Clinical Therapeutics/Volume 37, Number 1, 2015
LX4211 Therapy Reduces Postprandial Glucose Levels in Patients With Type 2 Diabetes Mellitus and Renal Impairment Despite Low Urinary Glucose Excretion Brian Zambrowicz, PhD; Pablo Lapuerta, MD; Paul Strumph, MD; Phillip Banks, MS, FRS; Alan Wilson, PhD; Ike Ogbaa, MD; Arthur Sands, MD, PhD; and David Powell, MD Lexicon Pharmaceuticals, Inc, The Woodlands, Texas ABSTRACT Purpose: We sought to assess the efficacy and safety profile of LX4211, a dual inhibitor of sodium-glucose cotransporter1 (SGLT1) and SGLT2, in patients with type 2 diabetes and renal impairment. Methods: Thirty-one patients with an estimated glomerular filtration rate (eGFR) o60 mL/min/1.73 m2 were randomly assigned to receive 400 mg of LX4211 or placebo for 7 days. The primary end point was the change from baseline to day 7 in postprandial glucose (PPG) levels. Other end points included changes in fasting plasma glucose levels, glucagon-like peptide 1 levels, urinary glucose excretion (UGE), and blood pressure. Findings: LX4211 therapy significantly reduced PPG levels relative to placebo in the total population and in patients with an eGFR o45 mL/min/1.73 m2, with a placebo-adjusted decrease in incremental AUCpredose–4 of 73.5 mg h/dL (P ¼ 0.009) and 137.2 mg h/dL (P ¼ 0.001) for the total population and the eGFR o45 mL/ min/1.73 m2 subgroup, respectively. There was a significant reduction in fasting plasma glucose levels relative to baseline of –27.1 mg/dL (P o 0.001). Total and active glucagon-like peptide 1 levels were significantly elevated relative to placebo with LX4211 dosing, and UGE was significantly elevated with placebo-subtracted measures of 38.7, 53.5, and 20.4 g/24 h (P r 0.007 for all 3) in the total population, eGFR 45 to 59 mL/min/1.73 m2, and eGFR o45 mL/min/1.73 m2 subgroups, respectively. Implications: The PPG effects were maintained in patients with an eGFR o45 mL/min/1.73 m2 despite the expected reduction in UGE, suggesting that dual SGLT1 and SGLT2 inhibition with LX4211 could prove useful for the treatment of patients with type 2 diabetes and renal impairment. ClinicalTrials.gov identifier: NCT01555008. (Clin Ther. 2015;37:71–82) & 2015 Elsevier HS Journals, Inc. All rights reserved.
January 2015
Key words: fasting plasma glucose, LX4211, pharmacokinetics, postprandial glucose, renal impairment, SGLT1, SGLT2, urinary glucose excretion.
INTRODUCTION Type 2 diabetes mellitus, a growing public health concern, affects approximately 29 million people in the United States alone, and an additional 86 million patients have prediabetes.1 Type 2 diabetes is a major risk factor for progressive chronic kidney disease (CKD), and since 1988 there has been a steady increase in the prevalence of patients in the United States with type 2 diabetes and moderate to severe CKD, defined as a glomerular filtration rate (GFR) o60 mL/min/1.73 m2.2 By 2006, 17% to 18% of US patients known to have type 2 diabetes, nearly 3 million people, were in this GFR range,2–4 as were an estimated 10% of US patients with prediabetes, defined as a fasting blood glucose level 4100 and o126 mg/dL.5 Patients with type 2 diabetes are at increased risk for cardiovascular (CV) disease, the leading cause of morbidity and mortality for these patients.6 This increased CV risk seems to be largely limited to the subgroup of patients with CKD as patients with type 2 diabetes and without nephropathy have a similar mortality risk compared with patients without These data have been presented in part, as a poster, at the 2014 American Diabetes Association meeting in San Francisco, CA and has been accepted for a poster presentation at the 2014 EASD meeting in September 2014. Accepted for publication October 29, 2014. http://dx.doi.org/10.1016/j.clinthera.2014.10.026 0149-2918/$ - see front matter & 2015 Elsevier HS Journals, Inc. All rights reserved.
