Clinical Endocrinology (1991). 34, 275280 AWNIS 030006649100051 D

Inter-relations between growth hormone, insulin, insulin-like growth factor4 (IGF-I), IGF-binding protein-I (IGFBP-1) and sex hormone-binding globulin in acromegaly

J. M. P. Holly, A. M. Cotterlii, R. C. Jemmott, D. Shears, S. Al-Othman, T. Chard and J. A. H. Wass Departments of Chemical Endocrinology, Reproductive Physiology and Endocrinology, Centre for Clinical Research, St Bartholomew's Hospital, London, UK

(Received 3 August 1990; returned tor revision 27 September 7990; tinally revised 72 November 1990; accepted 26 November 1990)

Summary

Acromegaly is characterized by a hypersecretion of GH, which in turn results in an excess of IGF-I, an important mediator of its actions. IGF-I itself is intimately related to insulin both in structure and function. IGF-I circulates associated with specific binding proteins which appear to have important effects on its actlvity. We have examined the inter-relations between GH, prolactin, Insulin, IGF-I and one of the binding proteins, IGFBP-1, in 62 patients with acromegaly of varying activity. Serum IGF-I levels were closely related to the logarithm of mean GH levels ( f = 0.76; n = 62; P < 0.001) but multiple regression analysis suggested that, after accounting for the variation due to GH, Insulin accounted for some of the additional variation of IGF-I. IGF-l concentrations were Independent of prolac- tin. Fasting insulin levels were high and unrelated to mean GH levels but correlated with those of IGF-I (r=0.542; n = 57; P < 0.001). This correlation coefficient was further improved by also accounting for variations in IGFBP-1 ( f = 0.684; n = SI; P < 0.001). Even in subjects whose acro- megaly was well controlled or cured, as indicated by GH levels of <1 mull or IGF-i levels of < 2 Ulml, fasting insulin levels remained significantly elevated in some individuals. The reason for this persistent abnormality Is not clear. Fasting IGFBP-1 levels were low and unrelated to mean GH but were lnveraely related to fasting insulin levels (r = - 0.593; n = 57; P < 0.001). We propose that a cascade of evenis occurs in acromegaly. GH hypersecre- tion causes chronic elevation of IGF-i levels which In turn could lead to increased pancreatic Insulin production. This raised Insulin could then have additional effects, raising IGF-i levels further and lowering IGFBP-1; the latter could then result in accentuated IGF-activity.

Correspondence: Dr J. M. P. Holly, Department of Chemical Endocrinology. St Bartholomew's Hospital, 5 1-53 Bartholomew Close. London EClA 7BE. UK.

Acromegaly has been studied for many years as a condition characterized by an over-secretion ofgrowth hormone (GH), usually from a pituitary tumour, and of the main mediator of its somatogenic actions, IGF-I. It has however become apparent that the physiology of the GH-axis is intimately integrated with that of insulin, in part due to the structural and functional homology of IGF-I and insulin and their receptors. In addition, insulin is an important regulator of hepatic IGF-I production (Daughaday et al., 1976) and, further, IGF-I has been implicated in the regulation of pancreatic insulin output (Swenne er al., 1987). The IGFs differ from insulin not only in their more widespread sites of production but also due to the presence of specific binding proteins with which the circulating IGFs are almost entirely associated. Recently much interest has focussed on one of the low molecular weight binding proteins, insulin-like growth factor binding protein-I (IGFBP- I) . This binding protein is specific for the IGFs and does not bind insulin but its circulating levels appear to be acutely regulated by insulin (Suikkari er al., 1988; Holly er al.. 1988). The levels of IGFBP-I have previously been reported to be low in acromegaly and an inverse relation with GH levels suggested (Povoa et al., 1984). We have re-examined the concentrations of this binding protein, its relations to GH. IGF-I and insulin levels, and the inter-relation between all these factors in a group of 62 patients with acromegaly of varying activity.

We have also measured SHBG levels in these subjects since in-vitro studies have implicated insulin and prolactin in the regulation of hepatic production of this binding protein (Plymate et af., 1988).

