1 INTRODUCTION ORIGINAL ARTICLE

Transcription

1 Received: July Revised: Month Accepted: 7 July DOI./dom.7 ORIGINAL ARTICLE Reduced peripheral activity leading to hepato-preferential action of basal insulin peglispro compared with insulin glargine in patients with type diabetes S. Mudaliar, R. R. Henry, T. P. Ciaraldi, D. A. Armstrong P. M. Burke J. H. Pettus P. Garhyan S. L. Choi M. P. Knadler E. C. Q. Lam M. J. Prince N. Bose N. K. Porksen V. P. Sinha H. Linnebjerg S. J. Jacober Veterans Affairs San Diego Healthcare System, San Diego, CA, USA Department of Medicine, Division of Endocrinology and Metabolism, University of California, San Diego, CA, USA Eli Lilly and Company, Indianapolis, IN, USA Eli Lilly and Company, Singapore, Singapore Corresponding Author: Scott J. Jacober, DO, Eli Lilly and Company, Indianapolis, IN 8 (jacober_scott_j@lilly.com) Aims: Basal insulin peglispro (BIL), a novel PEGylated basal insulin with a large hydrodynamic size, has a delayed absorption and reduced clearance that prolongs the duration of action. The current study compared the effects of BIL and insulin glargine (GL) on endogenous glucose production (EGP), glucose disposal rate (GDR) and lipolysis in patients with type diabetes. Materials and Methods: This was a randomized, open-label, four-period, crossover study. Patients received intravenous infusions of BIL and GL, each at two dose levels selected for partial and maximal suppression of EGP, during an 8 to h euglycemic clamp procedure with D- [- H] glucose. Results: Following correction for equivalent human insulin concentrations (EHIC), low-dose GL infusion resulted in similar EGP at the end of the clamp compared to low-dose BIL infusion (GL/BIL ratio of.) but a higher GDR (GL/BIL ratio of.), indicating similar hepatic activity but attenuated peripheral activity of BIL. Consistent with this, the EHIC-corrected GDR/EGP at the end of the clamp was.7-fold greater for GL than BIL following low-dose administration. At the lower dose of BIL and GL (concentrations in the therapeutic range), BIL produced less suppression of lipolysis compared with GL as indicated by free fatty acid and glycerol levels at the end of the clamp. Conclusions: Compared with GL, BIL restored the hepato-peripheral insulin action gradient seen in normal physiology via its peripherally restricted action on target tissues related to carbohydrate and lipid metabolism. KEYWORDS basal insulin, clinical trial, drug development, pharmacodynamics, phase III study INTRODUCTION Endogenous insulin is secreted into the portal vein where it first acts on hepatic metabolism before entering the systemic circulation. Insulin is significantly cleared by the liver (-8%), resulting in two- to fourfold greater levels in the portal vein and hepatic interstitium than in the systemic circulation. In patients with diabetes, subcutaneous Portions of this study were presented as abstracts at the 7th Scientific Sessions of the American Diabetes Association, Boston, MA, USA, 9 June, [and intravenous (IV)] insulin therapy is administered peripherally, resulting in similar systemic and portal levels of insulin. Euglycemia under normal treatment conditions may therefore be achieved at the expense of under-insulinization of the liver and over-insulinization of the periphery., In the fasting state, this may predispose patients to hypoglycemia and its associated complications, and insulin therapies which attempt to replicate the normal hepatic-to-peripheral insulin gradient may lessen this risk. In addition a hepato-preferentially acting insulin therapy could potentially lessen the weight effects of over-insulinization of the periphery. Diabetes Obes Metab ; 8 (Suppl. ): 7 wileyonlinelibrary.com/journal/dom John Wiley & Sons Ltd 7

