Clinical Pharmacology Applications of Real-World Data and Real-World Evidence in Drug Development and Approval-An Industry Perspective.

Since the 21st Century Cures Act was signed into law in 2016, real-world data (RWD) and real-world evidence (RWE) have attracted great interest from the healthcare ecosystem globally. The potential and capability of RWD/RWE to inform regulatory decisions and clinical drug development have been extensively reviewed and discussed in the literature. However, a comprehensive review of current applications of RWD/RWE in clinical pharmacology, particularly from an industry perspective, is needed to inspire new insights and identify potential future opportunities for clinical pharmacologists to utilize RWD/RWE to address key drug development questions. In this paper, we review the RWD/RWE applications relevant to clinical pharmacology based on recent publications from member companies in the International Consortium for Innovation and Quality in Pharmaceutical Development (IQ) RWD Working Group, and discuss the future direction of RWE utilization from a clinical pharmacology perspective. A comprehensive review of RWD/RWE use cases is provided and discussed in the following categories of application: drug-drug interaction assessments, dose recommendation for patients with organ impairment, pediatric plan development and study design, model-informed drug development (e.g., disease progression modeling), prognostic and predictive biomarkers/factors identification, regulatory decisions support (e.g., label expansion), and synthetic/external control generation for rare diseases. Additionally, we describe and discuss common sources of RWD to help guide appropriate data selection to address questions pertaining to clinical pharmacology in drug development and regulatory decision making.

Dose Rationale of Nasal Glucagon in Japanese Pediatric Patients with Diabetes Using Pharmacokinetic/Pharmacodynamic Modeling and Simulation.

BACKGROUND: Nasal glucagon (NG) 3 mg is approved in Japan to treat hypoglycemia in pediatric patients with diabetes, but an NG clinical study has not been performed in Japanese children because of practical and ethical concerns. OBJECTIVE: The aim of this study is to support the dose rationale for NG 3 mg in Japanese pediatric patients with diabetes using modeling and simulation. METHODS: We used a pharmacokinetic/pharmacodynamic bridging approach to extrapolate the available clinical data to Japanese pediatric patients. Population pharmacokinetic/pharmacodynamic modeling was performed using data from seven clinical studies, including five studies in non-Japanese adults, one study in Japanese adults, and one study in non-Japanese pediatric patients. Simulation was then used to estimate glucagon exposure and glucose response after NG 3-mg administration for three age categories of Japanese pediatric patients: 4 to < 8, 8 to < 12, and 12 to < 18 years. Treatment success was defined as an increase in blood glucose to ≥ 70 or ≥ 20 mg/dL from nadir within 30 min after administration of NG 3 mg. Safety was assessed in relation to the predicted maximum glucagon concentration of NG 3 mg using NG clinical trial data and published data on intravenous and intramuscular glucagon. RESULTS: The data showed a rapid and robust glucose response following NG 3 mg in Japanese and non-Japanese adults and non-Japanese pediatric patients, with some differences in glucagon exposure observed across studies. The pharmacokinetic/pharmacodynamic model described the observed clinical data well, and simulations indicated that > 99% of hypoglycemic Japanese pediatric patients in all three age groups would achieve treatment success. Predicted glucose responses to NG 3 mg in Japanese pediatric patients were comparable to those of intramuscular glucagon. Maximum concentration was not associated with the occurrence and severity of common adverse events (nausea, vomiting, and headache) in NG clinical studies. Furthermore, the predicted maximum concentration in Japanese pediatric patients, despite being higher than the observed maximum concentration in NG clinical studies, was substantially lower than the observed maximum concentration of 1 mg of intravenous glucagon, without serious safety issues. CONCLUSIONS: This analysis suggests NG 3 mg has robust efficacy without serious safety concerns in Japanese pediatric patients with diabetes.

Discovery, development, and clinical proof of mechanism of LY3463251, a long-acting GDF15 receptor agonist.

