The effect on intermediary metabolism of open-heart surgery with deep hypothermia and circulatory arrest in infants of less than 10 kilograms body weight. A preliminary study

The effects of hypothermic open-heart surgery with circulatory arrest upon intermediary metabolism and endocrine function in small children are still poorly understood. This report presents data obtained in a preliminary study in which seven children of <10 Kg body weight were investigated during and after such procedures. Frequent blood samples were taken from one day preoperatively to seven days postoperatively for estimation of hormone concentrations (insulin, growth hormone, glucagon, cortisol), and intermediary metabolites (glucose, lactate, pyruvate, alanine, glycerol, and 3- hydroxybutyrate).

Marked hyperglycaemia (34·8 ±3·3 mmol/L) was observed during cardiopulmonary bypass probably as a result of the glucose content of the pump prime. Moderate hyperglycaemia persisted for at least seven days postoperatively. A significant increase in lactate concentrations (p < 0·02) was observed during circulatory arrest and peak concentrations of 6·77 +0.87 mmol/L were measured at the end of the operation. Very high blood lactate concentrations were observed at the end of cardiopulmonary bypass in two patients who subsequently died. Peak insulin concentrations (20·7 ± 5·2 mU/L) were lower than those (30-40 mU/L) reported in adults undergoing similar procedures. Glucagon concentrations were significantly elevated during bypass ( p<0·05) to approximately twice the levels reported in adults. Unlike the other hormones, growth hormone concentrations revealed remarkably uniform change, similar to those reported in 'high responding' adults with peak values of 124 ± 26 mU/L, observed during cardiopulmonary bypass. Cortisol concentrations showed no significant changes throughout the study.

Thus, the response of small infants to these procedures differed both qualitatively and quantitatively from that reported in adults. The results suggest that the nature of the prime fluid may be of major importance in the metabolic consequences of such surgery and that glucose and lactate may be better avoided in the pump prime.

Circulating levels of pegvisomant and endogenous growth hormone during prolonged pegvisomant therapy in patients with acromegaly

OBJECTIVE

To investigate whether pegvisomant treatment in acromegaly induces gradual elevations in endogenous serum growth hormone (GH) levels and whether serum pegvisomant levels predict the therapeutic outcome.

PATIENTS AND METHODS

Seventeen patients (6 women), mean age 46·3 years (range: 23·2-76·2), were studied. For each patient, four hospital visits were identified including 'active disease' (no treatment) and last follow-up. At each visit, 12 blood samples were drawn during 3 h including an oral glucose tolerance test (OGTT). Eight patients received a somatostatin analogue in addition to pegvisomant on the last visit.

RESULTS

Median (range) pegvisomant doses (mg/day) were 10 (10-10), 15 (10-15) and 15 (10-15) at visits 2, 3 and 4, respectively, and the mean duration of pegvisomant treatment was 17·5 ± 3·2 (SEM) months. Serum IGF-I changed significantly during the treatment period with the highest level at baseline and lowest levels at visits 3 and 4. GH levels increased in a dose-dependent manner during pegvisomant treatment and decreased at visit 4. Changes in IGF-I levels correlated negatively with changes in serum pegvisomant levels between visits. Serum pegvisomant at each visit correlated with baseline growth hormone levels, whereas no associations between serum pegvisomant and either dose, gender, age or body weight were found.

CONCLUSIONS

(1) Serum GH levels increased initially, but remained stable during prolonged pegvisomant treatment in patients with acromegaly, (2) serum pegvisomant levels predicted the reduction in serum IGF-I during treatment and (3) the interindividual variation in serum pegvisomant levels seems not predicted by either age, gender or body composition.

Sustained low-dose growth hormone therapy optimizes bioactive insulin-like growth factor-I level and may enhance CD4 T-cell number in HIV infection

High-dose recombinant human growth hormone (rhGH) (2-6 mg/day) regimes may facilitate T-cell restoration in patients infected with human immunodeficiency virus (HIV) on highly active antiretroviral therapy (HAART). However, high-dose rhGH regimens increase insulin-like growth factor-I (IGF-I) to supra-physiological levels associated with severe side effects. The present study investigated whether lower doses of rhGH may improve T-cell restoration in patients infected with HIV following an expedient response of total and bioactive (i.e., free) IGF-I. A previous 16-week pilot-study included six HIV-infected patients on stable HAART to receive rhGH 0.7 mg/day, which increased total (+117%, P < 0.01) and free (+155%, P < 0.01) IGF-I levels. The study was extended to examine whether continuous use of low-dose rhGH (0.7 mg/day until week 60; 0.4 mg/day from week 60 to week 140) would maintain expedient IGF-I levels and improve CD4 T-cell response. Total and free IGF-I increased at week 36 (+97%, P < 0.01 and +125%, P < 0.01, respectively) and week 60 (+77%, P = 0.01 and +125%, P < 0.01) compared to baseline levels (161 +/- 15 and 0.75 +/- 0.11 microg/L). CD4 T-cell number increased at week 36 (+15%, P < 0.05) and week 60 (+31%, P = 0.01) compared to baseline levels (456 +/- 55 cells/microL). Following rhGH dose reduction, total IGF-I and CD4 T-cell number remained increased at week 88 (+44%, P = 0.01 and +33%, P < 0.01) and week 140 (+46%, P = 0.07 and +36%, P = 0.02) compared to baseline levels. These data support the notion that low-dose rhGH regimens may increase expediently total and bioactive IGF-I and improve T-cell restoration in patients infected with HIV on HAART.

Growth hormone and growth hormone secretagogue effects on nitrogen balance and urea synthesis in steroid treated rats

OBJECTIVES

Growth hormone (GH) reduces the catabolic side effects of steroid treatment via effects on the amino-nitrogen metabolism. Ipamorelin is a synthetic peptide with GH releasing properties. We wished to study the metabolic effects of Ipamorelin and GH on selected hepatic measures of alpha-amino-nitrogen conversion during steroid-induced catabolism.

DESIGN

Five groups of rats were included: (1) free-fed controls (2) pair-fed controls (3) prednisolone (delcortol, 4 mg x kg(-1) x day(-1)) (4) prednisolone and GH (1 mg x kg(-1) x day(-1)) (5) prednisolone and Ipamorelin (0.5 mg x kg(-1) x day(-1)). After seven days the hepatic capacity of urea-N synthesis (CUNS) was determined in parallel with measurements of liver mRNA levels of urea cycle enzymes, whole-body N-balance, and N-contents of various organs.