71
Clinical Therapeutics diabetes.7–9 Hyperglycemia is one of the modifiable risk factors for CKD, and studies suggest that glycemic control can reduce the risk of microvascular complications and reduce the progression of CKD.10–15 However, once renal impairment is present, patients with diabetes are at increased risk for hypoglycemia and drug-drug interactions, due to decreased kidney gluconeogenesis and decreased renal clearance of insulin and other drugs, thereby limiting their treatment options.16,17 The National Kidney Foundation (NKF) has released guidelines recommending a decreased dosage of some oral antidiabetes (OAD) agents and avoidance of others for patients with CKD.18 Options become increasingly restricted for patients with diabetes and a GFR o60 mL/min/1.73 m2. As a result, many of these patients with renal impairment use insulin therapy despite the risk of hypoglycemia. Notwithstanding recommendations, studies indicate that o50% of patients with type 2 diabetes and CKD are treated in accordance with NKF guidelines, and only 25.6% are treated in accordance with prescribing information.19–21 Discordance with treatment guidelines results in poorer glycemic control, increased risk of severe hypoglycemic events, and, in some studies, more inpatient hospitalizations and elevated treatment cost. Thus, there is a need for adherence to guidelines for currently available antidiabetes agents and new treatment options. The selective sodium-glucose cotransporter 2 (SGLT2) inhibitors represent a new class of OAD agents that work by blocking renal glucose reabsorption.22 By increasing urinary glucose excretion (UGE), these agents improve glycemic control by an insulinindependent mechanism while also decreasing weight and blood pressure (BP) through caloric loss and osmotic diuresis, respectively. However, because of their reliance on kidney function, UGE and the resulting reduction in the glycosylated hemoglobin (HbA1c) level are diminished as GFR is decreased. In contrast, LX4211 is a dual inhibitor of SGLT1 and SGLT2 that is designed to delay glucose absorption in the gastrointestinal tract, triggering elevated postprandial glucagon-like peptide 1 (GLP-1) and peptide YY release via SGLT1 inhibition and blocking renal glucose reabsorption via SGLT2 inhibition.23–26 Multiple-dose (28-day) administration of LX4211 as monotherapy in patients with diabetes decreased HbA1c, PPG, and fasting plasma glucose (FPG)
72
levels.23 In an additional 12-week study, dosing LX4211 on top of stable-dose metformin produced reductions in HbA1c concentrations, FPG levels, weight, and BP.27 As the dose increased in this study, further HbA1c reductions were produced in the absence of additional UGE, suggesting that the SGLT1 inhibition with LX4211 was clinically meaningful. Theoretically, any benefits obtained with LX4211 through inhibition of intestinal SGLT1 should not be diminished in patients with a decreased GFR. This study was conducted to determine whether the activity of LX4211 is maintained in patients with type 2 diabetes and moderate to severe renal impairment (defined by an estimated GFR [eGFR] o60 mL/min/ 1.73 m2) as assessed by the pharmacodynamic (PD) measures of PPG and FPG. In addition, the pharmacokinetic (PK) properties of LX4211 were characterized to determine whether the metabolism of LX4211 is modified in this patient population, and the study provided a first opportunity to assess the tolerability of LX4211 in patients with moderate to severe renal impairment.
PATIENTS AND METHODS Study Design This double-blind placebo-controlled study consisted of several study periods: Screening (Days –42 to Day 1), a 5-day Washout (Days –7 to -2) of any ongoing therapy for glucose, and a 7-day Treatment period. During Washout, patients discontinued any combination antidiabetes therapy and continued either a lifestyle or a diet-controlled regimen, basal (long-acting) insulin (ie, insulin glargine or insulin detemir only), or 1 permissible OAD agent (ie, a biguanide, a dipeptidyl peptidase 4 inhibitor, or thiazolidinedione). Patients who did not need medication washout were required to perform the 7-point finger-stick blood glucose (FSBG) test to demonstrate achievement of a daily fasting blood glucose level within the range Z100 and r270 mg/dL for Z2 consecutive days before day –2 and to be on a stable antidiabetes drug therapy regimen for Z5 consecutive days before day –2. LX4211 (2 x 200 mg) or placebo tablets were administered within 5 minutes before breakfast on day 1. At the investigator’s discretion, patients were released to return home on day 3, 4, or 5 with instructions to return for an outpatient visit on day 4. While outpatients, the patients continued
Volume 37 Number 1
B. Zambrowicz et al. dosing at home. Patients were contacted on day 5 via telephone to review FSBG measurements and to assess glucose ranges, the patient’s general condition, and any adverse events (AEs) that had occurred. Patients returned to the clinic for readmission by day 6. Patients were allowed to resume any prestudy antidiabetes drug therapy between days 8 and 16. Discharge from the clinic occurred on day 9, and the endof-study visit took place 1 week later on day 16. Sample size calculation was based on the assumption that the effect size in patients with renal impairment would be Z60% of the PPG reduction observed in a previous study of LX4211.23 This indicated a sample size of 26 patients for a statistical power estimate of 0.95. The randomization schedule was generated by ICON Development Solutions, LLC (Hanover, Maryland) following their standard operating procedures. Investigators, patients, Lexicon Pharmaceuticals, Inc, and the ICON study team were blinded to individual patients’ treatment during the study. Doses were selected based on the observation that the 400-mg dose of LX4211 produced the best efficacy in a dose-ranging study in patients with type 2 diabetes. This clinical study was conducted in accordance with the Declaration of Helsinki. The protocol was approved by the institutional review boards with jurisdiction over the sites, and all the patients provided written informed consent before study enrollment.
OBJECTIVES The primary objective of this study was to evaluate the effect of LX4211 therapy on PPG levels, measured as the change from baseline to day 7. Secondary objectives of this study were to evaluate tolerability, PD effects on FPG and GLP-1 levels from baseline to day 7, and PK effects of single and multiple doses. Several variables were examined in an exploratory manner, including 24-hour UGE, BP, mean FSBG levels, fractional excretion of calcium and phosphate, serum uric acid levels and fractional excretion of uric acid, fasting triglyceride levels, tumor necrosis factor α levels, leptin levels, and exogenous insulin dose.
