Intermittent fasting interventions to leverage metabolic and circadian mechanisms for cancer treatment and supportive care outcomes

Abstract

Intermittent fasting entails restricting food intake during specific times of day, days of the week, religious practice, or surrounding clinically important events. Herein, the metabolic and circadian rhythm mechanisms underlying the proposed benefits of intermittent fasting for the cancer population are described. We summarize epidemiological, preclinical, and clinical studies in cancer published between January 2020 and August 2022 and propose avenues for future research. An outstanding concern regarding the use of intermittent fasting among cancer patients is that fasting often results in caloric restriction, which can put patients already prone to malnutrition, cachexia, or sarcopenia at risk. Although clinical trials do not yet provide sufficient data to support the general use of intermittent fasting in clinical practice, this summary may be useful for patients, caregivers, and clinicians who are exploring intermittent fasting as part of their cancer journey for clinical outcomes and symptom management.

Dietary interventions can modulate cancer initiation, progression, treatment efficacy, symptom burden, and recurrence. These interventions have historically focused on diet composition and quantity of calories, however, more recently, there is mounting evidence that timing of calories can have powerful effects on oncologic and quality-of-life outcomes.

Intermittent fasting (IF) includes planned periods of little or no calorie consumption, either with or without additional restrictions placed on nutrient composition and caloric intake. The use of fasting as a therapeutic approach to improve health has been documented in early human history. Notably, in the fifth century, the Greek physician Hippocrates recommended fasting (water and medicinal tea only) to treat fevers and diseases in the acute crisis stage. Many religions have incorporated fasting for centuries, including a 25-hour fast during Yom Kippur in Judaism and daily eating regimens as part of some branches of Buddhism. In 1945, the first modern IF experiment was conducted in which rats were subjected to regular 24-hour fasting periods from 42 days of age until death. Rats who underwent IF lived longer and exhibited retarded breast tumor growth (1). This work was built on gradually until the 2010s [eg, (2-4)], when the benefits of IF for weight loss and metabolic regulation were further supported by more advanced molecular studies, and the biological plausibility for IF to benefit cancer populations materialized. Momentum subsequently increased [see (5)], and in 2020, in response to the growing interest in the benefits of fasting regimens, the National Cancer Institute proposed a “Provocative Question” to investigate IF on cancer incidence, treatment response, and outcomes (6).

There are several main types of IF (Table 1). Alternate-day fasting entails eating every other day and consuming no or minimal calories (<600 kcal) on the days in between (7). A modified alternate-day fasting protocol is a 5:2 pattern in which low-no calorie intake (approximately 75%-100% calorie restriction) occurs on 2 consecutive or nonconsecutive days per week; other modifications to the alternate-day fasting protocol include specification of meal frequency on the low-calorie (“fast”) day and macronutrient targets (8). Among the day-long to multiday-long IF regimens, the most well studied clinically is the fasting-mimicking diet (FMD), which entails several days of consuming 10%-50% of typical calorie intake (9). Thus far, the FMD has only been tested as a plant-based, low amino-acid substitution diet consisting of a prespecified formulation of soups, broths, liquids, and tea and has primarily been used before and after a chemotherapy infusion in patients with cancer (10-12). Time-restricted eating is when people consume all their food within a 4- to 12-hour window; prolonged nightly fasting is a special case of time-restricted eating in which the eating window coincides with daytime hours. Prolonged nightly fasting may hold additional health benefits in that it can entrain circadian rhythm (ie, sleep–wake cycle) or the body’s internal body clock (13,14). Ramadan fasting is another form of time-restricted eating that is a traditional Islamic practice. Muslims fast, without food or water, from sunrise to sunset for the 30 days of Ramadan. Thus, Ramadan fasting and prolonged nightly fasting are diametrically opposed with respect to timing of food intake alignment to light dark cycle and natural body rhythms. Most time-restricted eating regimens do not restrict or direct the macronutrient composition of food intake, though caloric restriction often occurs unintentionally as a consequence of the reduced number of eating hours (13,14). Because of the considerable heterogeneity across the multitude of IF protocols, a critical unmet need is the optimization of IF protocols for specific populations and clinical outcomes in the field of oncology and beyond.

Table 1.

Definitions of intermittent fasting protocols and mechanisms

Intermittent fasting regimen	Definition	Duration that regimen is applied for therapeutic benefit	Mechanistic modulation
Alternate-day fasting	Fasting or very low calorie for 24 h every other day, every 3 d, every 4 d, or 2 nonconsecutive days per wk	A few months	Caloric restriction, metabolism, ketosis
Fasting-mimicking diet	An extremely low-calorie diet	Several days leading up to a chemotherapy infusion and up to 24 h after infusion	Ketosis, metabolism to sensitize cancer cells to chemotherapy
Time-restricted eating	Restriction of calories to a specific time window during each 24-h period, typically less than 10-12 h	Sustainable, no time limit	Circadian rhythm, incidental caloric restriction, metabolism
Prolonged nightly fasting	A subset of time-restricted eating in which the eating window is during daytime hours	Sustainable, no time limit	Circadian rhythm, incidental caloric restriction, metabolism

Herein, we review mechanisms by which IF interventions may lead to health outcomes, summarize recent IF interventions in cancer-focused preclinical models and human studies, and finally, outline opportunities to evaluate IF to address clinical and supportive care outcomes in cancer.

Mechanisms

IF interventions are potent regulators of systemic and cellular pathways responsible for nutrient sensing and cellular stress responses such as autophagy and inflammation. Remodeling of these pathways has been implicated in the development and progression of cancer, aging, and a range of metabolic diseases (15). In addition, time-restricted eating interventions can modulate circadian rhythms and therefore affect many aspects of human behavior and physiology beyond metabolic pathways (Figure 1).

Figure 1.

The effects of intermittent fasting regimens on biological mechanisms, cancer treatment outcomes, and quality-of-life outcomes. IGF-1 = insulin-like growth factor 1; mTOR = mechanistic target of rapamycin.

Food intake rapidly initiates a cascade of biochemical responses to process the incoming nutrients. In contrast, during fasting, the body mobilizes stored energy. Specifically, the mechanistic target of rapamycin complexes 1 and 2 (mTORC1 and mTORC2) are regulated by nutrient availability, and loss of mTOR signaling partially mediates the cellular effects of IF interventions on cancer cells. In muscle, caloric intake activates mTOR signaling within 30 minutes of eating to yield an anabolic state (16), with peak mTORC1 activity at approximately 60-90 minutes. Subsequently, mTORC1 is gradually deactivated. Approximately 12-16 hours after food absorption, a metabolic switch is activated in which the primary source of energy shifts from glucose to fat and ketones (17). This metabolic switch, mediated by decreased mTOR signaling, is key for prolonged fasting (more than approximately 16 hours) and FMDs to be effective.

The activity of mTOR is regulated by complex and multilayered cellular signaling (18). In the fed state, sensors such as Sestrin2 for leucine; GTPase-activating protein activity toward Rags-1 (GATOR1) for leucine, arginine, and glutamine; and SAMTOR for S-adenosylmethionine (a methionine metabolite) promote activation of mTOR. Alternatively, fasting works through the same cascades to inhibit mTORC1 (19). In addition to amino acid availability, glucose availability also regulates mTOR. After 4-6 hours of fasting, gluconeogenesis is induced in the liver via transcriptional coactivator peroxisome proliferator-activated receptor γ coactivator-1α. Glucose-3-phosphate (G3P) is used as a substrate for hepatic gluconeogenesis, and the conversion of G3P to dihydroxyacetone phosphate (DHAP) maintains blood glucose levels during fasting. Recently, DHAP was shown to regulate mTORC1 (20). The tuberous sclerosis complex (TSC) is a strong inhibitor of mTORC1 that is also regulated by diverse energy sensors including insulin and insulin-like growth factor 1 (IGF-1) (18,21) as well as cellular stressors including hypoxia and DNA damage (22,23). However, in many cancers, TSC regulation of mTORC1 is often lost, resulting in constitutive mTORC1 activity. More research is needed to understand how fasting affects mTOR regulation in cancerous and nontransformed cells.