RESULTS

Compared to pair-fed controls, prednisolone increased CUNS (p<0.01) as well as the expression of urea cycle genes (p<0.01), and decreased N-balance (p<0.01) as well as organ N-contents (p<0.05). Compared to prednisolone treated animals, co-administration of GH reduced CUNS by 33% (p<0.01), normalized urea cycle gene expression, improved N-balance 2.5-fold, and normalized or improved organ N-contents. In prednisolone treated rats Ipamorelin reduced CUNS by 20% (p<0.05), decreased the expression of urea cycle enzymes, neutralised N-balance, and normalized or improved organ N-contents.

CONCLUSION

Accelerated nitrogen wasting in the liver and other organs caused by prednisolone treatment was counteracted by treatment with either GH or its secretagogue Ipamorelin, though at the doses given less efficiently by the latter. This functional study of animals confirms that the GH secretagogue exerts GH related metabolic effects and may be useful in the treatment of steroid-induced catabolism.

Elevated free IGF2 levels in localized, early-stage breast cancer in women

OBJECTIVE

Epidemiological studies imply an association between circulating IGF1 and breast cancer, whereas the role of IGF2, which also acts on the IGF1 receptor, is less settled. This study investigates the association between IGF2 and breast cancer in patients with localized disease.

DESIGN

The participants were women with well-characterized, early stage, localized breast cancer (n=43) and matched healthy women (n=38), from whom fasting serum levels of IGF-related peptides were measured.

RESULTS

In patients, mean free IGF2 was increased (+57%, P<0.001), in spite of reduced total IGF2 levels (-12%, P=0.003) when compared with controls. Similar changes were seen in free IGF1 (+28%, P=0.004) and total IGF1 (-16% P=NS). Pro-IGF2 and IGF-binding protein 1 (IGFBP1) were unchanged. IGFBP2 was reduced by 22% in the patients (P=0.004). The patients showed reduced IGFBP3 protease activity and accordingly increased levels of intact IGFBP3, whereas total IGFBP3 was unchanged.

CONCLUSION

Women with localized, early-stage breast cancer show elevated circulating free IGF1 and IGF2, reduced total IGF2 and alterations in IGFBPs. The changes observed despite minimal cancer disease suggest a role for the circulating IGF system in the progression of breast cancer in women.

Effects of a 3-day fast on regional lipid and glucose metabolism in human skeletal muscle and adipose tissue

AIM

Fasting is characterized by increased whole body lipolysis and lipid oxidation, decreased glucose oxidation and insulin resistance. To identify the regional sources and underlying mechanisms, we studied 10 healthy male volunteers post-absorptively and after 72 h of fasting.

METHODS

Each study comprised a 3-h basal period and a 3-h hyperinsulinaemic euglycaemic clamp and we used a combination of leg and forearm arteriovenous techniques, upper and lower body microdialysis and glucose and palmitate tracers.

RESULTS

In the basal state, plasma levels, fluxes and oxidation rates of free fatty acids all roughly doubled after fasting. Palmitate fluxes across the forearm and leg also increased by two to threefold and interstitial leg muscle glycerol concentrations doubled. Subcutaneous femoral glycerol concentrations and blood flows were unaltered, but abdominal subcutaneous blood flow increased by 50% in the presence of unchanged glycerol concentrations, indicating stimulated abdominal lipolysis. During the clamp, we observed whole body insulin resistance and glucose uptake across the leg and forearm decreased by 60%.

CONCLUSION

Our data show that fasting induces insulin resistance in upper and lower body muscles and suggest that increased lipolysis, is primarily due to the activation of lipolysis in muscle-associated fat (in the leg) and in upper body subcutaneous fat, whereas peripheral subcutaneous fat is spared.

Concomitant, specific determination of growth hormone and pegvisomant in human serum

Growth hormone (GH) and the GH receptor blocker, pegvisomant are usually circulating in high concentration in pegvisomant treated acromegalic patients. This and the close similarity between the peptides make determination of either difficult. In the present methodological study, endogenous GH in serum is initially isolated and determined in a slightly modified commercial immunometric assay, whereafter the now GH free medium allows measurement of pegvisomant. Inter-individual steady state levels of serum pegvisomant vary remarkably in both acromegalic patients and healthy controls, while the intra-individual variations are negligible.

[Medical co-treatment af acromegaly with a somatostatin analogue and a growth hormone receptor antagonist]

We tested the effect of co-treatment of acromegaly with a somatostatin analogue (SA) and a growth hormone receptor antagonist (GHA). Eleven patients underwent: 1) conventional treatment with SA, 2) discontinued treatment, 3) 6 weeks treatment with GHA (10 mg), 4) 6 weeks treatment with GHA (15 mg), 5) 3 months combined SA and GHA. Circulating IGF-I was lowered by GHA and more so with combined treatment. Treatment with GHA increased endogenous GH levels, which was partly reversed by combined treatment. Plasma glucose levels were highest during SA treatment and lowest with GHA. Co-treatment of acromegaly with SA and GHA is a promising concept.

Growth factors, glucose and insulin kinetics after low dose growth hormone therapy in HIV-lipodystrophy

OBJECTIVES

Low-dose growth hormone (GH) administration has been suggested as a treatment for HIV-lipodystrophy.

METHODS

Postglucose GH-secretion, kinetics of insulin-like growth factors (IGFs), insulin, and glucose metabolism were examined in six male HIV-infected lipodystrophic patients (two normal-weight patients with normal glucose-tolerance (NGT), two normal-weight with impaired glucose-tolerance (IGT), and two obese patients with diabetes (DM)) during a 16 weeks open-labelled pilot-study of low-dose GH, 0.7 mg/day.

RESULTS

DM, compared to NGT and IGT, displayed an impaired rebound of GH during a 5h oral glucose-tolerance test. Near lower normal limits in all patients before GH-therapy, total and free IGF-I increased between 87 and 152% during the GH-therapy (P<0.001), approaching upper normal limits in all patients with the highest incremental percentages shown in DM. A slight and temporary reduction in insulin sensitivity was caused by a reduction in non-oxidative glucose metabolism (n=5). GH-administration reduced hepatic extraction of insulin alleviating the demand for insulin secretion (n=5). No adverse effects of GH were detected.

CONCLUSIONS

As judged from effects on circulating IGF-I, glucose metabolism, and insulin kinetics, 0.7 mg/day of GH may be expedient for treatment of HIV-infected males with lipodystrophy. Whether the patients' glucose metabolic status matters for the IGF-response to low-dose GH-therapy awaits further investigation.