PD and PK Assessments Beginning on the first day of the washout period (and continuing through the end-of-study visit), all the patients performed 7-point FSBG checks daily, completed a daily study diary, and were encouraged to follow a low glycemic index diet. To evaluate the PD effects of LX4211, blood samples were collected at
January 2015
multiple time points for measurement of PPG and GLP-1 levels (Pacific Biomarkers, Seattle, Washington) and before breakfast for FPG levels on day –1 (baseline), day 1 after administration of the first dose of study medication, and day 7. The UGE was assessed from aliquots of 24-hour urine collections at baseline (day -1), day 1 after administration of the first dose, and day 7. After a minimum 10-hour overnight fast and administration of study drug, FPG levels were measured; blood glucose and GLP-1 (total and active) levels were measured at multiple time points after ingestion of a high glycemic index, mixed meal. Systolic and diastolic BP (SBP and DBP) were taken in triplicate in the seated, supine, orthostatic seated, standing, and change from seated to standing positions on days –1, 2, 3, 4, 6, 7, 8, 9, and 16.
Safety Assessments Safety assessments included monitoring of AEs, clinical laboratory tests (chemistry, hematology, lipid profile, and urinalysis), vital signs (BP, heart rate, respiratory rate, and oral temperature), 12-lead electrocardiograms, physical examinations, and blood glucose levels.
Statistical Methods Data are summarized descriptively by treatment group and by study day (where appropriate) for continuous variables using the total number, mean, SD, median, maximum, and minimum. Descriptive summaries of categorical variables were based on patient counts and their associated percentages. The primary analysis population for the PD variables was the modified intention-to-treat population that included all patients as randomized and who had valid measurements to derive a change from baseline difference score on the study day of interest. Analysis of the safety profile data was based on the safety population and consisted of patients who had received any fraction of study drug; treatment group status was based on the actual treatment received on day 1. A linear mixed-effects model was used to test LX4211 minus placebo group differences in the change from baseline in total AUC1–4 of PPG on days 1 and 7 separately and the mean response for these 2 days; the primary end point was specified as the day 7 comparison. Tolerability data are summarized by treatment group and study day, where appropriate, using descriptive statistics only. Full statistical methods are detailed in the Supplemental Appendix I in the online
73
Clinical Therapeutics version at http://dx.doi.org/10.1016/j.clinthera.2014. 10.026.
RESULTS Of 139 patients screened, 31 were enrolled in the study between October 31, 2012, and August 14, 2013, and were followed up through the end-of-study visit. Patient disposition is summarized in the Supplemental Figure in the online version at http://dx.doi.org/10.1016/j.clinthera. 2014.10.026. Patients met all the protocol-specific inclusion criteria (see the Supplemental Appendix II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.10.026); none of the exclusion criteria were met, and no exemptions were granted. Sixteen patients enrolled in the study had an eGFR of 45 to 59 mL/min/1.73 m2 (stage 3A), and 15 had an eGFR
o45 mL/min/1.73 m2 (12 with an eGFR of 30–44 mL/ min/1.73 m2 [stage 3B] and 3 with an eGFR o30 mL/ min/1.73 m2 [severe]). The demographic characteristics of LX4211- and placebo-treated patients were similar (Table I). Patients were randomly assigned to receive LX4211, 400 mg/d (n ¼ 16), or placebo (n ¼ 15) for 7 days; 30 patients were 100% compliant. One LX4211-treated patient was withdrawn by the investigator on day 3 owing to noncompliance. Most patients were taking some type of insulin and some were taking OAD agents at screening (see the Supplemental Table I in the online version at http:// dx.doi.org/10.1016/j.clinthera.2014.10.026). Although there was washout of either insulin or OAD agents in a subset of patients, the pattern of previous
Table I. Demographic and baseline characteristics of the safety population. Characteristic Age, y Mean (SD) Median (min, max) Sex, No. (%) Male Female Race, No. (%) Black or African American White Multiple Height, cm* Mean (SD) Median (min, max) Weight, kg* Mean (SD) Median (min, max) Body mass index* Mean (SD) Median (min, max) eGFR, mL/min/1.73 m2* Mean (SD) Median (min, max)
LX4211 Group (n ¼ 16)
Placebo Group (n ¼ 15)
Overall (N ¼ 31)
64.8 (8.53) 63.5 (48, 79)
68.1 (7.33) 69 (50, 79)
66.4 (8.02) 67 (48, 79)
8 (50.0) 8 (50.0)
9 (60.0) 6 (40.0)
17 (54.8) 14 (45.2)
5 (31.3) 10 (62.5) 1 (6.3)
4 (26.7) 11 (73.3) 0
9 (29.0) 21 (67.7) 1 (3.2)
170.79 (12.440) 171.25 (152.0, 191.8)
171.07 (10.343) 170 (157.0, 184.0)
170.93 (11.284) 170 (152.0, 191.8)
103.05 (15.170) 101.00 (81.9, 132.6)
98.12 (22.561) 101.00 (65.0, 137.2)
100.66 (18.944) 101.00 (65.0, 137.2)
35.24 (4.381) 36.00 (28.0, 43.4)
33.21 (5.553) 32.90 (24.2, 41.0)
34.26 (5.006) 34.50 (24.2, 43.4)
45.2 (10.15) 48.5 (21, 59)
41.5 (11.20) 39.0 (20, 59)
43.4 (10.66) 46.0 (20, 59)
eGFR ¼ estimated glomerular filtration rate; min ¼ minimum; max ¼ maximum. * Measured at screening.
