Dose-Response Effects of Insulin Glargine in Type 2 Diabetes

This single-center, randomized, double-blind, five-period study was approved by the Vanderbilt University Institutional Review Board. Twenty individuals (10 male and 10 female, aged 50 ± 3 years, with diabetes duration 8 ± 1 years, BMI 36 ± 2 kg/m², and A1C 8.3 ± 0.6%) were studied. None had major complications of diabetes, all were nonsmokers, and all had normal renal and hematological function. Some individuals had mild elevations of hepatic transaminases suggestive of nonalcoholic steatohepatosis that were <1.5 times the upper normal limit. All study subjects were receiving various combination therapy for glucose control. All study subjects were receiving metformin, and this was combined with sulfonylureas (6 patients), exenatide (6 patients), insulin (either glargine, detemir, Humalog, or NovoLog, 70–30 mixtures; 14 patients). Insulin doses ranged from 11 to 70 units/day (supplementary Table 1, available in an online appendix at http://care.diabetesjournals.org/cgi/content/full/dc09-2011/DC1).

The study consisted of five separate experiments, each lasting 26 h and separated by 6–8 weeks. In each experiment, individuals received subcutaneously into the abdomen either 0.5, 1.0, 1.5, or 2.0 units/kg glargine or an identical placebo injection (0 units/kg) in a randomized, double-blind fashion. Each limb of the study consisted of 12 individuals.

Exenatide was withheld for 24 h, and oral medications and long- and intermediate-acting insulins were withheld for 60 h before each study. Individuals continued their normal weight-maintaining diet and participated in no exercise for 3 days before each experiment. Glucose control was maintained if necessary with short-acting insulin on the day before each study. Individuals were admitted to the Vanderbilt General Clinical Research Center at ∼5:00 p.m. before an experiment. A hand vein was cannulated in a retrograde fashion and maintained in a heated box (55°C) so that arterialized blood could be sampled (10). In addition, a vein in the contralateral arm was cannulated so that insulin, 20% dextrose, and glucose tracer could be administered. An intravenous insulin infusion was started to control glucose during a standardized evening meal and a 9:00 p.m. snack and was constantly adjusted overnight to maintain plasma glucose between 5.0 and 6.0 mmol/l. After a 9-h overnight fast, a primed (10 μCi in 10 min) continuous infusion (0.092 μCi/min) of [³H-3]glucose was started at time −120 min and continued throughout the experiment. Plasma glucose was maintained at euglycemia during the 120-min isotope equilibration period by continuing the overnight insulin infusion. At time 0 min, a dose of glargine was administered subcutaneously into the abdominal area. The overnight insulin infusion was discontinued 45 min after the glargine injection. During the clamp period, plasma glucose concentrations were measured every 15–30 min, and a 20% dextrose infusion was used, when necessary, at a variable rate to maintain the plasma glucose concentration at the target value of 5.5 mmol/l (11). Blood for experimental parameters was sampled every 30 min–2 h throughout the study. The glucose clamp study was ended 24 h after subcutaneous injection of glargine.

Rates of glucose appearance, endogenous glucose production (EGP), and glucose utilization were calculated according to the method of Wall et al. (12). EGP was calculated as the total glucose production minus the exogenous glucose infusion rate. Total glucose production comprises EGP and any exogenous glucose that was infused to maintain the desired euglycemia. To maintain a constant specific activity (and reduce underestimates of glucose kinetics), isotope delivery was increased commensurate with increases in exogenous glucose infusion.

Pharmacodynamic parameters of insulin action were calculated as follows: 1) end of action, time at which plasma glucose was consistently (for at least 60 min) >7.0 mmol/l (126 mg/dl), and 2) maximum glucose infusion rate (μg/kg/min).

Plasma glucose concentrations were measured in triplicate using the glucose oxidase method with a glucose analyzer (Beckman, Fullerton, CA). Subcutaneously administered insulin glargine (A21-Gly-B31-Arg-B32-Arg-insulin) precipitates at the injection site from which it is gradually liberated into the bloodstream in three equivalently bioactive forms, insulin glargine, A21-Gly-insulin, and A21-Gly-des-B30-Thr-insulin, plus A21-Gly-B31-Arg-insulin, a minor metabolite of unknown biological potency (13). The virtually complete sequence homology among insulin glargine, its metabolites, and insulin facilitates immunoassay of the former compounds with methods developed for the assay of human insulin (14,15). However, these assays can only measure total immunoreactive insulin in human plasma. Consequently, the contribution of insulin glargine and its metabolites to total plasma insulin cannot be calculated unless each glargine component is known to react equally in the immunoassay and endogenous insulin can be discounted. An insulin radioimmunoassay (PI-12K; Millipore, St. Charles, MO) that uses recombinant human insulin as reference was used in the present study. The sensitivity of the assay is 0.05 ng (1.25 μ-I.U.)/ml, and the interassay coefficients of variation (CV) at 0.8 ng (20 μIU), 2.28 ng (57 μIU), and 6.0 ng (150 μIU)/ml are 3, 7, and 9%, respectively.