2 8 MUDALIAR ET AL. Basal insulin peglispro (BIL) is a novel, PEGylated, basal insulin with a large hydrodynamic size. Insulin peglispro has been shown to have a long duration of action with a half-life of - days. 7 Furthermore, BIL was associated with reduced rates of nocturnal hypoglycemia in all Phase and Phase clinical comparator trials and provided improved glycemic control compared to insulin glargine (GL) in these studies. 8 BIL was also associated with weight loss vs comparator GL. In a dog model of insulinopenic diabetes, BIL had a hepato-preferential effect of approximately twofold compared with peripherally administered, (IV) human insulin. 7 In that euglycemic clamp study, portal insulin levels remained below baseline despite substantive glucose infusion rates of to mg/kg/min, demonstrating the under-insulinization of the liver even with high rates of glucose disposal. We previously reported the hepato-preferential action of insulin peglispro in healthy subjects under euglycemic clamp conditions. 8 Dose response data from that study were utilized to select the doses for use in this study that was designed to investigate the hepatopreferential action of insulin peglispro in patients with type diabetes (TD). MATERIALS AND METHODS. Study design and participants This was a single-centre, randomized, open-label, -period, crossover study in patients with TD. The study consisted of a screening visit, four study visits occurring 7 to days apart, during which patients underwent a euglycemic clamp procedure, and a follow-up visit that occurred at least days after the last dose of study drug. Each patient underwent four clamps with primed, continuous IV infusions of two doses (low dose and high dose) each of BIL and GL given in a crossover design with six prespecified treatment sequences. For each insulin, doses were chosen such that the low dose would similarly produce partial suppression of endogenous glucose production (EGP), and the high dose would produce approximately % suppression of EGP. 8 The study was conducted in accordance with the International Conference on Harmonization Guidelines for good clinical practice and the Declaration of Helsinki. All procedures were approved by the institutional review board at the Veteran s Affairs San Diego Healthcare System, and all patients provided written, informed consent. Male patients 8 and years of age with a history of TD for at least year were eligible for the study. At screening, patients were required to have had a body mass index (BMI) of. to 9.9 kg/m c-peptide concentration <. ng/ml, hemoglobin Ac (HbAc) of to 7 mmol/mol (.% to 9.%), and have had no episodes of severe hypoglycemia in the prior months.. Study procedures Patients continued their routine basal insulin regimen until the morning of day (the day before the euglycemic clamp procedure). On day, patients using a twice-daily regimen of either insulin detemir or GL took only the morning dose; patients using a once-daily morning regimen of either insulin detemir or GL took % of their usual morning dose; patients using a once-daily evening/bedtime regimen of either insulin detemir or GL did not take their usual evening or bedtime dose; and patients using an insulin pump continued their pump basal infusion until approximately : PM, when the pump was discontinued. On day, patients continued their routine bolus insulin regimen. Patients were admitted to the clinical research unit (CRU) in the afternoon or evening of day, and after admission received a meal and their last bolus insulin dose until completion of the clamp procedure. Patients fasted from the end of the evening meal on day - (at least 8 h prior to start of the clamp) through the end of the clamp procedure. At approximately : PM on day, an infusion of regular insulin and % dextrose solution was initiated (if required) and titrated to maintain blood glucose in the range of approximately. to 7.8 mmol/l (8 to mg/dl) prior to the clamp. This infusion was discontinued at least hour prior to the start of study insulin infusion and the euglycemic clamp procedure. The low-dose BIL infusion was primed with a U IV bolus followed by a. mu/min continuous IV infusion and the high-dose BIL infusion was primed with an 8 U IV bolus followed by a 7. mu/min continuous IV infusion. The low-dose GL infusion was primed with a mu/m IV bolus followed by a mu/m /min continuous IV infusion and the high-dose GL infusion was primed with a mu/m IV bolus followed by a mu/m /min continuous IV infusion. The priming doses were administered over minutes. The euglycemic clamp method using radiotracers was previously described. 8 The sequence of events is diagrammed in Figure. Briefly, a continuous D-[- H] glucose (. μci/min) IV infusion was begun approximately hours before the start of the clamp and continued throughout the clamp. The study insulin during the clamp was administered beginning with the priming dose administered over minutes followed by a continuous IV infusion. The duration of the clamp procedure was approximately 8 hours except for clamps conducted at the high insulin peglispro dose level, where the duration was extended to approximately hours based on the results of the healthy subject clamps. 