Plasma prolactin, human chorionic gonadotropin, estradiol, testosterone, and progesterone in the ovarian hyperstimulation syndrome BASIL
Plasma prolactin, estradiol, progesterone, and testosterone, but not HCG-/3 levels, were higher in a patient who developed the ovarian hyperstimulation syndrome while undergoing ovulation induction with human gonadotropins than in two other women who also became pregnant after similar treatment without complications. These results suggest that hyperprolactinema, in association with elevated ovarian steroid levels, may be factors in the pathogenesis of this disorder. (AM. J. OESTET. GYNECOL. 133:316, 1979.)
1‘1i E o VA RI AN hyperstimulation syndrome remains a serious complication of ovulation induction with human chorionic gonadotropins (HCG) and, occasionally, clomiphene citrate.’ The syndrome is characterized by ovarian enlargement, ascites, hydrothorax, hypovolemia. hemoconcentration, and oliguria. In extreme cases, vascular thrombosis and death have been reported. The ovarian enlargement could be explained b) the effect of gonadotropins on the ovary. However, the pathogenesis of the systemic effects, particularly the profound third space fluid shifting phenomena observed during the acute phase of’ the illness, are not well understood. We recently observed a patient who developed the ovarian hyperstimulation syndrome. The endocrine data from this patient were contrasted with those of two other women who also conceived after human gonadotropin therapy without de\,eloping this syndrome. The findings in these patients extend previous observations on the endocrine alterations and Fwm thv Drpartment Faculty qf.Medicine,
oj Obstetrics and Gyw~~olo~~, Unizwrsity of British Cokrmhin.
Pwented at the Thirtyjourth .4nnual Meetitzg a/ The Socieq of Obstetricians and Gynaecologtits of Canada. Winnipeg. Manitoba. Canada,Junp 15-18. 197X. Reprint requests: Dr. B. Ho Yuen, Dupartmrnt O_ Ohstetric.~ a16d Gynaecology, Willow Paz&on, Varlcouzw General H~~spital, Vancouzler, Brztirh Columhza, Canada I’5Z lM9.
nlay provide additional this disorder.
insight into the pathogenesis
Methods and material Hormone radioimmunoassays (RIA). Estradiol was assayed after ether extraction of plasma without chromatography. Details of this rapid assay will be reported later. Progesterone was assayed after ether extraction and c-hn-(~~~~~~~o~~raph~on Sephadex-LH 20 micr-oconnin5 with isoortane : benzene : methanol (90: 5 : 5) ab solvent. Antisera for estradiol. progesterone. and I-SIP dioactke steroids were obtained from New England Nuclear, Lachinc, Quebec, Canada. Prolactin was assa\,ed by the double-antibody method. whereas the HCGP subunit assay employed polyethylene glycol solution to separate bound from free radioat tivit) ‘I-he materials for these polypeptide assays were obtained fiwm Serono, Boston, Massachusetts. Luteinizing hormone (LH) was assayed by a double-antibody RIA with materials obtained f’rom Bio-RIA, Montreal, Canada. .l‘he testosterone and follicle-stimulating hormone (FSH) RIA’s were performed by Dr. J. Frohlich. Division of Clinical Chemistry, \‘ancou\:er General Hospital. Patients. C. V., who developed the hype~~stimulatit~n syndrome. presented for initial evaluation at the age of 36 years in April, 1976. after 7 years of primary infertility. She experienced oligomenorrhea with irregular
Plasma hormones in ovarian hyperstimulation syndrome
menstrual cycles occurring at intervals up to 12 months apart. Unprotected sexual intercourse occurred frequently since discontinuation of oral contraceptives after- 2 months of llse in 1969. Physical and pelvic examinations were normal except for hirsutism. Hysrerosalpingogr~Ipll); and laparoscopy were normal sake for the oval-ian findings (including biopsy) which were consistenI with polycystic ovarian disease. Basal body temperatures were monophasic. Baseline plasma testosterone was 1 18 ngiml (normal = 10 to 70): estradiol, 52 pgiml (normal = 40 to 650); LH, 28 mIU/ml (normal = 2 to 30, second IRP HMG, excluding midcycle peak): FSH, 2.4 ngilnl (normal = 0.6 to 11, CEA-IRESorin). ‘rhe effective thyroxine ratio was normal. Her husband had sperm counts varying between 20 and 40 million/ml. His l&t varicocele was ligated before proceeding To ovulation induction. Clomiphene was ernployed without success in doses up to 200 mg daily for 5 days followed by HCG, 10,000 IU, 6 to 9 days after clomiphene withdra\val. The first pregnancy occurred when human menopausal gonadotropins (HMG, Pergonal. Cutter Laboratories), three ampules daily, were added on da)> ‘7. 