Clinical effects of sodium glucose transporter type 2 inhibitors in patients with partial lipodystrophy

Abstract

Objective:

Severe insulin resistance syndromes such as lipodystrophy lead to diabetes that is challenging to control. This study explores safety and efficacy of sodium glucose cotransporter 2 inhibitors (SGLT2i) in a series of 12 patients with severe insulin resistance due to partial lipodystrophy.

Research Design and Methods:

Retrospective chart review of safety (N=22) and efficacy (N=12) of SGLT2i in patients with partial lipodystrophy at our institution. Efficacy outcomes included HbA1c, insulin dose, fasting plasma glucose, C-peptide, lipid profile, 24-hour urinary glucose excretion, estimated glomerular filtration rate (eGFR), and blood pressure before and after 12 months of SGLT2i treatment.

Results:

HbA1c decreased after SGLT2i (baseline 9.2±2.0% [77.6±21.2 mmol/mol]; 12 months 8.4±1.8% [67.9±19.6 mmol/mol]; p=0.028). Significant reductions were also noted in systolic (p=0.011) and diastolic blood pressure (p=0.013). There was a trend toward decreased C-peptide (P=0.071). Fasting plasma glucose, lipids, and eGFR remained unchanged. Adverse effects included extremity pain, hypoglycemia, diabetic ketoacidosis (in a patient who was non-adherent to insulin), pancreatitis (in a patient with prior pancreatitis), and fungal infections.

Conclusions:

SGLT2i reduced HbA1c in patients with partial lipodystrophy, with a similar safety profile compared to type 2 diabetes. After individual consideration of risks and benefits, SGLT2i may be considered as part of the treatment armamentarium for these rare forms of diabetes, but larger trials are needed to confirm these findings.

CLINICAL SUMMARY :

This research explores the safety and efficacy of sodium glucose cotransporter 2 inhibitors in patients with partial lipodystrophy. With limited therapeutic options for these rare forms of diabetes, our findings that there was a reduction in HbA1c in patients with partial lipodystrophy treated with SGLT2i for 12 months. This holds true both for unadjusted HbA1c as well as HbA1c adjusted for change in insulin dose. SGLT2i should be considered as a part of their treatment armamentarium after careful consideration of risks and benefits for these drugs.

INTRODUCTION :

Insulin resistance is broadly defined as suboptimal response to a given concentration of insulin (1). Severe insulin resistance syndromes are a heterogenous group of rare hereditary or acquired disorders characterized by severe insulin resistance and hyperinsulinemia. Lipodystrophy syndromes, a group of disorders characterized by generalized or partial deficiency of adipose tissue, are associated with severe insulin resistance (2, 3). Most currently available therapies for management of diabetes associated with severe insulin resistance due to lipodystrophy are limited and nonspecific. Insulin, metformin, and other oral antidiabetic agents have been used to control hyperglycemia in patients with lipodystrophy; however, the efficacy of these treatments has not been studied systematically. Metreleptin, a pharmaceutical analog of the adipocyte-derived hormone leptin, ameliorates metabolic complications associated with generalized forms of lipodystrophy, but benefits are more modest and variable in partial lipodystrophy syndromes (4, 5). Thus, there is a clinical need to understand the efficacy of conventional diabetes therapies in patients with partial lipodystrophy.

Inhibition of the sodium glucose cotransporter type 2 (SGLT2) in patients with syndromes of severe insulin resistance is a rational therapeutic approach, as these drugs increase urinary glucose excretion, thus lowering blood glucose in an insulin-independent manner. However, the clinical effects of SGLT2 inhibitors (SGLT2i) in patients with severe insulin resistance remain largely unexplored. We therefore conducted a retrospective chart review of the safety and efficacy of SGLT2i in patients with severe insulin resistance due to partial lipodystrophy who were started on SGLT2i for clinical management of diabetes.

METHODS :

1. Subjects

Subjects were identified from a long-standing natural history study of patients with insulin resistance. Inclusion criteria for the efficacy analysis were clinically and/or genetically confirmed diagnosis of partial lipodystrophy, prescription for SGLT2i, available baseline clinical and laboratory data both prior to starting SGLT2i and after ≥6 months of continuous SGLT2i use. Exclusion criteria included failure to initiate SGLT2i treatment despite prescription. All patients included in the efficacy analysis, as well as patients who met the inclusion criteria above but did not have follow up laboratory data, were included in the safety analysis.

