Description of Studies

Fifteen manufacturer-sponsored randomized controlled trials (RCTs) plus five extensions were included in this systematic review, four of which were double blinded and the remainder open label. The studies featured different populations, T1DM and T2DM. Another four RCTs were excluded because either their small sample size or their short duration suggested they were unlikely to add any additional information beyond what was obtained in the studies summarized in the report.

DEVOTE was a double-blind RCT that compared IDeg with IGlar, both in a basal-bolus regimen, in a population of 7,637 patients with T2DM and cardiovascular disease. The primary outcome was the time to first major adverse cardiovascular event (MACE), testing the noninferiority of IDeg to IGlar, with a margin for noninferiority of 1.3 for the upper limit of the 95% confidence interval (CI) for the hazard ratio. DEVOTE was an event-driven study, targeting 633 MACEs before study completion, and it lasted a mean of 24 months. Confirmatory secondary end points, which all tested the superiority of IDeg to IGlar, included the number of confirmed hypoglycemic episodes and the occurrence of at least one hypoglycemic episode within a participant. The study had a two-week screening period, as well as a 30-day follow-up period at the end of study. DEVOTE is by far the largest of the studies included in this review.

The SWITCH studies, SWITCH-1 (T1DM) and SWITCH-2 (T2DM), employed a crossover design, with patients randomized to start on either IDeg or IGlar and then cross over to the other intervention after 32 weeks of therapy, resulting in a total treatment period of 64 weeks. The primary outcome of the SWITCH studies was the proportion of participants with severe or blood glucose–confirmed symptomatic hypoglycemic episodes during the maintenance period, that is, after 16 weeks of treatment. SWITCH-1 tested the noninferiority of IDeg to IGlar, with noninferiority confirmed if the upper bound of the 95% CI for the rate ratio was ≤ 1.10. SWITCH-2 tested the superiority of IDeg to IGlar. In each study, before testing of the primary outcome could proceed, noninferiority had to be confirmed for the secondary supportive end point of change from baseline in glycated hemoglobin (A1C) after 32 weeks of therapy. The margin for noninferiority was 0.4%, the same margin used in the BEGIN trials, described below. Each of the SWITCH studies had a two-week screening period and a one-week follow-up.

All of the BEGIN trials that compared IDeg with another basal insulin (described in more detail below) were noninferiority trials that tested the noninferiority of IDeg to a comparator for the primary outcome of change from baseline in A1C. Most of the trials had confirmatory secondary outcomes that compared the superiority of IDeg against the comparator. The extension trials, where they occurred, focused on the long-term safety of IDeg; therefore, there were no efficacy outcomes assessed. In the extensions, all patients continued in their originally randomized groups.

Populations

Inclusion and Exclusion Criteria

Participants in DEVOTE had T2DM, with an A1C of 7% or more (or below 7% if receiving current insulin therapy of at least 20 units daily). Participants were currently treated with one or more oral or injectable antidiabetes drugs. They had to be at least 50 years old and have evidence of cardiovascular disease or chronic kidney disease (Table 11).

In the trials in T1DM, participants had to have been treated on a basal-bolus regimen for at least 12 months, with an A1C of 10% or less and a BMI of 35 kg/m² or less. For a high-level summary of these trials, see Table 12; for detailed summaries, see Table 39 and Table 40.

Table 12

Type 1 Diabetes Mellitus and Type 2 Diabetes Mellitus Trial Details.

In the T2DM trials in insulin-naive patients, participants had to have had T2DM for at least six months, an A1C of between 7.0% or 7.5% and 10%, and a maximum BMI of 40 kg/m² to 45 kg/m². All were receiving OADs for at least three months before randomization in a regimen that typically featured metformin with or without another OAD. See Table 12 for a high-level summary of study designs; for detailed summaries, see Table 41, Table 42, Table 43, Table 44, and Table 45.