9. and 1 I. HCG, 10,000 IU (APL, Aperst Laboratories), \vas administered on day 12. This pregnancy ended in a spontaneous abortion 45 days after commencing treatment. Two months later, clomiphene, 200 mg daily with 20 pg of ethinyl estradiol, was administered from days 1 to 5 with HMG, three ampules daily, on clays 6 to 13 and on day 15; on day 16. HC(;. 10,OC)O IL!, was given. When menstruation ensued. this course of‘ treatment was repeated, with details as shown on Table I. Ten days after HCG, the patient was admitted M.ith florid hyperstimulation syndrome. She conlplained of dyspnea associated wi,th squeezing midabdominal and retrosternal pleuritic pains, appeared ill, with a pulse rate of 144/minute, blood pressure 1 lo/HO mm Hg, and respiratory rate of ‘LX/minute. There was abdominal pain with ascites, guarding, and tender masses in both lower quadrants. Respiratory distress was secondary to abdominal distention and a large right hydrothorax. There was a 5 pound increase in body weight. On admission, the hemoglobin was 16.7 gm/lOO ml, the white cell count was 18,600, and the hematocrit was 49%. The coagulation profilr ~cas normal except for a fibrinogen level of 587 mgi 100 ml (normal = up to 500 during pregnancy). The serum electrolytes were normal. Albumin infusions of’ 50 to 100 gm (total of 350 gm) were administered to enhance plasma oncotic pressure. A series of six aspirations from the right pleural cavity, obtaining a total of 8.530 ml over 14 days, was required to prevent respiratory distress. p-Scan ultrasound examinations showed enlarged cystic ovaries, 8 by 6 and 4 by 5 cm in
Days before HCG*
I c. v.s I. S.$ Y.F. W.$.
HMGi 3 0 0
3 3 1
3 3 1
3 3 1
dose (nmpoules) 3 3 1
*HCG (APL, Ayerst Laboratories) day 0. t HMG = Pergonal, Cutter Laboratories FSH and 75 ILJ ot‘ LH per ampule).
preceded by clomiphene
20 pg daily
2 2 1
3 2 1
3 2 1 to 10,000
0 2 0 IU on 75 IU of
200 mg and ethinyl es-
x 5 days.
diameter. Subsequent scans revealed the presence of a twin pregnancy. Plasma samples for HCG-/3 subunit, estradiol, progesterone, testosterone, and prolactin assays were obtained. After the pleural effusion and acute Huid shifts had subsided over the initial 2 weeks, there was peripheral edema which also subsided spontaneously with gradual decrease in size of the enlarged ovaries. She was discharged well, 37 days after the HCG injection. She was since delivered of normal twins (a male and female) at term by elective cesarean section. Two other subjects were treated at the same time. I. S. (age 32 years) with primary infertility of 18 months’ duration had polycystic ovarian disease. Her baseline testosterone level was 114 ngilO0 ml, LH 23 mIUim1, and FSH 2.9 rig/ml. Y. F. W. (age 24) with idiopathic hypothalamic amenorrhea, had a previous midtrimester spontaneous abortion followed by a successful HMG/HCG-induced ovulation and pregnancy. She returned for her second course of gonadotropin therapy. Baseline LH was 14 mIU/ml and FSH 2 rig/ml. Pretreatment prolactin levels in all three women were less than 15 rig/ml. Both the latter patients also were delivered of normal infants at term following spontaneous labor. Results The hormone levels in these patients are shown in Figs. 1 and 2. Fig. 2 also includes the plasma and pleural Huid protein levels from patient C. V. The levels of prolactin, estradiol, and progesterone were assayed in all samples of the pleural Huid and compared to levels in blood obtained prior to the thoracentesis. The concentrations of these hormones were slightly lower in the pleural effusion as compared to corresponding plasma samples (data not shown). Following ovulation and implantation. the levels of prolactin. estradiol, progesterone, and testosterone, but not HCG-/3 in
duction in ovarian size. Concomitant with this clinical improvement was the rapid decline in prolactin, escradiol. and testosterone levels. Progesterone lell some!vhat more slowly while HCG-p levels gradu;tlh increased, as in I. S. and Y. F. M’.
9876543210 Response Days
to HMG HCG
in the cycles of conception.