5' adenosine monophosphate (AMP)-activated protein kinase (AMPK) is another master regulator of cellular metabolism. Fasting regimens, including IF, increase the AMP: adenosine triphosphate (ATP) ratio, activating AMPK and leading to increased catabolism and suppression of anabolism (21,24-27). Specifically, AMP binding to the AMPK γ subunit activates the catalytic α subunit and promotes glucose and lipid metabolism, activates autophagy, and inhibits mTORC1 indirectly via activation of TSC and directly by phosphorylation of regulatory-associated protein of mTOR (RAPTOR). In opposition to mTORC1, AMPK is activated by Sestrin2. Further, inhibition of AMPK by mTOR is critical to the induction of mTORC1-driven anabolism and protein synthesis in response to amino acid availability.

The cellular ratio of nicotinamide adenine dinucleotide (NAD+) to its reduced form, NADH, is a key metric of the cellular redox state and is integrated by the activity of sirtuins, a specialized family of NAD-dependent protein deacetylases. The various sirtuin family members cooperate with mTOR (28) and AMPK signaling to direct cellular responses to various stressors (29). Indeed, sirtuin signaling induction has been implicated in the mediation of several IF-driven responses. Prolonged fasting as part of an IF regimen increases AMPK, reduces mTOR, and increases NAD-dependent deacetylase sirtuin-1 (SIRT1) activity. This signaling shift results in increased autophagy and mitophagy, DNA repair, antioxidant capacity, and stem cell self-renewal, as well as decreased senescence in normal cells (21).

Autophagy is a complex process that removes damaged organelles and misfolded proteins for recycling into anabolic substrates. Autophagy is activated by loss of mTOR activity and by activation of AMPK and is thus increased in the fasted state. Induction of autophagy in tumor cells has a complex relationship with survival of the cell. Autophagy often eliminates intracellular signals, such as damaged organelles or proteins, that may cause oncogenic transformation of normal cells, thus suggesting a cancer preventive role of autophagy induction. In contrast, autophagy confers a survival advantage in established tumors (30) by increasing resistance to hypoxia and disruption of growth factor signaling and by supporting paracrine signaling networks and the epithelial to mesenchymal transition (24). Systemic induction of autophagy in response to IF promotes gene expression that may promote longevity, as has been seen in model organisms. Several cell and animal models have implicated the induction of autophagy in the antitumor effects of IF; however, the role of autophagy in response to fasting in clinical trials is not yet clear (5). Mitophagy, defined as the selective removal of dysfunctional mitochondria via autophagic processes, is induced in several tissues in response to fasting; however, the role of such induction in cancer is yet unclear (31,32).

There is growing appreciation for how the above metabolic pathways interact with innate and adaptive immune responses to control tumor growth. Recent work demonstrated that a FMD intervention improved the response to immune checkpoint inhibition in mouse models of breast cancer otherwise nonresponsive to immune checkpoint inhibition (33). Of note, fasting was also associated with fewer instances of immunotoxicity and adverse events. Further, a FMD protected circulating T cells from chemotherapy-driven DNA damage while promoting therapeutic efficacy in patients with breast cancer (10).

IF protocols that include fasting for at least 24 hours (eg, alternate-day fasting) and FMD lasting 3-5 days lead to ketosis, which has been postulated to result in preferential sensitization of cancer cells to chemotherapy while protecting normal tissue. The fasting-induced switch from glucose to lipid metabolism is associated with changes in hormones and metabolites that reduce proliferation and metabolic activity in untransformed cells, thus increasing their resistance to chemotherapy. Cancer cells, however, demonstrate sustained dependence on glucose and a high proliferative rate, despite the lower abundance of available nutrients in the fasted state. This sustained growth in the absence of anabolic substrates may render cancer cells more sensitive to chemotherapy compared with untransformed cells (34). The difference in response to fasting between transformed and nontransformed cells, termed the differential stress response, underlies part of the rationale for pairing fasting regimens with chemotherapy and other forms of cytotoxic therapy in patients with cancer.

IF reshapes the gut microbial community and may impact tumor cell metabolism. In clinical trials and animal models, fasting for various lengths of time increases bacterial diversity of the gut microbiota, a result associated with improved health (35,36). Mechanistically, fasting promotes favorable changes in the microbial community leading to increased production of short chain fatty acids, which in turn can regulate insulin response and glucose metabolism, as well as decreased inflammation (35-37). Thus, modulation of the gut microbiota by IF regimens alters levels of several nutrients in the body, thereby changing the availability of substrates required for tumor cell metabolism.

Time-restricted eating IF regimens further remodel systemic metabolism through entrainment of the circadian rhythm (38). Circadian rhythms are 24-hour biological cycles that work in synchrony to regulate various physiological processes, and approximately half of protein-expressing genes operates on a circadian cycle (14). When eating behavior is in line with the daily active phase (ie, natural light in humans and dark in mice) and fasting occurs during the inactive phase, cells can optimally engage in physiological repair mechanisms and produce fewer free radicals. Sustained, robust circadian rhythms are beneficial to general health by improving cardiovascular risk factors and glucose regulation (39). Leveraging circadian processes has the potential to optimize drug dosing (40) and improve clinical outcomes and symptom management (39).

IF interventions in animal models

The number of clinical studies of IF in cancer is increasing but limited by short follow-up periods (especially in studies of cancers with good long-term outcomes), small sample size, limited statistical power, and compliance issues [eg, (10)]. Fortunately, animal models offer a more controlled means of examining tumor response to dietary intervention on an accelerated timescale and allow more rapid delineation of underlying mechanisms. Although preclinical studies of the effects of IF on cancer-related pathways and tumor outcomes have historically shown mixed results on tumor growth (41-47), modulation of IGF-1, insulin, ketone formation, and autophagy have been consistently observed as potential mechanisms underlying the anticancer response to IF (5,24,38,48-50). Here, we describe established and emerging responses to IF in recently published rodent models of cancer (Table 2).

Table 2.

