High risk pregnancies in hypopituitary women

Caroline E. Overton1, Colin J. Davis1, Christine West1, Melanie C. Davies1 and Gerard S. Conway2,3

1 University College London Hospitals NHS Trust, The Reproductive Medicine Unit and 2 Department of Endocrinology, The Middlesex Hospital, Mortimer Street, London W1N 8AA, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Various short papers have suggested that pregnancies in women with hypopituitarism are high risk but no formal assessment of pregnancy outcome has yet been reported. METHODS: An audit was carried out concerning the outcome of 18 pregnancies in nine women who underwent ovulation induction in a single centre over 20 years. RESULTS: The live birth rate was 61%, miscarriage rate 28% and mid-trimester uterine death rate 11% with no survivors from four sets of twins. The Caesarian section rate was 100% and half of the live births were on or below the 10th centile for weight. One woman successfully breast-fed. CONCLUSIONS: Women with hypopituitarism have high-risk pregnancies, perhaps because of a uterine defect secondary to endocrine deficiency. Fertility treatment must strive for singleton pregnancies with application of particularly strict criteria to avoid twin pregnancies. Early elective Caesarian section is probably warranted in this group.

Key words: maternal hypopituitarism/ovulation induction/pregnancy outcome


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Since the introduction of ovulation induction with gonadotrophins in the 1960s, women with established hypopituitarism can expect near normal fecundity. In the following 40 years of treatment programmes, only scant information has emerged on the outcome of pregnancies in these women (Van der Spuy and Jacobs, 1984Go). In two series that address this subject, only nine women were unequivocally panhypopituitary (Magyar and Marshall, 1978Go; Grimes and Brooks, 1980Go) and little information on fetal outcome of pregnancies was included. Nevertheless, these reports record a high rate of spontaneous abortion and maternal mortality.

We have audited the outcome of all pregnancies in women with panhypopituitarism attending the fertility clinics of University College London Hospitals between 1980 and 2000 with the purpose of determining whether spontaneous labour and lactation was possible in the absence of pituitary function and to document fetal outcome.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
A database which prospectively recorded all fertility treatment cycles identified nine women with panhypopituitarism who had successfully conceived with ovulation induction. The clinical details of these women are listed in Table IGo. The median age of the subjects was 35.5 years (range 29–42), their median height was 1.56 m (range 1.50–1.60) and their median body mass index was 23 kg/m2 (range 20–30). Six women booked for maternity care elsewhere, and their consultant was contacted to request details of the pregnancy on a proforma designed for the study. Complete data was obtained for eight of the nine women. All the women had a prior diagnosis of panhypopituitarism and received complete pituitary replacement with thyroxine, corticosteroids and estrogen replacement. In eight women we could confirm full pituitary testing with standard insulin tolerance, GnRH and thyroid releasing hormone. One woman subsequently had repeat pituitary testing and cortisol replacement has been successfully withdrawn. Four of the nine women were also taking vasopressin for diabetes insipidus.


View this table:
[in this window]
[in a new window]
 
Table I. Details of age, diagnosis and pituitary replacement regime in nine women with hypopituitarism
 
Two women (cases 5 and 7) had had uneventful earlier pregnancies with normal deliveries in case 5 and elective Caesarian sections in case 7. All nine women were amenorrhoeic and required ovulation induction in order to conceive. Two women showed a response to the GnRH pump and each one conceived—one ended in miscarriage and the other in a live birth. The remaining women received treatment with gonadotrophins and HCG.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Details of the 18 pregnancies, 14 (78%) singletons and four twin, are shown in Table IIGo.


View this table:
[in this window]
[in a new window]
 
Table II. Ovulation induction method, pregnancy outcome and pueperium details in 18 pregnancies in nine women with hypopituitarism
 
There were 11 (61%) live births, five (28%) first trimester miscarriages and two (11%) second trimester intrauterine fetal deaths. Of the singleton pregnancies, the first trimester miscarriage rate of 28% is within the normal expected range. The median birth weight was on the 25th centile and half of the babies were small for gestational age with birth weights on or below the 10th centile. Two babies were admitted to the neonatal intensive care unit because of prematurity, both at 34 weeks gestation. There were no reported congenital abnormalities. There were no survivors from the four sets of twins. Two sets of twins miscarried in the first trimester and two ended in intra-uterine fetal deaths in the second trimester.

Ten of the 11 live-born babies were delivered by Caesarean section and in one the mode of delivery was unknown. There were six emergency Caesarean sections; two for fetal distress, two for pre-term rupture of membranes and malpresentation, one for failure to progress and one failed induction of labour. There were four elective Caesarean sections, one for suspected intrauterine growth restriction, one for a high head at term and two repeat. All women received i.v. hydrocortisone replacement 50–100 mg three times daily for three doses during delivery. There was one case of puerperal psychosis.

