the bone mineral study. There were 59 males (mean age 54.0 years, range years) and 36 females (mean

Transcription

1 Reduced bone mineral content in adult patients with growth hormone deficiency Thord Ros\le'\n, Tommy Hansson, Hans Granhed, Janos Szucs and Bengt-\lA%o\keBengtsson Divisions of Endocrinology1, Orthopaedic Surgery2 and Radiophysics1. Sahlgrenska Hospital, Medical Faculty. University of Göteborg, Göteborg. Sweden Rosén T, Hansson T, Granhed H, Szucs J, Bengtsson B-Â. Reduced bone mineral content in adult patients with growth hormone deficiency. Acta Endocrinol 1993:129:21-6. ISSN Bone mineral content was measured in a follow-up study of adult patients with hypopituitarism and growth hormone deficiency. There were 95 patients (59 males, mean age 54. years, range 21\pn-\74 years; 36 females, mean age 53.5 years, range 31\pn-\73years). Routine replacement therapy with cortisone acetate and l-thyroxine was given. All males that were gonadal deficient were on proper testosterone therapy, except in four patients who were treated separately. Bone mineral content was measured using dual-photon absorptiometry in the third lumbar vertebra. Bone mineral content in the patients was compared with a control population (N413, 25\pn-\74years of age). Bone mineral content was significantly lower in males (N 55, p<.5) compared with controls. In females, bone mineral content was significantly lower both among the subjects with untreated gonadal deficiency (p<.1) and among those with treated gonadal deficiency and normal premenopausal gonadal function (p<.5) compared with controls. To summarize, patients with hypopituitarism on routine replacement therapy but not growth hormone have a lower bone mineral content than the controls. The reduced bone mineral content might be a result of untreated growth hormone deficiency. Thord Rosen, Endocrine Division, Department of Internal Medicine, Sahlgrenska Hospital, S Göteborg, Sweden Growth hormone (GH) has a profound effect on body composition through its anabolic, lipolytic and antinatriuretic actions (1). It also stimulates bone metab olism via direct effects, circulating IGF-I or locally produced IGF-I and IGF-II (2). Growth hormone stimu lates the production of 1,25-vitamin D (3), thereby increasing the availability of calcium and phosphate (4,5). Furthermore, GH receptors have been found on an osteoblast-like cell line (6). To investigate the role of GH on bone mineral content (BMC), we have conducted a follow-up investigation of adult patients with GH defi ciency. The BMC was determined in the third lumbar vetebra using dual-photon absorptiometry (DPA). Patients and methods Subjects Patients with pituitary disease diagnosed between 1956 and 199 were investigated in terms of possible GH deficiency. Each patient had been investigated as an inpatient at the Endocrine Unit at Sahlgrenska Hospital on at least one occasion between 1956 and 199. All the patients lived in the catchment area of the Endocrine Unit, which is western Sweden and includes the city of Göteborg. "Presented in part at the 9th International Congress of Endocrinology in Nice. France, September The patients received both written and verbal infor mation about the study and their written consent was obtained. The study was approved by the Ethics Com mittee of the Medical Faculty at the University of Göteborg. One hundred and forty-seven adult patients with a maximum age of 75 years were invited to participate in the study, 13 of whom accepted the invitation. One hundred and eleven of the 13 patients ( 71 males and 4 females) fulfilled the inclusion criteria in terms of the GH deficiency described below. Four patients with GH deficiency of childhood onset were excluded. One patient refused to take part in the BMC investiga tion and another patient was excluded because of a diagnosis of primary hyperparathyroidism, discovered at the follow-up. In 1 patients no reliable data on the BMC could be obtained because of obesity (six patients), skeletal deformities that made it impossible to examine vertebra L3 (in three patients) and technical problems with the equipment (one patient). Thus, of the 111 patients, 95 were left to take part in the bone mineral study. There were 59 males (mean age 54. years, range years) and 36 females (mean age years, range years). Routine replace ment therapy with cortisone acetate (mean dose mg/day), L-thyroxine (mean dose.12 mg/day), testost erone (mean dose 247 mg im every fourth week) and oestradiol (mean dose 1.8 mg/day) was given. Complete