C-peptide was determined using a radioimmunoassay kit (HCP-20HK; Millipore). The sensitivity of the assay is 0.1 ng/ml and the CVs at 0.34 and 1.41 ng/ml are 8.5 and 5%, respectively.

Plasma glucagon concentrations were measured using a modification of the method of Aguilar-Parada et al. (16) with an interassay CV of 12%. Plasma nonesterified fatty acid (NEFA) concentrations were determined using the WAKO kit. Plasma β-hydroxybutyrate was measured using the method of Lloyd et al. (17).

Results are expressed as means ± SE and were analyzed for statistical comparisons using one- and two-way ANOVA. P < 0.05 was accepted as a significant difference between doses. Calculations and statistical analyses were performed using SPSS (version 16.0; SPSS, Chicago, IL). Group differences were tested by a least significant difference multiple comparison test. The linear trapezoidal rule was used to calculate the AUC for plasma glucose, glucose infusion rate, and plasma insulin.

After placebo injection, plasma glucose levels increased to 8.7 ± 0.3 mmol/l at 24 h (Fig. 1). After 0.5 units/kg glargine, plasma glucose was maintained at <6.1 mmol/l until 18 h and then increased progressively (P < 0.05) to a peak of 7.2 ± 0.7 mmol/l at 24 h. In contrast, with higher doses of glargine (1.0, 1.5, and 2.0 units/kg), plasma glucose levels were maintained at 5.7 ± 0.3, 5.6 ± 0.2, and 5.3 ± 0.1 mmol/l at 24 h, which were significantly reduced (P < 0.05) compared with those for placebo.

The final 2 h of intravenous insulin administration were similar in each group (2.1 ± 0.6, 1.9 ± 0.3, 1.9 ± 0.6, 2.1 ± 0.2, and 2.3 ± 0.5 units/h for placebo and 0.5, 1.0, 1.5, and 2.0 units/kg, respectively). During the placebo (time control no exogenously administered insulin studies), basal insulin levels fell slowly and continuously over a 24-h period from 132 ± 18 to 94 ± 4 pmol/l (Fig. 1). After glargine administration, insulin levels were significantly increased (P < 0.05) compared with those for placebo (Fig. 1) and followed a physiological pattern (i.e., smooth slow decrease over 24 h) similar to that for placebo. Larger doses of glargine produced stepwise increases (P < 0.05) in mean insulin levels only up to 1.0 units/kg (137 ± 23, 216 ± 14, 254 ± 25, and 259 ± 27 pmol/l for 0.5, 1.0, 1.5, and 2.0 units/kg, respectively). No glargine dose produced a discernible peak in insulin concentration. Interindividual CVs for insulin levels were 45, 51, 54, and 46% for the 0.5, 1.0, 1.5, and 2.0 units/kg glargine doses, respectively. There were no differences in average plasma insulin levels between the 1.0, 1.5, and 2.0 units/kg doses.

Glargine administration significantly inhibited (P < 0.05) the release of C-peptide compared with placebo (Fig. 1). Incremental AUC C-peptide levels were significantly lower during the 1.0, 1.5, and 2.0 units/kg doses than during the 0.5 units/kg dose (P < 0.05). There were no differences in incremental AUC C-peptide values after 1.0, 1.5, and 2.0 units/kg doses. Plasma glucagon levels were significantly reduced (P < 0.05) compared with those for placebo by all glargine doses (supplementary Fig. 1, available in an online appendix).

Glucose infusion was not required after placebo. The incremental AUCs for the glucose infusion rate to maintain euglycemia were 64 ± 23, 149 ± 46, 213 ± 70, and 247 ± 60 μmol/kg for 0.5, 1.0, 1.5, and 2.0 units/kg, respectively. All were significantly increased (P < 0.05) compared with those for placebo, and the incremental AUCs during the 1.0, 1.5, and 2.0 units/kg doses were increased (P < 0.05) compared with that for the 0.5 units/kg dose. The peak glucose infusion rates were 2.8 ± 0.9, 5.5 ± 1.5, 6.6 ± 2.0, and 9.4 ± 2.0 μmol/kg/min in the 0.5, 1.0, 1.5, and 2.0 units/kg doses, respectively. All peak glucose infusion rates were significantly increased (P < 0.05) after glargine administration compared with those for placebo. Peak infusion rates were also higher (P < 0.05) in the 1.0, 1.5, and 2.0 units/kg doses than in the 0.5 units/kg dose. Glucose infusion rates at the end of the 24-h study for the four glargine doses in increasing order were 0.2 ± 0.17, 1.8 ± 1.0, 3.0 ± 1.6, and 3.7 ± 1.1 μmol/kg/min, respectively.

