Search tips
Search criteria 


Logo of jcoHomeThis ArticleSearchSubmitASCO JCO Homepage
J Clin Oncol. 2009 November 10; 27(32): 5439–5444.
Published online 2009 September 21. doi:  10.1200/JCO.2009.22.1135
PMCID: PMC2773226

Serum 25-Hydroxyvitamin D3 Levels Are Associated With Breslow Thickness at Presentation and Survival From Melanoma



A cohort study was carried out to test the hypothesis that higher vitamin D levels reduce the risk of relapse from melanoma.


A pilot retrospective study of 271 patients with melanoma suggested that vitamin D may protect against recurrence of melanoma. We tested these findings in a survival analysis in a cohort of 872 patients recruited to the Leeds Melanoma Cohort (median follow-up, 4.7 years).


In the retrospective study, self-reports of taking vitamin D supplements were nonsignificantly correlated with a reduced risk of melanoma relapse (odds ratio = 0.6; 95% CI, 0.4 to 1.1; P = .09). Nonrelapsers had higher mean 25-hydroxyvitamin D3 levels than relapsers (49 v 46 nmol/L; P = .3; not statistically significant). In the cohort (prospective) study, higher 25-hydroxyvitamin D3 levels were associated with lower Breslow thickness at diagnosis (P = .002) and were independently protective of relapse and death: the hazard ratio for relapse-free survival (RFS) was 0.79 (95% CI, 0.64 to 0.96; P = .01) for a 20 nmol/L increase in serum level. There was evidence of interaction between the vitamin D receptor (VDR) BsmI genotype and serum 25-hydroxyvitamin D3 levels on RFS.


Results from the retrospective study were consistent with a role for vitamin D in melanoma outcome. The cohort study tests this hypothesis, providing evidence that higher 25-hydroxyvitamin D3 levels, at diagnosis, are associated with both thinner tumors and better survival from melanoma, independent of Breslow thickness. Patients with melanoma, and those at high risk of melanoma, should seek to ensure vitamin D sufficiency. Additional studies are needed to establish optimal serum levels for patients with melanoma.


Vitamin D3 has pleiotropic effects, via the vitamin D receptor (VDR), of relevance to cancer. These include effects on cell growth, differentiation, and apoptosis1 and the regulation of tumor/immune system interaction.2 Vitamin D3 is synthesized in the skin as a result of exposure to sunlight and is present in a few foods and dietary supplements. Suboptimal 25-hydroxyvitamin D3 levels are widespread,3 and a role for vitamin D deficiency has been documented for a number of cancers.47

Known predictors of survival in patients with melanoma relate to histologic characteristics of the tumor,8 particularly the Breslow thickness9 (the distance from the granular cell layer of the skin to the deepest part of the tumor in millimeters), ulceration,10 age,11 tumor site, and sex,12 and to social deprivation.13,14 To identify other environmental predictors, we adopted a two-stage approach, involving a comparison of relapsed and nonrelapsed patients, then took a suggestive lead into a cohort study (hypothesis testing stage). For the initial retrospective study, we reasoned that by recruiting patients 3 or more years beyond initial diagnosis, we would remove some of the dominant effects attributable to the primary tumor, allowing subsequent environmental exposures to become more apparent.

In the pilot retrospective study comparing late relapsers and nonrelapsers, although not statistically significant, the analyses were consistent with a role for vitamin D in relapse. We have subsequently tested this in an independent prospective study in which lower levels of 25-hydroxyvitamin D3 were associated both with thicker, poorer-prognosis melanomas and with poorer outcome within Breslow thickness categories.


Studies were approved by the Multi-Centre Research Ethics and the Patient Information Advisory Groups.

Pilot Retrospective Study

Melanoma cases had all been diagnosed at least 3 years previously without relapse. Patients were then eligible as “relapsers” if relapse occurred or as “nonrelapsers” if they were free of relapse.15 Participants had a tumor Breslow thickness greater than 0.75 mm and were recruited between May 2000 and January 2005. Relapsers and nonrelapsers were frequency-matched for age, sex, and Breslow thickness. Participants were asked to complete a questionnaire (data collected on regular use of vitamins, minerals, fish oils, fiber, or other food supplements 1 year before interview) and give a single blood sample for DNA and for serum cryopreservation. In the United Kingdom, multivitamins usually contain 5 μg of vitamin D3 (range, 2.5 to 10 μg), and most patients who took fish oils took them as cod liver oil containing similar quantities of vitamin D3. These were therefore grouped in the analysis, which was based on use of any vitamin D supplementation compared with none.

Recruitment to the Prospective Cohort Study

The patients were population-ascertained incident patients with melanoma (stages I to IIIA) recruited in a geographically defined area of northern England from September 2000 through March 2008. Participants completed a questionnaire containing questions on drug intake and use of dietary supplements. Serum (single sample), plasma, lymphocytes, and DNA were stored. Recruitment (and therefore blood sampling) took place wherever possible within a period 3 to 6 months after diagnosis. Histopathology details of the primary tumor were extracted from pathology reports. Relapse/survival data were obtained from the patient by annual patient questionnaire, from the cancer registry, and by examination of clinical notes. Participants self-reported height and weight, and this was used to estimate the body mass index (BMI), as serum vitamin D levels vary with proportion of body fat, and vitamin D has complex metabolic interactions with fat cells.16,17

Serum 25-Hydroxyvitamin D2 and D3 Measurement

Concentrations of 25-hydroxyvitamin D2 and D3 (nmol/L) were measured in 100 μL of cryopreserved serum by liquid chromatography tandem mass spectrometry (Appendix, online only). D2 levels were almost uniformly undetectable (< 10 nmol/L), and D2 and D3 levels were therefore summed. We shall refer to this henceforth as serum vitamin D levels.