(1. \‘.. were higher than in I. S. and Y. F. MT. The A\-ation of prolactin, estradiol, progesterone, and testosterone was evident in the interval preceding the onset of the hyperstimulation syndrome. Following admission to hospital, when active intravenous fluid therap! with dextrose, sodium, potassium, and human albumin ~vas undertaken Lo maintain fluid, electrolyte, and protein balance, there appeared To he a marked suppt-ebsion ofprolactin levels during this interval. The prolactin levels in C. V. again increased when this intensive fluid therapy was discontinued. A further inct‘eilse in plasma estradiol occurred later. Further eff’usions of Huid into the peritoneal and pleural cavities necessitated the final two pleural aspirations, with subsequent spontaneous regression, as evidenced by clinical and radiologic examinations. Following each aspiration, there was a rapid reduction in the ascites. R! dav 33 af’ter the administration of HCG, these third space fluid shifts spontaneously subsided, with a re-
Elevated estrogen levels are associated with increases in pt-olactin secretion. This occurs during the tncnstrual cycle, when plasma prolactin concentrations peak at midcycle, during pregnancy, and following estrogen administration. The elevated prolactin we have observed in C. V. ma) therefore have resulted f‘rotn the marked increase in estradiol associated with the hjperstimulation svndrome. Nevertheless, nonspecific stress may increase pituitary prolaclin secretion and, because the decidua ma)- release prolactin,” the effects of’ stress and excessive decidualization need to he coilsidered. It seetns important to note that the prolactin increase occurred well before systemic symptoms and signs developed and also that the magnitude of the prolacriti inct-ease in C. \‘. was lip to 200 nghl. sipproxitnating levels observed at term. This would suggest that the stress-induced prolactin increase was not the sole f‘actor in the augmentation of prolactin in this patient. It was of’ interest that the prolactin le\,els \+ere lowered throughout the interval when intensive irttravenom fluid therapy with dextrose. saline, potassium, and albumin was administered. When intravenous therapy was discontinued, a further increase in prolactin levels was observed (Fig. 2). The possible effects of fluid loading on prolactin concentrations have been described b\- Huckman and Peake’ in the nonpregnant state but this has not been confirmed by othet- studi& Recent studies ha\:e demonstrated that prolac-tin ma\’ affect permeability of the guinea pig and human anttiotic metnbrane.“~ i In the guinea pig, prolactin increased Huid shift frotn the fetal to the maternal side. whereas vasopressin caused the opposite eflcr. Polishuk and Schenke? reproduced the hyperatimulation syndrome with HMC in female but not male tab bits, Jewelewicz and associates” suggested that massive ascites and resulting hypovolemia art‘ cardinal factors in this syndrome. Clinically. ovarian h\ pet-stimulation develops after ovulation has occurred and ma) he the result of high local concentrations of’ estrogen in the ovaries causing altered capillary permeabilit\ “I and ascttes. However. since exteriorization of the ovaries in the rabbit studies did not prevent this syndi-ome.” systemic effects from the ovaq. and elsewhere :11-c involved in the fluid shifts into the peritoneal and pleural cavities.
Plasma hormones in ovarian hyperstimulation syndrome
14 16 18 20 22 24 26 20 Days
30 32 34
36 38 40 42 44
Fig. 2. Hormone levels following administratron of human chorionic gonadotropin in the cycles of conception (HCG-/3 levels in semilog scale). Plasma and pleural fluid albumin levels are also shown. In C. V.. the onset of the syndrome was related to the time of implantation of the blastocyst, when endogenous HCG secreted from the trophoblast further stimulated and augmented the already high levels of ovarian hormones present at the time of implantation. The high plasma and urinary steroid levels observed with this syndrome, ‘3 ” coupled with the effect of estrogen in inducing fluid retention” and changes in vessel permeability,‘” suggest that this steroid plays an important role in the pathogenesis. The increase progesterone levels may cause a natriuretic effect, which
has been suggested as a possible stimulus to raise aldosterone secretion.‘” High urinary aldosterone excretion has been reported in the hyperstimulation syndrome.” Androgens may cause retention of water and sodium in human subjects. ” The slight increase in plasma testosterone levels in this patient may thus be relevant. Our results, taken together with the studies reviewed, suggest that the raised plasma levels of prolactin, estradiol, and progesterone are important factors in the pathogenesis of this disorder. The falling prola&n and ovarian steroid levels occurring with the
of’ the clinical signs and symptoms appealwilli this interpretation. Raised prolactin le\ els olmi~vctl in tlic premenstrual tension s! ndromc” aIs suggest that prolactin, in iissociation with o\arian aleroids, ma) ~‘lar ;I role ill these clinical disorders asociatcd I\-ith o\ul;irioii. Finall\. estrogen and testos(el’-
1. Ncuwirth. ,‘iM.
af’ftct specifv prolac-tin binding in tile Lidne>h arid adrenals of KILL “’ ‘I‘he possible role of’ prolactirl interacting ~\ith o\m iaii steroids in w:ttei~ ;i~itl elecrix~l~tc nictabolism during ttlC Ill~llStl.ll~ll c\ClC alltl })l‘C~~IIiIIl~~~ deserves ftwtlict. iinwtigation. one
R. S.. Tun-l\aq,
R. N.. and Vande Wiele, R. I..: 91:977,
‘1 Kohyn. C.: Itr