Conversion from insulin glargine U-100 to insulin glargine U-300 or insulin degludec and the impact on dosage requirements

Abstract

Background:

We wanted to determine whether basal insulin requirements change when patients transition from insulin glargine U-100 (Gla-100) to insulin glargine U-300 (Gla-300) or insulin degludec.

Methods:

This study involved subjects seen in the University of Colorado Health Endocrine Clinic who were transitioned from Gla-100 to either Gla-300 (n = 95) or insulin degludec (n = 39). The primary outcome was the difference between baseline Gla-100 dose and dose of Gla-300 or insulin degludec prescribed after first follow-up visit within 1–12 months. Secondary outcomes included changes in glycemic control and empiric dose conversion from Gla-100 to Gla-300 or insulin degludec on the day of transition. Wilcoxon rank sum tests evaluated changes in insulin doses, and paired t tests assessed changes in glycemic control using GraphPad statistical software.

Results:

Median daily basal insulin dose increased for individuals transitioned from Gla-100 to Gla-300 from 30 [19–60 interquartile range (IQR)] units at baseline to 34.5 (19–70 IQR) units after follow up (p = 0.01). For patients transitioned to insulin degludec, dose changes from baseline to follow up were not significantly different (p = 0.56). At the time of transition, the prescribed dose of Gla-300 or insulin degludec did not significantly differ from the previous dose of Gla-100 (p = 0.73 and 0.28, respectively), indicating that empiric dose adjustments were not routinely prescribed.

Conclusions:

Patients who transitioned from Gla-100 to Gla-300 had increased basal insulin requirements between visits, while basal insulin requirements for those transitioned from Gla-100 to insulin degludec were not significantly different.

Introduction

Around 29 million Americans are living with type 1 or type 2 diabetes, with an estimated 28% of these people using insulin.¹ Basal insulin is popular among prescribers and patients due to its long duration of action and ease of use. The ideal basal insulin would have a peakless pharmacokinetic profile, 24-h duration of action allowing for once-daily injections, and a minimized risk of hypoglycemia. Insulin glargine U-100 (Gla-100) is an acidic solution which forms a precipitate in the subcutaneous tissue upon injection and releases slowly over time.² With an average 3-h onset and 20-h duration of action, Gla-100 offers longer coverage and a less pronounced peak effect when compared with neutral protamine Hagedorn (NPH) insulin or insulin detemir.^2,3 Still, at lower doses of Gla-100, many patients require twice daily dosing to maintain adequate control throughout the day, and at higher doses, patients may experience discomfort from injecting large volumes.⁴ Therefore, certain subsets of people may benefit from a basal insulin with a differing pharmacokinetic profile.

The approvals of insulin glargine U-300 (Gla-300) and insulin degludec in 2015 opened the door to a new and unique selection of basal insulins.^5,6 Gla-300 is three times more concentrated than Gla-100, develops its onset over 6 h, and has a duration of action of 36 h.² These characteristics are thought to provide a more even distribution of insulin throughout the day compared with Gla-100. Gla-300 has also been shown to be equally safe and effective when injected 3 h before or after the scheduled dosing time.⁷ The Phase III EDITION trials showed subjects taking Gla-300 to have similar glycemic control and lower rates of nocturnal hypoglycemia compared with those taking Gla-100, although those receiving Gla-300 required doses 12–18% higher than those receiving Gla-100.^2,8–11 It is postulated that this higher dosing requirement may be due to a mildly lower bioavailability for Gla-300, perhaps related to a prolonged residence time in the subcutaneous tissue and subsequent tissue peptidase interactions.¹²

Insulin degludec is formulated with zinc and forms a depot composed of multihexamer chains after being injected into the subcutaneous space.¹³ As the zinc molecules dissociate, single monomers of insulin degludec are slowly released into the bloodstream over time. Insulin degludec has an average 42-h duration of action and provides steady insulin levels throughout the day.³ Its pharmacokinetics also allow for more flexible dosing, with subsequent injections permissible 8–40 h after the previous dose. Insulin degludec is currently available in U-100 and U-200 formulations. Despite differences in concentration, these two products exhibit equivalent pharmacokinetic and pharmacodynamic profiles.¹⁴ Insulin degludec has shown similar glycemic control with lower rates of nocturnal hypoglycemia compared with Gla-100.^15–18 To achieve these outcomes, the Phase III BEGIN studies showed that 11–12% lower doses of insulin degludec were required compared with Gla-100.