Single Nucleotide Polymorphism Methodology

Known and putative functional polymorphisms in VDR were genotyped (Appendix, online only).

Statistical Methodology

Pilot retrospective study.

Two-sided t tests were used for comparing continuous measures between relapsers and nonrelapsers. Associations between categoric variables were tested for using Pearson χ2 tests. Unconditional logistic regression models were used to estimate the odds ratio (OR) and 95% CI for the effect of self-reported vitamin D supplementation and VDR genotypes on risk of relapse. These analyses were carried out using STATA version 9 (Stata Statistical Software, Release 9, 2005; StataCorp, College Station, TX).

Prospective cohort study.

Adjusted means (least squares means in SAS) of serum vitamin D levels corrected for sex, age, BMI, and month sampled were calculated for different Breslow thickness groups using a general linear model. Relapse-free survival (RFS) and overall survival (OS) analyses were performed using Cox proportional hazards models to examine the associations between self-reported vitamin D supplement intake, measured serum vitamin D levels, and VDR status and risk of relapse and death from melanoma. Vitamin D supplement intake was categorized as a dichotomous variable according to whether the patient reported regularly taking multivitamins or cod liver oil a year before interview. Serum vitamin D levels were treated as a continuous variable in the analyses. In addition, serum vitamin D levels were categorized into tertiles for Kaplan-Meier survival analysis. Serum vitamin D levels were dichotomized using median levels as a cutoff point to investigate interactions with VDR genotype (coded as number of variant allele [0, 1, 2] in the additive models or presence of variant allele [0, 1] in the dominant models) in the survival analyses. Interaction was tested by comparing the model with main effects and an interaction term with a reduced model with only main effects, using the likelihood ratio test. Hazard ratios (HR) and their 95% CIs are presented. These analyses were carried out using the SAS/STAT statistical software version 9.1 for PC (SAS Institute, Cary, NC).


Pilot Retrospective Study

Overall, 131 (79%) of 165 eligible patients participated as relapsers, for whom 91 patients had complete supplement history, serum vitamin D level, and VDR data. We recruited 169 nonrelapsers (62% participation proportion) cared for by the same clinicians as the relapsers; 149 patients (88%) had complete data. Median time from diagnosis to relapse/interview for the relapsers was 6.6 years (range, 3.1 to 28.1 years) and for nonrelapsers was 7.4 years (range, 3.2 to 31.7 years). There was no effect of season of surgery on case-control status.

A total of 47% (70 of 149) of nonrelapsers and 38% (35 of 91) of relapsers reported regular use of any supplements before relapse (OR = 0.7; 95% CI, 0.4 to 1.2). Sixty-two (42%) of 149 nonrelapsers and 28 (31%) of 91 relapsers reported regular intake of supplemental vitamin D 1 year before interview (OR = 0.6; 95% CI, 0.4 to 1.1; P = .09; Table 1). Thirty-seven percent of both female and male patients took vitamin D supplements (χ21 = 0.005, P = .9). There was no association between relapse and the postal code–derived Townsend score,18 a measure of deprivation (t test, P = .5), or BMI.15

Table 1.
Distribution of Dietary Supplement Use by Case-Control Status in the Pilot Retrospective Study

There was no effect on vitamin D serum levels of duration of sample storage or of BMI (data not shown). Measured serum vitamin D levels were higher in patients who reported taking vitamin D supplements (mean, 54 nmol/L; 95% CI, 51 to 58 nmol/L) compared with those not taking supplements (mean, 43 nmol/L; 95% CI, 40 to 47 nmol/L; t test, P = .0001). Nonrelapsers had higher levels (mean, 49 nmol/L; 95% CI, 45 to 52 nmol/L) than relapsers (mean, 46 nmol/L; 95% CI, 41 to 50 nmol/L; P = .3), although the difference was not statistically significant (t test, P = .3).

There was marginal evidence of a main effect of VDR alone irrespective of supplementation, as heterozygosity for both Cdx2 and FokI was inversely associated with relapse (OR = 0.5, 95% CI, 0.3 to 1.0, and OR = 0.5, 95% CI, 0.3 to 1.0, respectively) compared with homozygotes for the ancestral allele. However, the effect was not seen for rare homozygotes. None of the other single nucleotide polymorphisms showed a genotype difference by relapse status (data not shown). The sample size was too limited to investigate the joint associations with VDR and serum vitamin D levels on relapse.

Prospective Cohort Study Analysis

A total of 1,132 patients with melanoma were recruited (63% participation proportion). The focus of this analysis is on the 872 participants with a tumor thickness in excess of 0.75 mm recruited before October 2006 (median follow-up, 4.7 years). There had been 173 relapses and 141 deaths by November 2008.

