R. J. M. Ross

Diabetes mellitus associated with a pathogenic point mutation in mitochondrial DNA

Family studies of diabetes mellitus (DM) show that patients are more likely to have affected mothers than affected fathers. Since the inheritance of mitochondrial (mtDNA), unlike nuclear DNA, is exclusively maternal, could it be that defect(s) in mtDNA account for some cases of DM? Such defects have been associated with rare neurological syndromes, in some of which DM has been an accompanying feature. We have looked for glucose intolerance and for a previously known point mutation of mtDNA in a family, some of whose members have a multisystem disorder with DM but not neurological involvement. DNA samples were obtained from fourteen family members. The point mutation (affecting position 3243 in the tRNA leucine mitochondrial gene) was found in all three diabetic patients and post mortem tissues in the proband; it was also found in seven offspring of female patients. It was not found in the two children of the male proband. The contribution of this mutation to DM in general is not known but clinicians ought to be aware of the possibility, especially in families with multisystem disease and maternal transmission.

GH FEEDBACK OCCURS THROUGH MODULATION OF HYPOTHALAMIC SOMATOSTATIN UNDER CHOLINERGIC CONTROL: STUDIES WITH PYRIDOSTIGMINE AND GHRH

We have studied the effect of increased cholinergic tone on the GH response to growth hormone‐releasing hormone (GHRH) and on GH feedback, using pyridostigmine, an acetylcholinesterase inhibitor. In six healthy male adult volunteers 120 mg oral pyridostigmine increased basal GH secretion compared to placebo and augmented the GH response to 100 μg i.v. GHRH (1‐29) NH2; the effect was more than the additive effect of pyridostigmine and GHRH when each was given alone. Pretreatment with 2 IU methionyl‐hGH given i.v. abolished the serum GH response to GHRH given 3 h later, demonstrating a negative feedback loop of GH on the response to GHRH; this inhibited response to GHRH was restored in subjects given pyridostigmine as well as methionyl‐hGH. The data demonstrate that enhanced cholinergic tone releases GH, augments the serum GH response to GHRH and unblocks the negative feedback effect of methionyl‐hGH pretreatment on the GH response to GHRH. These results suggest that GH negative feedback effects on its own secretion occur predominantly through increased hypothalamic somatostatin secretion; this somatostatin secretion is under inhibitory cholinergic control.

Acquired growth hormone resistance in patients with hypercatabolism

Sepsis, surgery and critical illness are associated with an increased catabolic rate, which if prolonged delays recovery and increases morbidity and mortality. There is evidence that changes in the GH/IGF-I axis are permissive to protein catabolism. Critically ill, septic patients have high basal levels of GH, low levels of IGF-I and its carrier binding protein IGFBP-3, high levels of an inhibitory binding protein, IGFBP-I, and increased serum protease activity which reduces the affinity of IGFBP-3 for IGF-I. Overall there is a reduction in the indirect IGF-I-mediated anabolic actions of GH and an increase in the direct catabolic actions of GH. These physiological changes may be adaptive when a sick patient is fasting; however, the availability of modern intensive care means that these changes are no longer an advantage. GH and IGF-I, in pharmacological doses, promote positive nitrogen balance, in both animal models and man. Preliminary studies with IGF-I in postsurgical patients suggest that it may provide a practical therapy. Future studies need to focus on outcome measures in relation to the use of GH and IGF-I as anabolic therapies.

Effects of growth hormone replacement on physical performance and body composition in GH deficient adults

Adults with GH deficiency complain frequently of low energy levels resulting in a low perceived quality of life. Body composition is altered, with increased fat mass and decreased lean body mass, and muscle strength is reduced. The aims of this study were to determine the effects of GH replacement on physical performance and body composition in GH deficient (GHD) adults.

