The CHRONO trial: Protocol for a randomized controlled trial of early time-restricted eating in patients with breast or rectal cancer

Abstract

Fasting combined with anti-cancer therapies improves tumor regression and survival in animals. Fasting sensitizes cancer cells to cytotoxic therapies while promoting stress resistance in healthy cells, a phenomenon known as differential stress resistance. Since prolonged water-only fasting is challenging, intermittent fasting has emerged as an attractive alternative. Here, we describe the protocol of a single-blind, randomized controlled trial testing whether time-restricted eating (TRE) can improve clinical, molecular, and behavioral outcomes in cancer patients. We will enroll 175 adults with newly diagnosed stage I-III human epidermal growth factor receptor 2-positive or triple-negative breast cancer or stage II-IV rectal cancer who are receiving neoadjuvant therapy. Patients will be randomized to early TRE (eTRE; ≤8-hour eating window early in the day) or to eat over a ≥12-hour window (control group) from around the start of treatment until surgery (typically 4–8 months). The primary outcomes are the pathologic complete response, the organ preservation rate, and patient-reported adverse events. Secondary outcomes include clinical response, tumor response, quality of life, and provider-reported adverse events. To test the differential stress resistance theory, we will measure the expression of proteins involved in cell growth, cell death, and stress resistance in tumor and adjacent normal tissue. Finally, we will determine the effects of eTRE on emotional, psychosocial, and lifestyle factors and how these factors influence adherence and clinical responsiveness. This trial will provide critical insight into whether intermittent fasting can potentially improve cancer outcomes, reduce toxicities from anti-cancer therapies, and enhance quality of life in cancer patients. The study is registered on ClinicalTrials.gov (NCT04722341).

1. Introduction

Approximately 1 in 2 men and 1 in 3 women in the United States will develop cancer in their lifetime [1]. Despite advances in early detection and treatment, cancer continues to be the second leading cause of death in the United States [2]. Moreover, treatment-related toxicities remain high, with 50% to 80% of patients experiencing grade ≥3 treatment-related toxicities [3–5], including persistent fatigue [6–9], gastrointestinal and genitourinary adverse events [3,4], neuropathies [10], and cognitive changes [9,11]. Therefore, novel approaches to improve tumor responses while simultaneously reducing side effects from treatment are greatly needed.

Fasting may fill this critical need. Fasting combined with chemotherapy induces tumor regression and long-term survival in animal models, while also protecting healthy cells [12–17]. Fasting decreases glucose, insulin, and growth factors, which increases stress resistance, decreases reactive oxygen species and DNA damage, and promotes survival in healthy cells [12–14,18–23]. Conversely, fasting causes an anti-Warburg effect in cancer cells, which slows cancer cell proliferation, produces more reactive oxygen species and DNA damage, and increases cancer cell death [15,24–27]. In preclinical models, this phenomenon of differential stress resistance (DSR) improves the effectiveness of chemotherapy, radiation, hormone therapy, and tyrosine kinase inhibitors, thereby enhancing tumor regression and promoting survival [12,13,15–22,24,25,28–37]. A few small studies in humans also partially support the DSR theory. Water-only fasting for 48 to 72 hours reduces gastrointestinal side effects, fatigue, and DNA damage in healthy cells [38–40]. However, cancer patients are typically unwilling or unable to do prolonged water-only fasting [41].

A promising alternative to prolonged water-only fasting is intermittent fasting, which is the practice of alternating periods of eating and extended fasting. In preclinical studies, intermittent fasting reduces tumor incidence [21,42–50], slows tumor growth [25,27,33,43,50–59], reduces metastases [58,60], and increases survival [21,25,33,44,55,59,61,62]. In a study of 131 patients with human epidermal growth factor receptor 2 (HER2)-negative stage II-III breast cancer, the fasting-mimicking diet (FMD; repeated cycles of 3–5 consecutive days of a very-low-calorie diet) increased the rate of radiologically complete or partial responses and reduced DNA damage in lymphocytes, while among adherent patients, it also reduced fatigue and nausea [63,64]. A subgroup analysis found that the FMD extended the survival of advanced triple-negative breast cancer patients by 15 months [65]. Other preliminary studies report that intermittent fasting improves antitumor immunity, decreases DNA damage, and reduces the number of chemotherapy dose reductions or delays [66–69], while findings on toxicities [5,40,67,68,70,71] and quality of life are mixed [63,64,66,68,72]. However, adherence to the FMD has been poorer than expected in cancer patients [63], and recently, the FMD was less effective at slowing tumor growth in mice than calorie-matched continuous energy restriction [73], highlighting the need to test other forms of intermittent fasting.

Time-restricted eating (TRE) is a form of intermittent fasting that limits caloric intake to a consistent ≤10-hour daily eating window, followed by a daily fast of ≥14 hours [74]. TRE does not require caloric restriction or water-only fasting and may be easier to adhere to. Clinical trials have reported high adherence to TRE (~70%-90%) [75–84] and improvements in body weight, cardiometabolic health, and quality of life [75–78,83–91]. Further, early TRE (eTRE), which involves eating earlier in the day to align with circadian rhythms in metabolism, may be especially effective, as it improves cardiometabolic health and lowers oxidative stress even when caloric intake is matched to the control group [74,92–95]. So far, research in murine cancer models indicates that TRE decreases tumor growth [25,55,57,96,97], tumor progression [52,57,98], lung metastases [60], and insulin-like growth factor-1 (IGF-1) concentrations [99] and even increases survival [25,55]. Preliminary research from our group suggests that eTRE tends to decrease IGF-1 levels, decreases oxidative stress, and increases autophagy [74,100], while a small study in cancer survivors found that TRE decreases fatigue [101]. However, no study to date has tested the effects of TRE on clinical cancer outcomes in patients undergoing active cancer treatment.

This randomized controlled trial (CHRONO) tests the effectiveness of 8-hour eTRE compared to a control ≥12-hour eating window in patients receiving neoadjuvant therapy for breast or rectal cancer. We focus on breast and rectal cancers because the neoadjuvant treatment regimen provides a unique opportunity to evaluate the combined effects of TRE and cancer treatment on tumor response, since surgery occurs following months of anticancer treatment. Our objectives are 2-fold: (1) determine whether combining eTRE with cancer treatment improves cancer outcomes, treatment-related toxicity, and/or quality of life in cancer patients; and (2) determine whether the underlying molecular mechanisms appear to support the DSR theory. We hypothesize that eTRE will enhance tumor regression, reduce toxicity, and improve quality of life. On a molecular level, we expect eTRE to reduce cell proliferation and induce cell death in tumor cells, while increasing antioxidant defenses and decreasing DNA damage in healthy cells, via modulation of IGF-1 signaling. Finally, the effects of behavioral dietary interventions, including TRE, are often heterogeneous and qualified by a variety of oft-neglected factors. This proposal is significant in considering how emotional, psychosocial, and lifestyle relate to adherence to TRE and, in turn, affect clinical outcomes.