Participants in Study 3668 (basal only) had to have had T2DM for at least six months and be on OAD monotherapy, insulin monotherapy, or a combination of the two. The only allowed OADs were metformin, insulin secretagogues, or pioglitazone. Participants on OAD alone had to have an A1C between 7% and 11%. Those on combination basal insulin and OAD or basal insulin monotherapy were to be between 7% and 10%. In the other basal-only study, participants were to have an A1C of at least 7.5% and be on metformin with or without another OAD. For a high-level summary of these two studies, see Table 12; for detailed summaries, see Table 46 and Table 47.

In the basal-bolus study, participants were to have hadT2DM for at least six months and could be on any insulin regimen, with or without OADs, for at least three months before randomization. Their A1C had to be between 7.5% and 11%, and their maximum BMI was to be 40 kg/m². For a high-level summary, see Table 12; for a detailed summary, see Table 48.

Baseline Characteristics

Trials with populations with T1DM tended to feature younger participants (early to mid-40s) versus studies that focused on T2DM (mid-50s to mid-60s), which aligns with the onset and progression of the disease subtypes. Among T2DM studies, the oldest populations were in DEVOTE (65 years of age) and SWITCH-2 (61 years of age) (Table 13, Table 14).

Table 13

Summary of Baseline Characteristics — DEVOTE.

Table 14

Summary of Baseline Characteristics — SWITCH-1 and SWITCH-2.

Across all studies, the majority of participants were male, and most were Caucasian, with the exception of Studies 3585 and 3586, where almost all participants were Asian (non-Indian), and Study 3587, where about two-thirds were Asian (non-Indian).

In the T1DM studies, between 14% and 29% of participants had diabetes complications at baseline. In DEVOTE, 86% of participants had established cardiovascular or chronic kidney disease, while in the other T2DM studies, the proportion of participants with diabetes complications varied widely at baseline, from a low of around 10% in Studies 3579 and 3580 to a high of about 40% in Study 3586.

Baseline characteristics were generally balanced between groups within studies. The most common baseline parameter to differ between groups was gender, with the largest difference between groups found in Study 3668, where 59% of participants in the IDeg-Flex (IDeg flexible dosing regimen) group and 48% of participants in the IGlar group were male.

Table 15

Summary of Baseline Characteristics — Type 1 Diabetes Mellitus (Studies 3770, 3583, and 3585).

Table 16

Summary of Baseline Characteristics — Type 2 Diabetes Mellitus, Insulin-Naive (Studies 3579, 3580, 3586, and 3672).

Table 17

Summary of Baseline Characteristics — Type 2 Diabetes Mellitus, Insulin-Naive (Studies 3587 and 3944).

Table 18

Summary of Baseline Characteristics — Type 2 Diabetes Mellitus, Basal Insulin (Studies 3668 and 3943).

Table 19

Summary of Baseline Characteristics — Type 2 Diabetes Mellitus, Basal-Bolus (Study 3582).

Interventions

The studies generally employed a treat-to-target strategy for insulin dosing. For example, in SWITCH-1, participants’ plasma glucose was titrated to a self-measured plasma glucose of 4.0 mmol/L to 5.0 mmol/L. A dose reduction was to be implemented if one or more of the pre-breakfast glucose values was < 4.0 mmol/L. Bolus insulin (insulin aspart) was titrated individually based either on carbohydrate counting or by using a sliding scale based on the lowest of three pre-meal or bedtime glucose values. Participants were typically given an algorithm they used to adjust their insulin regimens throughout the trial.

The majority of studies randomized participants in a 1:1 manner; however, Studies 3585, 3586, and 3587 randomized participants in a 2:1 manner (IDeg to comparator), and Studies 3579, 3583, and 3582 randomized 3:1. Two studies, 3770 and 3668, included an IDeg-Flex regimen in addition to the standard IDeg daily regimen, where participants’ injections were to be given in a rotating schedule with eight-hour to 40-hour intervals between doses. These studies randomized participants 1:1:1. If stratification was reported or performed, the most common variable was by region (Studies 3587, 3944, 3586, 3585). Study 3580 stratified by use of pioglitazone at screening, Study 3582 by prior insulin regimen, and Study 3668 by prior treatment (insulin, OAD, or both). The SWITCH studies and DEVOTE did not report whether stratification occurred.