8 Following initiation of dosing, a % glucose solution containing D-[- H] glucose was infused at a variable rate, adjusted to maintain euglycemia [approximately. mmol/l (9 mg/dl)] as described previously. 9 Upon completion of the clamp, patients received a meal and blood glucose was monitored at appropriate intervals to ensure stable blood glucose concentrations before the patient was discharged from the CRU.. Study drug BIL and GL were supplied as a -U/mL solution in a ml glass vial. Both study insulins were diluted as required to achieve study drug concentrations between. and. U/mL. Infusion rates were adjusted to achieve the required dosing rates.. Pharmacokinetic/pharmacodynamic sample and data collection Venous blood samples were collected at predose (time ), and minutes, and,,,,, 8, 9 and hours after the start of study insulin infusion for the determination of serum concentrations

3 MUDALIAR ET AL. 9 FIGURE Euglycemic clamp study design. Patients were admitted to the clinical research unit the afternoon/evening prior to receiving study drug. A low-dose infusion of regular insulin occurred between and - hour to maintain blood glucose at approximately mmol/l (9 mg/dl). A radioactive tracer (D-[ H] glucose;. μci/min) was continuously infused beginning at hours to label the glucose pool and achieve steady state, and continued through the end of the clamp. Administration of study drug began at time with infusion of a high or low priming dose over minutes followed by a constant infusion over the duration of the clamp. A % glucose solution containing a D-[ H] glucose tracer was infused to maintain euglycemia. At the end of the clamp the patients received a meal and prandial insulin; basal insulin was resumed at the discretion of the investigator. Glucose infusion and blood glucose monitoring continued until blood glucose values were stable and within a safe range. Low-dose basal insulin peglispro (BIL) =. U priming dose followed by a constant infusion of. mu/min; lowdose insulin glargine (GL) = mu/m priming dose followed by a constant infusion of mu/m /min; high dose BIL = 8. U priming dose followed by a constant infusion of 7. mu/min; high dose GL = mu/m priming dose followed by a constant infusion of mu/ m /min. *Clamp duration was hours following the high dose of BIL. CRU = clinical research unit. CRU admission Infusion of regular insulin or % glucose as needed to maintain blood glucose at ~.-7.8 mmol/l Tracer equilibration Variable glucose infusion % dextrose + D-[- H] glucose BIL/GL infusion (low/high dose) Continuous D-[- H] glucose (. µci/min) infusion Recovery (variable duration) % glucose CRU discharge Hour ~. 8 * ~ Clamp Blood sampling for glucose/insulin of BIL, GL, glucose, free fatty acid (FFA) and glycerol as appropriate. Additional samples were collected at and hours after the start of study insulin infusion during the -hour clamp performed with high-dose BIL infusion. Insulin peglispro and GL were measured as previously described. 8 The rate of glucose appearance and rate of glucose disappearance were calculated using the modified Steele equations for nonsteady-state conditions. 9, To adjust for differences in receptor binding between BIL and GL, equivalent human insulin concentrations (EHIC) were calculated by dividing each study insulin concentration by its receptor inhibitory binding constant (Ki) and multiplying by the Ki of human insulin. Free (non-esterified) serum fatty acids were determined using a validated, commercial - step acyl-coa synthase acyl-coa oxidase enzymatic assay. The lower limit of quantification (LLOQ) was.8 mg/dl. Free serum glycerol concentrations were determined using a validated glycerol phosphate-oxidase enzymatic assay. The LLOQ was. mg/dl.. Statistical analysis The endpoint EGP, % EGP suppression, glucose disposal rate (GDR) and GDR fold change from baseline (start of infusion) were calculated using the average of measurements from the last -time points of the infusion period as the endpoint. The EGP, GDR, FFA and glycerol concentrations (at baseline and endpoint), and