|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 6 (February 20), 2008: pp. 856-862 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.13.5947 VEGF Polymorphisms and Survival in Early-Stage Non–Small-Cell Lung Cancer
From the Massachusetts General Hospital; Harvard School of Public Health, Boston, MA; Princess Margaret Hospital, Toronto, Ontario, Canada Corresponding author: David C. Christiani, MD, MPH, Harvard School of Public Health, 665 Huntington Ave, Boston, MA 02115; e-mail: dchris{at}hsph.harvard.edu
Purpose Polymorphisms in the VEGF gene have been identified that are believed to have functional activity. We hypothesized that such polymorphisms may affect survival outcomes among early-stage non—small-cell lung cancer (NSCLC) patients. Patients and Methods We evaluated the relationship between VEGF polymorphisms and overall survival (OS) among patients with early-stage NSCLC treated with surgical resection at Massachusetts General Hospital from 1992 to 2001. We specifically investigated the VEGF polymorphisms +936C>T (rs3025039), –460T>C (rs833061), and +405G>C (rs2010963). Analyses of genotype associations with survival outcomes were performed using Cox proportional hazards models, Kaplan-Meier methods, and the log-rank test. Results There were 462 patients and 237 deaths. Patients carrying the variant C allele of the VEGF +405G>C polymorphism had significantly improved survival (crude hazard ratio [HR] = 0.70; 95% CI, 0.54 to 0.90; P = .006; adjusted HR = 0.70; 95% CI, 0.54 to 0.91; P = .008). Five-year OS for patients carrying the variant C allele of the VEGF +405G>C polymorphism was 61% (95% CI, 54% to 67%) versus 51% (95% CI, 43% to 59%) for those who had the wild-type variant. There was a trend toward improved survival among patients carrying the variant T allele of the VEGF +936C>T polymorphism (crude HR = 0.74; 95% CI, 0.53 to 1.03; P = .07; adjusted HR = 0.73; 95% CI, 0.52 to 1.03; P = .07). Moreover, patients with higher number of variant alleles of the +405G>C and +936C>T polymorphisms had better survival. There was no association found with the –460T>C polymorphism. Conclusion Polymorphisms in VEGF may affect survival in early-stage lung cancer.
Patients with early-stage lung cancer have a chance for cure, but many will experience disease recurrence and ultimately die as a result of their lung cancer. Although TNM stage is the best available clinical measure of a tumor's aggressiveness and prognosis, there are clearly important differences within a stage grouping. The ability to refine prognosis with molecular markers that are readily accessible would be helpful. Tumors require an adequate blood supply to grow.1 Vascular endothelial growth factor A (VEGF-A; also commonly called VEGF) is a critical angiogenic factor, and inhibition of VEGF has been correlated with suppression of tumor growth and angiogenesis.2-5 VEGF expression has been correlated with high microvessel counts as well as histologic de-differentiation,6-8 and increasing vascularity of tumors as measured by microvessel count and VEGF expression may be associated with worse prognosis in non—small-cell lung cancer (NSCLC).9-12 Polymorphisms in the VEGF gene have been identified that are believed to have functional activity. The variant allele of the +936C>T polymorphism, located in the 3'-untranslated region, has been associated with lower VEGF plasma levels.13,14 The —460T>C polymorphism is located in the promoter region and the T allele may be associated with decreased VEGF promoter activity.15 The C allele of the +405G>C polymorphism has been associated with lower VEGF levels as well.15,16 The +405G>C polymorphism is predicted to lie within an MZF1 transcription factor binding site, with the C allele reducing binding specificity; it is suggested that the lower VEGF production seen with the C allele is due to disruption of the MZF1 binding site.16 However, some studies have reported conflicting results for the function of the +405G>C polymorphism or no association with VEGF levels.17,18 We hypothesize that polymorphisms in VEGF may be associated with survival outcomes in NSCLC. Specifically, we hypothesize that the VEGF +405C, –460T, and +936T alleles, which may be associated with lower VEGF levels, are associated individually with improved overall survival in early-stage NSCLC.