2. Study design and population

2.1. Overview

We will conduct a multicenter, single-blind, randomized controlled superiority trial comparing the effectiveness of eTRE (consistent ≤8-hour daily eating window early in the day) with eating over a ≥12-hour eating period (control group) in cancer patients undergoing neoadjuvant therapy. We will target enrolling 175 women and men newly diagnosed with stage I-III HER2-positive or triple-negative breast cancer or stage II-IV rectal cancer who are undergoing neoadjuvant therapy at Cedars-Sinai Health System (CSHS) or its affiliate sites (Torrance Memorial Medical Center and Huntington Hospital) in Los Angeles, CA, and the University of Alabama at Birmingham (UAB) in Birmingham, AL. Participants will be randomly assigned to the control or eTRE groups in a 1:1 allocation ratio. Participants will be enrolled either shortly before or within 4 weeks after starting oncologic treatment and will follow their assigned eating schedule until the actual or planned time of surgery, typically 4–8 months later. During the study, participants will receive one-on-one behavioral counseling from a trained behavioral counselor. The primary endpoints are (1) a composite outcome of the percentage of patients who achieve a pathologic complete response (pCR) and/or organ preservation and (2) treatment-related toxicity, as assessed by the Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO–CTCAE) [102,103]. Secondary endpoints include clinical response, tumor response, quality of life, and provider-reported adverse events. Additional endpoints include molecular and behavioral mechanisms linking TRE to treatment tolerance and tumor response. The study was approved by the Institutional Review Boards of Cedars-Sinai Medical Center (STUDY00001187), the University of Alabama at Birmingham (IRB-300007335), and was preregistered on ClinicalTrials.gov (NCT04722341) a year before recruitment began.

2.2. Recruitment

Patients will be recruited through oncologist and surgeon referrals, weekly department/division meetings where patient cases are reviewed, and/or via queries of electronic medical records (EMR). We originally planned to enroll only patients with rectal cancer since rectal cancer has a moderate pCR rate of ~30% [104–106] and may be more responsive to therapeutic interventions than cancers with very low or high pCR rates. Moreover, most rectal cancer patients receive total neoadjuvant therapy, which gives us the rare opportunity to determine the longer-term effects of a dietary intervention on molecular markers in tumor tissue. We began recruiting patients with rectal cancer in January 2022, with an initial target of 300 patients. However, enrollment was slower than expected. To improve recruitment, we attempted to forge partnerships with local hospital systems, and we also tweaked some of our eligibility criteria and the protocol to reduce participant burden. Nonetheless, recruitment remained slow. Therefore, in December 2023, we expanded the eligibility criteria to include HER2-positive and triple-negative breast cancer, which have similar pCR rates (~40%-65%), thereby preserving statistical power for the primary endpoint of pCR. Despite this, enrollment remained slower than expected, so the target was eventually revised to 175 patients. Recruitment will continue until the budget runs out in 2026 or 2027.

2.3. Study population

Eligible participants will be adults (≥18 years) with a body mass index ≥18.5 kg/m² and pathologically-confirmed stage I-III HER2-positive or triple-negative breast cancer or stage II-IV rectal cancer per the American Joint Committee on Cancer (AJCC) criteria (version 8). To be eligible, patients with breast cancer must be treated with neoadjuvant therapy, while patients with rectal cancer must be treated with total neoadjuvant therapy with a 5’-fluouracil-based regimen. Patients with a prior history of radiation, chemotherapy within the last year, or a secondary malignancy will be excluded. The complete list of eligibility criteria is provided in Table 1. All participants must provide written informed consent prior to participating and will receive modest compensation for participating in the study.

Table 1 -.

Complete list of eligibility criteria for participants to enroll in the trial.

Inclusion criteria

• Histologically-confirmed stage I-III HER2-positive or triple-negative breast cancer or stage II-IV rectal cancer per AJCC criteria (version 8)

• Receiving either neoadjuvant therapy with curative intent (breast cancer patients) or total neoadjuvant therapy with a 5-fluorouracil-based regimen and curative intent (rectal cancer patients)

• Treatment-naïve or completed ≤4 weeks of oncologic treatment

• Aged ≥18 years

• Body Mass Index ≥18.5 kg/m2

Exclusion criteria

• Active secondary malignancy other than non-melanoma skin cancer or in situ cervical cancer within the last year

• Chemotherapy within the past 12 months

• Prior pelvic radiotherapy

• Allergic reaction to any of the treatment agents

• Pre-existing grade ≥3 neuropathy

• Uncontrolled congestive heart failure defined as NYHA Class III or greater

• Predisposing colonic or small bowel disorders with severe or rapidly worsening symptoms that are not related to current cancer symptoms

• Gastrointestinal tract perforation within the past 12 months

• Receiving parenteral nutrition or enteral feeding or using any similar nutritional supplement

• Currently receiving or recently received (£4 weeks) a new investigational drug, device, or treatment protocol

• Pregnant or breastfeeding

• Eats within a <10-hour period on most days

• Overnight shiftwork more than 1 d/wk on average

• Psychiatric, sleep, or substance abuse disorders that would interfere with participating in the study

• Medical condition or laboratory abnormality that could impact participant safety or data validity, in the opinion of the medical investigators

Abbreviations: AJCC, American Joint Committee on Cancer; HER-2, human epidermal growth factor receptor 2; NYHA, New York Heart Association.

3. Interventions

3.1. Cancer treatment

All participants will receive treatment for their cancer as directed by their physicians at their respective trial sites. Patients with rectal cancer must receive total neoadjuvant therapy, which consists of separate chemotherapy and radiotherapy regimens before surgery. All permutations of the treatment regimens are permitted, including radiotherapy before chemotherapy, radiotherapy after chemotherapy, and both short-course and long-course radiation. Patients with breast cancer must receive neoadjuvant therapy involving chemotherapy, which could be paired with either immune checkpoint inhibitors (triple-negative breast cancer) or HER2-targeted monoclonal antibodies (HER2-positive breast cancer). While there is heterogeneity in these treatment protocols, we needed to balance the desire for homogeneity against the pragmatic need to enroll sufficient patients. We will attempt to enroll most participants shortly before starting cancer treatment (typically 1–3 weeks), though patients may enroll later as long as they have completed ≤4 weeks of treatment.