Characterization of proteinuria and tubular protein uptake in a new model of oral L-lysine administration in rats

Intravenous infusion of basic amino acids is used experimentally and pharmacologically to prevent renal proximal tubular uptake of filtered proteins. Intravenously injected L-lysine is rapidly cleared from plasma and the effect on tubular protein reabsorption is transient. To obtain a more sustained effect, we developed a model of oral L-lysine administration and characterized this model by analyzing urinary protein excretion and proximal tubule uptake of filtered proteins. Rats placed in metabolic cages were treated with 20 mmol/kg/6 h of L-lysine, glycine, or water. Urines were analyzed for proteins by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, immunoblotting, and radioimmunoassay. Proximal tubule uptake of proteins and expression of apical membrane receptors were investigated by immunocytochemistry. In vitro uptake and receptor expression were studied using a yolk sac cell line. L-lysine administration produced increased urinary excretion of a large number of proteins while the effect on tubular accumulation of selected proteins was variable. L-lysine treatment induced changes in the localization of two receptors responsible for tubular endocytosis of filtered proteins. In conclusion, oral L-lysine treatment induced proteinuria, in particular albuminuria, as efficiently as previous reports on intravenous infusion. The effect on tubular protein accumulation was variable suggesting differential effects on tubular reabsorption and degradation of filtered proteins. Changes in tubular protein handling were accompanied by changes in the localization of the endocytic receptors, megalin, and cubilin. In vitro experiments supported the in vivo observations. The findings suggest that L-lysine may affect receptor trafficking in addition to possible effects on the direct binding of ligands to the receptors.

Cotreatment of acromegaly with a somatostatin analog and a growth hormone receptor antagonist

CONTEXT

Pegvisomant is a GH receptor antagonist that blocks the peripheral actions of GH in acromegaly. Pegvisomant, in contrast to somatostatin (SMS) analogs, does not suppress the activity of the GH-producing adenoma.

OBJECTIVE

We assessed the effects of cotreatment with pegvisomant and SMS in acromegaly on GH secretion, IGF-I levels, and glucose tolerance.

DESIGN, PATIENTS, AND INTERVENTIONS

Eleven patients with persistent disease despite previous therapy underwent the following fixed treatment algorithm: 1) on SMS therapy, 2) off therapy for 2 months, 3) 6-wk treatment with 10 mg/d pegvisomant, 4) 6-wk treatment with 15 mg/d pegvisomant, and 5) 3-month treatment with 15 mg pegvisomant plus SMS. Blood was sampled in the fasting state and during an oral glucose tolerance test.

RESULTS

Total serum IGF-I levels (micrograms per liter) decreased after pegvisomant, but the lowest levels were obtained with cotreatment [458 +/- 67 (SMS), 562 +/- 78 (active), 376 +/- 51 (10 mg), 269 (15 mg), 195 +/- 24 (combined) (P < 0.0001)]. Free and bioactive IGF-I changed in a similar pattern. Steady-state pegvisomant levels (micrograms per liter) were obtained, but SMS cotreatment increased pegvisomant levels by 20% (P = 0.02) [2631 +/- 616 (10 mg), 6536 +/- 1413 (15 mg), 8030 +/- 1914 (combined)]. Pegvisomant increased endogenous GH levels (micrograms per liter), which was countered by SMS cotreatment [5.1 +/- 1.3 (SMS), 8.9 +/- 2.9 (active), 14.6 +/- 4.9 (10 mg), 19.7 +/- 6.5 (15 mg), 11.8 +/- 2.8 (combined) (P < 0.01)]. Plasma glucose levels (millimoles per liter) were highest during SMS and lowest during pegvisomant 15 mg [2-h oral glucose tolerance test: 10.3 +/- 0.7 (SMS), 8.9 +/- 0.7 (active), 7.2 +/- 0.7 (10 mg), 6.5 +/- 0.5 (15 mg), 8.0 +/- 0.8 (combined) (P = 0.02)].

CONCLUSIONS

Dual blockade of the GH axis with pegvisomant and a SMS analog is feasible in acromegaly.

Testosterone and estradiol regulate free insulin-like growth factor I (IGF-I), IGF binding protein 1 (IGFBP-1), and dimeric IGF-I/IGFBP-1 concentrations

The present study tests the clinical postulate that elevated testosterone (Te) and estradiol (E2) concentrations modulate the effects of constant iv infusion of saline vs. recombinant human IGF-I on free IGF-I, IGF binding protein (IGFBP)-1, and dimeric IGF-I/IGFBP-1 concentrations in healthy aging adults. To this end, comparisons were made after administration of placebo (Pl) vs. Te in eight older men (aged 61 +/- 4 yr) and after Pl vs. E2 in eight postmenopausal women (62 +/- 3 yr). In the saline session, E2 lowered and Te increased total IGF-I; E2 specifically elevated IGFBP-1 by 1.5-fold and suppressed free IGF-I by 34%; and E2 increased binary IGF-I/IGFBP-1 by 5-fold more than Te. During IGF-I infusion, the following were found: 1) total and free IGF-I rose 1.4- to 2.0-fold (Pl) and 2.1-2.5-fold (Te) more rapidly in men than women; 2) binary IGF-I/IGFBP-1 increased 3.4-fold more rapidly in men (Te) than women (E2); and 3) end-infusion free IGF-I was 1.6-fold higher in men than women. In summary, E2, compared with Te supplementation, lowers concentrations of total and ultrafiltratably free IGF-I and elevates those of IGFBP-1 and binary IGF-I/IGFBP-1, thus putatively limiting IGF-I bioavailability. If free IGF-I mediates certain biological actions, then exogenous Te and E2 may modulate the tissue effects of total IGF-I concentrations unequally.

Residual beta-cell function more than glycemic control determines abnormalities of the insulin-like growth factor system in type 1 diabetes

The GH-IGF-I axis is disturbed in patients with type 1 diabetes. Our aim was to investigate whether abnormalities are found in patients in very good glycemic control and, if so, to estimate the role of residual beta-cell function. Patients with hemoglobin A1c (HbA1c) less than 6% (reference range, 3.6-5.4%) were selected for the study. Twenty-two men and 24 women, aged 41.3 +/- 13.8 yr (mean +/- SD), with a diabetes duration of 17.8 +/- 14.6 yr participated. Healthy controls (15 women and nine men), aged 41.3 +/- 13.0 yr, were also studied. Overnight fasting serum samples were analyzed for HbA1c, C peptide, free and total IGFs, IGF-binding proteins (IGFBPs), GH-binding protein, and IGFBP-3 proteolysis. HbA1c was 5.6 +/- 0.5% in patients and 4.4 +/- 0.3% in controls. Total IGF-I was 148 +/- 7 microg/liter in patients and 178 +/- 9 microg/liter in controls (P < 0.001). Free IGF-I, total IGF-II, IGFBP-3, and GH-binding protein were lower, whereas IGFBP-1, IGFBP-1-bound IGF-I, and IGFBP-2 were elevated compared with control values. Patients with detectable C peptide (> or =100 pmol/liter) had higher levels of total IGF-I, free IGF-I, and total IGF-II and lower levels of IGFBP-1 and IGFBP-2 than those with an undetectable C peptide level despite having identical average HbA1c. IGFBP-3 proteolysis did not differ between patients and controls. Despite very good glycemic control, patients with type 1 diabetes and no endogenous insulin production have low free and total IGF-I. Residual beta-cell function, therefore, seems more important for the disturbances in the IGF system than good metabolic control per se, suggesting that portal insulin delivery is needed to normalize the IGF system.

