Anti-Cancer Effects of Lycopene in Animal Models of Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis

Introduction Globally, hepatocellular carcinoma (HCC) is the sixth most diagnosed cancer and the third important cause of cancer-related death. As there are only two targeted drugs for the treatment of advanced HCC—that merely extend survival by a few months—the need for alternative treatments is inevitable. Lycopene, a carotenoid that is known to be most abundant in red tomatoes and tomato-based products, has been investigated for its anticancer activity in various types of cancers including HCC. This review was conducted to evaluate the effects of lycopene on HCC from animal models to pave the way for further clinical studies. Methods Electronic databases and search engines including PubMed, EMBASE, and Google Scholar were searched for original records addressing the anticancer effect of lycopene in animal models of HCC. Data were extracted using a format prepared in Microsoft Excel and exported to Stata 15.0 for analyses. A meta-analysis was performed using a random-effects model at a 95% confidence level for the outcome measures: tumor incidence, number, and growth (tumor volume and size). The presence of publication bias between studies was evaluated using Egger’s test and funnel plot. The study protocol was registered in the PROSPERO database with reference number: CRD42019159312. Results The initial database search yields 286 articles, of which 15 studies met the inclusion criteria. The characteristics of the included studies were a bit diversified. The studies involved a total of 644 animals (312 treatment and 332 control groups) and mice shared the majority (488) followed by rats (117) and ferrets (39). The meta-analysis showed that lycopene significantly reduced the incidence [RR 0.8; 95% CI 0.69, 0.92 (p=0.00); I2 = 30.4%, p=0.16; n=11], number [SMD-1.83; 95% CI -3.10, -0.57 (p=0.01); I2 = 95.9%, p=0.00; n=9], and growth [SMD -2.13; 95% CI -4.20, -0.04 (p=0.04); I2 = 94.6%, p=0.00; n=4] of HCC. Conclusions Administration of lycopene appears to inhibit the initiation and progression of cancer in animal models of HCC. However, more controlled and thorough preclinical studies are needed to further evaluate its anti-HCC effects and associated molecular mechanisms.

Introduction: Globally, hepatocellular carcinoma (HCC) is the sixth most diagnosed cancer and the third important cause of cancer-related death. As there are only two targeted drugs for the treatment of advanced HCC-that merely extend survival by a few months-the need for alternative treatments is inevitable. Lycopene, a carotenoid that is known to be most abundant in red tomatoes and tomato-based products, has been investigated for its anticancer activity in various types of cancers including HCC. This review was conducted to evaluate the effects of lycopene on HCC from animal models to pave the way for further clinical studies.
Methods: Electronic databases and search engines including PubMed, EMBASE, and Google Scholar were searched for original records addressing the anticancer effect of lycopene in animal models of HCC. Data were extracted using a format prepared in Microsoft Excel and exported to Stata 15.0 for analyses. A meta-analysis was performed using a random-effects model at a 95% confidence level for the outcome measures: tumor incidence, number, and growth (tumor volume and size). The presence of publication bias between studies was evaluated using Egger's test and funnel plot. The study protocol was registered in the PROSPERO database with reference number: CRD42019159312.

INTRODUCTION
Hepatocellular carcinoma (HCC) is a common type of primary liver cancer that has become a major public health problem ranked as the sixth most common cancer and the third leading cause of cancerrelated deaths (Bray et al., 2018). Surgical resection, liver transplantation, and locoregional therapies are potential curative treatment options for early or intermediate stage HCC (Kirstein and Wirth, 2020). For patients with advanced HCC, however, only two targeted drugs-multi-kinase inhibitor sorafenib, and the vascular endothelial growth factor-2 (VEGFR-2) antagonist ramucirumabthat only increase patient survival by few months exist (Lurje et al., 2019;Vogel and Saborowski, 2020). Therefore, the development of alternative methods of treatment needs urgent attention.
Lycopene, a bioactive deep-red pigment naturally synthesized by fruits and vegetables, is the most abundant carotenoid in red tomatoes and tomato-based products, including ketchup, tomato juice, and pizza sauce (Ilahy et al., 2019). Human and experimental studies have reported that lycopene can influence the development and progression of HCC (Mein et al., 2008;Ip and Wang, 2013;Stice et al., 2018;Thomas et al., 2020). For instance, in studies conducted using hepatoma cell lines lycopene inhibited cell growth, migration, and invasion (Huang et al., 2008;Yang et al., 2012;Jhou et al., 2017). Besides, orally administered lycopene attenuates liver-specific carcinogen diethylnitrosamine (DEN)-induced hepatocarcinogenesis in rodents (Ip et al., 2014;Sahin et al., 2014;Bhatia et al., 2018). To our knowledge, no published systematic review or metaanalysis that reported evidence vis-à-vis the effects of lycopene in animal models of HCC. Hence, we conducted a systematic review and meta-analysis to evaluate the anti-HCC effects of lycopene in animal models of HCC.

