Remember me
The aim of the present study was to compare the PK and GD variables of URLi in healthy subjects after SC injection into the abdomen, upper arm, or thigh, or after intravenous (IV) injection.
Participants and MethodsStudy DesignThis was a Phase I, open-label, 4-period, randomized, crossover, up to 10-hour euglycemic clamp study in healthy subjects. The study was conducted at a single center (Lilly-NUS Centre for Clinical Pharmacology, Singapore) in accordance with the principles of the Declaration of Helsinki (2000), the International Conference on Harmonisation Guidelines for Good Clinical Practice, and guidelines on bioavailability trials., Institutional review board approval and written informed consent from all subjects were obtained before any evaluations or study procedures. This study was registered at ClinicalTrials.gov as NCT03232983.An overview of the study design is outlined in Figure 1. Patients were randomized to 1 of 4 treatment sequences comprising single doses of 15 U of U100 URLi (Eli Lilly and Company) administered subcutaneously into the abdomen, thigh, or upper arm (deltoid), or by IV injection. The study included a screening period (≤28 days) followed by 4 inpatient treatment periods (3 days each) and a ≥14-day follow-up period. At least 3 days of washout occurred between each treatment period.Figure 1Study design. The trial consisted of 6 visits: screening (≤28 days before enrollment), 4 treatment visits for dosing and clamp procedures (periods 1–4) with a washout period of at least 3 days between visits, and follow-up visit with a washout period of at least 14 days after the last dose. A total of 24 of 28 subjects completed all study periods. Single doses of 15 U ultra rapid lispro (URLi) were administered either subcutaneously in the abdomen, thigh, or upper arm, or intravenously. An up to 10-hour euglycemic clamp procedure was performed after each dose. IV = intravenous; SC = subcutaneous.
ParticipantsEligible participants were overtly healthy men and women based on medical history, aged 21 to 65 years, body weight ≥45 kg, and body mass index 18.0 to 30.0 kg/m2. The main exclusion criteria were smoking, history of any medical or psychiatric illness, abnormal cardiac parameters or vital signs deemed clinically relevant by the investigator, allergies to any components of URLi, or use of prescription or nonprescription medication (apart from vitamin/mineral supplements, occasional paracetamol, thyroid replacement medication, or contraceptives).
Bioanalytical MethodsSerum-free insulin lispro was analyzed by using a validated ELISA specific for insulin lispro (Charles River Laboratories Montreal, Senneville, Quebec, Canada). Blood samples for PK analysis were collected at time 2.5, 5, 10, 15, 20, 25, 30, 40, 60, 70, 90, 120, 150, 180, and 210 minutes, and every 60 minutes from 240 to 600 minutes after drug administration. The lower limit of quantitation was 50.0 pg/mL (8.6 pmol/L), and the inter-assay accuracy (percent relative error) and inter-assay precision (percent relative SD) were ≤12%. Quantification of insulin lispro was not affected by the presence of lipemic serum, hemolyzed serum, treprostinil (1 ng/mL), or human insulin (1722 pmol/L).
Euglycemic Clamp ProcedureAt each dosing visit, subjects fasted for at least 8 hours before administration of URLi. After administration, subjects underwent a euglycemic clamp procedure for up to 10 hours. Baseline fasting blood glucose levels were calculated for each subject as the mean of blood glucose concentrations at 10, 20, and 30 minutes before URLi administration. After drug administration, the glucose infusion was initiated at the time that the blood glucose dropped by 5 mg/dL (0.3 mmol/L) below the individual subject's fasting baseline, which was then used as the target blood glucose level for the euglycemic glucose clamp procedure. The time when the clamp was initiated was defined as the onset of insulin action. The clamp procedure was not performed in subjects with baseline blood glucose targets ≤63 mg/dL (3.5 mmol/L). After onset of insulin action, the glucose infusion rate (GIR) was adjusted manually to maintain the predetermined target blood glucose concentration for each individual subject. Blood glucose was maintained at target by variable infusion of IV 20% d-glucose (dextrose) solution. The GIR was recorded throughout the clamp procedure by the clamp operator and reflected the activity of the administered dose of insulin. The clamp operator was blinded to randomization. Blood samples were collected and measured for blood glucose every 2.5 minutes for the first 30 minutes, every 5 minutes from 30 to 120 minutes, every 10 minutes from 120 to 480 minutes, and every 20 minutes from 480 to 600 minutes after URLi injection.