2. Study Design

This was a retrospective analysis of patients admitted to our institution. All studies were approved by the Institutional Review Board. All participants or their guardians provided written informed consent. Minors provided written assent. Data was collected for patients who met inclusion/exclusion criteria based on review of medical records. Due to the retrospective study design, no a priori sample size calculations were performed.

Clinical characteristics such as diagnosis, age, sex, BMI, and blood pressure were recorded. Historical data was collected by chart review, and included past medical history, medication use, side effects/adverse events from SGLT2i use and change in dose of SGLT2i. Total daily dose (TDD) of insulin was calculated if not explicitly noted in the medical record. TDD calculations were estimated based on blood glucose logs, home insulin prescriptions and history-based details about insulin compliance. Biochemical data was collected by chart review and included HbA1c, fasting blood glucose, C-peptide, 24-hour urinary glucose excretion, estimated glomerular filtration rate (eGFR), and fasting lipid profile. Standard methodology of the Department of Laboratory Medicine at the National Institutes of Health Clinical Center or other CLIA approved laboratories was used for all laboratory measurements. eGFR was calculated using CKD-EPI equation and reported as mL/min/1.73 square meter.

Visits were categorized as baseline visit (within one month prior to starting SGLT2i) and a follow up visit closest to 12 months of SGLT2i use (range, 6–15 months). Change in HbA1c, insulin dose, glucose, C-peptide, triglycerides, total cholesterol, LDL-C, HDL-C, eGFR, 24-hour urine glucose and blood pressure relative to baseline were calculated for the 12 month visit. Patients were classified as responders or non-responders to SGLT2i based on either improvement in HbA1c of ≥1% or decrease in the TDD of insulin by ≥20% after initiating SGLT2i.

3. Statistical analysis

Statistical analyses were performed using GraphPad Prism version 8.0.2 and SAS Enterprise Guide version 7.15 (SAS Institute, Cary, NC). For all outcomes, normally distributed data are reported as the mean ± SD and non-normally distributed data are reported as median [25^th, 75^th percentiles]. Non-normally distributed data were log-transformed prior to statistical analysis. P<0.05 was considered statistically significant. All statistical tests were two-sided. Outcomes were analyzed as change from baseline (prior to SGLT2i) adjusted for baseline values after 12 months of SGLT2i using linear mixed models. For HbA1c, an additional model was conducted including change in insulin dose for each individual subject as a covariate.

RESULTS

1. Subject Characteristics:

Review of medical records yielded 276 patients with lipodystrophy. 254 patients were excluded due to not meeting eligibility criteria. Among the 22 patients with partial lipodystrophy included in the study, 12 patients (58%) were analyzed for efficacy of SGLT2i as they had required clinical and laboratory follow up data and 10 (42%) were included only for safety analyses. While 6 patients had genetically confirmed partial lipodystrophy (LMNA [n=4], PPARγ [n=1] and PCYT1A [n=1]), genetic mutations were unknown for others. Baseline characteristics of the safety and efficacy cohorts are shown in Table 1. The mean age of the whole cohort was 40.9±13.2 years (efficacy group 37.4±14.1; safety-only group 45.2±11.1). Most patients were female (20 females, 2 males). Baseline BMI was 26.4±5.7 kg/m². Most patients were prescribed canagliflozin (n=15, 68%) or empagliflozin (n=6, 27%), and 1 patient was prescribed dapagliflozin. Among the efficacy cohort 10 patients (83.3%) were taking insulin and 10 (83.3%) were taking metformin. No new anti-diabetic medications were started on patients during the follow up period.

Table 1:

Baseline characteristics and safety profile of participants in the efficacy, safety only and combined cohort.