Low-dose growth hormone and human immunodeficiency virus-associated lipodystrophy syndrome: a pilot study

BACKGROUND

Treatment with high doses (2-6 mg day(-1)) of human growth hormone (hGH) in patients with human immunodeficiency virus (HIV)-associated lipodystrophy syndrome (HALS) has been shown to increase concentrations of total insulin-like growth-factor-I (IGF-I) more than twofold greater than the normal upper range and is accompanied by adverse effects such as joint pain and glucose intolerance.

MATERIALS AND METHODS

We performed a 16-week open-labelled prospective pilot study in six male HALS patients using a s.c. low-dose hGH, 0.7 mg day(-1), aiming to examine the impact on total and free IGF-I and fat distribution. Glucose metabolism was examined by oral glucose tolerance tests and hyperinsulinaemic euglycaemic clamps.

RESULTS

Total IGF-I increased twofold (P < 0.01) and free IGF-I increased 2.5-fold (P < 0.01) to the level of the normal upper range. HDL-cholesterol increased (P = 0.01). Patients reported improvements of lipodystrophy, which was supported by a decreased waist-to-thigh ratio (P = 0.01), and waist-to-hip ratio (P = 0.06). Ratio of peripheral to trunk soft tissue mass increased (P = 0.01, measured by dual-energy X-ray absorptiometry scans) and a trend towards reduction in percentage of trunk fat was suggested (P = 0.12). Total fat mass, exercise capacity, glucose tolerance, glucose disposal rate and immune status, respectively, did not change (all P > 0.5). The patients did not complain of arthralgia or other known GH-related side-effects.

CONCLUSIONS

Sixteen weeks' treatment of lipodystrophic HIV-infected patients with hGH, 0.7 mg day(-1), increased total and free IGF-I twofold and appeared safe and tolerable. The potential of low-dose hGH in the treatment of HIV-lipodystrophy awaits examination by placebo-controlled, randomized trials.

Gene expression of the GH receptor in subcutaneous and intraabdominal fat in healthy females: relationship to GH-binding protein

OBJECTIVE

Circulating GH-binding protein (GHBP) is produced by proteolytical cleavage of the extracellular part of the GH receptor (GHR) and is positively correlated to the amount of body fat. To test the hypothesis that adipose tissue may contribute to the production of circulating GHBP, we compared gene expression of two GHR isoforms in adipose tissue with serum GHBP concentrations in healthy females.

DESIGN

Twenty-two healthy females undergoing surgery for benign gynecological conditions were included in the study.

METHODS

During surgery, s.c. and intraabdominal fat biopsy samples were taken. Gene expression of the full-length GHR and a truncated GHR (GHRtr) was assessed by RT-PCR relative to the expression of beta-actin.

RESULTS

The full-length GHR was expressed to a much higher level than GHRtr in both tissues. The levels of both GHR and GHRtr mRNA were similar in intraabdominal and s.c. adipose tissues. Surprisingly, concentrations of circulating GHBP were negatively correlated to the levels of mRNA transcripts of both the full-length GHR and GHRtr in intraabdominal fat. Whole body resistance (as a measure of lean body mass) was positively correlated to mRNA levels for both GHRs in intraabdominal fat.

CONCLUSIONS

(i) The full-length GHR is expressed to a much higher level than GHRtr in s.c. as well as visceral abdominal fat; (ii) the observation of a significant correlation between GHR expression and GHBP levels further emphasizes the link between adipose tissue and GHBP; (iii) it remains, however, to be demonstrated whether circulating GHBP is produced to a significant degree by adipose tissue.

Acute hyperinsulinemia restrains endotoxin-induced systemic inflammatory response: an experimental study in a porcine model

BACKGROUND

Intensive insulin therapy in critically ill patients reduces morbidity and mortality. The current study elucidates whether acute hyperinsulinemia per se could attenuate the systemic cytokine response and improve neutrophil function during endotoxin (lipopolysaccharide)-induced systemic inflammation in a porcine model.

METHODS

Pigs were anesthetized, mechanically ventilated, randomized into four groups, and followed for 570 min: group 1 (anesthesia solely, n = 10), group 2 (hyperinsulinemic euglycemic clamp [HEC], n = 9), group 3 (lipopolysaccharide, n = 10), group 4 (lipopolysaccharide-HEC, n = 9). Groups 3 and 4 were given a 180-min infusion of lipopolysaccharide (total, 10 microg/kg). Groups 2 and 4 were clamped (p-glucose: 5 mM/l, insulin 0.6 mU.kg(-1).min(-1)) throughout the study period. Changes in pulmonary and hemodynamic function, circulating cytokines, free fatty acids, glucagon, and neutrophil chemotaxis were monitored.

RESULTS

Tumor necrosis factor alpha and interleukin 6 were significantly reduced in the lipopolysaccharide-HEC group compared with the lipopolysaccharide group (both P = 0.04). In the lipopolysaccharide-HEC group, the glucagon response was diminished compared with the lipopolysaccharide group (P < 0.05). Serum free fatty acid concentrations were decreased in animals exposed to HEC. Animals receiving lipopolysaccharide showed an increase in pulmonary pressure (P < 0.001), but otherwise, there were no major changes in pulmonary or hemodynamic function. Neutrophil function was impaired after lipopolysaccharide administration.

CONCLUSION

Hyperinsulinemia concomitant with normoglycemia reduces plasma concentrations of tumor necrosis factor alpha and the catabolic hormone glucagon in lipopolysaccharide-induced systemic inflammation in pigs. The finding strongly supports the role of insulin as an antiinflammatory hormone. Whether the effect to some extent operates via a reduced free fatty acid concentration is unsettled.

