The clinical problems requiring treatment in lipoatrophy syndromes are diabetes and hypertriglyceridemia. Although several classes of antidiabetic drugs are marketed in the United States, achieving good glycemic control (glycated hemoglobin <7.2%) is difficult for most patients with lipoatrophic diabetes. Apart from thiazolidinediones, 4 groups of pharmaceuticals are available to treat patients with diabetes (Clark, 1999).
Insulin
Because of their extreme insulin resistance, extremely high doses (eg, >1000 U/d) are often necessary in patients with various syndromes of lipoatrophy (Meyer, 1998). The volume of injections is more difficult to tolerate for younger patients. Nevertheless, insulin remains the only approved medication for the treatment of children with diabetes.
Insulin secretagogues
The principal mechanism of action for this class of drugs is to promote insulin secretion by beta cells (Nathan, 1988). Because patients with isolated insulin resistance may not have an independent defect in insulin secretion, sulfonylureas are frequently ineffective in patients with lipoatrophic diabetes.
Metformin
This drug primarily works by inhibiting hepatic glucose output. Metformin may help decrease insulin requirements in some patients with lipoatrophy/lipodystrophy. To this author's knowledge, no systematic trials are investigating the efficacy of this drug in patients with lipoatrophy/lipodystrophy. While pediatric experience with metformin is growing, it is still not approved by the US Food and Drug Administration (FDA) for use in children. Therefore, it must be administered by clinicians who have prior experience with treatment with this drug, ideally in facilities specialized in treating children with unusual forms of diabetes.
Acarbose
This drug interferes with GI absorption of many carbohydrates (Bressler, 1997). While this author is not aware of systematic studies on the efficacy of this drug in patients with lipoatrophy, it has relatively limited efficacy to decrease glucose levels in persons with type 2 diabetes mellitus.
Thiazolidinediones
This class is the newest addition to oral hypoglycemic agents. Troglitazone was the first member to be approved by the FDA. Members of this class are agonists of PPAR-gamma, thereby stimulating adipocyte differentiation (Tontonoz, 1994). In vivo, this class of drugs functions as insulin sensitizers, and they increase insulin-stimulated glucose uptake by the muscle (Inzucchi, 1998; Maggs, 1998). These effects of thiazolidinedione compounds seem ideally suited to treat the problems encountered in lipoatrophy/lipodystrophy syndromes. In addition, troglitazone has exerted favorable effects in an animal model of lipoatrophy (Burant, 1997).
This author undertook a treatment trial with troglitazone in patients with lipoatrophy. The primary endpoints were metabolic control and adipose tissue mass (Arioglu, 1999). Six months of troglitazone therapy led to significant improvement in metabolic control, with a reduction in glycosylated hemoglobin (HbA1c), triglyceride, and free fatty acid levels. In addition, a small but statistically significant increase in body fat was noted. Interestingly, the increase was predominantly in the subcutaneous compartment. Furthermore, body weight did not change significantly and liver size deceased.
Despite the favorable effects of troglitazone, a risk of hepatotoxicity was associated with therapy. The FDA withdrew troglitazone from the US market in March of 2000 because of concerns of hepatotoxicity. Despite the serious risk of hepatotoxicity, patients with lipoatrophy experienced significant clinical benefit from treatment with troglitazone.
The newer members of the class, rosiglitazone and pioglitazone, remain in clinical use. The preclinical studies with these 2 medications suggest that they do not have the same risk of hepatotoxicity. This author is continuing studies with the newer thiazolidinediones to determine whether they will exert similar metabolic effects in patients with lipoatrophy. Because the use of thiazolidinediones requires judicious monitoring for adverse effects, this therapy should be supervised by clinicians who are experienced in administering these medications and only after careful assessment of the risk-to-benefit ratio.
Treatment of dyslipidemia in lipoatrophic diabetes
The dyslipidemia of severe lipoatrophy is difficult to treat with currently available therapeutic interventions. This problem causes significant morbidity and mortality. Patients with triglyceride levels greater than 1000 mg/dL are at risk of developing acute pancreatitis. Recurrent episodes lead to pancreatic insufficiency. Chronically elevated triglyceride levels in the setting of hyperinsulinemia and low high-density lipoprotein (HDL) levels increase the predisposition for coronary artery disease. Severely elevated triglyceride levels may lead to painful cutaneous eruptive xanthomata.
Troglitazone is remarkable in its ability to improve hypertriglyceridemia in addition to improving glycemia. This author prefers to use fibrates and statins as opposed to niacin because of severe insulin resistance. These 2 classes may need to be used in combination, with careful monitoring for adverse effects. Again, the pediatric experience with these medications is extremely limited. Limitation of energy (caloric) intake may be useful for short-term results, but continuing this for long-term treatment is not possible. Medium-chain triglycerides and fish oil have not been tested in a systematic fashion, but they may prove useful. This author has attempted to use intestinal lipase inhibitors in a few severely affected pediatric patients, with some positive results.