One of the nine women did not have a live birth and we have no data for one woman. Of the remaining seven women, only one woman successfully breast-fed while another abandoned breast-feeding after one week. Five women bottle-fed their babies; three had `no milk', one had two pregnancies, both babies were born prematurely with `poor sucking' and one elected to bottle-feed. The three women with no milk production had prolactin levels of 124, 53 and <50 mIU/l (normal range 0–660).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
We have shown that women with panhypopituitarism have high-risk pregnancies in which fetal growth restriction is common and Caesarean section is the rule. The fact that there were no survivors from twin pregnancies may be evidence of poor placental function. It may be that preparation of the uterus before pregnancy requires greater focus, perhaps with more comprehensive pituitary replacement with growth hormone (GH) and higher doses of estrogen. We are not aware of any systematic assessment of non-pregnant uterine measurements in hypopituitarism which might help explain these observations.

Regimens of ovulation induction in hypopituitarism are worthy of comment. Where a structural lesion of the pituitary is the underlying pathology the GnRH pump is unlikely to be effective (Morris et al., 1987Go). In any instance of gonadotrophin deficiency treatment regimens should include preparations with LH activity in order to ensure an adequate estrogen response (Shoham et al., 1991Go). Further, there is anecdotal evidence that GH replacement might be beneficial in hypopituitarism when the response to conventional gonadotrophin therapy is poor (Salle et al., 2000Go).

The previous literature describing pregnancy outcome in panhypopituitarism consists of seven papers reporting on nine women (Polishuk et al., 1965Go; Corral et al., 1972Go; Bowers and Jubiz, 1972; Gemzell, 1973Go; Jorgensen et al., 1973Go; Grimes and Brooks, 1980Go; Golan et al., 1990Go). There were nine live births comprising seven singletons and two sets of twins. Five of the nine women were delivered by Caesarean section and four had vaginal deliveries. Seven women were unable to breast-feed, with no milk, and there was no comment on feeding for two of the women. Details of gestational age and birth weights were scant. Our single centre series of nine women may be more representative of pregnancy outcome in panhypopituitary women compared with case reports. Women ascertained on the basis of deficiency of GH alone have been reported to have had a much more favourable outcome than that reported here for panhypopituitarism, with a Caesarian section rate of 50% (de Boer et al., 1997Go) and low birth weight of 25% (Curran et al., 1998Go). Pituitary GH deficiency, however, is unlikely to be the main contributor to poor pregnancy outcome in this group because placental GH can compensate for its action (Verhaeghe et al., 2000Go).

The poor outcome for twin pregnancies reported here highlights the importance of ensuring unifollicular ovulation in this group of women, with careful ultrasonographic monitoring. More conservative criteria for the number of mature follicles allowed to develop may be required in this subgroup of infertile women compared with others.

Half of the live births were small for gestational age with birth weights on or below the 10th centile, which could be a reflection on maternal short stature as smaller, lighter mothers have lighter babies. This again raises the issue of intrauterine GH deficiency, as a role for pituitary GH in the regulation of fetal body weight during late gestation has been demonstrated in rats (Nogami et al., 1995Go). Indeed, one case report advocates the use of GH replacement during pregnancy in GH deficient women despite the action of placental GH (Muller et al., 1995Go).

The incidence of Caesarean section in our series is striking. It may be that these `precious pregnancies' trigger early obstetric intervention. Interestingly, in the two cases of fetal demise, delivery was vaginal indicating that labour and vaginal delivery are possible where fetal condition is not a consideration. Of the seven women, only one woman successfully breast-fed, presumably because of prolactin deficiency. However, successful breast-feeding has been reported despite undetectable prolactin concentrations postpartum (Franks et al., 1977Go).

We conclude that the obstetric outcome of these pregnancies suggests that they should be treated as high risk. It may be that greater attention to preparation of the uterus prior to conception, perhaps with the addition of GH, will improve the outcomes reported here. Twin pregnancy carries a particularly poor outcome and should be avoided at all costs with more conservative criteria for the triggering of ovulation than might be applied to women with normal pituitary function. Fetal growth should be followed by serial ultrasound to detect intrauterine growth restriction and delivery by elective Caesarean section discussed.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The authors would like to thank all who contributed to data for the follow-up of these women: Professor Howard Jacobs and Mr S.J. Steele, The Middlesex Hospital, London; Mr Brooks, Noble's Isle of Man Hospital; Professor Chris Redman and Dr Jane Moore, John Radcliffe Hospital, Oxford; Professor R. Kumar, Institute of Psychiatry at the Bethlem Royal and Maudsley NHS Trust and Mr David Howell, Luton and Dunstable Hospital.