Outcomes

The primary outcome of DEVOTE was time from randomization to first occurrence of an event adjudication committee (EAC)–confirmed 3-component MACE: cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke.

Confirmed hypoglycemic episodes consisted of episodes of severe hypoglycemia as well as minor hypoglycemic episodes with a confirmed plasma glucose value of < 3.1 mmol/L. Hypoglycemic episodes were defined as nocturnal if the time of onset was between 00:01 a.m. and 05:59 a.m., inclusive. Severe hypoglycemia was defined as an episode requiring the assistance of another person to actively administer carbohydrate or glucagon, or take other resuscitative actions.

Blood samples for A1C were analyzed using a Bio-Rad high performance liquid chromatography method at a central laboratory. A1C samples were collected at multiple visits in the main trial and extensions (where applicable). The assay method used was a National Glycohemoglobin Standardization Program–certified method. Blood samples for FPG were analyzed using a Roche enzymatic method at a central laboratory. FPG samples were collected at multiple visits in the main trials and extensions (where applicable). The patients were to attend these visits in a fasting state.

Within-patient variability as measured by coefficient of variation was to be derived from pre-breakfast plasma glucose values after 26 weeks of treatment. Logarithm-transformed self-measured plasma glucose values were to be analyzed as repeated measures in a linear mixed model with treatment, antidiabetes treatment at screening, sex, and region as fixed factors, age as a covariate, and patient as random factor. The model was to assume independent within-patient and between-patient errors with variances depending on treatment. Within-patient variability as measured by coefficient of variation for a treatment could be calculated from the corresponding residual variance. The CI for the coefficient of variation ratio between treatments was to be calculated using the delta method.

Changes in patients’ health-related quality of life (HRQoL) and treatment-related impacts of minor hypoglycemic episodes on patients’ daily function and well-being were evaluated using the Short Form (36) Health Survey, version 2.0 (SF-36v2) and Treatment-Related Impact Measure — Hypoglycemic Events (TRIM-HYPO) questionnaires, respectively. Responses for the SF-36v2 were measured on standardized scales from 1998 based on the US general population, with a mean of 50 and standard deviation of 10. Responses for the TRIM-HYPO were standardized to a scale of 0 to 100. In the SF-36v2 questionnaire, higher scores indicate a better HRQoL. In the TRIM-HYPO questionnaire, lower scores indicate better daily function and well-being for the patient.

SF-36 is a generic health assessment questionnaire that has been used in clinical trials to study the impact of chronic disease on HRQoL. SF-36 consists of 36 items representing eight dimensions: physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, and mental health. Item response options are presented on a 3-point to 6-point Likert-like scale. Each item is scored on a 0 to 100 range and item scores are averaged together to create the eight domain scores. SF-36 also provides two component summaries, the physical component summary and the mental component summary, which are created by aggregating the eight domains according to a scoring algorithm. On any of the scales, an increase in score indicates improvement in health status. Based on clinical anchor data, the SF-36 User’s Manual proposed the following minimal important differences for general use of the SF-36v2: a change of 2 to 4 points in each domain or 2 to 3 points in each component summary.²⁶ No minimal clinically important difference (MCID) in patients with T1DM or T2DM was found in the literature.

Treatment-Related Impact Measure — Diabetes (TRIM-D) is a diabetes-specific questionnaire developed to assess the full impact of diabetes treatment on patients’ quality of life. This patient-reported outcome measure consists of 28 items encompassed in five domains: treatment burden (six items), daily life (five items), diabetes management (five items), psychological health (eight items), and compliance (four items). Response options are presented on a 5-point Likert-like scale. An increase in score indicates an improvement in health state. Domains can be scored individually or the measure can be scored as a total of these domains.^27,28 No MCID has been determined for the TRIM-D.

