OBJECTIVE: The prognostic value of p53 codon72 polymorphism was analyzed in Brazilian women with cervical cancer. METHODS: The present study consists of 148 women diagnosed and treated for invasive cervical carcinoma (FIGO stages Ib-IIIb) between 1992 and 2002. Demonstration of p53 polymorphism was performed in DNA extracted from paraffin-embedded sections using the polymerase chain reaction (PCR). RESULTS: Among the 148 women, arg/arg was found in 99 (67%) and, arg/pro in 49 (33%). The overall survival (OS) curves (univariate) were different between arg/arg and arg/pro patients (P = 0.01). There was slightly increased risk of death for arg/arg patients (crude HR 2.2 CI 95% 1.2-4.0), which was not confounded by FIGO stages (adjusted HR 2.4 CI 95% 1.3-4.3). For disease-free survival (DFS), two situations were considered: (1) 124 women who received any treatment, and (2) 118 who received FIGO-recommended treatment. In the first group, 59% of arg/arg patients presented recurrence as compared to 32% in the arg/pro group (P = 0.02), whereas in the second group, 61% of the arg/arg and 34% arg/pro showed recurrence (P = 0.04). The risk of recurrence adjusted by FIGO stage for the 124 patients was 2.4 (CI 95% 1.0-3.7) and for the 118 it was 1.9 (CI 95% 1.0-3.4). These adjusted models showed no confounding and no interaction. CONCLUSIONS: Despite the prognostic significance of p53 polymorphism in univariate survival analysis, there was no or only marginal evidence on the independent prognostic value of p53arg/arg in multivariate analysis. The more ominous prognosis of the homozygous (arg/arg) patients was explained by the primary treatment, independent on the FIGO stage.
OBJECTIVE: The prognostic value of p53 codon72 polymorphism was analyzed in Brazilian women with cervical cancer. METHODS: The present study consists of 148 women diagnosed and treated for invasive cervical carcinoma (FIGO stages Ib-IIIb) between 1992 and 2002. Demonstration of p53 polymorphism was performed in DNA extracted from paraffin-embedded sections using the polymerase chain reaction (PCR). RESULTS: Among the 148 women, arg/arg was found in 99 (67%) and, arg/pro in 49 (33%). The overall survival (OS) curves (univariate) were different between arg/arg and arg/propatients (P = 0.01). There was slightly increased risk of death for arg/argpatients (crude HR 2.2 CI 95% 1.2-4.0), which was not confounded by FIGO stages (adjusted HR 2.4 CI 95% 1.3-4.3). For disease-free survival (DFS), two situations were considered: (1) 124 women who received any treatment, and (2) 118 who received FIGO-recommended treatment. In the first group, 59% of arg/argpatients presented recurrence as compared to 32% in the arg/pro group (P = 0.02), whereas in the second group, 61% of the arg/arg and 34% arg/pro showed recurrence (P = 0.04). The risk of recurrence adjusted by FIGO stage for the 124 patients was 2.4 (CI 95% 1.0-3.7) and for the 118 it was 1.9 (CI 95% 1.0-3.4). These adjusted models showed no confounding and no interaction. CONCLUSIONS: Despite the prognostic significance of p53 polymorphism in univariate survival analysis, there was no or only marginal evidence on the independent prognostic value of p53arg/arg in multivariate analysis. The more ominous prognosis of the homozygous (arg/arg) patients was explained by the primary treatment, independent on the FIGO stage.
Authors: Hayala Cristina Cavenague de Souza; Francisco Louzada; Pedro Luiz Ramos; Mauro Ribeiro de Oliveira Júnior; Gleici da Silva Castro Perdoná Journal: J Appl Stat Date: 2021-06-01 Impact factor: 1.416