Objective: To evaluate spontaneous regression curves of hCG serum positivity in patients with surgically treated molar pregnancies. Comparison of complete, partial and invasive mole. The study should result in optimalisation of follow up criteria of molar pregnancies in respect to their potential malignant change. Design: Retrospective comparative clinical study. Setting: Obst. Gyn. Dpt., Oncogynecology div., 2nd Medical Faculty, FNM, Charles University Prague, Pathology Dpt., 2nd Medical Faculty, Institute of Biology and Medical Genetics. Methods: Evaluation of spontaneous regression curves of serum hCG levels in 104 molar pregnancies. 46 patients with partial hydatiform mole, 48 patients with complete hydatiform mole, 10 patients with invasive mole. Serum hCG levels were detected by radioimunoassay (RIA) in the first period and imunochemoluminisent assay (LIA) in the second period. Regression curves of hCG positivity in particular moles were statistically evaluated by Fischer test and t-test. Results: There is statistically significant difference in spontaneous regression of hCG positivity in different types of molar pregnancies. Recommended criteria for gestational trofoblastic disease (GTD) diagnosis and follow up are fully applicable in clinical practice. There is exception in partial hydatiform moles, where plateau in hCG regression does not necessarily implicate chemotherapy in patient with good compliance. Conclusion: Early diagnosis of GTD predominantly due to the widespread use of ultrasonography changes classical clinical features of molar pregnancies. Spontaneous regression in hCG positivity in serum is more rapid in patients with partial hydatiform mole, slower in complete hydatiform mole and invasive mole. There is no significant change in malignant potential regarding early detection and treatment.

        Key words: partial hydatiform mole, complete, hydatidiform mole, invasive mole, regression of hCG, hCG persistancy, malignisation, follow up