Effects of Insulin-Like Growth Factor (IGF)-I/IGF-Binding Protein-3 Treatment on Glucose Metabolism and Fat Distribution in Human Immunodeficiency Virus-Infected Patients with Abdominal Obesity and Insulin Resistance

Abstract

Context: HIV-infected patients on antiretroviral therapy are at increased risk for excess visceral adiposity and insulin resistance. Treatment with GH decreases visceral adiposity but worsens glucose metabolism. IGF-I, which mediates many of the effects of GH, improves insulin sensitivity in HIV-negative individuals.

Objective: Our objective was to determine whether IGF-I, complexed to its major binding protein, IGF-binding protein-3 (IGFBP-3), improves glucose metabolism and alters body fat distribution in HIV-infected patients with abdominal obesity and insulin resistance.

Methods: We conducted a pilot, open-label study in 13 HIV-infected men with excess abdominal adiposity and insulin resistance to assess the effect of 3 months of treatment with IGF-I/IGFBP-3 on glucose metabolism and fat distribution. Glucose metabolism was assessed by oral glucose tolerance test and hyperinsulinemic-euglycemic clamp. Endogenous glucose production (EGP), gluconeogenesis, whole-body lipolysis, and de novo lipogenesis (DNL) were measured with stable isotope infusions. Body composition was assessed by dual-energy x-ray absorptiometry and abdominal computed tomography scan.

Results: Glucose tolerance improved and insulin-mediated glucose uptake increased significantly during treatment. EGP increased under fasting conditions, and suppression of EGP by insulin was blunted. Fasting triglycerides decreased significantly in association with a decrease in hepatic DNL. Lean body mass increased and total body fat decreased, whereas visceral adipose tissue did not change.

Conclusions: Treatment with IGF-I/IGFBP-3 improved whole-body glucose uptake and glucose tolerance, while increasing hepatic glucose production. Fasting triglycerides improved, reflecting decreased DNL, and visceral adiposity was unchanged.

IGF-I/IGFBP-3 treatment improved glucose tolerance and peripheral insulin sensitivity while increasing hepatic glucose production in HIV-infected men with insulin resistance.

Patients with HIV infection on antiretroviral (ARV) treatment are at increased risk for metabolic and morphological disturbances including visceral fat accumulation and abnormal glucose metabolism (1). Disturbances in body fat distribution may adversely impact quality of life and increase risk of cardiovascular disease (2).

Experimental therapies that have been used to ameliorate these disturbances include GH and GH-releasing factor. GH reduced visceral adipose tissue but worsened glucose metabolism (3,4). Because many patients with HIV infection on ARVs are insulin resistant, GH may lead to development of frank diabetes. GH-releasing factor reportedly did not worsen glucose metabolism, but neither did it improve insulin resistance (5) .

IGF-I mediates many of the anabolic effects of GH (6) but, unlike GH, improved glucose metabolism in patients with type 1 and type 2 diabetes or GH deficiency (7). IGF-I treatment also decreased fat and increased lean mass in patients with GH insensitivity syndrome while increasing lipolysis (8). Due to its short half-life, IGF-I normally requires continuous infusion or multiple daily injections. In this study, we used a novel preparation in which IGF-I is complexed to its major binding protein, IGF-binding protein-3 (IGFBP-3), to prolong its half-life (9). We performed a pilot, open-label study to determine whether 3 months of treatment with IGF-I/IGFBP-3 could improve glucose metabolism and reduce visceral fat in patients with HIV.

Subjects and Methods

Subjects and study design

Thirteen men with excess central fat (waist circumference >100 cm and waist-to-hip ratio >0.95) and insulin resistance [homeostasis model assessment of insulin resistance (HOMA-IR) (10) score >2.77] were enrolled. Subjects were required to be on a stable ARV regimen for at least 6 months and not on antidiabetic agents or other treatments that could affect fat distribution or glucose metabolism. The Committee on Human Research of the University of California, San Francisco, approved the study. Informed consent was obtained from each subject.

