Bioactive food components for colorectal cancer prevention and treatment: A good match

Abstract

Colorectal cancer (CRC) is the third most frequent cancer worldwide, accounts for about 10% of the total cancer cases, and ranks as the second cause of death by cancer. CRC is more prevalent in developed countries in close causal relation with occidental diets. Due to anatomy, the diet has a strong impact on CRC. High contents in meat are acknowledged risk factors whereas a diet rich in fruits and vegetables is an established CRC protective factor. Fruits and vegetables contain numerous Bioactive Food Components (BFCs), physiologically active food compounds, beneficial on health. Preventive and therapeutic benefits of BFCs in cancer have increasingly been reported over the past 20 years. BFCs show both chemopreventive and anti-tumor properties in CRC but more interestingly, abundant research describes BFCs as enhancers of conventional cancer treatments. Despite these promising results, their clinical transferability is slowed down by bioavailability interrogations and their poorly understood hormetic effect. In this review, we would like to reposition BFCs as well-fitted for applications in CRC. We provide a synthetic overview of trustworthy BFC applications in CRC, with a special highlight on combinatory approaches and conventional cancer treatment potentiation strategies.

Keywords:

• Bioactive food
• cancer prevention
• combination cancer treatment
• curcumin
• natural compounds
• resveratrol

Introduction

Colorectal cancer (CRC) is the third most frequent cancer worldwide, accounting for about 10% of the total cancer cases and ranks as the second cause of death by cancer. CRC is present especially in developed countries with about 60% of all CRC cases (Parkin et al. 2002). This high incidence was proven associated with occidental highly caloric diets (Ryan-Harshman and Aldoori 2007; Jin and Zhang 2020). Diet-linked CRC risk factors are multiple. Among them, important consumption of meat either red, processed, or cooked at high temperature (fried, broiled, or grilled) can increase the risk of developing CRC by up to 30 times (Imperial College London 2018). Conversely, regular physical activity and diets rich in fruits and vegetables are the main established CRC-specific protective factors (Imperial College London 2018).

Literature reports sufficient evidence to link obesity and higher risk of CRC, especially in men with low physical activity (Johnson and Lund 2007; Sawicki et al. 2021). The pathophysiological bases (Yang et al. 2022) include pro-oncogenic adipocyte metabolism (associated with lipid oxidation and energy storage) (Schwartz and Yehuda-Shnaidman 2014), low and chronic inflammation associated with pro-inflammatory cytokines secretion (Reilly and Saltiel 2017) and, stimulation of colorectal cell proliferation by adipokines and hormones (such as insulin) (Chen et al. 2011; Chen et al. 2018). In this context, unhealthy food consumption is closely related to obesity, ad increases the risk of CRC development.

The relationship between health and food regained intensive attention during the last 20 years to set the basis of numerous studies, first focusing on the cardiovascular impact and neurodegenerative diseases. The Mediterranean diet was early recognized as beneficial for health and nutrition (Altomare et al. 2013). Rich in fruits and vegetables, it prevents chronic cardiovascular diseases and several malignancies (reviewed in (Mentella et al. 2019; Widmer et al. 2015):). In parallel, the concept “French paradox” emerged, based on the observation that French people have a relatively low incidence of coronary heart diseases despite a rich diet, but including a moderate and regular consumption of red wine (Renaud and de Lorgeril 1992). Even if this theory was later refuted (Griswold et al. 2018), it triggered the identification of protective compounds in the wine and vine, and more broadly, in food.

The Bioactive Food Components (BFCs) are "physiologically active food compounds, derived from animal or plant sources, including compounds belonging to the food of a basic diet, for which a role has been shown to be healthy and the consumption of which is not harmful to health," according to the American Dietetic Association (Hasler et al. 2004). Noticeably, some BFCs have been long known in traditional medicines. For instance, Curcuma longa, with high content of curcumin, is widely used in India and China as a natural herbal remedy (reviewed in Hatcher et al. (2008)). Interestingly, most BFCs exhibit hormesis, characterized by opposite effects at low and high doses, because of their anti/pro-oxidative properties.

Preventive and therapeutic benefits of BFCs in cancer have increasingly been reported over the past two decades in particular with regards to CRC (Figure 1). Noticeably, BFCs showed both chemopreventive and anti-tumor properties in CRC (Yin et al. 2016; Ahmed et al. 2019) contributing to the general concept that a healthy way of life, including a balanced diet, is protective of CRC. Moreover, abundant literature describes BFCs as enhancers of conventional treatment efficiency in multiple cancers, suggesting a true biological relevance. Paradoxically, this profusion of information overflows important clues, hormesis adding confusion, which hinders their transfer to clinical applications. As our team gained expertise and improved comprehension of BFCs validity in cancer treatment, we would like to provide a synthetic overview of selected trustworthy applications in CRC deserving attention, with a special highlight on combinatory approaches.

Figure 1. BFCs and cancers. (A) Number of reported publications per year in PubMed using the key words “cancer” and “bioactive food component” from January 2000 to December 2020. (B) Number of total reported publications in PubMed including key words “breast cancer” or “lung cancer” or “colorectal cancer” or “leukemia” or “hepatocarcinoma” or “melanoma” or “pancreatic ductal adenocarcinoma” (PDAC) or “lymphoma” or gastric cancer” or glioblastoma” or bladder cancer” or sarcoma” or “cholangiocarcinoma” and “curcumin” or “quercetin” or “resveratrol” or “capsaicin” or “sulforaphane” from January 2000 to December 2020.

Literature and clinical trial searching process

Systematic literature searches of articles published between 2000 and October 2021 were performed in the electronic PubMed database from the American National Institute of Health and the Google Scholar database. The searches were conducted using medical subject headings(MeSH: “colorectal cancer” or “colon cancer” or “rectal cancer” combined with “resveratrol” or “sulforaphane” or “quercetin” or curcumin” or “capsaicin”) and free-text words and limited to articles published in the English language in indexed and peer-reviewed journals.

