acquired mutations and even microscopic in situ tumors.66 One study from Sweden compared patients who underwent bariatric surgery with a control group and found a 29% reduction in cancer incidence and 23% lower cancer mortality after a median 20 years of follow-up.67 The reduction of cancer incidence was associated with a reduced risk of overall female-specific cancers, including breast and gynecologic cancers, with greater benefit in patients who had higher baseline serum insulin levels.68 The Look AHEAD trial (ClinicalTrials.gov identifier NCT00017953) was a randomized trial of 4859 overweight, diabetic patients without a baseline cancer diagnosis who were randomized to an intensive lifestyle intervention that included a calorie goal of 1200 to 1800 kilocalories (kcal) daily (< 30% of calories from fat and >15% from protein).69 Differences in weight loss were most pronounced after 1 year, with an average weight loss of 8.7 kg compared with 0.75 kg in the intervention group.69 With a median follow-up of 11 years, patients in the intensive lifestyle intervention group had a 16% lower incidence of obesity-related cancers (including esophagus, colon, rectum, kidney, pancreas, stomach, liver, gallbladder, thyroid, uterine, ovarian, postmenopausal breast, and multiple myeloma), which the authors proposed was secondary to weight loss, but there was no difference in the incidence of cancers not associated with obesity.69

Multiple intersecting mechanisms are responsible for the protective effects of chronic daily CR on cancer development and progression.14 Animal and human studies have shown that energy restriction results in major sustained metabolic and hormonal adaptations associated with reduced cancer risk, including reduced insulin levels and improved insulin sensitivity,70, 71 increased IGFBP-1 and SHBG,72 reduced bioavailable testosterone and estrogen,73 and reduced inflammation74 and oxidative stress.75-77 These adaptations support the mechanistic data by which adiposity increases cancer risk. At the molecular level, long-term CR in rodents and humans activates DNA repair, autophagy, and antioxidant and heat-shock protein chaperone pathways while inhibiting cell proliferation and cell senescence biomarkers.78-80 Additional mechanisms include decreased production of growth factors and reactive oxygen species and enhanced anticancer immunity.14

Intermittent Fasting

Although studies of CR in cancer prevention are favorable, many individuals find CR difficult to sustain for prolonged periods. IF is being proposed as an alternative to chronic CR (in which daily food intake is reduced by 10%-25% but meal frequency is unchanged) because it may prove to be more sustainable. Fasting has a rich history rooted in religious traditions and has been practiced for thousands of years.81 Christianity, Judaism, Buddhism, and Islam have advocated various forms of fasting, although Islamic fasts, such as Ramadan, are most similar to secular IF regimens.82 Fasting has been studied by the medical community since the early 1900s for various conditions, including obesity,81 and has recently grown in popularity across many regions of the world.

IF refers to episodic periods of little to no calorie consumption. Variations of IF include every-other-day complete 24-hour fasting83, 84 or fasting on 1 or 2 nonconsecutive days per week, typically referred to as the 6:1 and 5:2 diets, respectively.85, 86 Many fasting programs advise no or small caloric intake (eg, 500 kcal daily) during the fasting period87 with an unlimited amount of calorie-free beverages such as water, coffee (without sugar or milk), bone broth, and diet soft drinks.88 A third method, TRF, requires limiting the consumption of calories to a window of time, typically between 4 and 12 hours daily.89 TRF may or may not include CR during the nonfasting period, which may have additional positive effects, including on circadian rhythm.90 Disruptions in circadian rhythm have been linked to increases in metabolic disorders associated with cancer risk, such as diabetes and obesity, as well as to breast, liver, colon, lung, skin, and prostate cancers.91, 92 A meta-analysis found that 5 years of night shift work increased the risk of breast cancer in women by 3.3%.93 Disruptions to the circadian rhythm are hypothesized to be involved in tumorigenesis through disruption of the expression in genes involved metabolism, autophagy, and DNA damage repair.91 Mouse models have shown that IF can reset circadian rhythms, although this depends on feeding time.94

Fasting in Preclinical Models Intermittent Fasting and Cancer Development and Growth in Rodent Models