Subjects were admitted to the Clinical Research Center (CRC) at San Francisco General Hospital for 5 d, consumed a constant metabolic diet, and underwent metabolic assessments including oral glucose tolerance testing, body composition measurements, and hyperinsulinemic-euglycemic clamp with stable isotope tracer infusions. After baseline testing, subjects started treatment with IGF-I/IGFBP-3 (mecasermin rinfate, iPLEX; Insmed, Inc., Richmond, VA). Ten subjects received 0.5 mg/kg · d sc and three subjects 1.0 mg/kg · d. The dose of IGF-I/IGFBP-3 was increased in the last three subjects after we found no important safety issues and little change in visceral adiposity with the lower dose. After 3 months of treatment, the inpatient assessments were repeated.

Glucose metabolism

For the oral glucose tolerance testing, subjects drank 75 g of glucose after fasting overnight (3); the area under the curve (AUC) for glucose, insulin, and HOMA-IR (10) was calculated. A 3-h hyperinsulinemic-euglycemic clamp was performed as reported previously (3).

Stable isotope tracer methods

Endogenous glucose production (EGP), gluconeogenesis (GNG), and whole-body lipolysis [rate of appearance of glycerol (Ra glycerol)] were measured under fasting and hyperinsulinemic conditions and hepatic de novo lipogenesis (DNL) under fasting and fed conditions using primed infusions of sodium [1-¹³C]acetate (0.34–0.40 g/h), [U-¹³C]glucose (1.2 mg/kg · h), and [2-¹³C]glycerol (15 mg/kg lean body mass/h) as described previously (11).

Body composition

Total and regional fat and lean mass were measured by dual-energy x-ray absorptiometry (DEXA) (Lunar DPX, Madison, WI) (3). Visceral and sc adipose tissue (VAT and SAT) was measured with a helical computed tomography (CT) scanner (General Electrics Medical Systems, Milwaukee, WI) using a single-slice at L4–L5 (12). Intramyocellular lipid content was measured in the soleus and tibialis anterior muscles as described previously (13).

Other laboratory measurements

Fasting total cholesterol and high-density lipoprotein (HDL)-cholesterol were measured using reagents from Fisher Scientific (Pittsburgh, PA), triglycerides (TG) from Sigma Chemical Co. (St. Louis, MO), and free fatty acids using Wako NEFA kit (Wako Chemicals, Richmond, VA). Low-density lipoprotein cholesterol was calculated (14). The San Francisco General Hospital CRC Core Laboratory measured serum insulin by RIA (Linco Research Inc., St. Charles, MO) and glucose by the glucose oxidase method (YSI STAT2300, Yellow Springs, OH). Fasting morning plasma cortisol was measured by competitive immunoassay (Siemens Centaur, Deerfield, IL). Free and total IGF-I and IGFBP-3 concentrations were measured by Insmed, Inc., using ELISA (Diagnostic Systems Laboratories, Webster, TX). Serum IGFBP-1 and IGFBP-2 were measured by RIA and GH by immunochemiluminometric assay at Esoterix Laboratories.

Statistical methods

Analyses were performed using the STATA version 9.0 software. Data are mean ± sem. Student’s paired t test was used to compare changes from baseline to month 3 using two-tailed P values.

Results

Patients and clinical course

Mean age was 53 ± 2.6 yr, duration of HIV infection 16 ± 1.7 yr, and CD4 count 460 ± 54 cells/μl (Supplemental Table 1, published on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org). All subjects had viral loads under 2000 copies/ml. ARV regimens consisted of at least two nucleoside reverse transcriptase inhibitors with at least one protease inhibitor or nonnucleoside reverse transcriptase inhibitor; one subject was also on a fusion inhibitor. Five subjects were taking medications for hyperlipidemia, and four were on replacement testosterone.

Of the 13 men who received at least one dose of IGF-I/IGFBP-3, one had an allergic reaction after the first dose and was discontinued from the study. Another withdrew at d 23 for an injection site reaction. A third, who had missed at least 20% of study doses, was considered unevaluable. Of the 10 evaluable subjects, seven were treated with 0.5 mg/kg · d of IGF-I/IGFBP-3 and three with 1.0 mg/kg · d.