Patients may transition from one basal insulin to another for many reasons, including their dosing flexibility and convenience, lower injection volume, and seemingly lower rates of hypoglycemia. While the choice of basal insulin should be made based on factors such as disease and drug characteristics, efficacy, and safety, nonmedical switching often occurs due to cost, insurance formulary preferences, and during care setting transitions.¹⁹ Although clinical studies found subjects to have different dose requirements when taking Gla-100 as opposed to Gla-300 or insulin degludec, the package insert for each of these new basal insulin products recommends switching patients previously taking Gla-100 by a 1:1 unit dosage conversion.^5,6 In addition, current guidelines do not offer specific recommendations for switching between basal insulin products.

Despite these recommendations and findings in the literature, it is unclear what happens in real world clinical practice when patients are transitioned from one basal insulin to another. If patients are transitioned to these alternative insulins with a 1:1 unit dosage conversion, they may experience changes in glycemic control prior to being seen again in clinic. Therefore, the purpose of this study was to retrospectively examine a population of patients with type 1 and type 2 diabetes managed in a diabetes specialty clinic, who were transitioned from Gla-100 to either Gla-300 or insulin degludec and assess their dosage conversion. Documentation of their first follow-up visit was also examined to determine whether or not doses required adjustment and to assess blood sugar control following this transition. If concluded that dose adjustments were necessary to maintain appropriate glycemic control, these findings could be used to help develop dosing recommendations when switching between these basal insulins.

Methods

This retrospective evaluation was conducted at the Endocrinology, Diabetes and Metabolism Clinic at University of Colorado Health (UCHealth) Anschutz Campus in Aurora, Colorado. This is an urban, multidisciplinary clinic which consists of 28 providers on an academic medical campus and provides comprehensive management of diabetes and other endocrine conditions. This study was approved by the Colorado Multiple Institutional Review Board. The objectives were to determine if dose adjustments were required after switching from Gla-100 to Gla-300 or insulin degludec and examine how doses of Gla-100 were converted to Gla-300 or insulin degludec in clinical practice. Data were collected from the electronic medical record through an automatic report identifying all patients with type 1 or type 2 diabetes seen in the UCHealth Endocrine Clinic and transitioned from Gla-100 to Gla-300 or insulin degludec (U-100 or U-200) from March 2015 to March 2017. Patients also required a follow-up visit in the UCHealth Endocrine Clinic 1–12 months after transition, in order to have complete data collected. Exclusion criteria included age < 18 years, pregnancy, and taking Gla-100 for less than 3 months prior to transition. Baseline glycated hemoglobin A1c (A1c) measurements were only included for analysis if drawn within 1 month prior to transition.

All data for the Gla-300 and insulin degludec groups were evaluated separately. The primary outcome was the difference between baseline Gla-100 dose and dose of Gla-300 or insulin degludec prescribed after the first follow-up visit within 1–12 months. Secondary outcomes included an analysis of how doses were converted from Gla-100 to Gla-300 or insulin degludec and a comparison of baseline Gla-100 dose with doses of Gla-300 or insulin degludec which participants reported taking when returning for follow up. Additional secondary outcomes included changes in A1c, fasting blood glucose readings and rates of hypoglycemia from time of transition to follow-up visit.

Baseline data collected at time of transition included Gla-100 dose, A1c, self-monitoring of blood glucose (SMBG) download data from the previous month, rate of hypoglycemic events per 7 days, weight, other diabetes medications and doses, and dose of new basal insulin at transition. At the patient’s first follow-up visit taking place 1–12 months after transition, A1c, SMBG download data from the previous month, rate of hypoglycemic events per 7 days, other diabetes medications and doses, and doses of Gla-300 or insulin degludec when reporting for visit, along with doses prescribed after this visit were recorded. If the patient’s first visit 1–12 months after transition did not include an A1c measurement, and the patient had a later visit within the applicable timeframe which did include this measurement then the subsequent visit was used for data collection. Documentation of SMBG readings and hypoglycemic events was taken from provider encounter notes. Average fasting blood glucose readings were manually verified with each subject’s glucometer reading upload. Dose changes were verified by comparing provider notes with medication prescription records. For descriptive purposes, insulin doses were differentiated as baseline Gla-100 dose (Point A), dose of Gla-300 or insulin degludec prescribed after initial visit (Point B), dose of Gla-300 or insulin degludec which the patient was taking when presenting for follow up 1–12 months after transition (Point C), and dose of Gla-300 or insulin degludec prescribed after follow-up visit (Point D).