other parameters (EGP/ EHIC, GDR /EHIC and EGP/GDR ratios at endpoint), were statistically compared at the low insulin doses ( mu/m /min GL was compared with. mu/min BIL) and the high insulin doses ( mu/ m /min GL was compared with 7. mu/min BIL) using a mixed effects model with treatment, period and sequence as fixed effects and subject fitted as a random effect. The least squares (LS) mean, standard error and P-values for the comparisons were calculated. The % EGP suppression and GDR fold change from baseline, and FFA and glycerol change from baseline at endpoint were compared between treatments using the same statistical model with baseline as an additional covariate. Because a large number of samples had FFA and glycerol values that were below the quantifiable lower limit of the assay (BQL), data that were BQL were imputed to half of the LLOQ. RESULTS A total of white, male patients with TD were enrolled in the study. One patient was discontinued from the study after a single clamp (low-dose BIL) due to lack of venous access. The mean age

4 MUDALIAR ET AL. [ standard deviation (SD)] was.7. years with a mean weight of kg and BMI of.. kg/m. The mean duration of diabetes was years and the mean HbAc was 9. mmol/mol (7..8%). The plasma concentrations of GL and BIL rapidly reached steady state during the IV infusions (Figure A,B). Equivalent human insulin concentrations at clamp completion appeared to be slightly higher for BIL compared to GL following low-dose infusions (97 vs pmol/l), but were substantially higher for BIL compared with GL following high-dose infusions ( vs pmol/l; Table ). Blood glucose levels stabilized within the target range of. to 7.8 mmol/l by approximately hours after the start of the clamp procedure following both low and high doses of BIL and GL (Figure C,D). For both primed infusions of GL administered over 8 hours, the glucose infusion rate (GIR) increased for approximately to hours before stabilizing (Figure E,F). When BIL was infused at the low dose for 8 hours, the GIR increased for approximately hour before reaching a plateau, whereas during the high-dose infusion, the GIR steadily increased over the -hour clamp duration (Figure E,F). Low-dose infusion of BIL or GL resulted in partial suppression of EGP (Figure A), while high-dose infusion of BIL or GL resulted in near % suppression of EGP (Figure B). Following low-dose infusions, which partially suppressed EGP, GDR remained relatively stable following GL infusion but decreased slightly following BIL infusion (Figure C). At the high dose of GL, which suppressed EGP nearly %, GDR increased to approximately twice that of baseline (Figure D, Table ). During the high-dose infusion of BIL, which also suppressed EGP nearly %, the GDR increased (Figure D) to a level approximately.-fold higher than baseline GDR (Table ). The EGP at baseline, the EGP at endpoint and EHIC-corrected EGP at endpoint were similar between insulin treatments for both the low and high doses (Table ). The absolute EGP and GDR values at baseline were similar between treatments for both the low- and high-dose infusions (Table ). The LS mean GDR at endpoint was higher following treatment with GL compared with BIL at both low [.7 vs. mg/kg/ min (ratio of LS means =.); P-value =. based on difference in LS means] and high doses [. vs. mg/kg/min (ratio.8); P =.]. This effect was further confirmed when EHIC-corrected (A) Mean (+/- SD) Serum GL Concentration (pmol/l) mu/m /min GL mu/m /min GL 8 8 (B) Mean (+/- SD) Serum BIL Concentration (pmol/l). mu/min BIL 7. mu/min BIL mg/dl mg/min/kg (C) (E) 8 mu/m /min GL. mu/min BIL 8 7 mmol/l 8 8 Mean (+/- SD) Blood Glucose Mean (+/- SD) GIR mu/m /min GL. mu/min BIL mg/dl (D) (F) mg/min/kg 8 Mean (+/- SD) Blood Glucose mu/m /min GL 7. mu/min BIL mu/m /min GL 7. mu/min BIL Mean (+/- SD) GIR 8 7 mmol/l FIGURE Basal insulin peglispro, insulin glargine, blood glucose and glucose infusion rate profiles. Mean serum concentrations of insulin glargine (GL) (A) or basal insulin peglispro (BIL) (B) were determined at the indicated time points over the duration of the clamp following low- or high-dose infusions. Blood glucose levels and glucose infusion rates during the clamps were determined following low-dose infusion (C, E) or highdose infusion (D, F) of BIL and GL. GIR = glucose infusion rate; SD = standard deviation.