2 Table 1. Deficiency in adrenal (A), thyroid (T) and gonadotrophin (G) function in 95 patients with growth hormone deficiency (GHD). Type of deficiency Males Females Total Isolated GHD 2 2 G T 2 2 A 1 1 G+T G+A T+A G + T + A Total Table 2. Causes of growth hormone deficiency in 95 patients. Cause No. of patients Idiopathic (no tumour) 7 Non-secreting adenoma 47 Prolactinoma 19 Craniopharyngioma 13 Meningioma 3 Dysgerminoma 1 Gangliocytoma 1 Astrocytoma grade I 1 Pylocytic optic nerve glioma 1 Undifferentiated tumour 2 Total 95 pituitary deficiency (in terms of the thyroid, adrenal and gonadal function) was found in 72 patients and partial deficiency in 21 patients (Table 1 ). Two patients suffered from isolated GH deficiency. Males with normal gonadal function (N 8) and males with gonadal deficiency on testosterone replace ment therapy (N 47) were treated as one group (group A), while males with untreated gonadal deficiency (N 4) were allocated to another group (group B). Females were considered to be gonadal deficient in the case of longstanding amenorrhoea before the age of 5 years. Females with normal gonadal function (N 6) and females on oestrogen replacement therapy (N 15) were treated as one group (group A), while female patients with a previous or current history of untreated gonadal deficiency (N15) were allocated to another group (group B). Growth hormone deficiency was diagnosed during: (i) a previously performed iv insulin or glucagon tolerance test, showing a maximum GH response of less than 5 mu/l(n37); (ii) a follow-up, showing a low concentration of plasma IGF-1 ( <.34 ku/1 for males and <.45 ku/1 for females) and a serum GH concentration of less than 1 mu/1 in all of three consecutive morning blood samples drawn at an interval of 1 h (N 58). The causes of pituitary deficiency are listed in Table 2. Pituitary deficiency secondary to pituitary adenomas was found in 66 patients; there were 19 patients with prolactinomas and 47 with non-secreting adenomas. Of these patients, 32 were operated on by the transcranial route and 21 by the trans-sphenoidal route. Hypopituitarism was attributable to extrapituitary tumours in 22 patients, 16 of whom had had transcra nial (12 subjects) or trans-sphenoidal (four subjects) operations. In a further seven patients, the cause of hypopituitarism was unknown. Of the patients who were operated on, 53 had received postoperative con ventional pituitary irradiation. In 15 patients with pituitary tumours ( 13 patients with pituitary adenomas and two patients with meningioma and craniopharyn gioma, respectively) who were not operated on, five received pituitary irradiation, three were treated with bromocriptine only, one patient had combined irradia tion and bromocriptine therapy and, finally, six patients received no active treatment for their pituitary tumours. The mean duration of disease at diagnosis was estimated from anamnestic data in the medical records. It averaged 4.1 years (-33 years) for all the patients (N 95), 4. years (-28 years) for the male patients (N59) and 4.4 years (-33 years) for the female patients (N36). The mean duration from the diagnosis of hypopitui tarism or pituitary disease to follow-up was 13.2 years for all 95 patients (range -44 years). The duration was 1.7 years (range -37 years) for the male patients (N59) and 17.1 years (range -44 years) for the female patients (N 36). Body weight, body height and body mass index Body weight (BW) was measured in the morning to the nearest.1 kg using a Stathmos balance after the patients had voided. Body height (BH) was measured barefoot and to the nearest.5 cm. Body mass index (BMÍ) was calculated from the formula: BMI body weight/body height2 (kg/m2) (7). Blood sampling and biochemical methods After one night's fasting and before the intake of the routine replacement therapy, blood samples were drawn to determine serum free thyroxine (T4), serum GH and EDTA-plasma IGF-I concentrations using the Vacutainer system (no indwelling cannulae, no intravenous heparin given during the hours preceding the sample). Another two blood samples for serum GH concentration were drawn after 2 h and 3 h, respectively. The serum free T4 concentration was determined by ligand-analogue a radioimmunoassay (Amerlex M, Amersham International pic, Amersham, Bucks, UK). The reference limits for adults are 9-23 pmol/1. The total coefficients of variation (CV) were 6.5%, 4.8% and 5.3%

