Cannabinoid Effects on Experimental Colorectal Cancer Models Reduce Aberrant Crypt Foci (ACF) and Tumor Volume: A Systematic Review

Objective Colorectal cancer represents a heavy burden for health systems worldwide, being the third most common cancer worldwide. Despite the breakthroughs in medicine, current chemotherapeutic options continue to have important side effects and may not be effective in preventing disease progression. Cannabinoids might be substances with possible therapeutic potential for cancer because they can attenuate the side effects of chemotherapy and have antiproliferative and antimetastatic effects. We aim to determine, through a systematic review of experimental studies performed on animal CRC models, if cannabinoids can reduce the formation of preneoplastic lesions (aberrant crypt foci), number, and volume of neoplastic lesions. Materials and Methods A systematic, qualitative review of the literature was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed, Embase, and Scopus databases were searched. We use the following Medical Subject Headings (MESH) terms in PubMed: “colorectal neoplasms,” “colonic neoplasms,” “colorectal cancer,” “polyps,” “rimonabant,” “cannabidiol,” “cannabinoids,” “azoxymethane,” “xenograft,” and “mice.” Only studies that met the eligibility criteria were included. Results Eight in vivo experimental studies were included in the analysis after the full-text evaluation. Seven studies were azoxymethane (AOM) colorectal cancer models, and four studies were xenograft models. Cannabidiol botanical substance (CBD BS) and rimonabant achieved high aberrant crypt foci (ACF) reduction (86% and 75.4%, respectively). Cannabigerol, O-1602, and URB-602 demonstrated a high capacity for tumor volume reduction. Induction of apoptosis, interaction with cell survival, growth pathways, and angiogenesis inhibition were the mechanisms extracted from the studies that explain cannabinoids' actions on CRC. Conclusions Cannabinoids have incredible potential as antineoplastic agents as experimental models demonstrate that they can reduce tumor volume and ACF formation. It is crucial to conduct more experimental studies to understand the pharmacology of cannabinoids in CRC better.


Background
Colorectal cancer (CRC) is the third most common cancer worldwide, only behind prostate and lung in males, and behind breast and lung in females [1]. It has high morbidity and mortality that represents a heavy burden for health systems worldwide. In the United States alone, with roughly 1.8 million new cases in 2018, healthcare costs exceed $14 billion annually [2]. In addition, it is the fourth cause of cancer-related deaths [3,4]. CRC represents a significant public health concern because temporal projections estimate that its global burden will increase by 60% to more than 2.2 million new cases and 1.1 million cancer deaths by 2030 [5].
CRC is a type of cancer with a complex and heterogeneous pathophysiology. It is the result of the transformation of healthy colonic epithelial cells into cancer [6]. is process, called "adenoma-carcinoma sequence," develops through an ordered series of events, in which the initial step is the transformation of normal colonic epithelium to aberrant crypt foci (ACF) [6]. ACF progress to CRC, in 10-15 years [7]. During this process, many risk factors play an essential role in pathogenesis, including unhealthy diet, smoking, alcohol use, physical inactivity, inflammatory bowel disease, and aging [2].
Breakthroughs in CRC therapy have decreased the mortality of patients with CRC. Current chemotherapeutic options continue to have important side effects due to cytotoxicity and may fail to prevent disease progression [8].
us, there is a great interest in new therapeutic approaches for CRC, including phytochemical agents.
Cannabinoids might be substances with possible therapeutic potential for cancer because of their chemotherapeutic effect and their ability to attenuate anorexia, pain, and emesis; these are common side effects of chemotherapy [9,10]. is has been proved in several experimental models of CRC, brain cancer, breast cancer, lung cancer, prostate cancer, leukemia, and melanoma [11]. However, to the best of our knowledge, cannabinoids have not been tested in humans as medicines for CRC.
Animal models and cell lines of CRC have tested cannabinoids. is study aims to conduct a systematic review of the research about the effect of cannabinoids on in vivo azoxymethane (AOM) or xenograft CRC models. e outcomes used to assess the effects of cannabinoids, compared with no cannabinoid therapy, were a decrease in the number of preneoplastic lesions (aberrant crypt foci), number, and volume of neoplastic lesions.