Endogenous glucose production was suppressed during the placebo time control (Fig. 2). Glargine at 0.5 units/kg produced suppression of EGP similar to that for placebo. Glargine doses of 1.0, 1.5, and 2.0 units/kg produced greater suppression (P < 0.05) of EGP than placebo. Glucose rates of disappearance were not significantly increased by any of the glargine doses compared with those for placebo (Fig. 2). Endogenous glucose production was suppressed by a relatively greater amount (P < 0.05) during placebo and all glargine doses compared with the respective changes in glucose disappearance (Table 1).

Plasma NEFA levels were suppressed (P < 0.05) in a stepwise fashion with increasing doses of glargine. Blood β-hydroxybutyrate followed a similar pattern with a significantly greater reduction (P < 0.05) of the ketone body with increasing doses of glargine (supplementary Fig. 1).

In this study, we examined the pharmacokinetic and pharmacodynamic dose-response effects of single subcutaneous injections of insulin glargine in obese type 2 diabetic individuals. Our study demonstrates that over a dose range of 0.5 to 2.0 units/kg, glargine lowers plasma glucose by a relatively hepatospecific mechanism. In obese insulin-resistant type 2 diabetic individuals, all of the glargine doses exerted metabolic effects throughout the 24-h clamp study. Circulating plasma insulin levels increased modestly despite large subcutaneous glargine doses (>200 U). Peak insulin levels plateaued after the 1.5 and 2.0 units/kg doses at 259 ± 27 pmol/l, which was barely double the insulin values after the 0.5 units/kg dose. The glucose infusion rates needed to maintain euglycemia were also modest with peak values of ∼9 μmol/kg/min required after the largest 2.0 units/kg dose.

Plasma glucose was maintained at or <7.0 mmol/l for ∼23 h after the 0.5 units/kg dose and for 24 h with the 1.0, 1.5, and 2.0 units/kg doses. In addition, NEFA, β-hydroxybutyrate, and C-peptide levels were all suppressed, relative to those for placebo for 24 h. This indicates that in this present group of obese, insulin-resistant type 2 patients single subcutaneous glargine doses of ≥0.5 units/kg can have a time action profile close to or longer than 24 h. Because in this present study, we investigated the effects of a single subcutaneous dose of glargine, we cannot determine whether repeated doses of glargine at 0.5 units/kg given over a longer period would also have resulted in a longer duration of action. This is a possibility, because Porcellati et al. (18) have demonstrated extended duration of action of glargine at a dose of 0.3 units/kg after 1 week of repeated daily use in a group of type 1 diabetic individuals. Luzio et al. (19) have also used the 24-h euglycemic clamp technique to investigate 0.5 units/kg of glargine in type 2 diabetic individuals (19). In their study, glargine at 0.5 units/kg had a 24-h duration of action. However, the type 2 diabetic individuals investigated by Luzio et al. were very different from the present study cohort because they were less obese, had better glycemic control, and were all treated with oral hypoglycemic agents. Thus, the subjects would have been predicted to be less insulin resistant and also seemed to have less advanced type 2 diabetes. Supporting this assumption is the finding that the maximal glucose infusion rate in the study of Luzio et al. was ∼9 pmol/kg/min, which was equivalent to the largest glucose infusion rate occurring in this study after a 4-fold higher dose of glargine (2.0 units/kg).

