Objective Cervical carcinoma (CC), a multifactorial cancer, is assumed to have a host genetic predisposition component that modulates its susceptibility in various populations. We investigated the association between CC risk in Saudi women and 6 single-nucleotide polymorphisms (SNPs) in hypothesis-driven candidate genes.

Methods A total of 545 females were included, comprising 232 CC patients and 313 age-/sex-matched control subjects. Six SNPs (CDKN1A C31A, ATM G1853A, HDM2 T309G, TGFB1 T10C, XRCC1 G399A, and XRCC3 C241T) were genotyped by direct sequencing.

Results Of the 6 SNPs studied, TGFB1 T10C (odds ratio, 0.74; 95% confidence interval, 0.57–0.94) and XRCC1 G399A (odds ratio, 1.45; 95% confidence interval, 1.11–1.90) displayed different frequencies in cancer patients and control subjects and showed statistically significant association in univariate (P = 0.017, P = 0.005, respectively) analysis. The Cochran-Armitage trend test had confirmed the results (P = 0.027 and P = 0.006, respectively), indicating an ordering in the effect of the risk alleles in CC patients. The 2 SNPs, TGFB1 T10C and XRCC1 G399A, showed also degrees of deviation from Hardy-Weinberg equilibrium in cancer patients (P = 0.001 and P = 0.083, respectively) but not in the control subjects. Furthermore, correction for multiple testing using multivariate logistic regression to assess the joint effect of all SNPs has sustained significant statistical association (P = 0.025 and P = 0.009, respectively).

Conclusions TGFB1 T10C and XRCC1 G399A SNPs were associated with CC risk in univariate and multivariate analysis and displayed allele-dosage effects and coselection in cancer patients. Patients harboring the majority allele TGFB1 T10 (Leu) or the variant allele XRCC1 399A (Gln) have approximately 1.5-fold increased risk to develop CC. Host SNPs genotyping may provide relevant biomarkers for CC risk assessment in personalized preventive medicine.