A clinical trial search was performed using the clinicaltrials.gov website. To select international studies of interest, “colorectal cancer” was selected as the condition of interest, and either “resveratrol” or “sulforaphane” or “quercetin” or curcumin” or “capsaicin” keywords were used in the “other terms” section.

Main phytochemical BFC categories and compounds

As the “Bioactive Food Components” term brings together a wide range of molecules, originating from both plant and animal sources, it seems vain and counter-productive to provide an exhaustive list. We chose to focus on three major categories of BFCs carrying the most convincing level of evidence for utility in CRC prevention and treatment.

Polyphenols, the most studied bioactive food group in the field of cancerology, are present in most diets, at around 1 gram per day. They are mainly provided by fruits, vegetables, wine, tea, and chocolate (Manach et al. 2004). They contribute to the organoleptic properties of the leaves, fruits, and processed products such as wine or olive oil. Three main classes are studied in oncology: Phenolic acids, Flavonoids, and Stilbenes. Phenolic acids are phenolic compounds with one or more carboxylic acid groups, present for instance in coffee, red fruits, and onions. These include Curcumin, a derived BFC of the spice turmeric, with promising applications in oncology (Giordano and Tommonaro 2019). Flavonoids are found in onions and broccoli and are constituted by two benzene rings linked by a heterocycle. Among them, Quercetin showed recurrent beneficial anti-cancer effects (Rather and Bhagat 2020). Stilbenes are chemically close to flavonoids. They include resveratrol and its derived such as hydroxylated derivatives (piceatannol), dimerized derivatives (ε-viniferin and δ-viniferin), and other oligomers (vitisin B) for example. They are present in some fruits like blackberries, peanuts, and more particularly in grapes and their derivative products such as red wine. The properties of resveratrol have been widely studied in cancer demonstrating both chemopreventive and anti-cancer activity (Vervandier-Fasseur and Latruffe 2019; Berretta et al. 2020).

Heterosides are glycosylated derivatives of terpenes, phenols, and, more rarely, alkaloids. Glucosinolates, present in cruciferous vegetables like cabbage and broccoli, are hydrolyzed into bioactive isothiocyanates, including sulforaphane, with antioxidant properties potentially relevant for the prevention of neurodegenerative diseases (Tarozzi et al. 2013) and cancers (Grabacka, Gawin, and Pierzchalska 2014).

Alkaloids are basic nitrogen compounds extracted from plants. They provide medicine like morphine or the vincristine and taxol chemotherapies. They also count BFCs such as quinine and caffeine. Capsaicin, naturally occurring in chili peppers, represents the most current edible alkaloid of interest in cancerology applications.

BFCs in colorectal cancer prevention

Prevention of CRC has important public health implications and the influence of the diet has now been documented for many years. Several risk factors have been highlighted such as important consumption of meat or chronic alcohol consumption (Imperial College London 2018). The lower CRC frequency in Mediterranean countries positioned the Mediterranean diet as chemopreventive (Reviewed in Farinetti et al. (2017)). Particularly rich in fruits and vegetables, this diet naturally contains many BFCs, especially polyphenols such as resveratrol and quercetin (Medina-Remón et al. 2017; Bagetta et al. 2020), conferring strong chemoprevention of CRC onset (Imperial College London 2018; Couto et al. 2011).

Anti-inflammatory properties

A wide range of BFCs is known to exert anti-inflammatory properties in many diseases and in particular cancer (Kim et al. 2009). Their activities take place at different levels of the inflammation process with for instance up-regulation of anti-inflammatory cytokines (Nguyen et al. 2006; Song et al. 2014) or downregulation of pro-inflammatory interleukin and enzyme synthesis, such as TNF-alpha and cyclooxygenase II (Manna, Mukhopadhyay, and Aggarwal 1950; Xagorari et al. 2001; Tong et al. 2005). Thus, BFCs can participate to maintain the balance between pro and anti-inflammatory mediators. They exhibit chemopreventive properties and could oppose CRC development by limiting chronic inflammation of intestine tissues (Arya, Kanthlal, and Linda 2020; Jantan et al. 2021). For instance, resveratrol and other dietary polyphenols downregulated inflammation-related enzymes and cytokines (Yin et al. 2016; Liang et al. 2001). The Association of curcumin and sulforaphane also exhibited a synergistic anti-inflammatory effect by induction of phase II antioxidant enzymes and down-regulation of pro-inflammatory cytokines like Tumor Necrosis Factor (TNF), nitric oxide (NO), or Prostaglandin E₂ (PGE2) (Cheung, Khor, and Kong 2009). In another study, colorectal inflammation levels in mice under a classic diet were compared to that of a diet enriched in 4 BFCs including curcumin. The BFC-enriched diet group showed significantly reduced number and size of solid lesions, decreased histological inflammation scores, lowered pro-inflammatory cytokine mRNA expression, decreased number of low-grade dysplasia and high-grade dysplasia areas, as compared to the classic diet group (Girardi et al. 2018), suggesting a systemic effect of the BFC-enriched diet. In a colitis mouse model, oral administration of microencapsulated quercetin, preventing its absorption by the small intestine, led to decreased intestinal neutrophil recruitment, as well as diminution of histological alterations and macroscopical damages (Guazelli et al. 2013). Reduction of Interleukin 33 (IL-33) and IL-10 production also contributed to the anti-inflammatory effect. Interestingly, these results were not observed with free quercetin. Thus, quercetin absorption (and potent metabolization) by enterocytes seemed to impede its anti-inflammatory properties in the colon. Thus, BFCs seem to be able to reverse the inflammatory imbalance of digestive tissues and contribute to the prevention of CRC development.

