Combining curcumin (C3-complex, Sabinsa) with standard care FOLFOX chemotherapy in patients with inoperable colorectal cancer (CUFOX): study protocol for a randomised control trial

Colorectal cancer (CRC) is the fourth most common cancer worldwide. It carries the second highest cancer-related mortality in Western countries with a lifetime incidence of ~1 in 15 [1,2]. Up to a fifth of CRC patients will present with metastatic disease, typically in the liver [3], and half of those undergoing resection for primary disease will develop metastases. The majority of these patients are not amenable to curative surgery [4] and experience a 5-year survival rate of less than 10%. The median survival for patients with un-resectable colorectal liver metastases in the absence of chemotherapy is approximately 8 months [5], compared with up to 24 months with chemotherapy in patients with KRAS (Kirsten rat sarcoma homologue) wild-type positive cancers [6,7]. Unfortunately, approximately 40% of such cancers display chemo-resistance from the outset and eventually all will fail to respond.

The mainstay of chemotherapy for CRC in most countries has been oxaliplatin-based, commonly with 5-fluorouracil (5-FU) and folinic acid, collectively known as FOLFOX. In the UK in 2002, the National Institute of Clinical Excellence first issued guidelines [8] for oxaliplatin to be combined with 5-FU/folinic acid and administered as first-line therapy. Compared with many cytotoxic regimens, FOLFOX is relatively well tolerated but dose-limiting side-effects are common (reviewed in National Institute of Clinical Excellence CG131 [8,9]). Diarrhoea, due to mucositis, can occur in up to a third of people receiving 5-FU [10], which can be life threatening. Patients may develop neutropaenia, increasing susceptibility to infection. The majority of patients (90%) will develop oxaliplatin-induced peripheral sensory neuropathy to some extent [11] and, after nine cycles, up to a third of patients will require dose reduction or cessation [12].

Strategies for overcoming chemo-resistance and/or augmenting the action of therapy are being investigated with the aim of extending the time that chemotherapy can be tolerated whilst maintaining efficacy and postponing resistance without increased drug toxicities. Naturally occurring dietary agents are attractive in this setting because they offer a potentially favourable side-effect profile with good patient tolerability. Interest in dietary agents has continued as evidence of anti-carcinogenic activity increases (reviewed in [13]). Curcumin is a derivative of the spice turmeric (Curcuma longa), typical to Asian cuisine. The use of turmeric in the field of medicine was described in Asia thousands of years ago (reviewed in [14]). Of note, Asian populations experience approximately an eighth of the incidence of bowel cancer than that of Western populations (30.8 versus 3.9 cases per 100,000 in the UK and India, respectively) [2] and diet is likely to be an important underlying factor.

Curcumin can interact beneficially with a wide variety of pathophysiological processes including maintenance of the cell cycle, carcinogenesis, wound healing and inflammation (reviewed in [15]). The effects of curcumin in combination with chemotherapy have been explored in a variety of cancer cell lines and in vivo models. There is a growing body of pre-clinical evidence repeatedly reporting reductions in tumour volume and metastatic features when curcumin is combined with chemotherapy, as compared to either agent alone [16]. We have previously reported that curcumin augments the effect of oxaliplatin against CRC cell lines, and restores efficacy in a cellular model of chemo-resistance [17]. Furthermore, these findings were reproduced using a xenograft model, where a reduction in tumour size of 53% was observed following treatment with curcumin and oxaliplatin compared to a 16% reduction following treatment with oxaliplatin alone. These pre-clinical data are compelling and have provided the rationale for clinical investigation. Presently, two trials have provided early evidence of efficacy for curcumin in patients with inflammatory bowel disease. Efficacy of curcumin in the cancer setting has not yet been effectively addressed. Clinical reports of curcumin in combination with any chemotherapy regimen are confined to three small trials (two in pancreatic cancer [18,19] and one in metastatic breast cancer patients [20]) plus a case report [21]. These were small studies, not powered for efficacy, and lacked a control arm, but have provided evidence that patients with advanced cancer receiving chemotherapy can tolerate daily curcumin for several months alongside their chemotherapy.

