The Effects of Recombinant Human Growth Hormone on Body ...

0013-7227/01/$03.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 86(8):3480 –3487 Copyright © 2001 by The Endocrine Society

The Effects of Recombinant Human Growth Hormone on Body Composition and Glucose Metabolism in HIV-Infected Patients with Fat Accumulation JOAN C. LO, KATHLEEN MULLIGAN, MUSTAFA A. NOOR, JEAN-MARC SCHWARZ, ROBERT A. HALVORSEN*, CARL GRUNFELD, AND MORRIS SCHAMBELAN Division of Endocrinology, Department of Medicine (J.C.L., K.M., M.A.N., J.-M.S., C.G., M.S.), and Department of Radiology (R.A.H), University of California–San Francisco, San Francisco, California 94143; and Department of Nutritional Sciences (J.-M.S.), University of California–Berkeley, Berkeley, California 94720 GH has been proposed as a therapy for patients with HIVassociated fat accumulation, but the pharmacological doses (6 mg/d) used have been associated with impaired fasting glucose and hyperglycemia. In contrast, physiologic doses of GH (⬃ 1 mg/d) in HIV-negative men reduced visceral adiposity and eventually improved insulin sensitivity, despite initially causing insulin resistance. We conducted an open-label study to evaluate the effects of a lower pharmacologic dose of GH (3 mg/d) in eight men with HIV-associated fat accumulation. Oral glucose tolerance, insulin sensitivity, and body composition were measured at baseline, and 1 and 6 months. Six patients completed 1 month and 5, 6 months of GH therapy. IGF-I levels increased 4-fold within 1 month of GH treatment. Over 6 months, GH reduced buffalo hump size and excess visceral adipose tissue. Total body fat decreased (17.9 ⴞ 10.9 to 13.5 ⴞ 8.4 kg, P ⴝ 0.05), primarily in the trunk region. Lean body mass increased (62.9 ⴞ 6.4 to 68.3 ⴞ 9.1 kg, P ⴝ 0.03). Insulin-mediated glucose disposal, measured by a euglycemic hyperinsulinemic clamp, declined at month 1 (49.7 ⴞ 27.5 to 25.6 ⴞ 6.6 nmol/kgLBM䡠min/pmolINSULIN/liter, P ⴝ 0.04); values improved at month 6 (49.2 ⴞ 22.6, P ⴝ 0.03, compared with

W

ITH THE CHANGING epidemiology of HIV disease and the advent of highly active antiretroviral therapy (HAART), there has been a dramatic reduction in HIVassociated morbidity and mortality (1). However, the optimism generated by overall improving health among patients on HAART has been tempered by concerns that these therapies may be associated with adverse metabolic effects, such as insulin resistance and hyperlipidemia, and changes in body composition (2–10). In the past 4 yr, several syndromes of regional fat accumulation have been reported in patients with HIV infection, including dorsocervical fat pad enlargement (“buffalo hump”), symmetric lipomatosis, and abdominal obesity (7, 8, 11–15), although not all of these patients were on HAART. In addition, loss of sc fat in the face, buttocks, and/or extremities, either alone or in combination with syndromes of fat accumulation, has been reported (2, 15, 16). The mechanisms underlying the abnormal distribution of fat and their relationship to specific antiretroviral theraAbbreviations: AUC, Area under the curve; CT, computed tomography; CV, coefficient(s) of variation; GCRC, General Clinical Research Center; HAART, highly active antiretroviral therapy; HbA1C, hemoglobin A1C; HDL, high-density lipoprotein; LDL, low-density lipoprotein; OGTT, oral glucose tolerance test; SAT, sc adipose tissue; VAT, visceral adipose tissue.

month 1) and did not differ significantly from baseline. Similarly, the integrated response to an oral glucose load worsened at month 1 (glucose area under the curve 20.1 ⴞ 2.3 to 24.6 ⴞ 3.7 mmol䡠h/liter, P < 0.01), whereas values improved at month 6 (22.1 ⴞ 1.5, P ⴝ 0.02, compared with month 1) and did not differ significantly from baseline. One patient developed symptomatic hyperglycemia within 2 wk of GH initiation; baseline oral glucose tolerance testing revealed preexisting diabetes despite normal fasting glucose. In conclusion, GH at 3 mg/d resulted in a decrease in total body fat and an increase in lean body mass in this open-label trial. While insulin sensitivity and glucose tolerance initially worsened, they subsequently improved toward baseline. However, the dose of GH used in this trial was supraphysiologic and led to an increase in IGF-I levels up to three times the upper normal range. Because there are known adverse effects of long-term GH excess, the effectiveness of lower doses of GH should be studied. We also recommend a screening oral glucose tolerance test be performed to exclude subjects at risk for GH-induced hyperglycemia. (J Clin Endocrinol Metab 86: 3480 –3487, 2001)

pies remain unclear. For many patients, these alterations have had a substantial negative impact on quality of life. Concerns have also been raised about the possible cardiovascular risks that these body composition changes, together with associated insulin resistance and dyslipidemia, may convey (17, 18). Although there are no approved drugs for fat distribution abnormalities, reports of treatments that have reversed these abnormalities have led patients and clinicians to use unproven, and in some cases, potentially deleterious therapies. Included among treatments that are being used is recombinant human GH, which was granted accelerated approval in 1996 for the treatment of HIV-associated wasting (19). Preliminary studies suggest that GH may have a therapeutic role in HIV-associated fat accumulation, based on its lipolytic effect on adipose tissue (20 –23). However, endogenous GH excess (i.e. acromegaly) and GH administration have been associated with insulin resistance and carbohydrate intolerance (24 –28). To date, no studies have systematically evaluated the effect of GH on glucose metabolism in patients with HIV-associated fat accumulation. On the other hand, in HIVnegative men with abdominal obesity, treatment with GH at a dose of 9.5 ␮g/kg䡠d (⬃1 mg/d) has been shown to reduce

