THE EPIDERMAL GROWTH-FACTOR RECEPTOR AS A PROGNOSTIC MARKER - RESULTS OF 370 PATIENTS AND REVIEW OF 3009 PATIENTS

Epidermal growth factor receptor (EGFR) and estrogen receptor (ER) were assayed by ligand binding in tumors from 370 patients with primary breast carcinoma with a median follow up of 18 months. Forty seven percent (175/370) and 57% (210/370) of tumors had >20 fmol/mg and >10 fmol/mg of EGFR and ER respectively. There was a highly significant inverse relationship between EGFR and ER (p=0.0032). There was also a significant association between EGFR and patient age (p=0.0006) but no correlation between EGFR and lymph node status, tumor grade, or tumor size (p=0.104, p=0.198, and p=0.085 respectively). In a univariate analysis of all patients, EGFR expression was not associated with a significant reduction in overall survival (OS). However, there was a significant decrease in relapse-free survival (RFS) and OS in node negative EGFR positive patients (p=0.03 and p=0.05 respectively). In a multivariate analysis (Cox proportional hazard model) of all patients, lymph node status was an independent prognostic indicator for OS and RFS (p<0.00005 and p=0.00005 respectively), ER status for RFS (p=0.0006), and EGFR (in the node negative model) for RFS (p=0.03). When all patients were stratified for EGFR and ER, there was a significant difference in RFS and OS such that EGFR positive and ER negative had the worst prognosis (p=0.0034 and p=0.005 respectively). A similar relationship was observed for OS in node negative patients (p=0.004) and for RFS in node positive patients (p=0.009). In a review of 3009 patients with follow-up, 11/16 series showed high EGFR was associated with shorter RFS or OS in univariate analysis, and 4 showed this in multivariate analysis. However, most series had inadequate follow-up time and most did not include multivariate analysis. This highlights the need for uniform criteria of reporting trials of prognostic factors.