Statistics were computed using GraphPad InStat, version 3.10, for Windows, GraphPad Software, San Diego, California, USA. Wilcoxon rank sum tests evaluated changes in median insulin doses across time points. Paired t tests assessed changes in A1c, average SMBG readings, and rates of hypoglycemic events per 7 days. Descriptive statistics assessed baseline characteristics and concomitant medication changes. Using data from the EDITION 2 trial (average insulin dose 0.67 units/kg/day, standard deviation 0.24), it was estimated that 101 participants were needed in each group in order to detect a 10% change with regard to the primary outcome, with an 80% power and an alpha value of 0.05.²⁰ Since each subject served as their own control in this study, 51 participants were therefore needed in each the Gla-300 and insulin degludec groups.

Results

A total of 131 subjects taking Gla-100 at baseline were included in the study. There were 95 patients prescribed Gla-300, and 39 were prescribed insulin degludec. In the Gla-300 group, 55 of the 95 subjects (58%) had a follow-up visit 1–12 months after transition and were still taking Gla-300 at that time. In the insulin degludec group, 17 of the 39 subjects (44%) had a follow-up visit 1–12 months after transition and were still taking insulin degludec. The mean time to follow up in those patients was 3.5 months in the Gla-300 group and 3.9 months in the insulin degludec group (Figure 1).

Figure 1.

Study enrollment.

Gla-100, insulin glargine U-100; Gla-300, insulin glargine U-300.

Baseline demographic information for subjects is listed in Table 1.

Table 1.

Demographic data.

	Gla-300	Insulin degludec
Total subjects	95	39
Men	48 (51%)	21 (54%)
Women	47 (49%)	18 (46%)
Type 1 diabetes	34 (36%)	18 (46%)
Type 2 diabetes	61 (64%)	21 (54%)
Average age (years)	57.7	54.8
Average disease duration (years)	19.5	19.5
Average weight (kg)	89.6	95.0
Average BMI (kg/m2)	31.2	31.0

Gla-300, insulin glargine U-300; BMI, body mass index.

With regard to the primary outcome, median daily basal insulin dose increased for patients transitioned to Gla-300 from 30 [19–60 interquartile range (IQR)] units at Point A to 34.5 (19–70 IQR) units at Point D (p = 0.01) (Table 2).

Table 2.

Change in daily median insulin dose after follow-up visit versus baseline.

	Baseline Gla-100 dose (Point A)	Point D	p value
Gla-300 (n = 54)
Units; median (IQR)	30 (19–60)	34.5 (19–70)	0.01
Units/kg; median (IQR)	0.35 (0.24–0.62)	0.39 (0.25–0.65)	0.01
Insulin degludec (n = 17)
Units; median (IQR)	60 (27–90)	66 (26–76)	0.56
Units/kg; median (IQR)	0.61 (0.34–0.99)	0.60 (0.30–0.86)	0.45

Gla-300, insulin glargine U-300; IQR, interquartile range.

There was no significant change in median daily basal insulin doses in the insulin degludec group, from 60 (27–90 IQR) units at Point A to 66 (26–76 IQR) units at Point D (p = 0.56). Results were similar when computed as total daily units or units per kilogram for each group (Table 2). Doses across time points for the initial and follow-up visits are shown in Figure 2.

Figure 2.

Change in insulin doses across time points.

^*p < 0.05 versus Point A.

^YFor Point A to Point C, the comparison was conducted with only 52 patients who had complete data points. For these comparisons, Point A = 29 (19–59 IQR) units and 0.34 (0.24–0.64) units/kg, respectively. There were similar findings when evaluated as units/kg.

