Timing of onset of growth hormone deficiency is a major influence on insulin-like growth factor I status in adult life.

OBJECTIVE: Several reports have suggested that IGF-I levels in patients with childhood-onset (CO) GH deficiency are lower than those observed in patients with adult-onset (AO) GH deficiency. However, these reports are unsatisfactory as there are several differences between the cohorts studied other than the timing of onset of GH deficiency; in particular, the patients were not matched for equal severity of GH deficiency. We have pursued this question further by examining the IGF-I standard deviation score (SDS) in patients with CO and AO GH deficiency, with equal degrees of severity of GH deficiency, as defined by the peak GH response to the insulin tolerance test (ITT). PATIENTS AND MEASUREMENTS: IGF-I SDS were compared in 146 non acromegalic patients (69 male), aged 15.7-76.6 years (median 33.4 years), bone mass index (BMI) 27.8 +/- 5.8 kg/m2, with severe GH deficiency (peak GH response < 9 mU/l to insulin-induced hypoglycaemia). Patients were subdivided by timing of onset of GH deficiency and the peak GH response to the ITT (GH response < or = 1 mU/l, group 1; > 1-3 mU/l, group 2; > 3-6 mU/l, group 3; > 6-8.9 mU/l, group 4). RESULTS: The IGF-I SDS (mean value +/- SD) in the CO group (n = 63) as a whole was significantly lower than that found in the AO group (n = 83) (-3.7 +/- 2.8 vs.-1.55 +/- 2.2, respectively; P < 0.0001). Despite this, there was no significant difference in the peak GH response to an ITT between the two cohorts (2.8 +/- 2.3 mU/l in the AO cohort and 2.6 +/- 2.2 mU/l in the CO cohort; P = 0.5). When the cohorts were subdivided by severity of GH deficiency, there remained a significant difference in IGF-I SDS in groups 1 (P < 0.0001), 2 (P = 0.05) and 3 (P < 0.05), but there was no significant difference between the AO and CO cohorts in group 4. The peak GH response to an ITT was similar in the AO and CO cohorts in all groups (P = 0.8, 0.8, 0.9 and 0.3 in groups 1-4, respectively). Although increasing severity of hypopituitarism was associated with a decline in IGF-I SDS in the CO cohort (P < 0.01), this was not the case in the AO cohort (P = 0.3). CONCLUSION: These data support the hypothesis that there is an innate difference between adult patients with either CO or AO GH deficiency that cannot be explained solely by variation in the severity of GH deficiency. A possible explanation is that childhood GH deficiency programmes the subsequent relationship between GH and IGF-I in adult life or that the body composition changes, which are more severe in AO GH deficiency, influence IGF-I status.

Low-dose GH replacement improves the adverse lipid profile associated with the adult GH deficiency syndrome.

OBJECTIVE: Adult growth hormone deficiency (AGHD) is associated with an adverse lipid profile. The majority of previous studies of GH replacement have used supraphysiological doses and reported favourable changes in the lipid profile. Whether this beneficial effect is the result of pharmacological GH therapy, or occurs in response to low-dose GH replacement aimed at normalization of the serum IGF-I, has not been fully elucidated. STUDY DESIGN: We studied 67 patients with GH deficiency using a low-dose individualized GH replacement regimen. GH was commenced at a dose of 0.27 mg/day and the GH dose titrated until the serum IGF-I was normalized. Serum lipids were assessed at baseline, 12 and 24 months. RESULTS: A reduction in total cholesterol (TC) was observed at 12 (6.01 vs. 5.77 mmol, P = 0.04) and 24 months (6.01 vs. 5.56, P = 0.09). The reduction in LDLC failed to reach significance at 12 months (3.97 vs. 3.8, P = 0.09), but was significant at 24 months (3.97 vs. 3.50, P = 0.02). Levels of HDLC did not change significantly at 12 or 24 months. Significant improvements in the TC/HDLC ratio were observed at both 12 (5.68 vs. 5.29, P = 0.01) and 24 months (5.68 vs. 4.86, P = 0.007). A significant fall in triglycerides (TG) was present at 12 months (2.07 vs. 1.83, P = 0.01), and was maintained at 24 months, but was no longer significant (2.07 vs. 1.89, P = 0.28). At 12 months there was no correlation between improvements in lipid parameters and either the change in IGF-I SD score or the GH dose. Using multivariate analysis the change in TC, LDLC and the TC/HDLC ratio with 12 months GH replacement were determined by the baseline TC, LDLC and TC/HDLC levels (R2 = 0.18, P = 0.004; R2 = 0.20, P = 0.006; and R2 = 0.33, P < 0.0001), respectively. CONCLUSIONS: Low-dose individualized GH replacement aimed at normalization of the serum IGF-I is associated with significant improvements in TC, LDLC, TGs and the TC/HDLC ratio. The greatest improvements are observed in patients with the most adverse lipid profiles at baseline. Improvements are independent of changes in the IGF-I SDS and GH dose.