Methods

Sixty-two acromegalic patients (34 males) aged 27-77 (mean f SE 5 1.3 & 17.7) years were studied. All subjects presented with symptoms and signs compatible with the clinical diagnosis and had elevated serum G H levels which failed to suppress to less than 4.0 mU/1 during a 50-g oral glucose tolerance test. At the time of study I 1 patients were receiving medical treatment with bromocriptine (total dose 5-60 mg taken orally two to four times daily), 15 patients with S.C. somatostatin analogue, octreotide (total dose 200- 1000 fig daily, in either two or three doses, and in one patient by continuous infusion), 23 patients had previously under- gone pituitary surgery (I9 transsphenoidal and four trans-

275

276 J. M. P. Holly et al.

frontal), 34 patients had received radiotherapy according to our normal protocol (Ciccarelli et al., 1989). Eleven patients were receiving hydrocortisone or prednisone. Thirty-one patients were not on drug therapy for the medical treatment of acromegaly when studied. Three patients studied had non- insulin-dependent diabetes associated with their acromegaly. Fasting samples were taken at 0900 h for the measurement of insulin, IGF-I, IGFBP-I, prolactin and SHBG. GH status was assessed by taking the mean of four serial GH concentra- tions measured during the day (Wass et ol., 1977). Hormone levels were compared to fasting levels measured in 45 normal adult volunteers (aged 17-48 (mean 28.5) years, 27 male).

Assays

Serum IGF-I levels were measured following acid ethanol extraction (Daughaday et al., 1982) by RIA using a well characterized rabbit antiserum developed by L. Underwood and J. J. Van Wyk (University of North Carolina, Chapel Hill, USA) and distributed by the Hormone Distribution Program of NIDDK through the National Institutes of Health, Bethesda, USA. The assay was standardized against a pool of normal serum which was defined as containing 1.0 U IGF-I/ml; this was equivalent to 159 ng/ml of a highly purified preparation of IGF-I (Morrell et al., 1986). The intra-assay coefficient of variation (CV) was 6.6% at 0.74 U/ ml and 4.8% at 1.48 U/ml; the interassay CV was 7.4% at 0.66 U/ml and 9.2% at 1.3 U/ml. In order to verify that changes in IGF-I levels in relation to changes in IGFBP-I levels were not due to IGFBP-1 interfering with the assay of IGF-I the effect of adding IGFBP-I to samples, up to 1000 &I, was investigated. IGFBP-I added directly into assay caused a dose-dependent reduction in counts presumed to be due to binding of tracer to IGFBP-I. However, whilst IGFBP-I was not efficiently excluded by the acid ethanol extraction, that which remained, when diluted in assay, was found to have no effect on the IGF-I assay with original IGFBP-I levels up to 1000 pg/l (high pathophysiological levels).

Serum GH was determined by standard RIA using MRC standard 66/2 17. The minimum detection limit of the assay was 1.0 mU/1 and the interassay CV was 8%.

Serum immunoreactive insulin was determined by a double antibody RIA. The interassay CV at 41 mU/1 was 8.2% and the intra-assay CV at 38 mU/I was 3.9% with a minimum detectable limit of 2 mU/I.

Serum levels of IGFBP-I were measured as described previously (Holly et af., 1988) using antisera and purified antigen kindly provided by Dr H. Bohn (Behringwerke, Marburg. FRG).

Serum SHBG levels were measured by a commercial

immunoradiometric assay that is based on a very specific monoclonal antibody with no cross-reactivity for albumin, corticosteroid-binding globulin, transferrin, and thyroxine- binding globulin (Farmos Diagnostica, Oulunsalo, Finland).

Statistics

The significance of the differences between groups was assessed using Student's t-tests. Multiple and linear regres- sion analyses were performed by the method of least squares analysis using Minitab (Minitab, PA, USA) and correlation coefficients calculated in order to assess associations between variables; P-values < 0.05 were considered significant. Natural logarithms of insulin, IGF-I and IGFBP-I concen- trations were generally used, unless otherwise stated, because these data were log-normally distributed; however, for ease of interpretation the untransformed data are presented in Fig. 2.