Gla-100, insulin glargine U-100; Gla-300, insulin glargine U-300; Point A = Gla-100 dose upon presentation to initial visit; Point B = Gla-300 or insulin degludec dose prescribed after initial visit; Point C = Gla-300 or insulin degludec dose at presentation to follow-up visit; Point D = Gla-300 or insulin degludec dose prescribed after follow-up visit; IQR, interquartile range.

The doses of Gla-300 at Point C were also significantly higher than baseline doses of Gla-100 at Point A when computed as either units or units per kilogram, indicating that participants had either self-titrated or been instructed to titrate their dose before being seen again (Table 3). This difference was not significant in the insulin degludec group.

Table 3.

Insulin dose by time point.

	Baseline Gla-100 dose (Point A)	Point B	Point C	Point D
Gla-300 (n = 54)
Units; median (IQR)	30 (19–60)	31.5 (18–60)	31.5 (19–61)	34.5 (19–70)
p value (versus Point A)		0.56	0.02*	0.01
Units/kg; median (IQR)	0.35 (0.24–0.62)	0.33 (0.25–0.58)	0.36 (0.25–0.63)	0.39 (0.25–0.65)
p value (versus Point A)		0.45	0.02*	0.01
Insulin degludec (n = 17)
Units; median (IQR)	60 (27–90)	60 (27–84)	60 (25–84)	66 (26–76)
p value (versus Point A)		0.27	0.24	0.56
Units/kg; median (IQR)	0.61 (0.34–0.99)	0.61 (0.33–1.05)	0.53 (0.30–0.87)	0.60 (0.30–0.86)
p value (versus Point A)		0.89	0.17	0.45

^*For Point A to Point C, the comparison was conducted with only 52 patients who had complete data points. For these comparisons, Point A = 29 (19–59 IQR) units and 0.34 (0.24–0.64) units/kg, respectively.

Point A = Gla-100 dose upon presentation to initial visit.

Point B = Gla-300 or insulin degludec dose prescribed after initial visit.

Point C = Gla-300 or insulin degludec dose at presentation to follow-up visit.

Point D = Gla-300 or insulin degludec dose prescribed after follow-up visit.

Gla-100: insulin glargine U-100; Gla-300: insulin glargine U-300; IQR: interquartile range.

The A1c, SMBG readings and rates of hypoglycemia per 7 days did not significantly differ between initial and follow-up visits for either the Gla-300 or insulin degludec groups (Table 4).

Table 4.

Changes in glycemic control.

	Baseline	Follow up	p value
Gla-300
Hemoglobin A1c % (mmol/mol) (n = 34)	7.9% (63 ± 14)	8.0% (64 ± 16)	0.50
Overall average SMBG (mmol/l) (n = 29)	9.19 ± 1.88	9.54 ± 2.15	0.25
Average fasting SMBG (mmol/l) (n = 28)	9.17 ± 2.29	9.35 ± 2.22	0.70
Hypoglycemia (per 7 days) (n = 45)	1.15 ± 1.8	0.86 ± 1.3	0.34
Insulin degludec
Hemoglobin A1c % (mmol/mol) (n = 12)	8.1% (65 ± 14)	7.6% (60 ± 7)	0.19
Overall average SMBG (mmol/l) (n = 9)	10.7 ± 1.56	10.2 ± 1.64	0.41
Average fasting SMBG (mmol/l) (n = 7)	9.91 ± 2.85	9.78 ± 1.97	0.89
Hypoglycemia (per 7 days) (n = 12)	1.25 ± 1.5	1.27 ± 1.4	0.94

Baseline: visit at which patient was converted from Gla-100 to Gla-300 or insulin degludec.

Follow up: first visit 1–12 months after conversion from Gla-100 to Gla-300 or insulin degludec. Data were only included for patients who were still taking Gla-300 or insulin degludec at this visit.

A1c, glycated hemoglobin A1c; Gla-300, insulin glargine U-300; SMBG, self-monitoring of blood glucose.

Baseline A1c measurements were 7.9% (63 ± 14 mmol/mol) and 8.1% (65 ± 14 mmol/mol) in the Gla-300 and insulin degludec groups, respectively.

At the time of transition, the prescribed dose of Gla-300 or insulin degludec at Point B did not significantly differ from the previous dose of Gla-100 at Point A (p = 0.73 and 0.28 respectively), indicating that empiric dose adjustments were not routinely prescribed (Figure 3).

Figure 3.

