Does Fingolimod have anticancer effects? A meta- analysis and systematic review based on experimental animal models of various cancers


 Background: Numerous studies have explored the anticancer effect of FTY720 (Fingolimod) in animal models, a sphingosine-1-phosphate (S1P) receptor antagonist and an immunosuppressant, but little clinical evidence guides the use of FTY720 in cancer patients.Methods: Strictly, only related published articles about the treatment with FTY720 for various cancers in vivo from January 1998 to January 2020 were selected from PubMed, Web of Science, Ovid, Embase, CNKI and Cochrane databases, and which were qualified. We acquired agreement through discussion. Then, we conducted meta-analysis, subgroups analysis, publication bias analysis and sensitivity analysis based on selected studies. In the last two sections, we summaried and compared side effects, drug combination effects and molecular pathways from selected studies.Results: In the 31 articles included from 2002 to 2019, FTY720 was found to reduce tumor volume (SMD =-2.58, 95% CI: -3.42, -1.75, Z = 6.09, P = 0.000), tumor weight (SMD = -3.69, 95% CI: -5.17, -2.21, Z = 4.88, P = 0.000) and body weight (SMD = -0.86, 95% CI: -1.61, -0.11, Z = 2.23, P = 0.025) in 14 types of cancer. Relevant frequent signal pathways include the Akt pathway, S1PRs-Caspase pathway and the STAT3-PP2A pathway. FTY720 has significant independent or in combination anticancer effects and a lower toxicity in renal cell carcinoma and neuroblastoma mice models. However, it should be noted that FTY720 achieved a significant therapeutic effect in immunodeficient mice, not in immunecompetent mice. Also, the dosage-safety of FTY720 alone in clinical use is a noteworthy issue. In mouse models, the mechanism of the FTY720 treatment of tumors lies in inducing the tumor cells apoptosis through important signaling moleculars.Conclusions: FTY720 alone or in combination exerted significant anti-tumor effects for neuroblastoma and renal cell carcinoma, however not for melanoma. Due to insufficient evidence, more specific studies of FTY720 only and in combination included in immunity, inflammation and melanoma should be carried out in the future preclinical and clinical studies.


Background
Sphingosine-1-phosphate, abbreviated as S1P, is an active form of sphingolipids and is exported to outside of cells after synthesised in nuclear. Some excellent reviews have summarized the roles of S1P in in ammatory and allergic responses and cancer (Hait and Maiti 2017;Maceyka et al. 2012;Pyne et al. 2016). S1P lyase or phosphatases and sphingosine kinases (SphKs) balance the S1P metabolism in plasama membrane, cytoplasm and outside of cells. S1P can target these signal molecules as Ras, PI3K, and Rho etc by binding S1P receptors, S1PR1-5, a kind of G protein-coupled receptors (GPCRs) (Ogretmen 2018).
There have been exciting animal experiments that explored sphingolipid metabolism and signaling involved in carcinogenesis and tumor metastasis. Recently, target sphingolipid signaling are developing in anti-cancer therapy, however unavailable SphK1 inhibitors were clinically adopted. FTY720 (chemical name: 2-amino-2-[2-(4-octylphenyl)-1,3-propanediol hydrochloride) is a prescription drug for multiple sclerosis (MS) treatment. S1PR1 is the main target of FTY720. Mechanismly, FTY720 stably downregulates S1PR1 in a unique way of interlization while transiently suppressing S1PR3, 4, 5. A transient heart rate decrease has been reported as a major clinical side effect of FTY720 on the rst day at 5 mg/day oral for 7 consecutive days (Koyrakh et al. 2005). Another experiment reported that FTY720 at 2.5 mg/kg/day with intraperitoneal injection (ip) for more than two weeks induced a moderate lymphopenia of nude mice (Pchejetski et al. 2010). In addition, weight loss or in ammation was not observed in healthy volunteers (5 mg/day for one week, oral) or in animals treated with FTY720 at 2.5 mg/kg/day (Kovarik et al. 2004).
Phosphorylated FTY720 inhibit lymphocytes egress of lymphoid tissues, acted as a potential immunosuppressant. However, three main ways of FTY720 in tumor supress were reported. Huwiler et al. and White C. et al. studies suggested that non-phosphorylated FTY720 displayed anti-cancer properties mainly via a SphK1 downregulation cascade reaction, inducing tumor cell apoptosis, inhibition of tumor cell proliferation, migration and tumor vessel density (Huwiler and Zangemeister-Wittke 2018;White et al. 2016). FTY720 also can be phosphated by SK2, and FTY720-P induces internalization of S1PR1 as an antagoist and exerts function of tumor suppress (Pyne and Pyne 2010). Non-phosphated FTY720 can target at non-receptor molecular such as PP2A to achieve its anti-cancer role (Perrotti and Neviani 2013). More interestingly, single FTY720 or combined with ge tinib treatment did not suppress the immune system of mice and conversely improved treatment in breast cancer (Martin et al. 2017). The anticancer effect of FTY720 in patients has rarely been reported. It is worth evaluating its potential anticancer function and the clinical safe dosage of oral FTY720. Therefore, we built this systematic review to assess how effective FTY720 is in treating diverse types of cancer and its dosage evaluation.