IF has been extensively studied in preclinical mouse models of cancer with promising yet mixed results (see Table 2).95-117 For example, IF did not inhibit spontaneous mammary cancer development and failed to slow tumor growth in DBA mice95; however, in xenograft mouse models of breast cancer, melanoma, and neuroblastoma, 2 cycles of 48-hour fasting alone were as effective as 2 cycles of chemotherapy at reducing tumor progression.101 In a small study of a xenograft LAPC-4 human prostate cancer model, an IF regimen composed of 2 separate 24-hour fasting periods showed trends (hazard ratio, 0.59-0.65; P > .05) toward delayed tumor growth and improved survival despite no differences in body weight.118 However, in the larger follow-up study, there was no difference in mouse survival or tumor volumes between mice in the IF cohort and the control groups.96 In cancer-prone, p53-deficient mice (mimicking the Li-Fraumeni syndrome in humans), a 1 day per week fasting regimen significantly delayed tumor onset (P = .001), reduced tumor metastasis (61% in the fasting group developed metastasis vs 75% in the ad libitum group), and increased overall survival (P = .039) compared with mice fed ad libitum, although to a lesser extent than chronic CR.119 In that study, feeding was controlled on nonfasting days to prevent overfeeding, resulting in a significant reduction in weight in the fasting group. Similarly, human lung, liver, and ovarian tumor-bearing mice undergoing periodic 1-day or 2-day per week fasting protocols experienced decreased tumor growth and metastases and improved survival compared with control mice.100 In a notable study with no weight change between groups, alternate-day fasting (ADF) for 2 weeks reduced tumor growth in a mouse model of colon cancer (P < .05).99 This was associated with increased expression of Atg5 and LC3II/I, which are markers of autophagy, suggesting one potential mechanism for the effect of ADF independent of weight loss in mice.