The main aims of this study are to: establish a safe and suitable dose of curcumin to administer to patients receiving up to 6 months of oxaliplatin-based chemotherapy; collect patient samples for biomarker and pharmacokinetic analysis; record patient-reported outcomes, including quality of life and peripheral neuropathy scores; observe progression-free survival (PFS) and overall survival (OS).

CUFOX (curcumin plus FOLFOX) is a Phase I dose-escalation study rolling in to a Phase IIa randomised controlled trial combining daily oral curcumin (C3-complex; Sabinsa Corp., UT, USA) with standard care FOLFOX-based chemotherapy in patients with histological diagnosis of metastatic CRC and disease measurable by response evaluation criteria for solid tumours (RECIST) 1.1, attending oncology clinics prior to chemotherapy. The study will not incorporate a placebo-controlled arm due to additional expense of sourcing an appropriate placebo. Identification of a suitable placebo is also complicated by the issue that curcumin (E100) is highly coloured and widely used as a food colouring. For Phase IIa, participants will therefore be randomised to FOLFOX + curcumin or FOLFOX only at a ratio of 2:1.

This study will recruit adults referred to the University Hospitals of Leicester NHS Trust Oncology Department. The study is sponsored by the University of Leicester and will be undertaken at the University Hospitals of Leicester NHS trust UK, as a single centre study.

Inclusion criteria are: histological or cytological diagnosis of metastatic CRC; inoperable colorectal liver metastases; disease measurable by RECIST 1.1; adequate haematological, hepatic and renal function; age ≥18 years; Eastern Cooperative Oncology Group performance status 0 or 1; recovered from effects of any recent major surgery; post-menopausal or willing to use contraception if applicable; informed consent; life expectancy estimated to be more than 12 weeks.

The exclusion criteria include: contraindications to FOLFOX chemotherapy; peripheral neuropathy (National Cancer Institute Common Terminology Criteria >1); liver failure, uncontrolled coronary heart disease, myocardial infarction within the previous 6 months; unwilling or unable to comply with the study protocol; pregnant or lactating or contemplating pregnancy; undergone chemotherapy (other than adjuvant for CRC) or participating in another drug study; previous cancer <5 years (other than colorectal, basal cell carcinoma, in-situ cervical cancer); major surgery within 4 weeks of starting the study; potentially operable liver disease without prior chemotherapy; bone marrow depletion; co-existing active infection or serious concurrent medical condition; significant cardiovascular disease; severe valvular heart disease; congestive heart failure requiring therapy (New York Heart Association grade ≥2); bone metastases; brain or leptomeningeal metastases; surgery or hospital admissions for symptomatic intra-abdominal adhesions; colitis.

Curcumin can produce diarrhoea in up to a third of patients. This is often mild and self-limiting. However, up to a quarter of patients can develop grade 3/4 diarrhoea due to 5-FU. Persisting grade 2 diarrhoea as a result of curcumin prior to receiving chemotherapy will require cessation or reduction of curcumin. If diarrhoea occurs after chemotherapy commences, 5-FU modification will be in keeping with local guidelines. Modification of curcumin dose will be as per the guidance described in Table 1.

Adherence will be monitored via means of a daily capsule and symptom diary over the first 4 weeks, combined with monitoring of capsule return over the study duration.

Participants in Phase I, or those randomised to the control group, will receive standard care chemotherapeutic intervention. Systemic FOLFOX-based chemotherapy will be administered every 2 weeks for up to 12 cycles via central venous catheter (body surface area capped at 2 m²). The scheduling will be as follows: Hour 0: oxaliplatin 85 mg/m² in 250 mL glucose 5%, 2-hour intravenous (IV) infusion; folinic acid 350 mg in 250 mL glucose 5%, 2-hour IV infusion. Hour +2: 5-FU bolus 400 mg/m², IV bolus; 5-FU continuous infusion 2,400 mg/m², 46-hour continuous IV infusion. Chemotherapy will be delayed (or reduced) until all of the following conditions are met: neutrophils >1 × 10⁹/L, platelets >75 × 10⁹/L, full recovery from stomatitis, National Cancer Institute Common Terminology Criteria for Adverse Events grade ≤2 diarrhoea, skin or other toxicity. Modifications to FOLFOX will be in line with local procedures, and will be modified or discontinued for one of the following reasons:

Any other treatment considered necessary for the patients’ safety and well-being may be given at the discretion of the investigators, and will be carried out in accordance with local practice.