A survey of references and Selection criteria
Our References between 1998 and 2020 came from PubMed, Ovid, Embase, Web of Science, CNKI and Cochrane databases. These references were no a language limit. This search used these mesh terms: " ngolimod" OR "FTY720" AND/OR "animals" AND/OR "cancer" in PubMed. Keyword "FTY720 cancer" was used in Ovid, Web of Science and Embase and "FTY720" in CNKI and the Cochrane library. We reviewed the titles, abstracts, main text and even references identi ed in the search and built the exclusion standard, including (a) reviews, letters, abstracts, editorials and expert opinions; (b) articles without data; (c) repetitive or similar studies; and (d) nonanimal based studies. Additionally, we set inclusion parameters, such as: (1) animals: experimental mouse without complication; (2) FTY720 only and in combination; (3) outcomes measures (at the end of experiment or the day of death of all animals in a group): tumor volume, tumor weight and body weight were evaluated. When different doses were used in the same experiment, the highest dose of FTY720 was adopted.

Meta-analysis data processing
Data extraction contained some routine parameter set, such as rst author, year, country, tumor type, animals' strain and age, dosage, frequency, duration, cell lines and numbers, location, administration and therapeutics outcomes (tumor volume, tumor weight) and side effect evaluation (body weight). We extracted the data by using standard form. Values for data were expressed in graph only, values were read off the graphics by using CorelDraw 2019 software.
When a control group shared more than one experiment group, we adjusted the numbers of animals in control group according to a practical guide on PROSPERO. When there was a lack of valid data in a study, review authors acquired the original data from the articles' author by E-mail or phone.

Assessment of data quality and risk of bias
The SYRCLE's risk for bias tool was utilized to assess the preclinical animal studies quality and evaluate the risk bias of individual animal model studies (Hooijmans et al. 2014). Considering for publication bias and selective results among studies, subgroup analysis, funnel plot analysis, and sensitive analysis were carried out. We evaluated the factors of bias for each study individually in this process.