Results

In the acromegalic patients mean GH levels varied from less than 1.0 mU/I up to 289 mU/I consistent with the nature of the subject group, ranging from those either surgically cured or clinically well controlled to untreated patients with very active disease. IGF-I levels similarly varied from 0-25 U/ml up to 5.64 U/ml. Untransformed IGF-I levels correlated with morning fasting levels of GH (r=0.527, n=62; P<O.OOI); this correlation was improved when mean daytime GH levels were used (r=0.607; n=62; PcO-OOI), but the closest relation was found between IGF-I levels and the logarithm of meanGH levels(Fig. 1, r=0.760;n-62; P<O-OOI). Multiple regression analysis revealed that, after accounting for the variation in IGF-I levels related to GH, some of the additional variation was significantly related to fasting insulin levels ( F I , ~ = 24.0; P < 0405), but IGF-I levels were unrelated to those of prolactin. The variation in IGF-I levels therefore correlated best with both GH and insulin (r=0*820; n-57; P<O.OOI compared to a correlation with GH alone of r-0.73 in the 57 subjects for whom insulin levels were obtained). Both GH and insulin were positively related to IGF-I levels.

Fasting insulin levels in the acromegalic patients were generally high in comparison with the normal control subjects, but no relationship was found with either the fasting or the mean daily GH levels. However, fasting insulin levels were found to correlate with IGF-I levels ( r =0.542; n=57; P<0401). More of the variation in insulin was accounted for when variations in IGFBP-1 levels were also considered (r=0.684; n = 56; P < 0.001). The effect of adding IGFBP-I was significant ( 4 . 5 , = 19.52; P < 0.005) with insu-

Insulin inter-relations in acromegaly 277

I IL L?

7

. I I I

-4 10 50 400 GH (mU/I )

Fig. 1 Relation between IGF-I and the logarithm of mean GH concentrations in the acromegalics, regression line represented by dashed line (r = 0.760; n = 62; P < 0.00 I).

lin being negatively related to IGFBP-1 levels. Furthermore, if the group of acromegalic patients was split into those with normal IGF-I levels ( < 2.0 U/ml, upper 95% confidence limit for normal group) and those with elevated IGF-I levels (> 2.0 U/ml) and then the other variables were compared; i t was apparent that the high fasting insulin levels were still present even in the successfully controlled acromegalics as judged by a normal IGF-I level (Fig. 2). Even in acromegalic patients in whom successful control or cure of disease activity was considered only by circulating G H levels being completely undetectable, persistent elevation of fasting insulin levels was still evident in some subjects (Fig. 2). This persistent abnormality in fasting insulin was not related to body weight. In addition, in these subjects IGFBP-I levels were significantly lower than controls and IGF-I levels were normal despite the complete suppression of G H levels (Fig. 2).

In the 15 patients treated with somatostatin the relation- ship between GH, IGF-I and insulin were unaltered. In the 11 patients receiving corticosteroids the IGF-I levels were consistently low (1.3 1 f 0.3 U/ml; mean f SE) and unrelated to G H levels, although no subject had a mean G H level higher than 16.2 mU/I. In this latter group of subjects, although fasting insulin levels were high (19.3 2 5 . 9 mU/l) and IGFBP-I levels were low (23+5 pg/l), neither was significantly different from the other groups of aeromegalic pa tien ts.

Fasting IGFBP-1 levels were low in the acromegalic patients compared to the normal adults. IGFBP-I levels were unrelated to mean GH levels and only weakly inversely related to IGF-I levels ( r = -0.345; n=56; Pc0.02). How- ever, IGFBP-I levels were inversely related to fasting insulin levels ( r = -0.593; n = 56; P c 0.00 I ) and multiple regression analysis revealed that after taking account of variations in insulin, IGFBP-I levels were unrelated to IGF-I levels.