The timeline of proposed interventions for this study can be found in Table 2.

This is the first study combining FOLFOX-based chemotherapy with curcumin, and so consequently there are no published data on which to power this study. This study will provide the first efficacy data for this regimen that can be used in the powering of future studies. The sample size for the Phase I dose escalation is pre-determined by the 3 + 3 + 3 design. Phase IIa will provide further information relating to long-term safety and tolerability of this regimen, and efficacy is a secondary outcome measure. Phase I will recruit between 9 and 18 participants, dependent on the observed rate of dose-limiting toxicity. Phase II will recruit 33 participants, 22 of whom will receive FOLFOX-based chemotherapy in combination with daily oral curcumin, and 11 of whom will be randomised to the control arm receiving standard care chemotherapy.

Potential participants will be identified via the weekly cancer multidisciplinary team meetings and referrals to oncology clinic. Participants who have been informed of their diagnosis and are being offered chemotherapy will be given or posted a patient information sheet (PIS). During their attendance at the oncology department they will be introduced to a trial investigator usually by the oncologist arranging the chemotherapy. The investigator will discuss the trial and PIS and answer any questions that potential participants may have. They will be given a minimum of 24 hours to consider the PIS before being invited to participate in the study. Patients can be seen at their second pre-chemotherapy clinic appointment to obtain written informed consent. Upon obtaining written informed consent from each participant, baseline evaluations can be undertaken. Baseline evaluations will include a complete physical examination (body weight, height; pulse, and assessment of Eastern Cooperative Oncology Group performance status). Body surface area must also be determined for FOLFOX dosing. Required haematological and clinical chemistry measurements will include haemoglobin, white blood cell count with differential if abnormal, platelet count, urea, creatinine, sodium, potassium, alkaline phosphatase, alanine aminotransferase, albumin, and bilirubin. Chorioembryonic antigen will be measured in all patients at baseline and, if elevated, every 12 weeks thereafter.

The sample size for Phase 1 is determined by the maximum dose desired (2 g/day based on previous trials within the local populace) and the number of tiers selected to step up to this dose (three tiers; 500 mg/day, 1 g/day and 2 g/day). In the absence of dose-limiting toxicity (DLT), only three patients per tier will be required (n = 9); however, the “worse-case” scenario is taken into consideration, allowing resources to repeat each tier once if necessary (n = 18).

Three patients meeting the inclusion criteria will receive one oral capsule (500 mg) per day of curcumin (Figure 1), commencing 7 days prior to chemotherapy to allow acclimatisation and any curcumin-induced adverse events (AEs) to manifest prior to chemotherapy. In the absence of curcumin-related toxicity after 1 week, patients will commence standard care FOLFOX-based chemotherapy (modified to capecitabine + oxaliplatin, or FOLFOX + bevacizumab where appropriate). After a 4-week enhanced early monitoring phase, and in the absence of curcumin-associated dose-limiting toxicity (DLT) in three consecutive patients, a further three participants will be recruited at 1 g/day oral curcumin (two capsules), and then to the maximum target dose of 2 g/day (four capsules). In the event of a DLT attributable to curcumin, the participant will cease involvement in the trial and a further three participants will be recruited at that dose level. If a participant withdraws for a reason unrelated to curcumin, only that participant will be replaced. Independent data monitoring will be convened prior to each dose escalation and prior to commencement of Phase II to ensure patient safety. De-escalation will be mandatory if more than two DLTs due to curcumin among six participants are observed within 4 weeks of commencing curcumin.