TRIM-HYPO is a patient-reported outcome measure developed to measure the impact of non-severe hypoglycemic events on patients’ HRQoL arising from the use of insulin to treat both forms of diabetes (T1DM and T2DM). TRIM-HYPO is a self-reported questionnaire comprising 33 Likert-like scale items (scored 1 to 5) in five domains: daily functioning, emotional well-being, diabetes management, work productivity, and sleep disruption. Domains are scored individually. A total score is also calculated, using three of the five domains (daily functioning, emotional well-being, diabetes management), as work productivity and sleep disruption do not apply to all patients. Lower scores on the TRIM-HYPO indicate a better health state. Raw scores are obtained by aggregating scale items into their respective domain scales. A weighted score is then generated, based by the number of non-severe hypoglycemic occurrences in the past 30 days: the higher the number, the greater the impact on the weighted score. This weighting helps account for the difference in HRQoL of patients experiencing few events versus those experiencing many hypoglycemic events. A standard algorithm method transforms the weighted scores into a 0 to 100 score.²⁹ No MCID has been determined for the TRIM-HYPO.

Statistical Analysis

All included studies carried out power calculations to determine sample size, and all studies randomized sufficient numbers of patients to ensure adequate power for assessing the primary end point.

Analysis Populations

Patient Disposition

The proportion of participants withdrawing varied greatly between studies, with the highest withdrawal rates seen in the SWITCH studies, ranging between 18% and 23% between groups, and the lowest in DEVOTE, around 2% (Table 20). Proportions of withdrawals above 20% were also seen in studies 3579 (22%) and 3580 (24%), although no differences in the proportion of withdrawals were evident between groups (Table 22). The largest difference in proportion of withdrawals was in Study 3944, with the IDeg + liraglutide group having a much lower proportion than the placebo + liraglutide group (8% versus 24%, respectively) (Table 23). In Study 3770, there was a numerically higher proportion of withdrawals in both IDeg groups versus IGlar (16% in each IDeg group versus 7% in IGlar) (Table 21).

Table 20

Patient Disposition — DEVOTE, SWITCH-1, and SWITCH-2.

Table 22

Patient Disposition — Type 2 Diabetes Mellitus, Insulin-Naive (Studies 3579, 3643, 3586, 3672, and 3580).

Table 23

Patient Disposition — Type 2 Diabetes Mellitus, Insulin-Naive (Studies 3587 and 3944).

Table 21

Patient Disposition — Type 1 Diabetes Mellitus (Studies 3770, 3583, and 3585 + Extensions).

Table 24

Patient Disposition — Type 2 Diabetes Mellitus, Basal Insulin (Studies 3668 and 3943).

Table 25

Patient Disposition — Type 2 Diabetes Mellitus, Basal-Bolus (Studies 3582 and 3667).

Table 26

Exposure to Study Treatments — Mean.

Exposure was generally similar between groups among studies, with the largest difference in Study 3944, which is also the study with the largest difference in withdrawal rates between groups.

Table 27

Basal Insulin Dose.

Internal Validity

All studies used an appropriate method of randomization using an interactive voice/Web response system with appropriate allocation concealment. Patient characteristics were generally well balanced in most studies, although in Study 3668 the baseline characteristics were not well balanced with respect to gender. With respect to blinding, only DEVOTE, SWITCH-1, and SWITCH-2, as well as Study 3944, were double blinded, and the insulins were provided in “visually identical” vials. All other studies were open labelled and therefore subject to bias, particularly for subjective outcomes such as HRQoL, which could be influenced by patient knowledge of their assigned treatment. A blinded EAC was employed in DEVOTE, and an interim analysis was prepared and submitted to the FDA.

The BEGIN studies that included an insulin comparator used a noninferiority design for testing the primary outcome (change from baseline to end of treatment in A1C), and all used the same margin for noninferiority of a change in A1C of 0.4%. The manufacturer provided a rationale for the choice of this noninferiority margin, and this rationale appeared to be reasonable. This margin is also suggested by the FDA, which considers an A1C reduction of > 0.3% to be clinically meaningful; therefore, a difference in A1C of 0.3 to 0.4% between treatments could be considered clinically significant.³⁶ Noninferiority was also tested for the primary outcome in DEVOTE; again, the rationale of the margin for noninferiority was described and the margin appeared reasonable. The SWITCH studies employed a noninferiority design in a hierarchy in order to determine testing of the primary outcome. In both studies, noninferiority for change from baseline in A1C, a secondary end point, had to be met before the primary end point and subsequent secondary end points in the hierarchy were tested.