3 at free T4 concentrations of 8.3, 15.9 and 46.1 pmol/i, respectively. The GH concentration (mu/1) was determined by an immunoradiometric assay according to the manufac turer's protocol (Pharmacia, Uppsala, Sweden), with the modification using sera from patients with pituitary deficiency as zero calibrator. The total CV were 3%, 4.% and 7.1% at GH concentrations of.26, 13.3 and 32.7 mu/1, respectively. The IGF-I concentration in EDTA-plasma was deter mined by a non-extraction radioimmunoassay (Nichols Institute Diagnostics, San Juan Capistrano, CA, USA). This method determines "low-affinity bound" analyte, i.e. the concentration in non-extracted plasma that is available to a known amount of antiserum to fgf-f under defined conditions. The reference limits for adults are: mean ku/1; women, ku/1 (U unit defined by the manufacturer; ku kilounits). The total CV were 8.4%, 6.7% and 5.9% at IGF-I concentrations of.26, 1.33 and 6.8 ku/1, respectively. Bone mineral content (BMC) The BMC was determined in the third lumbar vertebra (L3) using DPA and was expressed in g/cm. This unit was chosen instead of the more commonly used units of g/cm2 or g/cm3, because we have found that most mechanical properties of the lumbar vertebrae are better represented by the use of g/cm (8). Our dual-photon absorptiometer involves the use of two radionuclides, 241Am and lî7cs. The measurement takes place as a scanning procedure over the vertebra. An electronic control unit permits intermittent scanning in steps of 4 mm. The BMC at each measuring point, 33 in all, is estimated by assuming exponential attenuation of the radiation beam. The plotting of each point versus the position over the vertebra results in a bone profile curve. Points outside the bone on both sides are selected to form the endpoints of a baseline above which the bone profile curve is integrated, thereby yielding the BMC in units of g/cm (9). The accuracy and precision in the bone mineral determinations had been determined both in phantoms and in situ in the human body. The error in precision is less than 3%. The greatest error was obtained for determinations in situ, while the error in precision was even smaller when it was determined in phantoms (1). Control population As a reference population for the female patients we used a random sample of 214 women aged 35-8 years, selected from the Revenue Office Register in Göteborg (11). The participation rate ranged from 62% to 85%. The male control population consisted of 199 men from the city of Göteborg between 16and 79 years of age (12). The sample of men consisted of two subsamples; the participants aged 16-2 years (N38) were students who volunteered for the study and the men aged years (N 161 ) had been randomly selected to represent certain age strata. For each patient, a ratio of observed BMC value/ predicted BMC value (BMC %)( ± sd) was calculated. The predicted BMC value was derived from regression equa tions obtained from 413 healthy subjects (199 males, 214 females). The regression equation was y x for the male controls and y x for the female controls, where y indicates BMC and x indicates age (years). Statistics Conventional methods were used to calculate means, standard deviations and standard errors. Differences between groups were tested with Mantel's test (13), where the influence of age is eliminated. The correlation between two variables was calculated using Spearman's rank correlation test. Values are given two-tailed p values are quoted. as means and Table 3. Body weight, body height, body mass index (BMI) and concentrations of plasma insulin-like growth factor I (p-igf-i) and serum free thyroxine (s-free T4) in 95 subjects (59 males. 36 females) with growth hormone deficiency. N Body weight (kg) 59 Body height (cm) 59 BMI (kg/m2) 59 p-igf-i (ku/1) 52 s-free T4 (pmol/1) 57 Males Mean so 84.8 (±12.2) (±7.6) (±3.5) 36.27(±().I1) (±3.6) 36 Females Mean si) 67.5 (±1.8) (±5.6) 25.8(±3.8).26 (±.7) 15.4 (±5.) Table 4. Bone mineral content in 95 GH-deficient patients (59 males. 36 females) according to deficiency in adrenal (A), thyroid (T) and gonadal (G) function. Type of deficiency Isolated GHD" G T A G + T G + A T + A G + T + A Males Mean (±sn) 3.76 (±.9) 3.92 (±.25) 5.72 (-) 4.6 (±.54) 4.42 (±.33) 4.93 (±1.2) 4.17 (±.77) N Total " GHD: growth hormone deficiency. Females Mean (±sd) 3.7 (-) 3.6 (-) 3.7 (±.67) 3.18 (±.55)

