Abstract
Context
Familial partial lipodystrophy (FPLD) is most commonly caused by pathogenic variants in LMNA and PPARG. Leptin replacement with metreleptin has largely been studied in the LMNA group.
Objective
To understand the efficacy of metreleptin in PPARG vs LMNA pathogenic variants and investigate predictors of metreleptin responsiveness.
Design
Subgroup analysis of a prospective open-label study of metreleptin in lipodystrophy.
Setting
National Institutes of Health, Bethesda, Maryland.
Participants
Patients with LMNA (n = 22) or PPARG pathogenic variants (n = 7), leptin <12 ng/mL, and diabetes, insulin resistance, or high triglycerides.
Intervention
Metreleptin (0.08 to 0.16 mg/kg) for 12 months.
Outcome
Hemoglobin A1c (HbA1c), lipids, and medication use at baseline and after 12 months.
Results
Baseline characteristics were comparable in patients with PPARG and LMNA: HbA1c, 9.2 ± 2.3 vs 7.8 ± 2.1%; median [25th, 75th percentile] triglycerides, 1377 [278, 5577] vs 332 [198, 562] mg/dL; leptin, 6.3 ± 3.8 vs 5.5 ± 2.5 ng/mL (P > 0.05). After 12 months of metreleptin, HbA1c declined to 7.7 ± 2.4 in PPARG and 7.3 ± 1.7% in LMNA; insulin requirement decreased from 3.8 [2.7, 4.3] to 2.1 [1.6, 3.0] U/kg/d in PPARG and from 1.7 [1.3, 4.4] to 1.2 [1.0, 2.3] U/kg/d in LMNA (P < 0.05). Triglycerides decreased to 293 [148, 406] mg/dL in LMNA (P < 0.05), but changes were not significant in PPARG: 680 [296, 783] mg/dL at 12 months (P = 0.2). Both groups were more likely to experience clinically relevant triglyceride (≥30%) or HbA1c (≥1%) reduction with metreleptin if they had baseline triglycerides ≥500 mg/dL or HbA1c >8%.
Conclusion
Metreleptin resulted in similar metabolic improvements in patients with LMNA and PPARG pathogenic variants. Our findings support the efficacy of metreleptin in patients with the two most common genetic causes of FPLD.
Metreleptin treatment of 12 months led to similar improvements in HbA1c, insulin dose, and triglycerides in patients with familial partial lipodystrophy due to PPARG vs LMNA pathogenic variants.
Lipodystrophy syndromes are a heterogeneous group of disorders characterized by deficient adipose tissue and lack of adipose tissue‒derived hormones, including leptin. Lack of adiposity and deficiency of serum leptin lead to severe metabolic complications including insulin resistance, dyslipidemia, and nonalcoholic fatty liver disease (1). On the basis of the distribution of adipose tissue loss, lipodystrophy is classified into generalized and partial forms and can be further classified by etiology (genetic vs acquired).
Familial partial lipodystrophy (FPLD) is a genetic form of lipodystrophy characterized by deficiency of subcutaneous adipose tissue in the limbs, buttocks, and hips, with preserved or sometimes increased adipose tissue in the head, neck, and trunk. Abnormal adipose tissue distribution is typically noted at the time of puberty. FPLD usually follows an autosomal dominant inheritance pattern. Multiple genetic causes have been identified; the two most common are pathogenic variants in the LMNA gene (lamin A/C) on chromosome 1q22 (FPLD2; also called the Dunnigan variant) and in the PPARG gene (peroxisome proliferator activated receptor gamma; FPLD3) on chromosome 3p25.2 (2‒4). Much rarer genetic causes of FPLD include pathogenic variants in PLIN1, CIDEC, LIPE, AKT2, and PCYT1A (5, 6). FPLD caused by LMNA pathogenic variants is typically associated with more severe subcutaneous adipose tissue loss than FPLD with PPARG pathogenic variants. Consistent with their lower body fat, patients with LMNA pathogenic variants have lower serum leptin levels than patients with PPARG pathogenic variants (7).
Leptin replacement therapy with metreleptin (human recombinant methionyl leptin) has been shown to dramatically improve metabolic complications in patients with the generalized form of lipodystrophy, in whom serum leptin levels are markedly reduced (8). Our group previously demonstrated that leptin treatment is beneficial in patients with partial lipodystrophy (irrespective of genotype) with severe metabolic disease (serum triglyceride levels >500 mg/dL or hemoglobin A1c (HbA1c) >8%) and/or endogenous serum leptin level <4 ng/mL (9). However, most studies of metreleptin in patients with FPLD, including ours, had a preponderance of patients with LMNA pathogenic variants. To date, the effects of metreleptin treatment have been reported in only three patients with PPARG pathogenic variants, with variable efficacy. The primary aim of this study was to assess the response to metreleptin treatment in patients with FPLD caused by PPARG vs LMNA pathogenic variants. The secondary aim was to investigate the predictive value of baseline serum leptin, triglyceride, and blood HbA1c levels in determining responsiveness to metreleptin in these two groups.
Materials and Methods
Study design and subject population
This study was a subgroup analysis of a prospective, one-arm, open-label study of metreleptin in patients with lipodystrophy. Patients were enrolled in multiple clinical protocols, all with similar designs (NCT00005905, NCT00025883, NCT01778556, and NCT02262806). Studies were approved by the institutional review board of the National Institutes of Diabetes and Digestive and Kidney Diseases. Adult patients and legal guardians of patients <18 years of age provided written informed consent; minors provided written assent.
