Single Nucleotide Polymorphism of the TGF-B1 T869C Gene Associates Risks of Treatment Related Pneumonitis and Distant Metastasis in Patients with Non–small-cell Lung Cancer Treated with Definitive Radiotherapy
Z. Liao,X. Yuan,R. Komaki,L. Milas,Z. Liu,L. Wang,S. L. Tucker,L. Mao,X. S. Wang,M. Martel
DOI: https://doi.org/10.1016/j.ijrobp.2009.07.1001
2009-01-01
Abstract:Purpose/Objective(s)To investigate the association between single nucleotide polymorphisms (SNPs) in TGF-β1 gene and risks of treatment related pneumonitis (TRP), overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) in patients with NSCLC.Materials/MethodsUsing 217 available genomic DNA samples from NSCLC patients treated with definitive radiotherapy at our institution between 11/1998 and 1/2005, we genotyped three SNPs of the TGF-β1 gene (rs1800469:C-509T, rs1800471:G915C, and rs1982073:T869C) by polymerase chain reaction-restriction fragment length polymorphism method. We used Kaplan-Meier cumulative probability to assess the risk of Grade ≥1 TRP, OS, PFS, and DMFS. Cox proportional hazards analyses to evaluate the effect of TGF-β1 genotypes on such risk.ResultsThere were 124 men and 93 women in the study, with median age of 63 years. The median total radiation dose was 63 Gy (range, 50.4– 84.0 Gy) given in 1.8- to 2-Gy fractions once a day (151 patients) or in 1.2 Gy fractions twice a day (66 patients); 97% of patients received 60–70 Gy. The median follow-up times was 17 and 39 months for the entire group and living patients (range, 1–97 months), respectively. Grades ≥1, ≥2, and ≥ 3 TRP were observed in 151(69.5%), 88 (40.5%) and 40 patients (18.4%), respectively. Multivariate analysis found CT genotypes of TGF-β1 rs1982073:T869C to be associated with a statistically significantly lower risk of TRP Grades ≥ 1 (hazard ratio [HR] = 0.668; 95% CI, 0.475 to 0.957; p = .037) compared with the TT genotype. The median times of OS, PFS, and DMFS were 17, 8, and 10 months, respectively. Multivariate analysis showed that having the CT genotype in TGFβ1 rs1982073 predicted poor PFS (HR = 1.603, 95% confidence interval [CI] = 1.117–2.301, p = 0.01) and DMFS (HR = 1.823, 95% CI = 1.157–2.874, p = 0.01) in all patients and in those with brain metastases (HR = 3.491, 95% CI =1.525–7.989, p = 0.003). Having the CT genotype in TGFβ1 rs1800469 was also associated with shorter DMFS (HR = 1.569, 95% CI = 1.040–2.367, p = 0.032). OS was not associated with genotype at any of the three SNPs assessed. All HRs were adjusted for age, gender, race, Karnofsky performance score (KPS), smoking status, tumor histology, gross tumor volume(GTV), disease stage, concurrent chemoradiation therapy, and radiation dose received.ConclusionsOur finding suggests that TGF-β1 SNP rs1982073:T869C may be an important predictive biomarker for risk of toxicity and distant metastasis before treatment initiation, therefore used in selecting patients for personalization of treatments for NSCLC. Larger studies are needed for validation. Purpose/Objective(s)To investigate the association between single nucleotide polymorphisms (SNPs) in TGF-β1 gene and risks of treatment related pneumonitis (TRP), overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) in patients with NSCLC. To investigate the association between single nucleotide polymorphisms (SNPs) in TGF-β1 gene and risks of treatment related pneumonitis (TRP), overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) in patients with NSCLC. Materials/MethodsUsing 217 available genomic DNA samples from NSCLC patients treated with definitive radiotherapy at our institution between 11/1998 and 1/2005, we genotyped three SNPs of the TGF-β1 gene (rs1800469:C-509T, rs1800471:G915C, and rs1982073:T869C) by polymerase chain reaction-restriction fragment length polymorphism method. We used Kaplan-Meier cumulative probability to assess the risk of Grade ≥1 TRP, OS, PFS, and DMFS. Cox proportional hazards analyses to evaluate the effect of TGF-β1 genotypes on such risk. Using 217 available genomic DNA samples from NSCLC patients treated with definitive radiotherapy at our