Genetic stability

In non-cancer cells, resveratrol and other related polyphenols contributed to stabilizing histones H2AX and reduced replication stress-associated DNA double-strand breaks (DSB), resulting in increased genome stability of mouse embryonic fibroblasts and the prevention of gene mutations in the p53 pathway (Matsuno et al. 2020). In the same way, curcumin prevented telomere shortening and limited micronucleus formation in a transgenic model of Alzheimer’s disease (Thomas et al. 2009).

In CRC, mutant-APC mice fed with a curcumin-containing high-fat diet (HFD) displayed a significant reversion of HFD-induced polyps (Pettan-Brewer et al. 2011). Capsaicin was administrated to Wistar rats before and after treatment with 1,2-dimethylhydrazine (DMH), a DNA methylating agent known for its carcinogenesis properties (Caetano et al. 2018). The number and frequency of colorectal preneoplastic lesions were reduced in the capsaicin-treated group, suggesting that DMH genotoxicity was attenuated by capsaicin. In the same way, quercetin prevented DMH-induced colorectal inflammation and tumor development through the modulation of Wnt and nuclear factor erythroid-2-related factor 2 (NRF2) signaling pathways (Darband et al. 2020; Shree, Islam, and Sultana 2021). The chemopreventive role of a diet containing curcumin was also demonstrated in the healthy intestine. Mice receiving curcumin displayed fewer preneoplastic events, with reduced DNA damage in differentiated cells, possibly because DNA-damaged-Lgr5⁺ intestinal stem cells were more prone to apoptosis (Kim et al. 2019).

BFCs in colorectal cancer treatment

BFCs present direct antitumor properties on CRC, closely related to compound concentration and involving multiple cellular mechanisms. While low concentrations of BFCs seem to contribute to tumor prevention through their anti-oxidative properties, direct anti-tumor effects call for high compound concentrations and oxidative stress.

Upregulation of reactive oxygen species

Reactive oxygen species (ROS) are major determinants of the cell redox homeostasis and BFCs are potent modulators of ROS generation. ROS upregulation was commonly responsible for the colorectal antitumoral properties of BFCs. For instance, resveratrol and curcumin are pro-apoptotic for colorectal tumor cell lines in vitro in a ROS-dependent manner (Juan et al. 2008; Blanquer-Rosselló et al. 2017; Wang et al. 2017; Sritharan and Sivalingam 2021). Quercetin-induced ROS and derivatives positively modulated endoplasmic reticulum (ER) stress, especially by calcium transfer from the endoplasmic reticulum to mitochondria triggering intrinsic apoptosis (Khan et al. 2016).

Other cell death pathways may contribute, such as autophagy which was enhanced by ROS upregulation in human colon cancer cells treated with resveratrol (Miki et al. 2012). Oxidative stress induced by the flavonoid apigenin resulted in p53-independent senescence through p16, retinoblastoma protein (pRb), and p21 modulation (Banerjee and Mandal 2015).

BFCs contribute to ROS increase by direct regulation of enzymes involved in oxidative stress response. Resveratrol inhibited catalase and superoxide dismutase (SOD), both involved in ROS scavenging (Santandreu et al. 2011). By contrast, quercetin enhanced cyclooxygenase 2 (COX-2) activity, resulting in ROS generation upregulation and cytotoxicity increase on colorectal tumor cell lines (Raja et al. 2017).

DNA damages generation

One of the most deleterious effects of ROS is their ability to induce DNA damages when present at high concentrations in the nucleus. This observation set the basis of studies aiming at evaluating BFC-induced DNA damages. For instance, resveratrol was reported to increase the phosphorylation of histone gamma-H2AX, a chromatin modification mark at double-strand breaks (DSB), needed to initiate DSB DNA repair (San Hipólito-Luengo et al. 2017). In the same way, curcumin markedly increased DNA damage as measured by alkaline comet assay (Lu, Cai, and Ding 2011). When possible, DNA damages are resolved to preserve genomic integrity. BFCs hinder DNA repair, leading to synergistic toxicity by the direct increase of DNA damages and concomitant DNA repair inhibition. Sulforaphane analogs increase phosphorylation of histone gamma-H2AX, inhibit homologous recombination (HR), and non-homologous end joining (NHEJ) repair activities, notably by affecting the HAT/HDAC balance and histone acetylation status (Okonkwo et al. 2018). This cytotoxic synergy was also reported in other digestive cancers (Vendrely et al. 2019).

Cell cycle arrest

Because of unresolved DNA damage accumulation and inhibition of DNA repair pathways, BFCs also affect the cell cycle. The majority of studies found a typical accumulation of cells in the G2/M phase (Adams et al. 2005; Cheah et al. 2018; Wu, Yi, et al. 2019; Jozkowiak et al. 2020). The p53 pathway activation in stress conditions played a major role in cell cycle arrest. Resveratrol upregulated the Su(var)3-9, Enhancer-of-zeste and Trithorax (SET) domain-containing lysine methyltransferase 7/9 (SET7/9) in colorectal cell lines, which stabilized p53 by mono-methylation at lysine 372 (Liu et al. 2019). A resveratrol analog up-regulated p53 and p21 with subsequent cell cycle arrest and apoptosis (Cheah et al. 2018), as did capsaicin (Jin et al. 2014). Interestingly, looking deeper into the TP53/ATM/ATR DNA damage response pathway, curcumin-induced G2/M cell cycle arrest was not reversed by caffeine, an inhibitor of the ATM/ATR proteins, suggesting that additional pathways may be modulated (Lu, Cai, and Ding 2011).

