Type 2 diabetes mellitus (T2DM) is a progressive disease characterized by impaired glucose tolerance and deterioration of β-cell function [1]. With increase of the ageing population and changes in lifestyle, the estimated prevalence of diabetes in Korea was 16.7% in 2020 [2], and the mortality associated with T2DM was relatively higher than other countries from the Organisation for Economic Co-operation and Development (OECD) [3]. Given the increased morbidity, mortality, and health care expenditures associated with T2DM, the Clinical Practice Guidelines of the Korean Diabetes Association have emphasized the importance of individualised care and early initiation of oral antidiabetic drugs (OADs) followed by glucagon-like peptide-1 receptor agonists or insulins in parallel [4].

In general, basal insulin is administered as the first form of the insulin therapy for T2DM. The second-generation basal insulin analogues such as insulin degludec (IDeg), provides effective glycemic control with lower rates of hypoglycemia owing to their longer duration of action and flat pharmacodynamic profile [5-7]. When uncontrolled with basal insulin, the addition of prandial insulin or switching to premix insulin can be a feasible option of insulin intensification in routine clinical practices [4,5].

Insulin degludec/insulin aspart (IDegAsp) is a fixed ratio co-formulation of IDeg (70%) and insulin aspart (IAsp, 30%). It can be administered simply while each of insulin analogue’s components preserve their pharmacodynamic nature [8]. Given the shorter duration of action of basal insulin formulated in other premix insulins, IDegAsp can provide both better coverage of whole-day insulin needs with a lower day-to-day glucose variability (with an ultra-long duration of action with IDeg) and rapid mealtime onset (with relatively short duration of action with IAsp) for optimal glucose control with fewer injections [9]. These clinical benefits of the fixed ratio co-formulation of IDeg and IAsp were proven in several randomized controlled studies [10-14]. Although these randomized controlled studies found that IDegAsp was efficacious and safe, however, the generalizability of these findings into a real-world setting remains unclear. To date, several observational studies have been conducted but it remains uncertain whether IDegAsp results in better clinical outcomes in routine clinical settings due to heterogeneous baseline clinical and socioeconomic factors, previous treatments, and differences in study design. Recent real-world data from East Asia showed that even with the same ethnicity, glycemic control can be different across countries given different strategies of insulin intensification, practices of education and monitoring, and titration [15]. Furthermore, there is very limited data in patients uncontrolled with other premix insulins whereas 2021 Clinical Practice Guidelines in Korea positioned both basal insulin and premix insulin in parallel [4].

Therefore, this study aimed to investigate the efficacy and related clinical outcomes of switching uncontrolled patients with T2DM with basal or premix insulin to IDegAsp in routine clinical practice in Korea.

This study was a multicenter, retrospective study based on a medical chart review that consisted of two consecutive 26-week treatment periods. The medical records of participants who were switched to IDegAsp from March 6, 2018 to September 2, 2020, were reviewed and collected.

Patients with T2DM were enrolled if aged ≥19 years at the time of IDegAsp initiation; treated with basal or premix insulin±OADs for at least 26 weeks followed by changing the regimen to IDegAsp for at least 26 weeks. Patients treated by continuous subcutaneous insulin infusion prior to receiving IDegAsp were excluded from the study. The decision to change the regimen depended on each investigator’s clinical judgment in routine clinical practice in accordance with Korea Prescribing Information. IDegAsp was administered once or twice daily at each investigator’s discretion. Data were collected 26 weeks before (defined as the “pre-switching” period) and 26 weeks after switching to IDegAsp (defined as the “post-switching” period), allowing a time window of ±6 weeks for each period. The week 0 timepoint was defined as the timepoint where IDegAsp was initiated (also defined as the baseline) (Supplementary Fig. 1).

The primary endpoint was to compare the degree of glycosylated hemoglobin (HbA1c) change between the pre-switching and the post-switching period. The secondary endpoints were to compare the changes in fasting plasma glucose (FPG), body weight, daily total insulin doses, and the incidence of overall and severe hypoglycemia during each 26-week period. The proportions of patients achieving HbA1c <7%, and <7% without any hypoglycemia were also assessed. The severity of hypoglycemia was categorized as per the Korean Clinical Practice Guidelines, defined as either overall hypoglycemia (levels 1, 2, and 3) and severe hypoglycemia (level 3) [4]. Subgroup analysis was conducted according to the previous insulin regimen before switching to IDegAsp.

The data analysis was performed using IBM SPSS Statistics for Windows, version 21 (IBM Corp., Armonk, NY, USA). Categorical variables were reported as numbers (n) and percentages (%); continuous variables were presented as mean and standard deviation or confidence interval (CI). A Paired t-test was performed to assess changes in HbA1c, FPG, body weight, and daily total insulin dose between pre-switching and post-switching periods, and to determine the significance of the mean difference between these two periods. McNemar’s test was used to assess the proportion of patients achieving HbA1c <7%, and <7% without any hypoglycemia, and the incidence of overall and nocturnal hypoglycemia. In addition, subgroups of previous insulin regimens were compared using analysis of one-way variance (ANOVA) followed by Bonferroni’s post hoc analysis for pairwise differences. Multiple regression analysis was used to describe the relationship between HbA1c from week 0 to week 26 and the independent predictors. Missing data were not imputed, and a two-sided P<0.05 was considered significant. The P value was not adjusted for multiple testing.