GDF15 and its receptor GFRAL/RET form a non-homeostatic system that regulates food intake and body weight in preclinical species. Here, we describe a GDF15 analog, LY3463251, a potent agonist at the GFRAL/RET receptor with prolonged pharmacokinetics. In rodents and obese non-human primates, LY3463251 decreased food intake and body weight with no signs of malaise or emesis. In a first-in-human study in healthy participants, single subcutaneous LY3463251 injections showed a safety and pharmacokinetic profile supporting further clinical development with dose-dependent nausea and emesis in a subset of individuals. A subsequent 12-week multiple ascending dose study in overweight and obese participants showed that LY3463251 induced significant decreases in food intake and appetite scores associated with modest body weight reduction independent of nausea and emesis (clinicaltrials.gov: NCT03764774). These observations demonstrate that agonism of the GFRAL/RET system can modulate energy balance in humans, though the decrease in body weight is surprisingly modest, suggesting challenges in leveraging the GDF15 system for clinical weight-loss applications.

Evaluation of Insulin Lispro Pharmacokinetics and Pharmacodynamics Over 10 Days of Continuous Insulin Infusion in People With Type 1 Diabetes.

BACKGROUND: We evaluated the effect of meloxicam on insulin lispro pharmacokinetics and glucose pharmacodynamics over 10 days of continuous subcutaneous insulin infusion (CSII) at one infusion site in people with type 1 diabetes (T1D). METHOD: This phase 1, randomized, double-blind, single-center, two-way crossover study enrolled adults with T1D for ≥1 year on stable CSII for ≥3 months. Participants randomly received U100 insulin lispro and LY900027 (U100 insulin lispro + 0.25 mg/mL meloxicam). Primary end points were area under the insulin lispro curve from 0 to 5 hours (AUCIns.0-5h) after bolus administration prior to a mixed-meal tolerance test (MMTT) and maximum observed concentration of insulin lispro (CIns.max) on days 5, 7, and 10, versus day 3 (baseline). RESULTS: A total of 20 participants were randomized. Insulin absorption was accelerated for insulin lispro and LY900027 from days 1 to 7. The AUCIns.0-5h was significantly lower on day 10 versus day 3 for LY900027 (-19%) and insulin lispro (-14%); the AUCIns.0-5h did not differ significantly between treatments. The CIns.max increased with LY900027 and insulin lispro (by ~14%-23% and ~16%-51%) on days 5, 7, and 10 versus day 3. The CIns.max of LY900027 was ~14%-23% lower than insulin lispro CIns.max on days 7 and 10 (P ≤ .0805). Accelerated insulin absorption and a modest loss of total insulin exposure led to a loss of MMTT glycemic control at later time points. CONCLUSIONS: The pharmacokinetics of insulin changed over catheter wear time even when an anti-inflammatory agent was present. Postprandial glycemic control was adversely affected by the accelerated insulin absorption and decreased insulin exposure.

Optimization of trial duration to predict long-term HbA1c change with therapy: A pharmacometrics simulation-based evaluation.

Glycated hemoglobin (HbA1c) is the main biomarker of diabetes drug development. However, because of its delayed turnover, trial duration is rarely shorter than 12 weeks, and being able to predict long-term HbA1c with precision using data from shorter studies would be beneficial. The feasibility of reducing study duration was therefore investigated in this study, assuming a model-based analysis. The aim was to investigate the predictive performance of 24- and 52-week extrapolations using data from up to 4, 6, 8 or 12 weeks, with six previously published pharmacometric models of HbA1c. Predictive performance was assessed through simulation-based dose-response predictions and model averaging (MA) with two hypothetical drugs. Results were consistent across the methods of assessment, with MA supporting the results derived from the model-based framework. The models using mean plasma glucose (MPG) or nonlinear fasting plasma glucose (FPG) effect, driving the HbA1c formation, showed good predictive performance despite a reduced study duration. The models, using the linear effect of FPG to drive the HbA1c formation, were sensitive to the limited amount of data in the shorter studies. The MA with bootstrap demonstrated strongly that a 4-week study duration is insufficient for precise predictions of all models. Our findings suggest that if data are analyzed with a pharmacometric model with MPG or FPG with a nonlinear effect to drive HbA1c formation, a study duration of 8 weeks is sufficient with maintained accuracy and precision of dose-response predictions.

