Background: After DNA damage, the p53Arg allele of the Arg72Pro SNP (rs1042522:C>G) has a higher pro-apoptotic activity, while the p53Pro allele more efficiently induces cell cycle arrest. The G allele of the MDm2 SNP309 (rs2279744:T>G) expresses higher levels of the protein thus attenuating the p53 pathway. Both SNPs have been shown to affect lung cancer risk. We aimed to investigate their relationship with the clinical outcome of chemotherapy. We prospectively recruited 426 consecutive patients with advanced disease (352 NSCLC and 74 SCLC, 57% metastatic) referred for chemotherapy at our institution from July 1, 2002 to February 8, 2006. 82% of the patients were male, median age at diagnosis was 63 years, 56% current smokers and 11% never smokers, 81% received combined platinum-based chemotherapy. Median follow up time was 10.5 months. We designed specific TP53 and MDm2 primers for typing both SNPs using the Pyrosequencing® assay. Patients genotype frequencies were TP53 Arg/Arg 49%, Arg/Pro 42%, Pro/Pro 8%, MDm2 T/T 37%, T/G 40%, G/G 17% and were in Hardy-Weinberg equilibrium. At multivariable analysis (adjusted for gender, smoking status, type of chemotherapy, disease stage, and side effects) survival was significantly associated with Performance Status (PS) [HR 1.54 (1.2-2.0)], histology [SCLC vs. NSCLC - HR 1.51(1.1-2.1)] and objective response [yes vs. no HR 0.56 (0.4-0.7)] but with neither SNP. In contrast, both grade 3-4 toxicity and objective response were significantly associated with the two SNPs: p53Pro carriers more frequently experienced toxicity (HR 1.40, C.I. 1.1-1.8) and objective response (HR 1.44, C.I. 1.0-2.0) compared with p53Arg/Arg homozygotes. Conversely, toxicities and objective response were less frequent in MDm2309 GG vs. TT homozygotes (HR 0.57, C.I. 0.4-0.9 and 0.61, C.I.0.4-0.97 respectively). These findings are in agreement with the notion that MDm2 GG homozygous cells express higher levels of mdm2, thus attenuating the p53 pathway, but not with the alleged greater apoptotic potential of p53 Arg72. The reliability of the study was strengthened by the observation of significant associations of toxicity with platinum therapy (HR 0.59, C.I. 0.4-0.9), of objective response with histology (SCLC vs NSLC HR 2,16, C.I. 1.5-3.1) and of both outcomes with PS (toxicity: HR 1.64, C.I. 1.2-2.2; objective response HR 1.57, C.I. 1.1-2.2), in agreement with expectations. This prospective study provides preliminary evidence that germ-line p53 and MDm2 SNPs affect toxicity and objective response to therapy in lung cancer patients, probably by modulating the DNA damage response in the patients' tissues. In this study the assessed genotypes do not affect survival, suggesting that this is mainly determined by the tumor aggressiveness and somatic changes acquired during carcinogenesis (p53 mutations and MDm2 amplification), rather than by the constitutional genotype.
Prospective study assessing MDM2 309 and p53 Arg72Pro single nucleotide polymorphism (SNPs) and clinical outcome in advanced lung cancer
MANDRILE, Giorgia;GIACHINO, Daniela Francesca;NOVELLO, Silvia;GREGORI, Dario;DE MARCHI, Mario;SCAGLIOTTI, Giorgio Vittorio
2007-01-01
Abstract
Background: After DNA damage, the p53Arg allele of the Arg72Pro SNP (rs1042522:C>G) has a higher pro-apoptotic activity, while the p53Pro allele more efficiently induces cell cycle arrest. The G allele of the MDm2 SNP309 (rs2279744:T>G) expresses higher levels of the protein thus attenuating the p53 pathway. Both SNPs have been shown to affect lung cancer risk. We aimed to investigate their relationship with the clinical outcome of chemotherapy. We prospectively recruited 426 consecutive patients with advanced disease (352 NSCLC and 74 SCLC, 57% metastatic) referred for chemotherapy at our institution from July 1, 2002 to February 8, 2006. 82% of the patients were male, median age at diagnosis was 63 years, 56% current smokers and 11% never smokers, 81% received combined platinum-based chemotherapy. Median follow up time was 10.5 months. We designed specific TP53 and MDm2 primers for typing both SNPs using the Pyrosequencing® assay. Patients genotype frequencies were TP53 Arg/Arg 49%, Arg/Pro 42%, Pro/Pro 8%, MDm2 T/T 37%, T/G 40%, G/G 17% and were in Hardy-Weinberg equilibrium. At multivariable analysis (adjusted for gender, smoking status, type of chemotherapy, disease stage, and side effects) survival was significantly associated with Performance Status (PS) [HR 1.54 (1.2-2.0)], histology [SCLC vs. NSCLC - HR 1.51(1.1-2.1)] and objective response [yes vs. no HR 0.56 (0.4-0.7)] but with neither SNP. In contrast, both grade 3-4 toxicity and objective response were significantly associated with the two SNPs: p53Pro carriers more frequently experienced toxicity (HR 1.40, C.I. 1.1-1.8) and objective response (HR 1.44, C.I. 1.0-2.0) compared with p53Arg/Arg homozygotes. Conversely, toxicities and objective response were less frequent in MDm2309 GG vs. TT homozygotes (HR 0.57, C.I. 0.4-0.9 and 0.61, C.I.0.4-0.97 respectively). These findings are in agreement with the notion that MDm2 GG homozygous cells express higher levels of mdm2, thus attenuating the p53 pathway, but not with the alleged greater apoptotic potential of p53 Arg72. The reliability of the study was strengthened by the observation of significant associations of toxicity with platinum therapy (HR 0.59, C.I. 0.4-0.9), of objective response with histology (SCLC vs NSLC HR 2,16, C.I. 1.5-3.1) and of both outcomes with PS (toxicity: HR 1.64, C.I. 1.2-2.2; objective response HR 1.57, C.I. 1.1-2.2), in agreement with expectations. This prospective study provides preliminary evidence that germ-line p53 and MDm2 SNPs affect toxicity and objective response to therapy in lung cancer patients, probably by modulating the DNA damage response in the patients' tissues. In this study the assessed genotypes do not affect survival, suggesting that this is mainly determined by the tumor aggressiveness and somatic changes acquired during carcinogenesis (p53 mutations and MDm2 amplification), rather than by the constitutional genotype.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.