The acute leptin response to GH.

The effect of an acute bolus of GH on serum leptin in normal individuals and the factors affecting this response have not previously been studied. Seventeen healthy volunteers with normal body mass index, with ages ranging from 20.5-78.2 yr were studied. Each subject received three single doses of GH in random order at least 4 wk apart. Bioimpedence analysis was performed to provide estimates of fat and lean masses. Serum samples for leptin, insulin, and IGF-I were taken 0, 18, 24, 48, 72, and 120 h after each dose of GH. Leptin levels changed significantly after the 0.67- and 7-mg doses of GH, but not after the 0.27-mg dose. Compared with baseline, there was a significant elevation (P < 0.001) in serum leptin levels at 24 h, followed by a significant decrease (P < 0.01) at 72 h. Baseline and peak leptin levels were significantly determined by gender, fat mass, and log(10) insulin. Nadir leptin levels were significantly determined by gender and fat mass. In contrast, the increment in leptin levels was significantly determined by age, although this only accounted for 24% of the variability in the increment in leptin levels. We have demonstrated that administration of a single bolus dose of GH significantly increases serum leptin levels, followed by a significant nadir. This occurs not only after a supraphysiological dose of GH, but also after 0.67 mg, a dose within the physiological replacement range. The increment in leptin increases with advancing age, suggesting that at the level of the adipocyte, aging increases responsiveness to GH. However, this only partially explains the changes seen, and it is likely that another factor(s) is involved in the acute impact of GH on circulating leptin levels. The presence of a significant nadir after the peak in leptin levels supports the existence of a negative feedback loop, linking circulating leptin to its own biosynthesis in adipose tissue, mediated by peripheral leptin receptors. These data provide unequivocal evidence that GH can affect serum leptin levels in the absence of a change in body composition.

The impact of irradiation on growth hormone responsiveness to provocative agents is stimulus dependent: results in 161 individuals with radiation damage to the somatotropic axis.

GH provocative tests remain the mainstay for the diagnosis of GH deficiency and at present the insulin tolerance test (ITT) is the gold standard. There are, however, a variety of other stimulation tests used in clinical practice. Each necessitates the use of a specific cut-off derived from normative data, but there remains a widely held view that the implications from a "failed" test are independent of the nature of the stimulus. We sought to examine whether this is the case in individuals with evidence of radiation damage to the somatotropic axis. One hundred and sixty-one nonacromegalic patients were identified who had undergone an arginine stimulation test (AST) and an ITT within a 3-month period as part of routine testing between 1975 and 1999. They were divided into those tested before (n = 81; 48 males) and those tested after (n = 80; 36 males) completion of growth and puberty. Patients were considered for inclusion in the study if they had a history of cranial irradiation and a GH response to one provocative test of less than 8 microg/L, taken as indicating that some damage to the GH axis may have occurred. The patients were compared with 2 control groups. The first comprised 35 adults (18 males) and the second consisted of 16 prepubertal children (10 males). The median peak (range) GH response to the ITT was significantly greater (P < 0.0001) than that to the AST in the adult controls: 24.9 (4.1--76.9) vs. 12.2 (0.88--35.0) microg/L, respectively. However, in the patients the GH responses were similar (P = 0.28): 2.2 (0.2--25.7) vs. 1.4 (0.2--12.8) microg/L to the ITT and AST, respectively. In contrast to the pattern seen in the adult controls, the response to an ITT in childhood controls was of similar magnitude (P = 0.5) to that to the AST: 17.5 (8.1--40.0) vs. 19.4 (7.3--53.8) microg/L, respectively. However in the patients, the GH response to the AST was greater than that to the ITT (P < 0.0001): 4.3 (0.7--17.2) vs. 3.0 (0.4--18.1) microg/L, respectively. In summary, we have shown that the impact of irradiation on GH responsiveness to provocative agents is stimulus dependent. The GH response to an AST appears to be more resistant to the effects of irradiation than that to the ITT. When investigating the impact of irradiation on GH secretory status, the GH response to an AST may be a less sensitive guide to the functional ability of the GH axis.

