LEVELS OF TGF-ALPHA AND EGFR PROTEIN IN HEAD AND NECK SQUAMOUS-CELL CARCINOMA AND PATIENT SURVIVAL

Background: The most accurate predictor of disease recurrence in patie nts treated for head and neck squamous cell carcinoma is, at present, the extent of regional lymph node metastasis, Since elevated levels of epidermal growth factor receptor (EGFR) and of its ligand, transformi ng growth factor-alpha. (TGF-alpha), have been detected in primary tum ors of patients with head and neck squamous cell carcinoma, we determi ned whether tumor levels of these proteins were of prognostic importan ce. Methods: Monoclonal antibodies specific for EGFR and TGF-alpha wer e used for immunohistochemical detection of each protein in tissue sec tions of primary tumors from 91 patients who were treated by surgical resection, Levels of immunoreactive EGFR and TGF-alpha were quantified by use of a computerized image analysis system and were normalized to appropriate standards. The logrank test and proportional hazards regr ession analysis were used to calculate the probability that EGFR and T GF-alpha levels were associated with disease-free survival(i.e., no re currence of cancer) and cause-specific survival (i.e., patients do not die of their disease). All P values were two-sided. Results: When tum or levels of EGFR or TGF-alpha were analyzed as continuous variables, disease-free survival and cause-specific survival were reduced among p atients with higher levels of EGFR (both P = .0001) or TGF-alpha (both P = .0001), In a multivariate analysis, tumor site, tumor level of EG FR, and tumor level of TGF-ar were statistically significant predictor s of disease-free survival; in a similar analysis, regional lymph node stage and tumor levels of EGFR and of TGF-alpha were significant pred ictors of cause-specific survival. Conclusion: Quantitation of EGFR an d TGF-alpha protein levels in primary head and neck squamous cell carc inomas may be useful in identifying subgroups of patients at high risk of tumor recurrence and in guiding therapy.