Type 2 diabetes mellitus (T2DM) is a chronic disease characterised by a progressive decline of β-cell function and/or mass in the presence of insulin resistance that requires timely treatment intensification to achieve and maintain optimum metabolic control.
Currently, basal insulin represents the simplest and most effective method for controlling fasting hyperglycaemia. Nonetheless, only approximately half of all patients achieve target glycated haemoglobin (HbA1c) goals despite adequate dose titration and the achievement of near normoglycaemia, thus indicating a need for additional treatment to control postprandial glucose (PPG) excursions.
This can be achieved using different therapeutic modalities, including:
- A fixed combination of a rapid-acting insulin analog (RAA) and an intermediate-acting insulin, e.g. premixed insulins.
- A combination of a basal insulin and a glucagon-like peptide-1 (GLP-1) receptor agonist.
- A ‘basal-bolus’ regimen e.g. administration of a RAA to ongoing basal insulin before each meal.
More recently, a ‘basal-plus’ stepwise treatment regimen, e.g. a single injection of prandial insulin prior to the meal associated with the largest PPG excursion, has been proposed.
Several clinical trials have demonstrated the efficacy and safety of adding single bolus doses of insulin glulisine to basal insulin glargine, however, the effect of individual factors such as a patient’s age or body mass index (BMI) on the efficacy of this treatment strategy has yet to be clarified.
Therefore, a retrospective analysis of previous studies was performed to evaluate both the efficacy and safety of adding a single bolus of the ‘basal-plus’ regimen in patients with T2DM when stratified by age and BMI.
Study design and patient population
Patient-level data were pooled retrospectively from four randomised, controlled, multicentre parallel-group studies designed to evaluate the efficacy and safety of a ‘basal-plus’ regimen in patients with T2DM, 1-2-3, and a proof-of-concept study.
Participants aged ≥18 years who had a diagnosis of T2DM and who were poorly controlled (HbA1c ≥6.5% or 48mmol/mol) using basal insulin glargine in addition to oral antidiabetic agents (OADs), with both baseline and end of study HbA1c and BMI measurements available, were deemed eligible for inclusion in this pooled study population.
All included patients were initiated on a basal insulin glargine in addition to OADs, to which insulin glulisine was subsequently added once daily (‘basal-plus’ approach) for up to six months. Insulin glargine was titrated to protocol-defined fasting blood glucose (FBG) targets (with the exception of one study in which no titration was undertaken), while insulin glulisine was introduced and dose titrated to protocol-defined preprandial or PPG targets.
Outcomes and clinical end of studies
Demographic and clinical characteristics (gender, age, weight, height, BMI, duration of diabetes and age at first diagnosis of diabetes) as well as antidiabetic drug usage (duration of prior OAD and/or insulin use, age at initiation of OAD and/or insulin use and insulin glulisine administration time) were collected and analysed.
Efficacy of the ‘basal-plus’ insulin regimen was determined from the insulin dose and the change of HbA1c levels from baseline and proportion of patients achieving HbA1c <7% (< 53.0mmol/mol), change of FBG level from baseline and proportion of patients at <110mg/dL (<6.1mmol/L), change of PPG levels from baseline and proportion of patients at <180mg/dL (<10.0mmol/L) and change of weight and BMI over the study periods.
Safety measurements comprised the frequency of episodes of severe hypoglycaemia, nocturnal hypoglycaemia and symptomatic hypoglycaemia (as defined in each trial and determined from data collected during the respective trials.
Efficacy and safety measurements were then analysed following stratification by age (<55, 55–64 and ≥65 years) and BMI (<30kg/m2, <30–35kg/m2 and ≥35 kg/m2).
Due to the requirement for HbA1c and BMI data at baseline and at end of study, the total available number of patients was reduced (hereafter referred to as the ‘analysis population’).
A patient-level meta-analysis was conducted to allow for study-to-study differences.