J

.- = lo : <2 2oj Controls IGF-I IGF-I G H

'2.0 c2.0 cl.0 Acromegalics

Fig. 2 Concentrations of IGF-I, IGFBP-I and insulin in W, normal adult controls (n=45) and in acromegalic patients with

IGF-I levels (IGF-I c2.0 Ujml; n=36) and with 0. undetectable GH levels (GH c 1.0 mU/I; n =9). Bars represent means and standard errors, * represents a significant difference from the normal controls. Because of a bimodal distribution, individual insulin levels are shown for subjects with completely suppressed GH.

, raised IGF-I levels (IGF-Iz2.0 U/ml; n=21), with W. normal

No relation was found between SHBG levels and those of mean GH, IGF-I, insulin or prolactin. but there was a weak positive correlation between SHBG and IGFBP- 1 levels (r=0.316; n=42; Pc0.05) .

Discussion

We have investigated the interrelations between GH- and insulin-related peptides in a large group of patients with acromegaly. Whilst the primary disorder in acromegaly is a hypersecretion of GH. many of the somatogenic actions of G H are thought to be mediated by IGF-I. IGF-I levels were significantly correlated with the logarithm of G H levels in agreement with previousstudies(Barkanefa1.. 1988; Hulting el al., 1984; Rieu rf af., 1982; Roelfsema rf al., 1987). In addition to the prime effect of GH, the elevated insulin levels were also found to be a contributing Factor to the raised IGF-

270 J. M. P. Holly et al.

I levels. Insulin has been reported to stimulate IGF-I production at the cellular level (Daughaday et al., 1976) although this has been reported to be evident only at supraphysiological insulin levels (Scott et al., 1985). A physiologically more significant action may be a permissive effect of insulin on GH-stimulated IGF-I production, either by direct regulation of G H receptors (Baxter et a[., 1980) or by an effect on post GH-receptor events (Maes e f af., 1986). Many clinical observations would also support a role for insulin in IGF-I production. In malnutrition and diabetes, IGF-I levels are low despite raised G H levels (Holly et al., 1988), whilst in obesity, insulin and IGF-I levels are high despite low G H levels. In acromegaly the G H over-secretion and the resultant hyperinsulinaemia both appear to contri- bute to the elevated IGF-I levels. It is of interest that in the cured acromegalics, with undetectable circulating G H , the IGF-I levels had returned only to normal and were not lowered as would be expected from IGF-I levels in hypo- pituitarism and GH-defficiency. This may be due to the persistent elevation of insulin levels in contrast to the low levels of insulin seen in primary GH-deficiency. Alterna- tively, these cured acromegalic patients may have been not totally GH-deficient and may have reverted to a normal GH secretion pattern with nocturnal secretory episodes missed in our daytime samples. The observation that either G H o r insulin levels may be raised in acromegalic patients with a normal IGF-I level suggests that this latter measurement may not be adequate to define cure of acromegaly.

The association between carbohydrate intolerance and acromegaly is well established (Wass et al., 1980) and we confirm previous reports of elevated fasting insulin levels (Fineberg et al., 1970; Roelfsema & Frolich, 1985). The raised insulin levels might be a consequence of GH-induced insulin resistance, possibly at the insulin receptor level (Muggeo et al., 1983; Pav et al., 1986). However, in agreement with other reports (Roelfsema & Frolich, 1985; Sonksen et al., 1967). we found no relation between the raised insulin levels and mean G H levels although a surpris- ingly good correlation with fasting G H levels has been reported by one group (Muggeo et al., 1983). We did find a significant correlation between fasting insulin and IGF-I levels and even more of the variation in insulin was accounted for when variations in IGF-I and IGFBP-I were both considered. This is consistent with IGFBP-I being an important modulator of somatomedin activity (Taylor et al., 1990) and thus IGF-I and IGFBP-I levels together give a better index of somatomedin activity. These relations might merely reflect the insulin dependence of both IGF-I and IGFBP-I but there is now considerable evidence to suggest that IGF-I and/or G H are important factors in pancreatic insulin output.