Hormones and binding proteins

Total IGF-I levels increased in all evaluable subjects (P < 0.001; Fig. 1A), as did free IGF-I (P < 0.001; Fig. 1B). IGFBP-3 levels did not change significantly (Fig. 1C and Table 1). IGFBP-1 levels also did not change significantly, whereas IGFBP-2 increased by 78% (P = 0.007; Table 1). Cortisol levels increased by 38%, but this change did not reach significance (P = 0.09; Table 1). GH was undetectable in five subjects at baseline and eight subjects at month 3; in all subjects with detectable values at baseline, GH levels decreased at month 3 (Table 1).

Figure 1.

Effects of IGF-1/IGFBP-3 treatment on total IGF-I (164 ±17 vs. 594 ± 61 ng/ml) (A), free IGF-I (0.67 ± 0.17 vs. 4.35 ± 0.46 ng/ml) (B), IGFBP-3 (3865 ± 336 vs. 4082 ± 341 ng/ml) (C), and insulin-mediated glucose uptake (3.60 ± 0.46 vs. 5.86 ± 0.86 mg/kg · min · μIU/ml) (D). Individual results and group mean ± sem are shown. *, P < 0.001; **, P < 0.05. ○, 0.5 mg/kg · d dose; •, 1.0 mg/kg · d dose.

Table 1.

Effects of IGF-1/IGFBP-3 on glucose metabolism, lipid metabolism, body composition, and hormones

	Baseline	Month 3	Change	P value
Oral glucose tolerance test (n = 10)
Glucose (mg/dl) t = 0 min	105 ± 3	98 ± 2	−8 ± 2	0.005
Glucose (mg/dl) t = 120 min	166 ± 11	134 ± 13	−32 ± 9	0.008
Glucose AUC	481 ± 24	433 ± 23	−48 ± 20	0.04
Insulin (μIU/ml) t = 0 min	21.0 ± 2.6	15.1 ± 1.9	−5.9 ± 2.8	0.06
Insulin (μIU/ml) t = 120 min	160.5 ± 37.9	68.6 ± 14.9	−91.9 ± 32.7	0.02
Insulin AUC	390.8 ± 73.5	233.2 ± 38.8	−157.6 ± 58.4	0.03
HOMA-IR	5.45 ± 0.69	3.69 ± 0.54	−1.76 ± 0.73	0.04
Serum lipids (n = 10)
Total cholesterol (mg/dl)	165 ± 11	153 ± 12	−11.9 ± 6.3	0.09
TG (mg/dl)	228 ± 37	175 ± 25	−52.6 ± 23.1	0.049
HDL-cholesterol (mg/dl)	32 ± 3.4	32 ± 3.5	−0.5 ± 2.4	0.85
LDL-cholesterol (mg/dl)	87 ± 10	86 ± 11	−0.93 ± 5.1	0.85
Non-HDL-cholesterol (mg/dl)	133 ± 11	122 ± 11	−11.5 ± 5.4	0.06
Fasting FFA (mmol/liter)	0.48 ± 0.07	0.42 ± 0.05	−0.05 ± 0.03	0.14
DEXA scan (n = 10)
Weight (kg)	85 ± 2.6	84 ± 2.7	−0.9 ± 0.8	0.28
Lean body mass (kg)	57.1 ± 2.7	58.1 ± 2.7	1.0 ± 0.9	0.02
Total body fat (kg)	24.1 ± 2.1	22.4 ± 2.1	−1.7 ± 0.7	0.045
Truncal fat (kg)	16.3 ± 1.0	15.3 ± 1.1	−1.1 ± 0.4	0.04
Appendicular fat (kg)	6.7 ± 1.1	6.2 ± 1.1	−0.5 ± 0.3	0.20
CT scan (n = 10)
VAT (cm2)	327 ± 30	323 ± 35	−4 ± 19	0.80
SAT (cm2)	144 ± 30	138 ± 27	−3 ± 4	0.30
EGP (n = 9)
Fasting EGP (mg/kg · min)	2.03 ± 0.09	2.21 ± 0.09	0.18 ± 0.05	0.006
Hyperinsulinemic EGP (mg/kg · min)	0.35 ± 0.05	0.52 ± 0.06	0.17 ± 0.05	0.01
EGP suppression by insulin (%)	83.2 ± 2.1	76.6 ± 2.8	−6.6 ± 2.7	0.04
GNG (n = 9)
Fasting GNG (mg/kg/min)	0.53 ± 0.04	0.65 ± 0.01	0.12 ± 0.04	0.01
Hyperinsulinemic GNG (mg/kg/min)	0.03 ± 0.01	0.04 ± 0.01	0.01 ± 0.01	0.29
GNG suppression by insulin (%)	94.0 ± 1.1	93.6 ± 0.9	−0.4 ± 0.9	0.69
Glycogenolysisa (n = 9)
Fasting (mg/kg · min)	1.50 ± 0.06	1.56 ± 0.09	0.06 ± 0.06	0.33
Hyperinsulinemic (mg/kg · min)	0.32 ± 0.05	0.48 ± 0.05	0.16 ± 0.05	0.01
Suppression by insulin (%)	79.4 ± 2.6	68.6 ± 4.3	−10.7 ± 4.3	0.04
De novo lipogenesis (n = 10)
Fasting DNL (%)	8.64 ± 1.34	6.07 ± 1.05	−2.57 ± 1.2	0.06
Postprandial DNL (%)	12.68 ± 2.12	9.83 ± 1.98	−2.85 ± 1.7	0.14
IGFBPs (n = 10)
IGFBP-1 (ng/mL)	8.7 ± 1.8	9.0 ± 2.1	0.3 ± 1.9	0.88
IGFBP-2 (ng/mL)	409 ± 61	727 ± 129	318 ± 91	0.007
IGFBP-3 (ng/mL)	3865 ± 336	4082 ± 341	217 ± 139	0.15
Other hormones
Cortisol (μg/ml) (n = 10)	9.7 ± 1.9	13.4 ± 2.6	3.7 ± 2.0	0.09
GH (ng/ml)b	0.23 ± 0.96	0.075 ± 0.018	−0.16 ± 0.09	N/A