5 MUDALIAR ET AL. TABLE peglispro Insulin concentrations and glucodynamic parameters following low- and high-dose infusion of insulin glargine and basal insulin Low dose High dose GL ( mu/ m /min) BIL (. mu/min) P-value GL ( mu/m /min) BIL (7. mu/min) P-value Parameter N = N = N = N = Arithmetic mean standard deviation GL or BIL concentration (pmol/l) EHIC (pmol/l) Statistical analysis of glucodynamic parameters Least squares mean standard error (ratio of LS means for GL/BIL) EGP and GDR (mg/kg/min) EGP (mg/kg/min)..9 (.7) (.) EGP/EHIC (μg/kg/min/pmol/l).. (.) (.)...9 % EGP suppression GDR (mg/kg/min).7. (.)..... (.8)... GDR/EHIC (μg/kg/min/pmol/l)..8 (.)..8.. (.) (.)..8 <. GDR fold change from baseline.9. (.) (.)...8 EGP/GDR.8. (.7) (.87) EHIC-corrected 9.8. (.7) GDR/EGP Abbreviations: BIL = basal insulin peglispro; EGP = endogenous glucose production; EHIC = equivalent human insulin concentration; GDR = glucose disposal rate; GL = insulin glargine; LS = least squares. The P-values for comparison of GL and BIL are comparing low dose vs low dose and high dose versus high dose (differences in LS means). was considered to be the start of infusion; endpoint for EGP and GDR was the average of the last time points ( and 8 hours for all clamps except high dose BIL, which was 8 and hours). Statistical model: parameter = treatment + period + sequence + subject + error, where subject was fitted as a random effect. Statistical model: parameter = baseline + treatment + period + sequence + subject + error, where subject was fitted as a random effect. Following high-dose administration, the majority of EGP values were and GDR/EGP ratios were not calculated. N = ; data from one subject was excluded because of an extremely high and EHIC-corrected GDR/EGP due to a low EGP value. GDR was compared at low [. vs. μg/kg/min/pmol/l (ratio.); P =.] and high doses [. vs. μg/kg/min/pmol/l (ratio.); P <.; Table, Figure E,F]. The GDR fold change from baseline was also higher following treatment with GL compared to BIL at both low and high doses (Table ). Following low-dose infusions, the EHIC-corrected GDR/EGP was.7-fold greater (9.8 vs 7., P =.8) for GL compared to BIL. Following both GL and BIL high-dose infusions, concentrations of FFA and glycerol at endpoint (Figure B) were completely suppressed and similar; however, for the low-dose infusions, FFA and glycerol concentrations at endpoint (Figure A) were significantly higher (P <.) with BIL compared to GL. Low-dose BIL infusion resulted in an increase in FFA from baseline, but no statistically significant change in glycerol from baseline, in contrast to low-dose GL infusion, which resulted in a decrease for both FFA and glycerol. As a result, the change from baseline of both FFA and glycerol were significantly different following BIL infusion compared to GL infusion (Figure C). DISCUSSION The primary aim of this study was to compare the GDR achieved with GL to the GDR achieved with BIL at doses that partially suppress EGP and at doses that fully suppress EGP in patients with TD. In a dose-finding study using similar techniques in healthy subjects, we demonstrated that BIL had reduced effects on glucose uptake with similar effects on glucose production compared to GL. 8 In addition, reduced peripheral effects of BIL have been demonstrated in an acute dog model 7 under conditions where the design sought to achieve similar GIRs between insulins and, in contrast to the design of the human study, which sought to match suppression of EGP. When comparing BIL and GL, there are qualitative and quantitative differences on both glucodynamic and lipodynamic effects to consider. BIL compared with GL in the Phase trials resulted in improved glycemic control, lower nocturnal hypoglycemia and less weight gain (or weight loss in patients with TD), but greater triglyceride, FFA and hepatic fat levels.,, Given the pleiotropic effects of insulin, examining glucose production, glucose disposal and lipolysis in this acute study may provide insight into the differences observed in the Phase trials. The low-dose infusions of both insulins resulted in serum concentrations in their respective therapeutic range. 8 Although the percentage suppression of EGP was statistically different between insulins, this measure is dependent on baseline EGP. The baseline EGP in this study is not necessarily physiologic as it is dependent on the peripheral infusion of human insulin that results in a state of relative peripheral hyperinsulinism in patients with TD and is not analogous to the