Study Population Since 1992, patients with histologically confirmed NSCLC have been recruited prospectively onto a molecular epidemiology study at Massachusetts General Hospital (MGH; Boston, MA). Blood samples for genotyping and patient and demographic information (including age, sex, and smoking status) were collected at the time of recruitment. Informed consent was obtained to collect follow-up data. More than 85% of eligible patients were recruited in this cohort. We limited our analysis to all patients with incident cases of stage I or II NSCLC, enrolled between December 1992 and December 2001, who had their surgical resection performed at MGH and had follow-up data. There were 550 patients who met these criteria. Of these, 88 did not have adequate blood for genotyping and were excluded from the analysis. There were no significant differences in demographic information between those who had blood available for genotyping and those who did not. The study was approved by the institutional review boards of MGH and the Harvard School of Public Health (Boston, MA).
Genotyping
Outcomes Collection
Statistical Analysis
Patient Characteristics There were 462 patients. Demographic, tumor, and treatment characteristics are listed in Table 1. All patients were treated with surgical resection. Only three patients received adjuvant chemotherapy; the time period of this study was during a time when adjuvant chemotherapy was not of definite proven benefit. There were 32 patients (7%) who received adjuvant radiation; radiation therapy was closely linked to higher stage of disease. Genotype frequencies of the VEGF polymorphisms are as follows: for +936C>T, C/C 358 (78%), C/T 99 (21%), and T/T 5 (1%); for –460T>C, T/T 145 (32%), T>C 205 (44%), and C/C 112 (24%); for +405G>C, G/G 196 (42%), G>C 203 (44%), and C/C 63 (14%). Genotype was not associated with patient- or tumor-related factors such as smoking status, stage, or histologic subtype for any of the polymorphisms. The VEGF –460 T>C and +405G>C polymorphisms were closely linked (Lewontin's D' 0.97).
OS There were 237 deaths. On univariate analysis, age, sex, stage, histologic subtype, pack-years of smoking, and adjuvant radiation were all significantly associated with survival outcome (Table 2). The VEGF +405G>C polymorphism was significantly associated with survival (compared with G/G, crude hazard ratio [HR] for G/C = 0.68; 95% CI, 0.51 to 0.89; crude HR for C/C = 0.76; 95% CI, 0.52 to 1.13). Given that the direction of effect was the same in the G/C and C/C genotypes, and there were few participants in the homozygous variant group, we combined the G/C and C/C genotypes for additional analysis. With this dominant model, we found that patients carrying the variant C allele of the VEGF +405G>C polymorphism had significantly improved survival (crude HR = 0.70; 95% CI, 0.54 to 0.90).
After adjusting for age, sex, stage, and pack-years of smoking, the VEGF +405G>C polymorphism remained significantly associated with survival (Table 3). Given that histology and radiation had been significant on univariate analysis, we did initially include these markers in the adjusted model. However, neither histology nor radiation remained significant covariates in multivariate analysis, and therefore the final model did not include these factors. The adjusted HR point estimate or level of significance did not change with or without these factors. Five-year OS for patients carrying the variant C allele of the VEGF +405G>/C polymorphism was 61% (95% CI, 54% to 67%) v 51% (95% CI, 43% to 59%) for those who were had the wild-type variant (Table 3 and Fig 1).
There was a trend toward improved survival among patients carrying the variant T allele of the VEGF +936C>T polymorphism (crude HR = 0.74; 95% CI, 0.53 to 1.03; P = .07; adjusted HR = 0.73; 95% CI, 0.52 to 1.03; P = .07; Tables 2 and 3). Five-year survival for those with the wild-type C/C genotype was 54% (95% CI, 48% to 60%), compared with 67% (95% CI, 57% to 76%) for those who carried a variant T allele (Fig 2). There was no association found with the –460T/C polymorphism.
We performed an exploratory analysis investigating the combined effect of the VEGF +936C>T and +405G>C polymorphisms. Patients with a higher number of variant alleles had improved survival. Compared with the reference group of patients with zero variant alleles, those with one, two, and three variant alleles had HRs of 0.74, 0.64, and 0.55, respectively (Ptrend = .006; Table 4). There were no patients with a sum of four variant alleles. Adjusting for age, sex, stage, and smoking, these adjusted HRs were 0.76, 0.61, and 0.61, respectively (Ptrend = .006; Table 4). Five-year OS for those with three variant alleles was 73% (95% CI, 47% to 88%) as compared with 64% (95% CI, 51% to 74%), 58% (95% CI, 50% to 65%), and 50% (95% CI, 41% to 58%) for those with two, one, and zero variant alleles, respectively (P = .04; Table 4).