Preclinical studies of IF in models of cancer

Study	Mouse strain	Tumor model	Fasting schedule	Outcomes
Alternate-day fasting
Antunes et al. (105)	Athymic NOD/SCID (male)	SK-Mel-28 melanoma cells	Food was withheld for 24 h and was resupplied ad libitum; feed and fast cycles were performed only on weekdays.	Sorafenib with fasting led to a reduction in tumor growth compared with the control and either single treatment.
Chen et al. (60)	C57BL/6N (male) (for lung) BALB/c nude (male) (for colon)	LLC cells HCT116 colon cancer cells	Food was withheld 24 h from 6 pm to 6 pm of the following day when food was resupplied ad libitum.	The combination of selenite and alternate-day fasting more potently reduced tumor volume relative to either intervention alone. Selenite and fasting led to a stronger upregulation of SLC7A11, decrease in glutathione, and increase in ROS and NADP+/NADPH compared with either single treatment or the control.
Peng et al. (58)	Athymic BALB/c nude (female)	MGC803 gastric cancer cells	Food was withheld 24 h from 6 pm to 6 pm of the following day when food was resupplied ad libitum. Berberine was administered to ad libitum–fed mice, IF mice during their fed state, or IF mice during their fasted state.	Berberine and fasting were well tolerated without the occurrence of any adverse effects. Neither berberine nor fasting alone affected tumor size, but berberine with fasting decreased tumor mass and volume.
Fasting-mimicking diet
Caffa et al. (106)	BALB/c nude mice (female)	MCF7, T47D, and ZR-75-1 estrogen receptor–positive breast cancer cells	Fasting regimen 1: water-only fast for 48 h every wk Fasting regimen 2: FMD for 4 d every week (feeding cycle continued until body weight was regained)	Prolonged fasting and the FMD lowered circulating IGF-1, insulin, and leptin; inhibited AKT–mTOR signaling; and upregulated EGR1 and PTEN. FMD enhanced the activity of the endocrine therapies tamoxifen and fulvestrant, as seen in measures of tumor volume. Fulvestrant with albociclib (a cyclin-dependent kinase 4 and 6 inhibitor) with FMD promoted tumor regression and reverted acquired resistance to drug treatment.
Cortellino et al. (33)	NOD/SCIDgamma and BALB/c (female)	Breast cancer, low immunogenic model (ie, a model that mounts a low immune response)	Two cycles of FMD (FMD for 4 d) starting the third day after tumor implantation; 3 d of refeeding with standard diet between FMD cycles	FMD improved efficacy of anti–PD-L1 and anti-OX40. FMD promoted T-cell activation (CD8, CD4, Tregs, and gdT cells) and suppressed immunosuppressive cells (M2 macrophages, PMN-MDSC, M-MDSC) and prevented early exhaustion of effector T cells. FMD shifted cancer from glycolytic to respiratory metabolism and decreased lactic acid. FMD reduced collagen deposition. FMD prevented hyperactivation of the immune response. Anaphylaxis occurred in mice with drug alone and was prevented by adding FMD.
Pomatto-Watson et al. (65)	BALB/cJ (female)	4T1 breast cancer cells	FMD for 4 d with 10 d of refeeding chow in between cycles; low-calorie control followed same feeding cycle as FMD but fed with standard AIN-93G chow; calorie restriction consisted of daily caloric restriction isocaloric to the fasting conditions	Compared with daily caloric restriction, feeding cycles failed to protect against lung metastases, regardless of diet composition or treatment initiation period (pre- or post-4T1 injection). No difference in tumor area between FMD and low-calorie control fasting regimens. Protective effects of daily calorie restriction occurred in a dose-dependent manner up to 40% calorie restriction. Daily calorie restriction activated the immune system by reducing the number of protumor immune cells (CD11b+Gr1+) and upregulating antitumor (CD8+ and CD4+) immune cells in peripheral tissues.
Salvadori et al. (56)	NOD/SCIDγ (female) Balbc/Ola Hsd (female)	SUM159 and 4T1 triple-negative breast cancer cells	FMD for 4 d; weight loss was not able to exceed 20% during cycling; mice recovered to original body weight before cycling again.	FMD inhibited tumor-derived mammosphere propagation ex vivo, an effect that was reversed by supplementing sugar into drinking water of mice. Coupling FMD with a PI3K/AKT/mTOR inhibitor improved survival and reduced treatment toxicity and hyperglycemia.
Time-restricted feeding
Das et al. (38)	Ovariectomized C57BL/6J and MMTV-PyMT	Py230, E0771 breast cancer cells; spontaneous mammary tumor model	16-h fasting and 8-h feeding of high-fat diet; food given during the active phase.	TRF mice were leaner compared with ad libitum–fed mice. TRF improved insulin sensitivity. TRF entrained diurnal gene expression in the mammary tumor. TRF attenuated tumor growth in preventive- and treatment-oriented models. TRE attenuation of tumor growth mediated by insulin reduction.
Jawarneh and Talib (107)	BALB/c (female)	EMT6/P and EMT6/C breast cancer cells	18-h fasting and 6 h of ad libitum feeding	Fasting reduced blood glucose and increased β-hydroxybutyrate. Combination of IF, cisplatin, and Ashwagandha water extract reduced tumor size.
Turbitt et al. (41)	BALB/c (male)	Renca renal cancer cells	12-h fasting and 12-h feeding; food given during the active phase (6 pm–6 am).	TRF reduced renal tumor bioluminescence in 27-week-old, obese mice but did not affect excised tumor weights and intratumoral immune responses or improve response to anti-CTLA4 therapy.
Walker et al. (42)	BALB/c (female)	4T1 breast cancer cells	12-h fasting and 12-h feeding; food was given during either the active or inactive phase.	Feeding in the inactive phase resulted in higher body mass than those fed ad libitum. No effect of TRF on tumor volume. Mice fed during the inactive phase displayed statistically significantly disrupted daily activity and body temperature rhythms relative to other feeding regimens.
Yan et al. (108)	MMTV-PyMT (female)	Spontaneous mammary tumor model	12-h fasting and 12-h feeding of high-fat diet; food given during the active phase (6 pm–6 am).	No impact of TRF upon tumor metabolome. TRF attenuated expression of Hmgcr, Srebp1, Fads2, and Ppard in mammary tumors.
Yan et al. (43)	C57BL/6 (male)	LLC cells	12-h fasting and 12-h feeding of high-fat diet; food given during the active phase.	TRF shifted the metabolic profile of cancer-bearing mice toward that of noncancer-bearing controls (eg, changes in α-tocopherol, docosahexaenoic acid, cholesterol, several amino acids, lactate, and pyruvate).
Other fasting regimens
Ajona et al. (62)	Sv/129 (female), C57BL/6J (female), BALB/c	393P, LLC, Lacun3 lung cancer cells	Fasting was performed on days 14-16 (48 h), 19-21 (48 h), and 25-26 (24 h) following tumor cell implant; ad libitum food access when not fasted.	Short-term starvation synergizes with anti–PD-1 treatment to inhibit lung cancer progression and metastasis, partially negated with addition of recombinant IGF-1. Short-term starvation reduced IGF-1 and decreased IGF-1R signaling in tumor cells; suppression of IGF-1R in combination with anti–PD-1 decreases tumor volume.
Fu et al. (61)	BALB/c (female)	4T1 and 4T07 breast cancer cells	Two 48-h fasts interspersed with 8 d refeeding starting 14 d after tumor cell implant.	Fasting was as effective at reducing tumor volume as docetaxel. Fasting reduced a proportion of anti-tumor myeloid-derived suppressor cells in the spleen of tumor-bearing mice. The fasting intervention and 2-deoxyglucose treatment induced cleaved caspase 3 expression and reduced Ki67 expression in the spleen.
Tang et al. (59)	BALB/c (female) FVB/N-Tg (MMTV-PyMT)/Nju (female)	4T1 breast cancer cells	Two 48-h fasts interspersed with 8 d refeeding starting 16 d after tumor cell grafting.	Fasting suppressed 4T1 tumor growth in a SIRT7-dependent fashion; low glucose in the fasted state stabilizes SIRT7, which suppresses AKT activation. Fasting decreased lung metastatic tumor nodules. Combining doxorubicin, trametinib, and fasting synergistically decreased tumor growth.

^aFMD = fasting-mimicking diet; gdT = γδ T cells; IF = intermittent fasting; IGF-1 = insulin-like growth factor 1; IGF-1R = insulin-like growth factor 1 receptor; LLC = Lewis lung cancer; NADP⁺ = nicotinamide adenine dinucleotide phosphate (oxidized); NADPH = nicotinamide adenine dinucleotide phosphate (reduced); MRI = magnetic resonance imaging; mTOR = mechanistic target of rapamycin; PD-L1 = programmed cell death ligand 1; PTEN = phosphatase and tensin homolog; SIRT7 = sirtuin 7; Treg = regulatory T cells; TRF = time-restricted feeding.

Several different approaches to IF interventions have been employed in preclinical models of cancer, which vary in the frequency and duration of the fasting window as well as availability of food during the feeding period (ie, ad libitum access or caloric restriction) (48). The duration of fasting regimens ranges from 12 to 60 hours. Feeding windows are typically adjusted in accordance with the duration of the fasting window, with shorter fasting periods followed by shorter refeeding periods. For example, time-restricted feeding, which is the preclinical counterpart to human time-restricted eating, typically involves daily fasts of 12-16 hours. These fasts are interspersed by daily ad libitum food access. Models of intermittent calorie restriction follow a similar construct in which daily food allotments are generally consumed within 2-4 hours (51); thus, mice in these models undergo daily 20- to 22-hour fasts. Conversely, models of short-term starvation, which entail fasting periods of up to 60 hours, include multiple days of ad libitum refeeding to allow recovery of body weight. Finally, in FMDs, mice are fed a low-calorie broth enriched with vegetable powders and olive oil, typically with 1 day of 50% calorie restriction followed by 3 days of 90% calorie restriction. This method of fasting has been widely cited in the literature as an efficacious form of IF to reduce cancer incidence and increase the efficacy of antitumor treatments (34,52,53).