Invasion and metastasis inhibition

Inhibition of tumor cell dissemination takes a large part of the BFC literature. For instance, resveratrol strongly impacted the epithelial to mesenchymal transition (EMT) by interfering with AKT, Wnt, TNF-β, or TGF-β signaling pathways (Ji et al. 2013; Ji et al. 2015; Yuan et al. 2019; Buhrmann et al. 2019). Metastasis inhibition by resveratrol involved focal adhesion protein suppression (Buhrmann et al. 2017) and overexpression of the translation inhibitor RNA binding protein tristetraprolin (TTP) (Lee et al. 2018). Oxyresveratrol, a natural hydroxyl derivate of resveratrol, modulated EMT-related miRNAs, resulting in EMT inhibition and cell migration reduction (Lin et al. 2021). More recently, up-regulation of miR-200c and associated target EPM5, a DNA binding histone-like protein, was the effector of curcumin-related EMT downregulation (Wang, Cai, et al. 2020). Curcumin downregulated the transcription of oncogenic microRNA miR-21 to inhibit invasion and metastasis of HCT116 cells in a chicken-embryo-metastasis assay (Mudduluru et al. 2011). Likewise, the inhibition of the TGF-β pathway by quercetin compromised EMT resulting in invasion/metastasis inhibition (Feng et al. 2018). The antimetastatic effect of curcumin, quercetin and their derivatives were described in several studies (Kee et al. 2016; Li et al. 2018; Calibasi-Kocal et al. 2019).

The interest of combination for BFCs

Unlike the majority of the studies, only a few reported effects of molecules used in combination. This topic was reviewed recently for chemoprevention (Rizeq et al. 2020). BFCs share common antitumor effectors leading to ROS induction or apoptosis but the upstream sequences of events seemed compound-dependent. Thus, combining different BFCs may carry additive or synergistic activities. In this way, combining curcumin with different BFCs enhanced global antitumor efficiency in different CRC models (Ravindranathan et al. 2018; Wu, Koh, et al. 2019). With resveratrol curcumin induced a greater impairment of epithelial growth factor receptor (EGFR) constitutive activation (Majumdar et al. 2009) and upregulation of several genes involved in apoptosis (Gavrilas et al. 2019). Similarly, four BFCs (Lycopene, Sulforaphane, Quercetin, and Curcumin) displayed additive proliferation inhibition while no toxic effect was observed for normal colon epithelial cells (Langner, Lemieszek, and Rzeski 2019). In a clinical setting, the combination treatment of quercetin and curcumin significantly decreased polyp number and size after 6 months of treatment of adenomatous polyposis coli (APC) patients (Cruz-Correa et al. 2006), with minimal adverse side effects and no routine biological parameters abnormalities.

BFCs as adjuvant agents for potentiation of conventional CRC treatments

BFCs may enhance the efficiency of conventional therapies, in numerous cancers (Lin et al. 2020). Combinations of BFCs were tested in CRC with chemotherapy, radiotherapy, or targeted therapies.

Chemotherapy

Colorectal cancer treatment includes surgery and chemotherapy (Recio-Boiles and Cancer 2020). Indications depend on the location, stage, histology, and patient’s general condition. The common chemotherapies include 5FU, irinotecan, oxaliplatin, doxorubicin, and capecitabin, often combined for better efficiency. BFCs and chemotherapies target common tumor cell key functions (Figure 2). Improvement of chemotherapy antitumor activity by BFCs would greatly benefit patient management, especially if they do not carry systemic toxicity. This is particularly true for fragile patients for whom this strategy may allow to maintain disease control while reducing chemotherapy doses.

Figure 2. Bioactive food components (BFCs) as adjuvant agents for potentiation of conventional colorectal cancer (CRC) treatments. BFC and standard CRC treatments (chemotherapy/radiotherapy) target common tumor cell key functions. In particular, BFCs, as well as chemo and radiotherapy, induce oxidative stress associated with reactive oxygen species generation and DNA damages. BFCs are also able to modulate various signaling pathways resulting in, for instance, cellular cycle arrest, apoptosis induction and epithelial to mesenchymal transition (EMT) downregulation. Moreover, modulation of membrane transport proteins by BFCs can lead to reduce tumor drug resistance to standard chemotherapy treatments.

Several signaling pathways were regulated by BFCs with 5FU (Du et al. 2006; Marjaneh et al. 2018; Buhrmann et al. 2018; Chung et al. 2018; Zheng et al. 2021). For example, quercetin enhanced 5FU-induced apoptosis by overexpression of p53 in a CRC microsatellite instability (MSI) positive cell line. This was further supported by the fact the apoptosis enhancement disappeared in p53 deficient/knockdown CRC cell lines (Xavier et al. 2011). Up-regulation of cell-cell junction protein expression by resveratrol together with down-regulation of the NF-κB pathway led to inhibition of EMT and chemosensitization to 5FU (Buhrmann et al. 2015). Ginkgetin, a flavone BFC mainly isolated from Ginkgo biloba, and resveratrol suppressed VEGF-induced endothelial cell proliferation, migration, and invasion (Hu et al. 2019). They exerted a synergistic anti-tumor activity with 5FU by decreasing microvessel density inside the tumors. Concomitantly, combination treatment synergistically relieved the 5FU-induced inflammatory response by suppressing the expression of COX-2 and inflammatory cytokines. Quercetin and ellagic acid decreased human CRC cell proliferation, involving DNA damage and cell cycle modulation, to increase 5FU cell sensitivity. Moreover, tumors harboring BRAF mutations may be more responsive to vine pruning residues (VPE) than KRAS mutated tumors, suggesting that BFCs’ properties depend on CRC major oncogenic pathways (Jesus et al. 2020).

CRC oxaliplatin chemosensitization was reported mostly for curcumin and its derivates. Recently, oxaliplatin and curcumin co-treatment not only increased chemotherapy efficiency but also abolished oxaliplatin resistance (Patel et al. 2008; Chen et al. 2017; Han et al. 2020). Sulforaphane sensitized colon cancer cells to oxaliplatin-induced cell growth inhibition by both intrinsic and extrinsic apoptosis and programmed necrosis (Kaminski et al. 2011). WZ26, a curcumin synthetic analog, combined with cisplatin significantly inhibited tumor growth while attenuating body weight loss observed with cisplatin treatment (Zhang et al. 2019).