74
Volume 37 Number 1
B. Zambrowicz et al. medications and washout was similar in the treatment and placebo groups between patients with an eGFR of 45 to 59 mL/min/1.73 m2 and those with an eGFRo 45 mL/min/1.73 m2. In general, patients were returned to their previous medications at the end of the study. Regarding antihypertensive agents, most patients were taking medications at baseline (12 of 14 in the LX4211 arm and 13 of 14 in the placebo arm) and continued taking these medications throughout the treatment period.
Fasting Plasma Glucose Use of LX4211 produced a statistically significant reduction from baseline in FPG levels on day 7 of –27.1 mg/dL (P o 0.001) (see the Supplemental Table II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.10.026); however, there were no significant differences in reduction of FPG change from baseline comparing LX4211 with placebo (P ¼ 0.056). The difference in LS means between LX4211 and placebo was 27.3 mg/dL for the lower eGFR subgroup (P ¼ 0.07).
GLP-1 (Total and Active) Postprandial Glucose Compared with baseline (day –1), the mean plasma PPG concentrations decreased on days 1 and 7 after LX4211 dosing, with a greater decrease on day 7 compared with day 1, whereas the mean plasma PPG concentrations on days 1 and 7 did not change from baseline after placebo treatment (Figures A and B). LX4211 produced significant reductions from baseline for all 3 derived AUCs (AUC1–4, AUCpredose–4, and incremental AUCpredose–4) for PPG on day 7 and for incremental AUCpredose–4 on day 1 (P o 0.001), whereas placebo did not produce a statistically significant effect for any of the AUCs on either day 1 or 7 compared with baseline (see the Supplemental Table II in the online version at http://dx.doi.org/10. 1016/j.clinthera.2014.10.026). Compared with placebo, there was a significant reduction from baseline after LX4211 dosing for all 3 AUCs on day 7 (all, P r 0.010) but only for incremental AUC1–4 on day 1 (P ¼ 0.011). The least squares (LS) means indicated a greater reduction from baseline on day 7 compared with day 1 after LX4211 dosing, except for the incremental AUCpredose–4, which exhibited similar reductions from baseline on days 1 and 7. Levels of PPG were also examined in the eGFR subgroups of 45 to 59 mL/min/1.73 m2 and o45 mL/min/ 1.73 m2 (Figures C–F; see the Supplemental Table II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.10.026). In patients with an eGFR o45 mL/min/1.73 m2, PPG levels were significantly reduced on day 7 of LX4211 dosing compared with baseline (P o 0.001) and compared with placebo based on all 3 AUC measures (all, P r 0.002). In addition, for all comparisons, the peak PPG level seemed to be delayed in patients receiving LX4211 (Figures A, C, and E).
January 2015
For total and active GLP-1, the mean plasma concentrations increased on days 1 and 7 compared with baseline after LX4211 dosing (Figures G and I). There seemed to be a slightly greater increase on day 7 compared with day 1. The mean plasma concentrations of total and active GLP-1 seemed to be marginally greater than baseline levels after placebo treatment on days 1 and 7 (Figures H and J); however, the magnitude of increase was less than that after LX4211 dosing. LX4211 produced a statistically significant increase from baseline in all 3 derived AUCs (AUC1–4, AUCpredose–4, and incremental AUCpredose–4) for total and active GLP-1 on days 1 and 7 (see the Supplemental Table II in the online version at http://dx.doi.org/10.1016/j.clinthera. 2014.10.026). Total GLP-1 levels were significantly elevated relative to placebo on day 7 based on AUC1–4 and incremental AUCpredose–4, whereas active GLP-1 levels were not significantly elevated relative to placebo on day 7.
Urinary Glucose Excretion Use of LX4211 produced a significant UGE change from baseline on days 1 and 7 (P o 0.001); the increase was also significant compared with placebo on both study days (P o 0.001) (Table II). The LS means indicated a slightly greater increase in change from baseline on day 7 compared with day 1 after LX4211 dosing. The UGE was also significantly elevated versus placebo in the 2 eGFR subgroups, with a placebo-subtracted change from baseline of 53.5 g/24 h in the 45 to 59 mL/min/1.73 m2 group (P o 0.001) compared with 20.4 g/24 h in the o45 mL/min/1.73 m2 group (P ¼ 0.001) relative to placebo.