A promising approach involving IF and cancer involves synergistic combinations of fasting with pharmacological interventions. The idea of combining fasting with treatment regimens in cancer is not new—use of fasting to reduce the side effects of chemotherapy has been researched for several years (54,55). However, more recent work has highlighted the combination of fasting with drugs beyond chemotherapy to synergistically reduce tumor cell viability. For example, a FMD in a model of triple-negative breast cancer reduced glucose availability and protein kinase A signaling, thus making tumor cells more sensitive to treatment with phosphatidylinositol-3 kinase (PI3K)-AKT-mTOR inhibitors compared with tumors in ad libitum–fed mice (56). Similarly, in a model of colon cancer, alternate-day fasting (ie, 24-hour fasting followed by 24-hour feeding) reduced blood glucose concentrations compared with ad libitum–fed mice, and metformin treatment only reduced tumor size when paired with fasting (57). In a cell culture model, reduced glucose, but not reduced amino acids or serum, synergized with metformin treatment to reduce cell viability, suggesting that the fasting-mediated reduction in blood glucose underlies the antitumor effect of the cotreatment observed in vivo (57). In this model, metformin and fasting cotreatment activated the protein phosphatase 2A (PP2A)-glycogen synthase kinase-3β (GSK3β)-myeloid cell leukemia-1 (MCL-1) axis, resulting in cancer cell death (57). Alternate-day fasting similarly activated the PP2A-GSK3β-MCL-1 axis in a model of gastric cancer, suggesting that activation of this pathway occurs in response to fasting in multiple tumor types (58). Fasting-mediated reduction in glucose via short-term starvation (ie, 48-hour fast followed by 4 days refeeding) in a murine model of breast cancer–sensitized tumors to treatment with the mitogen-activated protein kinase kinase 1/2 (MEK1/2) inhibitor trametinib and the chemotherapy drug doxorubicin via stabilization of SIRT7 (59). Finally, reducing blood glucose through an alternate-day fasting protocol in models of lung and colon cancer increased sensitivity to selenite treatment, with the combination of fasting and selenite having stronger antitumor outcomes compared with either treatment alone (60). In this model, reduced glucose availability limited NAD hydrogen phosphate production from the pentose phosphate pathway, leading to impaired reactive oxygen species clearance and cytotoxicity following selenite treatment (60). Thus, strategically targeting glucose-dependent pathways and processes in combination with fasting regimens represent a powerful tool for treatment of cancer.

Remodeling the immune landscape is another emerging mechanism by which different IF interventions may prevent tumor progression. For example, 2 bouts of starvation (48 hours, interspersed with 8 days of ad libitum feeding) reduced the proportion of immunosuppressive CD205⁺ granulocytic myeloid–derived suppressor cells in the spleens of tumor-bearing mice in a model of mammary cancer (61). In a model of lung cancer, the combination of starvation and immune checkpoint blockade using anti–programmed death 1 increased the proportion of CD8⁺ T cells and decreased T-regulatory cells in the tumor microenvironment (62). Similarly, a FMD increased T-cell activation and depleted immunosuppressive cells, including M2 macrophages and granulocytic and monocytic myeloid–derived suppressor cells in a model of mammary cancer (33). In this model, tumors from mice treated with a FMD displayed a metabolic shift from a glycolytic to a respiratory phenotype, consequently resulting in reduced acidification of the tumor microenvironment (33). Given that acidic tumor microenvironments are immunosuppressive (63), the metabolic shift in tumor cell metabolism may play a direct role in remodeling tumor immune cell infiltration.

In addition to its role in modulating immune cell populations, IF may be able to improve tumor response to immunotherapy. Use of IF regimens improved the efficacy of anti–programmed cell death ligand-1 (PDL-1) and anti-OX40 in a mammary cancer model and anti–programmed cell death protein 1 (PD-1) therapy in a lung cancer model, a result that was further enhanced by the addition of an IGF-1 receptor inhibitor (33,62). Finally, IF limits hyperactivation of the immune system in response to immunotherapy (33), suggesting that a FMD may be a critical tool in overcoming current dose-limiting toxicities of immunotherapies. This finding complements previous research showing that a FMD limits chemotherapy-induced DNA damage in T cells (10). Collectively, these data indicate that a variety of fasting regimens may improve tumor outcomes by modifying immune populations in tumor-bearing mice, as well as sensitizing tumors to immunotherapy.

Time-restricted feeding has recently emerged in the IF literature and has shown mixed results on tumor outcomes (38,41,42). A recent report suggests that time-restricted feeding of a high-fat diet blocked mammary tumor development and progression through regulation of insulin (38); however, reports from other models have shown null effects of time-restricted feeding on cancer outcomes (41-43). These mixed results indicate that more work is required to identify the necessary parameters of time-restricted feeding, which underpin its purported beneficial effects. Interesting studies are underway that are exploring underlying mechanisms of time-restricted feeding, for example, to reduce cancer risk through optimization of mitochondrial energy production (eg, R01CA258221).

A future area of investigation will be understanding the importance of dietary composition during fasting regimens. Some evidence suggests that select diets, including FMD and ketogenic diets, may be beneficial to increase health via their ability to recapitulate physiological effects of fasting, for example, ketone body formation (64). However, other reports examining diet composition in rodent models indicate that diet composition does not modulate the effects of fasting on health or cancer-related outcomes. For example, administering either an FMD or a control diet in an intermittent fashion (4 days restriction, 10 days refeeding) had similar results in murine models of mammary cancer (65). Furthermore, comparing these 2 diet interventions with chronic calorie restriction revealed stronger anticancer effects in the chronic restriction group (65). Similarly, there was no difference in health outcomes such as insulin and homeostatic model assessment for insulin resistance between mice fed high-sucrose and low-sucrose diets when both diets were presented as a single meal once daily (66). These data may suggest that either reduced calorie intake, a prolonged daily fasting window, or a combination of the two, is more important than diet composition for improvement of cancer-related outcomes.

In conclusion, animal models provide necessary preclinical evidence that IF interventions may improve tumor outcomes in vivo. Accumulating data suggest that IF during chemotherapy sensitizes cancer cells while protecting normal cells from toxicity, and newer evidence suggests that fasting interventions may synergize with other types of cancer treatments to improve tumor outcomes. Further work is required to determine optimal diet composition of IF regimens. In addition, future research efforts should include head-to-head comparisons of fasting effects in multiple models of cancer to determine which tumor types are most responsive to fasting interventions.

IF trials in humans

Although preclinical work is necessary to deconstruct and understand underlying mechanisms, human work is imperative to determine safety, feasibility, efficacy, and practical dissemination practices. IF interventions in human clinical trials have recently been used to address obesity and associated chronic diseases, not only as an alternative to caloric restriction but also to rectify metabolic dysfunction, entrain circadian rhythm, and overall, improve physiology (13,15,67). A steadily increasing body of data indicates that IF may be a safe, accessible, low-cost, low-burden intervention with potential benefit to improve the efficacy of cancer treatments and prevent long-term sequelae including cardiometabolic toxicities, neurocognitive impairment, and other cancer-related outcomes. Safdie et al. (4) in 2009 showed that fasting for days around the chemotherapy infusion reduced fatigue, weakness, and gastrointestinal side effects and did not thwart the efficacy of chemotherapy. Additionally, Marinac et al. (68) in 2016 demonstrated that fasting less than 13 hours per night was associated with a 36% increase in the risk of breast cancer recurrence. Since then, more than 20 published human studies have examined the relationships between various IF regimens and cancer patient outcomes, including retrospective epidemiological studies, single-arm feasibility studies, and randomized controlled trials. Studies before 2020 have been summarized previously (5). Here we synthesize the clinical IF literature specific to the cancer population from January 2020 to August 2022 (Table 3).

Table 3.

Human clinical trials testing various fasting regimens on clinical and supportive care outcomes^a