Inclusion criteria for the overall study were a clinical diagnosis of lipodystrophy, serum leptin level <8 ng/mL for males and <12 ng/mL for females, and one or more metabolic derangements related to lipodystrophy, including diabetes defined by American Diabetes Association criteria, insulin resistance (fasting insulin level >30 µU/mL), or serum triglyceride levels >200 mg/dL. Patients with HIV infection, infectious liver disease, alcohol abuse, psychiatric disorders, pregnancy, or anorexigenic medications were excluded. Patients included in the current analysis had pathogenic variants in LMNA or PPARG, received metreleptin treatment for ≥6 months, and were adherent with ≥70% of metreleptin doses.
Metreleptin and concomitant medications
Patients received subcutaneous metreleptin injections, with the dose ranging from 0.08 to 0.16 mg/kg in one or two divided doses. Doses were adjusted for each individual to optimize metabolic control. Insulin doses were decreased as needed to minimize hypoglycemia. Increases in other antidiabetic medications and lipid-lowering therapies were not permitted per protocol during the first 12 months of metreleptin treatment; however, in two patients metformin was increased or added by outside physicians. Mean metreleptin dose during the treatment period was calculated for each patient. Adverse events (AEs) that occurred during the treatment period for the patients included in this analysis were tallied.
Biochemical analyses and procedures
Metabolic parameters were obtained at baseline and at 12 months (range, 6 to 19 months) following metreleptin treatment. Serum leptin was measured by RIA using a commercial kit (Linco Research) before initiation of metreleptin. Serum glucose, HbA1c, insulin, C-peptide, triglycerides, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and hepatic function tests were conducted in the National Institutes of Health Clinical Center laboratory according to standard methodology. Oral glucose tolerance testing was performed with 75 g of glucose with measurements of serum glucose, insulin, and C-peptide at −10, 0, 30, 60, 90, 120, and 180 minutes relative to the glucose load. Areas under the curve for serum glucose, insulin, and C-peptide were calculated using the trapezoidal method. Weight and height were measured at each visit, and body mass index (BMI) was calculated. Body composition was evaluated by dual-energy x-ray absorptiometry (Hologic QDR 4500). Patients met with a nutritionist at baseline and were instructed about a balanced macronutrient diet (∼30% fat, 20% protein, and 50% carbohydrate) with an emphasis on avoidance of simple carbohydrates. However, dietary modification was not required as part of the study, and adherence to the diet was not monitored.
Genetic studies
The LMNA exons, including the splice site regions, were amplified in 11 segments (10) and PPARG exons in seven segments (11) from 50 ng of genomic DNA using the PCR and exon-specific primers pairs (available on request). The resulting PCR products were analyzed in an agarose gel and purified using a PCR purification kit (Qiagen). The purified PCR products were sequenced using dye-terminator chemistry and an ABI 3730xl DNA analyzer. Sequence variants were verified by manually inspecting the chromatograms of both the wild-type and mutated products. Pathogenicity of each variant was assessed according to genotype-phenotype segregation, functional studies, and ClinVar predictions (Table 1).
Table 1.
Disease-Causing Heterozygous Variants in LMNA and PPARG in the Study Participants
| Gene | rs Numbera | Variant | No. of Patients | Pathogenicity Basis | ClinVarb | Reference(s) | |
|---|---|---|---|---|---|---|---|
| cDNA Level | Protein Level | ||||||
| LMNA | rs57920071 | c.1444C>T | p.R482W | 4 | Genotype-phenotype segregation | Pathogenic | (4) |
| LMNA | rs11575937 | c.1445G>A | p.R482Q | 14 | Genotype-phenotype segregation | Pathogenic | (3, 4, 12) |
| LMNA | rs61444459 | c.1622G>C | p.R541P | 1 | Genotype-phenotype segregation | Pathogenic/likely pathogenic | |
| LMNA | rs56657623 | c.1751G>A | p.R584H | 1 | None | Uncertain significance | (13) |
| LMNA | NA | c.1543A>G | p.K515E | 1 | None | Not reported | (14) |
| LMNA | rs267607543 | c.1488+5G>C | p.I497Vfs*20 | 1 | Functional studies | Pathogenic | (15) |
| PPARG | rs72551364 | c.1273C>T | p.R425C | 1 | Genotype-phenotype segregation | Pathogenic | (11, 16) |
| PPARG | rs121909246 | c.580C>T | p.R194W | 1 | Genotype-phenotype segregation | Pathogenic | (17, 18) |
| PPARG | rs121909244 | c.1484C>T | p.P495L | 1 | Functional study | Conflicting interpretations of pathogenicity | (2, 17, 19) |
| PPARG | NA | c.1159C>T | p.P387S | 2 | Functional study; genotype-phenotype segregation | Not reported | (17)c |
| PPARG | NA | c.1313A>C | p.Q438P | 1 | Functional study | Not reported | (17)d |
| PPARG | NA | c.1184A>G | p.K395R | 1 | Genotype-phenotype segregation | Not reported | (17)e |
LMNA: mRNA Accession Number NM_170707.3; Protein Accession Number NP_733821.1; PPARG: mRNA Accession Number NM_015869.4; Protein Accession Number NP_056953.2.
Abbreviation: NA, not available.
Per http://miter.broadinstitute.org/, this variant has 40% probability of causing FPLD3.
Per http://miter.broadinstitute.org/, this variant has 97.1% probability of causing FPLD3.
Per http://miter.broadinstitute.org/, this variant has <0.1% probability of causing FPLD3.
