Objective: It remains uncertain whether there is any disadvantage imposed upon women with pituitary disease who are GH-deficient and become pregnant. The aim of this study was to determine whether maternal GH deficiency adversely affects the outcome of pregnancy. Design: Retrospective study. Methods: The case notes of 77 female patients with known GH deficiency were examined. Sixteen patients (a total of 25 pregnancies) were identified who had been pregnant whilst known to be GH-deficient. Peak GH response to provocative testing prior to pregnancy, length of gestation, birth weight, maternal well-being and the incidence of maternal and fetal complications of pregnancy were documented. Results: Peak GH response to insulin tolerance test (n ¼ 21) or glucagon stimulation test (n ¼ 4) prior to pregnancy was 8.7 (<1 to 17.3) mU/l (peak Յ9 mU/l in 14 cases). There were 25 pregnancies resulting in 26 live births (including one set of twins and one set of quins) and 4 spontaneous first trimester abortions. Eight pregnancies were achieved by ovulation induction. Median gestation of live births was 39 (33 to 42) weeks. Median birth weight excluding multiple births (n ¼ 19), uncorrected for gestational age, was 3.09 (1.64 to 4.19) kg, and the numbers with birth weights below the 10th, between the 10th and 90th, and above the 90th centiles were five, nine and five respectively. Preeclampsia occurred in two pregnancies and post-partum haemorrhage after one pregnancy. There were three minor congenital abnormalities. Conclusions: Our data suggest that pregnancy in GH-deficient females is not detrimental to the fetus and the incidence of maternal morbidity is low. We conclude that GH replacement therapy is probably not essential for GH-deficient females during pregnancy.

89 Snajderova M, Mardesic T, Lebl J, et al. The uterine length in women with Turner syndrome reflects the postmenarcheal daily estrogen dose. Congenital subependymal giant cell astrocytoma diagnosed on fetal MRI A primiparous mother had an antenatal ultrasound at 21 weeks' gestation, which showed a mass in the left side of the brain arising from the intraventricular region and several cardiac tumours, most likely to be rhabdomyomas. An MRI scan at 24 weeks' gestation showed an intraventricular mass in the fronto-parietal part of the left cerebral hemisphere of the fetus (fig 1). The lesion was of high signal on T1 weighted sequence and low on T2. The boy was born at 36 weeks' gestation. Physical and neurological examination was normal. Postnatal cranial ultrasound and MRI confirmed a large mass lesion involving the left lateral ventricle; within the body of the right ventricle was a solitary subependymal nodule (fig 2). Based on the radiological abnormalities of the brain and heart, tuberous sclerosis (TS) was strongly suspected. The tumour in the left hemisphere fulfilled the neuroradiological diagnostic criteria for a subependymal giant cell astrocytoma (SEGA). 1There was no history or evidence of TS on clinical examination of family members. Genetic testing of the neonate showed the mutation for TS.At 2 months he developed right focal seizures and anticonvulsant therapy was commenced. EEG showed hypsarrhythmia. At 10 months of age he underwent surgical excision of the brain tumour, confirmed histologically as a SEGA. Now aged 1 year, his development is within normal limits and his seizures are controlled.SEGAs are rare brain tumours that occur typically in the walls of the lateral ventricles and are generally located near the foramen of Monro. They are nearly always associated with TS. 2 3 They are usually discovered in the second decade of life and only rarely occur in infancy. 4 To our knowledge this the first report of a SEGA diagnosed on fetal MRI. Although they are benign in nature and grow slowly, they represent a major cause of death due to raised intracranial pressure or haemorrhage in patients with TS. 5 N Hussain, A Curran