Development and Verification of a Body Weight-Directed Disease Trial Model for Glucose Homeostasis.

Weight loss has been associated with improvement in insulin sensitivity. It is consequently a cornerstone in the management of type 2 diabetes mellitus (T2DM). However, the strictly quantitative relationship between weight loss, insulin sensitivity, and clinically relevant glucose homeostasis biomarkers as well as changes therein as T2DM progresses is not yet fully understood. Therefore, the objective of our research was to establish a body weight-directed disease trial model for glucose homeostasis. To that end, we conducted a model-based meta-analysis using time course data of body weight loss (following lifestyle change or surgical procedure) and corresponding improvement of insulin sensitivity expressed as the Matsuda index. Changes in body weight were best described by a sigmoidal Emax model, whereas changes in the Matsuda index were best described by a linear model with a slope of 3.49. Once developed and verified, the model-based meta-analysis was linked to a disease-drug trial model for T2DM previously developed by our group to characterize and predict the impact of weight loss on clinically relevant glucose homeostasis biomarkers. The joint model was then used to conduct clinical trial simulations, which showed that weight loss can greatly improve clinically relevant glucose homeostasis biomarkers in T2DM patients.

Insulin, Not the Preservative m-cresol, Instigates Loss of Infusion Site Patency Over Extended Durations of CSII in Diabetic Swine.

Insulin infusion sets worn for more than 4-5 days have been associated with a greater risk of unexplained hyperglycemia, a phenomenon that has been hypothesized to be caused by an inflammatory response to preservatives such as m-cresol and phenol. In this cross-over study in diabetic swine, we examined the role of the preservative m-cresol in inflammation and changes in infusion site patency. Insulin pharmacokinetics (PK) and glucose pharmacodynamics (PD) were measured on delivery of a bolus of regular human insulin U-100 (U-100R), formulated with or without 2.5 mg/mL m-cresol, to fasted swine following 0, 3, 5, 7, and 10 days of continuous subcutaneous insulin infusion (CSII). In a subsequent study with the same animals, biopsies were evaluated from swine wearing infusion sets infusing nothing, saline, or U-100R either with or without 2.5 mg/mL m-cresol, following 3, 7, and 10 days of CSII. Exposure to m-cresol did not impact any PK or PD endpoints. PK and PD responses dropped markedly from Days 7-10, regardless of the presence of m-cresol. Histopathology results suggest an additive inflammatory response to both the infusion set and the insulin protein itself, peaking at Day 7 and remaining stable beyond.

Basal insulin peglispro increases lipid oxidation, metabolic flexibility, thermogenesis and ketone bodies compared to insulin glargine in subjects with type 1 diabetes mellitus.