3.2. Randomization

Participants will be randomly assigned in a 1:1 allocation ratio to eTRE (≤8-hour eating window) or a control eating schedule (≥12-hour eating window). Importantly, the pCR rate varies among cancer types/subtypes and treatment regimens. The pCR and clinical complete response rates for rectal cancer treated by total neoadjuvant therapy are ~30% [104–106] and 41% to 53% [104], respectively, while the pCR rates for triple-negative, triple-positive, and HER2-positive breast cancer are ~40%, ~40%, and ~65%, respectively [107–114]. Therefore, we will stratify randomization on cancer type, treatment regimen, and other salient factors. For patients with breast cancer, randomization will be stratified by trial location (California vs Alabama), molecular subtype (HER2-positive vs triple-negative), and stage (I vs II vs III). For patients with rectal cancer, randomization will be stratified by trial site, sex, treatment sequence (radiotherapy before vs after chemotherapy), and T stage (T1-T3 vs T4), where T stands for tumor and reflects the size and extent of the original primary tumor. We will stratify by treatment sequence because receiving radiation before chemotherapy may increase the pCR rate [115,116]. Randomization will be performed by the study statisticians (M.A.D. and J.S.R.) using computer-generated permuted blocks of 2 and/or 4. It will be concealed from participants until all baseline data have been collected. Participants will be randomized by their behavioral counselor during the first behavioral counseling session.

3.3. Early time-restricted eating

Participants randomized to the eTRE group will be instructed to eat within a consistent ≤8-hour eating window (≥16-hour daily fast) at least 6 days/week. To account for interindividual differences in sleep, social activities, and work obligations, the eating windows will be timed relative to sleep. Participants’ eating window must start within 2 hours of waking up and no later than 9 am. During their daily fasting period, participants may consume only water and zero-calorie, caffeine-free beverages, such as unsweetened tea. Aside from these prescriptions, no set number of snacks, meals, or calories will be prescribed, and participants will be asked to refrain from making changes in their usual diet. There are no restrictions on the types or amounts of food participants can eat.

3.4. Control eating schedule

Participants randomized to the control group will be required to eat over a ≥12-hour daily period at least 6 days/week. The ≥12-hour period does not need to be consistent and can vary from day to day. The control group was designed to mimic typical eating habits in the United States, as data suggest that the median American eats over a 12-hour daily period [117]. Aside from these prescriptions, no set number of snacks, meals, or calories will be prescribed, and participants will be asked to refrain from making changes in their usual diet. There are no restrictions on the types or amounts of food participants can eat.

3.5. Behavioral counseling

Participants in both groups will meet with a behavioral interventionist (overseen by S.J.S.) weekly for the first month and every other week thereafter. The overarching goals of these counseling sessions are to ensure that the study arms are implemented as intended and to foster adherence. During the initial 30-minute session, the interventionist will explain the assigned intervention, answer questions, and provide a written, personalized meal timing schedule based on each participant’s unique needs and preferences. The personalized handout displays the participants’ eating schedule and their preferred break day to accommodate social activities. Subsequent counseling sessions will be 15–30 minutes long and will focus on reviewing the participant’s daily meal timing schedule and providing support to promote adherence using a variety of strategies, including problem solving, shaping and goal setting, reminders, and enlisting support from others. The counseling sessions will also be used to recollect any missing data on the weekly meal timing surveys.

3.6. Adherence to eating windows

Participants in both groups will self-report their adherence via a weekly survey. Participants will record the time they start and finish eating every day and will document any reason(s) for nonadherence. Participants may complete the survey via REDCap software, by text message, or on paper, as preferred. The behavioral interventionists will review and discuss participants’ adherence during the counseling sessions to clarify and confirm recorded adherence. We will quantify adherence in 2 ways: (1) the percentage of adherent days and (2) the median eating duration. For the former metric, participants will be counted as adherent if they follow their assigned eating window within 30 minutes, while days with missing surveys will be scored as nonadherent. Adherence will be treated as both a tertiary outcome and as a moderator of the relationships between intervention arms and outcomes. See Sections 4.6.1 and 5.4 for further details.

3.7. Safety

Participants’ safety will be monitored by their health care providers as part of the standard of oncologic care. The behavioral counselors will monitor body weight during counseling sessions. Although the risk of cachexia in our specific patient population is unknown, we expect cachexia to be uncommon among our patients, based on provider experience. Should a participant lose >10% of initial body weight, the interventionists will notify the clinical care team to determine if early removal is warranted.

4. Assessments

4.1. Overview

All study data will be collected from 4 sources: (1) EMR, (2) in-person testing visits, (3) tumor specimens, and (4) weekly surveys (Table 2). From the EMR, we will extract data on medical history, treatment and surgical regimens, and clinical outcomes. From tumor specimens, we will measure several molecular endpoints pertinent to testing the DSR theory. On weekly surveys, patients will self-report their meal timing adherence and any treatment-related side effects. All other study data will be collected via 3 in-person testing visits.

Table 2 -.

Study assessment schedule and frequency from enrollment to final data collection.

Assessments	Baseline	Intermediate	Surgery
Covariates
Demographics	X
Medical history	X
Oncologic treatment		Continuous
Primary endpoints
Pathologic complete response and organ preservation rate			X
Patient-reported adverse events		Weekly
Secondary endpoints
Clinical response			X
Tumor response			X
Provider-reported adverse events		Continuous
Health-related quality of life	X	X	X
Molecular outcomes
Complete blood counts		Repeatedly
Insulin-like growth factor 1 and binding proteins	X	X	X
Protein expression related to cell growth/death	X		X
Total antioxidant capacity	X	X	X
Cardiometabolic outcomes
Body weight		Repeatedly
Blood pressure and heart rate	X	X	X
Psychosocial and behavioral outcomes
Adherence		Weekly
Fatigue	X	X	X
Physical functioning	X	X	X
Social functioning	X	X	X
Mood	X	X	X
Food intake	X	X	X
Appetite and eating-related behavior	X	X	X
Physical activitya	X		X
Sleepa	X		X
Exit interview			X

^aAccelerometry-derived estimates of physical activity and sleep are measured only at the baseline and surgery visits.