Serum vitamin D levels varied with season as expected (Fig 1), and in the majority of patients with melanoma, the levels were suboptimal (64%), particularly in the months of November through June, taking a conservative estimate of 60 nmol/L as optimal. Parathyroid hormone levels are reported to increase at a level of 75 nmol/L,19 so we took the 60 nmol/L level to be conservative. There was no effect of season of surgery on OS in an unadjusted comparison of winter compared with summer (P = .62). Serum vitamin D levels are known to vary with age and BMI, and in this study, levels were shown to be lower in younger individuals (test for trend, P = .001, adjusted for sex, month of venipuncture, and BMI) and those with higher BMI (test for trend, P = .005, adjusted for age, sex, and month). The mean serum vitamin D levels ranged from 54 nmol/L (95% CI, 51 to 56 nmol/L) for participants with a BMI less than 24.9, to 55 nmol/L (95% CI, 53 to 57 nmol/L) for BMI in the range 24.9 to 29.9, and 48 nmol/L (95% CI, 45 to 51 nmol/L) for BMI more than 29.9. Participants who reported use of vitamin D supplements had significantly higher serum vitamin D levels: 540 participants who were not taking supplements had a mean level of 50 nmol/L (95% CI, 48 to 52 nmol/L) compared with a mean of 60 nmol/L (95% CI, 57 to 63 nmol/L) for the 247 patients who reported taking supplements. The Spearman correlation between Breslow thickness and serum vitamin D levels was −0.10 (P = .001). Increased Breslow thickness was associated with lower serum vitamin D levels (test for linear trend P = .002, Table 2).

Fig 1.
Variation in serum vitamin D levels by season of venipuncture in the prospective cohort study cases (number of patients in parentheses). A conservative optimal level of 60 nmol/L is indicated by the gray line, and a higher optimal level based on stabilization ...
Table 2.
Association Between Serum Vitamin D Levels at Recruitment and Breslow Thickness in the Prospective Cohort Study

In univariate survival analysis (Appendix, online only), later age was associated with poorer survival (HR per year of age for RFS was 1.03; 95% CI, 1.01 to 1.04), and RFS was lower for male patients, with an HR of 1.85 (95% CI, 1.37 to 2.49; Appendix Table A2, online only). Patients with limb tumors were at reduced risk of relapse compared with those with tumors on the trunk (HR = 0.65; 95% CI, 0.46 to 0.93). Increases of 20 nmol/L in serum vitamin D levels were associated with a reduced risk of relapse (HR = 0.75; 95% CI, 0.64 to 0.90) and OS (HR = 0.80; 95% CI, 0.68 to 0.96) consistently across seasons (Appendix Table A2 and Fig 2). Reported use of vitamin D supplementation showed no statistically significant effect on outcome in univariate (RFS: HR = 0.81, 95% CI, 0.56 to 1.17; OS: HR = 0.71, 95% CI, 0.47 to 1.09; Appendix Tables A2) or in multivariate analyses (data not shown), being statistically consistent with both no effect and the estimate for the initial study.

Fig 2.
Kaplan-Meier curves of different serum vitamin D levels at interview (categorized based on tertile cutoff points) on relapse-free survival from melanoma in the cohort study.

Table 3 shows the results of multivariate analyses. After adjustment for age, sex, Townsend score, tumor site, Breslow thickness, and BMI, higher serum vitamin D levels were still seen to be protective for RFS and, to a lesser extent, for OS, consistently across the seasons (Table 3). The adjusted HRs for RFS and OS across seasons were 0.79 (95% CI, 0.64 to 0.96) and 0.83 (95% CI, 0.68 to 1.02), respectively, per 20 nmol/L increase in serum vitamin D levels. The inverse association with higher serum vitamin D levels, stratified by season, persisted through the range of levels detected based on tertile analysis. The HRs for RFS were 0.70 (95% CI, 0.42 to 1.14) and 0.57 (95% CI, 0.33 to 0.97) for the upper two tertiles of 41.3 to 61.4 nmol/L and more than 61.4 nmol/L, respectively, compared with the lowest tertile of ≤ 41.3 nmol/L group. Thus there was no evidence of an adverse effect of the highest observed serum levels on outcome.

Table 3.
Risk Factors for Relapse From Melanoma and Overall Death in the Multivariable Analysis From the Prospective Cohort Study

There was no clear effect of VDR genotype on outcome in the cohort study, apart from weak evidence that inheritance of the BsmI A allele was adversely associated with RFS. Assuming a dominant model, carriage of the A allele increased risk of relapse (HR = 1.44; 95% CI, 1.02 to 2.03), but there was no evidence of association using a general 2 df model (P = .12, Appendix Table A2).

An interaction between serum vitamin D levels and VDR single nucleotide polymorphisms on outcome was investigated. Inheritance of the BsmI A allele is associated with poorer outcome from melanoma in those with low vitamin D levels, but not in those with high vitamin D levels (P = .02 for interaction, Table 4).

Table 4.
Interaction of VDR With Serum Levels of Vitamin D on Relapse From Melanoma, Under Different Inheritance Modes in the Cohort Study


The pilot retrospective study was consistent with the hypothesis that vitamin D reduces the risk of relapse from melanoma. The subsequent cohort study tests and supports this hypothesis.