3480

Lo et al. • GH Treatment of HIV Fat Accumulation

The Journal of Clinical Endocrinology & Metabolism, August 2001, 86(8):3480 –3487 3481

total body fat and visceral abdominal fat mass (29). In these subjects, despite an initial worsening of insulin sensitivity during the first 6 wk presumably due to GH-induced insulin resistance, insulin sensitivity improved compared with placebo-treated subjects after 9 months of therapy. The authors speculated that the eventual improvement in insulin sensitivity might have been due to the reduction in visceral adiposity observed with GH treatment. A similar pattern in insulin action, but without improvement beyond baseline levels, has also been reported in GH-deficient adults treated with replacement doses of GH (27). In HIV-positive men with dorsocervical fat pad enlargement and/or abdominal obesity, treatment with pharmacologic doses of GH (4 – 6 mg/d) has been shown to reduce hump size, abdominal girth, waistto-hip ratio, and visceral adipose tissue volume (20 –22, 30). However, in one group of patients who received a dose of 6 mg/d, mean fasting glucose levels increased significantly into the impaired fasting glucose range (21), and frank hyperglycemia has been reported in several patients (21, 22, 30). We hypothesized that a lower dose of GH (3 mg/d) would reduce buffalo hump and abdominal adiposity in HIVinfected patients with fat accumulation and result in an initial impairment in glucose tolerance and insulin sensitivity, followed by improvement over time, similar to the pattern seen in HIV-negative individuals. We selected a supraphysiologic dose of GH based on prior evidence of relative GH resistance in HIV-infected patients with the wasting syndrome; in these patients, administration of a pharmacologic (100 ␮g/kg䡠d or ⬃6 mg/d) dose of GH resulted in a significantly lower IGF-I response when compared with healthy HIV-negative control subjects (31, 32). This report describes the first study of GH therapy in HIV- infected patients with regional fat accumulation in which insulin action has been quantified in conjunction with body composition measurements. Materials and Methods Subjects and study protocol Eight HIV-positive men who had developed an enlarged dorsocervical fat pad and/or abdominal obesity while on antiretroviral therapy were enrolled in the study. Patients with overt diabetes, fasting triglyceride levels of 11.29 mmol/liter or higher, or active malignancy were excluded. The study protocol was approved by the Committee on Human Research of the University of California–San Francisco, and informed consent was obtained from each subject. Subjects were admitted to the General Clinical Research Center (GCRC) at San Francisco General Hospital for 5 d and placed on a diet with fixed proportions of carbohydrate, fat, protein, and energy content designed to maintain body weight and minimize dietary influences on metabolism (33); the identical diet was given during subsequent inpatient metabolic evaluations. Fasting lipid studies were performed on d 2, an oral glucose tolerance test on d 3, body composition measurements on d 4, and a euglycemic hyperinsulinemic clamp on d 5. After the baseline evaluation, subjects began treatment with recombinant human GH (Serono Laboratories, Inc., Norwell, MA) at a dose of 3 mg/d (30 – 40 ␮g/kg䡠d) by sc injection on an outpatient basis. The same 5-d metabolic assessments were performed at months 1 and 6 while on GH therapy.

Laboratory measurements Hemoglobin A1C (HbA1C) levels were measured by ion exchange high-performance liquid chromatography and CD4 cell count by flow cytometry in the Clinical Laboratories at San Francisco General Hospital. HIV-1 RNA levels were quantified using the Amplicor HIV-1 Monitor test (detection limit 50 copies/ml; Roche Diagnostics, Branchburg, NJ)

in the Virology Core Laboratory of the Gladstone Institute of Virology and Immunology. Fasting triglyceride, total and high-density lipoprotein (HDL) cholesterol levels were measured by enzymatic colorimetric methods using reagents from Sigma Diagnostics (St. Louis, MO). HDL cholesterol was quantified after precipitation of Apo B-containing lipoproteins with phosphotungstic acid and MgCl2. The intra-assay coefficients of variation (CV) for triglyceride and total and HDL cholesterol are 3%, 3%, and 1%, respectively. Low-density lipoprotein (LDL) cholesterol levels were calculated based on the Friedewald equation (34). Direct LDL cholesterol levels were measured using reagents from Equal Diagnostics (Exton, PA) with an intra-assay CV of 1%. In addition, the following measurements were performed in the Core Laboratory of the GCRC at San Francisco General Hospital: whole blood and plasma glucose levels by the glucose oxidase method (YSI 2300 STAT-Plus Glucose/Lactate Analyzer, YSI Inc., Yellow Springs, OH); IGF-I levels by immunoradiometric assay (Nichols Institute Diagnostics, San Juan Capistrano, CA) with an intra-assay CV of 4.6%; and serum insulin levels by RIA (Coat-A-Count, Diagnostic Products Corporation, Los Angeles, CA) with an intra-assay CV of 7%, interassay CV of 14.1%, and 40% cross-reactivity with proinsulin.

Body composition Weight was measured on a calibrated scale each morning after an overnight fast with patients wearing only a hospital gown. Height was measured using a stadiometer. Body mass index was calculated as body weight (kilograms) divided by height (meters) squared. Anthropometric measurements were performed in triplicate, including waist circumference measured at the level of the umbilicus and hip circumference measured at the point of maximal gluteal width when viewed from the side. The dorsocervical fat pad was measured in length and width (along surface contours), and the size was estimated by multiplying length and width. Fat and lean body mass were measured by dual-energy x-ray absorptiometry (Lunar Corp. model DPX, Madison, WI) (35). Manual analyses of fat in the arms, legs, and trunk were performed using skeletal and soft-tissue landmarks, as described previously (11), using software version 3.65 in the extended analysis mode. In our laboratory, the coefficients of variation for repeated regional analyses of trunk, arm, and leg fat are 1.0, 2.7, and 1.4%, respectively. Abdominal adipose tissue was measured by computed tomography (CT) using a helical HiSpeed CT/i Scanner (General Electric Medical Systems, Milwaukee, WI). Patients were examined in the supine position with their arms stretched above their head. A preliminary scout image was obtained of the abdomen and pelvis in supine and lateral projections for anatomic localization. A single 7-mm slice obtained at the level of the L4 –L5 intervertebral disc space was used to quantify visceral and sc adipose tissue area (VAT and SAT). The axial image was displayed in a matrix of 512 ⫻ 512 pixels, and adipose tissue was identified by density mapping for pixel values between ⫺30 and ⫺190 Hounsfield units (36). A cursor line was traced, bisecting the abdominal musculature anterolaterally and outlining the transversalis fascia posteriorly (including the psoas muscles but excluding the vertebrae), to separate the visceral and sc compartments. All images were analyzed by one investigator (J.C.L.) with a CV less than 1% on repeated analyses.