Change in insulin dose at time of conversion.

Point A = Gla-100 dose upon presentation to initial visit.

Point B = Gla-300 or insulin degludec dose prescribed after initial visit.

Gla-100, insulin glargine U-100; Gla-300, insulin glargine U-300.

In a post hoc subgroup analysis, it was found that patients with type 1 diabetes who were switched to Gla-300 required significantly higher doses of basal insulin at Point D compared with Point A (Table 5). In patients with type 2 diabetes who were switched to Gla-300, basal insulin dose requirements were numerically higher, but not significantly different between these time points. There were no significant differences in basal insulin dose for patients with either type 1 or type 2 diabetes switched to insulin degludec from Point A to Point D.

Table 5.

Subgroup analysis: change in daily median insulin dose after follow-up visit versus baseline.

		Baseline Gla-100 dose (Point A)	Point D	p value
Gla-300 (n = 54)
Type 1 diabetes (n = 18)	Units; median (IQR)	17 (12–24)	17.5 (14–25)	0.04
	Units/kg; median (IQR)	0.23 (0.17–0.35)	0.26 (0.22–0.35)	0.04
Type 2 diabetes (n = 36)	Units; median (IQR)	44 (24–71)	49 (26–74)	0.06
	Units/kg; median (IQR)	0.41 (0.30–0.70)	0.49 (0.34–0.72)	0.08
Total subjects without changes in concomitant medications (n = 39)	Units; median (IQR)	28 (19–60)	33 (19–70)	0.03
	Units/kg; median (IQR)	0.32 (0.24–0.62)	0.36 (0.26–0.62)	0.03
Total subjects with changes in concomitant medications (n = 15)	Units; median (IQR)	34 (23–60)	35 (24–58)	0.18
	Units/kg; median (IQR)	0.41 (0.29–0.62)	0.42 (0.28–0.61)	0.08
Insulin degludec (n = 17)
Type 1 diabetes (n = 8)	Units; median (IQR)	26.5 (22–32)	26 (20–26)	0.16
	Units/kg; median (IQR)	0.34 (0.30–0.36)	0.30 (0.25–0.35)	0.11
Type 2 diabetes (n = 9)	Units; median (IQR)	90 (75–100)	76 (68–100)	0.94
	Units/kg; median (IQR)	0.99 (0.87–1.23)	0.86 (0.74–1.18)	0.99
Total subjects without changes in concomitant medications (n = 12)	Units; median (IQR)	70 (31–97)	68 (25–85)	0.70
	Units/kg; median (IQR)	0.74 (0.37–1.19)	0.63 (0.34–1.01)	0.56
Total subjects with changes in concomitant medications (n = 5)	Units; median (IQR)	32 (25–60)	27 (26–66)	0.81
	Units/kg; median (IQR)	0.34 (0.33–0.64)	0.37 (0.27–0.74)	0.63

Point A = Gla-100 dose upon presentation to initial visit.

Point D = Gla-300 or insulin degludec dose prescribed after follow-up visit.

Gla-100, insulin glargine U-100; Gla-300, insulin glargine U-300; IQR, interquartile range.

Additional antihyperglycemic medications were added or dose escalated at some point during the study period in 19% (10/54) of people taking Gla-300, and 50% of these changes were in bolus insulin regimens. Concomitant medications were discontinued or dose reduced in 9% (5/54) of subjects, and 80% of these changes were in bolus insulin regimens. Among the patients who did not experience any changes in concomitant medications, the increase in basal insulin dose requirements was similar to that seen in the overall Gla-300 population from Point A to Point D, and the change in dose was still statistically significant (Table 5). In the insulin degludec group, 18% (3/17) of participants had in increase in dose or addition of concomitant antihyperglycemic medication, while 12% (2/17) had a dose reduction or discontinuation. In patients who did not experience changes in concomitant medications, there was still no significant difference in basal insulin dose from Point A to Point D.