Testing visits will be conducted at 3 timepoints: baseline, an intermediate timepoint, and shortly before surgery. As shown in Fig. 1, the 3 testing visits will be timed as follows. The baseline testing visit will be conducted either shortly before oncologic treatment starts or no more than ≤4 weeks after starting treatment, with randomization typically occurring 1–2 weeks thereafter the baseline testing visit. The intermediate testing visit will be scheduled at the anticipated midpoint between the start of oncologic treatment and the anticipated date of surgery, typically 2–4 months after starting treatment. The final testing visit will be scheduled after treatment is completed and approximately 1–3 weeks before surgery (or when surgery would have been scheduled, if there is no longer any resectable tumor or the patient has declined surgery). The final testing visit will, therefore, typically be 4–8 months after the baseline testing visit. Patients must fast for at least 8 hours prior to each testing visit. At each testing visit, we will measure body weight, blood pressure (BP), and heart rate, and draw blood to perform a complete blood count and measure serum endpoints. We will also administer questionnaires and 3-day food records and affix accelerometers to measure physical activity and sleep. To reduce participant burden, participants may complete their questionnaires at home either, via REDCap or on paper, and they may mail study equipment and/or completed questionnaires back to the study team using prepaid envelopes. In addition, participants who live more than 1–2 hours away from the trial site or who face substantial burden may forgo the intermediate testing visit and instead complete the questionnaires at home (i.e., they may forgo the blood draw and anthropometric assessments).

Fig. 1 – Timeline of assessments in relation to eating timing intervention and oncologic treatment.

Testing visits will be conducted at 3 timepoints: baseline, an intermediate timepoint, and shortly before surgery. The baseline testing visit will be conducted no more than ≤4 weeks after starting treatment, with randomization occurring 1 to 2 weeks after the baseline testing visit. The intermediate testing visit will occur midpoint between the start of oncologic treatment and the anticipated date of surgery, 2 to 4 months after starting treatment. The final testing visit will be scheduled after treatment is completed and approximately 1 to 3 weeks before surgery, or when surgery would have been scheduled.

4.2. Covariates, anthropometrics, and vital signs

4.2.1. Covariates

During the first testing visit, we will collect self-reported demographic information, including age, race, ethnicity, marital status, employment status, education level, income level, health insurance status, and number of children. We will also collect self-reported data on prior history of cancer, concurrent diseases, family history of cancer, medication use, baseline meal timing habits, and smoking/tobacco use. From the medical record, we will extract data on medical history, including the date of diagnosis, date of biopsy, Eastern Cooperative Oncology Group (ECOG) Performance Status score, microsatellite instability status (rectal cancer), clinically actionable mutations (Kirsten rat sarcoma viral oncogene homolog [KRAS] [rectal cancer] and estrogen receptor [ER], progesterone receptor [PR], HER2, and antigen Kiel 67 [Ki-67] [breast cancer]), disease stage (tumor node metastatis [TNM] staging), histology, tumor stage, tumor grade, tumor size and/or location, number of affected lymph nodes, suspicious extramesorectal lymph nodes (rectal cancer), lymphovascular invasion, and perineural invasion. Finally, we will extract data on oncologic treatment, including chemotherapy (drugs, doses, dates, number of cycles, and dose reductions), radiation therapy (dates, course, technique, target location, sequence, number of fractions, total dose, boosts, dose interruptions, and concurrent chemotherapy), and surgical treatment (date, surgical procedures, and approaches).

4.2.2. Anthropometrics and vital signs

Body weight, BP, and heart rate will be assessed both as covariates and as cardiometabolic outcomes. At each in-person testing visit, we will measure metabolic weight in light clothing following at least 8 hours of fasting. In addition, interim (non-fasting) body weights will be extracted from patients’ medical records. We will also measure systolic and diastolic BP at each testing visit, using an automated BP monitor, following the 2019 American Heart Association and American College of Cardiology guidelines [118]. After a 5-minute rest, BP and heart rate will be measured in triplicate, with 1–2 minutes separating each measurement. We will use the average of the 2 closest measurements. If we cannot collect BP data at an in-person testing visit, we will instead extract it from the nearest patient visit reported in the EMR.

4.3. Primary outcomes

4.3.1. pCR and organ preservation rate

The first primary endpoint will be a composite outcome of pCR and/or organ preservation, which is assessed at the end of treatment. The composite outcome will be defined as the percentage of patients who achieve either a pCR (breast or rectal cancer) and/or organ preservation (rectal cancer only). A pCR will be defined as the absence of residual invasive cancer on hematoxylin and eosin evaluation of the complete resected surgical specimen and all sampled regional lymph nodes. Organ preservation will be defined as achieving a complete or near-complete clinical response at the end of neoadjuvant therapy and subsequently undergoing nonoperative management. Complete or near-complete clinical responses will be determined through physical exam, endoscopic, and magnetic resonance imaging (MRI) findings, using the criteria articulated in the National Comprehensive Cancer Network guidelines (see Section 4.4.1).

4.3.2. Patient-reported adverse events

The other primary endpoint is patient-reported toxicity. Patients will document the frequency, severity, and interference of a few dozen treatment-related adverse events (AEs)—including gastrointestinal, genitourinary, musculoskeletal, and neurologic—over the past week using the PRO–CTCAE (version 5) [119]. Table 3 shows the specific AEs that will be assessed; patients can also report other symptoms not listed. Studies have demonstrated the feasibility, validity, and reliability of the PRO–CTCAE in cancer patients [103,119–121]. Patients will complete the PRO–CTCAE at baseline before starting treatment and then weekly once treatment begins until the time of surgery. The weekly surveys will be timed relative to the start of each treatment segment. Data for each symptom will be converted into a single score using a composite grading algorithm, and then the individual scores will be summed into a single composite score for each timepoint [122].

Table 3 -.

List of patient-reported outcomes from the common terminology criteria for adverse events symptoms assessed

• Problems with tasting food or drink

• Decreased appetite

• Nausea

• Vomiting

• Heartburn

• Increased passing of gas (flatulence)

• Bloating of the abdomen (belly)

• Constipation

• Loose or watery stools (diarrhea)

• Pain in the abdomen (belly area)

• Loss of control of bowel movements

• Rash

• Dry skin

• Itching

• Acne or pimples on the face or chest

• Skin burns from radiation

• Numbness or tingling in hands or feet

• Headache

• Dizziness

• Problems with concentration

• Problems with memory

• Pain

• Aching muscles

• Aching joints (such as knees, elbows, shoulders)

• Insomnia (including difficulty falling asleep, difficulty staying asleep, or waking up early)

• Fatigue, tiredness, or lack of energy

• Anxiety

• Feel that nothing could cheer you up

• Sad or unhappy feelings

• Difficulty getting or keeping erection

• Decreased sexual interest

• Pain with sexual intercourse

• Unable to have orgasm

• Mouth/Throat sores

• Urinary frequency (urinating too often)