GROWTH HORMONE PRETREATMENT IN MAN BLOCKS THE RESPONSE TO GROWTH HORMONE-RELEASING HORMONE; EVIDENCE FOR A DIRECT EFFECT OF GROWTH HORMONE

The effect of pretreatment with biosynthetic methionyl human GH (hGH) on the GH response to GHRH has been studied in normal subjects. Eight volunteers were given either 4IU hGH or placebo s.c. 12‐hourly for 72 h before a GHRH test, or a single s.c. dose of 4 IU hGH 12 h before a GHRH test. Somatomedin‐C (Sm‐C) levels at the time of the GHRH tests were significantly elevated after treatment with hGH compared to placebo, and the GH response to GHRH was significantly attenuated. A further six subjects were given 2 IU hGH or placebo i.v., and i.v. GHRH 3 h later; there was no rise in Sm‐C for the 5 h of the study after either treatment; nevertheless, the response to GHRH was completely abolished by pretreatment with hGH. These results demonstrate that GH can regulate its own secretion independently of changes in Sm‐C levels, through a mechanism other than the inhibition of GHRH release. The attenuated response to GHRH in the presence of elevated Sm‐C levels may be related to Sm‐C, or be a more direct effect of the recently elevated GH levels.

THE EFFECT OF DOPAMINE AGONIST THERAPY ON LARGE FUNCTIONLESS PITUITARY TUMOURS

Fifteen patients (12 male) with large pituitary tumours and serum prolactin levels below 1000 mU/1 were given dopamine agonist therapy (bromocriptine, mesulergine or pergolide) for a mean of 9 months (range 3–36 months). Serum prolactin became undetectable in all. Despite this, significant suprasellar extensions and any associated neurological defect remained in 14 patients, who therefore were referred for surgery. In one patient there was evidence of spontaneous pituitary infarction unrelated to dopamine agonist therapy. At operation 12 patients had apparently functionless pituitary adenomas which failed to immunostain for prolactin, one had an epidermoid cyst and one a Rathkes pouch cyst. We conclude that patients with large pituitary tumours and only a mildly elevated serum prolactin are unlikely to have prolactinomas, and that such tumours are not likely to show significant tumour shrinkage with medical treatment with dopamine agonists.

THE RELATIONSHIP BETWEEN SERUM PROLACTIN AND IMMUNOCYTOCHEMICAL STAINING FOR PROLACTIN IN PATIENTS WITH PITUITARY MACROADENOMAS

We have studied the relationship between mean pretreatment levels of serum prolactin and the presence of positive immunohistochemical staining for prolactin in the pituitary tumours of 55 patients. Pretreatment serum prolactin was significantly higher in patients with tumours showing many prolactin immunostaining cells than in those with none (P>0·001). When the pretreatment serum prolactin exceeded 6000 mU/1, the tumours contained over 90% of prolactin positive cells; one patient was an exception who had received longterm high dose bromocriptine therapy, and her tumour showed only occasional cells with positive staining. When the pretreatment serum prolactin level was under 2500 mU/1, a tumour was found which showed either no cells or fewer than 1% of cells which stained for prolactin. There was no significant difference in pretreatment serum prolactin levels between 11 patients with craniopharyn‐giomas and 34 patients with pituitary macroadenomas showing no prolactin immunostaining. Seventy‐one percent (32) of the 45 patients with craniophar‐yngiomas or tumours with negative immunostaining for prolactin, had raised pretreatment serum prolactin levels (above 360 mU/1) although this was usually only slightly elevated; the levels exceeded 2500 mU/1 in six (13%) of them (two craniopharyngiomas, four pituitary tumours) but in none did the levels exceed 6000 mU/1. Four of the 55 pituitary tumours showed occasional cells (less than 1%) that stained positively for growth hormone. In none of the patients with these tumours was there evidence of acromegaly or pathologically elevated circulating growth hormone levels.

Depot-bromocriptine treatment for prolactinomas and acromegaly.

