Serum estrogen and tumor-positive estrogen receptor-alpha are strong prognostic classifiers of non-small-cell lung cancer survival in both men and women

Abstract

The role of tumor estrogen receptors (ERs) and serum estrogen in lung cancer is inconclusive. We investigated the hypothesis that ERs and functional single-nucleotide polymorphisms in the estrogen biosynthesis pathway are associated with poorer lung cancer survival. Lung cancer patients (n = 305) from a National Cancer Institute-Maryland (NCI-MD) case-case cohort in the Baltimore metropolitan area were used as a test cohort. To validate, 227 cases from the NCI-MD case-control cohort and 293 cases from a Norwegian lung cancer cohort were studied. Information on demographics, tobacco and reproductive histories was collected in an interviewer-administered questionnaire. Serum estrogen, progesterone, tumor messenger RNA expression of hormone receptors and germ line DNA polymorphisms were analyzed for associations with lung cancer survival. Patients in the highest tertile of serum estrogen had worse survival in all three cohorts (P combined < 0.001). Furthermore, the variant allele of estrogen receptor alpha (ER-α) polymorphism (rs2228480) was significantly associated with increased tumor ER-α levels and worse survival in all three cohorts [hazard ratio (HR) = 2.59, 95% confidence interval (CI): 1.20- 4.01; HR = 1.76, 95% CI: 1.08-2.87 and HR = 2.85, 95% CI: 1.31-4.36). Other polymorphisms associated with lower serum estrogen correlated with improved survival. Results were independent of gender and hormone replacement therapy. We report a significant association of increased serum estrogen with poorer survival among lung cancer male and female patients. Understanding the genetic control of estrogen biosynthesis and response in lung cancer could lead to improved prognosis and therapy.

Fig. 1.

The estrogen biosynthesis pathway. The pathway from cholesterol through estrogen synthesis and metabolism is shown with the enzymes involved. Bolded enzymes are those in which SNPs were investigated in relation to estrogen measures and lung cancer survival (20).

Fig. 2.

Kaplan–Meier 5 years of cancer survival curves for the association of serum estrogen with lung cancer survival. Serum estrogen is categorized by tertiles and the P-values were calculated for differences in survival curves using the log-rank test. (A) Serum estrogen and lung cancer survival in the NCI-MD case–case cohort (P = 0.02). (B) Serum estrogen and lung cancer survival in the NCI-MD case–control cohort (P < 0.001). (C) Serum estrogen and lung cancer survival in the Norwegian case-only cohort (P < 0.001). (D) Serum estrogen and lung cancer survival in all three cohorts combined (P < 0.001).

Fig. 3.

Kaplan–Meier 5 years of cancer survival curves for the association of ESR1_07 with lung cancer survival. The survival rates of patients with G/G genotypes were compared with patients with G/A genotypes and A/A genotypes. Differences between survival curves were calculated using the log-rank test. (A) ESR1_07 genotypes associated with lung cancer survival in the NCI-MD case–case cohort (P = 0.02). (B) ESR1_07 genotypes associated with lung cancer survival in the NCI-MD case–control cohort (P = 0.05). (C) ESR1_07 genotypes associated with lung cancer survival in the Norwegian case-only cohort (P = 0.03). (D) ESR1_07 genotypes associated with lung cancer survival in all three cohorts combined (P = 0.04).