Glucose tolerance and insulin sensitivity An oral glucose tolerance test (OGTT) was performed at 0800 h after a 10-h overnight fast and 200 g/d carbohydrate intake or higher during the previous 3 d. Patients were administered 75 g dextrose in 250 ml water, and blood samples were collected at 0, 30, 60, 90, 120, and 180 min for measurement of plasma glucose levels. The area under the curve (AUC) for glucose was calculated by the trapezoid method and used as an integrated measure of glucose tolerance. The following glucose tolerance criteria from the World Health Organization were used for clinical characterization (37): normal glucose tolerance (2-h glucose ⬍7.8 mmol/liter); impaired glucose tolerance (2-h glucose 7.8 –11.0 mmol/ liter); and overt glucose intolerance or diabetes (2-h glucose ⱖ11.1 mmol/liter). Peripheral insulin sensitivity was measured by the euglycemic hyperinsulinemic clamp method (38). After an overnight fast, insulin (Humulin R; Eli Lilly & Co., Indianapolis, IN) was administered as a primed continuous iv infusion at a rate of 40 mU/m2 䡠 min for 180 min. During this period, whole blood glucose levels were measured

3482

The Journal of Clinical Endocrinology & Metabolism, August 2001, 86(8):3480 –3487

every 5 min from a retrograde iv line placed in a hand that was warmed in a heated box at 50⫺55 C for arterialized blood sampling. A variable infusion of 20% dextrose was used to maintain euglycemia with a CV less than 5% during steady state. Blood samples were collected for post hoc determination of serum insulin concentration. The M value was calculated as a measure of peripheral insulin sensitivity according to De Fronzo and co-workers (38, 39), based on steady-state glucose infusion rates during the final 60 min of the clamp and lean body mass (kg) by dual-energy x-ray absorptiometry, and was adjusted for the steady-state insulin concentration achieved (M/I).

Statistical methods Data are expressed as mean ⫾ sd. Differences between baseline and 6 months were analyzed by Student’s paired t test for continuous outcomes. For measurements pertaining to glucose metabolism, differences between baseline, 1 month, and 6 months were initially analyzed using ANOVA for repeated measures. If the global comparison was statistically significant, pairwise comparisons were conducted between time points using the Student-Newman-Keuls test for multiple comparisons, based on our a priori hypothesis that GH treatment would result in an initial worsening followed by subsequent improvement in glucose metabolism. A two-sided P value less than 0.05 was considered statistically significant. All analyses were performed using SigmaStat software version 2.0 (SPSS Inc., Chicago, IL).

Results Baseline characteristics and clinical course

The clinical and biochemical characteristics of the eight men with HIV-associated fat accumulation are shown in Tables 1 and 2. Five patients had a buffalo hump, including two with associated lipomas at the base of the scalp posteriorly; one of these patients also had abdominal obesity. Three patients had only abdominal obesity. All patients with a buffalo hump had evidence of an elevated waist to hip ratio, consistent with our previous finding of increased trunk fat in HIV-infected men with a buffalo hump (11) and suggestive of a common underlying phenotype. The median duration of HIV infection was 12 yr (range, 5–16 yr). All patients had been on a stable antiretroviral regimen for at least 2 months before enrollment (patients 2– 8 for at least 6 months). Six patients were on combination therapy that included a protease inhibitor, while patient 7 was naı¨ve to protease inhibitor therapy. Patient 1 was previously on a protease inhibitor that had been discontinued 6 months before the baseline study due to virologic failure. All patients remained on a stable antiretroviral regimen during the study period, except patient 3, in whom abacavir was substituted for stavudine 1 wk after GH initiation due to persistent stavudine-associated neuropathy. Patient 5 had been on stable replacement doses of testosterone for 2 yr, and patient 2 on low-dose nandrolone decanoate for 4 months; both patients continued these ther-


apies at the same dose throughout the study. None of the other patients were on anabolic agents in the 12 months before or during the study. In addition, none of the patients had received systemic glucocorticoid or megestrol acetate therapy. Overall, GH was well tolerated. The most common complaints were arthralgias and nonpitting edema in the extremities that tended to improve during the course of the study. Patient 3 had persistent arthralgias that necessitated a reduction in the dose of GH to 1.5 mg/d at month 2 and for the remainder of the study, with improvement in symptoms. Patient 2 developed fasting hyperglycemia (glucose levels ranged 11.1–13.9 mmol/liter) within 2 wk of GH initiation, and treatment was discontinued at wk 3 (hence, there are no follow-up data from patient 2 presented here). The hyperglycemia persisted 7 wk after GH cessation (fasting glucose 12.7 mmol/liter, HbA1C 10%), and metformin treatment was required to normalize fasting glucose levels. In this patient, post hoc analysis of OGTT data obtained before GH treatment revealed overt glucose intolerance consistent with diabetes mellitus (2-h glucose 13.8 mmol/liter) despite a normal fasting glucose (5.3 mmol/liter) and HbA1C (6.3%) levels. Based on this observation, the OGTT results were evaluated in all subsequent patients (patients 3– 8) before GH initiation, and patients with overt glucose intolerance (2-h glucose ⱖ11.1 mmol/liter) were excluded. Two additional patients did not complete the study for reasons unrelated to GH treatment. Patient 1 developed antiretroviral drug-induced hepatitis and was withdrawn from the study before the month 1 assessment (hence, there are no follow-up data from patient 1 presented here). Patient 8 relocated after the month 1 assessment and was discharged from the study. The results reported below pertain to the subjects who completed at least 1 month of GH. Laboratory data