4 Results The BW, BH, BMI and the concentrations of EDTAplasma IGF-I and serum free T4 in the patients are given in Table 3. The separate data on BMC in the various subgroups of pituitary deficiencies are given in Table 4. As can be seen, the majority of the patients had a complete pituitary failure in terms of adrenal, thyroid and gonadal functions. Male patients in group A had a lower (N55, p<.5) BMC than controls and the four patients in group B all had values below the mean of the controls (Fig. 1). Similarly, female patients in group A (N 21, BMC BMC H 4. A H 2. (a) 1 J I_1 1_L_l_I_l l_i_i_l_ (b) 1. J '-'-.' Fig. 1. Bone mineral content (BMC) in male patients with growth hormone deficiency compared with the mean BMC (±2 sd) in the controls (N199, 2-79 years), (a) Patients with normal gonadal function or substituted gonadal deficiency (group A) (N 55). (b) Patients with untreated gonadal deficiency (group B) (N 4). The BMC was lower (p <.5) in group A compared with the controls. The four patients (group B) in (b) are all below the mean BMC of the corresponding control age groups. BMC 7. n 6. BMC H (a) (b) Fig. 2. Bone mineral content (BMC) in female patients with growth hormone deficiency compared with the mean BMC ( ± 2 so) in the controls (N 214, years), (a) Patients with normal gonadal function ( x ) (N 6) and patients with gonadal deficiency substituted with oestrogens ( ) (N15) (group A), (b) Patients with untreated gonadal deficiency (N 1 5) (group B). The BMC was lower in the patients of both group A (p<().()5) and group B (p<.1) compared with the controls.

5 Males p<.5 r-28 n-27 Females p-.6 n-13 n-8 Age < 55 > 55 < 55 > 55 years Fig. 3. The ratio of observed/predicted BMC (BMC%) in patients below and above the age of 5 5 years with growth hormone deficiency. Individual BMC% values for male patients with normal gonadal function and substituted gonadal deficiency (N 55) (group A) and for female patients with normal gonadal function and substituted gonadal deficiency (N 21 ) (group A) are given. The horizontal lines in the boxplots represent the 1th. 25th. 5th, 75th and 9th percentiles. respectively. p<().()()5) and group B (N 15. p<.1) also had a lower BMC than controls (Fig. 2). The ratio of observed/predicted BMC (BMC%) in the patients was decreased. In male groups A and B the average BMC% was 93.6% (± 19.3) and 82.5% (±6.4), respectively, and in female groups A and B, the average BMC% was 87.2%(± 18.2) and 81.8%(± 13.3) respect ively. In male patients (group A) of 55 years or younger the average BMC% was 85.8%(±15.4) but it was 11.7% (±19.8) (p<.5) in patients above the age of 55 years. Similarly, in female group A the average BMC% in patients of 55 years or younger was 81.5% (±12.6) compared with 96.5% (±22.8) in patients above the age of 55 years (p.6) (Fig. 3). There was no relationship between the BMC% and the estimated duration of disease at diagnosis, or the duration of disease to the time of follow-up. The BMC% was correlated positively (r.51, p<.1, N55) with age at diagnosis in the male patients but no correlation was observed in the female patients. Further more, there was no relationship between BCM% and IGF-I concentrations. Discussion This study shows that adult patients, both males and females, with hypopituitarism and GH deficiency have a reduced BMC compared with controls. We are confident that we included only patients with GH deficiency. For practical reasons we could not perform insulin tolerance tests on all patients. The majority of our patients were probably severely GH deficient because our criteria for including patients were stricter than those currently used to define GH defi ciency. Furthermore, within 6 months after the followup we