6 MUDALIAR ET AL. (A) EGP (mg/kg/min)..... Low Insulin Doses mu/m /min GL. mu/min BIL (B)..... High Insulin Doses mu/m /min GL 7. mu/min BIL (C) GDR (mg/kg/min) mu/m /min GL. mu/min BIL (D) mu/m /min GL 7. mu/min BIL (E) mg/kg/min TIme After Start of Infusion (hour) (F) Low GL High GL Low BIL High BIL p=. p=.77 p=.97 p=. µg/kg/min/pmol/l Low GL High GL Low BIL High BIL p<. p=. p=. p=.9 p=.88 p=.9 EGP GDR EGP GDR EGP/ EHIC GDR/ EHIC EGP/ EHIC GDR/ EHIC FIGURE Endogenous glucose production and glucose disposal rate following low- or high-dose infusion of basal insulin peglispro and insulin glargine. The endogenous glucose production (EGP) and glucose disposal rate (GDR) were measured over time after infusion of low (A, C) or high (B, D) insulin doses. Line plots represent the mean standard deviation for both parameters. (E) EGP and GDR at baseline and endpoint, where baseline was the -hour time point (start of infusion) and endpoint was the average of the last time points of the clamp period. (F) EGP and GDR were corrected for endogenous human insulin concentration (EHIC) to correct for differences in receptor binding affinity between the insulins. The P-values for each basal insulin peglispro (BIL)-insulin glargine (GL) comparison are indicated above the bars. Bars represent least squares mean standard error. fasting state in healthy subjects., In addition this insulin infusion was discontinued at least hour prior to the clamp and may have contributed to non-steady state conditions. For these reasons, the endpoint data are considered the best representation of the comparative effects of both insulins. GDR, EHIC-corrected endpoint GDR, GDR fold change from baseline, and endpoint EHIC-corrected GDR/EGP were all significantly higher with GL compared with BIL at the low dose, consistent with previous studies. 7,8

7 MUDALIAR ET AL. (A) p=. Low Insulin Doses p<. Low GL Low BIL (B) p=. High Insulin Doses High GL High BIL mg/dl mg/dl p=.9 p<. p=.7 p=. p=.9 Change from (mg/dl) Free Fatty Acids Glycerol (C) Low GL p<. Low BIL p<. Free Fatty Acids Glycerol (D) Change from (mg/dl) Free Fatty Acids Glycerol High GL High BIL p=.9 p=.8 Free Fatty Acids Glycerol FIGURE Analysis of free fatty acids and glycerol following low- or high-dose insulin infusion. Free fatty acid and glycerol concentrations were measured in serum samples at baseline and endpoint following low-dose (A) or high-dose (B) insulin infusion. Changes from baseline at endpoint of free fatty acids and glycerol following low-dose (C) and high-dose (D) insulin infusion are also shown. was the -hour time point (start of infusion) and endpoint was the last time point of the infusion period. The P-values for each comparison are indicated above the bars. Bars represent least squares mean standard error. Low GL = mu/m /min; low basal insulin peglispro (BIL) =. mu/min; high insulin glargine (GL) = mu/m /min; high BIL = 7. mu/min. Following low-dose infusion of BIL and GL, endpoint FFA and glycerol were higher with BIL suggesting less suppression of lipolysis with BIL. These data are consistent with the dog model study 7 and the observation on lipolysis noted with portal administration of human insulin compared with peripheral administration which resulted in a relative peripheral hyperinsulinism. The observed higher levels of lipolysis with BIL in this acute study may explain the higher levels of triglycerides and hepatic fat noted in the IMAGINE Phase trials of BIL.,, Following high-dose infusion of BIL and GL, FFA and glycerol levels were suppressed to a similar extent suggesting that a full peripheral effect on lipid metabolism was achieved at the higher doses. The high-dose infusions of both insulins completely suppressed EGP, and GL concentrations were within the therapeutic range, whereas BIL concentrations were in the supratherapeutic range. Although BIL EHIC was approximately fivefold greater than that of GL at the higher dose, BIL GDR was lower than GL GDR, consistent with the low-dose comparison. The EHIC-corrected GDR/EGP of GL compared with BIL was.7 following low-dose infusions, indicating that for approximately the same effect on EGP (EGP/EHIC ratio of GL compared with BIL of.), GL had nearly twice the effect of BIL on the periphery. The EGP/GDR ratio of GL compared to BIL at the low dose was.7 in the current study, consistent with a ratio of approximately. in the insulinopenic dog study. 7 This study was challenged by the need to accurately and prospectively predict an equivalent suppression of EGP between comparator insulins. Using an identical normalized unit insulin dose would not provide comparable results as the volume of distribution is very different for the two insulins when administered IV. 7 In addition, circulating GL and BIL concentrations are very different providing further challenges for comparison of dosing. Despite the assessment of the dose response (five doses of BIL and two doses of GL) in healthy subjects, 8 it remained difficult to more precisely match circulating levels of EHIC-corrected levels of BIL and GL in this study. These challenges were largely mitigated by normalizing circulating BIL and GL concentrations to human insulin concentration using the insulin receptor inhibitory binding constants.