Statistical analysis
Unpaired t tests or Mann-Whitney tests were used to compare baseline characteristics of patients with LMNA pathogenic variants and PPARG pathogenic variants. Changes in metabolic parameters after metreleptin were evaluated in the entire cohort (LMNA plus PPARG) and separately in each genetic group using paired t tests or the Wilcoxon signed rank test on the basis of data distribution. Mixed model analyses were used to assess interactions in metreleptin response by genetic group. Sensitivity analyses were conducted excluding the two patients in whom metformin was increased or added. Patients were defined as “responders” to metreleptin if they experienced either ≥30% reduction in serum triglyceride levels or ≥1% reduction in HbA1c value. The proportion of responders vs nonresponders in each genetic group was compared in patients with baseline serum triglyceride levels above vs below 500 mg/dL and in patients with baseline HbA1c value above vs below 8% using χ2 tests. Values are expressed as mean ± SD for normally distributed variables or median [25th, 75th percentile] for nonnormally distributed variables.
Results
Baseline characteristics
A total of 29 female patients with FPLD were included in the study. Of these, 22 had heterozygous LMNA pathogenic variants and seven had heterozygous PPARG pathogenic variants (Table 1). Eighteen patients had the two most common heterozygous LMNA variants (p.R482W or p.R482Q) associated with FPLD2 (3, 4, 12), whereas one each had heterozygous p.R584H, p.K515E, and p.I497Vfs*20 variants, previously reported to be linked to FPLD2 (13–15); one patient had a novel heterozygous p.R541P variant that is considered “pathogenic or likely pathogenic” according to ClinVar. Among those with heterozygous PPARG variants, one each had a previously reported p.R425C, p.R194W, or p.P495L variant associated with FPLD3 (2, 11, 16–19). Four patients had novel heterozygous PPARG variants; of these, the p.P387S variant (in two patients) and p.Q438P variant (one patient) are predicted to have a 40% and 97.1% probability of causing FPLD3, respectively, but the p.K395R variant has <0.1% probability of causing FPLD3 according to Majithia et al. (17). However, the PPARG p.K395R variant segregated with the phenotype in the pedigree, with both the patient participating in this study and her affected mother harboring it.
Baseline characteristics of patients are shown in Table 2. Age at presentation, BMI, and body fat distribution were similar in the two groups. The mean baseline serum leptin level was 6.3 ± 3.8 ng/mL in the PPARG variant group and 5.5 ± 2.5 ng/mL in the LMNA variant group (P = 0.5). Fasting glucose levels were comparable in the groups, whereas there was a trend toward a higher HbA1c value in patients with PPARG (9.3% ± 2.2%) compared with patients with LMNA pathogenic variants (7.8% ± 2.1%) (P = 0.1). The number of diabetic medications needed to achieve metabolic control was significantly higher in patients with PPARG than in patients with LMNA pathogenic variants (P = 0.02). There was a trend toward higher serum triglyceride levels in patients with PPARG compared with LMNA pathogenic variants: 1377 [278, 5577] mg/dL vs 332 [198, 562] mg/dL (P = 0.06); in addition, use of lipid-lowering medication was higher in the PPARG group (P = 0.04). Other lipid parameters were similar in the groups.
Table 2.
Baseline Characteristics in Female Patients With FPLD Caused by LMNA vs PPARG Pathogenic Variants
| Clinical Parameters | PPARG Variant Group (n = 7) | LMNA Variant Group (n = 22) | P Value |
|---|---|---|---|
| Demographic and Anthropometric Parameters | |||
| Age, y | 30.8 ± 12.1 | 37.3 ± 15.2 | 0.3 |
| Height, cm | 161 ± 9 | 164 ± 8 | 0.4 |
| Weight, kg | 64 ± 13 | 70 ± 13 | 0.3 |
| BMI, kg/m2 | 24.4 ± 2.2 | 25.7 ± 3.4 | 0.3 |
| Systolic blood pressure, mm Hg | 124 ± 14 | 124 ± 16 | 0.9 |
| Diastolic blood pressure, mm Hg | 74 ± 9 | 73 ± 11 | 0.9 |
| Left arm fat, % | 24.5 ± 8.3 | 21.3 ± 5.4 | 0.3 |
| Right arm fat, % | 24.9 ± 7.1 | 21.8 ± 5.5 | 0.3 |
| Trunk fat, % | 27.2 ± 7.0 | 27.8 ± 6.4 | 0.9 |
| Left leg fat, % | 20.0 ± 5.5 | 17.7 ± 3.7 | 0.2 |
| Right leg fat, % | 19.2 ± 4.9 | 18.2 ± 4.1 | 0.6 |
| Subtotal fat, %a | 22.0 [19.5, 26.5] | 26.8 [22.6, 28.2] | 0.1 |
| Head fat, % | 24.4 ± 0.6 | 24.1 ± 1.9 | 0.8 |
| Total fat, % | 24.6 ± 5.6 | 23.8 ± 5.0 | 0.7 |
| Serum leptin, ng/mL | 6.3 ± 3.8 | 5.5 ± 2.5 | 0.5 |
| Glycemic Parameters | |||
|---|---|---|---|
| Fasting glucose, mg/dL | 177 ± 97 | 180 ± 89 | 0.9 |
| Fasting insulin, µU/mLa | 70 [27, 144] | 15 [11, 42] | 0.06 |
| C-peptide, ng/mLa | 3.7 [0.9, 5.2] | 4.1 [3.2, 4.9] | 0.5 |
| Hemoglobin A1c, % | 9.2 ± 2.3 | 7.8 ± 2.1 | 0.1 |
| AUC glucose, mg/dL/190 min | 62,883 ± 21,340 | 47,423 ± 15,610 | 0.05 |
| AUC insulin, µU/mL/190 mina | 9542 [4675, 23,805] | 14,505 [5429, 30,745] | 0.5 |
| AUC C-peptide, ng/mL/190 mina | 1436 [572, 1491] | 1525 [1122, 2823] | 0.045 |
| Metformin dose, mg/d | 2558 ± 685 | 1705 ± 467 | 0.008 |
| Insulin dose, U/kg/da | 3.8 [2.7, 4.3] | 1.7 [1.3, 4.4] | 0.3 |
| Number of diabetic medications | 2.6 ± 0.8 | 1.6 ± 0.9 | 0.02 |
| Lipid Parameters | |||
|---|---|---|---|
| Total cholesterol, mg/dLa | 321 [155, 415] | 205 [153, 243] | 0.4 |
| Triglycerides, mg/dLa | 1377 [278, 5577] | 332 [198, 562] | 0.06 |
| HDL-C, mg/dL | 27 ± 5 | 34 ± 10 | 0.3 |
| LDL-C, mg/dL | 60 ± 14 | 98 ± 31 | 0.06 |
| Number of lipid-lowering medications | 1.9 ± 0.7 | 1.1 ± 0.8 | 0.037 |
| Hepatic Parameters | |||
|---|---|---|---|
| Alanine aminotransferase, U/La | 36 [20, 46] | 29 [23, 47] | 0.8 |
| Aspartate aminotransferase, U/La | 25 [18, 37] | 25 [19, 30] | 0.8 |
Data are presented as mean ± SD. P < 0.05 is considered statistically significant.