Management of pituitary tumours: strategy for investigation and follow-up.

Investigation of a patient presenting with evidence of a pituitary tumour has three main objectives: investigation of any hormonal hypersecretion; assessment of residual pituitary function, and examination of any mass effect of the tumour. A prolactin-secreting adenoma is often easily diagnosed by performance of a basal prolactin level. Biochemical assessment to exclude acromegaly or Cushing's disease should only be performed if clinically indicated. The standard investigations for acromegaly consist of establishing the degree of growth hormone (GH) suppression following a glucose load and estimating the basal insulin-like growth factor-I level. Before detailed investigation for Cushing's disease is initiated, the presence of Cushing's syndrome must be established. The second requirement is to determine the presence of any pituitary hyposecretion. Whilst the remainder of pituitary function can be assessed by baseline hormonal estimations, the evaluation of ACTH and GH secretion necessitates dynamic function testing. Lastly, the impact of the mass itself requires careful examination. Both neuroradiology, preferably magnetic resonance imaging at a centre specialized in examination of the pituitary fossa, and careful detailed clinical examination of the visual fields should be performed. The follow-up requirements in an individual patient are affected by a number of factors including the size and nature of the underlying tumour and any treatment administered. In patients with a hormone-secreting tumour, the hormone levels themselves provide a 'tumour marker' to aid follow-up. An important caveat, however, is that on some occasions tumour size and hormone levels do not change in parallel. Patients who have undergone pituitary surgery should have dynamic assessment of pituitary function performed approximately 6 weeks after surgery. There is no reason to suspect any further impairment of pituitary function after this date. In direct contrast, pituitary hormone deficiencies after radiotherapy are unlikely less than 6 months after treatment. Patients should undergo testing of pituitary reserve at 6 months, and then at yearly intervals for at least 10 years after radiotherapy, if they have not already developed panhypopituitarism. Even after this period, if patients develop new symptoms the possibility of further pituitary hormone deficits should be considered. Neuroradiology should be performed approximately 6 weeks to 3 months after surgery. If radiotherapy is not administered, neuroradiology should be performed yearly for at least 10 years. If the patient has received radiotherapy, tumour recurrence is much less likely and therefore in these individuals neuroradiology does not need to be performed with such regularity. In conclusion, when planning the investigation and follow-up of an individual patient, one should take into account the size and characteristics of the tumour, as well as the treatment modalities.

Effects of growth hormone on bone and muscle.