Most human diets provide little vitamin D, as it is primarily found in fatty fish. In some countries, foods are supplemented with the D2 form; dietary supplements usually contain the more potent D3 form.20 Most people in Europe are reliant on cutaneous synthesis as a result of sun exposure for the majority of their vitamin D, because supplementation of foods is infrequent, and fatty fish does not generally form a large part of the diet. There remains some controversy around ideal levels of vitamin D; in a study of Swedish women aged 75 years or more, vitamin D levels less than 50 nmol/L doubled the fracture risk.21 It is of note that the mean level among supplement takers in both the pilot and cohort studies was above this value, and the mean for those not taking supplements was lower. A higher level of 75 to 90 nmol/L is generally now thought to be optimal, because at this level, parathyroid hormone levels reach a nadir.20 That patients with melanoma in the United Kingdom have low levels is not surprising, as suboptimal levels of vitamin D have been reported in many “healthy” populations,20,22 and patients with melanoma often avoid sun exposure.

The difficulty of designing and implementing prospective studies for melanoma, which could address multiple hypotheses relating exposure to prognosis, prompted the novel retrospective study design in which only patients who had survived at least 3 years were recruited. Although it was itself logistically challenging, the comparatively short time frame of a retrospective study permits prioritization and focus of hypotheses, enabling longer-term cohort studies to be instigated. A major problem associated with interpreting retrospective studies involves recall bias, and in this study, this might potentially relate to reporting of vitamin D supplementation. However, serum vitamin D levels corroborated the history given by the participants, in that supplement takers had significantly higher levels, reducing this concern. The other major problem with the interpretation is confounding, such as from socioeconomic status, as deprivation is potentially related to both prognosis and prevalence of supplement taking; in this analysis, taking vitamin D did not vary by socioeconomic status, as indicated using the Townsend score. Supplement takers ate more fruit and less meat, but there was no effect of these dietary factors on relapse.15 Other possible sources of bias in a retrospective study are acknowledged, such as changed behavior after relapse and bias of ascertainment. Furthermore, it is possible that other unmeasured exposures, such as concurrent drug therapies, may also have an effect. Many of these concerns are removed in a prospective cohort study, which we then went on to carry out.

The vitamin D levels were higher in nonrelapsers from the pilot retrospective study than in relapsers although the difference was not statistically significant (which may be a consequence of limited power).

In the subsequent prospective cohort study, there was a strong inverse correlation between serum vitamin D levels and Breslow thickness. This relationship persisted when the data were corrected for factors known to be associated with lower vitamin D levels, such as age, although it is possible that the relationship is due to an unmeasured confounder. Serum vitamin D levels and BMI proved independently predictive of relapse even when corrected for Breslow thickness, supportive of the view that tumor progression after removal of the primary is also related to serum vitamin D levels.

Several previous case-control studies have reported an association between BMI and melanoma risk,2326 but not all.27 There are no previously published data suggesting a poorer outcome from melanoma for the obese, but there are for other cancers.28,29 It is interesting that better survival occurred in patients with an intermediate BMI (between 24.9 and 29.9), in which group serum vitamin D levels were highest.

Although analysis of the cohort study data provides strong evidence to support the hypothesis from the initial retrospective study, the possibility must be considered that serum vitamin D levels might merely be a marker for another causal relationship. Although vitamin D levels and prognosis might be confounded by some unmeasured exposure, the principle of Mendelian Randomization indicates that genotype assortment cannot be due to confounding. Thus the fact that the BsmI VDR variant A allele showed evidence of an interaction with serum vitamin D levels in the cohort study supports the suggestion that the association between vitamin D levels and prognosis is not due to confounding.

The level of serum vitamin D that would be most beneficial for patients with cancer is not known. Although there are some data that suggest that very high levels might be harmful as a result of suppression of dendritic cell function,30 it has been suggested that levels of 75 nmol/L are desirable.20 The cohort study provides no strong data about optimal levels, but there was no evidence for anything other than a beneficial effect for higher levels.

Sun avoidance after a diagnosis of melanoma is very common, and this study suggests that patients with melanoma have suboptimal levels of serum vitamin D. Protection from sunburn is important and should be continued, but efforts should be made to ensure vitamin D sufficiency, and these data suggest that this can be achieved using supplements. We have shown that vitamin D deficiency is common in patients with melanoma in the United Kingdom and that this may increase the risk of relapse from melanoma. This study suggests that, in the absence of medical contraindications, patients with melanoma who are avoiding sun exposure should take vitamin D3 supplements sufficient to ensure normal levels. More research is needed to identify optimal levels for patients with melanoma.


We thank Julian H. Barth and Helen P. Field from the Department of Clinical Biochemistry at Leeds Teaching Hospitals Trust for carrying out the measurement of serum vitamin D. We also thank Patricia Mack, Elaine Fitzgibbon, Linda Bale, and Kate Gamble for collecting data for the studies.