• Urinary urgency (strong, sudden urges to urinate)

• Painful urination

• Other symptom(s)

4.4. Secondary outcomes

4.4.1. Clinical response

Clinical response to anticancer treatment will be assessed at the end of treatment and categorized as complete, near-complete, partial, stable, or progressive disease. A complete clinical response is defined as the clinical absence of residual disease on physical examination, endoscopic examination, and/or repeat MRI following completion of neoadjuvant therapy. A near-complete response is defined as nearly all tumors have regressed, but tiny residual abnormalities remain. A partial response is defined as a ≥30% decrease in the size or extent of the tumor, while progressive disease is defined as a ≥20% increase in the size of the longest diameter of the tumor. To quantify clinical response, we will assess the objective response rate, which is the percentage of people with a complete response, near-complete, or partial response. In addition, we will report the percentages of patients who have complete responses, complete or near-complete responses, and progressive disease. As clinical responses are less commonly assessed in breast cancer patients, clinical responses for these patients may be estimated by study physicians and/or MRI technicians.

4.4.2. Tumor response

We will also assess tumor response. In patients with rectal cancer, tumor response will be assessed using the MRI Tumor Regression Grade, which is a 5-point scale quantifying the extent of remaining visible tumor and fibrosis, with 1 corresponding to a complete radiologic response. It will be assessed following the end of treatment. As a secondary measure, we will also use the Neoadjuvant Rectal (NAR) score, which is a validated surrogate endpoint reflecting treatment response [123]. It takes the pathologic nodal status and the clinical and pathologic T-stage as inputs and produces a value on a 0–100 scale. It has previously been shown to be a better predictor of oncologic outcomes than pCR [123–125]. NAR will be assessed at baseline and postintervention. In patients with breast cancer, tumor response will be assessed using the residual cancer burden (RCB), which reflects the amount of invasive cancer remaining after neoadjuvant chemotherapy. The RCB is a 5-point scale ranging from 0 to 4 derived from the area of the primary tumor bed, overall cancer cellularity, percentage of cancer that is in situ disease, the number of positive lymph nodes, and the diameter of the largest metastasis [126,127]. It quantifies the response to treatment, with 0 corresponding to a pathologic complete response, and is highly prognostic of survival.

4.4.3. Provider-reported adverse events

As a secondary measure of toxicity, we will analyze provider-reported AEs between the start of treatment and the day before surgery, which we will extract from the EMR. Provider-reported AEs will be graded using the CTCAE (version 5) [128]. We will analyze the incidence of total AEs, grade 1–2 AEs, and grade 3–4 AEs [129]. We will also calculate the Toxicity Index (TI) [130,131]. The TI accounts for the grade, frequency, and cumulative burden of all toxicities during the treatment. It is the weighted sum of the ordered toxicity grades, which can be divided into 2 components: an integer portion (the maximum grade) and a decimal portion (repeated and lower-grade AEs), and it ranges from 0 to 5.83. The TI has been validated in rectal and breast cancer patients receiving chemoradiation and standard chemotherapies and has been reported to provide greater statistical power than other traditional approaches (e.g., maximum grade) [132,133]. We will also report toxicities by organ or system, particularly gastrointestinal, genitourinary, endocrine, musculoskeletal, breast, skin, and hematologic.

4.4.4. Health-related quality of life

We will assess health-related quality of life at each of the 3 testing visits. We will administer the European Organization for Research and Treatment of Cancer Quality of Life core questionnaire C30 (EORTC QLQ-C30) [134] to all patients, the breast cancer-specific questionnaire (EORTC QLQ-BR23) [135] to breast cancer patients, and the colorectal cancer-specific questionnaire (EORTC QLQ-CR29) [136] to rectal cancer patients. The QLQ-C30 incorporates 9 multi-item scales: 5 functional scales (physical, role, cognitive, emotional, and social functioning), 3 symptom scales (fatigue, pain, and nausea and vomiting), and a global health-related quality-of-life scale. The QLQ-BR23 and QLQ-CR29 comprise a series of functional scales and symptom scales. For this study, we will focus our analysis on the most prevalent breast and rectal cancer treatment-related symptoms [137], including fatigue, pain, and gastrointestinal-related items.

4.5. Molecular outcomes

As tertiary outcomes, we will assess immune cell counts and molecular hallmarks of the DSR theory. As part of testing the DSR theory, we will perform mediation analyses to determine whether IGF-1 mediates the effects on clinical and molecular outcomes.

4.5.1. Complete blood counts (CBCs) and selected blood biomarkers

CBCs will be performed by each site’s Clinical Laboratory Improvement Amendments (CLIA)-certified hospital laboratory. If a patient has a CBC performed within the 2 weeks prior to a testing visit, those CBC values will be used in lieu of obtaining a fresh blood specimen. We will also use all CBC data collected between the time of consent and surgery in the analyses. Finally, we will extract all data on carcinoembryonic antigen (CEA) and carbohydrate antigen 19–9 (CA 19–9) from the time of consent until surgery for rectal cancer patients from the medical records.

4.5.2. Serum IGF-1 signaling

At each testing visit, we will draw blood to measure IGF-1 and its binding proteins 1 (IGFBP-1) and 3 (IGFBP-3). Analytes will be measured in serum using ELISA kits from American Laboratory Products Company (ALPCO; Salem, NH, USA).

4.5.3. Tumor tissue collection

Formalin-fixed paraffin-embedded tumor tissue will be obtained from available biopsies (baseline) and surgeries (postintervention). A pathologist will cut a 5-mm slice containing both tumor tissue and adjacent normal tissue and will place the tissue on unstained slides for analyses of protein expression.

4.5.4. Protein expression

To test the DSR theory, we will measure molecular markers representing cell growth and death and stress resistance, including Ki-67 (cell proliferation), caspase 3 (apoptosis), the ratio of microtubule-associated protein 1 light chain 3 I (LC3-I) to II (LC3-II) (autophagy), and gamma-phosphorylated histone H2A histone family member X (γ-H2AX; DNA damage). Protein expression will be measured in formalin-fixed paraffin-embedded tumor and adjacent normal tissue obtained from both the biopsy (baseline) and surgical resection (postintervention) using Nanostring’s GeoMx technology (Nanostring Technologies, Inc.; Seattle, WA, USA). As the assays will be optimized for this study, the methods will be described in future manuscripts.

