全文

Abstract

Cushing's syndrome during pregnancy is a rare metabolic condition that is associated with high maternal and foetal morbidity. Clinical symptoms may mimic those of normal pregnancy. A diagnosis is best made based on clinical presentation, laboratory and imaging findings as well as a high index of suspicion. Medical management with anti-steroidogenic agents such as metyrapone has been shown to be effective, but surgery is usually the recommended treatment option. Its main limitation is optimal timing of the procedure in late first trimester or early second trimester to prevent spontaneous termination of pregnancy. We describe our experience and management of a 39-year-old patient with uncontrolled hypertension at 25 weeks gestation which was later diagnosed as ACTH independent Cushing's syndrome and had a favourable pregnancy outcome. The role of medical therapy and its challenges, as well as its impact on pregnancy outcomes, were explored by a literature search conducted through Pubmed and Medline databases. A total of 12 patients with Cushing's syndrome during pregnancy were reported to have been managed with metyrapone, with ketoconazole being studied to a significant degree in three cases. Of these women, 53% delivered close to term and 20% developed pre-eclampsia. Despite two neonatal deaths and one stillborn reported, medical management appeared effective in controlling hypercortisolemia during pregnancy with strict monitoring of blood pressure and foetal surveillance. It remains the only active management in the setting of pregnancy-induced Cushing's syndrome, and has shown to be a viable option in controlling serum cortisol levels especially as an adjunct to surgery as reflected in four cases. A multidisciplinary approach towards an individualised management process is warranted with medical management to ensure a safe maternal and foetal outcome.

Keywords: Cushing's syndrome, Pregnancy, Hypercortisolemia, Metyrapone, Ketoconazole.

1. Introduction

Cushing's syndrome (CS) is a disorder caused by prolonged abnormal exposure to excess glucocorticoids, leading to significant consequences in the patient if left untreated [1]. Fortunately its association in pregnancy is rare, with only small and multiple case series reports in the literature. CS that occurs during pregnancy is a condition of great concern as it increases the risk of maternal and foetal morbidity [2]. Despite major advances in diagnosis and therapy, identifying CS is frequently a challenge for the clinician due to the normal physiological hypercortisolemia that accompanies pregnancy. Clinical features of weight gain, stretch marks, fatigue, back pain, mood changes and facial roundness are common in pregnancy but may also be symptoms of CS. In addition, the development of elevated blood pressure and blood glucose levels – both common in CS – may further affect the outcome of the pregnancy itself. Active management of hypercortisolemia with either medical or surgical therapy is associated with a trend towards live births [2], and should be individualised due to the rarity of disease and lack of treatment guidelines. The aim of the present paper is to share our experience in diagnosing and managing a case of CS in one of our patients while she was pregnant, and to present a review of the current literature which guided our treatment choice.

1.1. Case study

A 39-year-old Caucasian woman, gravida 3 para 2, was admitted at 25 weeks’ gestation from the antenatal clinic with hypertension. Her last two pregnancies had been uncomplicated, with the first delivery carried out through emergency caesarean section for a failed instrumental delivery at 39 weeks’ gestation, and the next a spontaneous vaginal delivery at 30 weeks’ gestation. This current pregnancy was of the same paternity as the previous two, and her medical history was significant for hypothyroidism that was being treated with thyroxine.

On presentation the patient's blood pressure was 173/99 mmHg. Clinically she exhibited a number of features associated with Cushing's syndrome: obesity (body mass index 31 kg/m2), an obvious buffalo hump with supraclavicular fat pads, abdominal striae and facial acne. Her laboratory investigations are described in Table 1. A magnetic resonance imaging (MRI) of the abdomen revealed a 40 mm diameter bi-lobed homogenous lesion in the left adrenal gland. She was diagnosed with adrenocorticotrophic hormone (ACTH) independent CS. Her ACTH and serum cortisol levels were measured twice-hourly for a 24 h period, with ACTH levels consistently less than 10 ng/L and cortisol levels between 980 and 1091 nmol/L. Given her late gestation and previous history of preterm labour, surgical intervention was considered inappropriate.