Statistical analysis
Our manuscript utilized the STATA software (Stata Corporation, College Station, TX, version 14.0). Tumor volume and weight were taken as an evaluation of the anticancer e cacy of FTY720, conversely body weight as an index of toxicity. The mean value standard difference (SD), standard mean difference (SMD) and 95% con dence interval (CI) in each group were collected. I² test illustrate the heterogeneity among studies and a funnel plot evaluated the possible publication bias. More accurately, Eggers's regression was nished following funnel plots. In our article, P < 0.05 was set as statistically signi cant by two-sided test.
Heterogeneity in over all types of cancer was I 2 = 91.1%, P = 0.000. No or low heterogeneity which were found in renal cell carcinoma (I² = 0.0%, P = 0.424) and neuroblastoma (I² = 46.8%, P = 0.153) revealed that FTY720 can signi cantly reduce tumor volume in these two cancers. Whereas I 2 of some cancers were missed due to there only being one study, including pancreatic cancer, ovarian cancer, mesothelioma, adrenocortical carcinoma, glioblastoma and acute myelogenous leukemia, many more studies should be carried out to verify the antitumor effect of FTY720 in these 6 types of cancer. Obvious heterogeneity existed in breast cancer (I² = 90.7%, P = 0.000), hepatocellular cancer (I² = 70.2%, P = 0.009), and prostate cancer (I² = 85.6%, P = 0.001). Hence, 7 potential heterogeneity factors were utilized to make a subgroup meta-analysis, including cell lines derivation, strain of animal, immunology function of animal model, administration, tumor location, three dosage range of FTY720 (≤25 mg/kg/w, 25 mg/kg/w~50 mg/kg/w, and ≥50 mg/kg/w) and animal age (Table S3). The results showed that cell lines derivation and animal age should be vital factors of heterogeneity in breast cancer mouse models.
Transplant tumor location is an obvious factor of heterogeneity in hepatocellular carcinoma mouse models. Due to the limited study number, we failed in nding the heterogeneous factors in prostate cancer mouse models.
More deeply, we summarized the toxicity reaction of FTY720 alone or in combination from every original study in Table S4. 16 of the total 31 articles related to side effect of FTY720, and showed no obvious side effects in mice and rats during their experiment (Alshaker et al. 2017;Chua et al. 2005;Gstalder et al. 2016;Hait et al. 2015;Ho et al. 2005;Katsuta et al. 2017;Lee et al. 2004;Lee et al. 2005;Leu et al. 2016;Li et al. 2017;Lifshitz et al. 2017;Martin et al. 2017;Pchejetski et al. 2010;Rosa et al. 2013;Szymiczek et al. 2017;Xu et al. 2015). The remaining 15 articles didn't document explicitly for the potential side effects of FTY720. Co-treatment with ge tinib, doxorubicin and docetaxel showed lower toxicity than by themselves.