Materials and Methods
e protocol for this study was registered in PROSPERO (International Prospective Register for Systematic Reviews) under CRD42019148356 [12].
is systematic review was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Supplementary file) [13].

Eligibility Criteria.
Population. e population should be animal species (no restrictions), used for in vivo models of CRC, either chemically induced (Azoxymethane or DSS) or by xenograft injection. Dose and time of exposure to azoxymethane were not exclusion criteria for this review. We excluded all studies that included only in vitro assessment and studies that evaluated species for noncolorectal cancer models.
Comparators. Studies had to include at least one comparator group of the same animal species used for the intervention group, with similar characteristics (weight, age, sex, exposure to the same environment, and feeding), without exposition to cannabinoid therapy.
Studies. Studies should be experimental in vivo studies of CRC in mice, with at least one control group. We excluded conference abstracts, narrative reviews, and systematic reviews.
Primary Outcome. ere should be a reduction in tumor volume (mm 3 ), number of aberrant crypt foci (ACF), and number of tumors comparing intervention and control group.

Study Selection and Data Collection.
e authors EOG and LLT conducted the search independently. Duplicate articles were moved to a different folder and registered in the flowchart. Before the selection process, a test was conducted to evaluate the agreement between evaluators. All titles and summaries of the articles were assessed by EOG and LLT independently based on a selection criterion. e full text of previously selected studies was then reviewed and analyzed. Any disagreement was discussed, and if not resolved, a third author (AVV) was consulted. All selected articles were summarized in a flowchart according to the PRISMA protocol. We used a standardized form with a pilot test to collect the following data: title, author, publication year, type of animal model, sample size, type of cannabinoid, the dose of cannabinoid, the dose of AOM, type of outcome measure, length of the experiment, reduction in ACF formation, reduction in the number of tumors, tumor volume reduction, pathway or function modified by cannabinoids, an increase of endocannabinoid levels, expression of apoptosis markers, and expression of proinflammatory markers.