	Efficacy cohort N=12	Safety only cohort N=10	Combined cohort N=22
Age (years)	37.4±14.1	45.2±11.1	40.9±13.2
Gender	10 Females, 2 males	10 Females	20 Females, 2 males
Race
White	11 (92%)	10 (100%)	21 (95%)
Black	1 (8%)	0	1 (4%)
Weight (kg)	79.9±23.6	66.4±7.9	74.2±19.5
Height (cm)	167.8±11.9	165.5±4.8	166.8±9.4
BMI (kg/m2)	28.1±6.9	24.2±2.5	26.4±5.7
Systolic blood pressure (mmHg)	122±11	121±16	122±13
Diastolic blood pressure (mmHg)	73±8	72±13	73±10
Type of SGLT2i
Canagliflozin	9 (75%)	6 (60%)	15 (68%)
Empagliflozin	3 (25%)	3 (30%)	6 (27%)
Dapagliflozin	0	1 (10%)	1 (4%)
Number of patients taking insulin	10 (83.3%)	7 (70%)	17 (77%)
Number of patients taking metformin	10 (83.3%)	8 (80%)	18 (82%)
Number of patients taking GLP-1 agonists	0	3 (30%)	3 (14%)
Number of patients taking DPP-4 inhibitors	1 (8%)	0	1 (4%)
Number of patients taking thiazolidinedione	0	1 (10%)	1 (4%)
Patients who encountered adverse events	4 (33%)	3 (30%)	7 (32%)
Patients who discontinued SGLT2i due to adverse events	1 (8%)	2 (20%)	3 (14%)
Duration of SGLT2i use before discontinuing* (months)	5.8	0.6±0.5	2.3±3.0
Adverse events
Urinary tract infection	0	1 (10%)	1 (4%)
Fungal infections	1 (8%)	0	1 (4%)
Diabetes ketoacidosis	1 (8%)	0	1 (4%)
Pancreatitis	0	1 (10%)	1 (4%)
Arm or foot pain	1 (8%)	1 (10%)	2 (9%)
Hypoglycemia	1 (8%)	0	1 (4%)

Data are shown as N (%) or mean±SD.

^*Among patients who discontinued due to adverse event.

2. Safety Analysis Group:

Seven patients (32%) experienced an adverse event (AE) with SGLT2i use. Arm/foot pain was the most frequent AE, noted in 2 patients (9%). Urinary tract infection, fungal infection, diabetic ketoacidosis, hypoglycemia and pancreatitis occurred in 1 patient each.

Three patients (14%) discontinued SGLT2i treatment due to an AE after a mean duration of 2.3 months. Adverse events leading to discontinuation of SGLT2i included pancreatitis occurring one month after initiating SGLT2i in a patient with history of recurrent pancreatitis, and urinary tract infection occurring one week after initiation of SGLT2i. One patient with insulin nonadherence developed diabetic ketoacidosis after 6 months of SGLT2i. None of these patients restarted SGLT2i.

3. Efficacy Analysis Group (Table 2):

Table 2.

Glycemic parameters before and after SGLT2i administration.

	Before SGLT2i initiation	12 months after SGLT2i initiation	P (change adjusted for baseline; baseline vs 12 months)
HbA1c (%) [mmol/mol] (N=11)	9.2±2.0 [77.6±21.2]	8.4±1.8 [67.9±19.6]	0.028
HbA1c (%) adjusted for % change in insulin dose	N/A	N/A	0.029
Total daily dose of insulin (units) (N=12)	161±179	119±134	0.17
Fasting plasma glucose (mg/dL) (N=9)	177±91	158±50	0.62
24-hour urine glucose excretion (g/24hrs) (N=2)	13.3±2.1	70.0±15.8	0.062
Fasting C-peptide (ng/ml) (N=5)	6.1 (4.2,10.8)	3.4 (2.8, 6.3)	0.071
Triglycerides (mg/dL) (N=10)	332 (113, 744)	181 (150, 1189)	0.99
Total cholesterol (mg/dL) (N=10)	150±37	165±40	0.25
LDL-C (mg/dL) (N=7)	68±35	90±36	0.26
HDL-C (mg/dL) (N=9)	30±7	31±7	0.40
eGFR (mL/min/1.73m2) (N=10)	105±24	104±34	0.72
Systolic blood pressure (mmHg) (N=6)	129±11	116±12	0.011
Diastolic blood pressure (mmHg) (N=6)	77±8	67±10	0.013

N/A, not applicable; LDL-C, low density lipoprotein cholesterol; HDL-C, high density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate

