Association between Blood 25-Hydroxyvitamin D Levels and Survival in Colorectal Cancer Patients: An Updated Systematic Review and Meta-Analysis

Previous meta-analyses have shown an improved survival with higher blood 25-hydroxyvitamin D (25(OH)D) concentrations in patients with colorectal cancer (CRC). However, a number of much larger studies have been published since then. We provide an updated meta-analysis to synthesize current evidence. PubMed and Web of Science databases were systematically searched for eligible studies. The dose-response relationships and pooled hazard ratios for overall and CRC-specific survival comparing the highest versus the lowest categories of blood 25(OH)D concentrations were assessed. Subgroup analyses based on study geographic location, year of publication, sample size, length of follow-up time and stage were conducted to explore potential sources of heterogeneity. Overall, 11 original studies with a total of 7718 CRC patients were included. The dose-response meta-analysis showed an improvement in survival outcomes with increasing blood 25(OH)D concentrations. Pooled hazard ratios (95% confidence intervals) comparing highest versus lowest categories were 0.68 (0.55–0.85) and 0.67 (0.57–0.78) for overall and CRC-specific survival, respectively. Associations were more prominent among studies conducted in Europe, with larger sample sizes, and including stage I–IV patients. This updated meta-analysis reveals robust evidence of an association between higher blood 25(OH)D concentrations and better survival in CRC patients. The potential for enhancing prognosis of CRC patients by vitamin D supplementation should be explored by randomized trials.


Search strategy
The reporting of meta-analyses of observational studies in epidemiology (MOOSE) guidelines were followed to perform this systematic review and meta-analysis (14) (Table A1). We carried out a systematic literature search in PubMed and Web of Science databases for articles reporting results of cohort studies conducted in CRC patients and assessing the association between blood 25(OH)D concentrations and overall and CRC-specific survival, using a comprehensive list of search terms (Table A2). The current literature search was restricted to articles published from 2013 until September 2017 with no language restrictions, thereby complementing our previous corresponding literature search of articles published up to 2013 (10).

Problem definition
To date, a number of reviews and meta-analyzes summarized results from studies that investigated the association between serum 25(OH)D levels and survival in patients with CRC. However, due to the small number of included studies, none of the previous reports have explored this association within subgroups. The endnote X7 software was used throughout the literature search process.
Use of hand searching (e.g. reference lists of obtained articles) Bibliographies of the retrieved papers (only the included studies) were hand searched for additional references List of citations located and those excluded, including justification Details of the literature search process are outlined in the flow chart ( Figure 1). The citation list of excluded articles is available in supplementary C.

Method of addressing articles published in languages other than English
We placed no restrictions on language; We were able to obtained all articles potentially eligible for inclusion in English language Method of handling abstracts and unpublished studies we excluded conference abstracts due to the insufficient data obtained from them. Description of any contact with authors none

Reporting of methods should include
Description of relevance or appropriateness of studies assembled for assessing the hypothesis to be tested We included articles reporting results of cohort studies conducted in CRC patients and assessing the association between serum 25(OH)D levels with overall and CRCspecific mortality. We excluded studies with (i) no longitudinal design; (ii) no CRC patients; (iii) no measurement of serum 25(OH)D and (iv) no reported measurement of the association between the exposure and the outcome of interest.
Rationale for the selection and coding of data (e.g. sound clinical principles or convenience) We used a standardized data extraction form to record study characteristics, information on study population as well as data about the time between serum 25(OH)D measurement and cancer diagnosis and the value of 25(OH)D concentrations presenting the mid-points of the ranges of the reported categories. Assessment of confounding (e.g., comparability of cases and controls in studies where appropriate) We conducted 5 subgroup analyses to explore the source of the heterogeneity across the studies Assessment of study quality, including blinding of quality assessors; stratification or regression on possible predictors of study results We assumed that studies that didn't adjust for at age, sex, and season as studies with low quality, and we assessed a sensitivity analysis to test the stability of the pooled estimates after their exclusion.

Assessment of heterogeneity
The heterogeneity among the included studies was investigated using I 2 and Q test statistics, with a significant evidence of heterogeneity for I 2 > 50% or a P value less than 0.05 from the Q-test Description of statistical methods (e.g. complete description of fixed or random effects models, justification of whether the chosen models account for predictors of study results, doseresponse models, or cumulative metaanalysis) in sufficient detail to be replicated We described the method used for each analysis and the software used.

Provision of appropriate tables and graphics
Flow chart: to show the method of studies identification. Table 1: showing characteristics of included studies Table 2: showing Results of Subgroup analysis Figure 2,3: forest plots of the main two meta-analyses conducted Figure 4: Hazards ratios and 95% confidence intervals for overall survival in patients with colorectal cancer according to circulating 25-hydroxyvitamin D (25(OH)D) serum concentrations. Figure 5: Pooled dose-response relationship between serum 25(OH)D levels and the two outcomes investigated.

Reporting of results should include
Graphic summarizing individual study estimates and overall estimate Figure 2, Figure 3 Table giving descriptive information for each study included Table 1 Results of sensitivity testing (e.g. subgroup analysis) Results section/sensitivity analysis sub-section Indication of statistical uncertainty of findings 95% CI intervals were presented for all analyses together with I 2 values for the two main meta-analyses Reporting of discussion should include Quantitative assessment of bias (e.g., publication bias)

Results of Funnel plot and risk of publication bias is highlighted in the results section and supplementary D. Justification for exclusion (e.g. exclusion of non-English-language citations)
Reasons for exclusion were reported mainly in the methods section

Assessment of quality of included studies
Study quality was judged based on the adjustment level. We considered age, sex, season and stage as important confounders. Implication of the subgroup analyses related to quality assessment was highlighted

Consideration of alternative explanations for observed results
Evidence from biological studies supports the role of vitamin D in cancer. vitamin D regulates the transcription of genes involved in the Inhibition of proliferation/angiogenesis and the transcription of genes responsible for the inducement of differentiation, apoptosis, and DNA repair mechanisms. Furthermore, several inflammatory processes and cytokines such as interleukin 6 and (IL6), IL8 and IL17 involved in CRC progression are regulated by vitamin D. Generalization of the conclusions (i.e. appropriate for the data presented and within the domain of the literature review) If a causal relationship between vitamin D status and CRC prognosis is demonstrated, the objective from using vitamin D as a potential therapeutic option for CRC patients may be related to socio-demographic factors.
Guidelines for future research Due to the low costs of mendelian randomization studies compared to RCTs, such method is warranted to test causality between vitamin D status and survival in patients with colorectal cancer. Disclosure of funding source none

15 and 19 and 23
List of excluded studies