Endocrine Journal 2006, 53 (5), 639–645 Long-term Effects of Recombinant Human Insulin-like Growth Factor I Treatment on Glucose and Lipid Metabolism and the Growth of a Patient with Congenital Generalized Lipodystrophy MARI SATOH, ATSUKO YOSHIZAWA, MASAROU TAKESUE, TSUTOMU SAJI AND SUSUMU YOKOYA* First Department of Pediatrics, Toho University School of medicine, Tokyo 143-8541, Japan *Department of Pediatrics, Toranomon Hospital, Tokyo 105-8470, Japan Abstract. Congenital generalized lipodystrophy (CGL) is a disease characterized by generalized lack of body fat, insulin resistance, hypertriglyceridemia, and fatty liver. We studied the long-term effects of recombinant human insulin-like growth factor I (rhIGF-I) treatment on glucose and lipid metabolism and the growth in a patient with CGL. During rhIGF- I treatment, the serum triglyceride level was maintained almost within the normal range, and the plasma glycosylated hemoglobin A1c (HbA1c) levels were maintained under 8.0% (5.8%–7.9%). Thus, rhIGF-I treatment was effective in lowering glucose and triglyceride levels over the long-term in a CGL patient. However, it was difficult to suppress the patient’s voracious appetite. Although serum total IGF-I levels were extremely high (1000–1700 ng/ml), growth was not accelerated after the start of rhIGF-I treatment, likely because of normal IGF binding protein 3 (IGFBP-3) levels. During rhIGF-I treatment, the patient developed a recurrence of mild hypertrophic cardiomyopathy and a mild elevation of intraocular pressure. Key words: Lipodystrophy, IGF-I, Diabetes mellitus, Hyperlipidemia, Growth (Endocrine Journal 53: 639–645, 2006) INSULIN-like growth factor I (IGF-I) has insulin-like Patient History effects on glucose metabolism. Therefore, recombinant human IGF-I (rhIGF-I) treatment has been reported to The patient was a 15-year-old female who had been be effective in lowering plasma glucose concentration born to non-consanguineous parents after an unevent- in patients with insulin resistance syndrome, including ful 40 week gestational period. At birth, her height was leprechaunism, type A insulin resistance, and congen- 50 cm (+0.58 SD), and body weight was 3112 g (+0.03 ital generalized lipodystrophy (CGL) [1]. However, SD). Her father’s height was 177 cm (+1.18 SD), her there are few reports dealing with the long-term effects mother’s height was 153 cm (–0.98 SD), and the pa- of rhIGF-I treatment on glucose and lipid metabolism tient’s target height was 158.5 cm (+0.12 SD). She and the growth of patients with CGL. In this study, we was referred to our hospital because of failure to thrive investigated the effects of 9 years of treatment with and hirsutism at the age of 4 months. At that time, her rhIGF-I in a female patient with CGL. height was 66.3 cm (+1.68 SD), her body weight was 5716 g (–1.11 SD), and the kaup score was 13.0 (nor- mal range, 15.0–18.0). She had the characteristic phe- notypic findings of CGL, including a generalized lack Received: January 20, 2006 Accepted: June 13, 2006 of body fat, muscle hypertrophy, hirsutism, and clitoro- Correspondence to: Mari SATOH, M.D., First Department of megaly (Fig. 1). Hypertriglyceridemia and a low se- Pediatrics, Toho University School of Medicine, 6-11-1 Oomori- rum high-density lipoprotein (HDL) cholesterol level nishi, Oota-ku, Tokyo 143-8541, Japan were noted (Table 1), and fatty liver was detected by 640 SATOH et al. Fig. 1. Phenotypic characteristics at 4 months of age (a, b) and 6 years of age (c). The patient shows characteristic phenotypes, including generalized lack of the body fat, muscle hypertrophy, and hirsutism. Table 1. Laboratory findings before the start of rhIGF-I treatment 4 months 6years Age 6 months 1 year 4 years (diagnosis) (start of rhIGF-I) Total-cholesterol (mg/dl) 243 247 158 164 166 HDL-cholesterol (mg/dl) 20 18 27 ND 37 Triglyceride (mg/dl) 1445 779 237 187 48 Free fatty acid (mEq/l) 0.89 1.05 0.79 ND 0.29 HOMA-R 1.54 9.26 9.48 9.82 1.76 Insulinogenic Index 0.55 3.03 0.64 ND 0.29 HbA1c (%) ND 4.1 4.2 5.1 6.1 Urinary C-peptide (µg/day) ND ND ND 133 25 Minimum glucose level (mg/dl) in insulin tolerance test O.1 U/kg, insulin ND 80 72 65 43 0.5 U/kg, insulin 25 48 52 32 18 IGF-I (ng/ml) ND ND ND 230 320 HDL, high-density lipoprotein; HOMA-R, homeostasis model assessment; HbA1c, glycosylated hemoglobin A1c; ND, not detected; IGF-I, insulin-like growth factor I Blood samples were obtained after an overnight fast. abdominal ultrasonography. Therefore, the patient