Estradiol supplementation modulates growth hormone (GH) secretory-burst waveform and recombinant human insulin-like growth factor-I-enforced suppression of endogenously driven GH release in postmenopausal women

The present study tests the mechanistic postulate that estrogen confers resistance to negative feedback by systemic IGF-I. To this end, eight postmenopausal women received a constant iv infusion of recombinant human (rh)IGF-I (10 micro g/kg.h x 6 h) and saline in randomized order on the 10th day of supplementation with oral estradiol (E(2)) and placebo (Pl). GH secretion was quantitated by 10-min blood sampling, immunochemiluminometry assay, and deconvolution analysis. Administration of E(2) compared with Pl followed by saline infusion: 1) stimulated pulsatile GH secretion ( micro g/liter.6 h), viz., 12 +/- 3.3 (Pl) and 18 +/- 4.6 (E(2)) (mean +/- SEM, paired comparison, P < 0.05); 2) halved the time latency (min) to achieve peak GH secretion after GHRH injection, 24 +/- 2.2 (Pl) and 12 +/- 2.1 (E(2)) (P < 0.01); and 3) did not alter the mass of GH secreted ( micro g/liter) in response to a maximally effective dose of GHRH, 30 +/- 7.2 (Pl) and 37 +/- 11 (E(2)). Exposure to E(2) compared with Pl followed by rhIGF-I infusion: 1) accelerated the rate of decline of GH concentrations by 3.3-fold, viz., absolute slope ( micro g/liter.1000 min), 3.8 (range, 2.5-5.0) (Pl) and 12 (range, 10-14) (E(2)) (P < 0.001); 2) augmented the algebraic decrement in GH concentrations ( micro g/liter) enforced by rhIGF-I infusion, 0.73 +/- 0.21 (Pl) and 1.6 +/- 0.25 (E(2)) (P < 0.01); 3) halved the time delay (min) to peak GHRH-induced GH secretion, 20 +/- 1.2 (Pl) vs. 10 +/- 1.3 (E(2)) min (P < 0.01). In contradistinction, E(2) did not alter: 1) the capability of rhIGF-I to suppress GHRH-stimulated GH secretory burst mass significantly, viz., by 50 +/- 8% (Pl) and 52 +/- 14% (E(2)) (P < 0.05 each vs. saline); 2) the hourly rate of rise of infused (total) IGF-I concentrations; and 3) total and ultrafiltratably free IGF-I concentrations ( micro g/liter) attained at the end of the two rhIGF-I infusions. In summary, compared with Pl, E(2) supplementation in postmenopausal women: 1) amplifies endogenously driven GH secretory-burst mass; 2) initiates rapid onset of GHRH-stimulated GH release; and 3) potentiates IGF-I-dependent suppression of unstimulated GH concentrations. Based upon companion modeling data, we postulate that E(2) facilitates the upstroke and IGF-I enforces the downstroke of high-amplitude GH secretory bursts in estrogen-replete individuals.

Placental growth hormone during pregnancy in a growth hormone deficient woman with type 1 diabetes compared to a matching diabetic control group

In pregnancy, pituitary growth hormone (GH) is gradually replaced by placental growth hormone (hPGH). GH deficient pregnant women may take advantage of GH substitution during pregnancy, but this issue still remains unresolved. Also, in pregnancy diabetes may cause macrosomia. The combination of GH deficiency, GH substitution therapy and type 1 diabetes mellitus may influence pregnancy in unforeseen ways. We present a case of pregnancy in a GH deficient woman with type 1 diabetes who continued on GH replacement until week 21. In gestational week 37 a thin and mildly small-for gestational-age (length 55 cm, +3 SD, 99th centile and weight 2445 g., -1.4 SD, 10th centile) but otherwise healthy boy was born. The patient had levels of serum hPGH at the lower end of the range of values found in a matching group of diabetic women. Serum IGF-I levels were at the upper end of the range of values in the control group. A positive correlation between serum hPGH and IGF-I values was seen in the control group when using the area-under-the-curve (r=0.84; p<0.001). The patient's child had lower birth weight and ponderal index, but was otherwise healthy. Serum IGF-I, but not hPGH, correlated to the absolute birth weight (r=0.63; p=0.015) and the birth weight z-score (r=0.55; p=0.039) in the control group. Serum hPGH and IGF-I declined rapidly after delivery. In conclusion, hPGH correlated to IGF-I in type 1 diabetes mellitus (DM), and IGF-I values correlated to the birth weight. Both type 1 diabetes mellitus and GH deficiency (with GH substitution therapy) may influence fetal growth, and in combination, the net influence may be difficult to predict.

Assessment of ghrelin

Four years ago Kojima and coworkers discovered ghrelin. Within this short lifespan ghrelin has become one of the "hottest topics" in metabolic research, and today more than 300 papers have emerged (PubMed search). The huge interest in ghrelin is partly due to its involvement in appetite regulation. Over-nutrition, obesity and type 2 diabetes are major burdens of health services in all Western countries, and the discovery of ghrelin opens for the development of antagonists that may make it possible to control appetite and food intake. At the other end of the nutritional scale, ghrelin agonists may be used in cachexia in e.g. anorexia nervosa and cancer. Thus, the potential clinical value of ghrelin research appears to be enormous. At the time of writing several in-house as well as commercial ghrelin assays have been developed. However, we still need to come to a consensus on measurement of circulating ghrelin levels. Up till now, blood ghrelin has been estimated by use of serum as well as various types of plasma, with or without extraction prior to assay. This may affect both results and conclusions. In the present paper we shall review the current literature on ghrelin, with special focus on measurements in human blood specimens.

The effect of growth hormone on the insulin-like growth factor system during fasting

The present study investigates the possible stimulatory effect of endogenous GH on IGF and IGF-binding protein (IGFBP) levels during fasting. Eight normal subjects were examined on four occasions: 1) in the basal postabsorptive state; 2) after 40 h of fasting; 3) after 40 h of fasting with somatostatin suppression of GH; and 4) after 40 h of fasting with suppression of GH and exogenous GH replacement. The two somatostatin experiments were identical in terms of hormone replacement (except for GH). Short-term fasting led to a 50% reduction in free IGF-I. The reduction in free IGF-I was paralleled by an increase in IGFBP-1, an increase in the complex formation of IGFBP-1 and IGF-I, and a modest reduction in IGFBP-3 proteolysis. GH deprivation during fasting led to a 35% reduction in total IGF-I and a 70% reduction in free IGF-I. GH replacement increased free and total IGF-I to levels similar to those observed during plain fasting and decreased IGFBP-1, however, without affecting IGFBP-1-bound IGF-I. Finally, IGFBP-3 proteolysis was slightly increased by GH replacement. In conclusion, the major new finding of the present study is that the GH hypersecretion seen during short-term fasting is not merely secondary to a reduction in IGF bioactivity.

