Combined effects of gender, skin type and polymorphic genes on clinical phenotype: use of rate of increase in numbers of basal cell carcinomas as a model system
Introduction
Various polymorphic genes have been linked with susceptibility and outcome in cancer [1], [2]. The identification of such genes prompts the question of how many loci, in combination, determine risk and, what is the relative importance of each gene in the presence of increasing numbers of other variables. Over recent years, we have used patients with cutaneous basal cell carcinomas (BCC) as a model system in which to study associations between genetic factors and disease risk and outcome [3], [4], [5], [6]. This particular cancer has the advantage that patients demonstrate marked, inter-individual diversity in several phenotypes including the rate at which further BCC develop during follow-up [3], [6], [7], [8]. We have described several phenotypes that are associated with how many BCC develop per year. For example, many patients develop no further BCC after first presentation with one tumor (mean 0.2 BCC/year). We termed this presentation pattern, single presentation phenotype 1 (SPP1). Other patients developed further tumors during follow-up (mean 0.6 BCC/year) but only presented with a single lesion at each presentation (SPP2). A further group comprising about 15% of patients demonstrated a presentational phenotype characterised by the appearance of multiple BCC (between 2 and 10 new tumours) at a presentation (multiple presentation phenotype, MPP). MPP patients developed relatively large numbers of BCC/year (mean 2.0) [4], [5].
The mechanisms that determine this marked inter-patient variability in numbers of BCC/year are unclear though we previously suggested that the rate of appearance of tumours is mediated by the efficiency of immune surveillance in recognising existing micro-tumours. This efficiency is in turn determined by environmental and host factors. We have attempted to identify host characteristics including allelic genes that are associated with numbers of BCC/year in patients with different phenotypes. For example, male gender (compared with females), skin type 1 (always burn on exposure to ultraviolet, UV, unable to tan) (compared with skin types 2–4) [4], and the polymorphic cytochrome P450 CYP2D6 EM, vitamin D receptor VDR TT and tumor necrosis factor-alpha (TNF-α) GG genes have been individually associated with the development of large numbers of BCC/year [4], [9].
The impact of a variable on the number of BCC/year is described by the rate ratio (RR). This approximates to the mean number of BCC per year in the test group/mean number of BCC per year in the reference category. A typical value for the RR for variables dichotomised as risk (e.g. male) and protective (e.g. female) is 1.3 [4]. While such values are associated with significant, between-group differences in BCC numbers/year, they are derived from the study of heterogenous groups. Thus, the RR for males is an average for a group comprising males with other risk (e.g. skin type 1, CYP2D6 EM, VDR TT, TNF GG) or, corresponding protective (e.g. skin types 2–4, other CYP2D6, VDR, TNF genotypes) variables. The reference category, female sex, is similarly heterogeneous. Thus, selection of males with progressively more risk variables will result in a smaller, increasingly homogeneous, high-risk group while selection of corresponding females with additional protective variables will define a smaller, more homogeneous low-risk, reference category. It is assumed that comparison of numbers of BCC/year in groups that comprise individuals with progressively more risk variables with that in the lower risk reference category will demonstrate increasingly larger values of RR.
Most studies of associations between allelic variants and disease phenotype have considered the impact of variables on phenotype in unstratified patient groups. The issue of heterogeneity within groups and, the extent to which addition of each further variable to combinations of factors affects risk has not generally been considered. We now describe studies to determine the effect of variables individually and in combinations on BCC numbers/year. Our aim was to define a combination of variables that describe a high risk group and corresponding lower risk reference category and, determine the extent of difference in RR between them. We used gender as the first variable [6], [7], [8] and then progressively stratified patients by skin type and CYP2D6, VDR, and TNF-α genotypes.
Section snippets
Patients
We studied 846 unrelated, Northern European Caucasians (29–90 years at first presentation) with one or more histologically proven BCC. They were recruited at first presentation or at annual follow-up in the North Staffordshire Hospital with Ethics Committee approval and informed consent between 1991 and 1998 [3], [4], [5], [6]. They represent about 20% eligible patients. We excluded subjects with basal cell naevus syndrome, xeroderma pigmentosum or BCC and any other malignancy but otherwise
Number of BCC/year and single variables
Table 1 shows BCC cases stratified by gender, skin type, CYP2D6, VDR, and TNF genotypes. We used negative binomial regression analysis to compare number of BCC/year in individual risk (males, skin type 1, CYP2D6 EM, VDR TT, TNF-α GG) and reference categories. The low-risk variables in the reference categories were grouped as skin types 2–4, CYP2D6 HET and PM, VDR Tt and tt and TNF GA and AA as numbers of BCC/year in subjects with these characteristics were similar [9]. When variables were
Discussion
We determined the impact of variables individually and in combination on the numbers of BCC developed per year. The variables were chosen as they have been previously associated with this model phenotype [11] and, their relatively favourable frequencies of risk/protective characteristics allowed combinations of gender, skin type and up to two genes. Clearly, the number of patients with some combinations of three or four variables was small and these findings should be interpreted cautiously.