The patient was medically managed with 250 mg of metyrapone, given twice daily from 27 weeks’ gestation onwards, where the dose was titrated in response to her weekly urinary free cortisol (UFC) levels to ensure a downward trend (Fig. 1). She was further commenced on dual anti-hypertensive treatment with methyldopa and nifedipine, and was subsequently diagnosed with gestational diabetes that was in turn managed with insulin. Pre-eclampsia was suspected at 32 weeks, with the patient undergoing successful vaginal delivery at 35 weeks after spontaneous onset of labour. She delivered a liveborn male infant weighing 2800 g (75th centile) with Apgar scores of 6 and 8 at 1 and 5 min, respectively.

Postnatally the patient's serum cortisol remained elevated (420–489 nmol/L) with no diurnal variation. Bone densitometry showed no evidence of osteoporosis. Metyrapone and nifedipine were continued until a laparoscopic adrenalectomy was performed six weeks post-partum. Histopathology confirmed a 40 mm benign adenoma comprised of large polygonal richly vascular cells with no evidence of malignancy. On follow-up over the next two years she has remained free of any signs or symptoms to suggest a relapse of her CS.

 

2. Materials and methods

A literature search of all case studies published in English was conducted using the Pubmed database. Following this, we reviewed all the reported cases of hypercortisolemia during pregnancy that had been managed with medical therapy, including those that were managed in preparation for surgery. The earliest publication dated back to 1975, with the most recent in 2011. Keywords used were “Cushing's syndrome”, “pregnancy”, “hypercortisolism”, “metyrapone”, and “ketoconazole”. Primary end points were maternal characteristics as well as foetal outcome.

 

3. Results

Twelve reports showing a total of 15 women with medically managed CS were identified (Table 2). Metyrapone was the most commonly used medical therapy (n = 12), with doses ranging from 250 mg to 3000 mg per day. The average number of treatment days was 107, with the majority being in the second to third trimester. Two women were treated with ketoconazole and one with both ketoconazole and cabergoline. Ketoconazole dosing ranged from 400 to 1000 mg daily.

IUFD, intrauterine foetal death; IUGR, intrauterine foetal growth restriction; PET, pre-eclampsia; NS, not stated.

Pregnancy-induced CS was the most common aetiology for medical treatment (n = 9). Over half of these women (66%) delivered before 34 weeks’ gestation, with pre-eclampsia and placental abruption warranting early deliveries. One woman delivered a stillborn infant, with no abnormalities identified on autopsy. Adrenal adenoma was the second most common aetiology (n = 5), and all women in this group successfully delivered between 34 and 37 weeks’ gestation, with no cases of pre-eclampsia being reported. The average gestational age at delivery was 33 weeks, and 80% of women (n = 12) required caesarean section for delivery. Wound complications were not reported. Two neonatal deaths occurred, with both infants delivered at 28 weeks’ gestation in the adrenal carcinoma and Cushing's disease group. The majority of neonatal complications were those of prematurity, including intensive care admissions, respiratory distress and neonatal jaundice. Of note, a single case of neonatal adrenal insufficiency was also reported. Ketoconazole was not associated with any neonatal abnormalities in the infants: one had exposure during the first trimester, and two women were treated within the third trimester of their pregnancy.

Table 3 describes the cases of five women who received both medical and surgical intervention. All women were treated medically prior to their surgical procedure. Four of these women underwent surgery in the second trimester and one at 31 weeks’ gestation. Of the five, one woman developed pre-eclampsia at 34 weeks. All five women had vaginal deliveries and live births.

4. Comments

Pregnancy dramatically affects the hypothalomo–pituitary–adrenal (HPA) axis and the endogenous secretion of cortisol, with total and free serum cortisol concentrations reaching levels higher than those compared to non-pregnant controls [3]. This was confirmed in a recent longitudinal study where increases in maternal total plasma cortisol, corticosteroid-binding globulin (CBG), plasma free cortisol, and 24-h UFC were demonstrated in normal pregnancy to reach peak levels during the third trimester using modern assays, with a 3-fold elevation that was consistent with previous studies [4].

The classification of CS in pregnancy is similar to the one used in the non-pregnant state, being either ACTH dependent or ACTH independent. The aetiology of CS in pregnancy is different, however, as demonstrated by the frequency of ACTH-independent cases that is increased in pregnant as compared to non-pregnant women. Approximately 60% of hypercortisolemia in pregnancy is caused primarily by adrenal adenomas [2], [5], and [6], as seen in our case study, with pituitary adenoma accounting for 15% and adrenal carcinoma for 9% [2] and [5]. The cause for such a difference is not known. It has been suggested that adrenal adenomas are purely cortisol-producing tumours with minimal androgen production, making ovulation and subsequent pregnancy possible, as opposed to a pituitary adenoma, where hyperandrogenism and hypercortisolemia will both suppress and impair ovulation [5], [7], and [8]. Other rare cases include pregnancy-induced CS which involves aberrant LH/hCG receptor expression. In such a phenomenon there is no identifiable pathology on medical imaging, with resolution of hypercortisolemia post-partum [9] and [10].