The circulating IGF system and its relationship with 24-h glucose regulation and insulin sensitivity in healthy subjects

OBJECTIVE AND DESIGN

It has been suggested that circulating free IGF-I participates in glucose homeostasis and that IGFBP-1 reflects changes in insulin sensitivity. To study this further, we examined 10 healthy, nonobese subjects under standardized conditions for 24 h with and without an intravenous infusion of glucose, the latter in order to augment insulin sensitivity. Serum was collected every 2 h for analysis of free and total IGFs, IGFBP-1, - 2 and - 3 and the acid labile subunit (ALS). Insulin sensitivity was estimated at the end of each 24-h study period by use of the hyperinsulinaemic euglycaemic clamp technique.

RESULTS

Glucose infusion resulted in mild hyperglycaemia (P < 0.0001), a reduction in IGFBP-1 by approximately 40% (P < 0.0003), and increased insulin and C-peptide levels (P < 0.0001). Glucose infusion also increased insulin sensitivity (P < 0.003). However, despite the reduction in IGFBP-1, glucose infusion did not increase free IGF-I over the control level, and free IGF-II was slightly reduced (P < 0.02). Irrespective of glucose infusion, free IGF-I and -II remained stable during daytime (i.e. they were unresponsive to meal-related changes in plasma glucose), but both free fractions decreased during the night, reaching nadir at 04.00 h. None of the other members of the IGF system showed any relationship with plasma glucose levels. Finally, we failed to observe any relationship between changes in insulin sensitivity and the circulating IGF system.

CONCLUSION

We found no evidence that the circulating IGF system is involved in meal-related blood glucose regulation or that it reflects short-term changes in insulin sensitivity in healthy, nonobese subjects. However, we cannot preclude that the observed changes in circulating IGFBP-1 may affect the glucose-lowering effect of IGF-I and -II at the local tissue level.

Low IGF-I and elevated testosterone during puberty in subjects with type 1 diabetes developing microalbuminuria in comparison to normoalbuminuric control subjects: the Oxford Regional Prospective Study

OBJECTIVE

To describe longitudinal variations in pubertal hormonal variables in subjects with and without microalbuminuria (MA).

RESEARCH DESIGN AND METHODS

Blood samples collected annually from subjects recruited at diagnosis of type 1 diabetes and followed prospectively through puberty (median follow-up 9.3 years, range 4.7-12.8) were analyzed for total and free IGF-I, IGF binding protein-1, testosterone, sex hormone-binding globulin, and HbA(1c). A total of 55 subjects who developed MA (MA(+) group) were compared with 55 age-, sex-, and duration-matched control subjects who did not develop MA (MA(-) group).

RESULTS

For female subjects, total IGF-I (MA(+) 1.2 mU/l vs. MA(-) 1.4 mU/l, P = 0.03) and free IGF-I levels (MA(+) 1,767 ng/l vs. MA(-) 2010 ng/l, P = 0.002) were lower, whereas the free androgen index (MA(+) 2.4 vs. MA(-) 2.0, P = 0.03) was higher in those with MA. These changes were less pronounced in male subjects. For both sexes, in a Cox model after adjusting for puberty, the presence of MA was associated with lower free IGF-I levels, higher testosterone standard deviation score (SDS), and poor glycemic control. We found that 22 of 55 case subjects (40%) developed persistent MA, whereas 33 (60%) had transient MA. In the persistent MA group compared with the transient and control groups, total IGF-I levels were lower (1.1 vs. 1.3 vs. 1.4 mU/l, P = 0.002) as were free IGF-I levels (1,370.9 vs. 1,907.3 vs. 1,886.7 ng/l, P < 0.001), whereas HbA(1c) levels were higher (11.8 vs. 10.3 vs. 9.9%, P < 0.001).

CONCLUSIONS

Poor glycemic control and differences in IGF-I levels and androgens, particularly in female subjects, accompany development of MA at puberty. These differences may in part account for the sexual dimorphism in MA risk during puberty and could relate to disease progression.

The relationship between the circulating IGF system and the presence of retinopathy in Type 1 diabetic patients

BACKGROUND

Patients with proliferative diabetic retinopathy (PDR) have increased vitreous levels of insulin-like growth factor (IGF)-I, IGF-II and IGF binding proteins (IGFBPs). This accumulation is probably caused by increased leakiness of the blood-retina barrier and influx of circulating IGFs and IGFBPs. To date, interest has focused on the role of circulating total IGF-I in the development of PDR, and there are only sparse data on circulating levels of free IGF-I and IGFBPs.

METHODS

We compared fasting serum samples from matched groups of Type 1 diabetic patients with no retinopathy (n = 29), non-PDR (n = 13) and PDR (n = 16). We also included matched controls (n = 26). Serum was analysed for free and total IGF-I and -II, free plus dissociable IGF-I, IGFBP-1, -2 and -3, IGFBP-1-bound IGF-I as well as IGFBP-3 proteolysis.

RESULTS

When compared with controls, diabetic patients (n = 58) showed reduced (P < 0.0005) levels of free and total IGFs, free plus dissociable IGF-I and IGFBP-3, whereas levels of IGFBP-1, IGFBP-1-bound IGF-I and IGFBP-2 were elevated. IGFBP-3 proteolysis remained unaltered. When comparing diabetic patients with different degrees of retinopathy, IGFBP-2 and IGFBP-1-bound IGF-I were the only parameters that differed significantly. Patients with retinopathy (non-PDR as well as PDR) had elevated IGFBP-2 (P < 0.03) and reduced IGFBP-1-bound IGF-I (P < 0.03), when compared with patients without retinopathy. Noteworthy, both parameters correlated (0.11 < r2 < 0.14, P < 0.02) with the severity of retinopathy.

CONCLUSION

Our study gives no evidence of a direct role of circulating free IGF-I in the development of PDR, whereas IGFBP-2 and IGFBP-1-bound IGF-I showed a relationship with the degree of retinopathy. However, further investigations are needed in order to clarify the basis and clinical relevance of this finding.

The effect of the somatostatin analogue lanreotide on the prevention of urethral strictures in a rabbit model

We evaluated the effect of the somatostatin analogue lanreotide on the development of surgically induced experimental strictures in the anterior urethra of the male rabbit. A total of 74 male rabbits were randomly allocated into four groups. Lanreotide was administered to the rabbits in groups 2 and 4 from day 0 to 14. To create a stricture, a resection was made in the urethra of the rabbits in groups 3 and 4 on day 2. On day 30, all rabbits were examined with urethrography, impedance planimetry and either histology or for collagen content. Urethrography and impedance planimetry demonstrated a urethral stricture in all operated animals. No difference was found between the two stricture groups, regardless of lanreotide administration, with respect to luminal cross-sectional area (CSA), circumferential tension-strain relation, histology or collagen content. The CSA of the urethra of the normal controls treated with lanreotide was smaller than the CSA of the normal controls not treated with lanreotide, however, no difference was found in histology or collagen content. Lanreotide had no measurable effect on the development of a surgically induced stricture in the male rabbit anterior urethra, however, lanreotide seems to exert an inhibitory effect on the normal growth of the urethra.