8 MUDALIAR ET AL. The design of the current study had specific limitations. Currently available assay methodology for FFA and glycerol limit interpretation of lipolysis data, as many of these measurements were below the LLOQ, and it was not possible to establish a dose-response relationship when there was complete suppression of FFA and glycerol. EGP and GDR are assumed to be equal at baseline when steady state is assumed. In patients with diabetes, however, these processes are unlikely to be at steady state and this may be a potential source of error. The use of a crossover design should limit the impact of this source of variability. Lastly, the GDR observations for BIL high-dose infusion may be limited, as interstitial steady state may not have been reached after hours, resulting in under-estimation of the effect of BIL on GDR. In conclusion, BIL restored the hepato-peripheral insulin action gradient seen in normal physiology compared to GL. Combined with previous results demonstrating hepato-preferential activity of BIL in healthy subjects 8 and in an acute dog model of insulinopenic diabetes, 7 the data suggests that BIL may provide a portal-peripheral insulin action gradient that is absent with most insulin replacement therapies via its peripherally restricted action on key target tissues involved in carbohydrate and lipid metabolism. ACKNOWLEDGEMENTS The authors wish to acknowledge the subjects who participated in this study, as well as the following individuals: Jane Pinaire, PhD, of Eli Lilly and Company, for project management, writing, and graphics support, Jamie Scism-Bacon, PhD, of Eli Lilly and Company, for project management support, and Jay Tichelaar, PhD for providing medical writing services on behalf of Covance, Inc. REFERENCES. Eaton RP, Allen RC, Schade DS, Standfer JC. Normal insulin secretion: the goal of artificial insulin delivery systems? Diabetes Care. 98;:7 7.. De Vos P, De Haan BJ, Vegter D, et al. Insulin levels after portal and systemic insulin infusion differ in a dose-dependent fashion. Horm Metab Res. 998;:7 7.. Meier JJ, Veldhuis JD, Butler PC. Pulsatile insulin secretion dictates systemic insulin delivery by regulating hepatic insulin extraction in humans. Diabetes. ;:9.. Edgerton DS, Lautz M, Scott M, et al. Insulin s direct effects on the liver dominate the control of hepatic glucose production. J Clin Invest. ;: 7.. Herring R, Jones RH, Russell-Jones DL. Hepatoselectivity and the evolution of insulin. Diabetes Obes Metab. ;: 8.. Gregory JM, Kraft G, Scott MF, et al. Insulin delivery into the peripheral circulation: a key contributor to hypoglycemia in type diabetes. Diabetes. ;:9. 7. Sinha VP, Choi SL, Soon DK, et al. Single-dose pharmacokinetics and glucodynamics of the novel, long-acting basal insulin LY in healthy subjects. J Clin Pharmacol. ;: Bergenstal RM, Rosenstock J, Arakaki RF, et al. A randomized, controlled study of once daily LY, a novel long-acting basal insulin, versus insulin glargine in basal insulin-treated patients with type diabetes. Diabetes Care. ;: Rosenstock J, Bergenstal RM, Blevins TC, et al. Better glycemic control and weight loss with the novel long-acting basal insulin LY compared with insulin glargine in type diabetes: a randomized, crossover study. Diabetes Care. ;: 8.. Garg S, Dreyer M, Jinnouchi H, et al. A randomized clinical trial