was at the age of 1 year, glucose intolerance was recog- diagnosed as having CGL. Since failure to thrive was nized, and the patient was diagnosed as having diabetes thought to be due to acute cardiac failure caused by hyper- mellitus at the age of 4 years. Therefore, her calorie in- trophic cardiomyopathy, a diuretic (furosemide) and a take was restricted to 80% of the calorie intake appro- β-blocker (propranolol hydrochloride) were started. priate for her chronological age. Since the patient’s The patient’s hypertriglyceridemia and fatty liver cardiac function and hypertrophic cardiomyopathy had improved gradually with only a low fat diet. However, gradually improved, furosemide and propranolol hy- IGF-I THERAPY IN A PATIENT WITH LIPODYSTROPHY 641 Table 2. Oral glucose tolerance test (1.75 g/kg) Time (min) Age 0 306090120 4 months Glucose (mg/dl) 78 103 88 106 108 IRI (µU/ml) 8.0 21.8 12.7 23.9 27.0 6 months Glucose (mg/dl) 85 130 136 131 120 IRI (µU/ml) 44.1 180.5 225.7 185.4 137.4 1 year Glucose (mg/dl) 79 105 135 151 145 IRI (µU/ml) 48.6 65.3 106.0 151.1 171.5 4 years Glucose (mg/dl) 114 219 204 203 174 IRI (µU/ml) 34.9 ND 151.3 ND 197.2 6 years Glucose (mg/dl) 115 235 265 204 156 IRI (µU/ml) 6.2 41.1 35.9 21.1 10.2 ND, not detected 10 years, the patient had a recurrence of hypertrophic cardiomyopathy, and an angiotensin converting en- zyme inhibitor (enalapril maleate) was started. The patient’s plasma HbA1c levels increased at the age of 13 years, because the patient could not maintain her low-calorie diet since she was about 11 years old (Table 3). The dosage of rhIGF-I was increased grad- ually and the maximum dose was 0.24 mg/kg/day. She developed localized fatty change of the liver and an elevated intraocular pressure, but tonsillar hypertrophy, renal hypertrophy, retinopathy, and polycystic ovaries were not present (Table 4). Fig. 2. Effect of rhIGF-I administration on plasma glycosylated From birth, she showed accelerated growth (Fig. 3). HbA1c and serum IGF-I. Although GH secretion was normal or slightly de- creased (Table 5), serum total IGF-I levels were high (Table 1) before the start of rhIGF-I. Serum total drochloride were stopped at the age of 4 years. At the IGF-I levels were extremely high after the start of age of 6 years, a subcutaneous injection of rhIGF-I rhIGF-I treatment (Table 3). However, the patient’s once a day before the breakfast at a dose of 0.15 mg/kg growth rate did not change after the start of rhIGF-I was started, since the plasma glycosylated hemoglobin treatment. The patient’s bone age was estimated using A1c (HbA1c) level had increased to 6.1%. The results the Tanner-Whitehouse 2 RUS method standardized of oral glucose tolerance tests are shown in Table 2. for Japanese children; accelerated bone maturation was The reason why we did not try the high dose insulin seen both before and after the start of rhIGF-I treat- treatment was that hyperinsulinemia was supposed to ment. Breast budding started at the age of 7 years and worsen hyperlipidemia and hirsutism. 6 months (mean age for Japanese girls, 10 years), and Due to re-evaluation of its effect on glucose metabo- she reached her adult height of 156.0 cm (–0.38 SD) lism, rhIGF-I treatment was discontinued 1 year later. at the age of 13 years and 6 months. Although her Plasma HbA1c increased from 6.0% to 7.9% 4 months menarche was at the age of 13 years (mean age for after the cessation of rhIGF-I treatment; as rhIGF-I Japanese girl, 12 years and 4 months), she had oligo- treatment appeared to be effective in lowering her plas- menorrhea. Although she had good social skills, she ma glucose concentration and decreasing the plasma was a poor student. As a junior high school student, HbA1c level, it was resumed (Fig. 2). At the age of her intellectual abilities were at the primary school level. 642 SATOH et al. Table 3. Laboratory findings during rhIGF-I treatment Age (Years) 7 9 11 13 15 After the treatment (years) 1 3 5 7 9 Total-cholesterol (mg/dl) 195 138 170 146 180 HDL-cholesterol (mg/dl) 39 32 46 44 46 Triglyceride (mg/dl) 183 192 123 67 83 Free fatty acid (mg/dl) 0.09 0.20 0.31 0.09 0.09 Fasting glucose (mg/dl) 148 131 131 157 166 HbA1c (%) 6.0 5.7 5.7 6.9 7.9 Total IGF-I (ng/ml) 1400 1200 1500 1300 1300 IGFBP-3 (µg/ml) ND ND ND 4.39 3.63 HDL, high-density lipoprotein; HbA1c, glycosylated hemoglobin A1c; ND, not detected; IGF-I, insulin-like growth factor I; IGFBP-3, IGF binding protein 3 Blood samples were obtained after 2 or 3 hours after lunch.