In the six patients who completed 1 month of GH therapy, a 4-fold increase in IGF-I levels was observed (21.0 ⫾ 3.3 to 88.7 ⫾ 12.2 nmol/liter, P ⬍ 0.001; Fig. 1). IGF-I levels were in the supraphysiologic range (up to three times the upper limit of normal) and remained elevated at 6 months, with the exception of patient 3, whose GH dose had been reduced to 1.5 mg/d. After 6 months of GH treatment, there was a trend toward reduced total cholesterol (6.03 ⫾ 1.09 to 5.03 ⫾ 1.18 mmol/liter, P ⫽ 0.09) and increased HDL cholesterol (0.72 ⫾ 0.07 to 0.81 ⫾ 0.12 mmol/liter, P ⫽ 0.09) levels, with a significant reduction in the total cholesterol to HDL cholesterol

TABLE 1. Baseline characteristics of the study subjects Patient no.

Age (yr)

CD4 (cells/␮l)

1 2 3 4 5 6 7 8

44 58 65 57 51 36 47 37

258 583 422 267 283 273 822 326

Antiretroviral treatment regimen

D4T, D4T, D4T, D4T, D4T, AZT, AZT, D4T,

DDI, hydroxyurea 3TC, RTV, SQV 3TC, NFV 3TC, IDV 3TC, IDV 3TC, NFV 3TC NEV, IDV

Abnormalities in body fat distribution

Dorsocervical fat pad Abdominal obesity Abdominal obesity Dorsocervical fat pad and symmetric lipomata Dorsocervical fat pad Dorsocervical fat pad and symmetric lipomata; abdominal obesity Abdominal obesity Dorsocervical fat pad

D4T, Stavudine; AZT, zidovudine; 3TC, lamivudine; NEV, nevirapine; NFV, nelfinavir; IDV, indinavir; RTV, ritonavir; SQV, saquinavir.



TABLE 2. Anthropometric and fasting metabolic measurements obtained at baseline Patient no.

1 2 3 4 5 6 7 8 Mean ⫾

SD

BMI (kg/m2)

Waist to hip ratio

Glucose (mmol/liter)

Insulin (pmol/liter)

CHOL (mmol/liter)

TRIG (mmol/liter)

HDL (mmol/liter)

LDL (mmol/liter)

23.0 31.1 23.0 26.2 25.0 29.0 29.5 25.5 26.5 ⫾ 3.0

0.96 1.09 1.00 0.98 0.99 1.00 0.99 0.99 1.00 ⫾ 0.04

4.9 5.3 4.7 5.0 5.2 4.8 4.9 4.3 4.9 ⫾ 0.3

118 271 90 204 93 224 60 153 152 ⫾ 75

3.72 5.53 7.55 6.15 6.31 5.53 4.58 7.99 5.92 ⫾ 1.42

1.19 8.30 2.99 3.71 4.85 3.64 1.26 7.14 4.13 ⫾ 2.55

1.09 0.54 0.80 0.65 0.78 0.67 0.72 0.75 0.75 ⫾ 0.16

3.08 3.10 4.86 3.65 3.41 3.10 2.72 3.36 3.41 ⫾ 0.65

BMI, Body mass index; CHOL, total cholesterol; TRIG, triglycerides. To convert glucose from mmol/liter to mg/dl, multiply by 18.01. To convert insulin from pmol/liter to ␮U/ml, multiply by 0.1394. To convert cholesterol from mmol/liter to mg/dl, multiply by 38.67. To convert triglycerides from mmol/liter to mg/dl, multiply by 88.57. TABLE 3. The effect of GH therapy on body composition measured by dual-energy x-ray absorptiometry

Lean body mass Total body fat Trunk fat Appendicular fat

Baseline (kg) (n ⫽ 6)

Change (kg) 0 –1 month

Change (kg) 0 – 6 months

Pa

62.5 ⫾ 5.8 16.9 ⫾ 10.1 10.6 ⫾ 5.8 5.2 ⫾ 4.3

⫹4.3 ⫾ 1.9 ⫺2.1 ⫾ 1.1 ⫺1.5 ⫾ 2.8 ⫺0.2 ⫾ 0.4

⫹5.4 ⫾ 3.4 ⫺4.4 ⫾ 3.6 ⫺3.7 ⫾ 2.5 ⫺0.5 ⫾ 1.4

0.03 0.05 0.03 NS

a Paired t test between baseline and 6-month values for patients 3–7. NS, Not significant.

FIG. 1. IGF-I levels in response to GH treatment at a dose of 3 mg/d. IGF-I levels increased 4-fold by month 1 and remained elevated, except in patient 3, in whom the GH dose was reduced from 3.0 to 1.5 mg/d after 2 months due to persistent arthralgias. Patient 3 (F); patient 4 (E); patient 5 (f); patient 6 (䡺); patient 7 (Œ); patient 8 (‚). To convert nmol/liter to ng/ml, multiply by 7.649.

ratio (8.3 ⫾ 1.3 to 6.3 ⫾ 1.3, P ⫽ 0.04). No significant changes were seen in direct or calculated LDL cholesterol levels and triglyceride levels. Importantly, hypertriglyceridemia was not exacerbated by GH in any patient during the study. There were no significant changes in CD4 cell counts. HIV viral load remained stable in all but one patient (viral load ⬍50 copies/ml in patients 3, 5, 6, and 8 and 1,590 – 8,218 copies/ml in patient 7). Patient 4 had evidence of virologic failure by the end of the study (viral load increased from 1,070 to 20,880 copies/ml between baseline and month 6) after having been stable on his antiretroviral regimen for the three previous yr. Body composition

