Objective: To study the correlation between serum and urine hCG levels in gestational trophoblastic disease (GTD) patients
with low hCG level. The correlation between serum hCG and results from urine pregnancy tests are evaluated along with
quantitative measurement.
Material and Method: In this prospective study, 86 cases of gestational trophoblastic disease patients with positive and low
level of serum hCG (<100 mIU/ml) were recruited. Quantitative serum hCG, urine hCG and urine pregnancy test were performed. The correlation coefficients between serum and urine hCG were then analyzed by SPSS 16.0. Furthermore, the
levels of serum hCG were compared to the results of the urine pregnancy test.
Results: From February 2006 to June 2008, 86 cases were recruited for this study. The correlation coefficient between serum and urine hCG levels in all cases was 0.44 (using Pearson correlation), p = 0.01. In subgroup analysis, the correlation
coefficient between serum and urine hCG levels in chemosensitive gestational trophoblastic neoplasia (GTN) patients (n = 27) was 0.73, p<0.01. The correlation coefficient in chemoresistant GTN patients (n = 38) was 0.29, p = 0.07; and the correlation coefficient in hydatidiform mole patients (n = 21) was 0.47, p = 0.03. A urine pregnancy test was positive only in 10 of 86 specimens.
Conclusion: The correlation coefficient between serum and urine hCG in GTD patients with low hCG level showed significant
correlation. However, patients with chemoresistance had less correlation than those with chemosensitivity and hydatidiform
mole. Urine pregnancy test had low correlation with urine hCG and was not useful in this group of patients.

Keywords: Gestational trophoblastic disease, Urine hCG, Human chorionic gonadotropin