Partial hydatidiform moles have impaired differentiated function (human chorionic gonadotropin and human placental lactogen secretion) in response to epidermal growth factor and 8-bromo-cyclic adenosine monophosphate Donald W. Morrish, MD, PhD,. Louis H. Honore, MB: and Damyanti Bhardwaj, MD" Edmonton, Alberta, Canada OBJECTIVE: The null hypothesis is that partial hydatidiform moles have normal differentiated function (human chorionic gonadotropin and human placental lactogen secretion) in response to epidermal growth factor and 8-bromo-cyclic adenosine monophosphate. STUDY DESIGN: Two complete moles, 10 partial hydatidiform moles, and 19 normal first-trimester placentas In monolayer culture were exposed to 10 ng/ml epidermal growth factor, 1 mmol/L 8-bromocyclic adenosine monophosphate plus 1 mmol/L theophylline, or control. Human chorionic gonadotropin and human placental lactogen secretion was measured. Frequency of response to stimuli was compared by x2 analysis, and hormone secretion was compared by analysis of variance. RESULTS: Partial moles demonstrated reduced frequencies of response of human chorionic gonadotropin and human placental lactogen to epidermal growth factor (partial moles 2/8 and 2/8, respectively; normal placentas 16/19 and 7118, respectively; p < 0.025) and of human chorionic gonadotropin to 8-bromocyclic adenosine monophosphate (partial moles 3/5, normal placentas 13/16; p < 0.005). CONCLUSION: Partial hydatidiform moles demonstrate impaired human chorionic gonadotropin and human placental lactogen secretory responsiveness to epidermal growth factor and cyclic nucleotides in comparison with normal first-trimester trophoblast. (AM J OBSTET GVNECOL 1992;166:160-6.)

Key words: Hydatidiform mole, human chorionic gonadotropin, human placental lactogen, epidermal growth factor, growth factors

Partial and complete hydatidiform moles, which are characterized by inappropriate hyperplasia of the villous tissue, demonstrate a variety of morphologic, antigenic, cytogenetic, and hormonal abnormalities. 1-13 Although ~-human chorionic gonadotropin (hCG) has been most useful as a disease marker, no information is available regarding the regulation of hCG or other hormones produced by molar tissue. Therefore we wished to determine if there was abnormal regulation of hCG and human placental lactogen (hPL) secretion in response to epidermal growth factor (EGF) and 8bromo-cyclic adenosine monophosphate (8-bromocAMP), two known stimulators of hormone secretion in normal trophoblast. From the Departments of MediCine" and Patholo/sy,' University of Alberta. Supported by the Medical Research Council of Canada. Received for publication December 18, 1990; reVISed April 22, 1991 ; accepted May 8, 1991. Reprint requests: Donald W. Morrish, MD, 362 Clinical Wing, Heritage Medical Research Centre, Department of Medzcine, University of Alberta, Edmonton, Alberta, Canada T6G 2S2. 6/1/30965

160

Material and methods The study protocol was approved by the institutional ethics review board. Normal first-trimester placentas and hydatidiform mole tissue from spontaneous or therapeutic abortions were obtained and transported in sterile saline solution to the laboratory. Membranes were removed and the entire piece of placenta without dissection was then subjected to eight sequential 10minute treatments with 0.25% trypsin II 0 VI ml deoxyribonuclease I as previously described. 14 Cells were collected after the third to eighth treatment, pooled, centrifuged at 300g, and resuspended in Dulbecco's modified Eagle medium containing 10% fetal calf serum and 50 Vlml penicillin, 50 f,Lg/ml streptomycin, and 0.2 mmollL glutamine. Cells were plated in 60 mm Petri dishes and allowed to attach for 2 hours, after which the medium was changed to Dulbecco's modified Eagle medium without serum for the remainder of the experiment (7 to 9 days). Medium was changed daily and frozen at - 20° C for subsequent hCG and hPL assay. After each medium change, plates received either 10 ng/ml EGF or bovine serum albumin (final concen-

\ 'O llll11