The decade since the initial availability of recombinant growth hormone (GH) has seen an increase in our understanding of the effects of GH on muscle and bone. Adult GH deficiency (GHD) is associated with osteopenia, the severity of which is related to three factors: the timing, age of onset and severity of GHD. Epidemiological data suggest that this osteopenia is associated with an increased risk of fracture. The impact of GH replacement therapy on bone mineral density (BMD) appears to be related to a large number of interrelated factors, including the dose and duration of therapy, timing of onset of GHD, skeletal site, degree of osteopenia at baseline, and age and gender of the patient. Overall, the effect of GH replacement on BMD in the majority of patients is beneficial. As yet, however, no data are available that demonstrate a reduction in fracture rate following GH therapy. In comparison with normal individuals, GH-deficient individuals have reduced lean body mass and muscle strength, both of which increase within 12 months of GH therapy. Therefore, the effects of GH replacement on muscle and bone in GH-deficient individuals are significant and beneficial, although the longer-term effects of GH replacement in terms of reducing the number of fractures and prevention of frailty in old age are not yet established. The effects of GH on bone and muscle in GH-replete individuals have been studied less fully. While GH therapy modulates markers of bone resorption and formation, its effects in patients with idiopathic osteoporosis are disappointing, with oestrogen therapy or bisphosphonates proving to be more effective in post-menopausal women. To date, however, there have been no GH treatment trials of adequate duration (longer than 18 months), and it remains possible that longer-term trials may demonstrate more profound effects. The effects of GH therapy on muscle have been examined in normal elderly individuals. Generally, the doses used have been supraphysiological and associated with an unacceptable incidence of side-effects. GH therapy has resulted in an increase in lean body mass, but functional ability and strength have not improved in the majority of studies. Thus, clear-cut beneficial effects of GH on muscle and bone in GH-replete individuals have not been demonstrated. It seems unlikely that normal elderly individuals will benefit significantly from GH therapy, but frail individuals or those with musculoskeletal or neuromuscular pathology are potential candidates for study.

Pre-treatment IGF-I level is the major determinant of GH dosage in adult GH deficiency.

OBJECTIVE: Severe GH deficiency in adults is a definite clinical entity, the effects of which can be reversed by administration of subcutaneous recombinant GH. The ideal dosing regimen and determinants of the maintenance dose have, however, yet to be elucidated. PATIENTS: In an open study of GH replacement we treated 65 GH-deficient adults of mixed adult- and childhood-onset, of mean age 35.5 (range 17-72) years, and comprising 38 females and 27 males, using an individualized low-dose titration regimen aimed at normalization of the serum IGF-I and induction of clinical improvement. RESULTS: Before initiation of GH therapy, median IGF-I SD was significantly lower in female than male patients (- 3.3 vs. - 1.9, P = 0.007) and in childhood-onset compared with adult-onset patients (- 3.9 vs. - 2.0, P < 0.001). Once maintenance dosage had been achieved, the median GH requirement was significantly greater in female than male patients (1.6 vs. 0.8 IU/day, P = 0.013) and childhood-onset compared with adult-onset patients (1.6 vs. 0.8 IU/day, P = 0.019). The median maintenance GH dose for the cohort overall was 1.2 (range 0.4-2.4) IU/day. By univariate analysis a significant negative correlation was observed between the maintenance GH dose and baseline IGF-I SD (r = - 0.63, P < 0.001). No significant correlation was demonstrated between maintenance GH dose and either age or weight. Multiple linear regression analysis using age, gender, weight, time of onset of GH deficiency, peak GH to the insulin tolerance test (ITT) and baseline IGF-I SD as independent variables demonstrated baseline IGF-I SD to account for 51% of the variation in GH dose required to normalize the IGF-I SD (P < 0.001). Those patients with the lower IGF-I SD at initiation of GH therapy required the greater GH dose. None of the other variables studied significantly influenced the maintenance dose. CONCLUSION: We have demonstrated that the GH dose required in an individual is dependent on the serum IGF-I SD before commencement of replacement. In contrast, the severity of GH deficiency as judged by the peak GH response to an ITT was unrelated to the maintenance GH requirement. The effect of age, gender and age at onset of GH deficiency on the final GH dose are accounted for by the lower pretreatment IGF-I SD in young, female and childhood-onset patients relative to older, male and adult-onset patients, respectively.

How many tests are required to diagnose growth hormone (GH) deficiency in adults?