AIMS: When treated with basal insulin peglispro (BIL), patients with type 1 diabetes mellitus (T1DM) exhibit weight loss and lower prandial insulin requirements versus insulin glargine (GL), while total insulin requirements remain similar. One possible explanation is enhanced lipid oxidation and improved ability to switch between glucose and lipid metabolism with BIL. This study compared the effects of BIL and GL on glucose and lipid metabolism in subjects with T1DM. MATERIALS AND METHODS: Fifteen subjects with T1DM were enrolled into this open-label, randomised, crossover study, and received once-daily stable, individualised, subcutaneous doses of BIL and GL for 4 weeks each. Respiratory quotient (RQ) was measured using whole-room calorimetry, and energy expenditure (EE) and concentrations of ketone bodies (3-hydroxybutyrate) and acylcarnitines were assessed. RESULTS: Mean sleep RQ was lower during the BIL (0.822) than the GL (0.846) treatment period, indicating greater lipid metabolism during the post-absorptive period with BIL. Increases in carbohydrate oxidation following breakfast were greater during BIL than GL treatment (mean change in RQ following breakfast 0.111 for BIL, 0.063 for GL). Furthermore, BIL treatment increased total daily EE versus GL (2215.9 kcal/d for BIL, 2135.5 kcal/d for GL). Concentrations of ketone bodies and acylcarnitines appeared to be higher following BIL than GL treatment. CONCLUSIONS: BIL increased sleeping fat oxidation, EE, ketone bodies, acylcarnitines and post-prandial glucose metabolism when switching from conventional insulin, thus, restoring metabolic flexibility and increasing thermogenesis. These changes may explain the previously observed weight loss with BIL versus GL.

Development and Qualification of a Drug-Disease Modeling Platform to Characterize Clinically Relevant Endpoints in Type 2 Diabetes Trials.

Type 2 diabetes mellitus (T2DM) is a chronic, progressive disease characterized by persistently elevated blood glucose concentration (hyperglycemia). We developed a mechanistic drug-disease modeling platform based on data from more than 4,000 T2DM subjects in seven phase II/III clinical trials. The model integrates longitudinal changes in clinically relevant biomarkers of glycemic control with information on baseline disease state, demographics, disease progression, and different therapeutic interventions, either when given alone or as add-on combination therapy. The model was able to simultaneously characterize changes in fasting plasma glucose, fasting serum insulin, and glycated hemoglobin A1c following administration of sulfonylurea, metformin, and thiazolidinedione as well as disease progression in clinical trials ranging from 16-104 weeks of treatment. The mechanistic components of this generalized mechanism-based platform, based on knowledge of pharmacology, insulin-glucose homeostatic feedback, and diabetes pathophysiology, allows its application to be further expanded to other antidiabetic drug classes and combination therapies.

Attenuated suppression of lipolysis explains the increases in triglyceride secretion and concentration associated with basal insulin peglispro relative to insulin glargine treatment in patients with type 1 diabetes.

AIMS: To test the hypothesis that, as well as lowering weight and increasing plasma triglyceride (TG) levels and hepatic fat compared with insulin glargine (GL) in patients with type 1 diabetes, the attenuated peripheral effects of basal insulin peglispro (BIL) may include increased free fatty acid flux to the liver, causing increased very-low-density lipoprotein (VLDL)-TG secretion and lipid oxidation, and decreased TG adipose tissue deposition. METHODS: In this open-label, randomized, 2-period crossover study, 14 patients with type 1 diabetes received once-daily, individualized, stable BIL or GL doses for 3 weeks. Palmitate flux was assessed using [9,10-3 H]palmitate infusion. VLDL-TG secretion, clearance and oxidation rate were assessed using primed-constant infusion of ex vivo labelled [1-14 C]VLDL-TG, while VLDL-TG storage rate was assessed using [9,10-3 H]VLDL-TG bolus injection. RESULTS: The VLDL-TG concentration and secretion rate, and palmitate flux were statistically significantly higher during BIL than during GL treatment (58%, 51% and 35%, respectively). The ratios of least squares (LS) geometric means for VLDL-TG clearance and oxidation were 0.92 (95% confidence interval [CI] 0.72, 1.17) and 1.31 (95% CI 0.91, 1.90), respectively. The difference in LS means for VLDL-TG storage rate was -0.36 (95% CI -0.83, 0.12). CONCLUSIONS: BIL-treated patients had higher effective lipolysis, VLDL-TG secretion and VLDL-TG concentration compared with GL-treated patients, explaining the increased plasma TG concentrations reported previously. Data support attenuated effects of BIL on lipolysis, in addition to the recently described hepato-preferential glucodynamic effects.