There is much experimental evidence to support an effect of G H on pancreatic insulin release (Sandler et al., 1987; Formby et al., 1988) and to suggest that this might be mediated by local IGF-I production (Swenne et al., 1987). Induced G H hypersecretion in rats produces pancreatic 8- cell hypertrophy and hyperplasia (Martin et a[., 1968). However, the most convincing evidence that IGF-I, and not GH, may be the important physiological regulator of insulin output in acromegaly is the effect of pharmacological administration of oestrogen to acromegalics. This results in improved carbohydrate tolerance and reduced insulin re- sponses to glucose (Mintz et al., 1967); this is accompanied by a fall in IGF-I levels but no change in GH levels (Clemmons et al., 1980). That GH/IGF-I has a direct effect on the pancreas rather than an indirect effect as a conse- quence of GH-induced insulin resistance is supported by a recent study of GH administration, for 1 year, to short normal children; this resulted in a significant rise in fasting and postprandial insulin and C-peptide concentrations but no change in insulin resistance determined by a euglycaemic clamp (Walker et al., 1989). An increase in pancreatic islet 8- cell mass due to chronic IGF-I elevation has been suggested to explain the abnormal insulin levels and insulin responses reported in acromegaly; such abnormalities were demon- strated in some subjects even 2 years after surgical treatment and apparent biochemical cure (Roelfsema & Frolich, 1985). Our results confirm such a persistent abnormality in insulin concentrations in some acromegalics after apparent bio- chemical cure.

A further possible reason for raised insulin levels is evident from the recent suggestion that IGF-I infusion in man leads to reduced hepatic insulin degradation (Guler et al., 1989; Keller et al., 1989). A reduced degradation of insulin due to the chronic elevation of IGF-I might be a contributing factor to the raised insulin concentrations found in acromegaly.

Fasting IGFBP-I levels were low in the acromegalic patients confirming previous observations (Povoa et al., 1984; Hall et al., 1988). An inverse relation with G H was originally suggested (Povoa et al., 1984) although further study revealed no correlation with G H concentrations (Hall et al., 1988). In a larger study we have also found no correlation with GH levels but we did obtain an inverse correlation with fasting insulin levels. This is consistent with the inverse regulation of IGFBP-1 levels by insulin which has now been described in many situations (Suikkari et al., 1988, 1989a; Holly et al., 1988. 1989). The potential significance of IGFBP-I levels arises from in-vitro studies which have shown IGFBP-I to inhibit IGF-I actions at the cellular level (Ritvos et al., 1988; Rutanen et al., 1988). This together with the marked circadian variation in circulating IGFBP-I levels (Baxter & Cowell, 1987; Cotterill e f al., 1988; Holly et al..

Insulin inter-relat ions in a c r o m e g a l y 279

1988) suggests t h a t IGFBP-I might be an impor tan t acute modula tor o f s o m a t o m e d i n activity. W e have recently shown tha t the nocturnal rise in serum IGFBP-I concentrat ions is associated with a fall a t this t ime in the somatomedin bioactivity in the serum (Taylor et al., 1990). The reduced IGFBP-I levels in acromegal ic pat ients might be a com- pounding factor for their disease activity, with t h e raised IGF-I levels being accompanied by low levels of a modula tor of its actions.

Recent evidence has suggested t h a t insulin might be a regulator of SHBG levels. SHBG product ion f rom a h u m a n hepatoma cell line is inhibited b y b o t h insulin and prolactin (Plymate er al., 1988). However , in this s tudy we found n o correlation between SHBG levels and those of insulin, prolactin, GH and IGF-I t h o u g h we did find a weak positive correlation between SHBG and IGFBP-I levels. A similar relation has been found in normal puber ty (Holly et al., 1989), obesity (Weaver er al., 1990) and polycystic ovarian disease (Suikkari er al., 1989b). This s tudy would be consistent with the suggestion t h a t there are c o m m o n regulation mechanisms for the hepat ic product ion of both these binding proteins (Holly er al., 1989).