Blood Pressure Clinically significant reductions in the mean seated SBP, with a day 8 reduction of 11.4 mm Hg
75
Clinical Therapeutics
B
Postprandial glucose
Postprandial glucose (mg/dL)
Postprandial glucose (mg/dL)
A
LX4211 All pts
275 250 225 200 175 150 125 100 –1
0
1
2
3
4
5
LX4211 eGFR ≥45
250 225 200 175 150 125 100 –1
0
1
2
3
4
5
200 175 150 125 100 –1
0
1
2 Hour
3
200 175 150 125 100 –1
0
1
2
3
4
5
275
Placebo eGFR ≥45
250 225 200 175 150 125 100 –1
0
1
2
3
4
5
4
5
15 10 5 1
2
3
7
4
6
J
5 4 3 2 1 1
225 200 175 150 125 100 –1
0
1
2 Hour
3
4
5
2 Hour
3
4
Day -1
15 10 5
Day 1
0
1
2
3
7
4
5
Placebo All pts
6 5 4 3 2 1 0 –1
5
Placebo All pts
20
0 –1
5
LX4211 All pts
0
Placebo eGFR
LX4211 Therapy Reduces Postprandial Glucose Levels in Patients With Type 2 Diabetes Mellitus and Renal Impairment Despite Low Urinary Glucose Excretion Brian Zambrowicz, PhD; Pablo Lapuerta, MD; Paul Strumph, MD; Phillip Banks, MS, FRS; Alan Wilson, PhD; Ike Ogbaa, MD; Arthur Sands, MD, PhD; and David Powell, MD Lexicon Pharmaceuticals, Inc, The Woodlands, Texas ABSTRACT Purpose: We sought to assess the efficacy and safety profile of LX4211, a dual inhibitor of sodium-glucose cotransporter1 (SGLT1) and SGLT2, in patients with type 2 diabetes and renal impairment. Methods: Thirty-one patients with an estimated glomerular filtration rate (eGFR) o60 mL/min/1.73 m2 were randomly assigned to receive 400 mg of LX4211 or placebo for 7 days. The primary end point was the change from baseline to day 7 in postprandial glucose (PPG) levels. Other end points included changes in fasting plasma glucose levels, glucagon-like peptide 1 levels, urinary glucose excretion (UGE), and blood pressure. Findings: LX4211 therapy significantly reduced PPG levels relative to placebo in the total population and in patients with an eGFR o45 mL/min/1.73 m2, with a placebo-adjusted decrease in incremental AUCpredose–4 of 73.5 mg h/dL (P ¼ 0.009) and 137.2 mg h/dL (P ¼ 0.001) for the total population and the eGFR o45 mL/ min/1.73 m2 subgroup, respectively. There was a significant reduction in fasting plasma glucose levels relative to baseline of –27.1 mg/dL (P o 0.001). Total and active glucagon-like peptide 1 levels were significantly elevated relative to placebo with LX4211 dosing, and UGE was significantly elevated with placebo-subtracted measures of 38.7, 53.5, and 20.4 g/24 h (P r 0.007 for all 3) in the total population, eGFR 45 to 59 mL/min/1.73 m2, and eGFR o45 mL/min/1.73 m2 subgroups, respectively. Implications: The PPG effects were maintained in patients with an eGFR o45 mL/min/1.73 m2 despite the expected reduction in UGE, suggesting that dual SGLT1 and SGLT2 inhibition with LX4211 could prove useful for the treatment of patients with type 2 diabetes and renal impairment. ClinicalTrials.gov identifier: NCT01555008. (Clin Ther. 2015;37:71–82) & 2015 Elsevier HS Journals, Inc. All rights reserved.
January 2015
Key words: fasting plasma glucose, LX4211, pharmacokinetics, postprandial glucose, renal impairment, SGLT1, SGLT2, urinary glucose excretion.
INTRODUCTION Type 2 diabetes mellitus, a growing public health concern, affects approximately 29 million people in the United States alone, and an additional 86 million patients have prediabetes.1 Type 2 diabetes is a major risk factor for progressive chronic kidney disease (CKD), and since 1988 there has been a steady increase in the prevalence of patients in the United States with type 2 diabetes and moderate to severe CKD, defined as a glomerular filtration rate (GFR) o60 mL/min/1.73 m2.2 By 2006, 17% to 18% of US patients known to have type 2 diabetes, nearly 3 million people, were in this GFR range,2–4 as were an estimated 10% of US patients with prediabetes, defined as a fasting blood glucose level 4100 and o126 mg/dL.5 Patients with type 2 diabetes are at increased risk for cardiovascular (CV) disease, the leading cause of morbidity and mortality for these patients.6 This increased CV risk seems to be largely limited to the subgroup of patients with CKD as patients with type 2 diabetes and without nephropathy have a similar mortality risk compared with patients without These data have been presented in part, as a poster, at the 2014 American Diabetes Association meeting in San Francisco, CA and has been accepted for a poster presentation at the 2014 EASD meeting in September 2014. Accepted for publication October 29, 2014. http://dx.doi.org/10.1016/j.clinthera.2014.10.026 0149-2918/$ - see front matter & 2015 Elsevier HS Journals, Inc. All rights reserved.