OBJECTIVE: The diagnosis of GH deficiency in adults relies on the results of GH provocative testing. Whilst in some patients the testing strategy is clear, this is not the case in all patients. The objective of this study was to further examine the concordance between the GH responses to two different provocative stimuli, to correlate this with the number of additional pituitary hormone deficits, and to produce guidelines as to which patients require two GH provocative tests and which require only one. STUDY DESIGN AND PATIENTS: The results of GH provocative tests were reviewed in 103 patients (mean age 28 years, 48 male), with documented or potential hypothalamic-pituitary disease and 35 normal volunteers (mean age 21 years, 18 male). All patients and normal volunteers underwent an insulin tolerance test (ITT) and an arginine stimulation test (AST). Severe GH deficiency was defined as a GH response to an ITT of < 5 mU/l and a GH response to an AST of < 2 mU/l, utilizing data from previous studies in this unit. Patients were divided into four groups according to the number of anterior pituitary hormone deficits present other than possible GH deficiency: no other pituitary hormone deficits (GHD0) or one, two or three other hormone deficits (GHD1, GHD2 or GHD3). RESULTS: The 103 patients were divided between the four groups as follows: 69 (67%) in GHD0, 15 (14. 6%) in GHD1, six (5.8%) in GHD2, and 13 (12.6%) in GHD3. There was a significant decline in the median GH peak to both the ITT and the AST with increasing numbers of other pituitary hormone deficits (P < 0.0001). If the magnitude of the difference between each individual's GH response to the ITT and AST is plotted against the mean GH value a clear trend is seen (Spearmans rank correlation = 0. 88, P < 0.0001) indicating that the magnitude of the difference between the GH responses to an ITT and AST increases with the underlying mean GH value. These data allow the estimation of the median ITT/AST ratio as 1.17 (CI 0.98, 1.39). None of the control subjects and 14.1% (10), 26.7% (four), 83% (five) and 92.3% (12) of groups GHD0, 1, 2 and 3, respectively, had severe GHD. The concordance between the AST and ITT (percent of patients in whom both tests confirmed or refuted the biochemical diagnosis of severe GHD) was 100%, 76.8%, 66.6%, 83.3%, and 92.3% in the controls, GHD0, 1, 2, and 3, respectively. Thus, 16/69 GHD0, 5/15 GHD1, 1/6 GHD2 and 1/13 GHD3 patients were misclassified by one or other test. CONCLUSION: We have demonstrated that a constant ratio links the GH response to an ITT and AST in an individual, rather than a constant difference, and that the difference between the GH responses to two provocative stimuli is greater in those patients with milder degrees of GH deficiency or insufficiency. These patients tend to have one or no additional pituitary hormone deficits and may be misclassified if a single GH provocative test is performed. We suggest that whilst a single GH provocative test can be used with confidence in patients with two or three additional pituitary hormone deficits, in patients with suspected isolated GH deficiency or with only one additional pituitary hormone deficit, two GH provocative tests should be performed.

Influences on quality of life in GH deficient adults and their effect on response to treatment.