    Notes
 
3 To whom correspondence should be addressed. E-mail: g.conway{at}ucl.ac.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Bowers, J.H. and Jubiz, W. (1974) Pregnancy in a patient with hormone deficiency. Arch. Intern. Med., 133, 312–4.[ISI][Medline]

Corral, J., Calderon, J., and Goldzieher, J.W. (1972) Induction of ovulation and term pregnancy in a hypophysectomized woman. A case report. Obstet. Gynecol., 39, 397–400.[ISI][Medline]

Curran, A.J., Peacey, S.R., and Shalet, S.M. (1998) Is maternal growth hormone essential for a normal pregnancy? Eur. J. Endocrinol., 139, 54–58.[ISI][Medline]

de Boer, J.A., Schoemaker, J., and van der Veen, E.A. (1997) Impaired reproductive function in women treated for growth hormone deficiency during childhood. Clin. Endocrinol. (Oxf), 46, 681–689.[ISI][Medline]

Franks, S., Kiwi, R., and Nabarro, J.D. (1977) Pregnancy and lactation after pituitary surgery. Br. Med. J., 1, 882.[Medline]

Gemzell, C. (1973) Induction of ovulation in patients following removal of a pituitary adenoma. Am. J. Obstet. Gynecol., 117, 955–961.[ISI][Medline]

Golan, A., Abramov, L., Yedwab, G., and David, M.P. (1990) Pregnancy in panhypopituitarism. Gynecol. Obstet. Invest., 29, 232–234.[ISI][Medline]

Grimes, H.G. and Brooks, M.H. (1980) Pregnancy in Sheehan's syndrome. Report of a case and review. Obstet. Gynecol. Surv., 35, 481–488.[Medline]

Jorgensen, P.I., Sele, V., Buus, O., and Damkjoer, M. (1973) Detailed hormonal studies during and after pregnancy in a previously hypophysectomized patient. Acta Endocrinol. (Copenh. ), 73, 117–132.[ISI][Medline]

Magyar, D.M. and Marshall, J.R. (1978) Pituitary tumours and pregnancy. Am. J. Obstet. Gynecol., 132, 739–751.[ISI][Medline]

Morris, D.V., Abdulwahid, N.A., Armar, A. and Jacobs H.S. (1987) The response of patients with organic hypothalamic-pituitary disease to pulsatile gonadotropin-releasing hormone therapy. Fertil. Steril., 47, 54–59.[ISI][Medline]

Muller, J., Starup, J., Christiansen, J.S., Jørgenson, J.O., Juul, A. and Skakkebaek N.E. (1995) Growth hormone treatment during pregnancy in a growth hormone deficient woman. Eur. J. Endocrinology, 132, 727–729.[ISI][Medline]

Nogami, H., Tachibana, T., and Ishikawa, H. (1995) Intrauterine growth retardation due to growth hormone deficiency in rats. Biol. Neonate, 68, 412–418.[ISI][Medline]

Polishuk, W.Z., Palti, Z., Rabau, E., Lunenfeld, B. and David A. (1965) Pregnancy in a case of Sheehan's syndrome following treatment with human gonadotrophins. J. Obstet. Gynaecol. Br. Commonw., 72, 778–780.[Medline]

Salle, A., Klein, M., Pascal-Vigneron, V., Dousset, B., Leclere, J. and Werhya, G. (2000) Successful pregnancy and birth after sequential co-treatment with growth hormone and gonadotropins in a woman with panhypopituitarism: a new treatment protocol. Fertil. Steril., 74, 1248–1250.[ISI][Medline]

Shoham, Z., Balen, A., Patel, A. and Jacobs, H.S. (1991) Results of ovulation induction using human menopausal gonadotropin or purified follicle-stimulating hormone in hypogonadotropic hypogonadism patients. Fertil. Steril., 56, 1048–1053.[ISI][Medline]

van der Spuy, Z.M. and Jacobs, H.S. (1984) Management of endocrine disorders in pregnancy. Part II. Pituitary, ovarian and adrenal disease. Postgrad. Med. J., 60, 312–320.[ISI][Medline]

Verhaeghe, J., Bougoussa, M., Van Herck, E., de Zegher, F., Hennen, G., and Igout, A. (2000) Placental growth hormone and IGF-I in a pregnant woman with Pit-1 deficiency. Clin. Endocrinol. (Oxf), 53, 645–647.[ISI][Medline]

Submitted on December 12, 2001; accepted on February 8, 2002.