I t therefore seems likely t h a t a cascade o f events occurs in acromegaly which aggravates the initial problem o f high GH levels. T h e primary hypersecretion o f GH causes chronic elevation o f IGF-I levels. This could then lead to increased pancreatic insulin product ion which has both a permissive effect enhancing IGF-I product ion in response to GH and also leads t o low IGFBP-I levels; this could then in turn result in reduced modula t ion of IGF activity, which would then be fur ther accentuated.

References

Barkan, A.L., Beitins, 1.Z. & Kelch, R.P. (1988) Plasma insulin-like growth factor-I/somatomedin-C in acromegaly: correlation with the degree of growth hormone secretion. Journal of Clinical Endocrinology and Metabolism. 67, 69-73.

Baxter. R.C., Bryson, J.M. & Turtle, J.R. (1980) Somatogenic receptors of rat liver: regulation by insulin. Endocrinology. 107, 1176-1181.

Baxter, R.C. & Cowell, C.T. (1987) Diurnal rhythm of growth hormone-independent binding protein for insulin-like growth factors in human plasma. Journal of Clinical Endocrinology and Metabolism, 65, 432-440.

Ciccarelli, E., Corsello. S.M., Plowman, P.N.. Jones, A.E., Touzel, R.. Rees, L.H.. Besser, G.M. & Wass, J.A.H. (1989) Long-term effects of radiotherapy for acromegaly on circulating prolactin. Acta Endocrinologica (Copenh.). 121,827-832.

Clemmons, D.R.. Underwood, L.E.. Ridgeway. E.C., Kliman. B., Kjellberg. R.N. & Van Wyk, J.J. (1980) Estradiol treatment of acromegaly: reduction of immunoreactive somatomedin-C and improvement in metabolic sea t us. American Journal of Medicine. 69. 571-575.

Cotterill, A.M..Cowell, C.T., Baxter, R.C., McNeil, D. &Silink, M. (1988) Regulation of the growth hormone-independent growth factor-binding protein in children. Journal of Clinical Endocrino- logy ond Metabolirm, 67,882-887.

Daughaday, W.H., Parker, K.A., Borowsky, S., Trivedi, B. & Kapadia, M. (1982) Measurement of somatomedin-related pep- tides in fetal, neonatal, and maternal rat serum by insulin-like growth factor (IGF) I radioimmunoassay, IGF-I1 radioreceptor assay (RRA), and multiplication-stimulating activity RRA after acid-ethanol extraction. Endocrinology, 1 10,575-58 I .

Daughaday, W.H., Phillips, L.S. & Mueller, M.C. (1976) The effects of insulin and growth hormone on the release of somatomedin by the isolated rat liver. Endocrinology, 98, 1214-1219.

Fineberg, S.E., Merimee, T.J., Rabinowitz, D. & Edgar, P.J. (1970) Insulin secretion in acromegaly. Journal of Clinical Endocrinology and Metabolism. 30,288-292.

Formby, B., Ullrich, A., Coussens. L., Walker, L. &Peterson, C.M. (1988) Growth hormone stimulates insulin gene expression in cultured human fetal pancreatic islets. Journal of Clinical Endocri- nology and Metabolism, 66, 1075-1079.

Guler, H.P., Schmid, Chr., Zapf. J. & Froesch, E.R. ( I 989) Effects of recombinant insulin-like growth factor I on insulin secretion and renal function in normal human subjects. Proceedings of the Nutional Academy of Sciences of the USA. 86, 2868-2872.

Hall, K., Lundin. G. & Povoa. G. (1988) Serum levels of the low molecular weight form of insulin-like growth factor binding protein in healthy subjects and patients with growth hormone deficiency, acromegaly and anorexia nervosa. Aria Endocrinolo- gicu. 118, 32 1-326.

Holly, J.M.P.. Biddlecome. R.A., Dunger, D.B., Edge, J.A.. Amiel, S.A.. Howell, R.. Chard, T., Rees. L.H. & Wass, J.A.H. (1988) Circadian variation of GH-independent IGF-binding protein in diabetes mellitus and its relationship to insulin. A new role for insulin? Clinical Endocrinology, 29. 667-675.