Publication bias and sensitivity analysis
First, SYRCLE's tool from PRPSPERO is set to evaluate quality of studies, detail data was listed in Table S2. Then, we used funnel plots to access publication for meta-analysis and tested by Egger's regression for tumor volume only, due to the limitation of the number of studies. It suggested that the majority of researches were approximately symmetrical (Fig. 4), and Egger's regression for all types of cancer in tumor volume (P = 0.235), tumor weight (P = 0.018) and body weight (P = 0.172) suggested that the publication bias were not obvious in tumor volume and body weight, but in tumor weight. For the sensitivity analysis, Fig. 5 showed that the pooled SMD in the rest of the studies were still signi cant although one individual research was deleted once.
Signaling pathway related with FTY720 anticancer mechanism Finally, the anticancer related molecular signals of FYP720 alone or in combination with chemical drug in total 26 articles were summarized in Table 1. The main molecular mechanism was focused on apoptosis which was discussed by 20 articles, including non-small lung cancer, pancreatic cancer, mesothelioma, breast cancer, renal cell carcinoma, colorectal cancer, adrenocortical carcinoma, acute myeloid leukemia, neuroblastoma, glioblastoma, prostate cancer, and hepatocellular carcinoma ( arrest (androgen-independent prostate cancer, acute myeloid leukemia, hepatocellular carcinoma) (Chua et al. 2005;Ho et al. 2005;Lee et al. 2004;Lee et al. 2005), four involved angiogenesis (hepatocellular carcinoma, Lewis lung cancer, androgen-independent prostate cancer, breast cancer) (Chua et al. 2005;Ho et al. 2005;Mousseau et al. 2012;Schmid et al. 2007), two involved invasion (non-small lung cancer, adrenocortical carcinoma) (Liu et al. 2015;Xu et al. 2015), one involved histone acetylation (breast cancer) (Hait et al. 2015) and one involved autophagy pathways (lung cancer) (Li et al. 2018) respectively. The frequent targeted molecules were SphK1, PP2A, S1PR1, STAT3, VEGF, EGFR, caspases. This evidence suggests that the sphingosine signaling pathway is one of importance in various types of cancer and mainly induces apoptosis of tumor cells after interfering it.
Next, our discussion focused on the side effects of FTY720 alone or in combination at a potent pharmaceutical dose. Only eight experiments of 31 articles discussed issues regarding reducing experimental animal body weight of FTY720 at the dosage range from 3 mg/kg, 2 weeks, biweekly to 10mg/kg /day, and only one study in prostate cancer reported a signi cant weight loss (SMD = -1.45, CI: -2.46, -0.44, Z = 2.81, P = 0.005) with low heterogeneity (I 2 = 0.0%, P = 0.920). Although total pooled values (SMD = -0.86, 95% CI: -1.61, -0.11, Z = 2.23, P = 0.025) suggested that there is a tendency in reducing body weight in the included eight experiments, therefore conclusions cannot be drawn in these studies, such as adrenocortical carcinoma, breast cancer, and mesothelioma due to the low number of studies. Additionally, because of the different combinations, we adopted the method of single literature summary to discuss the toxicity reaction of FTY720 in combination or alone in Table S4. 16 out of 29 articles did not suggest any side effects in mice and rats and conversely increased anticancer e cacy, especially co-treatment with ge tinib, doxorubicin and docetaxel. Actually, there exists some controversial conclusions about FTY720 side effects. For example, moderate lymphopenia was observed in prostate cancer models treated with FTY720 at the dose of 2.5 mg/kg/day (Pchejetski et al. 2011). Additionally, healthy volunteers treated by 5 mg/day FTY720 oral for one week showed no toxic effects, however FTY720 might cause skin cancers which was reported in clinical therapy (Michiels et al. 2019;Robinson and Guo 2016). Hence, the dosage-safety of FTY720 in clinical use is still a noteworthy issue. Another controversial results of animals immune system condition suggested signi cant anticancer effect in immunocompromised mice (SMD = -3.35, 95% CI: -4.17, -2.53, Z = 8.02, P = 0.000, I² = 87.7%, p = 0.000 for tumor volume), not in immunocompetent mice (SMD = -0.67, 95% CI: -2.80, 1.46, Z = 0.62, P = 0.537, I² = 94.5%, p = 0.000 for tumor volume) (Fig S1). The author analyze that This result shows FTY720 may have side effect in immune system against its anti-cancer function. Only one study paradoxically investigated more e cacy of FTY720 on immunocompetent mice than on immunocompromised mice, due to intratumoral T cells activiation (Martin et al. 2017). Con icting research demonstrated FTY720 in combination showed better therapeutic e cacy and did not inhibit the function of the immune system.
At last, to further understand the signi cance of sphingolipids, especially S1P in diverse cancers, we analyzed the molecular targets of FTY720 alone or in combination in detail in Table 1. 20 out of 31 articles explained that FTY720-induced cell cycle arrest of tumor cells and apoptosis are one of the main anticancer mechanisms in these cancer types, as is non-small lung cancer, pancreatic cancer, mesothelioma, breast cancer, renal cell carcinoma, colorectal cancer, adrenocortical carcinoma, acute myeloid leukemia, neuroblastoma, glioblastoma, prostate cancer, and hepatocellular carcinoma. These classic signal cascades mediated apoptosis including SphK1/S1PR1/STAT3/Ki67 or CycD1 or β-catenin or c-Myc or PP2A; FTY720-P/S1PR1/JAK2/STAT3/caspase-3 or PARP-1 or Bcl-2; SphK2/Akt or BAD /Cyt c; S1PR1/ERK; PI3K/Akt/caspases or MAPK/caspase cascades, et al, suggesting the close association between S1PR1 and important molecules such as Akt, STAT3, VEGF, EGFR, SphK 1, SphK 2, PP2A et al.

Conclusion
According to the data meta-analysis, we suggest that FTY720 alone or in combination may be a candidate for the treatment of certain cancers, especially for neuroblastoma and renal cell carcinoma, however not for melanoma.
Considering higher dosage of FTY720 in animal experiments and a few case reports of the side effects in MS, we recommend that more speci c studies of FTY720 only and in combination focus on immunity, in ammation and melanoma be carried out in the future preclinical studies.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article.
Competing Interests: The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper.
Funding This work was supported by the National Natural Science Foundation of China (grant no. 81672834). The funder had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.