Dose Unit Establishment for a New Basal Insulin Using Joint Modeling of Insulin Dose and Glycemic Response.

BACKGROUND: For new insulin analogs with properties that vary from human insulin, defining activity in units of human insulin based on glycemic lowering efficacy may be challenging. Here we present a new method that can be used to quantify a unit dose of an experimental insulin when the traditional euglycemic clamp method is not adequate. METHODS: Joint modeling of insulin dose and the glycemic outcome variable hemoglobin A1c (HbA1c), where both were response variables, was used to evaluate insulin unit potency for basal insulin peglispro (BIL). The data were from the Phase 3 program for BIL, which included greater than 5500 patients with type 1 or type 2 diabetes who were treated for 26 or 52 weeks with BIL or a comparator insulin. Both basal-bolus and basal insulin only studies were included, and some type 2 diabetes patients were insulin-naïve. RESULTS: The analysis showed that 1 unit of BIL, composed of 9 nmol of active ingredient, had similar or slightly greater potency compared to 1 unit insulin glargine or NPH insulin for all populations. CONCLUSIONS: Despite some limitations, the joint modeling of HbA1c and insulin dose provides a reasonable approach to estimate the relative potency of a new basal insulin versus an established basal insulin.

Hypoglycemia Risk Related to Double Dose Is Markedly Reduced with Basal Insulin Peglispro Versus Insulin Glargine in Patients with Type 2 Diabetes Mellitus in a Randomized Trial: IMAGINE 8.

BACKGROUND: Basal insulin peglispro (BIL) has a peripheral-to-hepatic distribution of action that resembles endogenous insulin and a prolonged duration of action with a flat pharmacokinetic/pharmacodynamic profile at steady state, characteristics that tend to reduce hypoglycemia risk compared to insulin glargine (GL). The primary objective was to demonstrate that clinically significant hypoglycemia (blood glucose ≤54 mg/dL [3.0 mmol/L] or symptoms of severe hypoglycemia) occurred less frequently within 84 h after a double dose (DD) of BIL than a DD of GL. METHODS: This was a randomized, double-blind, two-period crossover study in patients with type 2 diabetes (T2D) previously treated with insulin (N = 68). For the first 3 weeks of each of the two crossover periods, patients received an individualized dose of BIL or GL once nightly (stable dose for 2 weeks/period). Then, during a 7-day inpatient stay with frequent blood glucose monitoring and standardized meals, one DD of study insulin was given. Glucose was infused if blood glucose was ≤54 mg/dL (3.0 mmol/L) or for symptoms of severe hypoglycemia. RESULTS: Within 84 h after the DD, a significantly smaller proportion of patients experienced clinically significant hypoglycemia with BIL compared to GL (BIL, 6.6%; GL, 35.5%; odds ratio for BIL/GL 0.13 [95% confidence interval 0.04-0.39]; P < 0.001). Adverse event profiles were similar for the two insulins. Serum alanine aminotransferase and triglyceride levels were significantly higher with BIL versus GL. CONCLUSIONS: BIL has a markedly lower risk of hypoglycemia than GL when replicating a double-dose error in patients with T2D.

Treatment with LY2409021, a glucagon receptor antagonist, increases liver fat in patients with type 2 diabetes.