comparing basal insulin peglispro versus insulin glargine, in combination with prandial insulin lispro, in patients with type diabetes: IMAGINE study. Diabetes Obes Metab. ;8(S):.. Davies MJ, Russell-Jones D, Selam J-L, et al. Double-blind comparison of basal insulin peglispro and insulin glargine in insulin-naive patients with type diabetes: the IMAGINE clinical trial. Diabetes Obes Metab. ;8:.. Bergenstal RM, Lunt H, Franek E, et al. A randomized, double-blind clinical trial comparing basal insulin peglispro and insulin glargine, in combination with prandial insulin lispro, in patients with type diabetes: IMAGINE. Diabetes Obes Metab. ;8: Blevins T, Pieber TR, Vega G, Zhang S, Bastyr E Jr, Chang AM. A randomized, double-blind clinical trial comparing basal insulin peglispro and insulin glargine, in combination with prandial insulin lispro, in patients with type diabetes: IMAGINE. Diabetes Obes Metab. ;8: Buse JB, Rodbard HW, Trescoli-Serrano C, et al. Randomized clinical trial comparing basal insulin peglispro and insulin glargine in patients with type diabetes previously treated with basal insulin: IMAGINE. Diabetes Care. ;9:9.. Grunberger G, Chen L, Rodriguez A, Tinahones FJ, Jacober SJ, Bue-Valleskey J. A randomized clinical trial of basal insulin peglispro versus NPH in insulin-naive patients with type diabetes: IMAGINE. Diabetes Obes Metab. ;8(S):.. Jacober SJ, Rosenstock J, Bergenstal RM, Prince MJ, Qu Y, Beals JM. Contrasting weight changes with LY, a novel long-acting insulin, and insulin glargine despite similar improved glycaemic control in TDM and TDM. Diabetes Obes Metab. ;:. 7. Moore MC, Smith MS, Sinha VP, et al. Novel PEGylated basal insulin LY has a preferential hepatic effect on glucose metabolism. Diabetes. ;:9. 8. Henry RR, Mudaliar S, Ciaraldi TP, et al. Basal insulin peglispro demonstrates preferential hepatic versus peripheral action relative to insulin glargine in healthy subjects. Diabetes. ;(Suppl. ):A. 9. Mudaliar S, Mohideen P, Deutsch R, et al. Intravenous glargine and regular insulin have similar effects on endogenous glucose output and peripheral activation/deactivation kinetic profiles. Diabetes Care. ;:97.. Finegood DT, Bergman RN, Vranic M. Estimation of endogenous glucose production during hyperinsulinemic-euglycemic glucose clamps. Comparison of unlabeled and labeled exogenous glucose infusates. Diabetes. 987;:9 9.. Owens RA, Hansen RJ, Kahl SD, et al. In vivo and in vitro characterization of basal insulin peglispro: a novel insulin analogue. J Pharmacol Exp Ther. ;7:9.. Cusi K, Sanyal A, Zhang S, Hoogwerf BJ, Chang AM. Different effects of basal insulin peglispro and insulin glargine on liver enzymes and liver fat content in patients with type and type diabetes. Diabetes Obes Metab. ;8(S): 8.. Ginsberg H, Cariou B, Orchard T, et al. Lipid changes during basal insulin peglispro, insulin glargine, or NPH treatment in IMAGINE trials. Diabetes Obes Metab. ;8:88 9. How to cite this article: Mudaliar S, Henry RR, Ciaraldi TP, Armstrong DA, Burke PM, Pettus JH, Garhyan P, Choi SL, Knadler MP, Lam ECQ, Prince MJ, Bose N, Porksen NK, Sinha VP, Linnebjerg H and Jacober SJ. Reduced peripheral activity leading to hepato-preferential action of basal insulin peglispro compared with insulin glargine in patients with type diabetes, Diabetes Obes Metab, 8 (Suppl. ), 7. DOI:./dom.7