performed an insulin tolerance test on 26 patients who were diagnosed as GH deficient according to low IGF-I concentration; those patients all showed a maxi mum GH response below 5 mu/1. We have no reason to believe that the remaining patients differ in their GH status from the others. Moreover, all the patients were selected from a population of patients with a previously known history of pituitary disease. The BMC was determined in the third lumbar vertebra using DPA. This method measures both the trabecular bone in the vertebral body and the cortical bone in the posterior part of the spine. The mechanical strength of the vertebra is mainly determined by the amount of trabecular bone in the vertebral body. More than 9% of the variations in the compressive strength of a lumbar vertebra, for example, are determined by variations in the BMC (8). The BMC value might predict a future fracture, because recent studies have shown a connec tion between a lowered BMC and an increased risk of fractures (14, 15). Previous studies of adult subjects with GH deficiency have revealed a decreased BMC in patients with GH deficiency compared with controls (16-18). However, these patients were GH deficient since childhood. In contrast, all the patients in this study acquired their GH deficiency during adult life. Several factors influence BMC. In the present study we have not evaluated the diet regimen or alcohol con sumption, but we have no reason to believe that our patients differed in this respect from the controls. In fact, we have recently shown that their tobacco consump tion, a known risk factor for osteopenia, was lower than expected (19). Their body weight was also higher compared with normals (2), which is known to be beneficial to the bone mass. Adequate substitution therapy with corticosteroids and L-thyroxine was administered to the patients. No continuous overdosage of these medications was noted at the follow-up, but a previous minor overdosage that might have aggravated the bone loss cannot be excluded (21, 22). Gonadal deficiency is a major reason for low bone density in both males and females (23, 24). fn this study, all but four of the hypogonadal male patients were on testosterone replacement therapy. Not surprisingly, we noted that female patients on oestrogen replacement therapy had a higher BMC than female patients with untreated gonadal deficiency. Even so, these patients had a lower BMC compared with control subjects, thereby indicating that routine replacement therapy is not sufficient to optimize BMC. Previous studies on bone mineral status in GHdeficient adults of childhood onset have shown low BMC (18). This was mainly observed in the forearm. The cause of osteopenia in these patients could be deficient

6 build-up of bone mass during childhood and adoles cence. Deficient build-up of bone cannot explain the ostepenia observed in our study. Ageing is associated with decreasing GH secretion (25). Bone mass also decreases in normal subjects with increasing age. Low BMC was only observed in our younger adult patients and the ratio of observed/ predicted BMC (BMC%) did in fact increase with age. Thus, a similar BMC was observed in both patients and controls among older patients. Similarly, we observed recently in the same patients that body cell mass was reduced only in younger subjects (2). Recently performed short-term studies with recombi nant human GH (rhgh) to subjects with GH deficiency have shown promising results. Bone mineral density increased (16) with a parallel increase in the plasma levels of markers for bone turnover, such as alkaline phosphatase and osteocalcin, and an increase in hydroxyproline excretion (16, 26, 27). To summarize, we have shown that both male and female patients with adult onset GH deficiency have a lower BMC than controls. The clinical implications of this are still unknown, as we are not aware of any data regarding the incidence of fractures in these patients. Short-term studies with rhgh have shown promising results, but further studies are needed to ascertain whether long-term treatment can normalize the BMC in these patients. Acknowledgments. We thank Dr Anders Odén for statistical advice. Furthermore, we thank Professor Björn Sjögren and Dr Anders Nilsson for fruitful contributions. The study was supported by grants from the Swedish Medical Research Council (MFR No. B92-1 7X B). References 1. Cheek DB. Hill DE. Effect of growth hormone on cell and somatic growth. In: Creep RO. Astwood EB, eds. Handbook of physiology. Washington. DC: American Physiological Society. 