Descriptive statistics were used to measure and describe clinical characteristics and patient demographics as well as efficacy and safety outcome measurements. p values, unadjusted for study origin, were provided by χ2 test or analysis of variance (ANOVA) when appropriate.
Baseline and end of study efficacy measurements were compared with p values calculated using paired t-tests, a p value <0.05 was used to determine the level of statistical significance, again unadjusted for study origin.
A generalised linear model was used to assess the difference between end of study and baseline measurements for HbA1c, weight and BMI while adjusting for patient age, gender, duration of diabetes and different studies. A multivariate logistic regression model was used to assess the impact of patient characteristics on the risk of hypoglycaemia.
The outcomes were combined across studies using the random effects meta-analysis approach of DerSimonian and Laird.
A total of 711 patients comprised the analysis patient population. The majority (53.3%) were male, mean age 59.9 ± 9.5 years, and the mean known duration of T2DM was 11.0 ± 7.0 years. Prior to the study periods the mean duration of OAD use was 6.5 ± 5.7 years and the mean duration of basal insulin use was 2.2 ± 2.2 years.
The mean age at first OAD and insulin usage was 53.5 ± 9.7 years and 60.6 ± 9.2 years, respectively. The majority of patients received an injection of insulin glulisine at dinner time (41.3%). Of the remaining patients, 35.8% and 22.9% received their dose prior to breakfast and lunch, respectively. The majority of patients were receiving either metformin (77.7%) or sulfonylureas (69.0%) at baseline, with smaller numbers receiving thiazolidinediones, glinides, or other OADs.
At baseline and study endpoint the mean dose of insulin glulisine was 4.91 daily doses (U) and 13.21U, respectively. The mean insulin glargine dose was 36.78 U and 41.91 U, respectively.
The mean dose/weight of insulin glulisine was 0.06U/kg and 0.14U/kg, respectively. The mean insulin glargine dose was 0.40U/kg and 0.45U/kg, respectively.
For the subanalyses of patients by age (<55, 55–64 and ≥65 years) and BMI (<30, 30–34kg/m2 and ≥ 35kg/m2), a total of 711 patients were included. In general, baseline demographics and clinical characteristics were essentially similar across the two cohorts. However, there were significant differences within the two subgroups according to weight, BMI, duration of diabetes and age at first diagnosis.
The addition of a single injection of insulin glulisine at the main meal in patients already receiving an existing therapy of OADs and once-daily basal insulin resulted in significant decreases (mean ± SD) in HbA1c of −0.4 ± 0.1%, FBG 2.8 ± 3.7mg/dL and PPG −58.9 ± 9.1mg/dL over a six-month follow-up. These findings were confirmed in a meta-analysis of changes in HbA1c, FBG and PPG.
Furthermore, after six months of ‘basal-plus’ insulin therapy, more than twice as many patients achieved target HbA1c levels (<7% [<53.0mmol/mol]) at end of study compared with baseline (45.3% versus 20.3%).
Also, a much higher proportion of patients (80.4% versus 40.5%) achieved target PPG levels (<180mg/dL [<10.0mmol/L]), while there were no differences in the proportion of patients achieving target FBG at end of study compared with baseline (<110mg/dL [< 6.1mmol/L], 40.0% versus 44.4%).
Both overall daily doses of basal insulin glargine and insulin glulisine increased from baseline to end of study (+6.9 U to ±15.4 and +8.4 U to ±10.4, respectively). Similar results were observed when insulin doses were expressed per kg body weight (U/kg; +0.07 to ±0.14 and +0.05 to ±0.22, respectively). A small, though statistically significant, increase in body weight (+0.9 to ±4.0kg) and BMI (+0.3 ±1.4kg/m2) was observed.