Treatment with GH significantly increased lean body mass (62.9 ⫾ 6.4 to 68.3 ⫾ 9.1 kg, P ⫽ 0.03) in the five patients who completed 6 months of therapy (Table 3). Total fat mass decreased in all patients (17.9 ⫾ 10.9 to 13.5 ⫾ 8.4 kg, P ⫽ 0.05), primarily in the trunk region, including patient 3, whose dose was reduced to 1.5 mg/d. No significant changes were seen in appendicular fat or total body weight. There

was a trend toward visceral fat reduction (17166 ⫾ 8065 to 10863 ⫾ 2922 mm2, P ⫽ 0.08) for the group as a whole, an effect that was driven by the substantial (30 – 60%) decline in VAT in the three patients with the greatest amounts of visceral adiposity at baseline (Fig. 2). On the other hand, the patients with less VAT experienced minimal change after 6 months. There was no reduction in sc abdominal fat, except in patient 7, who had the largest amount of sc fat at baseline (Figs. 2 and 3). By the end of 6 months, patients 4 and 5 experienced 43% and 32% reduction in hump size, respectively. Patient 6 had a progressively enlarging dorsocervical fat pad at the time of study enrollment that continued to increase (40%) by month 1 but then decreased 31% in size by month 6. Glucose metabolism

Five of the six patients who completed at least 1 month of GH treatment had normal glucose tolerance at baseline. Patient 6 had mildly impaired glucose tolerance (2-h glucose 8.8 mmol/liter). In this group, selected for the absence of overt glucose intolerance (diabetes by OGTT criteria) at baseline, fasting glucose levels increased significantly at month 1 (4.8 ⫾ 0.3 to 5.4 ⫾ 0.3 mmol/liter, n ⫽ 6, P ⫽ 0.02) and tended to remain elevated at month 6 (5.3 ⫾ 0.3 mmol/liter, n ⫽ 5, P ⫽ 0.06) compared with baseline. Fasting insulin levels increased significantly at month 1 (137 ⫾ 67 pmol/liter to 220 ⫾ 85, P ⫽ 0.01), but declined by month 6 (159 ⫾ 76 pmol/liter, P ⫽ 0.05, compared with month 1); values at month 6 did not differ significantly from baseline. There was a small increase in HbA1C levels at month 6 (5.1 ⫾ 0.3 to 5.6 ⫾ 0.5%, P ⫽ 0.02) compared with baseline, but values remained within the normal range. The glucose AUC increased at month 1 in all six patients (20.1 ⫾ 2.3 to 24.6 ⫾ 3.7 mmol䡠h/ liter, P ⬍ 0.01); values declined by month 6 (22.1 ⫾ 1.5

3484



FIG. 2. The effect of GH treatment on VAT and SAT. The same symbols (see Fig. 1) are used to refer to patients 3–7.

mmol䡠h/liter, n ⫽ 5, P ⫽ 0.02, compared with month 1) and did not differ significantly from baseline (Fig. 4). However, a mild persisting impairment in oral glucose tolerance remained evident at month 6 (2-h glucose 8.2 ⫾ 1.0 vs. 6.9 ⫾ 1.1 mmol/liter at baseline, P ⫽ 0.02). Patient 3 developed overt glucose intolerance consistent with diabetes (2-h glucose 12.0 mmol/liter) at month 1 but then improved substantially by month 6 (as previously noted, his GH dose was reduced secondary to arthralgias). However, even with the exclusion of patient 3, this pattern in oral glucose tolerance remained apparent, with initial worsening at month 1 (glucose AUC 20.1 ⫾ 2.5 to 23.4 ⫾ 2.4 mmol䡠h/liter, n ⫽ 5, P ⬍ 0.001) and then improvement at month 6 (22.0 ⫾ 1.8, n ⫽ 4, P ⬍ 0.001, compared with month 1), although values at month 6 remained higher than baseline (P ⫽ 0.01). Insulin-stimulated glucose uptake (M/I) decreased at month 1 (49.7 ⫾ 27.5 to 25.6 ⫾ 6.6 nmol/kgLBM䡠min/pmolINSULIN/liter, n ⫽ 6, P ⫽ 0.04); values were improved at month 6 (49.2 ⫾ 22.6, n ⫽ 5, P ⫽ 0.03, compared with month 1) and did not differ significantly from baseline. Three of the five patients demonstrated this pattern in peripheral insulin sensitivity, with worsening at month 1 and improvement at month 6 (Fig. 4). In patient 4, who was the most insulin resistant at baseline, insulin sensitivity did not worsen at month 1 but rather improved slightly during the course of the study. Patient 6, who had impaired glucose tolerance at baseline, had a reduction in insulin sensitivity at month 1 and only minimal improvement by month 6. When patient 3 (whose dose of GH had been reduced) was excluded from the analysis, the overall differences in insulin-mediated glucose disposal did not reach statistical significance (44.2 ⫾ 26.8 at baseline, 23.9 ⫾ 5.8 at month 1, 45.3 ⫾ 24.1 at month 6, P ⫽ 0.11). Discussion

The results of this pilot study suggest that administration of GH at a dose of 3 mg/d reduced total body and trunk fat in patients with HIV-associated fat accumulation. Furthermore, GH treatment reduced buffalo hump size and excess