Pregnancies with CS are complicated by uncontrolled hypertension and gestational diabetes, two co-morbidities that are associated with adverse risks to both maternal and foetal wellbeing [2] and [5]. Elevated blood pressure caused by high cortisol levels is usually severe and when uncontrolled may lead to multiple complications including pre-eclampsia, pulmonary haemorrhage, and acute cardiac as well as renal failure [11]. The foetus is partially protected from hypercortisolemia in early pregnancy as the placental 11-beta-hydroxysteroid dehydrogenase type-2 enzyme converts the majority of maternal cortisol to the biologically inactive cortisone [7] and [12]. As the pregnancy progresses into the second and third trimesters, however, levels of cortisol increase drastically to abnormal levels which are harmful to the foetus, and have been associated with spontaneous pregnancy loss, prematurity, oligohydramnios, intrauterine growth restriction (IUGR) and intrauterine foetal demise [2], [5], and [13].

The diagnosis of CS in pregnancy is difficult due to the lack of defined reference ranges for the usual interpretive tests. Therefore a positive or negative result of a particular test will not necessarily exclude CS, as false positives may also occur, requiring all results to be interpreted with caution and applied clinical correlation [14].

Levels of total serum and free cortisol as well as UFC are increased in CS, complicating the normal screening of serum cortisol and 24-h UFC during pregnancy. The placenta produces corticotrophin-releasing hormone (CRH) and ACTH [3] and [15], which further increase serum cortisol levels, as well as interfering with measurements of plasma ACTH and the results of a CRH stimulation test. As such, ACTH values cannot be relied upon to distinguish ACTH-independent from ACTH-dependent forms of CS in pregnancy. Both low- and high-dose overnight dexamethasone suppression tests do not yield accurate results in patients during pregnancy, not only because glucocorticoids do not suppress placental ACTH, but also because dexamethasone may increase placental CRH and placental ACTH activity. Thus, dexamethasone testing increases the potential for false-positive results in pregnancy [16]. The diagnosis of ACTH-independent CS in our case was concluded based on the 3-fold elevation of 24 h UFC and loss of diurnal variation of serum cortisol, together with the low ACTH and MRI findings of the patient's abdomen. Night-time salivary cortisol (NSC) has been suggested as a useful screening test for CS but its efficacy has not yet been established. A physiological higher NSC has instead been reported due to normal higher level circadian rhythm that reduces the specificity of this test in pregnant women [17].

Imaging studies such as ultrasonography or MRI can be safely performed for adrenal or pituitary tumour detection in pregnancy, although a physiological enlargement of the pituitary or adrenal gland during pregnancy should be considered. The results of these imaging modalities must be placed in context with the clinical and the laboratory findings – particularly given the possibility of pregnancy-induced CS where no pathology is identified.

Active management of patients with CS during pregnancy, either surgically or medically, is associated with a reduction in maternal symptoms and neonatal complications [2] and [5]. It is therefore crucial that CS be actively treated when recognized. While surgery has been shown to be uniformly successful, medical therapy is an alternative when surgery is contraindicated, or when the pathology itself is unidentifiable.

Metyrapone is a steroidogenesis inhibitor that inhibits the enzyme 11-β-hydroxylase in the steroidogenesis pathway in the adrenal cortex, reducing serum cortisol levels [15] and [18]. This in turn stimulates ACTH, which increases the production of deoxycorticosterone and the accumulation of other mineralocorticoids, leading to hypertension [9] and [19]. Uncontrolled hypertension, meanwhile, predisposes the patient to pre-eclampsia and a high maternal mortality if left untreated. Early onset of HELLP syndrome has also been reported [17], [18], and [20], leading to disseminated intravascular coagulopathy and necessitating admission to the intensive care unit. The relationship between metyrapone administration and the incidence of pre-eclampsia is largely debatable. In a recent prospective study, it has been shown that maternal plasma CBG together with total and free cortisol concentrations in gestational hypertension and pre-eclampsia subjects are reduced when compared to controls [23]. Due to the small numbers of available cases of untreated CS during pregnancy, it remains unclear if metyrapone directly causes pre-eclampsia, although the existing studies do appear to point towards such an association. Nevertheless, it is imperative that anti-hypertensives such as methyldopa are used in conjunction with metyrapone in controlling blood pressure to ensure a safer outcome during pregnancy [15].