Treatment with SGLT2i was associated with significant reduction in HbA1c (Fig 1A), from 9.2±2.0% [77.6±21.2 mmol/mol] at baseline to 8.4±1.8% [67.9±19.6 mmol/mol] after 12 months (p=0.028). Total daily dose of insulin numerically decreased from mean of 161 units to 119 units but this change was not significant (p=0.17). After accounting for changes in insulin dose, HbA1c reduction remained significant (p=0.029). Fasting blood glucose did not decrease after starting SGLT2i (Fig 1B) (p=0.62). There was a trend toward decreased C-peptide (Fig 1C) (P=0.071). Significant reductions were also noted in systolic (p=0.011) and diastolic blood pressure (Fig 1D) (p=0.013). 24-hour urinary glucose excretion increased 5-fold in the 2 patients with follow-up measurements. Lipids and eGFR remained unchanged.

Figure 1.

Outcomes at baseline (time 0) and 12 months after initiation of SGLT2 inhibitors in patients with partial lipodystrophy.

A) HbA1c decreased after SGLT2i administration (p=0.028) without change in B) fasting glucose (P=0.62). There was a trend toward decrease in C) C-peptide (P=0.071). D) Both systolic (closed circles) and diastolic (open squares) blood pressure decreased (P=0.011 and 0.013, respectively).

4. Comparison of Responders vs Non-Responders:

Characteristics of patients who responded to SGLT2i based on either improvement in HbA1c or decrease in the requirement of insulin are summarized and compared with those of non-responders in Table 3. Responders (n=9, 75%) and non-responders had similar demographics such as age, weight, and BMI. Baseline HbA1c (9.3±2.4 [78.0±25.7] vs 8.0±0.5 [63.4±4.7]mmol/mol]), fasting blood glucose (175±98 vs 144±43.7mg/dl) as well as 24 hour urine glucose (47.8±59.1 vs 16.2±5.6 g/24 hrs) were numerically higher in responders vs non-responders, but these differences were not statistically significant.

TABLE 3:

Comparison of baseline characteristics of responders vs non-responders to SGLT2i

	RESPONDERS (N=9)	NON-RESPONDERS (N=3)	P
Age (years)	34.6±10.3	45.8±23.1	0.25
Weight (kg)	83.5±26.7*	70.3±10.115.6	0.44
BMI (kg/m2)	29.4±7.7*	24.6± 1.7	0.33
HbA1c(%) [mmol/mol]	9.3±2.4 [78.0±25.7]	8.0±0.5 [63.4±4.7]	0.37
Fasting plasma glucose (mg/dl)	175±98*	144±43.7	0.62
Fasting C-peptide (ng/ml)	3.9 (2.6,6.1)†	3.2 (1.8, 9.4)	0.95
24-hour urine glucose excretion (g/24 hrs.)	47.8±59.1‡	16.2±5.6	0.41

^*N=8,

^†N=7,

^‡N=4

DISCUSSION:

This study explores the potential role of SGLT2i in a series of 12 patients with severe insulin resistance due to partial lipodystrophy. Our key finding was a significant improvement in glycemia (0.8% [9.7 mmol/mol] reduction in HbA1c) after 1 year of SGLT2i administration. The reduction in HbA1c remained statistically significant after adjustment for changes in insulin dose, indicating that improvement in glycemia was likely attributable to SGLT2i use. We endeavored to identify patient characteristics that would predict response to SGLT2i, but no predictors were identified.

Therapeutic options in patients with severe insulin resistance syndromes include metreleptin, insulin and metformin. Metreleptin significantly improves glycemic control in patients with generalized lipodystrophy (6) but has limited efficacy in patients with partial lipodystrophy (7), with benefit seen only in the subgroup with baseline HbA1c >8% (4). Very few cases describe the therapeutic potential of SGLT2i in patients with severe insulin resistance. A case of successful treatment of SHORT syndrome (a genetic insulin resistance syndrome) with SGLT2i was reported from Japan (8). Another case report described regression of fatty liver and decreased insulin resistance in a patient with congenital generalized lipodystrophy after use of SGLT2i for one year (9).