In the analysis population, 1.7%, 12.4% and 37.9% of patients experienced a severe, nocturnal or symptomatic hypoglycaemia event (as defined in each study protocol), respectively. The mean number of events per year for severe, nocturnal or symptomatic hypoglycaemia was 0.03 (±0.2), 0.6 (±2.3) and 4.7 (±11.4), respectively. Nearly half the population (44.3%) achieved target HbA1c goals without experiencing severe or symptomatic hypoglycaemia (43.0%).
No predictors for an increased risk of severe hypoglycaemia could be identified. When nocturnal hypoglycaemia was considered, both female gender (odds ratio [OR] 1.81) and diabetes duration (OR 1.06) emerged as risk predictors. Moreover, insulin glargine dose (OR 1.02), female gender (OR 1.92) and diabetes duration (OR 1.04) were risk factors for symptomatic hypoglycaemia.
The forest plots from the meta-analyses show changes from baseline to end of study, demonstrating that HbA1c decreased, while weight and BMI increased slightly.
Both BMI and FPG declined across the three age categories, whereas age at diagnosis and diabetes duration increased. There were no differences for HbA1c.
The ‘basal-plus’ regimen resulted in significant reductions in mean HbA1c and PPG within each of the three age groups, with the smallest reduction in HbA1c occurring in the younger age group.
Conversely, there was an increase in mean FBG from baseline to end of study in all three age groups, with a significant change in the <55 years age group compared with the other age groups.
There were small but significant increases in both weight and BMI in all three age groups from baseline to end of study, with the magnitude of the weight and BMI gain markedly lower in the older (55–64 and ≥65 years) age groups.
In the three age groups (<55, 55–64 and ≥65 years), 2.0%, 0.8% and 2.3% of patients experienced a severe hypoglycaemic event (as defined in each study protocol), respectively. There was no significant difference between the three subgroups in the incidence of severe or symptomatic hypoglycaemic events. However, there were more nocturnal hypoglycaemic events in the two older age groups.
There were no significant differences between the BMI subgroups with respect to change in HbA1c, FBG and PPG. For both HbA1c and PPG, the observed decreases were significant in all three BMI groups. An increase in FBG levels was apparent in all three subgroups, which reached significance only in the BMI <30kg/m2 group.
As compared with baseline, the ‘basal-plus’ regimen was associated with an increased proportion of patients achieving target HbA1c levels (<7% [<53.0mmol/mol]), which was more apparent in patients with a higher BMI. Similar findings were observed with respect to FBG and PPG levels.
Both basal insulin glargine and insulin glulisine doses increased from baseline to end of study in all three BMI subgroups, with the highest dose increases observed for the ≥35kg/m2 group (both 12.3U).
Over the duration of the study, there were small but significant increases in both weight and BMI within each BMI group, with no significant differences between the three subgroups.
The percentage of patients experiencing hypoglycaemia was generally low in the three BMI groups (<30kg/m2, 30–34kg/m2 and ≥35kg/m2) at 2.5%, 1.8% and 0.0%, respectively, and virtually absent in those with the highest BMI. In particular, there were no significant differences between the three subgroups in the incidence (events per year) of severe or nocturnal hypoglycaemia. However, there were significant differences in the number of symptomatic hypoglycemic events, with the highest events rate observed in the ≥ 35 kg/m2 group.
The use of ‘basal-plus’ as an initial stepwise insulin treatment regimen, involving a single preprandial injection of insulin glulisine (on a background of basal insulin glargine) given before the meal associated with the largest PPG excursion, can achieve a good therapeutic response with a low risk of hypoglycaemia and weight gain, regardless of the patient’s age or BMI. The findings of our analysis demonstrate the efficacy and safety of a ‘basal-plus’ regimen in the different patient subsets studied, and provide clinicians with a relevant, alternative therapeutic option in this difficult-to-treat population with T2DM.
Reference: Lankisch MR, Del Prato SD, Dain M-P et al. Use of a basal-plus insulin regimen in persons with type 2 diabetes stratified by age and body mass index: A pooled analysis of four clinical trials. Primary Care Diabetes, 2016