As oxaliplatin, doxorubicin is known to reach limited efficacy in CRC due to multi-drug resistance. Resveratrol inhibited the P-glycoprotein efflux pump activity, a transmembrane protein responsible for the multiple drug resistance, leading to accumulation of doxorubicin in tumor cells. More interestingly, MDR1 gene (coding for P-glycoprotein) expression remained stable under resveratrol treatment, suggesting direct ATPase inhibition (Khaleel et al. 2016). Quercetin enhanced the doxorubicin antitumor effect by downregulating the expression of the P-glycoprotein (Zhou et al. 2020). In a C26 colon carcinoma murine model, a PEGylated liposomal curcumin and doxorubicin combination demonstrated enhanced antitumor activity through suppressive effects of angiogenesis, inflammation, invasion, and resistance to apoptosis (Sesarman et al. 2019). Moreover, dysregulation of the Th1/Th2 cell axis in the tumor microenvironment favored the antineoplastic phenotype.

In CRC and other cancer types, cancer stem cells (CSCs) are classically responsible for treatment resistance. Therefore, the impact of BFCs on CSCs was investigated. Quercetin targeted C133⁺ CSCs of the HT-29 CRC cell line, especially by promoting cell cycle arrest and apoptosis (Atashpour et al. 2015). Similarly, curcumin significantly attenuated chemoresistance in an irinotecan-resistant LoVo CRC cell line (Su et al. 2018). A significant reduction of the expression levels of CSC-specific markers, associated with more apoptosis, occurred only in curcumin-treated tumor cells.

As mentioned before, CRC chemotherapy classically combines several molecules for more efficiency. Rather than testing combinations of BFCs and individual chemotherapies, the impact of curcumin was assessed in combination with FOLFOX, the association of 5FU, oxaliplatin, and folinic acid, commonly used for the treatment of CRC. A sequence treatment of FOLFOX followed by curcumin greatly reduced survival of HCT116 and HT-29 CRC cells by particular targeting of FOLFOX-resistant cells through inhibition of Epidermal Growth Factor Receptor (EGFR) and Insulin Growth Factor 1 receptor (IGF-1R) (Patel, Gupta, et al. 2010). At the clinical level, conventional chemotherapeutic regimens associated with curcumin emerges as a potential strategy to prevent chemoresistance of colon cancer cells, as exemplified by the randomized phase IIa clinical study (Howells et al. 2019), with the primary outcome assessing the safety of curcumin for patients receiving either FOLFOX or CUFOX (curcumin + FOLFOX). Similar adverse event profiles were observed in both arms, confirming curcumin safety. Importantly, overall survival (OS) was improved in the CUFOX arm (502 versus 200 days). However, these results deserve careful consideration and external validation because of the small cohort size (28 patients).

Radiotherapy

Although radiotherapy has limited indications for colon cancer, it is almost systematically used in the management of rectal cancer. Neoadjuvant treatments classically associate radiotherapy and chemotherapy to reduce tumor volume before surgical procedure. In addition to chemosensitization, some BFCs exhibit CRC radiosensitization, notably by oxidative stress-associated DNA damage generation (Figure 2). Modulation of CRC radiosensitivity by curcumin was evidenced in 2 mouse models (Kunnumakkara et al. 2008; Yang et al. 2018). Both studies reported enhancement of radiotherapy efficacy when combined with curcumin through NF-kB pathway downregulation, apoptosis, angiogenesis, and DNA repair inhibition. Interestingly, curcumin also suppressed post-irradiation NF-kB activation and associated pro-survival response in human CRC cell lines (Sandur et al. 2009). Radiosensitization by BFCs like quercetin and resveratrol were recently confirmed in several tumor types (Vendrely et al. 2019; Lin et al. 2012; Amini et al. 2020).

Precision medicine

With the late development of precision medicine, the management of CRC becomes tailor-made, according to the tumor and patient specificities. The role of precision medicine in the management of digestive cancers was detailed in a recent publication (Matsuoka and Yashiro 2020). Regorafenib, a tyrosine kinase inhibitor approved by Food and Drug Administration (FDA) for CRC treatment, the antitumoral effect was enhanced by curcumin in the KRAS mutant HCT116 cell line, through autophagy and apoptosis. Curcumin acted as a MEK inhibitor. Interestingly, the KRAS wild-type HT-29 CRC cell was not responsive, suggesting that KRAS status influenced the curcumin potentiating effect (Wu, Wu, et al. 2019). Combination treatment of curcumin and dasatinib reduced cell growth, invasion, and colonosphere formation accompanied by chemosensitization of the cancer stem cell subpopulation (Nautiyal et al. 2011). Similarly, resveratrol sensitized CRC tumor cells to the anti-EGFR cetuximab (Wang, Wang, et al. 2020).

The immunomodulatory properties of BFCs seemed beneficial in combination with checkpoint inhibitors (Deng et al. 2020). For example, combination treatment of curcumin and sildenafil increased anti-PD1 antibody efficiency in a CRC mouse model (Dent et al. 2020). Moreover, BFCs combination reduced tumor expression of PD‐L1 and increased that of Class I human major histocompatibility complex (HMC). Resveratrol enhanced natural killer (NK) lymphocyte antitumor activity (Lee and Kim 2020; Lee, Shin, and Kim 2021), boosting the efficiency of checkpoint inhibitors (Chen and Musa 2021). These results also highlight the potentiality of BFCs as modulators of the tumor microenvironment (Nagaraju 2020).

Clinical trials

Few studies explored BFCs’ benefits in clinical setup for CRC. Results are not available for most of the trials, yet (see Table 1). Briefly, 12 studies tested curcumin (6 Phase I, 5 Phase II, and one Phase III), 4 Phase I studies tested resveratrol and one study tested quercetin. Among all of these 17 studies, 13 included CRC patients, 2 included APC patients, 1 included patients with a moderate risk of CRC, and 1 study included healthy patients.

