p53 mutations and survival in stage I non-small-cell lung cancer: Results of a prospective study

Steven A. Ahrendt, Yingchuan Hu, Martin Buta, Michael P. McDermott, Nicole Benoit, Stephen C. Yang, Li Wu, David Sidransky

Research output: Contribution to journalArticle

Abstract

Background: The p53 gene is frequently mutated in nonsmall-cell lung cancer (NSCLC); however, the effect of p53 gene mutations on patient prognosis remains unclear. Therefore, we initiated a prospective study to determine the association of p53 gene mutations with survival in patients with stage I NSCLC. Methods: Tumor samples were collected prospectively from 188 patients with operable NSCLC (stages I, II, and IIIA). p53 mutations were detected by direct dideoxynucleotide sequencing and p53 GeneChip analysis. Association of clinical and pathologic variables (e.g., alcohol consumption, sex, age, pathologic stage) with mutation of the p53 gene was determined by logistic regression. Associations between p53 mutation status, clinical and pathologic variables, and survival were assessed using a Cox proportional hazards regression model. All statistical tests were two-sided. Results: p53 mutations were detected in 55% (104/188) of tumors. These mutations were associated with nonbronchoalveolar tumors, a history of alcohol consumption, and younger patient age. The risk of death was statistically significantly higher in patients with p53 mutations in their tumors (hazard ratio [HR] = 1.6, 95% confidence interval [CI] = 1.0 to 2.4; P =.049) than in patients with wild-type p53 in their tumors. Tumor stage, the presence of a p53 mutation, and increasing patient age were statistically significant predictors of patient death in the entire patient group; however, the statistically significant prognostic effect of p53 mutation was limited to patients with stage I NSCLC (stage I HR = 2.8, 95% CI = 1.4 to 5.6; stage II HR = 1.8, 95% CI = 0.74 to 4.4; and stage III HR = 0.70, 95% CI = 0.32 to 1.5). Among patients with stage I NSCLC, actuarial 4-year survival was statistically significantly higher in those with wild-type p53 than in those with mutant p53 (78% versus 52%, respectively; difference in 4-year survival = 26%, 95% CI = 6% to 46%; P =.009, log-rank test). Conclusion: Tumor p53 mutations are a statistically significant predictor of poor outcome in patients with stage I NSCLC.

Original languageEnglish (US)
Pages (from-to)961-970
Number of pages10
JournalJournal of the National Cancer Institute
Volume95
Issue number13
DOIs
StatePublished - Jul 2 2003

    Fingerprint

ASJC Scopus subject areas

  • Oncology
  • Cancer Research

Cite this