Summary Basis of Decision for Levemir ®

3.3.1 Human Pharmacology

The PK and PD properties of insulin detemir were investigated in healthy subjects, subjects with Type I or Type II diabetes mellitus, and in special populations including: Japanese subjects, elderly subjects, children and adolescents, and subjects with hepatic or renal impairment. The clinical pharmacology development programme consisted of 32 clinical trials: 24 in Europe, 3 in the U.S., and 3 in Japan. This is in addition to one other trial presented as short-term for safety analyses as well as a pilot study to set up a technique for the evaluation of the PK/PD of insulin detemir.

The following issues were addressed: population, effect of injection sites, potential drug accumulation after multiple doses, effect of hepatic or renal impairment on PK, inter- subject variability and bioequivalence of the formulations used for clinical evaluation. Insulin detemir is considered a long acting insulin product similar to NPH insulin.

3.3.2 Pharmacodynamics

The dose-response relationship and duration of action of the to-be-marketed (100 U/mL) insulin detemir preparation (injected SC) was investigated in subjects with Type I diabetes mellitus under isoglycemic clamp conditions. The mean glucose infusion rate (GIR) profiles for five SC insulin detemir doses ranging from 0.1 U/kg to 1.6 U/kg and one NPH insulin dose (0.3 IU/kg) were considered. Both the area under the glucose infusion rate (AUC_GIR) and the maximum GIR increased proportionally with increased doses of insulin detemir.

To determine the insulin detemir dose equipotent to 0.3 IU/kg of NPH insulin, the equivalent insulin detemir dose was interpolated from estimates of
AUC_GIR(0-24h). In terms of AUC_GIR(0-24h), it was estimated that a dose of approximately 0.3 U/kg (0.29 U/kg) insulin detemir would give the same glucose lowering effect as 0.3 IU/kg NPH insulin with a GIR_max of
1.4 mg/kg/min.

The duration of action for insulin detemir was varied and apparently dose-dependent. It ranged from 6 hours at the lowest dose to approximately
24 hours at the highest for a range of doses including those that were clinically relevant.

3.3.3 Pharmacokinetics

A number of clinical pharmacology trials investigated PK after both IV and SC administration of insulin detemir in healthy and diabetic subjects.

Absorption, Distribution, and Metabolism

As was expected, due to the albumin binding properties of insulin detemir, clearance in the IV studies was lower (1.8 to 2.7 mL/min/kg across trials) than for human insulin (15.8 to 19.6 mL/min/kg across trials). Additionally, the initial half-life was longer (19 to 25 min - across trials) for insulin detemir versus 2.4 min for human insulin. The half-life after SC administration was considerably longer, being about 3 to 6 hours in healthy subjects and about 5 to 7 hours in subjects with Type I diabetes mellitus (in therapeutically relevant doses). The volume of distribution for insulin detemir was small (approximately 0.1 L/kg), which is expected for a highly protein bound drug and is in agreement with the albumin distribution space. The concentration gradient between serum and interstitial fluid was higher for insulin detemir.

Overall, the absorption PK profile of insulin detemir in subjects with Type I diabetes mellitus closely resembles that of healthy subjects. Insulin detemir was absorbed at a slightly slower rate in subjects with Type II diabetes mellitus than in those with Type I. Serum insulin detemir concentrations, as measured by AUC_(0-16h) and C_max, tended to be lower for subjects with Type II diabetes mellitus. AUC and C_max were greater after administration of 0.75 U/kg than 0.38 U/kg for both groups. The smaller AUC observed for subjects with Type II diabetes mellitus may be related to the relatively short evaluation period (16 hours). The apparent difference in total exposure may have been less pronounced if the entire concentration-time profile had been assessed.

Absorption was evaluated in 28 healthy subjects after SC administration in the three most commonly used injection sites: the thigh, abdomen, and deltoid. The resulting data indicates that insulin detemir has slower absorption, smaller peak concentration, longer half-life, and slightly lower bioavailability following SC administration in the thigh as compared to the abdomen or deltoid.