Study	Study type	Tumor type	Population	Fasting schedule	Duration of intervention	Outcome
Fasting-mimicking diet
Caffa et al. (106)	Single-arm phase II (NCT03595540)	Any solid or hematologic tumors	60 patients undergoing active treatment	Monthly 5-d FMD cycles; not as calorie restricted as NCT03340935	During active treatment: average 6.8 mo, up to 14 mo	No severe adverse events of grade 3 or higher. Body weight and hand grip strength were stable. Fat mass decreased and fat-free mass increased over time.
Vernieri et al. (12)	Single-arm prospective clinical trial (NCT03340935)	Different cancer types	101 patients with localized and advanced cancer	FMD for 5 d (4 d before and day of standard antitumor therapies) followed by 16-23 d of refeeding	Variable, up to 8 cycles	76% completed at least 3 FMD cycles; 20% completed all planned FMD cycles. Acute and pre-post reduction in plasma glucose, serum insulin, and plasma IGF-1; increase in urinary ketones. Acute decreases in BMI; no pre-post changes in BMI.
Ligorio et al. (109)	Case studies	Different cancer types	Five patients who had poor prognosis and achieved complete and long-lasting tumor responses [from Vernieri (12), NCT03340935]			FMD appeared to have additive or synergistic antitumor effects when combined with standard anticancer therapies.
Caffa et al. (106)		Patients with hormone receptor–positive breast cancer receiving estrogen therapy	Subset of patients from NCT03340935 NCT03595540			FMD reduced insulin, leptin, and IGF-1.
de Groot et al. (10)	Randomized phase II, DIRECT	HER2-negative breast cancer, stage II or III	131 patients receiving neoadjuvant chemotherapy	FMD or usual diet for 3 d prior to and the day of neoadjuvant chemotherapy	8 cycles of AC-T chemotherapy, once every 3 wk or 6 cycles of FEC-T chemotherapy, once every 3 wk	34% of those in the FMD group complied with FMD for more than 4 cycles, 20% complied for all cycles; 8% of those in the control group fasting during 1 or more cycles. No difference between grade 3 and 4 toxicity during chemotherapy between the FMD and the usual care groups. FMD cohort more likely to have Miller-Payne 4 or 5 pathologic response in a per-protocol analysis. FMD led to lower plasma insulin, lower plasma glucose, and higher ketone bodies prechemotherapy. No between-group differences for quality of life or distress. FMD reduced chemotherapy-induced DNA damage in T lymphocytes.
Lugtenberg et al. (110)	Same cohort as de Groot et al. (10), DIRECT					Within-group decrease in BMI for FMD group from pre- to postchemotherapy. No between-group differences in global health status. In a per-protocol analysis, patients adherent to FMD exhibited higher physical, role, emotional, cognitive, and social functioning; less fatigue, nausea, and insomnia; but no differences in other symptoms. Those in the FMD group felt they had a better understanding of chemotherapy-induced side effects.
Fay-Watt et al. (75)	Single-arm pilot implementation study	Prostate cancer	35 patients with features of metabolic syndrome	FMD for 4 d, once per month	3 mo	FMD was safe, had little or no toxicity, and had a high compliance (29 of 35 completed) . Pre-post decreases in body weight, abdominal circumference, and blood pressure. No pre-post differences in blood lipids of hemoglobin A1c.
Valdemarin et al. (11)	Single arm phase II	Solid tumor or hematologic malignancy (18 different types enrolled)	90 patients undergoing active treatment	FMD for 5 d (4 d prior to and the day of chemotherapy) or 1 d of FMD per month independent of chemotherapy schedule and individualized nutrition advice and a recommended muscle training routine; FMD occurred once per chemotherapy cycle, which was every month or every 3 wk at the most for the first 6 mo of chemotherapy.	Duration of treatment (2-21 cycles)	90% completed at least 1 FMD cycle, 72% completed all planned FMD cycles. FMD was safe (below prespecified threshold of adverse events). Pre-post increase in phase angle (indicator of nutrition status). Pre-post increase in total body water and intracellular water, decrease in fat mass (measured using bioimpedance analysis). No pre-post change in body weight or handgrip strength. No decreases in muscle or fat (measured using CT scans).
Vernieri et al. (12)	Single-arm prospective clinical trial, DigesT	Breast cancer	22 patients with surgically resectable cancer	FMD once for 5 d	12-15 d before surgery	Acute reduction in blood glucose, insulin, and IGF-1 and increase in urinary ketone bodies. Acute reduction in phosphorylated (ie, activated) IGF-1 receptor in tumor tissue. Acute increase in CD8+ T cells, CD8A gene expression, and CD8A:CD68 ratio gene expression from tumor tissue. RNA sequencing revealed acute increase in gene expression signatures associated with a favorable cancer prognosis.
Time-restricted eating (TRE) and prolonged nightly fasting (PNF)
Helo et al. (82)	Retrospective analysis of NHANES cohort study	Participants aged 40 years and older who participated in NHANES III (1988-1994)	7007 people with no history of cancer	Participants were asked how often they ate breakfast: every day, some days, weekends only, rarely, or never.	N/A	Participants who rarely consumed breakfast had a higher risk for all-cause and cancer-related mortality vs those who consumed breakfast every day.
Kirkham et al. (77)	Single-arm pilot study	Breast cancer	22 female cancer survivors, aged 60 years and older, BMI ≥ 25 kg/m2, completion of cardiotoxic treatment (anthracyclines within 1-6 y).	TRE; eat ad libitum between 12 and 8 pm on weekdays and any time of day on weekends; water, black tea, black coffee okay outside eating window.	8 wk	TRE was feasible. Pre-post decrease in Framingham cardiovascular risk. Pre-post decrease in visceral adipose tissue and whole-body fat mass.
Kleckner et al. (76)	Single-arm pilot study	Any type, majority breast cancer enrolled	39 cancer survivors 4-60 mo postcancer treatment with moderate-severe fatigue	TRE; self-selected 10-h fasting window	14 d	TRE was feasible and safe. Pre- to postintervention improvements in fatigue.
O’Donnell et al. (78)	Single-arm pilot study	Breast cancer, early stage (I-III)	40 female survivors, completed surgery, chemotherapy, and/or radiation at least 6 mo prior.	Prolonged nightly fasting for 13 h overnight for 12 wk	12 wk	Prolonged nightly fasting was feasible. BMI, anxiety, depression, and fatigue decreased pre- to postintervention. No pre-post changes in blood-based lipid profiles, HbA1c, leptin, adiponectin, C-reactive protein, or cytokines. No pre-post changes quality of life.
Palomar-Cros et al. (79)	Retrospective analysis of the Multi-Case Control (MCC) study	Prostate cancer	607 prostate cancer cases and 848 population controls	Night-time fasting duration calculated as the time between last eating episode before going to sleep and first episode the following day	N/A	A longer nighttime fast was associated with a reduced risk of prostate cancer An earlier breakfast was associated with a non-statistically significant reduced risk of prostate cancer.
Other fasting regimens
Harvie et al. (71)	Randomized controlled trial	Breast cancer, stage I to III	172 women scheduled for adjuvant or neoadjuvant chemotherapy	Intermittent energy restriction (IER) vs continuous energy restriction (CER) plus physical activity IER: 2 d of fasting (650-1000 kcal/d; scheduled before the chemotherapy on infusion weeks) and 5 d of Mediterranean diet. CER: Mediterranean diet to meet energy needs for weight loss or maintenance.	6 x 3-week chemotherapy cycles	There were no between-group differences in weight loss; when adjusting for total body water the IER group had a greater loss of weight and fat mass. Rates of severe toxicity were similar for cycles 1-3 but fewer in the IER vs CER group during cycles 4-6. There were few between-group differences in fitness measures. There were few differences in cardiovascular risk markers, though serum triglycerides increased in the CER group and were maintained in the IER group.
Mindikoglu et al. (83)	Observational study	Participants did not have a cancer diagnosis	14 adults with metabolic syndrome (8 males and 6 females) who planned to fast during Ramadan	Ramadan fasting; 14 h daily from dawn to sunset.	29 d	BMI, waist circumference, and blood pressure were lower at the end vs beginning of Ramadan. Levels of multiple tumor suppressor and DNA repair gene protein products were higher at the end of Ramadan vs before. Levels of tumor promoter gene protein products were lower at the end of Ramadan vs before. After Ramadan, there was an induction of an antidiabetes proteome response by upregulating key regulatory proteins of insulin signaling and an anti-aging proteome response by upregulating H2B histone proteins.
Riedinger et al. (72)	Randomized controlled trial	Gynecologic cancers (11 ovarian, 8 uterine, and 1 cervical cancer)	24 women undergoing at least 6 planned chemotherapy cycles	24-h water-only fast before and following chemotherapy for a total of a 48 h	For each of six 3-wk chemotherapy cycles	Fasting was feasible, with 95% adherence. There were no between-group differences in weight loss or hospital admissions. There were no adverse events grade 3 or higher in the fasting group. There were fewer treatment modifications in the fasting group. There were no between-group differences in quality of life scores, but there was a pre-post improvement in quality of life for those in the fasting group. There were no between-group differences in treatment outcomes (ie, complete response, partial response, disease progression).
Schreck et al. (69)	Single-arm phase II study	Astrocytoma stage II to IV	25 patients with stable disease after adjuvant chemotherapy	Glioma Atkins-Based Diet (GLAD) for 8 wk. Each week has 2 fasting d (<20% energy needs) interleaved between 5 modified Atkins diet days (≤20 g of carbohydrate per day).	8 wk	48% followed the diet as prescribed, and 72% followed it 6 of 7 days per week. The intervention was deemed safe and well tolerated. The diet led to increases in beta-hydroxybutyrate and acetone (ketone body) concentrations in the lesional and contralateral brain. There were pre-post decreases in hemoglobin A1c and serum insulin. There were pre-post decreases in glutamine processing and phosphocholine in the tumor. There were pre-post increases in glutamine processing in the contralateral side of the brain. There were no differences in lactate, N-acetylaspartic acid, or other metabolites over the course of the study.
Voss et al. (73)	Randomized controlled trial, ERGO2	Malignant glioma (glioblastoma, gliosarcoma, or malignant progression of a lower grade glioma on MRI)	50 patients undergoing reirradiation	A combination of calorie-restricted ketogenic diet (CR-KD) and intermittent fasting (intervention) or a standard diet (control)	3 d CR-KD, 3 d of zero calorie fasting, 3 d CR-KD; radiation occurred on days 4-8 of this 9-d protocol.	CR-KD induced ketosis (elevation in ketone bodies and reduction in glucose). CR-KD was well tolerated with no severe adverse events. CR-KD promoted statistically significantly more weight loss that the control group. Neither progression-free survival at 6 mo nor overall survival was different between the 2 groups. No between-group differences were observed for insulin, IGF-1, cholesterol, high-density lipoprotein cholesterol, or low-density lipoprotein cholesterol.
Voss et al. (74)	Same as cohort as Voss et al. (73), ERGO2					There were pre-post reductions in leptin and insulin and an increase in uric acid in the CR-KD group. There were no between-group differences in quality of life or cognition (eg, memory, attention, language).
Yassin et al. (84)	Retrospective	Chronic myeloid leukemia	49 patients who fasted during Ramadan of 2016, 2017, and 2018	Ramadan fasting, sunrise–sunset, no food or water	30 d	There were no statistically significant differences before vs after Ramadan in regard to white blood cells, neutrophils, basophils, or BCR-ABL1 transcripts (a CML diagnostic).
Zorn et al. (70)	Controlled cross-over	Breast or gynecological cancer (ovarian, cervical, endometrial)	30 women scheduled for chemotherapy	Fasting started 3 d prior to chemotherapy in the evening at 6 pm. The fasting period ended 1 d after chemotherapy at 6 pm, approximately 24 h after the end of drug administration, for a total of 96 ho of fasting (intervention 1), 6 d of a normocaloric ketogenic diet followed by the 96-h fasting regimen (intervention 2), or a normocaloric diet (control); participants were allocated to complete either intervention 1 or intervention 2 and the control.	2 or 3 (half of their total prescribed cycles) x 3- or 4-wk chemotherapy cycles	Fasting was safe and feasible. There was statistically significant loss in fat mass for the fasting cycles, resulting in statistically significant weight loss. Stomatitis, headaches, weakness, and the total toxicity scores were significantly lower in the fasting cycles vs the nonfasting cycles. Insulin and IGF-1 were lower after vs before the fasting period. The ketogenic diet addition to the fasting regimen did not affect outcomes.