Abbreviations: AUC, area under the curve; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.
Represents data as median [25th percentile, 75th percentile].
Response to metreleptin
The median duration of metreleptin therapy was 13 months (range, 11 to 15 months) in patients with PPARG pathogenic variants and 13 months (range, 6 to 20 months) in patients with LMNA pathogenic variants (P = 0.8). Mean metreleptin doses during the treatment period were similar between groups, equal to 0.10 ± 0.02 mg/kg per day in patients with PPARG pathogenic variants vs 0.12 ± 0.03 mg/kg per day in patients with LMNA pathogenic variants (P = 0.1). The maximum dose of metreleptin administered during the treatment period was 0.14 mg/kg per day in the PPARG pathogenic variants group vs 0.19 mg/kg per day in the LMNA pathogenic variants group.
Outcome measures for all treated subjects are listed in Table 3. Patients with LMNA pathogenic variants had a small but significant decrease in BMI with metreleptin (from 25.7 ± 3.4 to 24.9 ± 3.4 kg/m2; P = 0.01), whereas patients in the PPARG group had a nonsignificant increase in BMI (from 24.4 ± 2.2 to 25.2 ± 2.5 kg/m2; P = 0.3), with a significant difference between groups (P = 0.02).
Table 3.
Outcomes in Patients With FPLD With Pathogenic Variants in PPARG and LMNA Genes Before vs After 12 Months of Metreleptin
| PPARG Variant Group | LMNA Variant Group | |||||||
|---|---|---|---|---|---|---|---|---|
| Clinical Variables | Baseline (n = 7) | 12 Months (n = 7) | P Value | Baseline (n = 22) | 12 Months (n = 22) | P Value | P Value for Combined Group | P Value for Metreleptin × Genetic Interaction |
| Weight, kg | 64 ± 13 | 65 ± 14 | 0.5 | 70 ± 13 | 68 ± 13 | 0.006 | 0.036 | 0.06 |
| BMI, kg/m2 | 24.4 ± 2.2 | 25.2 ± 2.5 | 0.3 | 25.7 ± 3.4 | 24.9 ± 3.4 | 0.01 | 0.2 | 0.02 |
| Glycemic Parameters | ||||||||
|---|---|---|---|---|---|---|---|---|
| Fasting glucose, mg/dL | 177 ± 97 | 158 ± 81 | 0.9 | 180 ± 89 | 124 ± 49 | 0.0001 | 0.005 | 0.3 |
| Fasting insulin, µU/mLa | 70 [27, 144] | 76 [27, 105] | 1.0 | 15 [11, 42] | 18 [9, 56] | 0.6 | 0.7 | 0.9 |
| Fasting C-peptide, ng/mLa | 3.7 [0.9, 5.2] | 3.5 [1.6, 6.1] | 0.7 | 4.1 [3.2, 4.9] | 3.4 [2.2, 4.5] | 0.02 | 0.07 | 0.1 |
| Hemoglobin A1c, % | 9.2 ± 2.3 | 7.7 ± 2.4 | 0.02 | 7.8 ± 2.1 | 7.3 ± 1.7 | 0.01 | 0.0005 | 0.06 |
| Insulin dose, U/kg/da | 3.8 [2.7, 4.3] | 2.0 [1.6, 3.0] | 0.03 | 1.7 [1.3, 4.] | 1.2 [0.9, 2.3] | 0.02 | 0.0002 | 0.8 |
| Number of diabetic medications | 2.6 ± 0.8 | 2.6 ± 0.8 | 1.0 | 1.6 ± 0.9 | 1.4 ± 1.0 | 0.1 | 0.1 | 0.3 |
| Lipid Parameters | ||||||||
|---|---|---|---|---|---|---|---|---|
| Total cholesterol, mg/dLa | 321 [155, 415] | 202 [140, 209] | 0.08 | 205 [153, 243] | 175 [161, 218] | 0.1 | 0.01 | 0.01 |
| Triglycerides, mg/dLa | 1377 [278, 5577] | 680 [296, 783] | 0.2 | 332 [198, 562] | 293 [148, 406] | 0.001 | 0.002 | 0.1 |
| HDL-C, mg/dL | 27 ± 5 | 25 ± 5 | 1.0 | 34 ± 10 | 35 ± 10 | 0.3 | 0.3 | 0.3 |
| LDL-C, mg/dL | 60 ± 14 | 88 ± 58 | 0.4 | 98 ± 31 | 103 ± 35 | 0.7 | 0.5 | 0.9 |
| Number of lipid-lowering medications | 1.9 ± 0.7 | 1.6 ± 0.8 | 0.6 | 1.1 ± 0.8 | 1.1 ± 0.9 | 0.7 | 0.7 | 0.2 |
| Hepatic Parameters | ||||||||
|---|---|---|---|---|---|---|---|---|
| Alanine aminotransferase, U/La | 36 [20, 46] | 29 [23, 52] | 1.0 | 29 [23, 47] | 29 [19, 44] | 0.2 | 0.2 | 0.8 |
| Aspartate aminotransferase, U/La | 25 [18, 37] | 26 [17, 40] | 0.9 | 25 [19, 30] | 21 [18, 27] | 0.2 | 0.5 | 0.8 |
Data are presented as mean and ± SD. P < 0.05 is considered statistically significant.