We found that carrying the variant C allele of the VEGF +405G>C polymorphism was associated with significantly improved survival among early-stage NSCLC patients treated with surgical resection. In addition, there was a trend toward improved survival among patients carrying the variant T allele of the VEGF +936C>T polymorphism. Looking at combined effects, increasing numbers of variant alleles in the +936C>T and +405G>C polymorphisms were associated with significantly improved survival. This finding is biologically plausible, especially in light of the putative function of these polymorphisms.13-16 The +936T allele has been associated with decreased VEGF plasma levels and is thought to lead to loss of a potential binding site for transcription factor activator protein-4.13 The variant C allele of the +405G>C polymorphism has also been associated with lower VEGF production, and the presence of a C allele may disrupt the MZF1 transcription factor binding site.16 Stevens et al15 investigated functional consequences of haplotypes containing the VEGF 460/405 polymorphisms, and found that some haplotypes containing the 405G allele are high VEGF expressors, with the effect appearing to be dependent on collinearity with other 5' sequence polymorphisms. It should be noted, however, that the exact function of the +405G>C polymorphism is still debated, and some groups have reported higher VEGF level or no association with the +405C/C genotype.17,18 VEGF polymorphisms have been studied in relation to risk and histopathologic characteristics of various cancers with disparate results.19-27 The +936T allele has been associated with decreased risk of certain cancers including breast cancer,19-20 although reports have been conflicting.21,22 Studies have been conflicting whether the +405C allele is associated with a lower or higher tumor grade or aggressiveness in a variety of cancers, including melanoma, prostate, and breast cancer.21,23,24 Studies investigating the VEGF polymorphisms in relation to survival outcomes have been few and not consistent in their findings. Some studies have suggested worse survival with the VEGF +405C/C genotype in gastric and ovarian cancer.26,27 Tzanakis et al26 reported decreased OS with the +405C/C genotype; however, this study included all stages of gastric cancer with 60 of 100 patients having stage III or IV disease, and no information on treatment characteristics such as surgery or chemotherapy. Hefler et al27 also included all stages of ovarian cancer, and reported that the combination of +405C/C with two other polymorphisms in VEGF was associated with worse survival, but there was no association with the polymorphism by itself. In a study of more than 1,000 women with all stages of breast cancer, Lu et al28 found that the +405G>C polymorphism was significantly associated with survival, with the +405C/C genotype having the best survival, which is in agreement with our results. All of these studies included all stages of cancer, but little detailed information on treatment. Most recently, in colorectal cancer, Lurje et al29 found that patients carrying the variant T allele of the VEGF +936C>T polymorphism had significantly improved time to recurrence. This finding is in keeping with our results as well, where patients with the variant T allele of the VEGF +936C>T polymorphism seemed to have better survival. To our knowledge, there are no published studies of VEGF polymorphisms and lung cancer outcomes. We found improved survival among patients carrying the variant C allele of the VEGF +405G>C polymorphism, and a trend toward improved survival among those carrying the variant T allele of the VEGF +936C>T polymorphism. Those with an increasing number of variant alleles in these polymorphisms had improved survival, with 5-year survival of 73% in the highest variant allele group as compared with 50% in the lowest variant allele group. Survival differences on the order seen here, of a difference in 5-year survival of 10% for the VEGF +405 polymorphism, or of approximately 20% for a combined effect of VEGF +405 and +936, would be clinically meaningful in lung cancer, for which we routinely administer adjuvant chemotherapy for absolute survival benefits of 5% to 15%. Our findings are consistent with the published literature on function of the +936C>T polymorphism, and with reports of Watson et al16 and Stevens et al15 on the function of the +405G>C polymorphism. With lower VEGF levels, angiogenic capability of these early-stage tumors potentially would be limited, leading to improved overall survival. However, attributing functional causality to these findings should be limited, as there is still debate regarding the true function of some of these