Ketoconazole, an anti-fungal medication with anti-steroidogenic properties, has been used in a number of case reports with good maternal and foetal outcomes [24], [25], [26], and [27], but has also been implicated as a teratogen (FDA Category C). A recent population-based case–control study found no association between the use of ketoconazole and congenital anomalies, but the number of exposed cases and controls was very small [28]. Thus, at this time, more stratified studies are required to establish the safety profile of ketoconazole in pregnancy, and its use therapeutic use is therefore limited to individuals in need of emergency medical therapy.

Cyproheptadine, an anti-histamine with anti-serotonergic effects, has been used for CS but has been abandoned without further assessment of its efficacy [29]. Historically, mitotane, an adrenolytic drug that specifically destroys the adrenocortical cortex, had once been used but is now considered teratogenic [30].

The risks and benefits of medical treatment relative to surgery are not known due to the small number of reported cases. While surgical treatment has been shown to reduce perinatal mortality rate and maternal morbidity [5], the advantages of medical therapy are evident when surgery is contra-indicated – late gestational age, past history of pre-term delivery, absence of an experienced endocrine surgeon, or patient preference for non-invasive intervention. Medical therapy has also shown promising results as an adjunct to surgery, enabling optimal timing of an elective procedure in the late first trimester or early second trimester to prevent spontaneous termination of pregnancy. While evidence is still limited, metyrapone therapy appears effective in suppressing hypercortisolism in pregnancy without teratogenic effect on the foetus and represents the best pharmacological treatment available. Worsening of hypertension as well as risk of pre-term delivery should be anticipated and managed appropriately. In the setting of pregnancy-induced CS where no apparent surgical pathology is identified, medical therapy represents the only active management in controlling pathological hypercortisolism during pregnancy.

Given the extremely low prevalence of CS in pregnancy, it is unlikely that good quality randomised controlled trials will be conducted to establish the best treatment approach. Laboratory testing and imaging modalities provide guidance to the clinician, but by themselves are insufficient in making the diagnosis for CS. Moreover, there are very few data on normal levels of cortisol in pregnancy; assay variation will have an effect and the ideal level of cortisol to titrate to in response to medical therapy is not yet known. Although the case reports reviewed in the present paper provide some guidance for the treatment of CS in pregnancy, there is a paucity of good quality evidence that would better guide our practice.

 

5. Conclusion

Pregnancy is the only physiological state of sustained hypercortisolism in humans. In the setting of CS it poses a challenge to both the obstetrician and the physician due to its significant correlation with maternal and neonatal morbidity. Loss of diurnal variation of plasma cortisol levels appeared to be an important diagnostic feature of CS during pregnancy but should be considered together with imaging findings as well as a high index of clinical suspicion. Regular foetal surveillance, intensive control of maternal blood pressure with anti-hypertensives and control of blood glucose are mandatory interventions. Medical management is a viable option in controlling hypercortisolemia during pregnancy, especially in the setting of pregnancy-induced CS, but this remains an assertion based on limited case–control data of 15 reported cases. A multidisciplinary approach towards an individualised management process is clearly warranted to ensure a good outcome.

 

References

[1] S.T. Sharma, L.K. Nieman. Cushing's syndrome: all variants, detections and treatment. Endocrinology and Metabolism Clinics of North America. 2011;40:379-391

[2] M.A. Buesher, H.D. McClamrock, E.Y. Adashi. Cushing syndrome in pregnancy. Obstetrics and Gynecology. 1992;79:130-137

[3] M. Kita, M. Sakalidou, A. Saratzis, S. Ioannis, A. Avramides. Cushing's syndrome in pregnancy: report of a case and review of the literature. Hormones. 2007;6:242-246

[4] C. Jung, J.T. Ho, D.J. Torpy, et al. A longitudinal study of plasma and urinary cortisol in pregnancy and postpartum. Journal of Clinical Endocrinology and Metabolism. 2011;96:1533-1540