References (16)
- et al.
Glutathione S-transferase GSTM1 phenotypes and protection against cutaneous malignancy
Lancet
(1994) - et al.
Susceptibility and modifier genes in cutaneous basal cell carcinomas and their associations with clinical phenotype
J. Photochem. Photobiol.
(2001) - et al.
Dopamine formation from Tyramine by CYP2D6
Biochem. Biophys. Res. Commun.
(1998) - et al.
Analysis of 1,25-dihydroxyvitamin D(3) receptors (VDR) in basal cell carcinomas
Am. J. Pathol.
(1999) - et al.
Who's afraid of epistasis?
Nat. Genet.
(1996) - et al.
Complexity of lung cancer modifiers: mapping of 30 genes and 25 interactions in half of the mouse genome
J. Natl. Cancer Inst.
(2001) - et al.
Presentation with multiple cutaneous basal cell carcinomas: association of glutathione S-transferase and cytochrome P450 genotypes with clinical phenotype
Cancer Epidemiol. Biomarkers Prev.
(1999) - et al.
Basal cell carcinoma: tumor clustering is associated with increased accrual in high-risk subgroups
Cancer
(2000)
Cited by (43)
Sunlight, skin cancer and vitamin D
2023, Feldman and Pike's Vitamin D: Volume Two: Disease and TherapeuticsEndocrine actions of vitamin D in skin: Relevance for photocarcinogenesis of non-melanoma skin cancer, and beyond
2017, Molecular and Cellular EndocrinologyCitation Excerpt :These findings are well in line with a speculative function of Ptch as an efflux pump for vitamin D-related compounds that exert Hh-inhibitory potential (Biijlsma et al., 2006). Considering the relevance of the vitamin D endocrine system for carcinogenesis of skin cancer, low 25(OH)D serum concentrations and genetic variants (SNPs) of the vitamin D endocrine system have been identified as potential risk factors for occurrence and prognosis of skin malignancies (Köstner et al., 2009, 2012; Denzer et al., 2011; Ramachandran et al., 2003; Han et al., 2007; Caini et al., 2014). SNPs in the VDR gene can be associated with altered expression and/or function of the VDR protein (Köstner et al., 2009; Denzer et al., 2011).
Association of Prediagnostic Serum Vitamin D Levels with the Development of Basal Cell Carcinoma
2010, Journal of Investigative DermatologyCitation Excerpt :In mice, vitamin D receptor knockouts are more susceptible to chemically induced and UVR-induced skin tumors, suggesting that disruption of vitamin D receptor signaling predisposes to skin cancer (Zinser et al, 2002). Vitamin D receptor polymorphisms have been associated with increased BCC risk (Ramachandran et al. 2003). Immunohistochemical studies of human BCCs suggest that the vitamin D pathway may be important for the growth behavior of BCCs (Kamradt et al, 2003; Mitschele et al, 2004).
New Concepts for Basal Cell Carcinoma. Demographic, Clinical, Histological Risk Factors, and Biomarkers. A Systematic Review of Evidence Regarding Risk for Tumor Development, Susceptibility for Second Primary and Recurrence
2010, Journal of Surgical ResearchCitation Excerpt :Mononuclear cells from patients with a previous BCC demonstrated a significantly increased release of TNF-α upon stimulation with lipopolysaccharide compared with mononuclear cells from age-matched control subjects [33]. TNF-α GG genotype was reported to be associated with more BCC per y [27]. Also, HLA-DR4 (locus 6p21.3) was reported to be associated with multiple BCC, but the authors concluded that HLA-system might only play a minor role in the development of multiple BCC [34].
The beneficial role of vitamin D and its analogs in cancer treatment and prevention
2010, Critical Reviews in Oncology/HematologyClinical and prognostic factors in the development of basal cell carcinoma
2017, Clinics in DermatologyCitation Excerpt :An analysis of our study population on the basis of tumor number revealed that, compared with single BCCs, multiple BCCs tend to develop in older men, to be superficial, and to localize in the areas M and L. A greater prevalence of multiple BCCs in elderly men involving the head-neck and trunk regions was consistent with the findings of previous reports.6,11,12 A newly developed BCC was diagnosed in 30 of 363 patients (8.2%) who were accessed via telephone and attended follow-up on request.