^aAC-T = 4 cycles doxorubicin and cyclophosphamide followed by 4 cycles of docetaxel; BMI = body mass index; CML = chronic myeloid leukemia; CT = computed tomography; DIRECT = DIetary REstriction as an Adjunct to Neoadjuvant ChemoTherapy for HER2 Negative Breast Cancer; DigesT = Impact of Dietary Intervention on Tumor Immunity; ERGO2 = Calorie-restricted, Ketogenic Diet and Transient Fasting During Reirradiation for Patients With Recurrent Glioblastoma; FEC-T = 3 cycles of 5-fluorouracil, epirubicin, and cyclophosphamide followed by 3 cycles of docetaxel; FMD = fasting-mimicking diet; IGF-1 = insulin-like growth factor 1; NHANES = National Health and Nutrition Examination Survey; pre-post = pre-intervention to post-intervention; TRE = time-restricted eating.

Several prospective trials have implemented prolonged fasting around chemotherapy or radiation (69,70). Harvie et al. (71) conducted a randomized controlled trial comparing the effect of IF (2 days of 650-1000 kcal/day before chemotherapy infusion) vs continuous calorie restriction (both with a Mediterranean diet) with physical activity on body weight and chemotherapy toxicity. They recruited 172 women with breast cancer stages I-III scheduled for six × 3-week chemotherapy cycles. There was no statistically significant difference in weight change between the groups. The rates of severe toxicity were similar for cycles 1-3 but fewer in the IF vs calorie restriction group during cycles 4-6. Another randomized controlled trial of 24 women with gynecologic cancers found that a 24-hour water-only fast before and following chemotherapy (a total of a 48-hour fast) was feasible, with no grade 3 adverse events or higher and fewer treatment modifications in the fasting group (72). A longer fast was tested in a crossover study using a 96-hour fast (fasting 3 days prior to chemotherapy, the day of chemotherapy, and the day after) or a 6-day ketogenic diet followed by the 96-hour fasting regimen in 30 women with breast or gynecological cancer. The fasting was safe and feasible, and lower toxicity scores were reported during fasting cycles (70). Schreck et al. (69) tested the effects of a 5:2 eating:fasting regimen in which participants consumed 5 days of a modified Atkins diet (≤20 g carbohydrates per day) and 2 days of extremely low calorie (<20% of energy needs) for 8 weeks. They recruited 25 patients with stable astrocytoma after adjuvant chemotherapy. Fasting was feasible, safe, and effective at increasing circulating β-hydroxybutyrate and acetone (ketone body) concentrations. Here, IF also increased glutamine processing in the lesional and contralateral brain and decreased glutamine processing and phosphocholine in the tumor. Further, the diet decreased glycosylated hemoglobin (HbA1c) and serum insulin (69). Another trial combined fasting with a ketogenic diet before, during, and after radiation therapy in 50 patients with malignant glioma undergoing re-irradiation (42,73). The fasting combined with the ketogenic diet was well tolerated, effective at inducing ketosis, promoted weight loss (73), reduced leptin and insulin, and increased uric acid (74).

We identified 7 studies that investigated the use of a FMD (Table 3). All were performed by or in collaboration with Dr Valter Longo who has commercialized a 4-day FMD with a specific macro- and micronutrient composition (10). Specifically, the diet includes plant-based, low amino-acid components including soups, broths, teas, and other liquids. Participants consume approximately 1200 calories on day 1 (3.5:1:2 kcal ratio for carbohydrates:protein:fat) and approximately 200 calories on days 2-4 (80% calories from complex carbohydrates). Of the 5 clinical trials, 3 timed the FMD before chemotherapy (10-12), 1 timed 12-15 days before surgery (12), and 1 was not timed in relation to treatment (75). Collectively, these trials showed that the FMD was effective at increasing circulating ketone bodies, though compliance was variable (20%-83%). The FMD did not increase the rate of toxicities, and the FMD appeared to have additive or synergistic antitumor effects when combined with standard therapies. There were apparent improvements in metabolic measures such as plasma glucose, insulin, IGF-1, and in the longer duration studies, excess body fat. Direct comparative studies of different forms of FMD are needed to determine which components of this diet are necessary for efficacy. Also, FMD meets considerable challenges with patient adherence, so identifying alternative ways to achieve the therapeutic effect of FMD without a near-complete fast is critical to enabling more widespread adoption outside of clinical trials.

Time-restricted eating targets parameters regarding the timing of food consumption and, by extension, the timing of fasting. Timing of time-restricted eating protocols can be self-selected such that individuals define what their eating and fasting windows are (eg, fast for 12 hours at self-selected time of day or night). Prolonged nightly fasting is a distinct time-restricted eating protocol designating that fasting occurs in alignment with circadian rhythm (ie, sleep–wake cycle), such that food and caloric beverages are consumed during the waking (daylight) hours, and fasting occurs during the night. A series of recently published 2022 single-arm pilot studies exploring time-restricted eating among populations of mixed cancer survivors provide preliminary evidence for the feasibility of such interventions. Kleckner et al. (76) explored the use of a time-restricted eating intervention among 39 mixed cancer survivors who were 4-60 months postcancer treatment and were living with moderate to severe fatigue. Participants self-selected their 10-hour eating window, which was maintained across the 14-day trial. This study demonstrated feasibility and safety of intervention delivery and implementation. Further, fatigue decreased from pre- to postintervention. Kirkham et al. (77) recruited 22 female breast cancer survivors (aged older than 60 years, body mass index ≥ 25 kg/m²) to a single-arm, time-restricted eating intervention that consisted of ad libitum consumption between 12 pm and 8  pm on weekdays and anytime on the weekends for 8 weeks. Water, unsweetened tea, and black coffee were allowed during the fasting window. The intervention was feasible, and improvements were observed in visceral adipose tissue, whole-body fat mass, and cardiometabolic outcomes from pre- to post-intervention. O’Donnell et al. (78) conducted a prolonged overnight fasting study with 40 female breast cancer survivors (stages I-III) who had completed cancer treatment (ie, radiation, chemotherapy, surgery) at least 6 months prior. Participants fasted 13 hours overnight throughout the 12-week study. The study was feasible and, although there were no changes in HbA1c, leptin, adiponectin, C-reactive protein, or cytokines, there were improvements in anxiety, depression, and fatigue from pre- to post-intervention.