polymorphisms. Strengths of this study include its large sample size and detailed data regarding patients, tumor characteristics, treatment characteristics, and survival. However there are several limitations to this study. First is the choice of overall survival as the end point. In some ways this can be regarded as a strength, given that overall survival is a concrete verifiable end point that we were able to ascertain reliably for our patients using the resources described in our Methods section. Although time to recurrence would be an interesting end point to analyze, patients were not on a predefined surveillance schedule and were observed at the discretion of the treating physician. Therefore, although we have tried to collect recurrence data on these patients, there are limitations to the validity of investigating this end point due to the lack of predefined surveillance plan and schedule. OS is a much more stable end point and reliably measured by the methods we describe in Patients and Methods. A second limitation of our study is the fact that we took a candidate gene approach and selected only a few VEGF polymorphisms for study. Although such an approach benefits from being hypothesis-driven based on biologic plausibility, there are notable limitations. It is possible that the observed association is not due to the candidate polymorphism chosen, but rather to a linked polymorphism. In addition, there is debate about the true function of these polymorphisms. VEGF is an extremely polymorphic gene, and incorporation of other described polymorphisms as well as haplotype-tagging single nucleotide polymorphisms (SNPs) would be useful to characterize more fully the variation within VEGF. Finally, there are multiple other genes in addition to VEGF that play a role in angiogenesis, which is a complex interplay of many activating and inhibitory factors.30 A comprehensive pathway-based approach to angiogenesis would incorporate not only VEGF, but also polymorphisms in the VEGF receptors, as well as other non—VEGF-mediated angiogenic factors. The results presented in this article provide a firm basis and rationale on which to pursue these future studies. Other limitations of our study include the fact that the most significant effect was noted in the heterozygous G/C genotype of the VEGF +405 polymorphism. However, the C/C homozygous variant genotype had the same direction of effect, and the magnitude of the HR values is not dissimilar, possibly because of small number in the latter group. It is also not clear why the same effect was not seen in the VEGF –460 polymorphism when it is closely linked with the VEGF +405 polymorphism. It is possible that the true effect is due to a polymorphism that is highly linked to the +405 polymorphism, but not highly linked to the –460 polymorphism, rather than the +405 polymorphism itself. A potential role for the —460 polymorphism is not excluded by our findings, however, and it is also possible that a larger study would further elucidate this. This is the first report to our knowledge of VEGF polymorphisms and survival outcome in early-stage NSCLC. These results should be confirmed in independent, larger studies, as well as expanded to cover more completely the entire angiogenic pathway. Correlation with tumor microvessel density and VEGF levels in future studies would be warranted. The potential clinical implications of our findings are manifold. Angiogenesis is of great interest in lung cancer, particularly because antiangiogenic drugs are in active clinical use. Bevacizumab, a monoclonal antibody to VEGF, is currently being tested in the adjuvant setting in a large ongoing cooperative group trial, Eastern Cooperative Oncology Group trial 1505. This prospective clinical trial will randomly assign 1,500 patients with resected early-stage lung cancer to adjuvant platinum-based chemotherapy or adjuvant platinum-based chemotherapy with bevacizumab. Of interest will be whether VEGF or angiogenic pathway polymorphisms can be used to predict benefit from bevacizumab in the adjuvant setting. The ability to collect detailed and accurate recurrence and overall survival data in the prospective clinical trial setting will be of great importance in validating the findings in this study. VEGF polymorphisms should also be investigated in advanced-stage patients, in whom antiangiogenic treatment, whether with bevacizumab or small molecule inhibitors of VEGF receptors, are in active clinical use or testing. In summary, polymorphisms in VEGF may be associated with survival in early-stage resected NSCLC. Our findings should spur additional investigation of polymorphisms in VEGF as well as the entire angiogenic pathway in relation to lung cancer outcomes.