SGLT2i are an attractive therapeutic approach in patients with syndromes of severe insulin resistance because their mechanism of action bypasses defects in insulin signaling. The sodium-glucose cotransporter 2 located in the early proximal tubule of the kidney is the major cotransporter responsible for reabsorption of 90% of filtered glucose. SGLT2i promote renal excretion of this glucose thereby contributing to the therapeutic effect of lowering blood glucose in diabetes mellitus. Several clinical trials have been conducted to test efficacy of SGLT2i in patients with type 2 diabetes mellitus with baseline HbA1c ranging from 8–9%, leading to HbA1c reductions of 0.8% to 2% (10–14). The baseline HbA1c of 9.2% in our patients with partial lipodystrophy was higher than the patients studied in most clinical trials of type 2 diabetes mellitus. The observed 0.8% reduction in HbA1c% after 12 months of SGLT2i in patients with partial lipodystrophy is comparable to, albeit at the lower end of, efficacy estimates in patients with type 2 diabetes.

Consistent with known effects of SGLT2i to increase urinary glucose excretion, urinary glucose excretion increased 5-fold in 2 patients who had follow up measurement after 12 months of SGLT2i, to over 70 grams per day. Fasting blood glucose was not significantly reduced by SGLT2i treatment, likely reflecting day to day variability depending on food intake and exercise in this population with impaired beta-cell function and insulin dependence. The lipid lowering effect of SGLT2i treatment is controversial, with some studies suggesting decrease in total cholesterol, triglycerides and LDL-C (15) and others showing increased total cholesterol, LDL-C, HDL-C, and non-HDL-C (16). In our cohort, no change was found in the lipid profile over 12 months. Cardiovascular and renal benefits of SGLT2i treatment have been well documented in various trials and include reductions in 24-hour, daytime and nocturnal blood pressure in patients with diabetes and hypertension (17, 18). Consistent with this, there was a reduction in both systolic and diastolic blood pressure after SGLT2i treatment in our cohort.

This short-term retrospective analysis of SGLT2i use did not highlight any risks that clearly differed from those in T2D although the occurrence of diabetic ketoacidosis and pancreatitis in one patient each suggest that their use should be approached with caution in this population. Furthermore, long term effects of high glucosuria remain to be determined. The most common side effect of SGLT2i in our cohort was extremity pain, occurring in 9% of patients, which raised concern for vascular insufficiency. One patient taking insulin experienced hypoglycemia, which is consistent with knowledge that SGLT2 inhibitors cause hypoglycemia only in the presence of other concomitant therapy that can lower blood glucose levels, such as insulin. Although pancreatitis was also reported as an adverse event in one patient, its relationship to SGLT2i use is unclear as this patient had elevated triglycerides and prior history of recurrent pancreatitis. As expected based on inclusion criteria, the duration of SGLT2i use in patients who discontinued drug was much shorter in the safety only group vs the efficacy group.

This is largest study of SGLT2i use in patients with severe insulin resistance due to partial lipodystrophy. Although partial lipodystrophy is rare, thus limiting our cohort size, it is likely underdiagnosed, with newer prevalence estimates as high as 1 per 20,000 individuals (19). Thus, our findings may have relevance for many clinicians treating patients with diabetes who treat patients with partial lipodystrophy. Another limitation was the retrospective nature of the study with inadequacies in collection of historical and laboratory data as well as possibility of information bias due to reliance on accurate recordkeeping collected by others. Lack of follow up data further limited the number of patients that could be included in the efficacy analysis group.

We suggest that SGLT2 inhibitors might be a potential therapeutic option in a subset of patients with severe insulin resistance with uncontrolled diabetes and requiring high doses of insulin. In such patients, SGLT2i use may lead to better glycemic control. SGLT2i can be considered as part of the armamentarium for management of these patients after careful consideration of risks and benefits. Future prospective studies with larger sample sizes are warranted to confirm the clinical benefits and understand long term safety of SGLT2 inhibitors in treatment of syndromes of severe insulin resistance.

TEACHING POINTS:

• SGLT2 inhibitors reduced HbA1c in patients with partial lipodystrophy.

• SGLT2 inhibitors use in patients with partial lipodystrophy did not cause any adverse events that are different from those occurring in patients with type 2 diabetes.

• Use of SGLT2 inhibitors reduced systolic and diastolic blood pressure but fasting plasma glucose, lipids and eGFR remained unchanged in patients with partial lipodystrophy. After careful consideration of risks and benefits, SGLT2 inhibitors can be considered as part of the armamentarium for management of patients with partial lipodystrophy with uncontrolled diabetes and requiring high doses of insulin.