Quality Assessment.
e risk of bias was independently evaluated by two authors (EOG and LLT) following the Systematic Review Centre for Laboratory animal Experimentation (SYRCLE) risk of bias tool [14]. e domains considered were random sequence generation, baseline characteristics, allocation concealment, random housing, blinding, random outcome assessment, outcome assessor blinding, incomplete outcome data, selective outcome report, and other sources of bias (contamination, the influence of funders, and analysis of errors) [14]. We reviewed each article, and we sought if any of these biases were present. Any discrepancy was discussed between 2 authors (EOG and LLT), and if not resolved, a third investigator intervened (AVV). Evidence-Based Complementary and Alternative Medicine 2.5. Data Analysis. Proportions were used as descriptive statistics for primary outcomes. Secondary outcomes were described qualitatively. A meta-analysis or measures of consistency were not performed due to characteristics of the studies and heterogeneity of articles.
Δ9-Tetrahydrocannabinol (THC) and anandamide have the highest affinity for the CB1 receptor, while CBD exhibits low affinity for these receptors [20,21]. However, CBD has been proved to enhance endocannabinoid levels and indirectly activate CB receptors [8].
CB1 receptors constitute one of the most abundant receptors in the central nervous system. In the case of CB2 receptors, these are expressed in cells of the immune and hematopoietic system, spleen, and tonsils, modulating cytokine release and cellular immune migration [16]. Both receptors are metabotropic and belong to the G protein-coupled receptor (GPCR) family, and their activation produces inhibition of the adenylyl cyclase via G proteins (Gi/o) [16].
is decreases cAMP in the cell and activity of protein kinase A [16].
Cannabinoids may have alternative molecular targets other than classical CB1 and CB2 receptors [22]. Recently, orphan GPCRs like the GPR 55, GPR18, and GPR110 have been identified as new targets [22]. ere is also increasing evidence that they can interact with ionotropic receptors such as the transient receptor potential cation channel subfamily V member 1 (TRPV1), and the transient receptor potential cation channel subfamily M member 8 (TRPM8) [23]. e transient receptor potential vanilloid receptor 1 (TRPV1) and the transient receptor potential cation channel subfamily M member 8 (TRPM8) are ionotropic channels that allow Na + and Ca ++ entry to the cell [24]. Cannabidiol (CBD) and cannabigerol (CBG) close the TRPM8 channel, whereas CBD opens TRPV1 [24].
GPR55 is another GPCR, which is coupled to a Gα12/13 protein [25]. Several cancer lines like OVACAR3 (ovarian cancer cell line), PC-3, and DU145 (prostate cancer cell lines) exhibit expression of this orphan receptor [25]. Furthermore, Piñeiro et al. showed an autocrine activation of this receptor through his main endogenous agonist lysophosphatidylinositol (LPI) [25]. e receptor acts via activation of Gα12 and Gq family proteins, which activate Ras homolog gene family, member A (RhoA) kinase [26]. Overexpression of GPR55 produces increased levels of pERK in HEK-293, breast carcinoma, and glioma cells, while pAKT levels are increased in ovarian and prostate cancer cells [25,27]. e high expression of GPR55 is also linked to high proliferation indices in human breast tumors and Glioblastoma [26]. e best-studied cannabinoid with actions on the GPR55 in colonic tissue is O-1602 [28]. It is highly speculative that the compound exerts its antineoplastic effects on CRC tissue through the GPR55 receptor, as the cannabinoid has shown agonist activity on this receptor [28]. More research is needed before we can conclude the actions of cannabinoids on this receptor.
One of the main effects demonstrated in experimental models is apoptosis. is is proposed to be mediated through the upregulation of endoplasmic reticulum stress-related genes (ATF-4, TRB3), accumulation of reactive oxygen species (ROS), cell cycle arrest in a p-53 independent manner, and activation of proapoptotic proteins (BAX, caspase 3/9) [38,39,[42][43][44]. Ceramide enhanced production is another mechanism of apoptosis induction by cannabinoids through the mitochondrial pathway [45]. In these cases, proapoptotic proteins, usually sequestered in the intermembrane space, are released into the cytosol, assembling the "apoptosome" (formed by the binding of cytochrome c, Apaf-1, and procaspase-9) [46]. Procaspase 3 is cleaved by the apoptosome and causes morphological and biochemical changes seen in apoptosis [46].

Evidence-Based Complementary and Alternative Medicine
Cannabinoids also regulate different proliferation, growth, and survival pathways [45]. Different carcinoma cell lines treated with ∆9-THC also exhibited inhibition of the RAS-MAPK pathway and the phosphatidylinositol 3-kinase pathway (PI3k-AKT) pathway through CB1-receptor activation or ceramide accumulation [34,45].
In vitro studies have also reported that COX-2 metabolites of anandamide (PGE2-EA and PGD2-EA) have growth inhibitory effects in CRC [47][48][49]. In the case of rimonabant, this is an inverse agonist of CB1 receptors; however, it was demonstrated that this compound counteracts the Wnt/ β-catenin pathway by decreasing the activity of transcription factors T-cell factor/lymphoid enhancer-binding factor (TCF/ LEF) [50,51]. Other effects of cannabinoids are reported less frequently, including cytotoxic effects similar to 5-Fluorouracil and upregulation of estrogen receptors [52][53][54].
Contrarily to these findings, in some studies, CB receptors have been involved in CRC origin. CB1 and CB2 receptors were prognostic markers of survival in advanced CRC stages (IV) [36]. Moreover, Martínez et al. demonstrated a biphasic effect of synthetic cannabinoids on colon cancer-derived cell line HT29 [55]. Sub-micromolar concentrations (1 µm) of CB 2 -specific agonists stimulate the PI3K/AKT pathway and the transcription factor SNAIL, promoting cell proliferation in this in vitro model [55]. Figure 1 summarizes some of the main mechanisms of cannabinoids on CRC, described in the literature.