Table 1. Clinical trials interested in BFCs for chemoprevention and treatment of colorectal cancer.

Clinical trial title	Actual patients enrolment	Patient type	BFC	Administration way and doses	Main outcomes	Reference (ClinicalTrials.gov)
Phase I studies
A Pilot, Feasibility Study of Curcumin in Combination With 5FU for Patients With 5FU-Resistant Metastatic Colon Cancer	13	Patient with metastatic colorectal cancer that is not response to standard 5FU based therapies.	curcumin	Induction by per os curcumin 500 mg twice a day for 2 weeks before 6 weeks of curcumin (same doses) and 5FU (3 cycles)	Determine the safety using curcumin in patients with metastatic colon cancer Overall Response	NCT02724202
Curcumin in Preventing Colorectal Cancer in Patients Undergoing Colorectal Endoscopy or Colorectal Surgery	30	Patients with scheduled colorectal endoscopy or colorectal cancer surgery	curcumin	5 capsules daily for 14 days (dose not indicated) before colorectal endoscopy or colorectal cancer surgery	Concentration of curcumin in colorectal tissue after treatment	NCT00973869
Curcumin Chemoprevention of Colorectal Neoplasia	40	Patients with diagnosed colorectal Cancer	curcumin	Per os 4 g curcumin tablet daily × 28 days before colorectal biopsy	Gene expression Ribonucleic acid (RNA) level Apoptosis	NCT01333917
Effect of Curcumin on Dose Limiting Toxicity and Pharmacokinetics of Irinotecan in Patients with Solid Tumors	23	Patients with advanced Colorectal Cancer	curcumin	Per os Curcumin (1, 2, 3, or 4 g per day) for 4 days prior to irinotecan treatment	Maximum tolerated dose Pharmacokinetics of irinotecan	NCT01859858
Phase I Clinical Trial Investigating the Ability of Plant Exosomes to Deliver Curcumin to Normal and Malignant Colon Tissue	7	Patients with diagnosed colorectal Cancer	curcumin	Per os curcumin 3,6 g daily for 7 days or curcumin conjugated with plant exosomes daily for 7 days	Concentration of curcumin in normal and cancerous tissue Safety and tolerability of curcumin alone	NCT01294072
A Phase I/IIa Study Combining Curcumin (Curcumin C3-Complex, Sabinsa) With Standard Care FOLFOX Chemotherapy in Patients with Inoperable Colorectal Cancer.	41	Patients with Inoperable Colorectal Cancer.	curcumin	Per os curcumin at increasing doses escalating to 2 g during FOLFOX-based chemotherapy and continued for the duration of the chemotherapy course.	Completion of dose escalation over 2 cycles of therapy (Phase I)	NCT01490996
A Phase 1, Double-Blind, Randomized Clinical Study to Assess the Safety, Pharmacokinetics, and Pharmacodynamics of SRT501 (Resveratrol) in Subjects with Colorectal Cancer and Hepatic Metastases	9	Patients with colorectal cancer hepatic metastasis	Resveratrol	5 g of oral SRT501 (or placebo) once daily for 14 days before surgical removal of hepatic metastases	Safety, tolerability and pharmacokinetic profile of SRT501	NCT00920803
Phase I Repeat-Dose Study of Resveratrol in Colorectal Cancer Patients: Tolerability, Target Tissue Levels and Pharmacodynamics	20	Patients with diagnosed chirugicaly removable colorectal Cancer	Resveratrol	Oral administration 8 days before colorectomy (dose and posology not given)	Pharmacodynamics of resveratrol Concentrations of biomarkers in tumoral and healthy colorectal tissue (Ki67 and COX-2 expression, M1G concentration)	NCT00433576
Resveratrol for Patients with Colon Cancer	11	Patients with diagnosed colorectal Cancer and planned for surgical resection	Resveratrol	Resveratrol tablets at 20, 80, or 160 mg/day during 14 days before surgical resection	Test the hypothesis that resveratrol modulates Wnt signaling in vivo in colon cancer and normal colonic mucosa	NCT00256334
Phase I Biomarker Study of Dietary Grape-Derived Low Dose Resveratrol for Colon Cancer Prevention	30	Healthy patients	Resveratrol and others polyphenols	Seedless fresh red grapes at 3 different doses: 1 lb/day, 2/3 lb/day, and 1/3 lb/day	Define the minimum dietary achievable amount of resveratrol-rich fresh red grapes which are effective in inhibiting Wnt signaling in human colonic mucosa.	NCT00578396
Phase II studies
Evaluate First Line Avastin/FOLFIRI in Combination with Curcumin-containing Supplement in Colorectal Cancer Patients with Unresectable Metastasis	50	Colorectal Cancer Patients with Unresectable Metastasis	curcumin	Dietary supplement of nanostructured lipid curcumin particle 100 mg × 2 per os daily during Avastin/FOLFIRI chemotherapy	Progression free survival	NCT02439385
Phase II Double Blind Placebo-Controlled Trial of Curcuminoids’ Effect on Cellular Proliferation, Apoptosis and COX-2 Expression in the Colorectal Mucosa of Subjects with Recently Resected Sporadic Adenomatous Polyps	56	Patients with diagnosis for colon/rectal polyp resection, polypectomy	curcumin	Per os encapsulated 4 g curcumin or placebo daily for 4 months	Cellular proliferation and apoptosis in the colonic mucosa COX-2 expression and activity	NCT00118989
Curcumin for Treatment of Intestinal Adenomas in Familial Adenomatous Polyposis (FAP)	44	Patients with familial adenomatous polyposis	curcumin	Per os curcumin or placebo twice a day during 12 months	Polyp Number Polyp size Number of participants with a decrease in polyp burden	NCT00641147
Efficacy of Wholistic Turmeric Supplementation on Polyp Number and Size in Patients with Familial Adenomatous Polyposis A Randomized, Double Blinded, Placebo Controlled Study	40	Patients with familial adenomatous polyposis	curcumin	Per os 2 g curcumin or placebo twice a day during 6 months	Polyp Number Polyp size	NCT03061591
A Randomized Double Blinded Study of Curcumin with Pre-operative Capecitabine and Radiation Therapy Followed by Surgery for Rectal Cancer	45	Patient with T3,4 N0,1,2 or T2N1,2 adenocarcinoma of the rectum	curcumin	Per os curcumin twice a day during weeks 1–11 of capecitabine/radiation treatment	Pathologic complete response Change in curcumin level in tumor tissue and serum	NCT00745134
Phase III studies
Phase III Trial of Gemcitabine, Curcumin and Celebrex in Patients with Metastatic Colon Cancer	100	Locally advanced or metastatic colon cancer	curcumin	No information	Tumor progression	NCT00295035
Other studies
The Effect of Plant Phenolic Compounds on Human Colon Epithelial Cells	130	Patients at average risk or above average risk for development of colon cancer	curcumin, rutin and quercetin	Per os curcumin, rutin or quercetin, twice a day during 6 to 10 weeks	Determine colonic epithelium response to plant phenolics short term treatment Compare colonic mucosal response to plant phenolics with their response to sulindac	NCT00003365