Phase II will be an open-labelled, two-armed, randomised controlled trial/feasibility study, consisting of a FOLFOX only (control) group (11 participants) and a FOLFOX + curcumin group (22 participants). Patients will be managed identically to those recruited to the traditional escalation response model. Primary endpoints will remain focussed on drug safety, tolerance and the feasibility of undertaking a larger scale study. A group receiving chemotherapy alone is to be included principally to provide control samples for the proposed biomarker studies, but also to generate data informing on quality of life, disease response, PFS, OS and neuropathic side-effects.

Consecutive unique identifying numbers (1 to 33) will be assigned a group via an electronic random number generator at a ratio of 2:1 (number 1 = combination therapy; number 2 = control group receiving FOLFOX-based therapy only). Groupings will be placed in a sealed envelope bearing the unique identifying number. These will be blinded to the recruiting researchers and patients until informed consent is obtained, whereupon grouping can be revealed. The randomisation list is to be kept separate to the site file and a copy held by the trial pharmacist.

A neuropathic scoring questionnaire [24] will be completed by patients every two cycles throughout protocol chemotherapy, and subjects will also record the amount and timing of curcumin doses alongside any side-effects experienced as part of the early monitoring period. A 28-day diary has been designed to support monitoring during the early monitoring period. The European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire form C30 will be completed at the beginning of the trial and when participants cease protocol chemotherapy. Non-validated pre- and post-trial questionnaires relating to participant’s views on curcumin, its consumption, side-effects and the trial experience will be completed at the start and end of the curcumin course.

Data will be abstracted from the participant’s records onto paper case report forms. Data collection and archiving will be done in accordance with appropriate standard operating procedures. Case report forms are kept in secure storage within a locked room/building and personal data is kept to a minimum. Complete patient identifiable details will be stored on a separate paper recruitment log. Transcription accuracy will be cross-checked on a quarterly basis, with audit trail and responses filed in the Trial Master Site File.

Data will be presented in summary format. This will predominantly take the form of counts or percentages for categorical outcome data with mean/median, range and standard deviations for continuous outcome data and with 95% confidence intervals where appropriate. Additional independent samples t-test or unpaired non-parametric tests will be performed with P < 0.05 taken as significant. Estimated survival times will be calculated from clinical outcome data. Statistical comparison can be made between groups; however, this study is not primarily powered to be a biomarker or efficacy-driven study and the data generated herein is intended for the design of future clinical trials. The statistical software at the University of Leicester is SPSS/PASW (Chicago, IL, USA). Statistical analyses will be supported by the University of Leicester’s Bioinformatics and Biostatistics Hub.

An independent data monitoring committee will be convened to provide safety monitoring prior to dose escalation in Phase I. The committee will subsequently meet on a quarterly basis to review any AEs, SAEs, SUSARs (Suspected Unexpected Serious Adverse Reaction) and other safety issues arising in Phase IIa. This committee may be convened at an earlier date in the event of significant safety concerns relating to the study drug. The study will audited by an external auditor every 6 months, in keeping with Sponsor and Institutional policies.

Ethical approval for this study was granted by the East Midlands (Derby 1) regional ethics committee (11/EM/0263). Clinical trials approval was obtained from the Medicines and Healthcare Products Regulatory Agency. The trial is registered at ClinicalTrials.gov (NCT01490996) and with the European Drug Regulating Authorities (EudraCT 2011-002289-19), and will be conducted in accordance with good clinical practice and the principles of the Declaration of Helsinki.

Written informed consent will be obtained by delegated members of the participant’s healthcare team. Consent will be sought from participants to allow sections of their medical records to be looked at in strict confidence by responsible individuals from the study team, the Sponsor, Research Ethics Committee, NHS Trust or from regulatory authorities where relevant.

Research findings will be reported at local levels on University Hospitals of Leicester NHS Trust and University of Leicester websites. National and international dissemination of results will take place at oncology-based conferences such as those hosted by the National Cancer Research Institute, British Association for Cancer Research, and American Society of Clinical Oncology. Furthermore, results will be disseminated in both lay and scientific summary format via the websites of major funders for this study including Hope against Cancer, The Royal College of Surgeons, the Bowel Disease Research Foundation, the Cancer Research UK Experimental Cancer Medicine Centre Network and the Cancer Research UK Leicester Centre.