visceral adiposity. These results need to be confirmed in a larger randomized, placebo-controlled study. While changes in appendicular fat were not observed in the small number of patients enrolled, other studies have reported that appendicular fat stores may be reduced with GH treatment (40, 41). Thus, while GH may be an effective therapy for patients with regional fat accumulation, it would not be appropriate for patients whose primary complaint is lipoatrophy. Consistent with results in HIV-negative individuals, oral glucose tolerance initially worsened after 1 month of GH therapy and then improved toward baseline after 6 months. However, plasma glucose measurements obtained at 6 months were slightly higher than baseline levels and consistent with a mild persisting impairment in glucose tolerance. In the one patient who, in retrospect, had overt glucose intolerance at baseline, GH therapy led to fasting hyperglycemia and symptomatic diabetes within 2 wk of treatment initiation. The effect of GH on insulin-mediated glucose disposal followed a similar pattern with initial worsening and then improvement to baseline values, as evident in the majority of the subjects. It should be noted that this biphasic response occurred in patients selected for the absence of overt glucose intolerance (diabetes by OGTT) at baseline and does not apply to those with abnormal glucose tolerance. Overall, the initial detrimental effects on glucose metabolism observed after 1 month of GH treatment are consistent with previous observations of GH action attributed to impairment of insulin action in the liver and peripheral tissues (24, 26 –28). In the current study, the specific changes in hepatic glucose regulation and intermediary carbohydrate metabolism remain to be elucidated. The subsequent improvement in glucose metabolism toward baseline values might have been mediated, in part, by the reduction in body fat. However, we were not able to establish a causal relationship between visceral fat reduction and improvement in insulin sensitivity in this small number of patients. It is also possible that other effects of GH, including the action of GH-induced increases in IGF-I levels on glucose transport



FIG. 3. Abdominal CT in patient 7 before (A) and after (B) GH treatment. Note the marked reduction in sc and visceral fat; adipose tissue has relatively low attenuation and is represented by the dark gray area.

(42), may have contributed to these findings. Among the patients with normal glucose tolerance, the degree of fasting hyperinsulinemia at baseline did not reliably predict the GH-induced changes in peripheral insulin sensitivity or response to glucose challenge. In our current study, a 4-fold increase in IGF-I levels (up to three times the upper normal range) was observed with administration of GH at a pharmacologic dose of 3 mg/d. These results are in contrast to previous studies in patients with HIV-associated wasting demonstrating less than a 2-fold increase in IGF-I concentrations following treatment with GH at an approximate dose of 6 mg/d (100 ␮g/kg䡠d)

(19, 31), indicating that the GH resistance in HIV-infected patients with wasting is not apparent in patients with fat accumulation. Indeed, the dose of GH used in the current study would be expected to result in supraphysiologic IGF-I levels, except in individuals with GH resistance. Our observations underscore the need for further investigation of lower doses in the treatment of affected patients to avoid negative effects of long-term GH excess. Evidence that obesity and excess visceral fat are associated with blunted GH secretion in HIV-negative individuals (43– 45) may provide an additional rationale for low-dose GH therapy in patients with fat accumulation syndromes. Indeed, recent data indi-

3486


FIG. 4. The effect of GH treatment on (a) oral glucose tolerance and (b) insulin-mediated glucose uptake. The same symbols (see Fig. 1) are used to refer to Patients 3– 8.

cate that visceral adiposity seems to be a strong predictor of reduced GH concentrations in patients with HIV-related lipodystrophy (46), and future studies should assess whether individuals with excess visceral fat might respond best to low-dose GH treatment. Furthermore, a dose-ranging study may be necessary to determine the optimal dose of GH to reduce visceral adiposity, improve insulin sensitivity and minimize complications. Within our sample of men with HIV-associated fat accumulation, in whom baseline HDL cholesterol concentrations were low and fasting triglyceride levels were elevated in the majority of cases, GH treatment was associated with a more favorable lipoprotein profile. After 6 months of treatment, a significant reduction in the total cholesterol to HDL cholesterol ratio was apparent, as well as a tendency toward decreased total cholesterol and increased HDL cholesterol levels. In addition, hypertriglyceridemia was not exacerbated


despite the pharmacologic doses of GH used. Additional studies are needed to confirm these preliminary observations and to determine the potential metabolic pathways involved. Although one patient experienced virologic failure during the course of study, previous studies have shown no deleterious effect of GH on HIV replication as measured by HIV RNA levels in plasma, immune-complex dissociated p24 antigen levels, and titers of infectious viremia in plasma, particularly in patients with high viral burden (19). Concerns about a potential effect of GH on HIV replication were initially raised when in vitro data showed augmented release of p24 antigen from phytohemagglutinin-activated, HIVinfected mononuclear cells after incubation with GH (47). However, this effect was variable and inhibited by the addition of zidovudine to the culture media. All patients in the current study continued antiretroviral therapy while receiving GH. The fact that one patient experienced virologic failure after maintaining the same antiretroviral regimen for more than 3 yr is not unexpected (48, 49). In conclusion, treatment with GH at 3 mg/d resulted in a decrease in total body fat and an increase in lean body mass in this open-label trial. The most common side effects seen during the study were arthralgias and edema, probably reflecting the pharmacologic dose of GH used and the resultant supraphysiologic levels of IGF-I. Given the known negative aspects of chronic supraphysiologic GH and IGF-I levels, notably acromegaly with increased morbidity and mortality, long-term exposure to these high doses cannot be recommended at this time as treatment for HIV-infected men with fat accumulation. Rather, future studies should assess the efficacy of lower doses of GH that result in serum IGF-I levels within the normal range. Even with lower doses, we recommend performing an OGTT to exclude individuals with glucose intolerance before GH initiation. Testing may be particularly relevant as a significant proportion of HIV-infected subjects with fat distribution abnormalities have evidence of abnormal glucose tolerance despite normal fasting glucose levels (50, 51). In these individuals, significant hyperglycemia and frank diabetes mellitus can be precipitated by GH therapy. Finally, appropriate caution with monitoring of IGF-I levels, fasting glucose concentrations, and subsequent oral glucose tolerance testing are advised for patients continued on maintenance doses of GH therapy to avoid the long-term adverse consequences associated with hyperglycemia and GH excess. Acknowledgments We thank Scott Evans, M.D., Kyle Yu, M.D., and Rafael Ibarra, R.T., for efforts in developing the radiology CT protocol; the GCRC Nursing and Dietary Staff for help in performing the inpatient metabolic studies; Serono Laboratories, Inc. for the provision of GH; and Barbara Chang, Joy Hirai, Natalie Patterson, Carlynn Yee-Hicaiji, and Viva Tai for technical assistance. Received December 5, 2000. Accepted April 6, 2001. Address all correspondence and requests for reprints to: Joan C. Lo, M.D., Division of Endocrinology, San Francisco General Hospital, Building 30, Room 3501-K, San Francisco, California 94110. E-mail: jlo@itsa. ucsf.edu. This work was supported by the NIH (Grants DK45833 and DK54615) and the Universitywide AIDS Research Program (F98-SF-054). J.C.L. is a recipient of a Clinical Associate Physician Award from the National



Center for Research Resources. All studies were conducted in the General Clinical Research Center at San Francisco General Hospital with support by the National Center for Research Resources, NIH (Grant RR00083). This work was presented in part at the 2nd International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV, Toronto, Canada, September 2000. * Present address: University of Medicine and Dentistry of New Jersey, Newark, New Jersey 07103.