Abbreviations: AUC, area under the curve; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.
Represents data as median [25th percentile, 75th percentile].
Mixed model analyses did not show any significant differences between patients with PPARG and those with LMNA pathogenic variants regarding the effect of metreleptin on any glycemic parameter. HbA1c value declined in both groups, from 9.2% ± 2.3% to 7.7% ± 2.4% in those with PPARG pathogenic variants (P = 0.02) and from 7.8% ± 2% to 7.3% ± 1.7% in those with LMNA pathogenic variants (P = 0.01). The change in HbA1c value for each individual is illustrated in Fig. 1. Mean fasting glucose level declined from 180 ± 89 to 124 ± 49 mg/dL (P = 0.0001) in patients with LMNA pathogenic variants, whereas the decline was not significant in those with PPARG pathogenic variants (from 177 ± 97 to 158 ± 81 mg/dL; P = 0.9). The fasting insulin level did not change in either group, whereas the C-peptide level decreased in patients with LMNA but not in patients with PPARG pathogenic variants. The number of antidiabetic medications did not change in either group; six of seven patients with PPARG and seven of 22 patients with LMNA pathogenic variants required insulin treatment before metreleptin treatment, and none were able to discontinue insulin after metreleptin. However, the average daily insulin requirement decreased from a median of 3.8 [2.7, 4.3] to 2.1 [1.6, 3.0] U/kg/d in patients with PPARG and from 1.7 [1.3, 4.4] to 1.2 [1.0, 2.3] U/kg/d in patients with LMNA pathogenic variants (P < 0.05). Although no patient in either group had more antidiabetic medications or an increased insulin dose, one patient was started on metformin during metreleptin treatment, and one had an increase in metformin dose by 17% (both patients with LMNA pathogenic variants). After these two patients were excluded from the analysis, the remaining patients with LMNA pathogenic variants continued to demonstrate a significant decline in HbA1c, fasting glucose, and C-peptide levels after metreleptin treatment; however, the reduction in insulin dose was no longer statistically significant.
Figure 1.
Changes in (A) HbA1c and (B) serum triglyceride levels in patients with familial partial lipodystrophy due to pathogenic variants in PPARG and LMNA at month 0 (M0) prior to metreleptin and month 12 (M12) after metreleptin treatment. The mean reduction in HbA1c value was 1.5% in the PPARG variant group (P = 0.02) and 0.6% in the LMNA variant group (P = 0.01). The median reduction in triglyceride level was 2170 mg/dL in the PPARG variant group (P = 0.2) and 560 mg/dL in the LMNA variant group (P = 0.001).
Mixed model analyses showed that the effects of metreleptin on lipid parameters were not statistically different in patients with LMNA vs PPARG pathogenic variants, with the exception of total cholesterol. Total cholesterol reduction in the PPARG group (from 321 [155, 415] to 202 [140, 209] mg/dL; P = 0.08) was greater than in the LMNA group (from 205 [153, 243] to 175 [161, 218] mg/dL; P = 0.1; P = 0.01 between groups). Metreleptin significantly reduced serum triglyceride levels in patients with LMNA pathogenic variants from 332 [198, 562] mg/dL at baseline to 293 [148, 406] mg/dL at 12 months (P = 0.001), whereas the change in triglyceride levels after metreleptin was not significant in patients with PPARG pathogenic variants (1377 [278, 5577] mg/dL at baseline and 680 [296, 783] mg/dL at 12 months; P = 0.2) (Fig. 1). There were no changes in high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, and lipid-lowering medications in either group.
Serum alanine aminotransferase and aspartate aminotransferase values did not change after metreleptin in either group.
Among patients with the LMNA pathogenic variants, five of seven patients (71%) with baseline serum triglyceride levels ≥500 mg/dL were classified as metreleptin responders (i.e., had serum triglyceride reduction of ≥30%), whereas four of 13 patients (31%) with baseline serum triglycerides <500 mg/dL were responders (P = 0.08). Among patients with PPARG pathogenic variants, all four (100%) with baseline serum triglyceride levels ≥500 mg/dL were classified as metreleptin responders, whereas only one of three patients (33%) with baseline triglyceride levels <500 mg/dL was a responder (P = 0.053). Among patients with LMNA pathogenic variants, five of eight patients (63%) with baseline HbA1c value ≥8% were classified as metreleptin responders (i.e., had HbA1c reduction ≥1%), whereas zero of 11 patients (0%) with baseline HbA1c value <8% were responders (P = 0.002). Among patients with PPARG pathogenic variants, two of four patients (50%) with baseline HbA1c value ≥8% were classified as metreleptin responders, and one of three patients (33%) with baseline HbA1c value <8% was a responder (P = 0.66); the one responder with baseline HbA1c value <8% had a baseline HbA1c value of 7.9%. Leptin level above vs below 4 ng/mL was not predictive of metreleptin responsiveness (based on either triglyceride or HbA1c lowering) in either genetic subgroup.