OBJECTIVE: Studies of the effect of GH on quality of life (QOL) in growth hormone deficient (GHD) adults have reported conflicting results. Recently, however, we have demonstrated that by selecting only those patients with impaired QOL the efficacy of GH replacement on QOL can be greatly improved. The improvement in QOL was observed to correlate significantly with that recorded before commencing GH therapy. This study aims to assess if demographic variables affect QOL in untreated GHD adults or the improvement in QOL following GH therapy. DESIGN: An open study of GH replacement, initiating treatment with a dose of 0.8 IU/day and titrating the dose by 0.4 IU increments to normalize the IGF-I SDS between - 2.0 and + 2.0 SD of the age related normal range. PATIENTS: 65 severely GHD patients (peak GH < 9 mU/l to provocative testing), mean age 38.7 (range 17-72) years. Inclusion criterion was that of subjectively poor quality of life on clinical interview. MEASUREMENTS: Blood was taken for insulin-like growth factor 1 (IGF-I). The Psychological General Well-Being Schedule (PGWB) and Adult Growth Hormone Deficiency Assessment (AGHDA) self-rating questionnaires were used to assess quality of life at baseline, three and eight months after commencing GH. RESULTS: The patients were subgrouped on the basis of gender, age of onset of GHD, pathology and presence of additional pituitary hormone deficits. The cohort consisted of 40 females and 25 males, 45 of adult-onset (AO) and 20 of childhood-onset (CO). GH deficiency resulted from a hypothalamo-pituitary pathology, or treatment thereof, in 36 patients and as a result of cranial irradiation for a primary brain tumour or prophylaxis in acute lymphoblastic leukaemia in 29 patients. Isolated GH deficiency (IGHD) was present in 25 patients, and 32 patients were demonstrated to have at least two additional pituitary hormone deficits (MPHD). No significant difference was detected between baseline PGWB scores of the subgroups. Multiple linear regression analysis revealed the age of onset of GHD to be a significant determinant of both the baseline PGWB (P = 0.05) and AGHDA (P = 0.025) scores, AO patients perceiving the greater distress. A significant improvement, from baseline, in both QOL scores was observed in all subgroups at three months, and in all subgroups at eight months except IGHD, where a trend towards improvement in the AGHDA score was observed but failed to reach significance. The mean improvement in the PGWB following GH therapy was not significantly different between subgroups. Multiple linear regression analysis confirmed baseline PGWB and AGHDA scores to be the most important variable in prediction of the level of improvement in respective scores following GH therapy. Age of onset was also observed to be a significant determinant of the PGWB scores following GH therapy (P = 0.02), the CO cohort experiencing the greater improvement. A similar relationship between age of onset and AGHDA scores was not observed (P = 0.22). CONCLUSIONS: Baseline QOL as assessed by self-rating questionnaires is influenced by the age of onset of the GH deficiency, adult onset patients expressing the greater distress. Improvements in QOL scores are influenced by both baseline score and to a lesser extent the age of onset of GHD, the greater improvement being observed in childhood onset patients. The degree of improvement was observed to be independent of gender, pathology and number of pituitary hormone deficits. In a cohort selected by subjectively impaired QOL, we have demonstrated childhood onset GHD patients perceive themselves to have less impairment of QOL pretreatment. In contrast to previous data in unselected cohorts, however, we have shown that those childhood onset GHD patients in whom QOL is significantly reduced, show a capacity for improvement that is equal to, if not greater, than that seen in adult onset-GHD patients.

Dose titration and patient selection increases the efficacy of GH replacement in severely GH deficient adults.

OBJECTIVE: Previous studies of GH replacement in adults have used unselected cohorts of GH deficient (GHD) adults and weight-based dosing regimens resulting in supraphysiological serum IGF-I levels and a high frequency of side-effects and withdrawal from these studies. By choosing patients with a high level of morbidity at baseline and using a low dose GH titration regimen we aimed to avoid over-replacement and increase the efficacy of treatment. DESIGN: An open study of GH replacement, initiating treatment with a dose of 0. 8 U/day and titrating the dose by 0.4 U increments to normalize the IGF-I SDS between - 2.0 and + 2.0 SD of the age-related normal range. PATIENTS: 65 severely GHD patients (peak GH < 9 mU/l to provocative testing), 25 males, of mixed adult and childhood-onset and mean age 38.7 (range 17-72) years. Inclusion criterion was that of subjectively poor quality of life on clinical interview. MEASUREMENTS: Height, weight, waist and hip circumference were measured to allow calculation of body mass index (BMI) and waist-hip ratio (WHR). Bone mineral density (BMD) was measured by dual X-ray absorptiometry (DEXA). Serum haemoglobin A1C (HbA1C), lipid profile and insulin like growth factor 1 (IGF-I) were measured. The Psychological General Well-Being Schedule (PGWB) and Adult Growth Hormone Deficiency Assessment (AGHDA) self-rating questionnaires (SRQ) were used to assess quality of life. RESULTS: Baseline characteristics were consistent with those previously described in severely GHD adults; mean IGF-I SDS - 2.4 (+/- 2.7), BMI 28.8 (+/- 5. 4) kg/m2, total cholesterol 6.17 (+/- 1.2) mmol/l, reduced BMD z-scores at the lumbar spine (- 0.8 +/- 1.2) and femoral neck (- 0. 44 +/- 1.4), and SRQ scores considerably lower than reported in previous studies of GH deficient adults and normal controls. Following initiation of GH serum IGF-I SDS was increased significantly from baseline to a mean level of 0.15 +/- 2.7 (P < 0. 001) and 0.31 +/- 2.0 (P < 0.001) at three and eight months, respectively. The mean PGWB score increased from 59.7 +/- 19.9 to 75. 8 +/- 15.0 (P < 0.001) and 73.7 +/- 19.5 (P = 0.001) at three and eight months, respectively. An increase of 14 points represents the largest improvement in quality of life, using this index, that has been reported in GHD adults. The mean AGHDA score also demonstrated considerable improvement, falling from 15.3 +/- 6.0 to 10.4 +/- 6.2 (P < 0.001) and 9.8 +/- 6.5 (P < 0.001) at three and eight months, respectively. The changes observed in both the PGWB and AGHDA scores between baseline and at both three and eight months were shown to correlate significantly with the respective baseline score. A significantly greater improvement was observed in the PGWB following GH replacement in those with a baseline PGWB score of < 60 than in those with a score > 60. This observation was significant at both three (27.1 vs 6.7, P = 0.0001) and eight (25.6 vs 3.3, P = 0.0003) months. All PGWB subscales showed significant improvement though that of vitality was of greatest magnitude. A strong correlation was observed between the generic and disease-specific SRQ (r = - 0.73, P < 0.001). CONCLUSIONS: The observed improvement in quality of life in GH deficient adults is proportional to the degree of impairment before commencing therapy. The use of low-dose titration and selection of a population with greater morbidity reduces the occurrence of over-replacement and increases the efficacy of treatment. This allows direction of resources to those in greatest need.