Fifteen patients with hyperprolactinaemia and pituitary macroadenomas (5 patients), microadenomas (6 patients), or acromegaly (4 patients) were given a single intramuscular injection of 50 mg bromocriptine bound to polylactic acid microspheres, depot‐bromocriptine. None of the patients had any short‐term or long‐term discomfort from the injection. In the 11 patients with prolactinomas, serum prolactin fell to minimum levels 12–72 h post‐injection; nine patients were highly responsive to depot‐bromocriptine, with a mean serum prolactin of 12·9% of basal levels 24 h post‐injection, rising to 19% at 28 days. Two patients with prolactinomas were resistant to both depot‐bromocriptine, and large doses of oral dopamine agonists. Initiating side‐effects (nausea, vomiting, symptomatic postural hypotension) were seen in five patients in the first 24 h post‐injection, but were minimal or absent thereafter. Five of six patients previously intolerant of oral dopamine agonists were able to be transferred successfully to bromocriptine 5 mg daily at 4 weeks. Of the four patients with acromegaly, raised prolactin levels were successfully lowered to normal for 4 weeks after injection; serum GH was also partially lowered, but returned to baseline levels at 2–4 weeks. In one patient serum GH was resistant to suppression by both depot bromocriptine and high doses of oral bromocriptine. One patient with a large tumour and visual field defects showed a rapid and maintained improvement in visual fields and acuity after depot‐bromocriptine, and was successfully transferred to high‐dose oral bromocriptine at 4 weeks. In the total group of 15 patients, computed tomography (CT) scans showed clear tumour shrinkage in seven patients 4 weeks post‐injection. It is concluded that depot‐bromocriptine is a significant improvement in the initiation of treatment for hyperprolactinaemia and acromegaly. It allows the rapid diagnosis of bromocriptine resistance, and the initiation onto oral dopamine agonist therapy of some patients said to be intolerant of such therapy. Furthermore, it may well be the treatment of choice for rapid shrinkage of large pituitary tumours sensitive to dopamine agonist therapy.

The Use of Growth Hormone-Releasing Hormone in the Diagnosis and Treatment of Short Stature

We have assessed the role of growth hormone-releasing hormone (GHRH) as a diagnostic test in 40 children and young adults with growth hormone deficiency (GHD), principally using the GHRH(1-29)NH2 analogue. Following 200 micrograms GHRH as an acute intravenous bolus, serum GH rose to normal or just subnormal levels in 13 out of 17 children with structural lesions, and in 8 of 14 patients with idiopathic GHD or panhypopituitarism. Of 9 children (mean age 12 years) with GHD following treatment with cranial irradiation for nonendocrine tumours, all responded acutely to GHRH. 12- and 24-hour infusions with GHRH(1-29)NH2, and 1- and 2-week treatments with twice-daily subcutaneous GHRH(1-29)NH2, showed persistent stimulation of GH release. It is concluded that many children with GHD of diverse aetiology will respond both acutely and chronically to treatment with GHRH.

Serum leptin and leptin binding activity in children and adolescents with hypothalamic dysfunction

Marked disturbance in eating behaviour and obesity are common sequelae of hypothalamic damage. To investigate whether these were associated with dysfunctional leptin central feedback, we evaluated serum leptin and leptin binding activity in 37 patients (age 3.5-21 yr) with tumour or trauma involving the hypothalamic-pituitary axis compared with 138 healthy children (age 5.0-18.2 yr). Patients were subdivided by BMI <2 SDS or > or = 2 SDS and healthy children and children with simple obesity of comparable age and pubertal status served as controls. Patients had higher BMI (mean 1.9 vs 0.2 SDS; p <0.001), a greater proportion had BMI > or = 2 SDS (54% vs 8%; p <0.001) and higher serum leptin (mean 2.1 vs 0.04 SDS; p <0.001) than healthy children. Serum leptin (mean 1.1 vs -0.1 SDS; p = 0.004) and values adjusted for BMI (median 0.42 vs 0.23 microg/l:kg/m2; p = 0.02) were higher in patients with BMI <2 SDS. However, serum leptin adjusted for BMI was similar in patients with BMI > or = 2 SDS compared to corresponding controls (1.08 vs 0.95; p = 0.6). Log serum leptin correlated with BMI SDS in all subject groups but the relationship in patients with BMI <2 SDS was of higher magnitude (r = 0.65, slope = 0.29, p =0.05 for difference between slopes) than in healthy controls (r = 0.42, slope = 0.19). Serum leptin binding activity (median 7.5 vs 9.3%; p = 0.02) and values adjusted for BMI (median 0.28 vs 0.48 % x m2/kg; p <0.001) were lower in patients than in healthy children. The markedly elevated leptin levels with increasing BMI in non-obese patients with hypothalamic-pituitary damage are suggestive of an unrestrained pattern of leptin secretion. This along with low leptin binding activity and hence higher free leptin levels would be consistent with central leptin insensitivity.