Serum prolactin was measured pretherapeutically and sequentially there
after using immunoradiometric assay method in 37 male patients with ad
vanced tongue cancer and compared with 23 healthy, age-matched control
s. Prolactin levels were correlated with age, various clinicopathologi
c parameters, overall survival, and patients with response and those w
ith progressive disease. Patients with advanced tongue cancer had high
er prolactin levels than controls (P < 0.02), but intergroup variation
in prolactin was not observed when considering the age, site of the l
esion, disease stage, histologic grade, and keratin. Of the patients,
30% had hyperprolactinemia (prolactin >15.0 ng/ml). To assess the prog
nostic significance of pretherapeutic prolactin level, the patients we
re divided according to the cutoff level of prolactin (15.0 ng/ml). Hy
perprolactinemic patients had more unfavourable prognosis than patient
s with prolactin <15.0 ng/ml (X(2) = 2.91, df = 1, P < 0.0037). In mon
itoring disease course, patients who responded to treatments had decre
ased prolactin levels at the end of 18 months as compared to their pre
therapeutic levels (P < 0.01). In patients who subsequently developed
progressive disease within 18 months, prolactin levels reduced initial
ly at response, whereas with disease progression, prolactin levels inc
reased significantly (P < 0.05). The positive and negative predictive
value of prolactin was 100%. Immunohistochemical localization confirme
d the ectopic production of prolactin by tongue tumors. In conclusion,
our data indicate that hyperprolactinemia may be an independent predi
ctor of short-term prognosis; circulating prolactin may be used as a m
arker for monitoring disease course in patients with advanced tongue c
ancer, and prolactin is produced ectopically by tongue tumors. (C) 199
4 Wiley-Liss, Inc.