Holly, J.M.P., Smith,C.P., Dunger, D.B., Howell, R.J.S.,Chard,T., Perry, L.A., Savage, M.O., Cianfarani, S.. Rees, L.H. & Wass, J.A.H. (1989) Relationship between the fall in sex hormone binding globulin and insulin-like growth factor binding protein-I. A synchronized approach to pubertal development? Clinical Endocrinology, 31,277-284.

Hulting. A.-L.. Hall, K. & Werner, S. (1984) Somatomedin-A increments are diminished in acromegaly with concomitant hyperprolactinaemia. Aria Endocrinologica, 106, 305-3 10.

Keller, S., Schmid. Chr., Zapf. J. & Froesch. E.R. (1989) Inhibition of insulin degradation by insulin-like growth factors I and II in human hepatoma (HepG2) cells. Acra Endocrinologica, 121,279- 285.

Maes, M., Underwood, L.E. & Ketelslegers, J.-M. (1986) Low serum somatomedin-C in insulin-dependent diabetes: evidence for a postreceptor mechanism. Endocrinology, 118.377-382.

Martin, J.M., Akerblom. H.K. & Garay, G. (1968) Insulin secretion in rats with elevated levels of circulating growth hormone due to MtT-WI5 tumor. Diabetes, 17, 661-667.

Mintz, D.H.. Finster, J.L. & Josimovich. J.B. (1967) Effect of estrogen therapy on carbohydrate metabolism in acromegaly. Journal of Clinical Endocrinology. 27, I32 1 - 1327.

Morrell. D.J.. Ray, K.P., Holder, A.T.. Taylor, A.M., Blows, J.A.. Hill, D.J., Wallis. M. & Preece. M.A. (1986) Somatomedin C insulin-like growth factor 1. Simplified purification procedure and biological activities of purified growth factor Journul of Endocri- nOlOgy. 110, 151-158.

280 J . M. P. Holly et al.

Muggeo. M., Saviolakis, G.A., Businaro, V., Valerio. A., Moghetti, P. & Crepaldi, G. (1983) Insulin receptor on monocytes from patients with acromegaly and fasting hyperglycaemia. Journal 01 Clinical Endocrinology and Metabolism, 56, 733-738.

Pav, J., Hilgertova, J.. Marek, J. & Sramkova, J. (1986) Insulin binding to erythrocyte receptors in acromegalic patients in relation to the activity of acromegaly and to concomitant diabetes mellitus. Experimental and Clinical Endocrinology, 88, 76-80.

Plymate, S.R., Matej, L.A., Jones, R.E. & Friedl, K.E. (1988) Inhibition of sex hormone binding globulin production in the human hepatoma (Hep G2) cell line by insulin and prolactin. Journal of Clinical Endocrinology and Metabolism, 67,460464.

Povoa, G., Roovete, A. & Hall, K. (1984) Cross-reaction of serum somatomedin-binding protein in a radioimmunoassay developed for somatomedin-binding protein isolated from human amniotic fluid. Acta Endocrinologica, 107, 563-570.

Rieu. M., Girard, F., Bricaire, H. & Binoux, M. (1982) The importance of insulin-like growth factor (somatomedin) measure- ments in the diagnosis and surveillance of acromegaly. Journal of Clinical Endocrinology and Metabolism, 55, 147- 153.

Ritvos, 0.. Ranta.T.. Ja1kanen.J.. Suikkari,A.-M., Voutilainen, R., Bohn, H. & Rutanen, E-M. (1988) Insulin-like growth factor (IGF) binding protein from human decidua inhibits the binding and biological action of IGF-I in cultured choriocarcinoma cells. Endocrinology, 122, 2 150-2 157.

Roelfsema, F. & Frolich, M. (1985) Glucose tolerance and plasma immunoreactive insulin levels in acromegalics before and after selective transsphenoidal surgery. Clinical Endocrinology, 22,

Roelfsema, F., FrBlich, M. &van Dulken, H. (1987) Sornatomedin- C levels in treated and untreated patients with acromegaly. Clinical Endocrinology, 26, 137-144.