References 1. Palella Jr FJ, Delaney KM, Moorman AC, et al. 1998 Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 338:853– 860 2. Carr A, Samaras K, Burton S, et al. 1998 A syndrome of peripheral lipodystrophy, hyperlipidaemia and insulin resistance in patients receiving HIV protease inhibitors. AIDS 12:F51–F58 3. Noor MA, Lo JC, Mulligan K, et al. 2001 Metabolic effects of indinavir in healthy HIV-seronegative men. AIDS 15:F11–F18 4. Behrens G, Dejam A, Schmidt H, et al. 1999 Impaired glucose tolerance, beta cell function and lipid metabolism in HIV patients under treatment with protease inhibitors. AIDS 13:F63–F70 5. Walli R, Herfort O, Michl GM, et al. 1998 Treatment with protease inhibitors associated with peripheral insulin resistance and impaired oral glucose tolerance in HIV-1 infected patients. AIDS 12:F167–F173 6. Mulligan K, Grunfeld C, Tai VW, et al. 2000 Hyperlipidemia and insulin resistance are induced by protease inhibitors independent of changes in body composition in patients with HIV infection. J Acquir Immune Defic Syndr 23:35– 43 7. Miller KD, Jones E, Yanovski JA, Shankar R, Feuerstein I, Falloon J 1998 Visceral abdominal-fat accumulation associated with use of indinavir. Lancet 351:871– 875 8. Roth VR, Kravcik S, Angel JB 1998 Development of cervical fat pads following therapy with human immunodeficiency virus type 1 protease inhibitors. Clin Infect Dis 27:65– 67 9. Purnell JQ, Zambon A, Knopp RH, et al. 2000 Effect of ritonavir on lipids and post-heparin lipase activities in normal subjects. AIDS 14:51–57 10. Pe´riard D, Telenti A, Sudre P, et al. 1999 Atherogenic dyslipidemia in HIVinfected individuals treated with protease inhibitors. Circulation 100:700 –705 11. Lo JC, Mulligan K, Tai VW, Algren H, Schambelan M 1998 “Buffalo hump” in men with HIV-1 infection. Lancet 351:867– 870 12. Lo JC, Mulligan K, Tai VW, Algren H, Schambelan M 1998 Body shape changes in HIV-infected patients (Letter). J Acquir Immune Defic Syndr 19: 307–308 13. Miller KK, Daly PA, Sentochnik D, et al. 1998 Pseudo-Cushing’s syndrome in human immunodeficiency virus-infected patients. Clin Infect Dis 27:68 –72 14. Hengel RL, Watts NB, Lennox JL 1997 Benign symmetric lipomatosis associated with protease inhibitors. Lancet 350:1596 15. Saint-Marc T, Partisani M, Poizot-Martin I, et al. 2000 Fat distribution evaluated by computed tomography and metabolic abnormalities in patients undergoing antiretroviral therapy: Preliminary results of the LIPOCO study. AIDS 14:37– 49 16. Mallal SA, John M, Moore CB, James IR, McKinnon EJ 2000 Contribution of nucleoside analogue reverse transcriptase inhibitors to subcutaneous fat wasting in patients with HIV infection. AIDS 14:1309 –1316 17. Safrin S, Grunfeld C 1999 Fat distribution and metabolic changes in patients with HIV infection. AIDS 13:2493–2505 18. Henry K, Melroe H, Huebsch J, et al. 1998 Severe premature coronary artery disease with protease inhibitors (Letter). Lancet 351:1328 19. Schambelan M, Mulligan K, Grunfeld C, et al. 1996 Recombinant human growth hormone in patients with HIV-associated wasting. A randomized, placebo-controlled trial. Ann Intern Med 125:873– 882 20. Torres R, Unger K The effect of recombinant human growth hormone on protease-inhibitor-associated fat maldistribution syndrome. 6th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, 1999; abstract 675 21. Engelson ES, Glesby M, Sheikhan J, et al. 1999 Effect of recombinant human growth hormone in the treatment of visceral fat accumulation in HIV infection: interim analysis. Antivir Ther 4(Suppl 2):19 (Abstract) 22. Wanke C, Gerrior J, Kantaros J, Coakley E, Albrecht M 1999 Recombinant human growth hormone improves the fat redistribution syndrome (lipodystrophy) in patients with HIV. AIDS 13:2099 –2103 23. Mauss S, Wolf E, Jaeger H 1999 Reversal of protease inhibitor-related visceral abdominal fat accumulation with recombinant human growth hormone (Letter). Ann Intern Med 131:313–314 24. Rizza RA, Mandarino LJ, Gerich JE 1982 Effects of growth hormone on insulin action in man. Mechanisms of insulin resistance, impaired suppression of glucose production, and impaired stimulation of glucose utilization. Diabetes 31:663– 669 25. Hansen I, Tsalikian E, Beaufrere B, Gerich J, Haymond M, Rizza R 1986 Insulin resistance in acromegaly: defects in both hepatic and extrahepatic insulin action. Am J Physiol 250:E269 –E273 26. Bratusch-Marrain PR, Smith D, DeFronzo RA 1982 The effect of growth

27. 28. 29.