AUC_{(0- ∞)} and C_max increased proportionally with administered doses ranging from 0.1 to 1.6 U/kg. Steady state with insulin detemir was reached quite rapidly (by the end of 24 h after two doses administered at 12 hour intervals) and trough concentrations at this point were almost identical to those after 7 days of daily administration twice per day. No further accumulation was evident after the second dose on the first treatment day.

Special Populations

A cross-trial analysis demonstrated that there was no difference in PK between genders, however, one study did appear to show that absorption of insulin detemir was slower in males than in females following SC administration in the thigh. A trial investigating differences in PK in the elderly showed higher AUC for elderly subjects (69-85 years of age). This difference is not considered to be relevant to dosing recommendations. No consistent differences were found between the PK properties of insulin detemir in healthy Caucasian and Japanese subjects.

There did not appear to be any relevant effect of renal impairment on the PKs.

Hepatic impairment impacted the extent of bioavailability as estimated by AUC_{(0- ∞)}. There was a decrease in bioavailability with increasing degree of hepatic impairment (as shown in four groups: healthy, mild, moderate, and severe) where the AUC_{(0- ∞)} for healthy subjects was 58% higher than that of subjects with severe hepatic impairment. The PK results were confounded, however, by unbalanced bodyweight between healthy subjects and subjects with moderate to severe liver impairment. A >40% smaller AUC_(0-∞) and 35-47% faster clearance were observed, however, in patients with severe/moderate liver insufficiency as compared to healthy subjects. The requirements for Levemir^® may need to be adjusted in patients with hepatic impairment.

3.3.4 Clinical Efficacy

The efficacy of Levemir^® was studied in multi-centre, randomized, parallel comparison, open label trials. Subjects had to have been diagnosed with Type I or Type II diabetes mellitus for >12 months and had to be >18 years old for Type I diabetics or >35 years for Type II diabetics. Subjects were required to have a BMI <35 kg/m₂, an Hb_A1c level < 12.0%, an ability to measure glucose independently, and a basal insulin requirement of >30% of total daily dose. The starting dose for each subject was based upon each patient's previous insulin requirements. All doses were administered by SC injection.

Type I Diabetes Mellitus Trials

Three pivotal trials (6 months, 4 months, and 4 months) were conducted to compare the efficacy of insulin detemir to NPH insulin in a basal/bolus regimen, administered mainly twice daily. The total data base included 1038 insulin detemir patients and 517 NPH insulin patients.

The primary efficacy endpoint was the level of Hemoglobin _A1c (Hb_A1c) at the end of the trials. The rationale for this measurement is based on the fact that glucose binds readily to subtype Hemoglobin A forming the glycosylated Hb_A1c. The decomposition of this process is relatively slow and overall Hb_A1c levels change gradually over approximately 10 weeks. The Hb_A1c level therefore represents the average BG level of roughly the last 8-12 weeks and can be used as an indicator of the average recent BG level.

The secondary efficacy endpoints included the fasting plasma glucose (FPG), 9 point BG profiles, variation in FPG, frequency of high and low BG excursions and fluctuations, bodyweight changes, and quality of life.

Table 1: Pivotal Type I Diabetes Mellitus Trial Summary

In all three pivotal studies, insulin detemir was non-inferior to NPH insulin with respect to the primary efficacy endpoint (Hb_A1c at end of trials). The FPG and variation in FPG were reduced in the insulin detemir groups in all three studies. The overall risk of hypoglycaemia was similar in the test and control groups but the frequency of nocturnal hypoglycaemia tended to be lower in the insulin detemir groups. Insulin detemir performed better with respect to nocturnal hypoglycaemia when it was given in the morning and at bedtime than when it was given at 12 hour intervals. Weight gain was less in all groups receiving insulin detemir but no explanation could be provided to clarify this phenomenon. It was noted however, that this was not due to poorer control of diabetes mellitus in these groups.