Rutanen, E.-M., Pekonen. F. & Makinen, T. (1988) Soluble 34K binding protein inhibits the binding of insulin-like growth factor 1 to its cell receptors in human secretory phase endometrium: evidence for autocrine/paracrine regulation of growth factor action. Journal of Clinical Endocrinology and Metabolism, 66,

Sandler. S., Andersson. A., Korsgren, O., Tollemar. J., Petersson, B., Groth, C.-G. & Hellerstrom. C. (1987) Tissue culture of human fetal pancreas: growth hormone stimulates the formation and insulin production of islet-like cell cluster. Journal ofClinical Endocrinology and Metabolism, 65, I 154- 1 158.

Scott, C.D., Martin, J.L. & Baxter, R.C. (1985) Rat hepatocyte insulin-like growth factor I and binding protein: effect of growth hormone in uitro and in uiuo. Endocrinology, 116, 1102-1 107.

53 1-537.

173- 180.

Sonksen, P.H., Greenwood, F.C., Ellis, J.P.. Lowry, C., Rutherford. A. & Nabarro, J.D.N. (1967) Changes of carbohydrate tolerance in acromegaly with progress of the disease and in response to treatment. Journal of Clinical Endocrinology and Metabolism. 21, 14 18- 1430.

Suikkari, A.-M.. Koivisto, V.A., Rutanen, E.-M., Yki-Jarvinen. H., Karonen, S.-L. & Seppall, M. (1988) Insulin regulates the serum levels of low molecular weight insulin-like growth factor-binding protein. Journal of Clinical Endocrinology and Metabolism. 66,

Suikkari, A.-M., Koivisto. V.A., Koistinen, R., Seppali, M. & Yki- Jarvinen H. (l989a) Dose-response characteristics for suppression of low molecular weight plasma insulin-like growth factor- binding protein by insulin. Journal of Clinical Endocrinology and Metabolism, 68, 135-140.

Suikkari,A.-M., Ruutiainen. K., Erkkola, R. &Sepplla. M. (1989b) Low levels of low molecular weight insulin-like growth factor- binding protein in patients with polycystic ovarian disease. Human Reproduction. 4, 136-1 39.

Swenne, I . , Hill, D.J., Strain, A.J. & Milner, R.D.G. (1987) Growth hormone regulation of somatomedin C/insulin-like growth factor I production and DNA replication in fetal rat islets in tissue culture. Diabetes, 36, 288-294.

Taylor, A.M., Dunger, D.B., Preece, M.A., Holly, J.M.P.. Smith, C.P., Wass, J.A.H.. Patel, S. & Tate. V.E. (1990) The growth hormone independent insulin-like growth Pactor-l binding protein BP-28 is associated with serum insulin-like growth factor-I inhibitory bioactivity in adolescent diabetics. Clinicul Endocrino-

Walker, J., Chaussain, J.L. & Bougntres. P.F. (1989) Growth hormone treatment ofchildren with short stature increases insulin secretion but does not impair glucose disposal. Journal oJClinical Endocrinology and Metabolism. 69, 253-258.

Wass, J.A.H.. Cudworth, A.G., Bottauo, G.F.. Woodrow, J.C. & Besser. G.M. (1980) A n assessment of glucose intolerance in acromegaly and its response to medical treatment. Clinical Endocrinology. 12, 53-59.

Wass. J.A.H., Thorner, M.O.. Morris. D.V., Rees. L.H., Stuart Mason, A., Jones, A.E. & Besser. G.M. (1977) Long-term treatment of acromegaly with bromocriptine. British Medical Journal. i, 875-878.

Weaver, J.U., Holly. J.M.P., Kopelman, P.G., Noonan, K., Gia- dom, C.G., White, N., Virdee. S. & W a s , J.A.H. (1990) Decreased sex hormone binding globulin (SHBG) and insulin-like growth factor binding protein (IBP-I) in extreme obesity. Clinical Endocrinology. 33,4 15-422.

266-272.

logy. 32, 229-239.