Physical capacity influences the response of insulin-like growth factor and its binding proteins to training

The influence of initial training status on the response of circulating insulin-like growth factor (IGF) and its binding proteins (IGFBP) to prolonged physical training was studied in young men. It was hypothesized that highly standardized training would result in more extensive changes in the circulating IGF system in untrained subjects because of lower fitness level. Seven untrained (UT) and 12 well-trained (WT) individuals performed 11 wk of intense physical training (2-4 h daily). Fasting serum samples were analyzed for total and free IGF-I and -II, for IGFBP-1 to -4, as well as for IGFBP-3 proteolysis. Eleven weeks of physical training resulted in decreased levels of total IGF-I, free IGF-I, and IGFBP-4 in both the UT and WT groups. In the UT group, IGFBP-2 increased, IGFBP-3 decreased [from 4,255 +/- 410 (baseline) to 3,896 +/- 465 (SD) microg/l (week 4); P < 0.05], and IGFBP-3 proteolysis increased [from 28 +/- 8% (baseline) to 37 +/- 7% (week 4) and 39 +/- 12% (week 11); P < 0.05], whereas no significant changes were found in the WT group. In conclusion, intense physical training results in a marked influence on the IGF system and its binding proteins with generally more extensive changes seen in the untrained individuals. Also, prolonged physical training resulted in increased IGFBP-3 proteolysis in previously untrained individuals only, indicating that intense physical training affects trained and untrained individuals differently.

Acute growth hormone administration causes exaggerated increases in plasma lactate and glycerol during moderate to high intensity bicycling in trained young men

We studied the acute effects of a single, sc GH dose on exercise performance and metabolism during bicycling. Seven highly trained men [age, 26 +/- 1 yr (mean +/- SEM); weight, 77 +/- 3 kg; maximal oxygen uptake, 65 +/- 1 ml O(2).min(-1).kg(-1)] performed 90 min of bicycling 4 h after receiving 7.5 IU (2.5 mg) GH or placebo in a randomized, double-blinded, cross-over design trial. A standardized pre-exercise meal was given 2 h before exercise. Blood was sampled at rest and during exercise and analyzed for GH, IGF-I, glucose, lactate, insulin, glycerol, and nonesterified fatty acids (NEFA). In the placebo trial, all subjects completed the exercise protocol without any difficulties. In contrast, two subjects were not able to complete the exercise protocol in the GH trial, and one subject barely managed to complete the protocol. In addition, GH administration resulted in exaggerated increases in plasma lactate concentrations during exercise (P < 0.0001). The combined lipolytic effect of GH and exercise, evidenced by increased plasma glycerol and serum NEFA concentrations, was 3-fold greater than the effect of exercise alone (P < 0.0001), but this increased substrate availability did not result in increased whole body fat oxidation (indirect calorimetry). Plasma glucose was, on average, 9% higher during exercise after GH administration compared with placebo (P < 0.0001). We conclude that a single, relevant GH dose causes exaggerated increases in plasma lactate and glycerol as well as serum NEFA during 90 min of subsequent bicycling at moderate to high intensity. The exaggerated increase in plasma lactate may be associated with substantially decreased exercise performance.

Ghrelin immunoreactivity in human plasma is suppressed by somatostatin

OBJECTIVE

Ghrelin was recently identified as a specific endogenous ligand for the growth hormone secretagogue receptor (GHS-R). This new hormone was isolated from rat and human stomach and was reported to circulate in human plasma, but the regulation and physiological significance of ghrelin in humans have not been clarified. The present study was undertaken to test the following hypotheses: (1) prolonged fasting, which is known to stimulate GH secretion, is associated with changes in ghrelin immunoreactivity; (2) somatostatin in the systemic circulation regulates ghrelin secretion; and (3) GH affects ghrelin levels.

DESIGN AND PATIENTS

The study population included normal subjects investigated on three occasions (fasting alone, fasting and somatostatin infusion +/- GH); GH-deficient adults investigated after 12 and 36 h of fasting +/- GH, as well as patients with active acromegaly before and after somatostatin analogue treatment.

RESULTS

Somatostatin infusion lowered ghrelin levels 70-80% (P < 0.0001), whereas continued fasting +/- GH did not significantly affect ghrelin levels. In active acromegaly, suppression of plasma ghrelin levels was recorded after a single subcutaneous octreotide injection as well as during prolonged administration of slow-release octreotide.

CONCLUSIONS

(1) Amplification of GH release during prolonged fasting is not caused by an increase in ghrelin immunoreactivity, (2) systemic somatostatin suppresses plasma ghrelin levels independently of GH status, and (3) the feasibility of measuring ghrelin in the circulation provides an opportunity for studying the interaction between hormones and nutrition.

Development and clinical evaluation of a novel immunoassay for the binary complex of IGF-I and IGF-binding protein-1 in human serum

Correlation studies have suggested that IGF-binding protein (IGFBP)-1 is a dynamic regulator of free IGF-I. To further study this, we developed a monoclonal immunofluorometric assay specific for the binary complex of IGF-I and IGFBP-1 in human serum. An IGFBP-1 antibody, which recognizes all phospho-forms of IGFBP-1, was used for coating. An europium-labeled IGF-I antibody served as tracer. Assay incubation was performed at conditions approaching those in vivo (i.e. pH 7.4, 37 C). The assay was highly specific: no signal was obtained unless both IGF-I and IGFBP-1 were present and neither IGFBP-2, -3, -4, nor IGF-II caused any cross-reaction. The linear standard curve covered 3 orders of magnitude, and within and in-between assay coefficients of variation were less than 5 and 15%, respectively. To study the dynamic relationship between free IGF-I and binary complex formation, seven healthy subjects were fasted for 72 h. Samples were collected every 3 h. During fasting, free IGF-I was reduced by two thirds (P < 0.0001). IGFBP-1 and the binary complex increased in parallel (P < 0.0001), and levels correlated positively in all subjects (0.89 < or = r < or = 0.98; P < 0.0001). Free IGF-I correlated inversely with IGFBP-1 (-0.81 < or = r < or = -0.48; 0.0001 < or = P < or = 0.05) and the binary complex (-0.79 < or = r < or = -0.41; 0.0001 < or = P < or = 0.05). To study overnight fasting levels, we compared healthy controls and patients with type 1 diabetes and chronic renal failure (n = 10), because these patients show profound alterations in their IGF-system. In both groups, the binary complex was increased about 2.5-fold (P < 0.0001), whereas IGFBP-1 was increased by 5- to 6-fold (P < 0.0001). Accordingly, free IGF-I was severely reduced (P < 0.0001). In conclusion, the assay enables us to study the role of IGFBP-1 as a dynamic regulator of free IGF-I. Our results clearly show that IGFBP-1 and free IGF-I are tightly associated peptides. Furthermore, it has now become possible to compare levels of IGF-I carried within the binary complex IGFBP-1:IGF-I in different (patho-) physiological conditions.