Study Protocol
This study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al., 2009). The completed checklist is provided as supplementary material (Additional file 1: Table  S1). The study protocol is registered on PROSPERO with reference number CRD42019159312 and available at: https:// www.crd.york.ac.uk/prospero/#recordDetails

Data Sources and Search Strategy
We conducted an electronic literature search in PubMed and EMBASE from onset to February 29 th , 2020 using the following keywords and indexing terms: "lycopene", "carotenoids", "hepatocellular carcinoma", and "animal model". Reference lists of identified citations and Google Scholar were also searched to identify additional studies. Boolean operators (AND, OR) or its meta and truncation were used when appropriate to increase the number of relevant articles. An example of the search strategy employed in PubMed is provided as supplementary material (Additional file 2: Table S2).

Study Selection and Inclusion Criteria
Two reviewers independently reviewed the articles for inclusion through scanning of titles and abstracts for relevance. Articles deemed important or for which decisions were difficult to exclude were retained for full-text review. Similarly, two reviewers independently reviewed the articles against predefined inclusion criteria: used animal models of HCC, reported anticancer effects directly generated by lycopene treatment, and controlled studies with a separate control group, published in English. There was no restriction based on the year of publication, or sample size. Disagreement during the full-text review was resolved through discussion until unanimity.

Data Extraction and Quality Assessment
By using a format prepared in Microsoft Excel the following items were extracted: author, year, characteristics the animal models (species/strain, age, sex, weight, and type), number of animals in the experimental and control group, the dose of lycopene, timing, duration and route of administration, and outcomes measures including tumor size (TS), tumor volume (TV), tumor number, tumor incidence (TI), and molecular markers.
The quality of the included studies was evaluated by using the Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies (CAMARADES) checklist (Macleod et al., 2004). The items consisted of (1) publication in a peer-reviewed journal; (2) statement of control of temperature; (3) randomization to treatment or control; (4) allocation concealment; (5) blinded assessment of outcome; (6) use of a suitable animal model of HCC; (7) sample size calculation; (8) avoidance of anesthetics with marked intrinsic properties; (9) statement of compliance with regulatory requirements; and (10) statement of potential conflicts of interest. A point was given for each criterion reported. The Abbreviations: CDK, Cyclin-dependent kinase; GPx, Glutathione peroxidase; GST, Glutathione-S-transferase; HO-1, Heme oxygenase-1; IL, Interleukin; NFkb, Nuclear factor-kappa B; Nrf2, Nuclear factor E2-related factor 2; SMD, Standardized mean difference; SOD, Superoxide dismutase; STAT, Signal transducer and activator of transcription; TI, Tumor incidence; TN, Tumor number; TNFa, Tumor necrosis factor alpha; TS, Tumor size; TV, Tumor volume; VEGFR-2, Vascular endothelial growth factor-2. potential score ranges from 0 to 10, with higher scores indicating greater methodological rigor. Two independent investigators performed a quality assessment and disagreements were resolved through discussion by reviewing the full text together.