Glucose levels were measured by using glucose analyzers (YSI 23000 STAT Plus Glucose and Lactate Analyzer, YSI Inc, Yellow Springs, OH, USA). The clamp was discontinued if the GIR fell to 0 for at least 30 minutes.
The variability (%CV) in the glucose target during the euglycemic clamp was consistent between treatment arms and similar to data reported for short-acting insulins (Supplemental Table 1).23Heise T Zijlstra E Nosek L Heckermann S Plum-Mörschel L Forst T. Euglycaemic glucose clamp: what it can and cannot do, and how to do it.PK AnalysesPK analyses were conducted by using standard noncompartmental methods of analysis with Phoenix version 7.0 (Certara USA Inc, Princeton, NJ, USA) and S-PLUS version 8.2 (TIBCO Software, Palo Alto, CA, USA).
Free serum insulin lispro concentrations were used to calculate PK parameters, including Cmax, AUC0–∞, and Tmax. Insulin lispro absorption was characterized by time to early half-maximal drug concentration (early 50% Tmax), and the onset of appearance, defined as time that serum insulin lispro reached the lower limit of quantitation. The determination of onset of appearance used a linear interpolation between the time of dosing (zero insulin lispro concentration) and the time of the first quantifiable insulin lispro.
The absolute bioavailability was calculated by using the ratio of the insulin lispro AUC0–∞ after SC injection into the abdomen, or upper arm, or thigh to the insulin lispro AUC0–∞ after IV administration.
GD AnalysesGD parameters were derived from the GIR during the glucose clamp procedure, using Phoenix version 6.4 and S-PLUS version 8.2. A locally weighted scatterplot smoothing function using a span of 0.2 was applied to all individual GIR versus time profiles in each treatment group using S-PLUS software version 8.2. The fitted data for each subject were used to calculate the following GD parameters: maximum GIR (Rmax), time of maximum GIR (TRmax), total amount of glucose infused over the duration of the clamp procedure (Gtot), total amount of glucose infused over the first 30 minutes (Gtot[0–30min]), and total amount of glucose infused over 1 hour (Gtot[0–1h]). Time to onset of insulin action was based on the raw observed GIR data.
Safety and Tolerability AssessmentsSafety and tolerability assessments included adverse events, clinical laboratory parameters, vital signs, ECGs, and hypoglycemic events. In addition, injection site assessment (erythema, pain, induration, edema, and itching) was performed at protocol-defined time points: immediately after injection, and 1, 4, and 10 hours after injection. Any injection site reaction findings were also recorded as adverse events.
Statistical AnalysisGeneral ConsiderationsUnless otherwise specified, testing for significance was done at an α level of 0.1 with two-sided CIs. Statistical significance was claimed if the P value of a test was <0.1. For selected analyses in which no P values were calculated, statistical significance was claimed if the corresponding 90% CI did not contain 0 (difference between treatment groups) or did not contain 1 (ratio of treatment groups).
Sample Size ConsiderationsA sample size of 22 completing subjects was estimated to provide the two-sided 90% CIs of the ratios of geometric means for AUC0–∞ after SC injection into the thigh and upper arm, compared with the abdomen, to be within ∼0.8 to 1.25 when the observed ratio is 1. This calculation is based on the assumption of a log-normal distribution and an estimate of intrasubject log-scale SD of 0.2.
PK and GD Statistical MethodsAll relevant PK and GD parameters were log transformed, except for the time parameters. For log-transformed end points AUC0–∞, Cmax, Gtot, and Rmax, geometric least squares means (LSM), ratios of geometric LSM, and their corresponding 90% CIs were estimated for each injection site using the statistical model that includes injection site, period, and sequence as fixed effects, and subject within sequence as a random effect. Data from the upper arm, thigh, and abdomen injection sites were included in the same model. The same model without log transformation was used for the analysis of the time parameters Tmax and TRmax. The LSM, injection site differences in LSM, and the corresponding 90% CIs for the injection site differences were estimated from the model. The ratios between injection sites and 90% CIs for the ratios were calculated by using Fieller's theorem. A nonparametric approach was used to test the median of injection site differences for time parameters using the Wilcoxon signed rank test and approximate 90% CIs.