71
Clinical Therapeutics diabetes.7–9 Hyperglycemia is one of the modifiable risk factors for CKD, and studies suggest that glycemic control can reduce the risk of microvascular complications and reduce the progression of CKD.10–15 However, once renal impairment is present, patients with diabetes are at increased risk for hypoglycemia and drug-drug interactions, due to decreased kidney gluconeogenesis and decreased renal clearance of insulin and other drugs, thereby limiting their treatment options.16,17 The National Kidney Foundation (NKF) has released guidelines recommending a decreased dosage of some oral antidiabetes (OAD) agents and avoidance of others for patients with CKD.18 Options become increasingly restricted for patients with diabetes and a GFR o60 mL/min/1.73 m2. As a result, many of these patients with renal impairment use insulin therapy despite the risk of hypoglycemia. Notwithstanding recommendations, studies indicate that o50% of patients with type 2 diabetes and CKD are treated in accordance with NKF guidelines, and only 25.6% are treated in accordance with prescribing information.19–21 Discordance with treatment guidelines results in poorer glycemic control, increased risk of severe hypoglycemic events, and, in some studies, more inpatient hospitalizations and elevated treatment cost. Thus, there is a need for adherence to guidelines for currently available antidiabetes agents and new treatment options. The selective sodium-glucose cotransporter 2 (SGLT2) inhibitors represent a new class of OAD agents that work by blocking renal glucose reabsorption.22 By increasing urinary glucose excretion (UGE), these agents improve glycemic control by an insulinindependent mechanism while also decreasing weight and blood pressure (BP) through caloric loss and osmotic diuresis, respectively. However, because of their reliance on kidney function, UGE and the resulting reduction in the glycosylated hemoglobin (HbA1c) level are diminished as GFR is decreased. In contrast, LX4211 is a dual inhibitor of SGLT1 and SGLT2 that is designed to delay glucose absorption in the gastrointestinal tract, triggering elevated postprandial glucagon-like peptide 1 (GLP-1) and peptide YY release via SGLT1 inhibition and blocking renal glucose reabsorption via SGLT2 inhibition.23–26 Multiple-dose (28-day) administration of LX4211 as monotherapy in patients with diabetes decreased HbA1c, PPG, and fasting plasma glucose (FPG)
72
levels.23 In an additional 12-week study, dosing LX4211 on top of stable-dose metformin produced reductions in HbA1c concentrations, FPG levels, weight, and BP.27 As the dose increased in this study, further HbA1c reductions were produced in the absence of additional UGE, suggesting that the SGLT1 inhibition with LX4211 was clinically meaningful. Theoretically, any benefits obtained with LX4211 through inhibition of intestinal SGLT1 should not be diminished in patients with a decreased GFR. This study was conducted to determine whether the activity of LX4211 is maintained in patients with type 2 diabetes and moderate to severe renal impairment (defined by an estimated GFR [eGFR] o60 mL/min/ 1.73 m2) as assessed by the pharmacodynamic (PD) measures of PPG and FPG. In addition, the pharmacokinetic (PK) properties of LX4211 were characterized to determine whether the metabolism of LX4211 is modified in this patient population, and the study provided a first opportunity to assess the tolerability of LX4211 in patients with moderate to severe renal impairment.
PATIENTS AND METHODS Study Design This double-blind placebo-controlled study consisted of several study periods: Screening (Days –42 to Day 1), a 5-day Washout (Days –7 to -2) of any ongoing therapy for glucose, and a 7-day Treatment period. During Washout, patients discontinued any combination antidiabetes therapy and continued either a lifestyle or a diet-controlled regimen, basal (long-acting) insulin (ie, insulin glargine or insulin detemir only), or 1 permissible OAD agent (ie, a biguanide, a dipeptidyl peptidase 4 inhibitor, or thiazolidinedione). Patients who did not need medication washout were required to perform the 7-point finger-stick blood glucose (FSBG) test to demonstrate achievement of a daily fasting blood glucose level within the range Z100 and r270 mg/dL for Z2 consecutive days before day –2 and to be on a stable antidiabetes drug therapy regimen for Z5 consecutive days before day –2. LX4211 (2 x 200 mg) or placebo tablets were administered within 5 minutes before breakfast on day 1. At the investigator’s discretion, patients were released to return home on day 3, 4, or 5 with instructions to return for an outpatient visit on day 4. While outpatients, the patients continued
Volume 37 Number 1
B. Zambrowicz et al. dosing at home. Patients were contacted on day 5 via telephone to review FSBG measurements and to assess glucose ranges, the patient’s general condition, and any adverse events (AEs) that had occurred. Patients returned to the clinic for readmission by day 6. Patients were allowed to resume any prestudy antidiabetes drug therapy between days 8 and 16. Discharge from the clinic occurred on day 9, and the endof-study visit took place 1 week later on day 16. Sample size calculation was based on the assumption that the effect size in patients with renal impairment would be Z60% of the PPG reduction observed in a previous study of LX4211.23 This indicated a sample size of 26 patients for a statistical power estimate of 0.95. The randomization schedule was generated by ICON Development Solutions, LLC (Hanover, Maryland) following their standard operating procedures. Investigators, patients, Lexicon Pharmaceuticals, Inc, and the ICON study team were blinded to individual patients’ treatment during the study. Doses were selected based on the observation that the 400-mg dose of LX4211 produced the best efficacy in a dose-ranging study in patients with type 2 diabetes. This clinical study was conducted in accordance with the Declaration of Helsinki. The protocol was approved by the institutional review boards with jurisdiction over the sites, and all the patients provided written informed consent before study enrollment.
OBJECTIVES The primary objective of this study was to evaluate the effect of LX4211 therapy on PPG levels, measured as the change from baseline to day 7. Secondary objectives of this study were to evaluate tolerability, PD effects on FPG and GLP-1 levels from baseline to day 7, and PK effects of single and multiple doses. Several variables were examined in an exploratory manner, including 24-hour UGE, BP, mean FSBG levels, fractional excretion of calcium and phosphate, serum uric acid levels and fractional excretion of uric acid, fasting triglyceride levels, tumor necrosis factor α levels, leptin levels, and exogenous insulin dose.