AIMS: To evaluate whether treatment with LY2409021, a novel, selective glucagon receptor antagonist, is associated with changes in hepatic fat and other safety variables related to the benefit-risk profile for chronic use in patients with type 2 diabetes (T2D). METHODS: Safety and efficacy were assessed in patients with T2D taking metformin and sulphonylurea who were randomized to LY2409021 20 mg (n = 65), placebo (n = 68), or sitagliptin 100 mg (n = 41). Key endpoints included change from baseline to month 6 in hepatic fat fraction (HFF), assessed by magnetic resonance imaging; hepatic aminotransferases; blood pressure; lipid profile; fasting plasma glucose; and glycated haemoglobin (HbA1c). RESULTS: A significant increase in HFF was seen with LY2409021 vs sitagliptin (least squares [LS] mean difference 3.72%; P < .001) and placebo (4.44%; P < .001), accompanied by significant elevations in alanine aminotransferase levels with LY2409021 vs sitagliptin (6.8 U/L; P = .039) and vs placebo (10.7 U/L; P < .001). No patients had concomitant elevations in bilirubin levels. LY2409021 treatment showed significant HbA1c reductions vs placebo (LS mean difference -0.77%; P < .001) but not sitagliptin (-0.20%; P = .383). Similar results were observed for fasting plasma glucose. LY2409021 was also associated with significant increases in systolic blood pressure vs sitagliptin (4.9 mm Hg; P = .030) and vs placebo (4.3 mm Hg; P = .029), as well as significant increases in body weight and total cholesterol. All effects of LY2409021 were reversible. CONCLUSION: In this cohort of patients with T2D, chronic glucagon receptor antagonism with LY2409021 was associated with glucose-lowering but also demonstrated increases in hepatic fat, hepatic aminotransferases, and other adverse effects.

Quantification of Basal Insulin Peglispro and Human Insulin in Adipose Tissue Interstitial Fluid by Open-Flow Microperfusion.

BACKGROUND: Restoration of the physiologic hepatic-to-peripheral insulin gradient may be achieved by either portal vein administration or altering insulin structure to increase hepatic specificity or restrict peripheral access. Basal insulin peglispro (BIL) is a novel, PEGylated basal insulin with a flat pharmacokinetic and glucodynamic profile and altered hepatic-to-peripheral action gradient. We hypothesized reduced BIL exposure in peripheral tissues explains the latter, and in this study assessed the adipose tissue interstitial fluid (ISF) concentrations of BIL compared with human insulin (HI). METHODS: A euglycemic glucose clamp was performed in patients with type 1 diabetes during continuous intravenous (IV) infusion of BIL or HI, while the adipose ISF insulin concentrations were determined using open-flow microperfusion (OFM). The ratio of adipose ISF-to-serum concentrations and the absolute steady-state adipose ISF concentrations were assessed using a dynamic no-net-flux technique with subsequent regression analysis. RESULTS: Steady-state BIL concentrations in adipose tissue ISF were achieved by ∼16 h after IV infusion. Median time to reach steady-state glucose infusion rate across doses ranged between 8 and 22 h. The average serum concentrations (coefficient of variation %) of BIL and HI were 11,200 pmol/L (23%) and 425 pmol/L (15%), respectively. The ISF-to-serum concentration ratios were 10.2% for BIL and 22.9% for HI. CONCLUSIONS: This study indicates feasibility of OFM to measure BIL in ISF. The observed low ISF-to-serum concentration ratio of BIL is consistent with its previously demonstrated reduced peripheral action.

Treatment with the glucagon receptor antagonist LY2409021 increases ambulatory blood pressure in patients with type 2 diabetes.

AIMS: To assess the effect of LY2409021 on systolic blood pressure (SBP) in patients with type 2 diabetes. MATERIALS AND METHODS: This 6-week, randomized, crossover study evaluated the effects of once-daily administration of LY2409021 20 mg vs those of placebo on SBP, diastolic BP (DBP), and mean arterial pressure (MAP) using 24-hour ambulatory BP monitoring (ABPM) in 270 subjects treated with diet/exercise ± metformin. Other measures included changes in glycemic control, serum lipids, and hepatic safety markers. RESULTS: At 6 weeks of LY2409021 treatment, 24-hour mean SBP was increased, with a least squares mean (LSM) difference of 2.26 mm Hg vs placebo (95% CI: 1.11, 3.40; P < .001). The 24-hour mean DBP and MAP also increased, with LSM differences of 1.37 mm Hg (95% CI: 0.66, 2.08; P < .001) and 1.67 mm Hg (95% CI: 0.86, 2.47; P < .001) vs placebo, respectively. At week 6, LY2409021 treatment reduced glycated hemoglobin (HbA1c) levels, with an LSM difference of -0.49% (-5.4 mmol/mol) (95% CI: -0.56%, -0.42% [-6.1, -4.6 mmol/mol]; P < .001) vs placebo. Mean HbA1c at baseline was 7.3% (56 mmol/mol). Small but significant changes in serum lipid and aminotransferase levels were observed with LY2409021 treatment (all P < .05 vs placebo). CONCLUSIONS: Statistically significant increases in BP, MAP and serum lipid levels were observed with LY2409021 treatment at a dose that lowered HbA1c and glucose levels. These effects may limit the clinical utility of LY2409021 as a chronic treatment for type 2 diabetes.