ResultsDemographic CharacteristicsA total of 28 healthy male subjects, aged between 24 and 63 years (mean [SD], 39.8 [8.6] years) and with a mean (SD) body mass index of 24.3 (3.08) kg/m2, participated in this study. All subjects were Asian.
Twenty-four subjects completed the study. Four subjects discontinued the study: one due to subject's decision, one after experiencing redness and swelling over the cannulation site (physician's decision), one after cannulation issues (physician's decision), and one due to work commitments. No subject discontinued the study due to an adverse event considered related to the study treatment.
PK ParametersThe mean insulin lispro concentration versus time profiles after SC injection into the abdomen and upper arm are superimposable, whereas the profile was marginally broader after injection into the thigh (Figure 2). Overall insulin lispro exposure (AUC0–∞) was consistent between all 3 injection sites. The 90% CI for the ratio of AUC0–∞ mean values between (thigh or upper arm) and the abdomen included 1 and were contained within 0.8 to 1.25 for both comparisons (Table I). Cmax was similar between the abdomen and upper arm but slightly lower (17%–20%) for the thigh. The 90% CI treatment ratio between the upper arm and abdomen for Cmax was contained within 0.8 to 1.25; however, the lower limit of the 90% CI of the ratio between thigh and abdomen was 0.737 and fell outside the 0.8 to 1.25 limits. The median (minimum–maximum) time of Cmax (Tmax) occurred at 0.67 hour (0.25–2.5 hours) for the abdomen, 0.67 hour (0.25–2.0 hours) for the upper arm, and 1.5 hours (0.17–2.5 hours) for the thigh.Figure 2Arithmetic mean (SE) serum concentration versus time profiles for single subcutaneous injections of 15 U ultra rapid lispro into the abdomen, upper arm, or thigh of healthy subjects.
Table IStatistical comparison between subcutaneous injection sites (upper arm/abdomen and thigh/abdomen).
Gtot = total amount of glucose infused over the duration of the clamp procedure; Rmax = maximum glucose infusion rate.
Mean treatment ratios (%CV) for AUC0–∞ for the abdomen, upper arm, or thigh compared with IV administration were 0.65 (18), 0.65 (14), and 0.64 (19), respectively. Thus, absolute bioavailability of URLi was ∼65% after SC administration in all 3 injection regions. The observed mean PK time profiles after IV administration or SC administration into the abdomen, upper arm, and thigh of URLi are shown in Supplemental Figure 1.
GD ResponsesAfter SC injection of URLi into the abdomen, thigh, or upper arm, the mean GIR profiles were comparable (Figure 3). The maximum GIR (Rmax) and total insulin action (Gtot) were similar across injection sites with the 90% CI for the ratio of means including 1 and contained within 0.8 to 1.25 for both comparisons between (thigh or upper arm) and the abdomen (Table I). The median (minimum–maximum) TRmax occurred at 96 minutes (30–216 minutes) for the abdomen, 123 minutes (36–228 minutes) for the upper arm, and 168 minutes (30–234 minutes) for the thigh.Figure 3Arithmetic mean locally weighted scatterplot smoothing–fitted glucose infusion rate profiles during euglycemic clamps for single SC injections of 15 U ultra rapid lispro into the abdomen, upper arm, or thigh of healthy subjects.
Early Insulin Lispro Exposure and Insulin ActionThe insulin lispro concentration time profiles for the different injection sites overlapped during the first 30 minutes after dosing (Figure 2). The time to early half-maximal drug concentration (early 50% Tmax) across injection sites was similar, occurring at ∼10 minutes after SC injection (Table II). The median onset of appearance was 0.79 minute for the abdomen, 0.48 minute for the upper arm, and 1.16 minutes for the thigh. Although there were differences in the medians between injection sites, the range for the onset of appearance for these injection sites overlapped. Thus, the average onset of appearance of insulin lispro across all injection sites following URLi SC injection was ∼1 minute.Table IISummary of the pharmacokinetic parameters for insulin absorption and glucodynamic parameters for early insulin action after a single subcutaneous injection of 15 U ultra rapid lispro into the abdomen, upper arm, or thigh of healthy subjects. Data are %CV geometric mean unless otherwise stated.