A hierarchical testing procedure was employed to account for type I error in all studies that included confirmatory secondary end points, and the hierarchy was adhered to. The studies that did not include confirmatory secondary end points were Studies 3770, 3668, 3943, and 3944.

In the SWITCH-1 study, a per-protocol sensitivity analysis of the primary end point, a test of noninferiority, does not appear to have been conducted, as a per-protocol population was not defined in the study. The use of a per-protocol population is a recommended approach for noninferiority trials; thus, differences between the groups could have been masked, particularly given the high withdrawal rates in these studies.

With the exception of DEVOTE, no studies conducted a true intention-to-treat analysis; however, the small number of participants excluded from the main analysis is unlikely to have biased results.

The proportion of participants withdrawing varied greatly between studies, with the highest withdrawal rates seen in the SWITCH studies, ranging between 18% and 23% between groups. The direction of bias is confounded by the crossover design; however, given the high proportion of withdrawals, it is likely that the results were affected in some way, as the composition of the original randomized population would have been altered significantly throughout the trial. Proportions of withdrawals above 20% were also seen in studies 3579 (22%) and 3580 (24%), although generally no differences in proportion of withdrawals were evident between groups. The largest difference in proportion of withdrawals was in Study 3944, with the IDeg + liraglutide group having a much lower proportion than the placebo + liraglutide group (8% versus 24%, respectively). In Study 3770, there was a numerically higher proportion of withdrawals in both IDeg groups versus IGlar (16% in each, versus 7%). A high proportion of withdrawals may understate important outcomes such as hypoglycemic events, for example; and a higher proportion of withdrawals in one group versus another may bias results in favour of the group with more withdrawals, as they have less exposure to risk of hypoglycemia. Additionally, extensive withdrawals are a concern, given the noninferiority study designs employed across the studies.

The included studies typically accounted for missing data using an LOCF approach. Sensitivity analyses were also performed and appeared to support the results of the primary analysis. The LOCF approach can introduce bias into the results, and the risk of bias would be expected to increase with higher proportions of withdrawals and when there are differential withdrawals between groups within studies; both of these phenomena were seen among the included studies. The fact that the sensitivity analyses supported the conclusions of the LOCF results does allay some concern about the use of this approach for the imputation of missing data; however, a major assumption in the sensitivity analysis is that the data were missing at random, which is rarely the case and could also bias the results. The DEVOTE study, which was an event-driven study, employed a tipping-point analysis to account for missing data due to early withdrawals.

All studies employed a treat-to-target design with respect to dosing; therefore, differences in A1C would not be expected. This approach is recommended by the FDA for assessing differences in safety, tolerability, and clinical utility when insulin dosing and efficacy are maximized. However, these studies have limited utility for evaluating treatment efficacy.

External Validity

There were numerous clinical trials included in this review, with representation across the globe. Across all the included studies, there was a relatively low proportion of Indigenous participants (< 1% across the studies). The consistent majority of participants were Caucasian, with the exception of studies conducted in Asia. The lack of representation of Indigenous populations is a generalizability issue for Canada, given the relatively high proportion of Indigenous peoples diagnosed with diabetes mellitus. The clinical expert also noted that participants had a relatively long duration of disease at baseline, but did not note any other generalizability issues.

The trials largely focused on IGlar as a comparator, with IDet a comparator in only one of the studies. These are the two most commonly used intermediate-acting and long-acting insulins; however, NPH is still a popular option in some patients due to its lower cost. Thus, at least one trial comparing IDeg with NPH may have provided some additional useful insight into the comparative efficacy and harms of IDeg.

DEVOTE had the longest follow-up of all the included studies, with a mean exposure of approximately 24 months, and was powered to assess clinical outcomes in a T2DM population. However, there were no trials that similarly assessed diabetes complications in T1DM. Such trials would likely require a much longer follow-up than those included in this review. Other included trials focused on hypoglycemia as a primary outcome (SWITCH studies); the majority of included studies (BEGIN trials) focused on A1C, a widely used surrogate marker of disease in diabetes mellitus.