Because glargine can have a variable activation time (18), we decided, a priori, to maintain the overnight infusion of insulin for 45 min in all studies. This is the reason that there was a small increase in plasma glucose levels after glargine administration at the start of the clamp studies. In addition, because the initial period of the clamps would have reflected both subcutaneous glargine and intravenous insulin administration, we did not report plasma insulin levels until 2 h into the clamp studies. The insulin profiles from each of the injected doses did not display any demonstrable peak. In addition, the elevations in circulating plasma insulin were dramatically truncated and were not proportional to the linear increases of the subcutaneously injected insulin glargine. Thus, insulin levels were only ∼30 pmol/l (∼5 μU/ml) higher than those for placebo after the 0.5 units/kg dose. Insulin levels increased ∼80 pmol/l (14 μU/ml) between the 0.5 and 1.0 units/kg doses. Thereafter, there was only a mean nonsignificant 38 pmol/l (∼6 μunits/ml) increase between the 1.0 to 1.5 units/kg doses and no difference in insulin levels between the 1.5 and 2.0 units/kg doses. In fact, the mean maximal insulin levels after the 1.5 and 2.0 units/kg doses were equivalent at only 312–318 pmol/l (52–53 μU/ml). It is also worth noting that, based on work by Ciaraldi et al. (20), this level of glargine has binding similar to that of skeletal muscle IGF-I receptors compared with equivalent levels of human insulin. This result provides reassurance that glargine levels at even very high clinical doses do not have increased mitogenic potential. When the insulin levels are interpreted, some additional points may warrant consideration. The circulating insulin levels represent a combination of exogenous and endogenous insulin. The endogenous C-peptide area under the curve was suppressed by a greater amount for the 24-h study after the largest doses of glargine compared with that for placebo or the 0.5 units/kg dose. Thus, it may be assumed that the suppressed endogenous insulin levels may have contributed a relatively lower amount to the total plasma insulin level during the higher dose glargine clamp studies. In addition, measurement of insulin levels after subcutaneous glargine administration is complicated by circulating metabolites of the molecule that have both differential metabolic effects and abilities to cross-react in a conventional insulin immunoassay. Thus, the plasma insulin levels after glargine may represent an overestimate of the total circulating glargine species present in the plasma (13). Nevertheless, what is evident from this study is that the chemical formulation of the glargine molecule dramatically limits absorption of insulin from the subcutaneous tissue and acts as a buffer to limit circulating insulin levels (19,21). We believe that the formulation of insulin glargine is a major determinant of the pharmacokinetic profile of the molecule as the interindividual CVs of insulin levels in the present study ranged from 45 to 54% and are similar to values reported previously with nonanalog extended-action insulins (22).

Glargine lowered plasma glucose by a relatively hepatospecific mechanism. At each dose level, endogenous glucose production was suppressed by a greater amount compared with any relative increase in glucose disposal. In fact, none of the glargine doses increased glucose disposal by a significantly greater amount than placebo. Contributing to the relatively hepatospecific action was the effect of glargine on glucagon and NEFA levels. Glucagon is known to be a significant contributor to basal EGP (23). Thus, the effects of glargine to suppress glucagon during the glucose clamps could also have been a contributory mechanism for the relative hepatospecific action. In addition, the effects of glargine to reduce circulating NEFA levels would also have effects to lower EGP. The reduced β-hydroxybutyrate levels after glargine administration also support the hepatic action of the molecules. Conversion of NEFA to ketone bodies largely occurs in the liver, and this was significantly reduced during the present studies. During fasting, EGP falls as glycogen stores are used (22). Furthermore, physiologically endogenous basal insulin has a relatively hepatospecific effect. This can be clearly seen in the placebo experiments. However, what the present studies have demonstrated is that insulin glargine doses up to 2.0 units/kg (with injected insulin doses >200 U) also lower fasting glucose in type 2 diabetes in a “physiological” hepatospecific manner.

Although the doses of glargine used in type 1 diabetes practice are usually quite low (<0.5 units/kg), the amount of glargine used in managing type 2 diabetes is often considerably higher. Therefore, in this present study, we have studied the acute 24-h pharmacokinetic and pharmacodynamic effects of large doses of glargine (0.5–2.0 units/kg) in type 2 diabetes. We should point out that the study was performed in the southeastern U.S. where there is a very high prevalence of type 2 diabetes and accompanying obesity. Our study population had a mean BMI of 36 ± 2 kg/m² and was clearly obese. Additional studies investigating the pharmacokinetic and pharmacodynamic dose-response characteristics of glargine using lower doses of the insulin in less obese individuals would also be useful. In addition, we should also mention that in this study, glargine was administered in the morning, which occurs in clinical practice but not as commonly as evening dosage. Furthermore, the pharmacodynamics of glargine may be influenced by the time of administration. The dawn phenomenon with its attendant changes in plasma glucose and insulin sensitivity can occur 4–8 h after evening dosage but ∼20 h after a morning injection. Thus, the pattern of glucose kinetics might have been altered if glargine had been administered in the evening.

In summary, in this study we have investigated the pharmacokinetics and pharmacodynamics of large doses (0.5–2.0 units/kg) of insulin glargine in obese type 2 diabetic individuals. A single subcutaneous injection of glargine can have a duration of action of at least 24 h. Very large doses result in modest increases in glucose flux with no discernible peak action. This seems to be due to the limited, peakless, and continuous release of the insulin from the subcutaneous tissue. Glargine lowers plasma glucose by a predominantly hepatospecific action (i.e., inhibiting endogenous glucose production) with minimal effects on stimulation of glucose disposal.

Funding for this investigator-initiated study was provided by sanofi-aventis. Support was also was provided by National Institutes of Health grants M01-RR-000095 and P60-DK-020593.

S.N.D. has served as a consultant for sanofi-aventis. No other potential conflicts of interest relevant to this article were reported.

We thank the Vanderbilt General Clinical Research Center staff for their expert care. We also thank Susan Hajizedah for her technical excellence and Jan Botts Hicks for her expert editorial assistance.