The effect of growth hormone on insulin-like growth factor I and bone metabolism in distraction osteogenesis

Limb lengthening in the left tibia of 30 mature female Yucatan micropigs was performed using distraction osteogenesis. A treatment group of 15 animals received recombinant porcine growth hormone (r-pGH) (100 microg/kg/day) while the others served as controls. Serial serum measurements of total insulin-like growth factor I (IGF-I), free IGF-I, IGF binding proteins -1, -2, -3 and -4 (IGFBP-1 to -4) were performed. Bone-specific alkaline phosphatase (bone-ALP) and the serum carboxyl-terminal telopeptide of type I collagen (ICTP) were measured as bone turnover markers. The GH-treated animals showed a significant increase in total IGF-I, free IGF-I and IGFBP-3 after surgery (P<0.001). Similarly, the treated animals showed a significantly higher level of bone-ALP (P<0.001) throughout the experiment compared to the controls. There was a significant correlation between bone-ALP and total IGF-I (r=0.76) in the GH-treated group and an even higher correlation for free IGF-I (r=0.90). There was no difference in the ICTP serum levels between the two groups. These data indicate that the application of species-specific growth hormone results in a stimulation of bone formation in distraction osteogenesis which may be mediated by IGF-I. The stronger correlation between free IGF-I and bone-ALP indicates that the anabolic effect of IGF-I may be regulated through the IGFBPs by binding and inactivating IGF-I.

Impaired fibrinolysis determines the outcome of percutaneus transluminal coronary angioplasty (PTCA)

BACKGROUND

Coronary artery stenosis lesions dilated by percutaneus transluminal coronary angioplasty (PTCA) show a disappointingly frequent recurrence of stenosis. We have investigated the possible role of fibrinolysis and various platelet-release factors - specifically in the locality of the affected vessel - by following 19 patients for 6 months after PTCA.

METHODS

PTCA was performed on 19 patients with a significant primary coronary stenosis, proven by quantitative CAAS analysis. Blood for measurement of local fibrinolysis and platelet activity was drawn from the aortic root and the coronary sinus, at three times: just before PTCA, 10 min after it, and 6 months later.

RESULTS

The incidence of restenosis at the 6 months follow-up was 37%. PTCA almost doubled the platelet-derived growth factor level (PDGF) in coronary sinus blood in all patients. The seven restenosis patients had a substantially higher tissue plasminogen activator inhibitor antigen (PAI-1ag) level in the aortic root before PTCA than the 12 who remained stenosis-free (mean 62.4 +/- 31.6 ng mL -1 compared with 33.1 + 25.3; P < 0.04) and a lower tissue plasminogen activator activity (t-PAac) level (mean 0.32 +/- 0.19 IU mL-1 compared with 0.68 +/- 0.34; P < 0.03). This was corroborated by the levels of tissue plasminogen activator inhibitor activity (PAI-1ac). At reassessment after 6 months, the restenosis patients had developed, in coronary sinus blood, a large rise of PAI-1ac (7.7 +/- 4.8 IU mL-1 rising to 15.7 +/- 13.9, P < 0.04) and a large rise of of PAI-1ag (48.8 +/- 31.3 ng mL-1 vs. 72.4 +/- 47.2; P < 0.03). But no such increase occurred in the patients who remained stenosis-free. Conclusion Our results indicate that the minor balloon injury, which is inseparable from PCTA, stimulates the local release of PDGF. We suggest that, in those patients whose fibrinolytic activity is inherently low, this rise of PDGF could be a major causative factor in restenosis. We also discuss the possibility that the preoperative level of PAI-1ac could provide a limited but useful prediction of the outcome of PTCA.

Determination of free insulin-like growth factor-I in human serum: comparison of ultrafiltration and direct immunoradiometric assay

Two fundamentally different methods are currently used for the determination of free insulin-like growth factor-I (IGF-I): ultrafiltration by centrifugation (UF) and direct immunoradiometric assay (IRMA). The aim was to evaluate a commercial IRMA (DSL, Webster, TX, USA) and to compare it with UF. In the IRMA it is recommended that samples be incubated for 2 h at 5;C. When comparing samples (n = 8) incubated for 1 and 2 h, levels increased by 27 +/- 5% (P< 0.0001). When incubating samples at 22;C instead of 5;C, levels increased by 192 +/- 32% (P< 0.0001). Addition of IGF-binding protein-1 (IGFBP-1) to normal sera (n = 6) dose-dependently decreased ultrafiltered free IGF-I only (P< 0.0007). Similarly, UF was more sensitive than IRMA to addition of IGFBP-2 (P< 0.05). In healthy subjects (n = 35) IRMA yielded 20% higher levels than UF (1.09 +/- 0.09 vs 0.91 +/- 0.12 microg/L; P< 0.0001). IRMA and UF yielded similar results in healthy subjects treated with IGF-I (n = 5) or growth hormone (n = 7) and in acromegalic patients (n = 6) before and after somatostatin analogue treatment. However, marked differences were observed in conditions with elevated IGFBP-1 and -2. In type-1 diabetics (n = 23) ultrafiltered free IGF-I was more reduced than IRMA free IGF-I (38 +/- 9 vs 76 +/- 7% of matched controls (n = 13); P< 0.0001). In patients with chronic renal failure (n = 25), IRMA free IGF-I was identical to control levels (n = 13), whereas ultrafiltered free IGF-I was decreased by 51 +/- 7% (P< 0.0001). Similarly, women with anorexia nervosa (n = 9) studied before and after weight gain showed significant changes in ultrafiltered free IGF-I only (P< 0.03). In conclusion, IRMA was not very robust with respect to variations in sample incubation and this may bias results. IRMA generally yielded higher levels than UF, in accordance with the knowledge that IRMA measures free plus readily dissociable IGF-I. IRMA was less affected than UF by added IGFBP-1 and -2, and reductions in free IGF-I were better revealed by UF than IRMA.