PD and PK Assessments Beginning on the first day of the washout period (and continuing through the end-of-study visit), all the patients performed 7-point FSBG checks daily, completed a daily study diary, and were encouraged to follow a low glycemic index diet. To evaluate the PD effects of LX4211, blood samples were collected at
January 2015
multiple time points for measurement of PPG and GLP-1 levels (Pacific Biomarkers, Seattle, Washington) and before breakfast for FPG levels on day –1 (baseline), day 1 after administration of the first dose of study medication, and day 7. The UGE was assessed from aliquots of 24-hour urine collections at baseline (day -1), day 1 after administration of the first dose, and day 7. After a minimum 10-hour overnight fast and administration of study drug, FPG levels were measured; blood glucose and GLP-1 (total and active) levels were measured at multiple time points after ingestion of a high glycemic index, mixed meal. Systolic and diastolic BP (SBP and DBP) were taken in triplicate in the seated, supine, orthostatic seated, standing, and change from seated to standing positions on days –1, 2, 3, 4, 6, 7, 8, 9, and 16.
Safety Assessments Safety assessments included monitoring of AEs, clinical laboratory tests (chemistry, hematology, lipid profile, and urinalysis), vital signs (BP, heart rate, respiratory rate, and oral temperature), 12-lead electrocardiograms, physical examinations, and blood glucose levels.
Statistical Methods Data are summarized descriptively by treatment group and by study day (where appropriate) for continuous variables using the total number, mean, SD, median, maximum, and minimum. Descriptive summaries of categorical variables were based on patient counts and their associated percentages. The primary analysis population for the PD variables was the modified intention-to-treat population that included all patients as randomized and who had valid measurements to derive a change from baseline difference score on the study day of interest. Analysis of the safety profile data was based on the safety population and consisted of patients who had received any fraction of study drug; treatment group status was based on the actual treatment received on day 1. A linear mixed-effects model was used to test LX4211 minus placebo group differences in the change from baseline in total AUC1–4 of PPG on days 1 and 7 separately and the mean response for these 2 days; the primary end point was specified as the day 7 comparison. Tolerability data are summarized by treatment group and study day, where appropriate, using descriptive statistics only. Full statistical methods are detailed in the Supplemental Appendix I in the online
73
Clinical Therapeutics version at http://dx.doi.org/10.1016/j.clinthera.2014. 10.026.
RESULTS Of 139 patients screened, 31 were enrolled in the study between October 31, 2012, and August 14, 2013, and were followed up through the end-of-study visit. Patient disposition is summarized in the Supplemental Figure in the online version at http://dx.doi.org/10.1016/j.clinthera. 2014.10.026. Patients met all the protocol-specific inclusion criteria (see the Supplemental Appendix II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.10.026); none of the exclusion criteria were met, and no exemptions were granted. Sixteen patients enrolled in the study had an eGFR of 45 to 59 mL/min/1.73 m2 (stage 3A), and 15 had an eGFR
o45 mL/min/1.73 m2 (12 with an eGFR of 30–44 mL/ min/1.73 m2 [stage 3B] and 3 with an eGFR o30 mL/ min/1.73 m2 [severe]). The demographic characteristics of LX4211- and placebo-treated patients were similar (Table I). Patients were randomly assigned to receive LX4211, 400 mg/d (n ¼ 16), or placebo (n ¼ 15) for 7 days; 30 patients were 100% compliant. One LX4211-treated patient was withdrawn by the investigator on day 3 owing to noncompliance. Most patients were taking some type of insulin and some were taking OAD agents at screening (see the Supplemental Table I in the online version at http:// dx.doi.org/10.1016/j.clinthera.2014.10.026). Although there was washout of either insulin or OAD agents in a subset of patients, the pattern of previous
Table I. Demographic and baseline characteristics of the safety population. Characteristic Age, y Mean (SD) Median (min, max) Sex, No. (%) Male Female Race, No. (%) Black or African American White Multiple Height, cm* Mean (SD) Median (min, max) Weight, kg* Mean (SD) Median (min, max) Body mass index* Mean (SD) Median (min, max) eGFR, mL/min/1.73 m2* Mean (SD) Median (min, max)
LX4211 Group (n ¼ 16)
Placebo Group (n ¼ 15)
Overall (N ¼ 31)
64.8 (8.53) 63.5 (48, 79)
68.1 (7.33) 69 (50, 79)
66.4 (8.02) 67 (48, 79)
8 (50.0) 8 (50.0)
9 (60.0) 6 (40.0)
17 (54.8) 14 (45.2)
5 (31.3) 10 (62.5) 1 (6.3)
4 (26.7) 11 (73.3) 0
9 (29.0) 21 (67.7) 1 (3.2)
170.79 (12.440) 171.25 (152.0, 191.8)
171.07 (10.343) 170 (157.0, 184.0)
170.93 (11.284) 170 (152.0, 191.8)
103.05 (15.170) 101.00 (81.9, 132.6)
98.12 (22.561) 101.00 (65.0, 137.2)
100.66 (18.944) 101.00 (65.0, 137.2)
35.24 (4.381) 36.00 (28.0, 43.4)
33.21 (5.553) 32.90 (24.2, 41.0)
34.26 (5.006) 34.50 (24.2, 43.4)
45.2 (10.15) 48.5 (21, 59)
41.5 (11.20) 39.0 (20, 59)
43.4 (10.66) 46.0 (20, 59)
eGFR ¼ estimated glomerular filtration rate; min ¼ minimum; max ¼ maximum. * Measured at screening.