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a "U" are those for which no compensation was received; those relationships marked with a "C" were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: David C. Christiani, Gentra (C) Stock Ownership: None Honoraria: None Research Funding: None Expert Testimony: None Other Remuneration: None
Conception and design: Rebecca Suk Heist, Geoffrey Liu, Wei Zhou, David C. Christiani Financial support: Thomas J. Lynch, David C. Christiani Administrative support: Thomas J. Lynch, David C. Christiani Provision of study materials or patients: Rebecca Suk Heist, Li Su, Thomas J. Lynch, John C. Wain, David C. Christiani Collection and assembly of data: Rebecca Suk Heist, Rihong Zhai, Li Su, Kofi Asomaning, Thomas J. Lynch, John C. Wain, David C. Christiani Data analysis and interpretation: Rebecca Suk Heist, Rihong Zhai, Geoffrey Liu, Wei Zhou, Xihong Lin, Li Su, Kofi Asomaning, Thomas J. Lynch, John C. Wain, David C. Christiani Manuscript writing: Rebecca Suk Heist, Rihong Zhai, Geoffrey Liu, Wei Zhou, Xihong Lin, Li Su, Kofi Asomaning, Thomas J. Lynch, John C. Wain, David C. Christiani Final approval of manuscript: Rebecca Suk Heist, Rihong Zhai, Geoffrey Liu, Wei Zhou, Xihong Lin, Li Su, Kofi Asomaning, Thomas J. Lynch, John C. Wain, David C. Christiani
Supported by National Institutes of Health Grants No. CA074386 (D.C.C.), CA092824 (D.C.C.), CA090578 (D.C.C.), and K12CA087723 (R.S.H.), and the Flight Attendants Medical Research Institute Young Clinical Scientist Award (W.Z.). Terms in blue are defined in the glossary, found at the end of this article and online at www.jco.org. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Carmeliet P, Jain RK: Angiogenesis in cancer and other diseases. Nature 407:249-257, 2000[CrossRef][Medline] 2. Jain RK: Tumor angiogenesis and accessibility: Role of vascular endothelial growth factor. Semin Oncol 29:3-9, 2002 (suppl 16)[Medline] 3. Ferrara N, Alitalo K: Clinical applications of angiogenic growth factors and their inhibitors. Nat Med 5:1359-1364, 1999[CrossRef][Medline] 4. Asano M, Yukita A, Matsumoto T, et al: Inhibition of tumor-growth and metastasis by an immunoneutralizing monoclonal antibody to human vascular endothelial growth factor vascular permeability factor. Cancer Res 55:5296-5301, 1995 5. Kim KJ, Li B, Winer J, et al: Inhibition of vascular endothelial growth factor induced angiogenesis suppresses tumor growth in vivo. Nature 362:841-844, 1993[CrossRef][Medline] 6. Stefanou D, Goussia AC, Arkoumani E, et al: Expression of vascular endothelial growth factor and the adhesion molecule E-cadherin in non-small cell lung cancer. Anticancer Res 23:4715-4720, 2003[Medline] 7. Oshika Y, Nakamura M, Tokunaga T, et al: Expression of cell-associated isoform of vascular endothelial growth factor 189 and its prognostic relevance in non-small cell lung cancer. Int J Oncol 12:541-544, 1998[Medline] 8. Fontanini GM, Mussi A, Angeletti CA, et al: Expression of vascular endothelial growth factor mRNA in non-small cell lung carcinomas. Br J Cancer 79:363-369, 1999[Medline] 9. O'Bryne KJ, Koukourakis MI, Giatromanolaki A, et al: Vascular endothelial growth factor, platelet-derived endothelial growth factor and angiogenesis in non-small cell lung cancer. Br J Cancer 82:1427-1432, 2000[Medline] 10. Fontanini G, Lucchi M, Vignati S, et al: Angiogenesis as a prognostic indicator of survival in non-small cell lung carcinoma: A prospective study. J Natl Cancer Inst 89:881-886, 1997 11. Yuan A, Yu CJ, Kuo SH, et al: Vascular endothelial growth factor 189 mRNA isoform expression specifically correlates with tumor angiogenesis, patient survival, and postoperative relapse in non-small cell lung cancer. J Clin Oncol 19:432-441, 2001 12. Han H, Silverman JF, Santucci TS, et al: Vascular endothelial growth factor expression in stage I non-small cell lung cancer correlates with neoangiogenesis and a poor prognosis. Ann Surg Oncol 8:72-79, 2001[Medline] 13. Renner W, Kotschan S, Hoffman C, et al: A common 