Specific to the context of meal timing modification, recent epidemiological studies have suggested that the dietary pattern of eating late in the day is associated with elevated risk of cancer (79-81). For example, in a 2018 prospective cohort study, Srour et al. (80) demonstrated that circadian rhythm misalignment or disruption—a probable carcinogenic among humans—may be the result of eating late in the day (ie, late timing of last caloric intake/eating episode) and may be involved in the carcinogenesis of tumor locations (ie, breast, prostate). Helo et al. (82) examined time-restricted eating from an epidemiological standpoint for cancer prevention. They analyzed data from 7007 people with no history of cancer in the US National Health and Nutrition Examination Survey database. They demonstrated that skipping breakfast had a higher risk for all-cause and cancer-related mortality compared with those who consumed breakfast every day. Although skipping breakfast suggests a longer overnight fast, data were obtained by participants responding to the question, ‘How often do you eat breakfast?’ and it is unclear if people considered drinking coffee or tea, which could contribute substantial calories, as eating breakfast. Palomar-Cros et al. (79) conducted a similar analysis on a large, Spanish, multicase-control study. They demonstrated that prolonged nightly coupled with an early breakfast time—aligned with biologically driven circadian rhythms or sleep–wake cycle—may reduce the risk of prostate cancer. Thus, the specific timing of food intake (ie, earlier in the day aligned with diurnal circadian rhythms) may reduce cancer risk.

Muslim populations fast from sunrise to sunset during the 30-day Ramadan holiday, which is essentially a time-restricted eating pattern. Thus, Ramadan provides a unique opportunity for retrospective and prospective epidemiological analyses assessing cancer-related outcomes (83,84). Yassin et al. (84) performed a retrospective study among 49 patients with chronic myeloid leukemia who fasted during Ramadan in 2016, 2017, and 2018 and found no statistically significant differences in neutrophils, basophils, or BCR-ABL1 transcripts (a chronic myeloid leukemia diagnostic) before vs after Ramadan. In an observational study that analyzed data from 14 adults with metabolic syndrome who did not have a cancer diagnosis, Ramadan fasting decreased body weight and waist circumference, improved blood pressure, increased levels of proteins involved in tumor suppression and DNA repair, and decreased proteins involved in tumor promotion (83).

In summary, time-restricted eating interventions have shown the highest degree of adherence and are suggested to be the most sustainable compared with other forms of IF. Thus far, time-restricted eating interventions have been tested for supportive care outcomes but have not yet been tested to enhance the efficacy of chemotherapy. FMD and prolonged fasting diets (> approximately 16 hours) are effective at promoting ketosis, which contributes to enhanced antineoplastic chemotherapy and radiation efficacy. These therapeutic diets tend to have lower rates of adherence but have demonstrated safety and promising outcomes including similar or improved antitumor markers, similar or fewer treatment toxicities, improved energy metabolism, and improved symptom management. All the human clinical trials performed to date have been secondary analyses, observational trials, retrospective studies, or phase I or II randomized trials; none have been confirmatory phase III clinical trials. However, several well-powered studies are underway to address outstanding questions (6) (eg, NCT05114798, NCT05132686, NCT04722341). More extensive data are needed before IF can be incorporated into evidence-based clinical guidelines and routine clinical use outside of closely monitored trials.

A high-priority knowledge gap: the intersection of fasting and cachexia

Cancer cachexia is a multidimensional syndrome characterized by progressive loss of muscle and adipose tissue that is common in patients with advanced disease (85). Patients with cancer cachexia experience reduced physical performance and quality of life, as well as increased toxicity and reduced efficacy during anticancer therapy, leading to shorter survival. The underlying mechanisms are not well understood but include tumor-secreted humoral factors and systemic inflammation, both of which promote well-defined catabolic processes including lipolysis, protein degradation, and anorexia (86). The prevalence of cachexia differs depending on the definition applied and population studied, but estimates range from 50% to 80% of cancer patients (87); patients with pancreatic, esophageal, and lung cancer typically are at the highest risk. There are currently no effective lifestyle or pharmacologic treatments to reverse cachexia. Treatments to mitigate cachexia emphasize nutritional support and supplementation, appetite stimulants, anti-inflammatories, and adapted exercise (85). Although several studies of IF have highlighted beneficial effects on tumor outcomes, IF is often associated with calorie restriction. Therefore, fasting is understandably controversial and understudied in patients with advanced cancer and/or at high risk for cachexia.

Much remains unknown regarding fasting regimens in this population. On the side of potential benefit, some IF regimens alter the timing of nutrient intake, thereby reinforcing circadian rhythms, which are themselves deranged in cachexia and regulate many of the metabolic abnormalities characteristic of cachexia (eg, altered glucose and lipid metabolism) (88). In addition, various IF regimens can enhance chemotherapy efficacy and protect nontumor cells from off-target effects (5). This presents a potential benefit given that toxicity is a major obstacle to life-prolonging treatment of patients with or at risk for cancer cachexia and that off-target effects of chemotherapy can exacerbate cancer cachexia (89,90). Further, in individuals without cancer, muscle and lean body mass are often preserved in IF regimens even if weight loss occurs (7,39,91,92). Yet, although the benefits of fasting interventions often hinge on metabolic adaptations to fasting, metabolic responses to nutrient supply are markedly abnormal in patients with cachexia. Further, fasting regimens often increase autophagy (86,90), which is upregulated in cachexia already.

Moving forward, research in this area is promising but must be undertaken with great care and attention to patient safety. Research is needed to determine the impact of various fasting regimens on metabolic function, quality of life, chemotherapy tolerance, and maintenance of lean body mass in patients at risk for cancer cachexia. A licensed dietitian should be actively involved in patient care to monitor nutritional status and provide appropriate nutrition interventions. A potential next step would be to enroll patients with advanced cancer at risk for cachexia—but without active cachexia—who are beginning systemic chemotherapy. Such patients have a longer anticipated survival and higher functional status than those for whom systemic chemotherapy is not recommended. It would be imperative that the study team carefully monitor potential adverse effects of IF, including loss of weight or lean mass, reductions in caloric intake, and reductions in functional status. These studies would be designed to test hypotheses that IF would reduce chemotherapy toxicity, improve quality of life, and/or slow the progression to precachexia or cachexia.

Supportive care, survivorship, and quality of life

Successful advances in curative cancer treatment over the last few decades have led to a new and pressing need to address quality-of-life challenges in long-term survivorship. Cognitive impairment, fatigue, neuropathy, cardiotoxicity, and other symptoms can be debilitating and persist for years or even decades although the person is cancer-free. The underlying mechanisms for treatment-related toxicities and side effects are complex and many are poorly understood, and research to address these concerns remains a priority for the National Cancer Institute (93). Although IF regimens chiefly dictate the timing of eating, there can be either intentional or unintentional changes in diet quality and quantity. With obesity as a risk factor for at least 11 types of cancer (94), IF regimens that result in a calorie deficit can help achieve loss of excess body weight, and more sustainable regimens such as time-restricted eating can help maintain body weight. Maintenance of a healthy weight can reduce the risk of recurrence as well as address persistence side effects of treatment such as fatigue (95). Without changes in calorie intake, prolonged nightly fasting has pleiotropic effects via its mechanisms to entrain circadian rhythm and rectify central and peripheral pathophysiology that underlies cognitive impairment, fatigue, low appetite, and poor sleep quality, as these symptoms tend to cluster (39,96). Reducing overall late morbidity from treatment toxicity can greatly increase quality of life in survivorship.

Other Considerations

IF may be a beneficial dietary strategy for a number of health-related outcomes within and beyond cancer, however, this approach may be contraindicated and inappropriate for some clinical populations (Box 1) (97). Specifically, IF may not be appropriate for individuals at risk for hypoglycemia and/or those who have a history of or an active eating disorder. All patients with cancer undergoing IF should have a dietitian specialized in oncologic care on their team, and this is most important for patients who are undergoing a new IF regimen and at risk for becoming underweight and have or are at risk for malnutrition (98). The experience of a trained dietitian is extremely valuable in how to recognize and prevent symptoms that may arise during IF and whether certain dietary strategies are helpful to mitigate these.