AEs
During 12 months of treatment with metreleptin, 69 AEs occurred in 17 patients (58%), including seven serious adverse events (SAEs) in four patients. All SAEs were judged to be unrelated to metreleptin treatment. AEs that were assessed as related to metreleptin included hypoglycemia (seven events in two patients), fatigue (three events in three patients) and hair loss (one event). The most common AEs were fatigue (10% of patients), hypoglycemia (7%), generalized muscle aches (7%), headache (7%), and insomnia (7%).
Discussion
In this study, we present the largest data set of patients with genetically proven FPLD due to heterozygous pathogenic variants in LMNA and PPARG genes who were treated with metreleptin. As in prior studies, LMNA pathogenic variants were much more common than PPARG pathogenic variants, and only seven patients with PPARG met the inclusion criteria for this study, thus limiting statistical power in this genetic group. Despite the small sample size, we found that metreleptin lowered HbA1c values and insulin doses in both the LMNA and PPARG pathogenic variants cohorts, whereas a significant triglyceride-lowering effect was observed only in patients with LMNA pathogenic variants. Responder analyses supported the idea that patients with both PPARG and LMNA pathogenic variants are more likely to experience clinically relevant triglyceride (≥30%) or HbA1c (≥1%) reductions with metreleptin if they had severe elevations of these parameters at baseline (triglycerides ≥500 mg/dL or HbA1c >8%), although most analyses did not meet thresholds for statistical significance given the small sample size.
At baseline, patients in our cohort with PPARG pathogenic variants required more lipid- and glucose-lowering therapies to achieve a similar degree of metabolic control as those with LMNA pathogenic variants, suggesting that metabolic disease may actually be more severe in PPARG than in LMNA pathogenic variants. Larger sample sizes are needed to confirm this finding.
Because LMNA variants were the most commonly reported abnormalities or the only reported molecular abnormality in previous cohorts, studies investigating metreleptin efficacy in FPLD largely reflect outcomes in patients with LMNA pathogenic variants. Effects of metreleptin were previously reported in only three patients with PPARG pathogenic variants. One patient exhibited substantial improvement in glycemic parameters, including fasting blood glucose level, HbA1c value, and oral glucose tolerance, as well as a significant reduction in the severity of hypertriglyceridemia after 18 months of metreleptin treatment (16, 20). Another study including two patients with PPARG pathogenic variants showed a decrease in HbA1c value after 12 months of metreleptin treatment in both subjects; however, the change in serum triglyceride levels was not consistent among subjects, with one patient having a 30% increase in triglyceride level at the end of 1 year of therapy (20). Our study suggests that patients with PPARG pathogenic variants respond to metreleptin similarly to those with LMNA pathogenic variants, particularly regarding glycemic improvements. These improvements in the PPARG variant cohort were clinically relevant, with a 1.5% reduction in HbA1c value and an almost 50% reduction in insulin dose.
Improved diabetes control after metreleptin results primarily from improved insulin sensitivity (8, 21, 22), with variable reported effects on β cell function (23, 24). In this study, C-peptide levels decreased only in patients with LMNA pathogenic variants, perhaps because of the larger sample size, but reduced insulin doses in both the LMNA and PPARG groups were consistent with improved insulin sensitivity. Our sample size was not large enough to confirm that metreleptin reduced serum triglyceride levels in patients with PPARG pathogenic variants. Given the rarity of this subtype of FPLD, it may be challenging to attain adequately powered study cohorts to verify whether metreleptin improves triglycerides.
The efficacy of metreleptin in partial forms of lipodystrophy appears to correlate with the severity of leptin deficiency and the severity of metabolic disease (6). A previous study from our group demonstrated the efficacy of metreleptin in patients with partial lipodystrophy who had severe metabolic derangements with an HbA1c value >8% or triglyceride level >500 mg/dL at baseline (9). The current analyses support the use of baseline HbA1c and triglyceride levels to predict metreleptin response in patients with both PPARG and LMNA pathogenic variants.
Data regarding the utility of endogenous leptin to predict metabolic improvements with metreleptin have been mixed. The prior study from our group demonstrated beneficial responses to metreleptin in patients with partial lipodystrophy and severe hypoleptinemia (<4 ng/mL). However, another study comparing the response to metreleptin in patients with FPLD due to LMNA pathogenic variants with severe hypoleptinemia (mean, 1.9 ng/mL) vs those with LMNA pathogenic variants with moderate hypoleptinemia (mean, 5.3 ng/mL) did not show between-group differences in triglyceride reduction regardless of the severity of hypoleptinemia; no HbA1c reduction was seen in either group (25). In contrast to prior reports, we did not see any difference in body composition and hence endogenous leptin, in our cohorts of patients with LMNA and PPARG pathogenic variants, and baseline serum leptin levels had no power to predict metreleptin responders in either genetic group. This suggests that baseline metabolic disease is a more important predictor of metreleptin responsiveness than endogenous levels of serum leptin, at least within the range of endogenous leptin levels in patients eligible for inclusion in this study (i.e., <12 ng/mL).
In conclusion, we found that metreleptin use resulted in similar metabolic improvements in patients with FPLD caused by LMNA and PPARG pathogenic variants. The confirmation that metreleptin improved glycemia in patients with PPARG pathogenic variants is particularly timely given the recent European Medical Authority approval of metreleptin for partial lipodystrophy. The findings of our study should affirm to clinicians that patients with either of the major genetic variants causing FPLD and with major elevations in triglyceride or HbA1c level are candidates for metreleptin treatment. Metreleptin was generally well tolerated in this study. Weight loss due to appetite suppression is an expected on-target effect of metreleptin, particularly in patients with generalized lipodystrophy and very low leptin levels (26). Appetite was not measured in the current study, but clinically significant weight loss was not observed. However, it is important to note that SAEs associated with metreleptin have been reported, including the development of neutralizing antibodies to leptin. These neutralizing antibodies are of particular concern in patients with partial lipodystrophy who make meaningful amounts of endogenous leptin, as the antibody may block endogenous leptin as well as exogenous metreleptin. Blockade of endogenous leptin has the potential to lead to hyperphagia and weight gain and resultant worsening of metabolic disease. No neutralizing antibody activity was observed in participants reported in this analysis.