Results
We identified 94 records in electronic databases: 57 in Scopus, 27 in PubMed, and 10 in Embase. Duplicate records were removed, and 76 studies remained. After the title and abstract review, 18 studies were selected, of which only 8 were included in the analysis after the full-text evaluation ( Figure 2).
Six studies were excluded because they only included in vitro assessments, three were abstracts from a conference,  Evidence-Based Complementary and Alternative Medicine and their full-text article could not be found, and one study assessed a noncannabinoid compound.

Characteristics of Included
Studies. e characteristics of the included studies are summarized in Table 1. Seven studies were AOM-colorectal cancer models [7,8,28,35,38,56,57]. e protocol for 4 of the studies with this model was four single doses of 10 mg/kg (40 mg/kg in total, intraperitoneally) at the beginning of the first, second, third, and fourth week [7,8,38,56]. Two studies performed a model, in which 3 mg/ kg was given intraperitoneally at days 1 and 5 during the first week, 3 mg/kg at days 1 and 5 during the third week, and 2 mg/kg at days 1 and 5 during the 17 th week [35,57]. In Kargl et al.'s research, animals received a single intraperitoneal injection of AOM (10 mg/kg) [28].
However, CBD was the most common, being evaluated in two of the studies [7,8]. Romano et al. tested different cannabis extracts, with a high content of CBD [8].

Risk of Bias.
Measures to reduce performance bias were not reported in any of the publications, making them highly likely to have performance bias [7,8,28,35,38,[56][57][58]. ere was a high risk of detection bias in all studies because there was no random outcome assessment [7,8,28,35,38,[56][57][58]. We found that other potential sources of bias were present in half of the studies as pharmaceutical companies funded them [7,8,38,57]. Some of the domains had to be scored as unclear risk of bias, as there were few details about the methods used by each author [7,8,28,35,38,[56][57][58] (Table 2).
Borrelli et al. demonstrated that CBG antagonizes TRPM8 through inhibition of intracellular calcium increase, with a rise in apoptosis markers [38]. Pharmacological enhancement of endocannabinoid levels was observed in three of the included studies [7,35,56]. Izzo [8].
Two studies demonstrated that angiogenic factors (VEGF, FGF, and HIF) were downregulated by the direct action of URB602 and the synthetic cannabinoid LYR-8 [56,58]. In the case of rimonabant, it provoked a phenomenon called "mitotic catastrophe" as its mechanism of action [50]. Mitotic catastrophe is a phenomenon that has been studied in MCF7 and HeLa cells, in which cell cycle progression is inhibited due to an arrest of cells in the G2/M phase of the cell cycle [59]. is is provoked by high amounts of damaged DNA (including high amounts of cells with polyploidy) and apoptosis induction through activation of p53 [59]. is molecular effect is characteristic of the antineoplastic medicine paclitaxel [59]. Other reported effects of cannabinoids include the reduction of proinflammatory markers [7,28]. Kargl et al. showed that SW480 and HT-29 cells treated with O-1602 could reduce the expression of TNF-α, phosphorylation of NFκβ, and STAT3. TNF-α is a cytokine, which exerts a dual role in immunomodulation (inflammation, immune surveillance, and hematopoiesis) and tumorigenesis [60], while the transcription factor, NFκβ, regulates proinflammatory cytokines (TNF-α and IL-1), and it has a pathogenic function in cancer and inflammation [61,62]. In the case of STAT3, this belongs to a family of proteins, activated mainly by IL-6, which are linked to inflammation-associated tumorigenesis [63].  Evidence-Based Complementary and Alternative Medicine 9