The study was conducted in accordance with the Declaration of Helsinki [16]. The study protocol was reviewed and approved by the Institutional Review Boards (IRBs) in all institutions (IRB no. CBNUH 2021-02-001-001, no. SCHCA 2021-02-001, no. KBSMC 2021-02-001, no. VC21RSDI0016, no. 3-2020-0531, no. B-2102/669-102). The study is registered with ClinicalTrials.gov, number NCT04656106.

The present study is a multicenter, retrospective observational study investigating the efficacy and safety of switching to IDegAsp from either basal or premix insulin in patients with T2DM. The findings from this study demonstrated that the switch to IDegAsp led to an improvement in glycemic control without an increase in body weight or hypoglycemia in real-world clinical practice. A 26-week treatment with basal or premix insulin therapy (pre-switching period) was associated with an increase in HbA1c with no significant change in FPG levels, and significantly small increase in body weight (0.63 kg). In contrast, switching to IDegAsp resulted in significant reductions in HbA1c and FPG, and greater proportions of patients achieved a target of HbA1c <7% with no increase of body weight. Importantly, no differences were reported in the incidence of overall, severe, and nocturnal hypoglycemia.

The high baseline HbA1c (8.78%) and FPG (163.03 mg/dL) in this study highlights inadequate glycemic control under real-world conditions in patients receiving insulin regimen. This coupled with other characteristics like long standing T2DM, insulin treatment >6 years, implies a strong unmet need for newer insulin analogues in this patient population.

Previous real-world studies demonstrated the effectiveness and safety of IDegAsp for the treatment of T2DM [15,17-19]. In two studies, compared to basal insulin, IDegAsp showed a significant reduction in HbA1c with no difference in FPG levels [17,18]. However, the changes in total daily insulin doses were inconsistent across studies, which may be attributed to different treatment approaches across countries under real-world clinical practice [15,17-19]. In the a ryzodeg initiation and switch effectiveness (ARISE) study for East Asian population, not only HbA1c but also FPG reduction was demonstrated after switching to IDegAsp. However, in contrast to the present study, the ARISE study included all patients with any antidiabetic medications, except IDegAsp before the switch, and the participants had higher HbA1c levels at baseline compared to the present study [15].

The current study confirms the previous effect of IDegAsp for improving the glycemic parameters in patients with uncontrolled T2DM previously treated with basal or premix insulin (±OADs). Furthermore, our findings examined clinical factors associated with improved HbA1c which were not investigated in real-world data to date. Older age and higher baseline HbA1c levels were the predictors for greater HbA1c reduction at week 26 after switching to IDegAsp. Sex, body mass index, duration of T2DM, duration of insulin therapy, or baseline insulin doses were not associated with HbA1c reduction of IDegAsp.

In the subgroup analysis by insulin regimen in the pre-switching period, significant HbA1c reduction changes between the two periods were shown in patients treated with ‘basal insulin’ or ‘premix insulin’ whereas these changes were not significant in the group previously treated with ‘basal-prandial insulin’ which has limited number of study population (n=20). To the best of our knowledge, there is only one observational study that evaluated the benefits of switching to IDegAsp from basal-prandial insulin therapy, which makes it difficult to compare the results with our study. Different levels of HbA1c at baseline (9.8% for the ARISE study [15] vs. 8.78% for the current study), and differences in titrations and monitoring practices might have contributed to varying results. However, considering the treatment complexity and number of injections, IDegAsp might also provide benefits to those using basal-prandial insulin therapy.

There are several unique points in our study. The study was designed to compare the clinical outcomes 6 months before and after the treatment with IDegAsp, which enabled the investigators to assess the robust benefits of IDegAsp for a given treatment period of intensification under real-world clinical practice. Also, the predictors associated with HbA1c improvement will be able to provide further insights into clinical decisions. The subgroup analysis by previous insulin regimen before switching to IDegAsp may warrant larger, longer-term observational studies to further dissect our findings.

This study also has potential limitations. First, due to the nature of retrospective design, First, due to the nature of retrospective design, there is the possibility that overall and nocturnal hypoglycemia may not have been adequately reported, with a recall bias. Also, the reasons of clinical decision for switching to IDegAsp were not assessed in the present study. It is pertinent to note that they may align with reasons elucidated in previous studies [15,17,19], including the needs for further glycemic control, reduction of hypoglycemia, and improvement of compliance. Second, factors such as insulin adherence, dietary habits, and physical activity levels, that could impact overall glycemic control, were not documented. Finally, subgroup analysis by previous insulin regimen was based on low sample size which may limit our findings in these subgroups. Nonetheless, given that limited data are available on clinical outcomes of IDegAsp in Korean adults with T2DM, our study provides important real-world insights on the use of IDegAsp in insulin intensification. Future investigations with larger sample size and various clinical parameters such as insulin secretion capacity are needed to further elucidate the effectiveness and safety of IDegAsp in the Korean population.

To summarize, the current study demonstrated that when T2DM subjects are uncontrolled with conventional basal insulin therapy or premix insulin therapy, switching to IDegAsp can provide additional clinical benefits including improved glycemic control, with no difference in weight gain or incidence of hypoglycemia.