The generalizability of this study was limited by inclusion of only female participants. This is typical for studies of partial lipodystrophy and likely relates to difficulty in diagnosing males phenotypically, as well as more severe metabolic complication in females (27). This study was also limited by small sample size, particularly in the PPARG cohort; however, it is the largest cohort of patients with PPARG pathogenic variants ever treated with metreleptin. Our findings should provide guidance to clinicians considering metreleptin treatment of patients with the two most common genetic causes of FPLD.
Acknowledgments
This work was supported by the intramural research programs of the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Diabetes and Digestive and Kidney Diseases. We wish to acknowledge Carmel Tovar for help preparing the table of pathogenic variants.
Financial Support: This work was supported by the intramural research programs of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (to H.S.) and the National Institute of Diabetes and Digestive and Kidney Diseases (to R.J.B.) and by grants from the National Institutes of Health (R01-DK-105448; to A.G.) and the Southwestern Medical Foundation (to A.G.).
Clinical Trial Information: ClinicalTrials.gov nos. NCT00005905 (registered 12 June 2000), NCT00025883 (registered 29 October 2001), NCT01778556 (registered 29 January 2013), and NCT02262806 (registered 13 October 2014).
Disclosure Summary: A.G. consults for Aegerion Pharmaceuticals and Regeneron and has received grant support from Aegerion Pharmaceuticals, Quintiles, Ionis Pharmaceuticals, and Intercept Pharmaceuticals. R.J.B and E.C. have received scientific writing support from Aegerion Pharmaceuticals. The remaining authors have nothing to disclose.
Glossary
Abbreviations:
- AE
adverse event
- BMI
body mass index
- HbA1c
hemoglobin A1c
- FPLD
familial partial lipodystrophy
- PPARG
peroxisome proliferator activated receptor gamma
- SAE
serious adverse event
References
- 1. Garg A. Acquired and inherited lipodystrophies. N Engl J Med. 2004;350(12):1220–1234. [DOI] [PubMed] [Google Scholar]
- 2. Barroso I, Gurnell M, Crowley VE, Agostini M, Schwabe JW, Soos MA, Maslen GL, Williams TD, Lewis H, Schafer AJ, Chatterjee VK, O’Rahilly S. Dominant negative mutations in human PPARγ associated with severe insulin resistance, diabetes mellitus and hypertension. Nature. 1999;402(6764):880–883. [DOI] [PubMed] [Google Scholar]
- 3. Cao H, Hegele RA. Nuclear lamin A/C R482Q mutation in Canadian kindreds with Dunnigan-type familial partial lipodystrophy. Hum Mol Genet. 2000;9(1):109–112. [DOI] [PubMed] [Google Scholar]
- 4. Shackleton S, Lloyd DJ, Jackson SN, Evans R, Niermeijer MF, Singh BM, Schmidt H, Brabant G, Kumar S, Durrington PN, Gregory S, O’Rahilly S, Trembath RC. LMNA, encoding lamin A/C, is mutated in partial lipodystrophy. Nat Genet. 2000;24(2):153–156. [DOI] [PubMed] [Google Scholar]
- 5. Garg A. Clinical review: Lipodystrophies: genetic and acquired body fat disorders. J Clin Endocrinol Metab. 2011;96(11):3313–3325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Brown RJ, Araujo-Vilar D, Cheung PT, Dunger D, Garg A, Jack M, Mungai L, Oral EA, Patni N, Rother KI, von Schnurbein J, Sorkina E, Stanley T, Vigouroux C, Wabitsch M, Williams R, Yorifuji T. The diagnosis and management of lipodystrophy syndromes: a multi-society practice guideline. J Clin Endocrinol Metab. 2016;101(12):4500–4511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Akinci B, Onay H, Demir T, Savas-Erdeve Ş, Gen R, Simsir IY, Keskin FE, Erturk MS, Uzum AK, Yaylali GF, Ozdemir NK, Atik T, Ozen S, Yurekli BS, Apaydin T, Altay C, Akinci G, Demir L, Comlekci A, Secil M, Oral EA. Clinical presentations, metabolic abnormalities and end-organ complications in patients with familial partial lipodystrophy. Metabolism. 2017;72:109–119. [DOI] [PubMed] [Google Scholar]
- 8. Oral EA, Simha V, Ruiz E, Andewelt A, Premkumar A, Snell P, Wagner AJ, DePaoli AM, Reitman ML, Taylor SI, Gorden P, Garg A. Leptin-replacement therapy for lipodystrophy. N Engl J Med. 2002;346(8):570–578. [DOI] [PubMed] [Google Scholar]
- 9. Diker-Cohen T, Cochran E, Gorden P, Brown RJ. Partial and generalized lipodystrophy: comparison of baseline characteristics and response to metreleptin. J Clin Endocrinol Metab. 2015;100(5):1802–1810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Patni N, Li X, Adams-Huet B, Vasandani C, Gomez-Diaz RA, Garg A. Regional body fat changes and metabolic complications in children with Dunnigan lipodystrophy-causing LMNA variants. J Clin Endocrinol Metab. 2019;104(4):1099–1108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Agarwal AK, Garg A. A novel heterozygous mutation in peroxisome proliferator-activated receptor-gamma gene in a patient with familial partial lipodystrophy. J Clin Endocrinol Metab. 2002;87(1):408–411. [DOI] [PubMed] [Google Scholar]