Discussion
It was observed across the studies that cannabinoids can reduce the development of preneoplastic lesions (ACF) and tumor growth of colorectal cancer in chemically induced CRC and xenograft models. ACF arise from a stem cell in the colonic crypts and represent abnormalities before polyp formation, being regarded as the earliest preneoplastic lesions of CRC [64]. e most common genetic changes involved in ACF formation are mutations of the protein KRAS and microsatellite instability; both are key events in the two main pathways of CRC genesis [64]. Cannabinoids (CBD, AA-5HT, VDM11, HU210, and rimonabant) have chemopreventive potential, as they were able to attenuate colon carcinogenesis in vivo [7,8,35,57]. e MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay determines the mitochondrial activity an indirect measure of the number of viable cells [65]. CBD and CBG have cytotoxic effects on CRC tissue and cells, as they were tested either by MTT assay or 3 H-thymidine incorporation with a significant decrease of cell proliferation [7,8,38]. is makes them possible candidates for further research on CRC cytotoxic drugs.
In a prospective human study, researchers achieved a tumor volume reduction rate (measured by MRI-volumetric techniques), of 51.7% in patients treated with four cycles of FOLFOX [66]. In this systematic review, we found that various cannabinoids could achieve similar reduction rates of the tumor volume (33.18%-52%, for in vivo research).
Tumor growth and metastasis depend on angiogenesis because, in the absence of vascular support, tumors may become necrotic or even apoptotic [67]. URB-602 and LYR-8 have antiangiogenic mechanisms, as proven in their studies, with similar actions to bevacizumab (inhibiting VEGF) [56,58]. e strengths of this review include being the first study to our knowledge that has systematically reviewed the experimental evidence of cannabinoids in colorectal cancer, as there is no study in humans to date. e review used a comprehensive search strategy in a wide range of registries and data sources and used a well-known quality assessment (SYRCLE's risk bias tool) [14], and the manuscript was registered in the PROSPERO [12]. However, some limitations were found. e main limitation of this study was that animal models, outcomes, and cannabinoids tested were not the same across studies [7,8,28,35,38,56,57]. e quality assessment indicated that all the studies had methodological limitations and were at risk of bias [7,8,28,35,38,56,57]. e eight studies did not explicitly state blinding and allocation concealment on their methods [7,8,28,35,38,56,57].
us, it was impossible to determine if certain risks were present. Four studies were founded by pharmaceutical companies, which represent a significant bias [7,8,38,57]. We believe that these were not significant drawbacks, as we did not perform a metaanalysis. Current chemotherapy generates high costs and high toxicity for CRC treatment [68]. CRC therapy includes regimes such as FOLFOX (5-fluorouracil, leucovorin, and oxaliplatin), FOLFIRI (5-fluorouracil, leucovorin, and irinotecan), recombinant monoclonal antibodies like cetuximab, and bevacizumab [68]. It has been reported that FOLFOX and FOLFIRI have a high incidence of neutropenia (41.7% and 24%, respectively); additionally, a high percentage of patients (18%) under FOLFOX develop neurologic symptoms due to toxicity, and in a high percentage of patients treated with cetuximab, gastrointestinal symptoms like diarrhea are present (13-25%) [69]. Moreover, chemotherapy during eight weeks in the USA for CRC is estimated to generate high costs for FOLFOX/bevacizumab ($21033) and FOLFIRI/cetuximab ($30675) [70].
New molecules as effective as the current chemotherapy need to be synthesized, with less cytotoxic effects. Cannabinoid molecules are not as expensive as monoclonal antibodies and are proven that they have antiproliferative, proapoptotic, and chemopreventive effects on CRC. e next step is to conduct more in vivo research of these compounds, before proceeding to human beings, to determine whether these results are reproducible.

Conclusions
Overall, there is no robust evidence so far to introduce cannabinoids to the clinical practice as adjuvants of chemotherapy or the main treatment of CRC because no human models have tested these compounds, and only eight in vivo experimental models have been conducted and reported in the literature, during the assessed period. Nevertheless, current literature findings demonstrate that cannabinoids might have potential as antineoplastic agents because they can reduce tumor volume and ACF formation. Induction of apoptosis through several mechanisms is the main action of cannabinoids on CRC, while inhibition of angiogenesis and mitotic catastrophe were also reported. It is crucial to conduct more experimental studies before conducting research on humans, providing high-quality evidence on the efficacy and safety of cannabinoids for treating CRC. Data Availability e datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Supplementary Materials
Supplementary file 1: the PRISMA statement of this study. Supplementary file 2: the search strategy used for this study. (Supplementary Materials)