4.5.5. Total antioxidant capacity

As a proxy for stress resistance in healthy cells, we will measure total antioxidant capacity in serum and/or peripheral blood mononuclear cells (PBMCs) collected at each testing visit. PBMCs will be isolated from whole blood using Leucosep tubes (STEMCELL Technologies Inc.; Cambridge, MA, USA), following the manufacturer’s instructions, as described previously [138]. Total antioxidant capacity will be measured by running colorimetric assays (Catalog #ab65329; Abcam Limited; Cambridge, MA, USA) in duplicate on a microplate reader.

4.6. Psychosocial and behavioral outcomes

Psychosocial and behavioral outcomes will be assessed at each of the 3 testing visits. The only exceptions to this general rule are as follows: (1) adherence will be assessed daily and (2) accelerometry will be performed only twice (baseline and postintervention).

4.6.1. Adherence

Participants will self-report their adherence to their assigned eating window from randomization through the day before surgery. We will quantify adherence in 2 ways: (1) the percentage of adherent days and (2) the median eating duration. Adherence will be treated as both a tertiary outcome and as a moderator of the relationships between intervention arms and outcomes. As a tertiary outcome, adherence data will be interpreted along a continuum, ranging from excellent (~6 days/week), good (~5 days/week), acceptable (~4 days/week), fair (~3 days/week), to poor (≤2 days/week). As a moderator, we will explore how adherence moderates the effects of the intervention on clinical and behavioral outcomes, as described in Section 5.4.

4.6.2. Mood, fatigue, and physical and social functioning

We will use the National Institute of Health Patient-Reported Outcomes Measurement Information System (PROMIS) subscales to assess anxiety, fatigue, and physical function (e.g., ability to complete chores, going up/down the stairs, walking for 15 minutes, running errands). PROMIS was developed to standardize the assessment of patient-reported outcomes related to physical, mental, and social health using Likert-scale questions [139–141]. Total scores are converted to age-standardized T scores, for which the normative mean is 50 and the standard deviation is 10. To assess social functioning, we will administer the University of California at Los Angeles (UCLA) 3-Item Loneliness Scale [142] and the PROMIS Satisfaction with Participation in Discretionary Social Activities (Short Form 7a). Finally, we will measure the severity of depressive symptoms using the Patient Health Questionnaire8 (PHQ-8). We will assess changes in mood, fatigue, and physical and social functioning across intervention arms (eTRE vs control) as tertiary outcomes.

4.6.3. Dietary intake

Dietary intake will be assessed using 3-day food records in conjunction with each of the 3 testing visits. Participants will record all foods and beverages consumed, and a dietitian will review the logs with participants to ensure accuracy and completeness. Data will be analyzed using the Nutrition Data System for Research software, developed by the Nutrition Coordinating Center (NCC) at the University of Minnesota. We will assess changes in total energy intake, macronutrient composition, and the Healthy Eating Index (HEI 2015) [143,144], as well as changes in key food groups and nutrients, such as added sugar, saturated fat, fiber, fruits, vegetables, solid fats, alcoholic beverages, and sodium.

4.6.4. Appetite and eating-related behavior

We will use the Dutch Eating Behavior Questionnaire to assess eating behaviors along 3 dimensions: emotional, external, and restrained eating [145]. To assess appetite, participants will retrospectively report how frequently they experienced various degrees of hunger and fullness over the last week on a 5-point Likert scale. They will also report their typical levels of hunger and fullness during their eating and fasting windows on a 100-point visual analog scale (0 = not at all to 100 = extremely) adapted from Flint et al. [146].

4.6.5. Physical activity and sleep

We will use an accelerometer and questionnaires to measure physical activity and sleep. Participants will wear an accelerometer (ActiGraph GT9X Link wearable device; ActiGraph, LLC; Pensacola, FL, USA) on their nondominant wrist for 1 week at baseline and postintervention. We will use an open-source and validated set of algorithms to assess sleep and physical activity outcomes (GGIR package for R) [147–149]. Physical activity will be assessed using metabolic equivalent units and the percent of time spent in sedentary, light, moderate, and vigorous activity, as assessed by the Friedson algorithms [150]. We will also measure self-reported physical activity using the General Physical Activity Questionnaire, sleep quality using the Pittsburgh Sleep Quality Index, and sleep duration and timing using the Munich Chronotype Questionnaire.

4.6.6. Qualitative experiences with assigned meal timing intervention

We will conduct semi-structured interviews with participants at the end of the meal timing interventions. These interviews will present open-ended questions probing participants’ experiences with their assigned eating schedule in the context of their cancer treatment, symptoms, and side effects, as well as their roles, values, and routines. Participants will also be asked to discuss facilitators and challenges they faced (if any) in adhering to their assigned eating window, as well as any positive or negative experiences. Interviews will be conducted via teleconference and will last up to 45–60 minutes. Interviews will be audio-recorded and transcribed verbatim, and Dedoose software (Los Angeles, CA, USA) will be used to analyze the qualitative data. Data analysis will be informed by a social constructivist, interpretive framework consistent with grounded theory [151,152], and interpretive description [153]. Two researchers trained in qualitative methodology will independently code the transcripts and develop a codebook using a priori codes from background concepts relevant to the intervention (e.g., social components of eating) and in vivo codes that emerged during the interviews. Any discrepancies in coding will be resolved by a third member of the study team with qualitative expertise. Themes that we uncover from this qualitative investigation will inform the interpretation of acceptability and feasibility data.

5. Data management and statistical methods

5.1. Data management

To manage study data, we will use REDCap electronic data capture tools hosted at Cedars-Sinai Medical Center (CSMC) [154].

5.2. Blinding

All data will be collected and analyzed blinded, except for (1) adherence and body weight data, which are collected and reviewed by the behavioral counselors (overseen by S.J.S.); and (2) dietary intake data, which is collected by a dietitian. All other research staff, health care providers, study investigators, and technicians will be blinded.