Data are mean ± sem, and P values are by Student’s t test. Values in bold are statistically significant (P < 0.05). The mean and sem values were calculated by setting the undetectable values at 0.05 ng/ml. Statistical analysis was not performed due to the few subjects with detectable values (n = 2) at month 3. LDL, Low-density lipoprotein. 

^aGlycogenolysis is a calculated value (EGP − GNG). 

^bGH levels were undetectable (i.e. <0.05 ng/ml) in 50% of subjects at baseline and 80% of subjects at month 3. 

Glucose metabolism

Fasting glucose, HOMA-IR, and 2-h glucose and insulin decreased significantly, as did glucose and insulin AUC (Table 1). Insulin-stimulated glucose uptake (M/I) increased by 63% (P = 0.02; Fig. 1D). Fasting EGP increased by 10% (P = 0.006), and suppression of EGP by insulin was blunted during treatment (P = 0.04; Table 1). A 23% increase in fasting GNG (P = 0.01) explained most of the increase in fasting EGP.

Lipid metabolism

Fasting TG levels decreased by 23% (P = 0.049), whereas HDL-cholesterol was unchanged (Table 1). Both total and non-HDL-cholesterol levels tended to decrease. Fasting free fatty acids and whole-body lipolysis, measured both fasting and during hyperinsulinemia, did not change significantly. Fasting and postprandial rates of fractional DNL decreased by 30% and 20%, respectively (Table 1), but neither achieved statistical significance.

Body composition

Lean body mass increased and total body and truncal fat decreased, whereas changes in limb fat were minor and not significant (Table 1). VAT and SAT did not change significantly; there was no apparent difference between the two dosing levels (Table 1). There was no significant change in the intramyocellular lipid content/water ratio in either the soleus (5.1 ± 1.3 vs. 4.5 ± 1.1%, P = 0.55) or tibialis anterior (0.5 ± 0.2 vs. 0.7 ± 0.1%, P = 0.37).

Other laboratory data

Standard chemistries and liver function tests were not altered during treatment. There was a mild, nonsignificant decrease in hemoglobin (15.2 ± 0.3 vs. 14.7 ± 0.4 g/dl, P = 0.15) and white blood cell count (6.5 ± 0.6 vs. 5.5 ± 0.4 × 10³ cells/μl, P = 0.15), and a statistically significant decrease in platelet count (266 ± 18 vs. 237 ± 19 × 10³ cells/μl, P = 0.03). Total lymphocyte count decreased (2.0 ± 0.1 vs. 1.7 ± 0.1 × 10³ cells/μl, P = 0.03), as did absolute CD4 count (565 ± 61 vs. 516 ± 67 cells/μl, P = 0.21).

