Storey et al.1 reported that a p53 gene polymorphism which results in the substitution of arginine for proline at codon 72 significantly increases the susceptibility of p53 protein degradation by the human papillomavirus E6 oncoprotein and suggested that individuals homozygous for the arginine residue may be at increased risk of developing cervical cancer following exposure to human papillomavirus 16 or 18. Several subsequent reports on British, Scandinavian, Dutch and Japanese populations2–4 have failed to confirm an association between codon 72 arginine homozygosity and cancer of the cervix. It is well documented that the p53 codon 72 genotypes vary according to ethnicity, with black Africans having a markedly lower prevalence of the arginine allele5, and consequently there have been calls for studies in other populations groups.
This study examined the p53 codon 72 genotype in blood samples taken from 121 black South African women with histologically proven squamous carcinoma of the cervix. In 100 of these women freshly biopsied tumour tissue was also available for human papillomavirus subtyping. The control group comprised 251 healthy race-matched women attending a contraceptive services facility. The p53 genotype analysis was performed by allele specific amplification on both blood and tumour DNA, homozygote samples being repeated to exclude possible non-amplification. Previously sequenced heterozygous blood and tumour DNA was included in each run as a control. Tumour DNA was amplified using human papillomavirus MY11/09 consensus primers, followed by direct nucleic acid sequencing, and matched against genbank sequences.
Table 1 shows that there were no statistically significant differences in the frequency of the homozygous arginine genotype between patients with cancer of the cervix, irrespective of human papillomavirus status, and healthy controls. A higher arginine allele frequency was seen in patients without a human papillomavirus infection and in the presence of non-human papillomavirus 16 or 18 viral DNA, but the differences did not reach statistical significance using Yates corrected χ2 test. Blood and tumour genotypes in the 100 patients in whom both analyses were performed were concordant, with the exception of three individuals whose tumours appeared to have lost the proline allele. In all groups the distribution of alleles fits the Hardy-Weinberg equilibrium.
|Genotype distribution||Allele frequency|
|Cancer patients (n= 121)||20 (17)||56 (46)||45 (37)||96 (40)||146 (60)|
|Tumour HPV 16/18 (n= 57)||7 (12)||26 (46)||24 (42)||40 (35)||74 (65)|
|Tumour HPV other (n= 19)||5 (26)||8 (42)||6 (32)||18 (47)||20 (53)|
|Tumour HPV negative (n= 24)||4 (17)||12 (50)||8 (33)||20 (43)||28 (57)|
|Controls (n= 251)||27 (11)||108 (43)||116 (46)||162 (32)||340 (68)|
The lower prevalence of the p53 arginine allele and the resultant low proportion of arginine homozygotes previously reported in black African populations are clearly confirmed in this study. The similarity in the p53 genotype distribution between women with cervical tumours exposed to high risk human papillomavirus 16 and 18, and normal healthy controls suggests that the arginine allele does not predispose towards viral tumorigenesis in this population and supports the pattern of findings in other ethnic groups. Infection by human papillomavirus occurs frequently in black South African women while cancer of the cervix is the most commonly occurring malignancy. Given the low incidence of arginine homozygosity in this group, a significant correlation between this genotype and the risk of developing cancer of the cervix in human papillomavirus exposed women would thus be difficult to explain. Perhaps of greater interest in this study is the markedly higher frequency of the arginine allele in the human papillomavirus negative patients and in those with lower risk human papillomavirus infections. While the numbers of patients in each of these subgroups is still small, results so far suggest that the arginine p53 allele may contribute to the risk of developing cervical tumours through some mechanism other than an increased susceptibility to p53 protein degradation.