Two studies had extensive run-in periods, which can suggest enrichment of the study population. In Study 3943, the run-in period was to establish that the study population was one requiring high-dose insulin (participants were included only if they failed to reach target while on high-dose IGlar); this does seem appropriate, given the study objective. In Study 3944, participants were initiated on liraglutide in the 15-week run-in, which they then continued on in the study. The purpose of the study was to assess the combination of IDeg with liraglutide versus liraglutide alone (i.e., placebo plus liraglutide) in patients on metformin. A large proportion of participants were screened out during the run-in (in most cases due to failure to reach A1C targets); this was consistent with the planned 941 participants in the run-in versus 320 that were to be randomized. The study set a relatively narrow target for A1C (7.0% to 9.0%) for inclusion in the study. The rationale was that, as it was placebo-controlled, this narrow range would reduce the risk of intensified treatment being needed during the 26-week treatment phase.

HRQoL was consistently assessed in the included studies, but not as a confirmatory (i.e., high-priority) secondary outcome and not at all in the largest study, DEVOTE. Thus, HRQoL appears to have been given a lower priority in the included studies than would be expected based on the importance that patients place on this outcome.

Morbidity

In DEVOTE, the primary composite cardiovascular outcome occurred in fewer IDeg than IGlar participants (8.5% versus 9.3% of participants), for a hazard ratio of 0.91 (95% CI, 0.78 to 1.06) (Table 32). The upper bound of the 95% CI was below 1.3, confirming noninferiority of IDeg relative to IGlar with respect to cardiovascular safety (P < 0.001). Other signs of morbidity where there was no difference between IDeg and IGlar were non-fatal myocardial infarction (hazard ratio 0.85; 95% CI, 0.68 to 1.06; P = 0.15), non-fatal stroke (hazard ratio 0.90; 95% CI, 0.65 to 1.23; P = 0.50), and unstable angina leading to hospitalization (hazard ratio 0.95; 95% CI, 0.68 to 1.31; P = 0.74).

Table 32

Key Efficacy Outcomes — DEVOTE.

Type 1 Diabetes Mellitus

There were few MACEs in any of the three studies in T1DM (Studies 3770, 3583, and 3585) and no clear differences between groups (Table 28).

Table 28

BEGIN Trials — Patients With an Adjudicated Major Adverse Cardiovascular Event.

Health-Related Quality of Life

HRQoL was assessed using the SF-36 and TRIM (both TRIM-D and TRIM-HYPO) instruments.

HRQoL was not assessed in DEVOTE. In SWITCH-1 and SWITCH-2, there was no statistically significant difference between IDeg and IGlar in any of the SF-36 subscales. In SWITCH-2, the manufacturer noted that there was an improvement in TRIM-HYPO for IDeg versus IGlar; however, this outcome was not part of the confirmatory end points and no P values were reported (Table 33). There is no MCID for the TRIM-HYPO.

Table 33

Key Efficacy Outcomes — SWITCH-1 and SWITCH-2.

Glycated Hemoglobin

Type 1 Diabetes Mellitus

Change in A1C from baseline to end of treatment was the primary outcome of Studies 3770, 3583, and 3585. IDeg was noninferior to IGlar (Studies 3770 and 3583) and to IDet (Study 3585) in these studies. In Study 3583, the treatment difference between IDeg and IGlar was −0.01 (95% CI, −0.14 to 0.11), and in Study 3585, after 52 weeks, the treatment difference between IDeg and IDet was −0.09 (95% CI, −0.23 to 0.05). In Study 3770, only the flexible IDeg group was tested versus IGlar, and the treatment difference was 0.17 (95% CI, 0.04 to 0.30), which was judged to be noninferior, as the upper limit of the 95% CI for the estimated treatment difference was > 0 and ≤ 0.4% (Table 29).

Table 29

Primary Outcome — Change From Baseline in Glycated Hemoglobin (BEGIN Trials).

Blood Glucose (Fasting)

Type 1 Diabetes Mellitus

Change in FPG was a confirmatory outcome in Studies 3585 and 3583, and as part of the statistical hierarchy, was not tested. In Study 3770, FPG was not part of a statistical hierarchy and also was not tested (Table 30).