Box 1.

Intermittent fasting may be contraindicated in participants with these clinical characteristics

• History of an eating disorder

• Pregnancy

• At risk for hypoglycemia

• With or at risk for cachexia

• With or at risk for sarcopenia

• With or at risk for frailty

• With or at risk for malnutrition

Future directions

Future work should establish optimal IF strategies depending on where a person is on the cancer continuum (prevention, active treatment, survivorship, stable disease, metastatic disease). IF interventions have been proposed theoretically for primary and secondary cancer prevention (5,39,99). Although numerous clinical studies are still ongoing, future collaborative consortia-based studies can leverage these to understand how dietary interventions may impact cancer risks and outcomes. These large datasets will be instrumental in gathering epidemiological evidence to support large nationwide studies evaluating the effects of IF on cancer incidence and recurrence. Specifically, it will be important not only to use demographic, oncologic, and dietary information but also to incorporate genetic testing (germline and tumor) and treatment course information in individual-level meta-analysis to improve the power of being able to detect possible benefits from fasting-based interventions.

Features of IF interventions

Details regarding moderators for IF remain undefined. Is time-restricted eating, alternate-day fasting, or a FMD best for the desired outcome? How long should fasting windows be to incur benefits? How frequently should the fasting occur? When should the fasting occur in relation to chemotherapy or other treatment procedures? For time-restricted eating, when should the eating window start? Should it be synced with daylight or sleep patterns? How much nonadherence or laxity is allowable without nullifying the effects of IF? To provide beneficial effects, what duration should the fasting intervention be? What is the sustained effect of the intervention on the outcomes if a person abandons their IF regimen? Should noncaloric sweeteners, caffeine, herbal tea, coffee, vitamins, minerals, and/or supplements be allowed (or encouraged) outside the eating window?

Mechanisms

Fasting-induced alterations in glycemic regulation, circadian rhythm, and nutrient sensing pathways have emerged as interacting mechanisms underlying IF effects on the cancer outcomes, though it will be important to more precisely define and quantify these mechanisms. Measures of eating patterns, morning or evening insulin resistance, actigraphy, cortisol, psychological factors, as well as omics approaches in different tissues, will help delineate and define mechanisms underlying the effects of fasting on metabolism and other clinically meaningful outcomes. Understanding these mechanisms will aid in optimizing IF interventions for specific outcomes and populations.

Comparing IF interventions with established therapies and in combination

Effective phase II clinical trials will lead to phase III clinical trials that quantify the effects of IF compared with more well-established interventions on metabolism, circadian rhythm, and other desired outcomes. These results are necessary before IF can be safely and appropriately incorporated into clinical guidelines. Comparator interventions may include other interventions that focus on the quantity of food (eg, caloric restriction), composition of food (eg, Mediterranean diet, low glycemic index diet), and timing of food (eg, other IF protocols), as well as those that entrain circadian rhythm (eg, bright light therapy) or have pleotropic effects on metabolism (eg, aerobic and/or resistance training exercise). Further research is needed to determine whether IF interventions have synergistic or additive effects with the above lifestyle interventions or pharmaceutical interventions, such as metformin.

Primary prevention

Prevention of cancer is where any intervention could have the largest impact. Although it is premature to dedicate large resources toward a prospective randomized controlled trial on the effects of an IF protocol on disease incidence, we can and should leverage large cohort studies that incorporate food logs with time stamps, as Marinac et al. (68,100) did with the Women’s Healthy Eating and Living trial. Self-administered dietary assessment tools such as the National Cancer Institute’s ASA24 (web-based) (101) and myCircadianClock (app-based) (14) facilitate time-stamped food intake data collection and can be useful in a variety of clinical trials and in the free-living population. In future epidemiological studies, standardization of data collection on mealtime and frequency will be important, for example, using the American Cancer Society Cancer Prevention Study-3 grid tool (102).

Population

Patient characteristics that determine the effect size of IF interventions on clinical outcomes are undefined. Effect modifiers could include biological variables and circadian behaviors such as age, sex, body mass index, sleep (timing, duration, quality, schedule regularity [ie, shift work vs regular schedule vs inconsistent schedule]), regularity of eating schedule, physical fitness and activity, comorbidities, baseline metabolic measures (eg, fasting glucose and insulin, insulin sensitivity), clinical characteristics (cancer site and stage), medications (over the counter, prescription), severity of symptoms, and chronotype (see next paragraph). Properly powered research studies are required to determine which patient populations will benefit the most from IF interventions. Addressing this research gap will allow clinicians to recommend IF to those who will benefit most based on demographics, clinical factors, or biomarkers. This will also save time and resources associated with dietary consultation and monitoring and, importantly, reduce unnecessary burden on patients who are unlikely to benefit from IF.

Chronotype research and its ability to modify the effect of IF is a rich area of exploration to maximize the clinical utility of interventions aimed to entrain circadian rhythm. Chronotype, also referred to as circadian typology, is an expression of the reflection for individual preference with respect to the time of day for the sleep–wake, or activity–rest cycle (103,104). Categorically broken down into 3 subtypes—morning, intermediate (ie, neither), evening—these preferences and subsequent behaviors further dictate various biological mechanisms and outcomes (104). A strong body of research suggests that evening-type individuals frequently engage in dietary habits and eating behaviors associated with weight gain and obesity (eg, excessive caloric consumption late in the day, food consumption prior to bed, longer eating duration) (104) as well as potentially cancer, in part through mechanisms of circadian rhythm misalignment (103). In the context of IF (ie, time-restricted eating), individual chronotype may act as an effect modifier such that those with a preference toward eveningness may benefit to a greater degree from manipulation of food timing as compared with those with a preference toward morningness (ie, majority of caloric consumption earlier in the day).

Outcomes and analysis considerations

Clinical trials that implement IF regimens should select outcomes that are pertinent to the population. For populations who have obesity or diabetes, measures could include body weight, body composition, blood pressure, blood lipid profile, triglycerides, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, HbA1c, fasting blood glucose and insulin, insulin sensitivity and/or resistance (eg, homeostatic model assessment for insulin resistance), use of medications, and patient-reported symptoms (39). For the cancer population specifically, measures could include cancer progression or recurrence, tumor markers, treatment-related toxicities (eg, chemotherapy-related cognitive impairment), and other supportive care outcomes. Because age is a risk factor for cancer, other important measures could include frailty, epigenetic markers of accelerated aging, functional assessments, and other measures. Confounding and/or effect modifying variables described above should be carefully considered in study design, data collection, and statistical analyses.

Summary and conclusions

There is immense potential for IF regimens to be leveraged in cancer, though the benefits and drawbacks of IF are currently being defined. As such, IF has not yet been adequately vetted to be included into any clinical guidelines. Therefore, it is important that patients work with their physician and an oncology dietitian to set goals and monitor progress. There is no doubt that the next decade of research will produce advances in how IF can be used to modulate the physiology of tumor and nontumor cells, improve the efficacy of anticancer treatments, and improve quality of life for those whom cancer has affected.

Data availability

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

Author Contributions

Amber S. Kleckner, PhD (Conceptualization; Writing—original draft; Writing—review & editing), Faiza Kalam, PhD (Conceptualization; Visualization; Writing—original draft; Writing—review & editing), Dara L. James, PhD, MS (Conceptualization; Writing—original draft; Writing—review & editing), Yun Rose Li, MD, PhD (Conceptualization; Writing—original draft; Writing—review & editing), Michael F. Coleman, PhD (Conceptualization; Writing—original draft; Writing—review & editing), Violet A. Kiesel, PhD (Writing—original draft; Writing—review & editing), Elizabeth M. Cespedes Feliciano, ScD, SM (Writing—original draft; Writing—review & editing), Stephen D. Hursting, PhD, MPH (Writing—original draft; Writing—review & editing), and Dorothy D. Sears, PhD (Writing—original draft; Writing—review & editing)

Funding

This work was funded in part by the University of Maryland Baltimore, Institute for Clinical & Translational Research via grant no. UL1TR003098 (pilot funds to ASK), the Maryland Department of Health’s Cigarette Restitution Fund Program (ASK), NIH NCI R35CA197627 (SDH), NIH NCI K01CA226155 (EMCF), NIH NCI K12CA001727 (YRL), the NIH Office of the Director DP5OD033424 (YRL), and NIH NCI T32CA193193 (support to FK).

Conflicts of interest

No authors declare any conflicts of interest.