TABLE 2. Selected Studies of Intermittent Fasting in Mouse Models of Cancer STUDY MOUSE STRAIN TUMOR MODEL FASTING SCHEDULE OUTCOME MECHANISMS Intermittent fasting in cancer prevention and treatment Tannenbaum 195095 Female DBA inbred strain Spontaneous mammary tumors 24-h fasting every Mon and Thurs compared with ad libitum (ad lib) (n = 104) No difference in the % of mice forming mammary carcinoma and the mean time of appearance of the tumors Neg results likely due to insufficient weight loss per author Thomas 201096 Male CB-17 SCID LAPC-4 prostate cancer 24-h fasting every Mon and Thurs compared with ad lib (n = 100) No difference in tumor volume at any time point; serum insulin and IGFBP-3 similar; IGF-1 and IGF-1/IGFBP-3 higher in the fasting arm Neg results likely due to insufficient weight loss per author; higher urine ketone levels in fasting mice gradually declined later on Lu 201797 Irradiated SCID mice N-Myc B-ALL, Notch1 T-ALL, MLL-AF9 AML Six cycles of 1-d fasting followed by 1-d feeding; alternative, 2-d fasting followed by 2-d feeding Completely inhibited B-ALL or T-ALL development, at both early and late stage, prolonged survival, but no effect on AML model; fasting decreased circulating glucose and insulin levels, IGF-1, and leptin and increased IGFBP-1 Effects of fasting on leukemia development are cancer type-dependent; fasting-attenuated LEPR signaling in ALL development and maintenance Descamps 200598 Aged OF1 mice Spontaneous lymphoma Alternate-d fasting (ADF) compared with ad lib Fasting significantly reduced the incidence of lymphoma (0% vs 33% for controls) Fasting exerted a beneficial antioxidant effect, absence of weight loss Sun 201799 Female BALB/c CT26 colon ADF for 2 wk Fasting inhibited tumor growth Fasting altered cancer immune microenvironment without weight loss Chen 2012100 Female athymic BALB/c and Beige-nude mice Human A549 lung, HepG-2 liver, SKOV-3 ovarian 4 wk of periodic 1-d fasting or 2-d fasting per wk Fasting led to tumor growth arrest, regression, reduced metastasis, and improved survival Fasting led to NK cell reactivity and IGFBP-3 increase Lee 2012101 BALB/c, C57BL/6 for mouse tumor, nude mice for human tumor 4T1 breast, B16 melanoma, GL26 glioma, neuroblastoma NXS2, MDA-231, neuroblastoma ACN, OVCAR3 48-h fasting every wk × 2 Fasting was as effective as chemotherapy in delaying progression of different tumors and increased the effectiveness of these drugs against melanoma, glioma, and breast cancer cells There is tumor growth upon refeeding; fasting differentially regulated translation and proliferation genes and increased oxidative stress, caspase-3 cleavage, DNA damage, and apoptosis Prolonged fasting before and during chemotherapy Raffaghello 2008102 A/J, CD-1, Nude/nude mice Neuroblastoma NXS2 48-h to 60-h fasting then etoposide Fasting protects host more than protecting tumor Fasting causes differential stress resistance in normal and cancer cells Shi 2012103 CD-1 female nude mice Lung adenocarcinoma A549, mesothelioma ZL55 Fasting started 32 h before and 16 h after CDDP once weekly × 3 Fasting protects normal cells but not cancer cells from cisplatin Activation of ATM/Chk2p53 Pietrocola 2016104 Female C57Bl/6, BALB/c, and nude athymic mice MCA205 fibrosarcoma Fasting 48 h followed by chemotherapy (MTX, oxaliplatin, CDDP) Improved chemotherapy antitumor effect in immunocompetent mice but not in athymic mice Fasting induced autophagy, inhibition of regulatory T cells; fasting can be replaced with caloric restriction mimetics such as hydroxycitrate Safdie 2012105 C57BL/6N SC or intracranial murine GL26 glioma, rat C6, human U251 LN229 and A172 glioma Fasting 48 h before temozolomide or radiation Sensitized glioma but not glia cells to chemotherapy and radiation efficacy Fasting reduced glucose and IGF-1 Saleh 2013106 Female Balb/c Orthotopic 4T1, 67NR ADF and radiation ADF added to radiotherapy, reduced tumor growth; greater effect if involved caloric restriction Downregulation of the IGF-1R pathway Bianchi 2015107 Balb/c CT26 colon 48-h fasting followed by oxaliplatin Fasting potentiated the effects of oxaliplatin Fasting downregulated aerobic glycolysis, and glutaminolysis while increasing oxidative phosphorylation; fasting promoted anti-Warburg effect by increased oxygen consumption but failed to generate ATP, resulting in oxidative damage and apoptosis Huisman 2016108 Male Balb/c CT26 colon 3 d of fasting followed by irinotecan Fasting prevented toxicities but did not enhance the efficacy of chemotherapy Fasting induced a lower systemic exposure to SN-38, which may explain the absence of adverse side effects, while tumor levels of SN-38 were unchanged Jongbloed 2019109 Male Balb/c CT26 colon 3 d of fasting followed by irinotecan Fasting reduced chemotherapy-induced side effects Fasting activated a protective stress response in normal tissue but not in cancer Tinkum 2015110 B6(Cg)-Tyrc-2J/J Fasting for 24 h followed by etoposide Fasting preserved small intestinal (SI) architecture, improved survival Fasting maintained SI stem cell viability and SI barrier function; DNA repair and DNA damage response genes were elevated, with DNA damage more efficiently repaired Fasting-mimicking diets (FMDs) as adjuvant cancer treatment Di Biase 2016111 Female Balb/c, female C57BL/6 4T1 breast in Balb/c, B16 melanoma in C57BL/6 FMD or fasting and chemotherapy (doxorubicin, or cyclophosphamide) FMD is as effective as fasting alone or in combination with chemotherapy in reducing tumor progression but is not effective in nude mice FMD reduces IGF-1, increases the levels of bone marrow common lymphoid progenitor cells and cytotoxic CD8-positive tumor-infiltrating lymphocytes; this effect is partially mediated by downregulating HO-1 Caffa 2020112 NOD/SCIDγ MCF7 xenograft FMD in combination with hormonal therapy (tamoxifen or fulvestrant) and CDK4/6 inhibitor (palbociclib) FMD improved efficacy of tamoxifen and fulvestrant and CDK4/6 inhibitor; FMD prevented tamoxifen-induced endometrial hyperplasia FMD lowered circulating IGF-1, insulin, and leptin and inhibits AKT-mTOR signaling by upregulation of EGR1 and PTEN Time-restricted feeding (TRF) in cancer development and treatment Das 2021113 Female C57BL/6 J, ovariectomized, or chemically induced ovotoxicity; transgenic PyMT female mice Orthotopic Py230 and E0771 breast cancer cells, tail vein injection of E0771 cells, MMTV-PyMT spontaneous breast cancer TRF from 10 pm to 6 am daily with high-fat diet (HFD) TRF abrogates obesity-enhanced postmenopausal mammary tumor growth in the absence of calorie restriction or weight loss and reduces metastasis in the lung; inhibition of insulin with diazoxide mimics TRF, but insulin pump reverses the effect of TRF TRF increases insulin sensitivity, reduces hyperinsulinemia, restores diurnal gene expression rhythms in the tumor, and attenuates tumor growth and insulin signaling Sundaram & Yan 2018114 MMTV-PyMT mice (FVB) MMTV-PyMT spontaneous breast cancer TRF of HFD at dark phase (12 h) between 12 and 24 h TRF mitigates HFD-enhanced mammary tumorigenesis TRF reduced the HFD-induced parameters, including plasma leptin, MCP-1, PAI-1, hepatocyte growth factor, and angiogenic factors Yan 2019115 Male C57BL/6 mice Subcutaneous Lewis lung carcinoma xenograft TRF of HFD at dark phase (12 h) between 12 and 24 h TRF prevented HFD-enhanced lung metastasis TRF prevented HFD-induced increase in plasma glucose, insulin, cytokines, and angiogenic factors Turbitt 2020116 Balb/c mice Orthotopic renal tumor cells TRF of HFD at dark phase for 12 h TRF did not alter tumor weight, lung, or metastasis and failed to improve anti–CTLA-4 efficacy TRF did not alter excised renal tumor weights or intratumoral immune response Abbreviations: ALL, acute lymphoblastic leukemia; AML, acute my