Current (October 2021) clinical trials registered on ClinicalTrials.gov and combining key words “colorectal cancer” and “resveratrol” or “curcumin” or “quercetin” or “sulforaphane” or “capsaicin.”

A phase II clinical trial, including 44 patients with 8 or more aberrant crypt foci (ACF) identified by colonoscopy, evaluated the preventive action of curcumin on the occurrence of colorectal neoplastic lesions. Patients were given an oral daily 2 or 4 g treatment of curcumin for 1 month. A significant 40% reduction in ACF number was observed with the 4 g dose (p < 0.005) but not with 2 g (Carroll et al. 2011). Importantly, both doses were well tolerated. Curcumin impact on intestinal adenomas was tested in a cohort of patients with familial adenomatous polyposis (FAP) (Cruz-Correa et al. 2018). There was no difference in the mean number or size of lower intestinal tract adenomas between curcumin (3 g/day for 12 weeks) and placebo. By contrast, significant FAP number and size decreased after 6 months with BFCs supplementation including quercetin (20 mg 3 times/day) and curcumin (480 mg 3 times/day) (Cruz-Correa et al. 2006). These contradictory results may be explained by the beneficial combinatory action of two BFCs and/or by a long-term effect with longer treatment.

Curcumin bioavailability and tolerability were studied in phase I trials (Sharma et al. 2001; Sharma et al. 2004; Garcea et al. 2005), finding no toxicity while biological effects were observed. Curcumin concentrations were measured in tumor and healthy tissues. As curcumin presents low bioavailability after oral administration, a liposomal form was recently developed and tested in a phase I study including metastatic CRC patients (Greil et al. 2018). The administration was carried out intravenously daily for 8 weeks. The authors considered the dose of 300 mg/m² over 6 h as the maximum tolerated dose. Over this dose, some patients presented hemolysis or hemoglobin drops. Levels of resveratrol and its Phase II metabolites were higher in the right-side colon compared to the left side, in direct relation with the anatomy portion after the ileum, suggesting that resveratrol is not fully absorbed in the intestine or that a portion is released back into the colon (Patel, Brown, et al. 2010). Moreover, daily oral doses of resveratrol at 0.5 g or 1 g during 10 to 21 days were sufficient to obtain resveratrol and associated metabolites concentrations in the gastrointestinal tract to elicit anti-carcinogenic effects. Alternative administration routes have been developed and tested for enhanced resveratrol systemic availability. In a randomized phase, I trial, assessment of safety, pharmacokinetics, and pharmacodynamics of micronized resveratrol was performed in CRC patients with hepatic metastases eligible for hepatectomy (Howells et al. 2011). Treatment was well tolerated and plasmatic levels of resveratrol were increased as compared to non-micronized formulations. Additionally, authors reported a 39% significant increase of the apoptosis biomarker caspase 3, in malignant hepatic tissues with resveratrol treatment compared with tissues from the placebo group.

The tolerability of chemotherapy in association with BFCs was also evaluated. In a randomized clinical study, 72 patients with metastatic CRC cancer received leucovorin, 5-fluorouracil, and oxaliplatin in combination with either MB-6, a botanical preparation containing grape seeds and curcumin extracts, or a placebo for 16 weeks (Chen et al. 2014). Patients’ follow-up (70 weeks after treatment) revealed no significant difference in best overall response rate and OS between groups. However, patients in the MB-6 group exhibited a significant lower disease progression rate (0.0% vs 15.8%, p = 0.026). Moreover, no higher incidence of adverse events was observed in the MB-6 group compared to the placebo group. Curcumin was also tested in combination with FOLFOX in a phase I dose-escalation study (James et al. 2015). Authors reported curcumin to be a safe and tolerable adjunct to FOLFOX at doses up to 2 g daily. Concordant results were recently combining FOLFOX with oral curcumin (2 g daily) or placebo (Howells et al. 2019). Similar adverse event profiles were reported for both arms. More randomized trials considering OS and progression-free survival (PFS) as primary endpoints are needed to assess the antitumor benefit of BFCs in CRC.

