Thymidylate synthase expression and response to neoadjuvant chemotherapy in patients with advanced head and neck cancer.

Abstract

Thymidylate synthase (TS), an essential enzyme in DNA synthesis, is a target for the fluoropyrimidines, an important group of antineoplastic agents used widely in the treatment of head and neck cancer. We evaluated relationships between the level and/or pattern of tumor TS expression and response to fluorouracil (5-FU)-based neoadjuvant chemotherapy in patients with advanced head and neck cancer. Tumor specimens from 86 patients were available for this retrospective analysis. The patients were enrolled in four consecutive phase II studies that tested combinations of 5-FU, leucovorin, and cisplatin with or without added methotrexate plus piritrexim or interferon alfa-2b (IFN alpha-2b). TS protein expression in the tumors was assessed by use of the TS 106 monoclonal antibody and standard immunohistochemical staining techniques. TS immunostaining was classified according to its level of intensity (TS 0-1 = low, TS 2 = intermediate, and TS 3 = high) and according to its extent (focal pattern = less than 25% of tumor cells positive; diffuse pattern = greater than or equal to 25% of tumor cells positive). Data from 79 patients were available for an analysis of tumor TS expression and patient/tumor characteristics; 70 patients were assessable for their response to neoadjuvant chemotherapy. There was a statistically significant association between the level of tumor TS expression and the degree of tumor differentiation; a higher proportion of patients whose tumors exhibited TS 0-1 immunostaining had undifferentiated or poorly differentiated tumors than patients whose tumors exhibited TS 2 or TS 3 immunostaining (P = .04, Jonckheere-Terpstra trend test). Among the 70 patients who were assessable for response to neoadjuvant chemotherapy, TS 3 tumor immunostaining was associated with a lower rate of complete response (i.e., complete disappearance of clinically detectable disease for a minimum of 4 weeks from time of initial determination) than was TS 2 or TS 0-1 immunostaining, but this association was not statistically significant (P = .09, exact trend test); among the 39 patients who were treated with regimens that included 5-FU, leucovorin, cisplatin, and IFN alpha-2b, this inverse association between TS immunostaining intensity and response was statistically significant (P = .02, exact trend test). Tumor TS immunostaining intensity and overall survival were not found to be associated. Patients with tumors exhibiting a focal pattern of TS immunostaining have experienced significantly longer survival than patients with tumors exhibiting a diffuse pattern; for the 53 patients with diffuse tumor TS immunostaining, the median survival was 24.7 months, whereas the median survival has not yet been reached for the 22 patients with focal tumor TS immunostaining (P = .04, two-tailed logrank test). However, the survival advantage for the focal versus diffuse TS immunostaining pattern was limited to patients whose tumors also exhibited a TS 3 level of immunostaining intensity. Characterization of tumor TS expression may be of value in identifying patients with advanced head and neck cancer who would benefit from fluoropyrimidine-based neoadjuvant chemotherapy.