74
Volume 37 Number 1
B. Zambrowicz et al. medications and washout was similar in the treatment and placebo groups between patients with an eGFR of 45 to 59 mL/min/1.73 m2 and those with an eGFRo 45 mL/min/1.73 m2. In general, patients were returned to their previous medications at the end of the study. Regarding antihypertensive agents, most patients were taking medications at baseline (12 of 14 in the LX4211 arm and 13 of 14 in the placebo arm) and continued taking these medications throughout the treatment period.
Fasting Plasma Glucose Use of LX4211 produced a statistically significant reduction from baseline in FPG levels on day 7 of –27.1 mg/dL (P o 0.001) (see the Supplemental Table II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.10.026); however, there were no significant differences in reduction of FPG change from baseline comparing LX4211 with placebo (P ¼ 0.056). The difference in LS means between LX4211 and placebo was 27.3 mg/dL for the lower eGFR subgroup (P ¼ 0.07).
GLP-1 (Total and Active) Postprandial Glucose Compared with baseline (day –1), the mean plasma PPG concentrations decreased on days 1 and 7 after LX4211 dosing, with a greater decrease on day 7 compared with day 1, whereas the mean plasma PPG concentrations on days 1 and 7 did not change from baseline after placebo treatment (Figures A and B). LX4211 produced significant reductions from baseline for all 3 derived AUCs (AUC1–4, AUCpredose–4, and incremental AUCpredose–4) for PPG on day 7 and for incremental AUCpredose–4 on day 1 (P o 0.001), whereas placebo did not produce a statistically significant effect for any of the AUCs on either day 1 or 7 compared with baseline (see the Supplemental Table II in the online version at http://dx.doi.org/10. 1016/j.clinthera.2014.10.026). Compared with placebo, there was a significant reduction from baseline after LX4211 dosing for all 3 AUCs on day 7 (all, P r 0.010) but only for incremental AUC1–4 on day 1 (P ¼ 0.011). The least squares (LS) means indicated a greater reduction from baseline on day 7 compared with day 1 after LX4211 dosing, except for the incremental AUCpredose–4, which exhibited similar reductions from baseline on days 1 and 7. Levels of PPG were also examined in the eGFR subgroups of 45 to 59 mL/min/1.73 m2 and o45 mL/min/ 1.73 m2 (Figures C–F; see the Supplemental Table II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.10.026). In patients with an eGFR o45 mL/min/1.73 m2, PPG levels were significantly reduced on day 7 of LX4211 dosing compared with baseline (P o 0.001) and compared with placebo based on all 3 AUC measures (all, P r 0.002). In addition, for all comparisons, the peak PPG level seemed to be delayed in patients receiving LX4211 (Figures A, C, and E).
January 2015
For total and active GLP-1, the mean plasma concentrations increased on days 1 and 7 compared with baseline after LX4211 dosing (Figures G and I). There seemed to be a slightly greater increase on day 7 compared with day 1. The mean plasma concentrations of total and active GLP-1 seemed to be marginally greater than baseline levels after placebo treatment on days 1 and 7 (Figures H and J); however, the magnitude of increase was less than that after LX4211 dosing. LX4211 produced a statistically significant increase from baseline in all 3 derived AUCs (AUC1–4, AUCpredose–4, and incremental AUCpredose–4) for total and active GLP-1 on days 1 and 7 (see the Supplemental Table II in the online version at http://dx.doi.org/10.1016/j.clinthera. 2014.10.026). Total GLP-1 levels were significantly elevated relative to placebo on day 7 based on AUC1–4 and incremental AUCpredose–4, whereas active GLP-1 levels were not significantly elevated relative to placebo on day 7.
Urinary Glucose Excretion Use of LX4211 produced a significant UGE change from baseline on days 1 and 7 (P o 0.001); the increase was also significant compared with placebo on both study days (P o 0.001) (Table II). The LS means indicated a slightly greater increase in change from baseline on day 7 compared with day 1 after LX4211 dosing. The UGE was also significantly elevated versus placebo in the 2 eGFR subgroups, with a placebo-subtracted change from baseline of 53.5 g/24 h in the 45 to 59 mL/min/1.73 m2 group (P o 0.001) compared with 20.4 g/24 h in the o45 mL/min/1.73 m2 group (P ¼ 0.001) relative to placebo.
Blood Pressure Clinically significant reductions in the mean seated SBP, with a day 8 reduction of 11.4 mm Hg
75
Clinical Therapeutics
B
Postprandial glucose
Postprandial glucose (mg/dL)
Postprandial glucose (mg/dL)
A
LX4211 All pts
275 250 225 200 175 150 125 100 –1
0
1
2
3
4
5
LX4211 eGFR ≥45
250 225 200 175 150 125 100 –1
0
1
2
3
4
5
200 175 150 125 100 –1
0
1
2 Hour
3
200 175 150 125 100 –1
0
1
2
3
4
5
275
Placebo eGFR ≥45
250 225 200 175 150 125 100 –1
0
1
2
3
4
5
4
5
15 10 5 1
2
3
7
4
6
J
5 4 3 2 1 1
225 200 175 150 125 100 –1
0
1
2 Hour
3
4
5
2 Hour
3
4
Day -1
15 10 5
Day 1
0
1
2
3
7
4
5
Placebo All pts
6 5 4 3 2 1 0 –1
5
Placebo All pts
20
0 –1
5
LX4211 All pts
0
Placebo eGFR