Data Processing and Statistical Analysis
The extracted data were analyzed using Stata 15.0. Risk ratio (RR) was used as the effect measure for the outcome of tumor incidence and the standardized mean difference (SMD) for the outcome measures of tumor growth (TV, TS) and number. Considering the variation in true effect sizes across the population, the random-effects model was applied for the analysis at a 95% confidence level. The significance of heterogeneity of the studies was assessed using I 2 statistics based on Cochran's Q test, I 2 returns, and the percent variation across studies. Subgroup analysis was performed based on type of HCC model, species/strain, and lycopene dose. The presence of publication bias was evaluated using Egger's regression (Egger et al., 1997) tests and presented with funnel plots. A statistical test with a p-value of less than 0.05 was deemed to be significant.
The median quality score checklist items scored was 6 out of 10 (interquartile range, 5-7; Additional file 3: Table S3). All the included studies were peer-reviewed publications, used a suitable animal model of HCC, and avoided the use of anesthetics with marked intrinsic properties. However, none of the studies reported the detailed protocol regarding the conducted experiment, and all of the studies did not report sample size calculations except one study (Sahin et al., 2014).

Meta-Analysis
Tumor Incidence Seven studies with 11 independent experiments were included to investigate the effect of lycopene treatment on HCC incidence. The pooled estimate indicated a significant decrease in tumor incidence (RR 0.8; 95% CI 0.69, 0.92 (p=0.00), and between-study heterogeneity was not evident (I 2 = 30.4%, p=0.16; Figure 2

Tumor Growth
Four studies with 4 independent experiments were included to investigate the effect of lycopene treatment on HCC growth. The pooled estimate suggested a significant inhibition of HCC growth [SMD -2.13; 95% CI -4.20, -0.04 (p=0.04); I 2 = 94.6%, p=0.00] (Figure 4). Due to the limited availability of data, it was not possible to conduct subgroup analyses.

Publication Bias
The publication bias evaluation for the meta-analysis of tumor incidence (11 studies) is shown in Figure 5. No significant publication bias was observed for TI (Egger' Sensitivity analysis showed that no studies substantially influenced the overall RR/SMD after removing one study at a time ( Table 3).

DISCUSSION
This systematic review and meta-analysis was conducted to evaluate the effect of lycopene against animal models of HCC. We found that lycopene proved to generate a robust antitumor activity in animal models of HCC. Various mechanisms have been pointed out to explain the antitumor activity of lycopene against HCC development and progression. Although several in vitro and in vivo studies reported beneficial effects of lycopene against HCC, to the best of our knowledge, this is the first metaanalysis that assessed the anticancer activity of lycopene in animal models of HCC.
Lycopene has been shown to inhibit liver tumor initiation by diverse mechanisms including suppressing the expression of cytochrome p450 2E1 enzymes (involved in the activation of procarcinogens), scavenging oxygen free radicals or upregulating the antioxidant defense system (Wang et al., 2010;Stice et al., 2015;Aizawa et al., 2016). The latter due to its ability to stimulate the nuclear factor E2-related factor 2 (Nrf2) that is known to be a crucial regulator of the cellular response to oxidative stress (Wang et al., 2010). Nrf2 has been shown to induce the expression of antioxidant or detoxifying enzymes such as superoxide dismutase (SOD), glutathione-Stransferase (GST), glutathione peroxidase (GPx), heme oxygenase-1 (HO-1), and catalase (Gupta et al., 2013b;Bhatia et al., 2018).
The strength of this review is the use of evidence generated from controlled randomized experiments and prospectively  collected data. However, our review also has some limitations that should be considered during interpretation of findings. First, the use of different species of animals made controlling the contribution of inter-species differences. We addressed this through sub-group analysis for each species. But given the overall small number of experiments, such sub-group analysis may not yield strong evidence. Second, the overall small number of animals in the study and its corresponding small number of experiments should be considered.

CONCLUSION
The current meta-analysis suggests that lycopene has a potential to prevent HCC development and progression through its effect on the incidence, number and growth of HCC. This occurs as a result of antioxidant, anti-inflammatory, antiproliferative, and apoptosis inducing actions of lycopene in animal models of HCC. However, further controlled studies are required to support these mechanisms.

AUTHOR CONTRIBUTIONS
AM, BH, KM, and BM were involved in the conception, design, analysis, interpretation, and manuscript writing. AT, JA, and MS were involved in the design, analysis, and critically reviewing the manuscript. All authors contributed to the article and approved the submitted version.