936 C/T mutation in the gene for vascular endothelial growth factor is associated with vascular endothelial growth factor plasma levels. J Vasc Res 37:443-448, 2000[CrossRef][Medline] 14. Zhai R, Gong MN, Thompson TB, et al: Genotypes and haplotypes of VEGF gene are associated with higher ARDS mortality and lower VEGF plasma levels. Thorax 62:718-722, 2007 15. Stevens A, Soden J, Brenchely PE, et al: Haplotype analysis of the polymorphic human vascular endothelial growth factor gene promoter. Cancer Res 63:812-816, 2003 16. Watson CJ, Webb NJA, Bottomley MJ, et al: Identification of polymorphisms within the vascular endothelial growth factor (VEGF) gene: Correlation with variation in VEGF protein production. Cytokine 12:1232-1235, 2000[CrossRef][Medline] 17. Awata T, Inoue K, Kurihara S, et al: A common polymorphism in the 5' untranslated region of the VEGF gene is associated with diabetic retinopathy in type 2 diabetes. Diabetes 51:1635-1639, 2002 18. Balasubramanian SP, Cox A, Cross SS, et al: Influence of VEGF-A gene variation and protein levels in breast cancer susceptibility and severity. Int J Cancer 121:1009-1016, 2007[CrossRef][Medline] 19. Krippl P, Langsenlehner U, Renner W, et al: A common 936 C/T gene polymorphism of vascular endothelial growth factor is associated with decreased breast cancer risk. Int J Cancer 106:468-471, 2003[CrossRef][Medline] 20. Kataoka N, Cai Q, Wen W, et al: Population based case control study of VEGF gene polymorphisms and breast cancer risk among Chinese women. Cancer Epidemiol Biomarkers Prev 15:1148-1152, 2006 21. Jin Q, Hemminki K, Enquist K, et al: Vascular endothelial growth factor polymorphisms in relation to breast cancer development and prognosis. Clin Cancer Res 11:3647-3653, 2005 22. Jacobs EJ, Feigelson HS, Bain EB, et al: Polymorphisms in the vascular endothelial growth factor gene and breast cancer in the Cancer Prevention Study II cohort. Breast Cancer Res 8:R22, 2006[CrossRef][Medline] 23. Howell WM, Bateman AC, Turner SJ, et al: Influence of vascular endothelial growth factor single nucleotide polymorphisms on tumour development in cutaneous malignant melanoma. Genes Immun 3:229-232, 2002[CrossRef][Medline] 24. Sfar S, Hassen E, Saad H, et al: Association of VEGF genetic polymorphisms with prostage cancer risk and clinical outcome. Cytokine 35:21-28, 2006[CrossRef][Medline] 26. Tzanakis N, Gazouli M, Rallis G, et al: Vascular endothelial growth factor polymorphisms in gastric cancer development, prognosis, and survival. J Surg Oncol 94:624-630, 2006[CrossRef][Medline] 27. Hefler LA, Mustea A, Konsgen D, et al: Vascular endothelial growth factor polymorphisms are associated with prognosis in ovarian cancer. Clin Cancer Res 13:898-901, 2007 28. Lu H, Shu XO, Kataoka N, et al: Association of genetic polymorphisms in the VEGF gene and breast cancer survival. Cancer Res 65:5015-5019, 2005 29. Lurje G, Schultheis AM, Hendifar AE, et al: VEGF and VEGFR2 gene polymorphisms predict tumor recurrence in stage II and III colon cancer. J Clin Oncol 25:164s, 2007 (suppl; abstr 4004) 30. Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 86:353-364, 1996[CrossRef][Medline] Submitted July 20, 2007; accepted October 10, 2007.
Related Correspondence
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 tests (for categoric variables) and Kruskal-Wallis tests (for continuous variables), where appropriate. The associations between VEGF polymorphism status and survival were estimated using the method of Kaplan and Meier and assessed using the log-rank test. Cox regression models were used to adjust for potential confounders, with VEGF genotypes fitted as indicator variables. Joint effects of polymorphisms were analyzed based on the number of variant alleles, where the wild-type genotype has zero variant allele, heterozygous genotype has one variant allele, and homozygous variant genotype has two variant alleles. All statistical testing was done at the two-sided .05 level, and SAS software version 9.1 (SAS Institute, Cary, NC) was used. 