The following recruited patients to the studies: Mr J. Ausobsky, Yorkshire Clinic; Dr A. Carmichael, Mr M. Coady, Dr S. Shehade, Mr H. Siddiqui, Mr K. Allison, Mr K. Erdinger, Mr Ramanathan, Mr Toby Muir, James Cook University Hospital; Dr A. S. Highet, Mr K. R. Mannur, Mr M. Telfer, Dr K. Thomson, Mr G. Miller, Mr J. M. Hayward, Mr J. Taylor, Mr A. Coatesworth, Dr A. E. Myatt, Dr J. Schofield, Dr Callum Lyon, York District Hospital & Scarborough Hospital; Dr Alison Layton, Harrogate District Hospital; Dr Anthony Maraveyas, Dr S. Walton, Dr N. Alexander, Mr Alistair Platt, Mr N. B. Hart, Mr P. M O'Hare, Mr P. Stanley, Mr M. Riaz, Mr Ramakrishnan, Castle Hill Hospital and Princess Royal Hospitals Hull; Dr Brunt, Mid Staffordshire General; Dr David Fyfe, Nottingham City Hospital; Dr E. Marshall, Whiston Hospital; Dr D. Seukeran, Friarage Hospital; Dr Bruce Pollock, Dr E. D. A. Potts, Dr S. Clark, Dr S. MacDonald Hull, Mr L. Le Roux Fourie, Miss O. M. B. Austin, Mr S. Southern, Mr O. M. Fenton, Mr S. Majumder, Mr A. R. Phipps, Pinderfields Hospital; Dr D. Cowan, Dr H. Hempel, Dr J. Holder, Dr M. Cheesbrough, Dr H. Galvin, Mr D. Sutton, Huddersfield Royal Infirmary, Dr I. Barbar, Calderdale Royal Hospital, Dr J. A. A. Langtry, Dr S. Natarajan, Dr Verlangi, Sunderland Royal Hospital; Dr J. M. Eremin, Lincoln County Hospital, Dr Lewis, Worcester Royal Infirmary; Dr M. L. Wood, Rotherham District General Hospital; Dr Maria Marples and Mr P. W. Griffiths, Weston Park Hospital, Sheffield; Dr Mark Middleton, Churchill Hospital, Oxford; Dr Marshall, Whiston Hospital; Dr Neil Cox, Cumberland Infirmary; Dr Neil Steven, Dr Peake, Queen Elizabeth Hospital Birmingham; Dr P. C. Lorrigan, Christie Hospital, Manchester; Dr Stewart, Walsall Manor Hospital; Dr T. Watts, Lincoln County Hospital; Dr W. Tucker, Alexandra Hospital; Dr Ilchyshyn, Walsgrave; Mr A. R. Groves, George Eliot Hospital; Mr R. Debono, University Hospital North Durham; Mr Jeremy Roberts, North Staffordshire City General; Mr Murray, Wordsley Hospital & Selly Oak Hospital; Mr P. Baguley, Middlesborough General Hospital; Mr Paul Davison, North Staffordshire City General; Mr R. Matthews, George Eliot Hospital; Mr R. B. Berry, Mr M. Erdmann, Mr N. McLean, Mr S. Rao, University Hospital North Durham; Mr S. L. Knight, Mr C. Fenn, Dr M. Marples, Mr Mark Liddington, Mr Simon Kay, Mr Howard Peach, Mr Andrew Batchelor, Dr Michelle Cronk, Dr R. Sheehan-Dare, Dr Mark Goodfield, Dr Poulam Patel, Dr Victoria Goulden, Dr Alison Humphreys, Mr K. Horgan, Mr Chips Browning, Dr Graeme Stables, Dr Sabine Sommer, Dr Caroline Wilson and Dr S. M. Wilkinson Leeds Teaching Hospitals Trust; Mr Sugden, University Hospital Hartlepool, Mr T. E. E. Goodacre, Radcliffe Infirmary, Oxford, Dr Andrew Wright, Mr Al Ghazal, Mr S. F. Worrall, Mr R. M. Antrum, Mr Ivan Foo, Mr D. Watt, Dr K. London, Dr D. J. Barker, Prof D. T. Sharpe, Mr M. Timmons, Bradford Royal Infirmary and Airedale Hospital, Dr G. P. Ford, Dr G. Taylor, Dr M. Shah, Dewsbury and District Hospital. These pathologists also assisted Dr A. Clarke, York District Hospital. Dr A. Gledhill, Harrogate, Dr A. Roy, Hull. Drs Sara Edwards, Andrew Boon and Will Merchant, Leeds Teaching Hospitals Trust. Dr Srikantiah Nagara, Sunderland Hospital, Dr Paul Barrett, University Hospital North Durham, Dr David Henderson, Friarage Hospital, Dr Hugh Cochrane, Sunderland Hospital. Airedale General, Dr John J. O'Dowd and Dr Phillip Batman, Bradford Royal Infirmary, Dr Patricia W. Gudgeon and Dr U. Raja and Dr I. W. Claire MacDonald, Dewsbury/Pontefract, Dr George D. H. Thomas, Huddersfield Royal Infirmary, Dr Alan Padwell, Calderdale, Dr Scott Sanders, Birmingham.

Supplementary Material

Single nucleotide polymorphism methodology.