Safety parameters

Apart from the aforementioned two subjects who were unable to complete the study due to allergic reactions, side effects were minimal. One subject developed mild arthralgias and lower-extremity edema that improved spontaneously. A subject with a preexisting papule on his philtrum underwent a biopsy of the lesion while on study and was diagnosed with local squamous cell carcinoma. There were no instances of hypoglycemia.

Discussion

In this pilot study, 3 months of IGF-I/IGFBP-3 treatment was associated with significant improvements in glucose tolerance and peripheral insulin sensitivity. These improvements were observed in subjects who received both the lower (0.5 mg/kg · d) and higher (1.0 mg/kg · d) doses of IGF-I/IGFBP-3 and are consistent with results of other studies using IGF-I/IGFBP-3 (15). However, the improvement in extrahepatic insulin sensitivity was accompanied by worsening hepatic glucose control as evidenced by increases in EGP and GNG. In diabetics, 7 d of treatment with IGF-I alone decreased EGP (16), as did 2 d of treatment with IGF-I/IGFBP-3 (15). However, in our study, which had a longer period of treatment, administration of IGF-I/IGFBP-3 was associated with less suppression of EGP and GNG during the clamp. Increased GNG may be a compensation for the decrease in glucose levels; it may also be a consequence of increased cortisol (17). We observed a 38% increase in plasma cortisol levels; however, there was no correlation with the increase in GNG or EGP. GH decreased in all five subjects with detectable levels, which may help explain the observed increase in cortisol levels. IGFBP-1, which may reflect portal insulin levels (18), was not significantly changed. Finally, IGF-I/IGFBP-3 may alter hepatic substrate fluxes, resulting in increased substrate channeling into the gluconeogenic pathway, as discussed further below.

As with GH, IGF-I/IGFBP-3 tended to improve lipid metabolism, including a significant decrease in fasting TG levels and fractional hepatic DNL. IGF-I/IGFBP-3 also decreased truncal and total body fat while increasing lean body mass. However, unlike GH, IGF-I/IGFBP-3 did not decrease VAT. DEXA and CT measure different fat compartments, so the decrease in truncal fat could occur because sc fat was lost in a region of the trunk not adequately captured in a single-slice CT scan, such as upper trunk fat, which is independently associated with insulin resistance (19). A decrease in upper trunk fat could contribute to the changes in metabolism seen here.

IGF-I/IGFBP-3 decreased fasting DNL by 30% (P = 0.06). We previously observed a comparable decrease in DNL and concomitant increase of GNG and EGP after GH treatment (11). These simultaneous changes in hepatic fluxes are consistent with the hypothesis that during treatment with IGF-I/IGFBP-3, fewer gluconeogenic substrates are channeled into hepatic DNL, leading to a modest surge into the gluconeogenic pathway and consequent increase in glucose production (11) as has been observed in other hyperinsulinemic states (20,21). In our study, the reduction of hepatic DNL is consistent with the reduction of fasting insulin.

Our detailed metabolic studies in HIV-infected patients with central fat accumulation have shown a number of similarities and differences between the effects of GH and IGF-I/IGFBP-3 (3,11). Within the liver, both IGF-I/IGFBP-3 and GH treatment increased glucose production and GNG while simultaneously decreasing DNL. Unlike GH, IGF-I/IGFBP-3 significantly improved insulin-mediated glucose uptake in the periphery. GH increased lipolysis and reduced visceral fat, whereas IGF-I/IGFBP-3 treatment did not appreciably alter VAT or lipolysis.

In conclusion, 3 months of treatment with IGF-I/ IGFBP-3 in HIV-infected patients with excess visceral adiposity and insulin resistance improved glucose tolerance and peripheral insulin sensitivity. Interestingly, EGP increased, whereas hepatic DNL and serum TG decreased. Further studies are needed to clarify the effects of IGF-I/IGFBP-3 on hepatic glucose and lipid metabolism.