Table 30

Key Secondary Outcome — Change in Fasting Plasma Glucose (BEGIN Trials).

Blood Glucose (Variability)

Blood glucose variability was reported in the T1DM trials, and was part of the statistical hierarchy in studies 3585 and 3583; however, testing had been halted by this time, so no P value was reported (Table 31).

Table 31

Key Secondary Outcome — Blood Glucose Variability (BEGIN Trials).

In the T2DM studies, blood glucose variability was part of the statistical hierarchy, but testing had been halted once this end point was reached, or the outcome was not part of the statistical hierarchy, and no P values were reported.

Adverse Events

There were no clear differences in the overall proportion of participants with an adverse event in DEVOTE or in the SWITCH studies. The most common adverse event was nasopharyngitis (Table 37).

Table 37

Harms — DEVOTE, SWITCH-1, and SWITCH-2.

In the BEGIN trials, in Study 3770, there was a numerically lower proportion of IDeg (68%) participants who were on a flexible dosing regimen who experienced an adverse event versus those on a regular IDeg (76%) regimen or on IGlar (72%) (Table 59). In Study 3586, 59% of IDeg-treated and 65% of IGlar participants experienced an adverse event (Table 64). Otherwise, there were no clear differences in the proportion of IDeg versus IGlar participants experiencing an adverse event in Study 3583 (or extension) or Study 3579 or extension (Study 3643), or between IDeg and IDet in Study 3585, while in the extension (Study 3725), 82% of IDeg and 78% of IDet participants had experienced an adverse event by 104 weeks There was no clear difference in the proportion of participants with an adverse event with IDeg compared with sitagliptin in Study 3580.

Serious Adverse Events

There were no clear differences between IDeg and IGlar in the proportion of participants experiencing a serious adverse event in DEVOTE or in the SWITCH studies (Table 37).

In the BEGIN trials, there was no clear difference in the proportion of participants with a serious adverse event between IDeg and IGlar (Table 34). In Study 3725, the extension to Study 3585, 12% of IDeg versus 7% of IDet participants had a serious adverse event. There was no clear difference in proportion of participants with a serious adverse event with IDeg compared with sitagliptin in Study 3580 or with IDeg + liraglutide versus placebo + liraglutide in Study 3944.

Table 34

Harms — BEGIN Trials.

Withdrawals Due to Adverse Events

There were no clear differences in the proportion of participants who withdrew due to an adverse event in DEVOTE or in the SWITCH studies (Table 37). In the BEGIN trials, there were numerically more IDeg participants who withdrew due to an adverse event compared with sitagliptin (4% versus 1%) (Table 34).

Notable Harms

BEGIN trials

Confirmed hypoglycemic events, both overall and nocturnal events, were consistently reported across the BEGIN trials as confirmatory secondary end points. In two of three studies where it was tested in T1DM and in all four comparisons versus IGlar in T2DM, insulin-naive patients, IDeg was not shown to be superior to IGlar (five studies) or IDet (one study) (Table 35). The only study that reported a statistically significant decrease in the risk of hypoglycemia with IDeg versus comparator (IGlar) was Study 3582, with a treatment ratio of 0.82 (95% CI, 0.69 to 0.99; P = 0.018) (Table 35). Data were collected for this outcome in other studies but not as a confirmatory end point, and no P values were reported.

Table 35

Key Secondary Outcome — Confirmed Hypoglycemic Events (BEGIN Trials).

Confirmed hypoglycemic episodes consisted of episodes of severe hypoglycemia as well as minor hypoglycemic episodes with a confirmed plasma glucose value of < 3.1 mmol/L. Hypoglycemic episodes were defined as nocturnal if the time of onset was between 00:01 a.m. and 05:59 a.m., inclusive. Hypoglycemic episodes occurring during sleep in the extended time range from 10:01 p.m. to 07:59 a.m. were also analyzed.

Table 36

Key Secondary Outcome — Confirmed Nocturnal Hypoglycemic Events (BEGIN Trials).