Early 50% Tmax = time to early half-maximal drug concentration; Gtot (0–30min) = total amount of glucose infused over the first 30 minutes; Gtot (0–1h) = total amount of glucose infused over the first hour; Onset of appearance = time that serum insulin lispro reached the lower level of quantification; Tonset = time to onset of insulin action.
The early insulin action was also similar between SC injection sites, as shown by the overlap of the GIR during the first hour after injection (Figure 3). Correspondingly, the onset of insulin action occurred at approximately the same time (abdomen, 22.8 minutes; upper arm, 22.8 minutes; and thigh, 19.8 minutes). In addition, the amount of glucose infused in the first 30 minutes (Gtot [0-30min]) and in the first hour (Gtot [0-1h]) of the clamp was comparable across the injection sites (Table II).Endogenous Insulin Secretion (C-peptide)Mean C-peptide profiles during the euglycemic clamp were similar across the SC injection sites after a 15 U dose of URLi (Supplemental Figure 2).
Safety and TolerabilityA total of 75 treatment-emergent adverse events (TEAEs) were reported by 23 subjects. All TEAEs were mild in severity, and no serious adverse events were reported. The frequency of TEAEs was comparable between the three SC injection sites and IV administration. The majority of TEAEs were related to study procedures, mainly bruising, pain, swelling, and erythema associated with blood sampling cannulation and infusion catheter sites.
Five subjects reported TEAEs deemed to be related to study treatment: mild headache was reported by 1 subject; mild transient injection site pain was reported by 3 subjects after SC administration (abdomen and upper arm); and mild transient injection site pruritus was reported by 1 subject.
No hypoglycemic events were reported during the study. No clinically relevant alterations in laboratory parameters, vital signs, or ECGs were observed.
DiscussionThis study is the first, to the best of our knowledge, to compare the PK and GD effects of a single SC injection of URLi at different injection sites of the body. After a single SC injection of URLi into the upper arm, thigh, or abdomen, the insulin exposure and insulin action were similar across these injection sites. Importantly, the rate of absorption of insulin lispro and early insulin action of URLi were maintained regardless of SC injection site.
The inclusion of an IV study arm enabled the determination of absolute bioavailability of URLi, which was ∼65% for all three SC injection sites. This is consistent with the reported absolute bioavailability of Humalog, which was 55% to 77% for SC doses between 0.1 and 0.2 U/kg.After SC injection of URLi into the abdomen, upper arm, or thigh, the total insulin lispro exposure was similar across the injection sites. This is consistent with other mealtime insulins, such as Humalog,25Braak EWT Woodworth JR Bianchi R et al.Injection site effects on the pharmacokinetics and glucodynamics of insulin lispro and regular insulin. regular insulin, NovoRapid® (Novo Nordisk, Bagsværd, Copenhagen),26Mudaliar SR Lindberg FA Joyce M et al.Insulin aspart (B28 asp-insulin): a fast-acting analog of human insulin: absorption kinetics and action profile compared with regular human insulin in healthy nondiabetic subjects. Fiasp® (Novo Nordisk),27Hövelmann U Heise T Nosek L Sassenfeld B Thomsen KMD Haahr H. Pharmacokinetic properties of fast-acting insulin aspart administered in different subcutaneous injection regions. and Apidra® (sanofi-aventis U.S. LLC, Bridgewater, NJ, USA). Consistent with a similar total insulin lispro exposure, the total insulin action after URLi was also similar across the injection sites.The Cmax was lower after injection into the thigh compared with abdomen and upper arm injections, but this was not observed with the GD response, as the Rmax was similar across these injection sites.