- 12. Speckman RA, Garg A, Du F, Bennett L, Veile R, Arioglu E, Taylor SI, Lovett M, Bowcock AM. Mutational and haplotype analyses of families with familial partial lipodystrophy (Dunnigan variety) reveal recurrent missense mutations in the globular C-terminal domain of lamin A/C. Am J Hum Genet. 2000;66(4):1192–1198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hegele RA, Cao H, Anderson CM, Hramiak IM. Heterogeneity of nuclear lamin A mutations in Dunnigan-type familial partial lipodystrophy. J Clin Endocrinol Metab. 2000;85(9):3431–3435. [DOI] [PubMed] [Google Scholar]
- 14. Chirico V, Ferraù V, Loddo I, Briuglia S, Amorini M, Salpietro V, Lacquaniti A, Salpietro C, Arrigo T. LMNA gene mutation as a model of cardiometabolic dysfunction: from genetic analysis to treatment response. Diabetes Metab. 2014;40(3):224–228. [DOI] [PubMed] [Google Scholar]
- 15. Morel CF, Thomas MA, Cao H, O’Neil CH, Pickering JG, Foulkes WD, Hegele RA. A LMNA splicing mutation in two sisters with severe Dunnigan-type familial partial lipodystrophy type 2. J Clin Endocrinol Metab. 2006;91(7):2689–2695. [DOI] [PubMed] [Google Scholar]
- 16. Guettier JM, Park JY, Cochran EK, Poitou C, Basdevant A, Meier M, Clément K, Magré J, Gorden P. Leptin therapy for partial lipodystrophy linked to a PPAR-γ mutation. Clin Endocrinol (Oxf). 2008;68(4):547–554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Majithia AR, Tsuda B, Agostini M, Gnanapradeepan K, Rice R, Peloso G, Patel KA, Zhang X, Broekema MF, Patterson N, Duby M, Sharpe T, Kalkhoven E, Rosen ED, Barroso I, Ellard S; UK Monogenic Diabetes Consortium, Kathiresan S; Myocardial Infarction Genetics Consortium, O’Rahilly S; UK Congenital Lipodystrophy Consortium, Chatterjee K, Florez JC, Mikkelsen T, Savage DB, Altshuler D. Prospective functional classification of all possible missense variants in PPARG. Nat Genet.48(12):1570–1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Monajemi H, Zhang L, Li G, Jeninga EH, Cao H, Maas M, Brouwer CB, Kalkhoven E, Stroes E, Hegele RA, Leff T. Familial partial lipodystrophy phenotype resulting from a single-base mutation in deoxyribonucleic acid-binding domain of peroxisome proliferator-activated receptor-γ. J Clin Endocrinol Metab. 2007;92(5):1606–1612. [DOI] [PubMed] [Google Scholar]
- 19. Tsai YS, Kim HJ, Takahashi N, Kim HS, Hagaman JR, Kim JK, Maeda N. Hypertension and abnormal fat distribution but not insulin resistance in mice with P465L PPARγ. J Clin Invest. 2004;114(2):240–249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Chong AY, Lupsa BC, Cochran EK, Gorden P. Efficacy of leptin therapy in the different forms of human lipodystrophy. Diabetologia. 2010;53(1):27–35. [DOI] [PubMed] [Google Scholar]
- 21. Ebihara K, Kusakabe T, Hirata M, Masuzaki H, Miyanaga F, Kobayashi N, Tanaka T, Chusho H, Miyazawa T, Hayashi T, Hosoda K, Ogawa Y, DePaoli AM, Fukushima M, Nakao K. Efficacy and safety of leptin-replacement therapy and possible mechanisms of leptin actions in patients with generalized lipodystrophy. J Clin Endocrinol Metab. 2007;92(2):532–541. [DOI] [PubMed] [Google Scholar]
- 22. Brown RJ, Valencia A, Startzell M, Cochran E, Walter PJ, Garraffo HM, Cai H, Gharib AM, Ouwerkerk R, Courville AB, Bernstein S, Brychta RJ, Chen KY, Walter M, Auh S, Gorden P. Metreleptin-mediated improvements in insulin sensitivity are independent of food intake in humans with lipodystrophy. J Clin Invest. 2018;128(8):3504–3516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Muniyappa R, Brown RJ, Mari A, Joseph J, Warren MA, Cochran EK, Skarulis MC, Gorden P. Effects of leptin replacement therapy on pancreatic β-cell function in patients with lipodystrophy. Diabetes Care. 2014;37(4):1101–1107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Vatier C, Fetita S, Boudou P, Tchankou C, Deville L, Riveline J, Young J, Mathivon L, Travert F, Morin D, Cahen J, Lascols O, Andreelli F, Reznik Y, Mongeois E, Madelaine I, Vantyghem M, Gautier J, Vigouroux C. One-year metreleptin improves insulin secretion in patients with diabetes linked to genetic lipodystrophic syndromes. Diabetes Obes Metab. 2016;18(7):693–697. [DOI] [PubMed] [Google Scholar]
- 25. Simha V, Subramanyam L, Szczepaniak L, Quittner C, Adams-Huet B, Snell P, Garg A. Comparison of efficacy and safety of leptin replacement therapy in moderately and severely hypoleptinemic patients with familial partial lipodystrophy of the Dunnigan variety. J Clin Endocrinol Metab. 2012;97(3):785–792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. McDuffie JR, Riggs PA, Calis KA, Freedman RJ, Oral EA, DePaoli AM, Yanovski JA. Effects of exogenous leptin on satiety and satiation in patients with lipodystrophy and leptin insufficiency. J Clin Endocrinol Metab. 2004;89(9):4258–4263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Garg A. Gender differences in the prevalence of metabolic complications in familial partial lipodystrophy (Dunnigan variety). J Clin Endocrinol Metab. 2000;85(5):1776–1782. [DOI] [PubMed] [Google Scholar]