Growth hormone status following treatment for Cushing's syndrome.

OBJECTIVE: Both pituitary surgery and radiotherapy for Cushing's disease can lead to growth hormone (GH) deficiency. Studies to date have, however, described the incidence of impaired GH secretion and not the incidence of severe GH deficiency following treatment of Cushing's disease. Furthermore, following cure of Cushing's disease and resolution of hypercortisolaemia, recovery of GH secretory status is seen, thus creating uncertainty as to the persistence of any documented GH deficiency. This study has two aims; to determine the incidence of severe persistent GH deficiency following treatment of Cushing's disease and to assess the time scale of any recovery of GH secretory status following surgical cure of Cushing's disease. DESIGN AND PATIENTS: The case notes of 37 patients either cured or in clinical and biochemical remission following treatment for Cushing's syndrome were reviewed to determine the incidence of severe GH deficiency. Of 34 patients with Cushing's disease, 20 were treated by pituitary surgery, and 14 with radiotherapy. Three patients with adrenal adenomas underwent unilateral adrenalectomy. MEASUREMENTS: GH secretory status was assessed by provocative testing using an insulin tolerance test (ITT, 85% of all tests), glucagon stimulation test (GST) or arginine stimulation test (AST). RESULTS: Thirty-six percent (5/14) of radiotherapy treated patients demonstrated severe GH deficiency at a mean time of 99 months following remission. Fifty-nine percent (10/17) of surgically treated patients assessed in the two years following remission demonstrated severe GH deficiency, whilst only 22% (2/9) of patients assessed beyond two years following remission demonstrated severe GH deficiency. This latter cohort is biased, with patients in whom severe GH deficiency had been demonstrated on earlier tests being over-represented. It is more accurate to estimate the incidence of persistent severe GH deficiency following surgically induced remission of Cushing's disease by incorporating data from patients in whom original testing demonstrated adequate GH reserve. Collating such data, 13% (2/15) of patients had persistent severe GH deficiency. Across all time periods five surgically treated patients demonstrated recovery of GH secretory status over a median time course of 19 months. In the surgically treated cohort, seven (35%) patients had anterior pituitary hormone deficits other than GH deficiency: 14% (2/14) of patients with normal GH secretory status at the last assessment, 83% (5/6) of patients with severe GHD at the last assessment. Of the 5 patients who demonstrated recovery of GH secretory status 40% (2) had additional anterior pituitary hormone deficits. Within the radiotherapy treated cohort 14% (2/14) of patients demonstrated additional anterior pituitary hormone deficits: 11% (1/9) of patients with normal GH secretory status and 20% (1/5) of patients with severe GH deficiency. None of the patients with adrenal adenomas treated by unilateral adrenalectomy demonstrated any abnormality of GH secretory status CONCLUSIONS: The incidence of severe persistent GH deficiency following surgically induced or radiotherapy induced remission of Cushing's disease is lower than has been suggested by previous studies, although these latter studies have assessed GH insufficiency and not severe GH deficiency. In the presence of additional pituitary hormone deficits severe GHD is common and is likely to be persistent. Recovery of GH secretory status is seen in a high proportion of patients reassessed, at a median time of 19 months following surgically induced remission of Cushing's disease. Thus, we recommend that definitive assessment of GH secretory status is delayed for at least two years following surgical cure of Cushing's disease. This has important implications for patients in whom GH replacement therapy is being considered.