It has generally been observed that insulin absorption is slower in the thigh than from the abdomen.20Vora JP Burch A Peters JR Owens DR. Relationship between absorption of radiolabeled soluble insulin, subcutaneous blood flow, and anthropometry. This has been attributed to differences in absorption rate between injection sites, which have been shown to vary due to differences in SC blood flow and anthropometry. Within this study, the median Tmax and TRmax occurred later in the thigh compared with the abdomen or upper arm after injection of URLi, but the range for both the Tmax and TRmax overlapped across these injection sites. Importantly, numerous clinical studies with ultra-rapid mealtime insulins have shown that timing of the Tmax is not correlated with postprandial glucose lowering.29Heise T Hövelmann U Zijlstra E Stender-Petersen K Jacobsen JB Haahr H. A comparison of pharmacokinetic and pharmacodynamic properties between faster-acting insulin aspart and insulin aspart in elderly subjects with type 1 diabetes mellitus., 30Heise T Hövelmann U Brøndsted L Adrian CL Nosek L Haahr H. Faster-acting insulin aspart: earlier onset of appearance and greater early pharmacokinetic and pharmacodynamic effects than insulin aspart., 31Kapitza C Leohr J Liu R et al.A novel formulation of insulin lispro containing citrate and treprostinil shows faster absorption and improved postprandial glucose excursions vs. Humalog in patients with T1DM., 32Kapitza C Leohr J Liu R et al.A novel formulation of insulin lispro containing citrate and treprostinil shows significantly faster absorption and an improvement in postprandial glucose excursions vs. Humalog in patients with T2DM. Rather, the faster rate of insulin absorption, as measured by the timing of the early 50% Tmax, or the amount of early insulin exposure, resulted in a greater postprandial glucose lowering. Consistent with this finding, Tmax was not significantly different after SC injection of URLi compared with Humalog in patients with T1DM31Kapitza C Leohr J Liu R et al.A novel formulation of insulin lispro containing citrate and treprostinil shows faster absorption and improved postprandial glucose excursions vs. Humalog in patients with T1DM. and was significantly later with URLi compared with Humalog in patients with T2DM.32Kapitza C Leohr J Liu R et al.A novel formulation of insulin lispro containing citrate and treprostinil shows significantly faster absorption and an improvement in postprandial glucose excursions vs. Humalog in patients with T2DM. In contrast, the timing of the early 50% Tmax occurred earlier and there was a greater amount of insulin exposure within the first hour after dosing with URLi versus Humalog.12Leohr J Dellva MA LaBell E et al.Pharmacokinetic and glucodynamic responses of ultra rapid lispro vs lispro across a clinically relevant range of subcutaneous doses in healthy subjects., 13Linnebjerg H Zhang Q LaBell E et al.Pharmacokinetics and glucodynamics of ultra rapid lispro (URLi) versus Humalog® (lispro) in younger adults and elderly patients with type 1 diabetes mellitus: a randomised controlled trial., 14Leohr J Dellva MA Coutant DE et al.Pharmacokinetics and glucodynamics of ultra rapid lispro (URLi) versus Humalog® (lispro) in patients with type 2 diabetes mellitus: a phase I randomised, crossover study.,33Heise T Linnebjerg H Coutant D et al.Ultra rapid lispro lowers postprandial glucose and more closely matches normal physiological glucose response compared to other rapid insulin analogues: a phase 1 randomized, crossover study. This has resulted in a consistent improvement in postprandial glucose lowering with URLi compared with Humalog in adults with T1DM or T2DM.16Klaff L Cao D Dellva MA et al.Ultra rapid lispro improves postprandial glucose control compared with lispro in patients with type 1 diabetes: results from the 26-week PRONTO-T1D study.,17Blevins T Zhang Q Frias JP Jinnouchi H Chang AM. Randomized double-blind clinical trial comparing ultra rapid lispro with lispro in a basal-bolus regimen in patients with type 2 diabetes: PRONTO-T2D.,31Kapitza C Leohr J Liu R et al.A novel formulation of insulin lispro containing citrate and treprostinil shows faster absorption and improved postprandial glucose excursions vs. Humalog in patients with T1DM.,32Kapitza C Leohr J Liu R et al.A novel formulation of insulin lispro containing citrate and treprostinil shows significantly faster absorption and an improvement in postprandial glucose excursions vs. Humalog in patients with T2DM. Similarly, a faster rise in the insulin lispro concentration in the first 15 minutes was observed after URLi administration in adults with T1DM compared with endogenous insulin levels in healthy subjects following the same test meal.33Heise T Linnebjerg H Coutant D et al.Ultra rapid lispro lowers postprandial glucose and more closely matches normal physiological glucose response compared to other rapid insulin analogues: a phase 1 randomized, crossover study.
Comments (0)