5.3. Power calculation

We powered the study to detect changes in the pCR and organ preservation rate and PRO–CTCAE scores, using 2-sided tests and a type I error rate of α = 0.05 for each individual outcome. Based on preliminary data, we estimated the pCR and organ preservation rate for the control group to be 40% [150,151]. Using Fisher’s exact test, a sample size of 158 completers (n = 79 completers per group) provides 80% power to detect a 23% between-group difference in the pCR and organ preservation rate (i.e., increase from 40% to 63%). From a clinical perspective, we are powered only to detect large improvements in the pCR and organ preservation rate. For the PRO–CTCAE score, since the expected mean differences and variances are unknown, we assumed asymptotic normality and calculated the minimum detectable effect size for the group effect (i.e., the adjusted marginal means for the group effect across all other factors in a linear mixed-effects model). Using an independent samples t test with equal variance, the minimum detectable effect size for the PRO–CTCAE score is d = 0.49, which is a medium effect size. We expect to be able to detect even smaller effect sizes for the PRO–CTCAE since we will analyze the data using repeated measures methods. In addition, we performed a power calculation for one of the secondary outcomes: health-related quality of life as assessed by the EORTC QLQ-C30. Assuming a within-group standard deviation of 23.9 [152] and using an independent samples t test with equal variance, we will have 80% power to detect differences of 7.3, 8.3, 9.2, and 10.0 for intraclass correlation coefficients of 0.2, 0.4, 0.6, and 0.8, respectively, to estimate the effective sample size with 3 observations per person. Assuming a 10% dropout rate, we will therefore enroll 175 patients in order to have 158 completers. Akin to other trials [155,156], we did not adjust for multiplicity in the primary outcomes, and the combined false-positive rate remains well controlled at 9.75%. Since we are powered to detect large differences in the pCR, organ preservation rate, and small-to-medium differences in the PRO–CTCAE score, we view our trial as being somewhere between a confirmatory and an exploratory trial, but not fully either type of trial.

5.4. Statistical analyses

All analyses will be performed blinded, using 2-sided tests and α = 0.05 in the software program R [157]. The main analyses for the primary outcomes and secondary outcomes (except for quality of life). These main analyses will be supplemented by per-protocol analyses with participants who completed the study. Participants will not be excluded based on adherence to eTRE or Control. Rather, adherence will be treated as both a tertiary outcome and as a moderator of the relationships between intervention arms and outcomes as described below.

Descriptive statistics will be presented for both quantitative and qualitative variables. Contingency tables for treatment and adverse event data aggregated by type, grade, system, or term will be summarized using correspondence analysis [158].

We will interpret the intervention’s effectiveness by collectively considering the primary and secondary outcomes, as they are intertwined in affecting cancer prognosis and quality of life. We will interpret these outcomes on a continuum based on their importance to providers and patients, their statistical significance, and the clinical magnitude of the effects (e.g., insignificant, moderate, large). The primary outcomes will be analyzed as follows. We will test for differences in the pCR and organ preservation rate using unadjusted generalized linear mixed models with main effects and an interaction term for treatment and cancer type/subtype, accounting for trial site as a random effect. Clinical and tumor responses will be analyzed using similar methods. For self-reported adverse events, we will analyze PRO–CTCAE items at specific time points as the response variable of a multivariable linear mixed model using group, time, and the interaction between group and time as covariates. Random effects will be used to describe the correlation within repeated measures and sites. We will include covariates such as baseline PRO–CTCAE values, sex, age, race, cancer type/subtype, stage, the timing of baseline data collection (before versus after the start of treatment), and treatment-related parameters, as statistically merited. We will also test the interaction of group with sex, age, cancer type/subtype, stage, the timing of baseline data collection, treatment-related parameters, and other clinically relevant variables. Finally, we will also perform exploratory subgroup analyses by cancer type and other relevant clinical variables.

As a secondary measure of toxicity, we will assess provider-reported adverse events extracted from the EMR. We will analyze the frequencies of all, grade 1–2, and grade 3–4 adverse events between groups using negative binomial regression. We will also calculate the toxicity index using methods proposed in Rogatko et al. [131]. Depending on the distribution, we will either fit a multivariable probabilistic index model [159], which is a generalization of the Mann–Whitney U test [160], where the toxicity index is the response variable, with potential adjustment by sex, age, race, cancer type/subtype, stage, and/or treatment-related parameters, following Gresham et al. [132], or use another distribution-appropriate model. Lastly, we will analyze subcategories of treatment-emergent adverse events by organ or system (e.g., gastrointestinal, genitourinary), if merited.

For all other secondary and tertiary outcomes, we will use multivariable linear mixed models. In particular, these models will be fitted using group, time, and the interaction between group and time as fixed effects. As appropriate, random effects will be added to consider the correlation from repeated measures and different sites. We will adjust for any relevant and impactful covariates, such as baseline values, sex, age, race, cancer type/subtype, stage, and treatment-related parameters as fixed effects. Also, we will perform exploratory subgroup analyses by cancer type and other relevant clinical variables, since TRE may different effects in these subpopulations, as the mechanisms may sufficiently differ.

We will also perform mediation and moderation analyses. As part of testing the DSR theory, we will perform mediation analyses to test the hypothesis that IGF-1 mediates the changes in molecular endpoints. Formal mediation analyses will test whether the relationship between the intervention and each molecular endpoint is reduced by a statistically significant amount when IGF-1 (the potential mediator) is included in the main regression model. If evidence suggests that the causal chain linking the intervention to the outcomes works via a change in IGF-1, we will conclude that the effects of the intervention are mediated by IGF-1. Covariates will be considered in the model to strengthen the validity of sequential ignorability.

We will use moderation analyses to test adherence as a moderator of the relationship between intervention arms and clinical outcomes. For moderation analyses, we will add interaction terms to the models described above. If the interaction effects are statistically significant, then we will interpret the data using the “pick-a-point” approach [161]. This approach involves selecting representative high and low values (mean ± 1 SD) of the moderator variable and then estimating the effect of the focal predictor at those values.

Analyses of primary and secondary outcomes will be done without accounting for multiple testing. For families with larger numbers of outcomes and where at least one false positive is expected, we will also report results using the Benjamini–Hochberg method with a family-wise false discovery rate of 10%, where a family of tests will comprise all tests on related outcomes.

6. Protocol changes and amendements

The study protocol was initially approved by IRB on April 8, 2021. In January 2022, prior to enrolling the first participant, we made the following changes in categorizing the study outcomes: (1) moved provider-reported AEs, quality of life, and clinical response from being primary to secondary outcomes, (2) moved molecular outcomes from secondary outcomes to tertiary outcomes, and (3) added additional clinical outcomes (e.g., tumor response, NAR score) as secondary outcomes. These changes were approved by the IRB on February 1, 2022. In addition to revising eligibility criteria to include patients with breast cancer (October 2023), we further expanded eligibility to stage IV rectal cancer with curative intent (January 2023), and to patients enrolled in other trials but assigned to a control arm without exposure to investigational treatment or device (September 2025).

7. Discussion

The CHRONO trial tests the effects of eTRE on clinical, molecular, and behavioral outcomes among patients with stage I-III HER2-positive or triple-negative breast cancer or stage II-IV rectal cancer undergoing neoadjuvant therapy. Roughly 40% to 65% of these patients are expected to achieve a pathologic complete response or organ preservation.