The VDR single nucleotide polymorphisms (SNPs; Cdx2, rs11568820, G>A; GATA, rs4516035, A>G; FokI, rs10735810, C>T; BsmI, rs1544410, G>A; ApaI, rs7975232, C>A; and TaqI, rs731236, T>C) were chosen on the basis of the limited published data on function (supplementary data). The functional consequences of the Cdx2 and FokI variants have been extensively characterized, whereas the functionality of the common BsmI-ApaI-TaqI haplotypes is less clear. The Cdx2, FokI, GATA, and ApaI SNPs were amplified using allele-specific differentially labeled fluorescent polymerase chain reaction. The BsmI and TaqI SNPs were examined using restriction enzyme (New England Biolabs, Hitchin, United Kingdom) digest of differentially labeled fluorescent polymerase chain reaction products. (Primer sequences are listed in Appendix Table A1). Amplification was performed using Hotstar Mastermix (Qiagen, Crawley, West Sussex, United Kingdom) according to supplied protocols with an annealing temperature of 55°C. All products were detected using a 3130XL ABI Automated Sequencer (Applied Biosystems, Warrington, United Kingdom) and analyzed using Genemarker v1.5 genotyping software (Softgenetics, State College, PA).

Measurement of vitamin D methodology.

Modifications of the sample preparation and liquid chromatography tandem mass spectrometry conditions described by Maunsell et al (Clin Chem 51:1683-1690, 2005) were used. Deuterated 25-OH vitamin D3 was used as internal standard. The Waters Quattro Micro MS/MS was operated in multiple reaction monitoring mode with the transitions m/z+ 413.2>82.7 for 25-OH vitamin D2, m/z+ 401.2>159.1 for 25-OH vitamin D3 and m/z+ 407.2>159.1 for hexadeuterated 25-OH vitamin D3. Calibrators were prepared by spiking 25-OH vitamin D2 and D3 into vitamin D depleted serum. Pools of serum samples were also spiked with 25-OH D2 or D3 as necessary to prepare quality control samples. The working range of the assay is 10 to 250 nmol/L. Interassay imprecision for both metabolites was less than 10% at concentrations ranging from 12 to 159 nmol/L. Recovery of added 25-OH vitamin D2 and D3 into pooled samples ranged from 85% to 116%.

Choice of VDR SNPs.

At the 5′end of the VDR gene, the FokI polymorphism results in the creation of an alternative start site (ACG>ATG), which produces a variant protein (427a.a) that is three amino acids longer than the wild-type protein (424a.a). The Cdx2 (rs11568820, G>A) polymorphism is located in the exon 1α promoter region within a Cdx2 (a caudal-related homeodimer transcription factor) binding element. At the 3′end of the gene, the intron 8 BsmI (rs1544410, G>A) and ApaI (rs7975232, C>A) SNPs and a synonymous TaqI (rs731236, T>C, I352I) SNP in exon 9 are in tight linkage disequilibrium. Recently an additional promoter polymorphism has been described 1012-bp upstream of the exon 1a transcription start site in a melanoma population (GATA or A-1012G, rs4516035, A>G)(100). This variant seems to result in the destruction of a strong GATA-3 core-binding site.

The A allele of Cdx2 has been shown to be more “active” than the G allele, having a stronger binding affinity with the Cdx2 transcription factor and increased transcriptional activity (Yamamoto H, Miyamoto K, Li B, et al: J Bone Miner Res 14:240-247, 1999; Arai H, Miyamoto K, Taketani Y, et al: J Bone Miner Res 12:915-921, 1997). The BsmI-ApaI-TaqI region of linkage disequilibrium extends well into the 3′UTR of the VDR gene.4 The 3′UTR is known to be involved in regulation of gene expression, especially via mRNA stability. Overall in vitro studies suggest that the G-C-T (baT) haplotype is more active associated with increased VDR expression, increased mRNA stability, and increased transactivation levels, although this is by no means consistent in the literature, reviewed by Uitterlinden et al (Gene 338:143-156, 2004).

In studies reported here, VDR allele frequencies were comparable to those given in the National Center for Biotechnology Information SNP database for white populations (

Table A1.

Details of Oligonucleotide Primers

PCRPrimerLabel-Sequence (5′→ 3′)Product Sizes (bp)Concentration (μmol/L)
MP1Cdx2-ConFFAM - AAA GGA ATG AAA GAG GGA AG187–Con 155–WT 81-Mut10
ApaI-ConFFAM - GAT GTA CGT CTG CAG TGT GT218–Con 107–WT 143-Mut4
MP2FokI-ConFHEX - CCT TCA CAG GTC ATA GCA TT180–Con 147–WT 79-Mut10

NOTE. Allele-specific bases are shown in bold. Wobble bases to improve specificity are shown in lowercase. GATC tags were used to increase product size where appropriate for purposes of multiplex reactions.

Abbreviations: PCR, polymerase chain reaction; WT, wild type; Mut, mutation.

Table A2.