Type II Diabetes Mellitus Trials

Three pivotal trials were conducted (6 months, 6 months, and 22 weeks). Two were basal/bolus regimes comparing insulin detemir with NPH Insulin, once or twice daily. The other considered insulin detemir combined with an oral antihyperglycemic (Metformin or similar). The total data base included 845 insulin detemir patients and 521 NPH insulin patients. The primary efficacy endpoint and secondary efficacy endpoints were the same as for the Type 1 diabetes mellitus trials.

Table 2: Pivotal Type II Diabetes Mellitus Trial Summary

Two of the three pivotal studies demonstrated non-inferiority of insulin detemir in comparison with NPH insulin. In the third study, insulin detemir was inferior to NPH insulin with respect to Hb_A1c levels at the end of the trial (primary efficacy endpoint). This latter trial was a combination of insulin detemir (or NPH Insulin) with an oral antihyperglycemic (Metformin or other). Based on these results, a statement was included in the Product Monograph that insulin detemir is intended for use in a basal/bolus regimen and that studies in combination with Metformin or other oral antihyperglycemic agents are ongoing. The secondary efficacy criteria were much the same in the test and control groups except for the subject variation in FBG which was reduced in the insulin detemir groups.

Non Pivotal Efficacy Trials

One trial was conducted. Due to the fact that it employed a 'not to be marketed' preparation it was only considered for safety and not for efficacy.

3.3.5 Clinical Safety

Type I Diabetes Mellitus Trials

Four pivotal studies (6 months, 4 months, 4 months, and 26 weeks - Table 1) were conducted. The total data base included 1199 insulin detemir patients, 517 NPH insulin patients, and 159 insulin glargine (glArg) patients. The safety criteria included the frequency of hypoglycaemia, presence of nocturnal hypoglycaemia or adverse events (AEs) and the formation of anti-insulin antibodies.

Treatment emergent AEs were equally distributed in frequency and severity between the test and control groups and no safety concerns specific to insulin detemir were identified in any of the pivotal trials. In these studies, Levemir^® and NPH insulin had a similar overall rate of hypoglycaemia. Nocturnal glucose profiles were flatter and smoother, and the risk of nocturnal hypoglycaemia was reduced (by 22%) for Levemir^® as compared to NPH insulin. Mild local reactions at the injection site appeared to be more frequent with insulin detemir but were not considered a significant problem. Specific and non-specific antibody formation occurred in the insulin detemir groups but apparently did not impair insulin action. No studies were performed on subjects <18 years of age.

Four non-pivotal trials were conducted comparing insulin detemir with NPH insulin in a basal/bolus regimen. Each trial was 6 months in duration and two were extensions of the other two. The safety criteria were the same as those for the pivotal trials. The total data base included 907 insulin detemir patients and 603 NPH insulin patients. Treatment emergent adverse events were equally distributed between the test and control groups and no safety concerns specific to insulin detemir were identified.

Type II Diabetes Mellitus Trials

The three pivotal trials are described in Table 2. The safety criteria were the same as those of the Type I Diabetes mellitus trials.

Due to an apparently greater number of deaths in the insulin detemir groups the sponsor commissioned an independent meta-analysis to explore a possible relationship between cardiovascular and cerebrovascular events in the patients in the studies that used insulin detemir versus NPH Insulin. No relationship to trial medication was found. The clinical reviewer has perused the causes of death and has concluded that they can be attributed to existing medical conditions and could not be specifically attributed to insulin detemir. No other safety concerns were identified in the studies.

One non-pivotal trial was conducted. The duration of this study was 6 months and it included a total data base of 221 insulin detemir and 218 NPH insulin patients. No safety concerns specific to insulin detemir were identified.

Insulin detemir has been approved in the U.S. and in the European community. Safety studies did not identify any major concerns related to insulin detemir or significant differences between insulin detemir and NPH insulin.

Contraindications, Warnings, and Precautions

Levemir^® is contraindicated during episodes of hypoglycaemia and in patients hypersensitive to Levemir^® or one of its excipients. As with insulins in general, concomitant use of other drugs may influence insulin requirements.