Limits and future directions

Overall BFCs research is a worthy research domain because it holds promising chemopreventive or antitumor actions. However, attention should be given to their hormetic properties that could defeat their benefits. Indeed, enhancement of ROS production leading to antitumor effect was typically reported for high doses or concentrations of BFCs but low doses were cytoprotective. For instance, low concentrations of resveratrol (1 to 10 μM) enhanced the proliferation of the HT-29 CRC cell line, while higher concentrations were cytotoxic (San Hipólito-Luengo et al. 2017). Similarly, low doses of capsaicin promoted metastasis of CRC cell lines in vitro and in a murine preclinical model by triggering ROS generation and activation of the Akt-mTOR pathway (Yang et al. 2013). Similarly, in combination with 5-FU, a high concentration of resveratrol (200 μM) increased chemotherapy-triggered apoptosis of the HCT116 cell line (Chan et al. 2008). By contrast, lower doses (25 or 50 μM) inhibited the pro-apoptotic action of 5-FU, especially in p53 proficient cells. In the same way, sulforaphane annihilated resveratrol and capsaicin combined antitumor action in vivo (Vendrely et al. 2017). Sulforaphane inhibited T cell-mediated immune response, which may, in turn, interfere with checkpoint inhibitors’ efficiency (Liang et al. 2019). Therefore, BFCs association with conventional cancer therapies requires a deep understanding of the positive or negative synergy of combinations (Roell, Reif, and Motsinger-Reif 2017; Fernando, Rupasinghe, and Hoskin 2019).

The recurrent concern about using BFCs in routine clinical practice is their challenged bioavailability. A large majority of compounds present low absorption rates or are actively transformed into less potent metabolites by digestive tract tissues and liver upon oral administration (Figure 3). CRC may actually avoid this pitfall since non-metabolized oral-given BFCs can reach the tumors. More work, comparing the toxic effects of BFCs according to administration routes is however required.

Figure 3. Colorectal tumor anatomy allows direct intake of unmetabolized active BFCs after oral administration with entero-hepatic metabolization cycle shunt. 1. After oral administration, a part of the BFCs is metabolized by microbiota; 2. BFCs and metabolites can be directly uptake by normal intestine and colon tissues but also by colorectal tumors; 3. Enterocytes metabolize a part of the BFCs and release BFCs and their metabolites into portal circulation to the liver; 4. After hepatic metabolization, BFC metabolites are released into the general blood circulation; 5. BFCs metabolites can reach either colorectal tumor or any other tumor site through blood circulation. Unlike distant tumor sites receiving only metabolized non-active BFCs (blue path), colorectal tumors are directly exposed to active non-metabolized BFCs (green path). This figure was made using Servier Medical Art support (https://smart.servier.com).

To overcome bioavailability limits, researchers developed various galenic formulations of BFCs. Encapsulation into nanoparticles, either by physical, chemical, or physicochemical methods were frequently tested, as recently reviewed for polyphenols (Witika et al. 2021; Teng et al. 2021). The antitumor effect of curcumin on CRC cell lines was enhanced when encapsulated as Gemini surfactant nanoparticles (Ebrahimi et al. 2021) and nano encapsulated chrysin-curcumin combination showed a synergistic antitumor effect on the SW480 CRC cell line (Bagheri, Sanaat, and Zarghami 2018). This synergy was not observed with the free forms. Similarly, synthetic quercetin-caffeic-acid phenethyl ester co-loaded nanoparticles exhibited stronger anti-proliferative, anti-migration, and apoptotic properties than the free form of the combination (Colpan and Erdemir 2021). Recently, nanoparticles were used to combine curcumin with a silencer of the long non-coding (lnc) RNA colon cancer-associated transcript-1 (CCAT1) involved in CRC oncogenesis (Jia et al. 2021). The authors reported the synergistic anti-proliferative effect of the encapsulated combination better than drugs encapsulated individually. Thus, encapsulation can both enhance antitumor activity (by increasing bioavailability) as well as synergism of BFCs.

Excluding curcumin, for which several clinical studies have been carried out, BFC inclusion in clinical trials remains limited, and this is despite intense academic research. We believe that considering the high cost of clinical trials, the unpatented availability of BFCs is not attractive for pharmaceutical companies or private promoters. Patents may become possible with the development of new galenic forms or combination treatments. The time gap between basic research and clinical transfer is to deplore, especially for the patients, for whom the benefit may be important.

The toxicity of radiotherapy and chemotherapy is a major concern for oncologists and patients. Unlike standard treatments, BFCs do not induce major systemic toxicity, as shown by numerous phase I and II clinical trials. Surprisingly, very few studies explored this absence of toxicity on healthy cells compared to their tumor cell counterparts. Curcumin exerted a greater pro-apoptotic impact on hepatocarcinoma cell lines compared to healthy hepatocytes (Syng-Ai, Kumari, and Khar 2004), in close relation with intrinsic levels of the radical scavenger Glutathione. Curcumin also differentially modulated morphologic and elastic properties of mammalian cancerous and normal epithelial cells with a selective negative impact on tumor cells (Saab et al. 2013). In the same way, resveratrol was pro-oxidative in human astrocytoma cells, but not in primary healthy astrocytes (Gran et al. 2021). Thus, some BFCs exert tumor-specific cytotoxicity while sparing surrounding healthy tissue. This interesting property reinforces the cause of using BFCs in cancer, as reviewed for resveratrol (Mortezaee et al. 2020). Moreover, in combination with conventional treatments, BFCs could help maintain tumor control for fragile or comorbid patients necessitating treatment de-escalation. For instance, the combination of resveratrol and capsaicin associated with a gemcitabine dose reduced by 1/3 rescued the full dose anti-tumor effect (Vendrely et al. 2017).

In conclusion, the mechanisms ruling BFCs chemoprotective and antitumor activities in CRC are still unclear, especially accounting for their systemic metabolism and tumor bioavailability. Another important aspect of their use will no doubtedly explore pharmaceutical and galenic developments to optimize their oral bioavailability. Despite this, the current state of knowledge positions BFCs as important players for preventing and treating CRC, particularly because they are part of most diets, and because they do not carry systemic toxicity. However, clinical studies are still needed to fully reveal their health potential.

Declaration of interest statement

The authors have no conflicts of interest to declare.

Funding

This work was supported by SIRICBRIO (Bordeaux Recherche Intégrée en Oncologie).