30.

31.

32. 33.

34. 35. 36. 37. 38. 39.

40.

41.

42. 43. 44.

45. 46. 47. 48. 49. 50.

51.

hormone on glucose metabolism and insulin secretion in man. J Clin Endocrinol Metab 55:973–982 Fowelin J, Attvall S, Lager I, Bengtsson B-Å 1993 Effects of treatment with recombinant human growth hormone on insulin sensitivity and glucose metabolism in adults with growth hormone deficiency. Metabolism 42:1443–1447 Butler P, Kryshak E, Rizza R 1991 Mechanism of growth hormone-induced postprandial carbohydrate intolerance in humans. Am J Physiol 260:E513–E520 Johannsson G, Mårin P, Lo¨nn L, et al. 1997 Growth hormone treatment of abdominally obese men reduces abdominal fat mass, improves glucose and lipoprotein metabolism, and reduces diastolic blood pressure. J Clin Endocrinol Metab 82:727–734 Furrer H, Nguyen Q-V, Malinverni R Treatment of HAART associated fat accumulation disease with recombinant human growth hormone: results of a randomised double blind placebo controlled crossover trial. XIII International AIDS Conference, Durban, South Africa, 2000; abstract LbPp114 Mulligan K, Grunfeld C, Hellerstein MK, Neese RA, Schambelan M 1993 Anabolic effects of recombinant human growth hormone in patients with wasting associated with human immunodeficiency virus infection. J Clin Endocrinol Metab 77:956 –962 Lieberman SA, Butterfield GE, Harrison D, Hoffman AR 1994 Anabolic effects of recombinant insulin-like growth factor-I in cachectic patients with the acquired immunodeficiency syndrome. J Clin Endocrinol Metab 78:404 – 410 Schwarz JM, Neese RA, Turner S, Dare D, Hellerstein MK 1995 Short-term alterations in carbohydrate energy intake in humans. Striking effects on hepatic glucose production, de novo lipogenesis, and whole-body fuel selection. J Clin Invest 96:2735–2743 Friedewald WT, Levy RI, Fredrickson DS 1972 Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 18:499 –502 Mazess RB, Barden HS, Bisek JP, Hanson J 1990 Dual-energy x-ray absortiometry for total-body and regional bone-mineral and soft-tissue composition. Am J Clin Nutr 51:1106 –1112 Kvist H, Chowdhury B, Grangård U, Tyle´n U, Sjo¨stro¨m L 1988 Total and visceral adipose-tissue volumes derived from measurements with computed tomography in adult men and women: predictive equations. Am J Clin Nutr 48:1351–1361 World Health Organization 1985 Diabetes mellitus: Report of a WHO study group. Geneva, World Health Organization (Technical Report Series, No. 727) DeFronzo RA, Tobin JD, Andres R 1979 Glucose clamp technique: A method for quantifying insulin secretion and resistance. Am J Physiol 237:E214 –E223 Matsuda M, DeFronzo RA 1997 In vivo measurement of insulin sensitivity in humans. In: Draznin B, Rizza R, eds. Clinical research in diabetes and obesity. Volume 1: Methods, assessment, and metabolic regulation. Totowa, NJ: Humana Press, Inc.; 23– 65 Tai V, Mulligan K, Algren H, et al. Effects of growth hormone on regional fat distribution in patients with HIV-associated wasting. American Society for Parenteral and Enteral Nutrition, 24th Clinical Congress, Nashville, TN, 2000; abstract P0028 Engelson ES, Glesby M, Sheikan J, Mendez D, Wang J, Kotler DP Body composition changes during and after growth hormone therapy for lipodystrophy with truncal adiposity. XIII International AIDS Conference, Durban, South Africa, 2000; abstract ThPpB1437 Lund S, Flyvbjerg A, Holman GD, Larsen FS, Pedersen O, Schmitz O 1994 Comparative effects of IGF-1 and insulin on the glucose transporter system in rat muscle. Am J Physiol 267:E461–E466 Mårin P, Kvist H, Lindstedt G, Sjo¨stro¨m L, Bjo¨rntorp P 1993 Low concentrations of insulin-like growth factor-1 in abdominal obesity. Int J Obes 17:83– 89 Veldhuis JD, Liem AY, South S, et al. 1995 Differential impact of age, sex steroid hormones, and obesity on basal versus pulsatile growth hormone secretion in men as assessed in an ultrasensitive chemiluminescence assay. J Clin Endocrinol Metab 80:3209 –3222 Vahl N, Jørgensen JOL, Skjærbæk C, Veldhuis JD, Ørskov H, Christiansen JS 1997 Abdominal adiposity rather than age and sex predicts mass and regularity of GH secretion in healthy adults. Am J Physiol 272:E1108 –E1116 Rietschel P, Hadigan C, Corcoran C, et al. 2001 Assessment of growth hormone dynamics in human immunodeficiency virus-related lipodystrophy. J Clin Endocrinol Metab 86:504 –510 Laurence J, Grimison B, Gonenne A 1992 Effect of recombinant human growth hormone on acute and chronic human immunodeficiency virus infection in vitro. Blood 79:467– 472 Ledergerber B, Egger M, Opravil M, et al. 1999 Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss HIV Cohort Study. Lancet 353:863– 868 Carpenter CCJ, Cooper DA, Fischl MA, et al. 2000 Antiretroviral therapy in adults. Updated recommendations of the International AIDS Society-USA Panel. JAMA 283:381–390 Carr A, Samaras K, Thorisdottir A, Kaufmann GR, Chisholm DJ, Cooper DA 1999 Diagnosis, prediction, and natural course of HIV-1 protease-inhibitorassociated lipodystrophy, hyperlipidaemia, and diabetes mellitus: a cohort study. Lancet 353:2093–2099 Hadigan C, Corcoran C, Basgoz N, Davis B, Sax P, Grinspoon S 2000 Metformin in the treatment of HIV lipodystrophy syndrome. A randomized controlled trial. JAMA 284:472– 477