Endocrine Morbidity in Midline Brain Defects: Differences Between Septo-Optic Dysplasia and Related Disorders

Major Points and Findings:

This study does not concern stem cells. The authors report the largest single-centre follow-up of children with septo-optic dysplasia (SOD) published to date. It covers 171 children with SOD, compared with 35 children with isolated optic nerve hypoplasia (ONH) and 53 with multiple pituitary hormone deficiency (MPHD) but no eye or midline brain involvement, all looked after at Great Ormond Street Hospital and University College London Hospitals. We include it because it documents the hormonal course of SOD and ONH over many years. It reports which deficiencies appear and how late, what happens at puberty, how far MRI predicts the course, and how often children died.

The study is retrospective and comes from a specialist referral centre, so the children are probably more severely affected than average. The authors acknowledge this.

Aim:

To describe endocrine illness and mortality in a large group of children and adolescents with SOD, and to compare them with children who have MPHD without midline defects and children with isolated ONH.

Methods:

The cohort comprised 259 children diagnosed between 1994 and 2015. SOD was defined in the usual way, as two or more of ONH, pituitary deficits and midline brain defects. The SOD group was split into SOD+ (with hypopituitarism, n = 132) and SOD- (pituitary function preserved, n = 39). Median follow-up was 8.0 years for SOD, 6.6 for MPHD and 6.9 for ONH, to a mean age of about 9 years (range up to 21).

Assessments were as follows:

• Growth hormone, TSH, ACTH and gonadotrophin deficiencies and central diabetes insipidus were diagnosed according to current guidelines. Some children also had an overnight growth hormone profile (samples every 20 minutes for 12 hours).
• Severity was summarised with an Endocrine Morbidity Score from 1 (one deficit) to 5 (panhypopituitarism).
• All MRI scans were reviewed by one experienced neuroradiologist.
• Time to each deficiency was compared using Cox proportional hazards models.

Results:

In the SOD group, 64.9% had all three features of the triad and 89.9% had ONH, which was bilateral in 87.4%. Median age at diagnosis was 1.03 years for SOD and 1.68 years for isolated ONH, against 0.44 years for MPHD.

Only 39 of 171 children with SOD (22.8%) kept normal pituitary function, over a median of 6.25 years and up to 14 years of follow-up.

The timing and pattern of deficiencies were as follows:

• In SOD+, the median age at first deficiency was 1.78 years, and 90% had their first deficiency by 8.54 years. In MPHD the figures were 0.18 and 4.80 years (hazard ratio 0.62, 95% CI 0.44 to 0.86).
• Hazard ratios for SOD+ against MPHD were 0.63 for growth hormone, 0.48 for TSH, 0.55 for ACTH and 0.28 for gonadotrophins. Every anterior pituitary deficit was thus less frequent and later in SOD. Diabetes insipidus showed the opposite tendency (2.27, not significant).
• Children with SOD continued to develop new deficits through adolescence, with a first deficiency as late as 14 years, ACTH deficiency at 16 and diabetes insipidus at 17. The authors believe theirs is the first report of late-onset diabetes insipidus in SOD. It never occurred alone.
• The commonest combination was growth hormone plus TSH plus ACTH deficiency (30.7% of SOD+). No typical sequence emerged.
• Growth hormone deficiency was not universal, and 11 children with SOD+ (8.3%) had normal growth hormone secretion despite other deficits. Five others passed the standard stimulation test but had an abnormal overnight profile.

Puberty: Of children old enough to assess, 81.8% with SOD and 92.8% with ONH entered puberty spontaneously at the normal age (MPHD 46.7%). Precocious or rapidly progressing puberty was treated in 7.0% of SOD and 8.6% of ONH and in no child with MPHD. Among boys with SOD, 42.7% were born with undervirilised genitalia (small penis, hypospadias, undescended testes or a combination).

Weight: Obesity at the last visit was recorded in 31.1% of SOD, 41.2% of MPHD and 12.1% of ONH. Leanness was seen only in SOD (5.6%) and ONH (6.1%), and extreme obesity in three children with SOD. All children with SOD who had early puberty, leanness or extreme obesity had bilateral ONH.

MRI: A small anterior pituitary was seen in 81.5% of SOD and in half of the children with isolated ONH, who nevertheless still had normal pituitary function after a median of 10 years. An ectopic posterior pituitary (41.6% of SOD+, 2.7% of SOD-, 0% of ONH) and stalk abnormalities predicted earlier anterior pituitary deficits. However, 23% of children with diabetes insipidus had a normal-looking posterior pituitary. Septum pellucidum and corpus callosum abnormalities, and whether ONH was in one or both eyes, showed no clear link with when deficiencies appeared.

Of 144 children with SOD who had genetic testing, a variant was found in 6 (4.2%).

Deaths: Mortality data were available for 144 SOD, 31 ONH and 50 MPHD patients. Six children with SOD (4.2%, all of whom had ONH) and one with isolated ONH (3.2%) died. No child with MPHD died. In no case was the hormone disorder recorded as the cause. All had complex problems involving the heart, nervous system, bones or lungs.

Conclusions:

The authors regard SOD as a spectrum and not a single disease. Its hormone problems are dynamic and sequential, in contrast to the early, more uniform pituitary failure of MPHD. Isolated ONH appears to sit at one end of the same spectrum and may share a degree of hypothalamic dysfunction. About half of these children already have a small pituitary and may go on to meet criteria for SOD. Because midline brain findings had little bearing on hormone outcomes, the authors propose renaming the condition “hypothalamo-pituitary-optic dysplasia” (HPOD). MRI predicts the course only partly, so they call lifelong regular surveillance essential in every group.

The authors state several limitations. The design was retrospective. The cohort was followed in childhood, so deficiencies arising in adulthood are missed. Most children had not reached puberty, so gonadotrophin deficiency is probably under-counted. Referral bias is likely, since 65% had the full triad and 77% had hypopituitarism, against 24 to 30% and 50 to 66% in earlier studies.

Background Information:

The paper cites an incidence for SOD of 10.9 per 100,000, with causative mutations found in fewer than 10% of cases.

The findings are relevant before any elective treatment. A child with SOD or ONH may have, or may later develop, cortisol deficiency or diabetes insipidus. In standard endocrine practice both need to be recognised and covered with replacement treatment, particularly around illness, fasting and sedation. The authors note that the low mortality in their cohort was seen in children whose hormones were, as far as they could tell, properly replaced. The paper did not test any restorative therapy, and we found no published clinical study of cell therapy in SOD. The authors’ main recommendation is careful, continuing endocrine follow-up for these children.