Predictors of Relapse-Free Survival and Overall Survival in the Univariate Analysis of the Cohort Study

HR95% CIHR95% CI
Age, per year1.031.01 to to 1.06
Sex, male v female1.851.37 to 2.491.971.41 to 2.74
    Quartile 2 v quartile 11.260.79 to 1.751.140.68 to 1.92
    Quartile 3 v quartile 11.400.89 to 2.001.310.79 to 2.15
    Quartile 4 v quartile 11.480.95 to 2.211.661.03 to 2.68
    Head and neck v trunk1.160.74 to 1.821.030.63 to 1.66
    Limbs v trunk0.650.46 to 0.930.520.35 to 0.76
    Others v trunk1.721.06 to 2.801.110.62 to 1.99
Breslow thickness, per mm1.341.29 to 1.401.311.25 to 1.36
    24.9-29.9 v < to 1.451.180.78 to 1.78
    > 29.9 v < 24.91.350.91 to 2.011.510.97 to 2.35
Vitamin D supplement, yes v no0.810.56 to 1.170.710.47 to 1.09
Serum 25-dihydroxyvitamin D2/D3 (per 20 nmol/L) increase
    January to March0.700.54 to 0.900.720.56 to 0.94
    April to June0.830.68 to 1.040.830.67 to 1.06
    July to September0.740.62 to 0.890.800.67 to 0.96
    October to December0.750.60 to 0.940.830.67, 1.04
VDR genotype
        GA v GG0.890.63 to to 1.53
        AA v GG1.270.68 to 2.371.500.78 to 2.91
        AG v AA0.940.67 to 1.340.970.66 to 1.44
        GG v AA0.830.52 to 1.330.810.48 to 1.39
        CT v CC1.000.71 to 1.421.040.69 to 1.50
        TT v CC1.280.82 to to 1.53
        GA v GG1.461.02 to to 1.75
        AA v GG1.360.85 to 2.200.850.48 to 1.51
        CA v CC1.310.85 to 2.030.990.63 to 1.55
        AA v CC1.631.02 to 2.601.090.66 to 1.80
        TC v TT1.360.95 to 1.931.140.78 to 1.68
        CC v TT1.280.80 to 2.050.870.50 to 1.52

NOTE. Findings significant at the 5% level are highlighted in bold.

Abbreviations: HR, hazard ratio; BMI, body mass index.


Supported by the following: The collection of samples in the melanoma cohort study was funded by Cancer Research UK (project Grant No. C8216/A6129 and program award C588/A4994) and by the National Institutes of Health (Grant No. R01 CA83115). Recruitment was facilitated by the UK National Cancer Research Network. The relapse study was funded by Cancer Research UK (Programme Grant No. C588/A4994). It was also in part funded by a grant from the Skin Cancer Research Fund, Frenchay Hospital, Bristol, United Kingdom.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.


The author(s) indicated no potential conflicts of interest.


Conception and design: Julia A. Newton-Bishop, Samantha Beswick, Paul Affleck, Clarissa Nolan, Jennifer H. Barrett, D. Timothy Bishop

Financial support: Julia A. Newton-Bishop, David E. Elder, D. Timothy Bishop

Administrative support: Julia A. Newton-Bishop, Paul Affleck, Susan Leake, Birute Karpavicius, Sue Haynes, Linda Whitaker, Edwina Gerry, Clarissa Nolan, Chandra Bertram, Jerry Marsden

Provision of study materials or patients: Julia A. Newton-Bishop, Samantha Beswick, Juliette Randerson-Moor, Kairen Kukalizch, Sharon Jackson, Chandra Bertram, Jerry Marsden

Collection and assembly of data: Julia A. Newton-Bishop, Samantha Beswick, Juliette Randerson-Moor, Yu-Mei Chang, Paul Affleck, May Chan, Susan Leake, Birute Karpavicius, Sue Haynes, Kairen Kukalizch, Linda Whitaker, Sharon Jackson, Edwina Gerry, Clarissa Nolan

Data analysis and interpretation: Julia A. Newton-Bishop, Samantha Beswick, Juliette Randerson-Moor, Yu-Mei Chang, Faye Elliott, May Chan, Jennifer H. Barrett, D. Timothy Bishop

Manuscript writing: Julia A. Newton-Bishop, Samantha Beswick, Juliette Randerson-Moor, Yu-Mei Chang, Paul Affleck, Faye Elliott, May Chan, Linda Whitaker, David E. Elder, Jennifer H. Barrett, D. Timothy Bishop

Final approval of manuscript: Julia A. Newton-Bishop, Samantha Beswick, Juliette Randerson-Moor, Yu-Mei Chang, Paul Affleck, Faye Elliott, May Chan, Susan Leake, Birute Karpavicius, Sue Haynes, Kairen Kukalizch, Linda Whitaker, Sharon Jackson, Edwina Gerry, Clarissa Nolan, Chandra Bertram, Jerry Marsden, David E. Elder, Jennifer H. Barrett, D. Timothy Bishop