[5] J.R. Lindsay, J. Jonklaas, E.H. Oldfield, L.K. Nieman. Cushing's syndrome during pregnancy: personal experience and review of the literature. Journal of Clinical Endocrinology and Metabolism. 2005;90:3077-3083

[6] S. Murakami, M. Saitoh, T. Kubo, Y. Kawakami, K. Yamashita. A case of mid-trimester intrauterine fetal death with Cushing's syndrome. Journal of Obstetrics and Gynaecology Research. 1998;24:153-156

[7] L. Vilar, M.D.C. Freitas, L.H.C. Lima, R. Lyra, C.E. Kater. Cushing's syndrome in pregnancy: an overview. Arquivos Brasileiros de Endocrinologia e Metabologia. 2007;51:1293-1302

[8] N. Polli, F. Pecori Giraldi, F. Cavagnini. Cushing's syndrome in pregnancy. Journal of Endocrinological Investigation. 2003;26:1045-1050

[9] N. Achong, M. D’Emden, N. Fagermo, R. Mortimer. Pregnancy-induced Cushing's syndrome in recurrent pregnancies: case report and literature review. Australian and New Zealand Journal of Obstetrics and Gynaecology. 2011;52:96-100

[10] V. Hána, M. Dokoupilová, J. Marek, R. Plavka. Recurrent ACTH-independent Cushing's syndrome in multiple pregnancies and its treatment with metyrapone. Clinical Endocrinology. 2001;54:277-281

[11] W.J. Choi, T.S. Jung, W.Y. Paik. Cushing's syndrome in pregnancy with a severe maternal complication: a case report. Journal of Obstetrics and Gynaecology Research. 2011;37:163-167

[12] B. Guilhaume, M.L. Sanson, L. Billaud, X. Bertagna, M.H. Laudat, J.P. Luton. Cushing's syndrome and pregnancy: aetiologies and prognosis in twenty-two patients. European Journal of Medicine. 1992;1:83-89

[13] P.C. De Groot, I.L. Van Kamp, E.J. Zweers, C.D. De Kroon, W.J. Van Wijngaarden. Oligohydramnios in a pregnant woman with Cushing's syndrome caused by an adrenocortical adenoma. Journal of Maternal-Fetal and Neonatal Medicine. 2007;20:431-434

[14] Nieman LK. Cushing's syndrome in pregnancy. In: Basow DS, editor. UpToDate. Available at http://www.uptodate.com/contents/cushings-syndrome-in-pregnancy [accessed 25.02.12].

[15] G. Bednarek-Tupikowska, E. Kubicka, T. Sicińska-Werner, et al. A case of Cushing's syndrome in pregnancy. Polish Journal of Endocrinology. 2011;62:181-185

[16] L. Cousins, L. Rigg, D. Hollingsworth, et al. Qualitative and quantitative assessment of the circadian rhythm of cortisol in pregnancy. American Journal of Obstetrics and Gynecology. 1983;145:411-416

[17] A. Viardot, P. Huber, J.J. Puder, H. Zulewski, U. Keller, B. Muller. Reproducibility of nighttime salivary cortisol and its use in the diagnosis of hypercortisolism compared with urinary free cortisol and overnight dexamethasone suppression test. Journal of Clinical Endocrinology and Metabolism. 2005;90:5730-5736

[18] R.F. Castro, F.F. Maia, A.R. Ferreira, et al. HELLP syndrome associated to Cushing's syndrome—report of two cases. Arquivos Brasileiros de Endocrinologia e Metabologia. 2004;48:419-422

[19] J.M. Connell, J. Cordiner, D.L. Davies, R. Fraser, B.M. Frier, S.G. McPherson. Pregnancy complicated by Cushing's syndrome: potential hazard of metyrapone therapy. British Journal of Obstetrics and Gynaecology. 1985;92:1192-1195

[20] T. Delibasi, I. Ustun, Y. Aydin, et al. Early severe pre-eclamptic findings in a patient with Cushing's syndrome. Gynecological Endocrinology. 2006;22:710-712

[21] C. Wallace, E.L. Toth, R.Z. Lewanczuk, K. Siminoski. Pregnancy-induced Cushing's syndrome in multiple pregnancies. Journal of Clinical Endocrinology and Metabolism. 1996;81:15-21

[22] C. Cabezón, O.D. Bruno, M. Cohen, S. García, R.A. Gutman. Twin pregnancy in a patient with Cushing's disease. Fertility and Sterility. 1999;72:371-372