1974:4: Brixen K, Kassem M, Eriksen EF, Nielsen HK, Flyvberg A, Mosekilde L. Growth hormone (GH) and adult bone remodeling: the potential use of GH in treatment of osteoporosis. J Ped Endocrinol 1993;6: Gray RW, GarthwaiteTL. Activation of renal 1.25-dihydroxyvitamin D3 synthesis by phosphate deprivation: evidence for a role for growth hormone. Endocrinology 1985:116: Chipman JJ. Zerwekh J, Nicar M, Marks J. Pak CYC. Effect of growth hormone administration: reciprocal changes in serum 1,25-dihydroxyvitamin D and intestinal calcium absorption. ] Clin Endocrinol Metab 198:51: Boullon R. Growth hormone and bone. Horm Res 1991 ; 36(suppl l): Barnard R. Ng KW. Martin TJ. Waters. MJ. Growth hormone (GH) receptors in clonal osteoblast-like cells mediate a mitogenic response to GH. Endocrinology 1991:128: Thomas AE, McKay DA. Cutlip MB. A nomograph method for assessing body weight. Am J Clin Nutr 1976;29: Hansson T. Roos B. Nachemson A. The bone mineral content and ultimate compressive strength of lumbar vertebrae. Spine 198;5: Roos B. Sköldborn H. Dual photon absorptiometry in lumbar vertebrae. I. Theory and method. Acta Radiol 1974;3: Roos B, Hansson T, Sköldborn. H. Dual photon absorptiometry in lumbar vertebrae: an experimental study of the base line error. Acta Radiol Oncol 198:19: Hansson T, Roos B. Age changes in the bone mineral content of the lumbar spine in normal women. Calcif Tissue Int 1986; 38: Granhed H. The changes with age in the bone mineral of the spine in normal men (Thesis). Göteborg. Sweden: University of Göte borg, 1988: Mantel M. Chi-square tests with one degree of freedom; extensions of the Mantel-Haenszel procedure. J Am Stat Assoc 1963; 58: Gärdsell P. Osteoporosis. Prediction of fragility fractures (Thesis). Malmö, Sweden: University of Malmö Melton LJ. Eddy DM, Johnston CC. Screening for osteoporosis. Ann Intern Med 199:112: Degerblad M, Elgindy N, Hall K, Sjöberg H-E. Thorén M. Potent effect of recombinant growth hormone on bone mineral density and body composition in adults with panhypopituitarism. Acta Endocrinol 1992:126: Elgindy N, Grunditz R. Thorén M. Degerblad M. Sjöberg H-E, Ringertz H. Long term follow-up of metacarpal cortical thickness and bone mineral density in panhypopituitarism. Radiol Diagn 1991:32: Kaufman J-M, Taelman P, Vermeulen A. Vandeweghe M. Bone mineral status in growth hormone-deficient males with isolated and multiple pituitary deficiencies of childhood onset. J Clin Endocrinol Metab 1992:74: Rosen T, Edén S, Larson G, Wilhelmsen L. Bengtsson B-A. Cardiovascular risk factors in adult patients with growth hormone deficiency. Acta Endocrinol 1993:129: Rosen T. Bosaeus I. Tölli J. Lindstedt G. Bengtsson B-Â. Increased body fat mass and decreased extracellular fluid volume in adults with growth hormone deficiency. Clin Endocrinol 1993:38: Hahn TJ. Corticosteroid-induced osteopenia. Arch Intern Med 1978;138: Kung AWC, Pun KK. Bone mineral density in premenopausal women receiving long-term physiological doses of levothyroxine. JAMA 1991;265: Francis RM. Peacock M, Aaron JE. Osteoporosis in hypogonadal men: role of decreased plasma dihydroxyvitamin D, calcikum malabsorption, and low bone formation. Bone 1986:7: Aitken JM. Osteoporosis and its relation to oestrogen deficiency. In: Campbell S, ed. The management of the menopause and postmenopausal years. MTP Press, 1976: Rudman D, Kutner MH. Rogers CM, Lubin MF. Fleming GA. Bain RP. Impaired growth hormone secretion in the adult population. J Clin Invest 1981:67: Bengtsson B-Â. Edén S. Lönn L, Kvist H. Stokland A. Lindstedt G et al. Treatment of adults with growth hormone deficiency with recombinant human growth hormone. J Clin Endocrinol Metab 1993:76: Whitehead H, Boreham C, McIIrath EM, Sheridan B, Kennedy L. Atkinson AB et al. Growth hormone treatment of adults with growth hormone deficiency: results of a 13-month placebo controlled cross-over study. Clin Endocrinol 1992:36: Received January 8th Accepted May 12th, 1993