Novel hepato-preferential basal insulin peglispro (BIL) does not differentially affect insulin sensitivity compared with insulin glargine in patients with type 1 and type 2 diabetes.

AIMS: Basal insulin peglispro (BIL) is a novel PEGylated basal insulin with a flat pharmacokinetic and glucodynamic profile and reduced peripheral effects, which results in a hepato-preferential action. In Phase 3 trials, patients with T1DM treated with BIL had lower prandial insulin requirements, yet improved prandial glucose control, relative to insulin glargine (GL). We hypothesized that this may be because of an enhanced sensitivity to prandial insulin with BIL resulting from lower chronic peripheral insulin action. MATERIALS AND METHODS: Two open-label, randomized, 2-period crossover clinical studies were conducted in 28 patients with T1DM and 24 patients with T2DM. In each study period, patients received once-daily, individualized, stable, subcutaneous doses of BIL or GL for 5 weeks before a euglycaemic 2-step hyperinsulinemic clamp procedure (with [6,6- 2 H2 ]-glucose in 12 of the patients with T1DM). M-values were derived from the clamp procedure for all patients, with rate of glucose appearance (Ra) and disappearance (Rd) and insulin sensitivity index (SI) determined from the clamps with [6,6- 2 H2 ]-glucose. RESULTS: There were no statistically significant differences between BIL and GL in key measures of hepatic (% Ra suppression during the low-dose insulin infusion; 78.7% with BIL, 81.8% with GL) or peripheral (M-value and M/I during the high-dose insulin infusion, Rd and SI) insulin sensitivity in patients with T1DM or T2DM. CONCLUSIONS: The need to reduce prandial insulin observed with BIL during phase 3 trials cannot be explained by the differential effects of BIL and GL on sensitivity to prandial insulin in either T1DM or T2DM.

A Comprehensive Review of Novel Drug-Disease Models in Diabetes Drug Development.

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disease, which affects millions of people worldwide. The disease is characterized by chronically elevated blood glucose concentrations (hyperglycaemia), which result in comorbidities and multi-organ dysfunction. This is due to a gradual loss of glycaemic control as a result of increasing insulin resistance, as well as decreasing ß-cell function. The objective of T2DM drug interventions is, therefore, to reduce fasting and postprandial blood glucose concentrations to normal, healthy levels without hypoglycaemia. Several classes of novel antihyperglycaemic drugs with various mechanisms of action have been developed over the past decades or are currently under clinical development. The development of these drugs is routinely supported by the application of pharmacokinetic/pharmacodynamic modelling and simulation approaches. They integrate information on the drug's pharmacokinetics, clinically relevant biomarker information and disease progression into a single, unifying approach, which can be used to inform clinical study design, dose selection and drug labelling. The objective of this review is to provide a comprehensive overview of the quantitative approaches that have been reported since the 2008 review by Landersdorfer and Jusko in an increasing order of complexity, starting with glucose homeostasis models. Each of the presented approaches is discussed with respect to its strengths and limitations, and respective knowledge gaps are highlighted as potential opportunities for future drug-disease model development in the area of T2DM.