Growth hormone replacement in an adult with mild growth hormone deficiency and hereditary motor and sensory neuropathy: growth hormone restores independent mobility.

We present the case of an adult patient with growth hormone (GH) insufficiency and hereditary motor and sensory neuropathy type 1. Stopping GH replacement at the attainment of final height was associated with a marked reduction in power and mobility, resulting in the patient becoming wheelchair bound. GH replacement was assessed in a double-blind placebo-controlled trial. During the GH replacement arm of the trial, the patient's mobility and independence returned to previous levels. We suggest that the indications for GH replacement in adults should take account of other medical problems, in particular neuromuscular disorders, as well as the degree of GH deficiency.

The outcome of surgery for acromegaly: the need for a specialist pituitary surgeon for all types of growth hormone (GH) secreting adenoma.

OBJECTIVE: Acromegaly is associated with reduced life expectancy, while therapeutic 'cure' (defined by achievement of GH levels < 5 mU/l) is associated with normalization of life expectancy. Surgery remains the treatment of choice but in those in whom operative 'cure' is not achieved, radiotherapy and/or medical treatment are valuable treatment modalities. The chance of subsequent 'cure' with radiotherapy or somatostatin analogue therapy is increased if the post-operative GH level is reduced below 30 mU/l. Using strict criteria for cure and a single dedicated pituitary surgeon, two large European studies reported 'cure' rates of 42% and 56%. In the Manchester region, surgery for these patients has been performed by a number of neurosurgeons, with no specific designated pituitary surgeon dominating the picture. We wished to examine the impact of this surgical strategy on cure rates and the incidence of a post-operative GH level below 30 mU/l. DESIGN: We reviewed the GH results between 1974 and 1997 for every acromegalic who had been referred to the endocrine departments of the two Manchester hospitals responsible for the majority of pituitary disease referrals in Manchester and who had been subsequently referred for pituitary surgery. PATIENTS AND MEASUREMENTS: Seventy-three (33 male) patients had had GH status assessed before and after surgery by an OGTT or GH profile. The patients were aged between 19 and 70 (mean 43) years at surgery. Seventy-one underwent transsphenoidal and 2 transfrontal surgery. Nine surgeons performed operations. RESULTS: Eighteen (24.7%) had microadenomas and 51 (69.9%) macroadenomas. In 4 patients (5.5%) insufficient data were available to size the adenoma. 17.8% of patients were cured by surgery, 38.8% with microadenomas and 11.8% with macroadenomas. In addition, of 52 patients whose GH levels were > 30 mU/l before surgery, only 27 (51.9%) had GH levels below 30 mU/l post-operatively (81.8% of microadenomas, 43.2% of macroadenomas). CONCLUSION: In comparison with other series, the cure rate in this study is significantly lower. The success in reducing GH levels below 30 mU/l post-operatively is difficult to compare with previously published studies, as few groups have analysed their data in this manner. Nonetheless, of our acromegalic patients with a pretreatment GH level in excess of 30 mU/l, nearly 50% have similar GH status postoperatively, thereby rendering them less amenable to cure by alternative therapeutic modalities. This highlights the importance of a specialist pituitary surgeon, not only for GH secreting microadenomas but also for GH secreting macroadenomas. If these patients are not 'cured', the cost of continuing therapy becomes a significant burden on health-care costs. In addition, if the postoperative GH levels remain above 30 mU/l the chances of achieving adequate control of GH levels are greatly reduced, thereby increasing mortality rates as well as morbidity in these patients.