1. Hansen CM, Binderup L, Hamberg KJ, et al. Vitamin D and cancer: Effects of 1,25(OH)2D3 and its analogs on growth control and tumorigenesis. Front Biosci. 2001;6:D820–D848. [PubMed]
2. Tosetti F, Ferrari N, De Flora S, et al. Angioprevention': Angiogenesis is a common and key target for cancer chemopreventive agents. FASEB J. 2002;16:2–14. [PubMed]
3. Holick MF. High prevalence of vitamin D inadequacy and implications for health. Mayo Clin Proc. 2006;81:353–373. [PubMed]
4. Holick MF. Vitamin D: Its role in cancer prevention and treatment. Prog Biophys Mol Biol. 2006;92:49–59. [PubMed]
5. Chen TC, Holick MF. Vitamin D and prostate cancer prevention and treatment. Trends Endocrinol Metab. 2003;14:423–430. [PubMed]
6. Kricker A, Armstrong B. Does sunlight have a beneficial influence on certain cancers? Prog Biophys Mol Biol. 2006;92:132–139. [PubMed]
7. Berwick M, Armstrong BK, Ben-Porat L, et al. Sun exposure and mortality from melanoma. J Natl Cancer Inst. 2005;97:195–199. [PubMed]
8. Clark WH, Jr, Elder DE, Guerry Dt, et al. Model predicting survival in stage I melanoma based on tumor progression. J Natl Cancer Inst. 1989;81:1893–1904. [PubMed]
9. Breslow A. Prognostic factors in the treatment of cutaneous melanoma. J Cutan Pathol. 1979;6:208–212. [PubMed]
10. Balch CM, Buzaid AC, Soong SJ, et al. Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol. 2001;19:3635–3648. [PubMed]
11. Lasithiotakis K, Leiter U, Meier F, et al. Age and gender are significant independent predictors of survival in primary cutaneous melanoma. Cancer. 2008;112:1795–1804. [PubMed]
12. Elder D, Murphy G. Washington, DC: Armed Forces Institute of Pathology; 1991. Malignant tumors (melanomas and related lesions), in Atlas of Tumor Pathology: Melanocytic Tumors of the Skin 2 (series 3) pp. 103–205.
13. MacKie RM, Hole DJ. Incidence and thickness of primary tumours and survival of patients with cutaneous malignant melanoma in relation to socioeconomic status. BMJ. 1996;312:1125–1128. [PMC free article] [PubMed]
14. Downing A, Newton-Bishop JA, Forman D. Recent trends in cutaneous malignant melanoma in the Yorkshire region of England: Incidence, mortality and survival in relation to stage of disease, 1993-2003. Br J Cancer. 2006;95:91–95. [PMC free article] [PubMed]
15. Beswick S, Affleck P, Elliott F, et al. Environmental risk factors for relapse of melanoma. Eur J Cancer. 2008;44:1717–1725. [PMC free article] [PubMed]
16. Sun X, Zemel MB. 1Alpha, 25-dihydroxyvitamin D and corticosteroid regulate adipocyte nuclear vitamin D receptor. Int J Obes (Lond) 2008;32:1305–1311. [PubMed]
17. Li J, Byrne ME, Chang E, et al. 1alpha,25-Dihydroxyvitamin D hydroxylase in adipocytes. J Steroid Biochem Mol Biol. 2008;112:122–126. [PMC free article] [PubMed]
18. Reading R, Jarvis S, Openshaw S. Measurement of social inequalities in health and use of health services among children in Northumberland. Arch Dis Child. 1993;68:626–631. [PMC free article] [PubMed]
19. Lips P. Vitamin D physiology. Prog Biophys Mol Biol. 2006;92:4–8. [PubMed]
20. Cantor I. Shedding light on vitamin D and integrative oncology. Integr Cancer Ther. 2008;7:81–89. [PubMed]
21. Gerdhem P, Ringsberg KA, Obrant KJ, et al. Association between 25-hydroxy vitamin D levels, physical activity, muscle strength and fractures in the prospective population-based OPRA Study of Elderly Women. Osteoporos Int. 2005;16:1425–1431. [PubMed]
22. Reginster JY. The high prevalence of inadequate serum vitamin D levels and implications for bone health. Curr Med Res Opin. 2005;21:579–586. [PubMed]
23. Dennis LK, Lowe JB, Lynch CF, et al. Cutaneous melanoma and obesity in the Agricultural Health Study. Ann Epidemiol. 2008;18:214–221. [PMC free article] [PubMed]
24. Thune I, Olsen A, Albrektsen G, et al. Cutaneous malignant melanoma: Association with height, weight and body-surface area—A prospective study in Norway. Int J Cancer. 1993;55:555–561. [PubMed]
25. Samanic C, Chow WH, Gridley G, et al. Relation of body mass index to cancer risk in 362,552 Swedish men. Cancer Causes Control. 2006;17:901–909. [PubMed]
26. Gallus S, Naldi L, Martin L, et al. Anthropometric measures and risk of cutaneous malignant melanoma: A case-control study from Italy. Melanoma Res. 2006;16:83–87. [PubMed]
27. Freedman DM, Sigurdson A, Doody MM, et al. Risk of melanoma in relation to smoking, alcohol intake, and other factors in a large occupational cohort. Cancer Causes Control. 2003;14:847–857. [PubMed]
28. Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle factors on mortality in women with breast cancer. Int J Cancer. 2008;123:2188–2194. [PubMed]
29. Caan BJ, Kwan ML, Hartzell G, et al. Pre-diagnosis body mass index, post-diagnosis weight change, and prognosis among women with early stage breast cancer. Cancer Causes Control. 2008;19:1319–1328. [PMC free article] [PubMed]
30. Taher YA, van Esch BC, Hofman GA, et al. 1alpha,25-dihydroxyvitamin D3 potentiates the beneficial effects of allergen immunotherapy in a mouse model of allergic asthma: Role for IL-10 and TGF-beta. J Immunol. 2008;180:5211–5221. [PubMed]

Articles from Journal of Clinical Oncology are provided here courtesy of American Society of Clinical Oncology