[23] J.T. Ho, J.G. Lewis, P. O’Loughlin, et al. Reduced maternal corticosteroid-binding globulin and cortisol levels in pre-eclampsia and gamete recipient pregnancies. Clinical Endocrinology. 2007;66:869-877

[24] J. Berwaerts, J. Verhelst, C. Mahler, R. Abs. Cushing's syndrome in pregnancy treated by ketoconazole: case report and review of the literature. Gynecological Endocrinology. 1999;13:175-182

[25] M. Boronat, D. Marrero, Y. López-Plasencia, M. Barber, Y. Schamann, F.J. Nóvoa. Successful outcome of pregnancy in a patient with Cushing's disease under treatment with ketoconazole during the first trimester of gestation. Gynecological Endocrinology. 2011;27:675-677

[26] A.P. Prebtani, D. Donat, S. Ezzat. Worrisome striae in pregnancy. Lancet. 2000;355:1692

[27] J.A. Amado, C. Pesquera, E.M. Gonzalez, M. Otero, J. Freijanes, A. Alvarez. Successful treatment with ketoconazole of Cushing's syndrome in pregnancy. Postgraduate Medical Journal. 1990;66:221-223

[28] Z. Kazy, E. Puho’, A.E. Czeizel. Population-based case–control study of oral ketoconazole treatment for birth outcomes. Congenital Anomalies. 2005;45:5-8

[29] A.S. Khir, J. How, P.D. Bewsher. Successful pregnancy after cyproheptadine treatment for Cushing's disease. European Journal of Obstetrics, Gynecology, and Reproductive Biology. 1982;13:343-347

[30] S. Leiba, R. Weinstein, B. Shindel, et al. The protracted effect of o,p’-DDD in Cushing's disease and its impact on adrenal morphogenesis of young human embryo. Annales d Endocrinologie. 1989;50:49-53

[31] V. Mundra, C.C. Solorzano, P. DeSantis. Cushing syndrome: a rare occurrence in pregnancy. Endocrinologist. 2010;20:165-167

[32] A. Kasperlik-Załuska, B. Migdalska, W. Hartwig, et al. Two pregnancies in a woman with Cushing's syndrome treated with cyproheptadine. Case report. British Journal of Obstetrics and Gynaecology. 1980;87:1171-1173

[33] C.F. Close, M.C. Mann, J.F. Watts, K.G. Taylor. ACTH-independent Cushing's syndrome in pregnancy with spontaneous resolution after delivery: control of the hypercortisolism with metyrapone. Clinical Endocrinology. 1993;39:375-379

[34] M.J. Gormley, D.R. Hadden, T.L. Kennedy, D.A. Montgomery, G.A. Murnaghan, B. Sheridan. Cushing's syndrome in pregnancy—treatment with metyrapone. Clinical Endocrinology. 1982;16:283-293

[35] C. Blanco, E. Maqueda, J.A. Rubio, et al. Cushing's syndrome during pregnancy secondary to adrenal adenoma: metyrapone treatment and laparoscopic adrenalectomy. Journal of Endocrinological Investigation. 2006;29:164-167

[36] J.A. Shaw, D.W. Pearson, Z.H. Krukowski, P.M. Fisher, J.S. Bevan. Cushing's syndrome during pregnancy: curative adrenalectomy at 31 weeks gestation. European Journal of Obstetrics, Gynecology, and Reproductive Biology. 2002;105:189-191

[37] T.J. Hanson, L.B. Ballonoff, R.C. Northcutt. Letter: amino-glutethimide and pregnancy. JAMA. 1974;230:963-964

 

Footnotes

a Women's and Children's Division, Lyell McEwin Hospital, South Australia, Australia

b School of Paediatrics and Reproductive Health, University of Adelaide, South Australia, Australia

c Endocrine and Metabolic Unit, Royal Adelaide Hospital, South Australia, Australia

d School of Medicine, University of Adelaide, South Australia, Australia

e Department of Obstetrics & Gynaecology, Flinders Medical Centre, South Australia, Australia

  Corresponding author at: Department of Obstetrics and Gynecology, Lyell McEwin Hospital, Haydown Road, Elizabeth Vale, 5112 South Australia, Australia. Tel.: +61 08 8182 9586; fax: +61 08 8282 1646.