Other potential therapies from the laboratory to clinical trials
Two encouraging interventions in mouse models of lipoatrophy are worth mentioning because they suggest novel therapeutic approaches for the treatment of patients with lipoatrophy. One intriguing experiment is transplantation of white fat into transgenic mice with severe lipoatrophy (Gavrilova, Marcus-Samuels, and Graham, 2000). This model exerted very severe metabolic abnormalities and had the least amount of white adipose tissue among the viable animal models of lipoatrophy. Transplantation of fat ameliorated the diabetes in a dose-dependent manner. This is currently a difficult therapy to undertake in humans because of the requirement for long-term immunosuppression and technical difficulties. Nevertheless, this experiment unequivocally establishes that insulin resistance and diabetes in lipoatrophy are caused by the absence of fat tissue.
The reason why replacement of fat reverses the diabetes remains unknown. One hypothesis is that the absence of adipocytes causes an absence of hormones produced by these cells. Therefore, replacement of these hormones may lead to reversal of the metabolic findings. Shimomura et al administered one such factor, leptin, to lipoatrophic mice. These mice exhibited diabetes, hypertriglyceridemia, and hepatic steatosis in addition to severe leptin deficiency. Interestingly, 3 weeks of leptin therapy ameliorated all the metabolic abnormalities, ie, normalizing glycemia and triglyceridemia and decreasing the degree of steatosis in the liver (Shimomura, 1999).
The results with leptin administration in another model of lipoatrophy are less dramatic, but they still demonstrate a significant effect on plasma insulin levels (Gavrilova, Marcus-Samuels, and Leon, 2000). These results from animal models warranted a clinical trial, which has been underway for approximately 1 year. The results from this study will be available soon. The latter study is likely to inspire a similar set of studies with a replacement approach of adipocyte-specific hormones, opening an era of rediscovery of the adipocyte as an endocrine organ.
Table 1. Genetic Syndromes of Lipoatrophy
Syndrome Lipoatrophy Gene/Locus Inheritance OMIM*
Primary Lipoatrophy Syndromes
Congenital generalized lipoatrophy Generalized 9q34 AR 269700
(Seip-Berardinelli) See text for details (Congenital generalized lipoatrophy)
Gene unknown† - -
Dunnigan syndrome Familial partial
See text for details (Familial partial lipodystrophy)
1q21-22
Lamin A/C AD 151660
Others Numerous distributions Unknown AD/AR N/A
Complex Syndrome Associated with Lipoatrophy
Mandibuloacral dysplasia Congenital, partial
Involves extremities Unknown AR 248370
Werner syndrome Congenital, partial Involves extremities 8p12
Werner helicase AR 277700
Cockayne syndrome Congenital, partial
Involves extremities 5
CSA AR 216400
Carbohydrate-deficient
glycoprotein syndrome Transient, partial
Buttocks 16p.13.3
PMM 1 and 2‡ AR 212065
SHORT syndrome Generalized, congenital Unknown AR 269880
AREDYLD§ syndrome Generalized, congenital Unknown - 207780
*Online Mendelian Inheritance of Man; database providing information about genetic syndromes
†Evidence for genetic heterogeneity
‡Phosphomannomutase 1 and 2
§Not clear if this is a variation of Berardinelli-Seip syndrome
Table 2. Characteristics of Patients With Generalized and Partial Lipoatrophy
Generalized Partial
General characteristics
Male/female 2/8 4/23
Age range, y 6-19 26-65
Genetic/other 4/6 10/17
Clinical characteristics
Diabetes/IGT*/IR† 9/1/0 15/5/3
Acanthosis nigricans, % 70 56
Hyperandrogenism, % 50 70
Central hypogonadism, % 50 0
NASH, % 70 33
Laboratorycharacteristics Reference Range
HbA1c, % 9.9 ±3.5 7.2 ±0.8 <6.2
Triglycerides‡, mg/dL 2015 ±1463 426 ±164 35-155
Free fatty acids, µmol/L 926 ±156 693 ±115 50-550
HDL cholesterol§, mg/dL 21.4 ±10.3 22.7 ±11.9 35-85
Leptin||, ng/mL 1.8 ±1.0 10.5 ±6.3 Variable
Body fat by DEXA, %
7.7 ±2.6 23.8 ±2.5 Variable
*Impaired glucose tolerance
†Insulin resistance
‡Conversion factor to mmol/L: 0.01129X
§Conversion factor to mmol/L: 0.02586X
||Conversion factor to nmol/L: 0.08X
Dual-energy x-ray absorptiometry