We hypothesize that eTRE will improve pCR and organ preservation rate as primary outcomes. Should our findings indicate no statistically or clinically significant change in these outcomes, we will consider our hypotheses disconfirmed. However, we believe that a strict confirmatory trial may be premature at this stage, considering the paucity of data on the efficacy of eTRE during active cancer treatment. Accordingly, in determining the success of the trial, we take a patient- and provider-oriented approach, such that any clinical improvement in the patient’s health would constitute a success. The degree of success depends on the specific health outcome and the magnitude of the effect. For instance, if our data indicate that eTRE reduces treatment-related toxicities, with no detectable change in pCR and organ preservation rate (or vice versa), we would still deem the intervention promising in addressing key outcomes important to patients and providers, with the potential to improve oncology care, patients’ quality of life, and/or long-term prognosis.

This work is clinically significant and innovative in several ways. First, while there is preliminary evidence that short-term, water-only fasting reduces treatment-related toxicities, improves blood cell counts, and decreases DNA damage in healthy cells [38–40], acute fasting is generally challenging for patients with cancer, and only a small fraction of cancer patients are willing to do water-only fasting for 48–72 hours. While other intermittent fasting approaches have been proposed—such as the FMD—they typically involve severe calorie restriction (~800 kcal/day), and adherence has been poorer than expected in cancer trials. By comparison, TRE is ostensibly more feasible, as it requires much shorter bouts of fasting and does not involve energy or macronutrient restriction. We expect adherence to TRE to surpass that of the FMD (i.e., >50%). Since adherence to TRE typically averages 80% (5.6 days/week) in noncancer populations, we predict that the mean adherence in our population of cancer patients will fall somewhere between 50% and 80% (~3.5–5.6 days/week). If successful, eTRE may provide an easier-to-follow alternative to both water-only fasting and substantial energy restriction. eTRE’s simplicity is also attractive from a provider perspective, as eTRE can be readily adopted and implemented by oncology teams, with minimal training and no certification required.

Second, testing the effects of intermittent fasting during neoadjuvant therapy allows us to collect tumor specimens before and after several months of treatment. This provides us with a rare, nearly unparalleled opportunity to investigate the molecular effects of a dietary intervention lasting several months on tumor cells. We will test a handful of key molecular players indicative of cell growth and death, which will not only allow us to test the DSR theory but may also provide novel mechanistic insights into tumor biology and metabolism.

Third, another strength of this study is its inclusion of patients with multiple tumor types and molecular subtypes. This will increase the generalizability of our findings. It will also allow us to collect preliminary data on the impact of intermittent fasting on different tumor types and the interaction between fasting and different treatment regimens. The major drawback, however, is that the more heterogeneous population will increase the variability of responses, making it harder to detect differences between groups. To reduce the impact of heterogeneity on the outcomes, we will adjust for important covariates in the analyses, as detailed in the statistics section. Another possible issue is that the underlying pathogenic mechanisms behind rectal versus breast cancer are sufficiently different, and that, as a result, TRE will be more effective in one population than the other. Although there is not much we can do about this—aside from adjusting for cancer type as covariate—we will conduct subgroup analyses to explore whether the effects differ by cancer type, which will provide valuable preliminary information for the field.

Fourth, effects of behavioral and dietary interventions, including intermittent fasting, are often heterogeneous and influenced by a variety of moderators. For instance, social commitments and roles may affect adherence, which, in turn, may moderate the effect of intermittent fasting on clinical outcomes. This study considers often neglected lifestyle and socio-structural factors that may affect the relationships between the interventions and clinical responses. The multi-site nature of the trial, being conducted in California and Alabama, further makes it possible to explore whether the acceptability of meal timing interventions varies by geographic region, socioeconomic status, race/ethnicity, or other factors. These considerations are important for understanding how eTRE and the standard of oncology care interact with cultural, emotional, psychosocial, and lifestyle factors to determine cancer outcomes, as well as how social determinants of health influence the willingness to enroll, adherence, and retention.

Despite these many strengths, there are also potential challenges. The intervention duration will vary across participants more than usual in dietary trials, due to the heterogeneity of cancer types, treatment regimens, and patients’ unique medical needs and circumstances. While we will attempt to control the heterogeneity by stratifying randomization on several factors and by including relevant covariates in the analyses, this variability may be reflected in our findings. Second, since we are enrolling individuals shortly after diagnosis, the shock of diagnosis often leaves individuals overwhelmed, anxious, and afraid as they grapple with the implications of their cancer for their loved ones and themselves. Many patients understandably feel too overwhelmed to consider participating in a clinical trial until they have regained some sense of empowerment. Thus, recruitment may be a challenge. Indeed, we have already revised our initial enrollment target from 300 patients to 175 patients. Third, participants may not be able to isolate how their assigned eating schedule affects their family and social life, sleep, and overall energy levels independent of their treatment, since the intervention and their treatment are concurrent. While we hypothesize that eTRE will improve clinical outcomes and reduce toxicities, it is possible that the social, emotional, and professional disruptions associated with having cancer and receiving cancer treatment may overshadow the potential benefits of intermittent fasting on quality of life, fatigue, pain, mood, and psychosocial outcomes. Fourth, since we can collect resected tumor tissue only in individuals who do not have a complete response, selection bias may affect most of our molecular endpoints, and we may underestimate the effects of intermittent fasting on molecular endpoints. Fifth, we purposefully selected a patient population where cachexia is uncommon, and therefore, TRE may not be appropriate for cancer populations where cachexia is more common, as TRE can induce mild energy restriction and could exacerbate cachexia. Finally, we acknowledge that although we are well-powered to detect differences in the PRO–CTCAE scores and quality of life, we are powered to detect only larger increases in the pCR and organ preservation rate. We will not have sufficient power to detect clinically meaningful improvements of 10%-15% in pCR and organ preservation rates with our sample size. That said, a FMD study did report statistically significant differences in the objective response rate, a measure of clinical response [63], despite being underpowered to detect differences in the pCR rate.

In summary, promising preclinical data indicate that fasting and intermittent fasting enhance the effectiveness of cancer treatment, while also reducing treatment-induced adverse events and improving long-term survival. A few preliminary clinical trials in humans have also reported some positive findings. This innovative trial will build on these promising findings to determine whether intermittent fasting interventions, such as eTRE, can improve cancer outcomes. If successful, this trial has the potential to impact clinical cancer care and improve patients’ quality of life, as well as open new lines of research investigating dietary interventions to improve cancer-related outcomes.

Abbreviations:

BP

blood pressure

CBCs

complete blood counts

EMR

electronic medical record

PBMCs

peripheral blood mononuclear cells

pCR

pathologic complete response