Ba. Goff et al., IMMUNOPHENOTYPE OF OVARIAN-CANCER AS PREDICTOR OF CLINICAL OUTCOME - EVALUATION AT PRIMARY SURGERY AND 2ND-LOOK PROCEDURE, Gynecologic oncology (Print), 70(3), 1998, pp. 378-385
Objective. This study was undertaken to evaluate whether immunophenoty
ping of advanced epithelial ovarian cancer could predict response to i
nitial chemotherapy and whether tumor immunophenotype changed after ch
emotherapy. Study design. Fifty-four patients with stage III and IV ov
arian cancer, treated at the University of Washington Medical Center,
had pathology specimens evaluated. A subset of 23 patients also had sp
ecimens from a secondary surgery evaluated. Using immunocytochemistry,
tumors were immunostained for overexpression of c-erb-B-2, epidermal
growth factor receptor (EGFR), p53, and expression of the Ki67-defined
antigen (a marker of cellular proliferation), tumor necrosis factor a
lpha (TNF alpha), estrogen receptor (ER), progesterone receptor (PR),
and P-glycoprotein (P170, a marker of multidrug resistance). Twenty-fo
ur patients had a good response to chemotherapy (defined as a negative
, or microscopically positive second look), and 30 had a poor response
(defined as grossly positive second look or progressive disease). Res
ults. Comparison of tumor markers from the initial and the secondary s
urgeries revealed that the only significant change was in the Ki67-def
ined cell proliferation rate, which showed a marked reduction in those
with a good response to chemotherapy (P = 0.002). Comparison of tumor
markers at initial surgery between good and poor responders revealed
a correlation with p53 expression. Good responders were less likely to
have p53 overexpression compared to poor responders, and this result
approached significance (P = 0.058). Comparison of tumor markers at se
condary surgery revealed a significant reduction in Ki67-defined cell
proliferation rate in good responders compared to poor responders (P =
0.01). No significant differences were found between good and poor re
sponders for the other tumor markers evaluated. Conclusions. The only
tumor markers to predict for response to chemotherapy were p53 at init
ial surgery (P = 0.058) and Ki67 indices at secondary surgery (P = 0.0
01). Expression of steroid hormone receptors, TNF alpha, and P-glycopr
otein and overexpression of c-erb-B-2 or EGFR are not associated with
chemoresistance. (C) 1998 Academic Press.