institution between 11/1998 and 1/2005, we genotyped three SNPs of the TGF-β1 gene (rs1800469:C-509T, rs1800471:G915C, and rs1982073:T869C) by polymerase chain reaction-restriction fragment length polymorphism method. We used Kaplan-Meier cumulative probability to assess the risk of Grade ≥1 TRP, OS, PFS, and DMFS. Cox proportional hazards analyses to evaluate the effect of TGF-β1 genotypes on such risk. ResultsThere were 124 men and 93 women in the study, with median age of 63 years. The median total radiation dose was 63 Gy (range, 50.4– 84.0 Gy) given in 1.8- to 2-Gy fractions once a day (151 patients) or in 1.2 Gy fractions twice a day (66 patients); 97% of patients received 60–70 Gy. The median follow-up times was 17 and 39 months for the entire group and living patients (range, 1–97 months), respectively. Grades ≥1, ≥2, and ≥ 3 TRP were observed in 151(69.5%), 88 (40.5%) and 40 patients (18.4%), respectively. Multivariate analysis found CT genotypes of TGF-β1 rs1982073:T869C to be associated with a statistically significantly lower risk of TRP Grades ≥ 1 (hazard ratio [HR] = 0.668; 95% CI, 0.475 to 0.957; p = .037) compared with the TT genotype. The median times of OS, PFS, and DMFS were 17, 8, and 10 months, respectively. Multivariate analysis showed that having the CT genotype in TGFβ1 rs1982073 predicted poor PFS (HR = 1.603, 95% confidence interval [CI] = 1.117–2.301, p = 0.01) and DMFS (HR = 1.823, 95% CI = 1.157–2.874, p = 0.01) in all patients and in those with brain metastases (HR = 3.491, 95% CI =1.525–7.989, p = 0.003). Having the CT genotype in TGFβ1 rs1800469 was also associated with shorter DMFS (HR = 1.569, 95% CI = 1.040–2.367, p = 0.032). OS was not associated with genotype at any of the three SNPs assessed. All HRs were adjusted for age, gender, race, Karnofsky performance score (KPS), smoking status, tumor histology, gross tumor volume(GTV), disease stage, concurrent chemoradiation therapy, and radiation dose received. There were 124 men and 93 women in the study, with median age of 63 years. The median total radiation dose was 63 Gy (range, 50.4– 84.0 Gy) given in 1.8- to 2-Gy fractions once a day (151 patients) or in 1.2 Gy fractions twice a day (66 patients); 97% of patients received 60–70 Gy. The median follow-up times was 17 and 39 months for the entire group and living patients (range, 1–97 months), respectively. Grades ≥1, ≥2, and ≥ 3 TRP were observed in 151(69.5%), 88 (40.5%) and 40 patients (18.4%), respectively. Multivariate analysis found CT genotypes of TGF-β1 rs1982073:T869C to be associated with a statistically significantly lower risk of TRP Grades ≥ 1 (hazard ratio [HR] = 0.668; 95% CI, 0.475 to 0.957; p = .037) compared with the TT genotype. The median times of OS, PFS, and DMFS were 17, 8, and 10 months, respectively. Multivariate analysis showed that having the CT genotype in TGFβ1 rs1982073 predicted poor PFS (HR = 1.603, 95% confidence interval [CI] = 1.117–2.301, p = 0.01) and DMFS (HR = 1.823, 95% CI = 1.157–2.874, p = 0.01) in all patients and in those with brain metastases (HR = 3.491, 95% CI =1.525–7.989, p = 0.003). Having the CT genotype in TGFβ1 rs1800469 was also associated with shorter DMFS (HR = 1.569, 95% CI = 1.040–2.367, p = 0.032). OS was not associated with genotype at any of the three SNPs assessed. All HRs were adjusted for age, gender, race, Karnofsky performance score (KPS), smoking status, tumor histology, gross tumor volume(GTV), disease stage, concurrent chemoradiation therapy, and radiation dose received. ConclusionsOur finding suggests that TGF-β1 SNP rs1982073:T869C may be an important predictive biomarker for risk of toxicity and distant metastasis before treatment initiation, therefore used in selecting patients for personalization of treatments for NSCLC. Larger studies are needed for validation. Our finding suggests that TGF-β1 SNP rs1982073:T869C may be an important predictive biomarker for risk of toxicity and distant metastasis before treatment initiation, therefore used in selecting patients for personalization of treatments for NSCLC. Larger studies are needed for validation.