Emj. Schutter et al., Urinary gonadotropin peptide (UGP) and serum CA 125 in gynaecologic practice, a clinical prospective study, ANTICANC R, 19(6C), 1999, pp. 5551-5557
Background: Beta human chorionic gonadotropin (beta-hCG) is expressed in hu
man fetal tissue and cancer cells of various histologic types. It is degrad
uated to the beta-core fragment (beta cf-hCG) which is concentrated in urin
e, and is known urinary gonadotropin peptide (UGP). The objective of this s
tudy was to assess 1) the value of urinary gonadotropin peptide (UGP) as a
single test and the combination of UGP with CA 125 as a diagnostic test in
predicting the benign or malignant origin of gynecologic disease, 2) the in
fluence of surgical removal of the tumor on the levels of UGP and CA 125, 3
) the influence of the urinary concentration oil the UGP levels in relation
to the test results. Patients, Materials, Methods and Statistics: Serum an
d urine were collected from 31 gynecological patients with malignant and no
n-malignant disease, preoperatively: and 1 week and 6 weeks after surgery.
Optimal cut-off levels were determined by Receiver Operating Characteristic
- curves (ROC). Sensitivity (SENS), specificity (SPEC), positive (PPV) and
negative predictive value (NPV) and overall rest accuracy (ACC) for their a
bility to discriminate benign fi from malignant masses were calculated Logi
stic regression analysis was performed to calculate the contribution of CA
125, UGP and UGP/creatinine (UGP/creat) to a model predicting malignancy. R
esults: The optimal cut-off level for UGP was found 1 fmol/l, for UGP/creat
1.33 fmol/mg creatinine and for CA 125 100 kU/L. The distribution of the u
rinary creatinine values varied considerably (median = 8.3 mmol/l, range 0.
6 - 25.8 mmol/l). The correlation coefficient (r) between log UGP and log C
A 125 was 0.44 (p = 0.001) and between log UGP/creat and log CA 125 0.53 (p
< 0.0001). Conclusions: UGP may be used as a tumor maker in gynecological
disease. However, CA 125 as single test discriminates malignant from benign
disease better than UGP or UGP/creat. In a logistic model the combination
of CA 125 with UGP and UGP/creat predicts the benign ol malignant character
in 89% of the study population. Significant changes in UGP and UCP/creat l
evels were seen after removal of benign tumors, however; this was not found
in ovarian cancer patients. Correction of the UGP values for the urinary c
oncentration improved the results slightly.