Discussion

Patients transitioned from Gla-100 to Gla-300 had increased basal insulin requirements, while basal insulin requirements for those transitioned from Gla-100 to insulin degludec were not significantly different. The 15% increase in median dosage requirements seen in the Gla-300 group was in concordance with phase III trials which found participants taking Gla-300 to have 12–18% higher dosage requirements compared with those taking Gla-100.^2,8–11 Clinical trials found subjects taking insulin degludec to have 11–12% lower dosage requirements compared with those taking Gla-100, and additional randomized, cross-over trials published after the initiation of this study found participants with type 1 and type 2 diabetes to have 3% and 4% lower dosage requirements when taking insulin degludec than when taking Gla-100, respectively. ^{15–18,21,22} However, our study was not powered to detect the anticipated dosage change in the insulin degludec group due to small sample size. In our study, doses were converted at an average 1:1 basis for both Gla-300 and insulin degludec, as is recommended in the package insert for each respective insulin. If people who are transitioned from Gla-100 to Gla-300 or insulin degludec have variable new insulin requirements, as our study and previously completed trials suggest then converting these doses on a 1:1 basis could be problematic because patients may be subject to a period of inadequate glycemic control. ^{2,8–11,15–18,21,22} No significant changes were seen with regard to glycemic control in our study, although these analyses were also limited by low subject numbers.

In this study, 54% of individuals prescribed Gla-300 or insulin degludec either did not fill the prescription or did not have a follow-up visit in the endocrine clinic within 1–12 months after transitioning. In most cases, the absence of follow up was due to insurance not covering the prescribed insulin. Of individuals who filled the prescription and had a follow-up visit within 1–12 months, the average time to follow up in the endocrine clinic was around 3–4 months. This standard follow up may have indicated that providers felt comfortable with their patients’ glycemic control at the time of transition and did not expect there to be significant changes when switching from Gla-100 to Gla-300 or insulin degludec. While the average A1c at the time of transition was around 8% (64 mmol/mol), given that many people seen in a specialized endocrine clinic have complicated and difficult-to-control disease, this may have demonstrated adequate control based on their individual goals of care. The time to follow up would also potentially be longer if insulin conversion occurred in a primary care clinic which does not specialize in diabetes management. Additionally, as shown in Table 3, there was a significant difference in insulin dose when comparing Point A and Point C in the Gla-300 group, which indicated patients had already increased their dose prior to being seen for follow up. If patients had already titrated their dose prior to their visit then these patients may have found it permissible to delay following up in clinic.

These findings must be interpreted with caution, as this was a retrospective analysis based on data collected from electronic medical records and is therefore subject to bias. Many of the data for SMBG readings and hypoglycemic episodes were extracted from provider encounter notes, which may not have been comprehensive. External factors such as changes in concomitant medications and medication adherence could have impacted these results, and a number of patients were lost to follow up. There was great variability in baseline insulin doses, likely as a result of the study assessing both patients with type 1 and type 2 diabetes. This contributed to the non-normal distribution of data and limited statistical analyses. Post hoc subgroup analyses were conducted for patients with both type 1 and type 2 diabetes, but they were not adequately powered to draw conclusions. Our power calculation was conducted using insulin dosing data from a previously published study, and it is possible that demographics for the patients in this study differed from those in our clinic. The anticipated sample size was not met in the insulin degludec group for the primary outcome measure, and it was difficult to assess changes in glycemic control because many individuals’ insurances did not cover the newly prescribed insulin. Therefore, patients were not always transitioned, despite being written a prescription for Gla-300 or insulin degludec. Interestingly, this indicated that many providers in our clinic chose to prescribe Gla-300 or insulin degludec for clinical reasons (e.g. lower risk of nocturnal hypoglycemia or dosing flexibility) rather than out of obligation due to insurance formulary preferences.

Despite these limitations, the findings from this study support evidence from phase III trials which found patients taking Gla-300 to have higher dosage requirements than those taking Gla-100. As the 15% dose increase seen in our study was in line with the 12–18% increases reported in the literature for Gla-300, these findings strengthen support for empiric dose increases when switching from Gla-100 to Gla-300. Dose requirements for individuals transitioned to insulin degludec warrant further study, as we were not adequately powered to detect changes in these doses. Regardless, if patients are converted on a 1:1 basis from Gla-100 to either of these insulins, it is important that providers ensure timely follow up or instruct patients regarding dose titration to prevent lapses in control.

Conclusion

Patients who transitioned from Gla-100 to Gla-300 had increased basal insulin requirements between visits, while basal insulin requirements for those transitioned from Gla-100 to insulin degludec were not significantly different.