Lixisenatide in type 1 diabetes: A randomised control trial of the effect of lixisenatide on post‐meal glucose excursions and glucagon in type 1 diabetes patients

Abstract Aims The GLP1 agonist lixisenatide is glucagonostatic and reduces post‐prandial blood glucose (PPBG) in type 2 diabetes. This study investigates its impact in type 1 diabetes (T1D). Methods In a blinded, crossover trial, 25 patients with T1D were randomised to 4 weeks adjunctive treatment with lixisenatide (L) or placebo (P), with a 4‐week washout period. The primary outcome was percentage of 3 hours PPBG in target (4‐10 mmol/L) assessed by CGM before and after treatment. Participants also underwent post‐treatment standardised mixed meal test (MMT, n = 25) and hyperinsulinaemic hypoglycaemic clamp (n = 15). Results PPBG CGM readings in target were similar between L vs P (Mean % ± SE, breakfast 45.4 ± 6.0 vs 44.3 ± 6.0, P = .48, lunch 45.5 ± 5.8 vs 50.6 ± 5.3, P = .27 and dinner 43.0 ± 6.7 vs 47.7 ± 5.6, P = .30). HbA1C was similar between L vs P (64.7 ± 1.6 vs 64.1 ± 1.6 mmol/mol, P = .30). Prandial insulin fell after lixisenatide (dose change −0.7 ± 0.6 vs +2.4 ± 0.7 units/d, P = .004), but basal insulin dose was similar between groups. The post‐MMT glucose area under the curve (AUC) was lower with L than P (392.0 ± 167.7 vs 628.1 ± 132.5 mmol/L × min, P < .001), as was the corresponding glucagon AUC (140.0 ± 110.0 vs 304.2 ± 148.2 nmol/L × min, P < .001). Glucagon and counter‐regulatory hormone values at a blood glucose of 2.4 mmol/L during the hypoglycaemic clamp were similar between L and P. Conclusion In T1D, PPBG values were not altered by adjunctive lixisenatide although prandial insulin dose fell. Glucose and glucagon level during an MMT were significantly lower after lixisenatide, without affecting counter‐regulatory response during hypoglycaemia.


| INTRODUC TI ON
Only 30% of patients with type 1 Diabetes (T1D) achieve a glycaemic goal of HbA1C <7.5% (58 mmol/mol). 1 Post-prandial hyperglycaemia is common in all types of diabetes and may have a role in overall glycaemic control. 2 It is generally agreed that reducing glucose excursion after meal improves overall glycaemic control, 3 although recent reviews have cast a doubt on the effect of short glucose variability on long-term diabetes complications. 4 Glucagon levels are physiologically suppressed at high plasma glucose concentrations. In the absence of diabetes and in T2D, glucagon level is found to be higher after oral glucose compared with isoglycaemic intravenous glucose infusion, although more insulin is secreted when glucose is administered by the oral route.
This has been attributed to the glucagonotropic effect of GIP. 5 In people with T1D with no detectable β cell function, paradoxically high levels of glucagon have been noticed after 50 g oral glucose compared with isoglycaemic glucose infusion. 6 Post-prandial hyperglucagonaemia following a MMT has been found to worsen progressively in the first year after a new diagnosis of T1D, as C-peptide levels decline. 7,8 Nonsuppression of glucagon contributes to post-prandial hyperglycaemia in T1D and therefore may have a role in treatment.
In this study, we investigate its effect in patients with T1D.

| Study design
A single centre, double blind, placebo-controlled crossover trial was performed ( Figure 1). Patients were enrolled from the outpatient clinic at the Oxford Centre for Diabetes Endocrinology and Metabolism. The study was approved by the clinical ethics committee of UK, registered with ISCRTN (No. 00290196), performed according to Good Clinical Practice and and externally monitored. Following informed consent, participants were randomised by a computer-generated programme to receive treatment for 4 weeks with lixisenatide (10 µg/d titrated up to 20 µg/d in 2 weeks if tolerated) or placebo in the morning, along with their usual insulin, in random order with a washout period of 4 weeks between treatments. During the treatment period, the usual dose of insulin was reduced (−20% basal insulin, −50% bolus at breakfast, and −20% bolus at lunch). Participants were advised on insulin titration to maintain blood glucose between 6 and 9 mmol/L, guided by investigators. They received a phone call at the beginning of week 3 to check on side effects, and to advise titration of the trial drug to 20 µg/d. They were advised to follow their usual daily routine.

| Laboratory measurements
Electrolytes and glucose concentrations were measured at the CPA-accredited Oxford University Hospitals NHS Trust biochemistry laboratory (ADVIA 2400 general chemistry analyser).
C-peptide and cortisol were analysed using chemiluminescence immunoassay (ADVIA Centaur analyser using Siemens Healthcare Diagnostics Ltd.). HbA1C was measured using ion-exchange chromatography (Menarini 8160 Diagnostics). Adrenaline, noradrenaline and pancreatic polypeptide were measured using ELISA.
Glucagon was measured using a sandwich ELISA, which uses antibodies to both C-and N-terminal antiglucagon antibodies, and eliminates cross-reactivity with elongated or truncated forms of glucagon peptide. 10

| Statistical analysis
The sample size was calculated based on the percentage of readings between 4 and 10 mmol/L derived using CGM for 3 hours post-prandial period. As no data were available for lixisenatide in any previous cohort, best estimates derived from similar data sets 11 have been used for our power calculation. The mean of the F I G U R E 2 Screening, randomisation and retention

| Subjects and effect on baseline characteristics
Patients were recruited from outpatient clinics of Oxford University Hospitals NHS Trust by the investigators, from January 2014 to August 2016. Recruitment was closed after 30 subjects had been screened, and 27 were randomised ( Figure 2). A total of 25 participants completed the trial as two stopped after the first treatment period (one who became pregnant, and one who withdrew consent). Baseline characteristics of the participants are summarised in Table 1.
There was no difference in mean HbA1c after treatment between lixisenatide or placebo. There was no difference in mean BMI or body weight when comparing groups, before or after treatment.
However while the body weight did not change after treatment with placebo, a small but significant reduction was noticed after lixisenatide (Table 2).

| Primary end-point
Continuous glucose monitoring data were analysed in all 25 data sets. After excluding incomplete data, comparison between groups  Figure 3A, Table S1).

| Insulin requirement
There was a numerically lower total basal insulin dose after treatment in both groups, with a greater reduction seen after lixisena- this was not significant between groups before or after treatment ( Figure 3B, Table S1).  Figure 3B, Table S1).  Table S2).

| Hyperinsulinaemic hypoglycaemic clamp
A total of 15 patients agreed to undergo the hypoglycaemic clamps, but one individual only completed one clamp and so this data was excluded, leaving 14 paired clamps to be analysed. Mean plasma glucose level at the start of the clamp was the same between groups, and fell steadily during the clamp in both groups to euglycaemic level between 90 to 135 minutes ( Figure 5A, Table S3).

| Adverse events
A total of 840 adverse events were reported in 27 patients during the study (Table 3). One serious adverse event was reported during treatment with placebo, when loss of consciousness resulted in hospital attendance. Blood glucose testing at site confirmed that this was not related to hypoglycaemia. None of the adverse effects resulted in withdrawal from the study. Another participant became pregnant during treatment with placebo did not enter the lixisenatide arm and was excluded from the study.
The majority of the adverse events (95% of total) reported by participants related to hypoglycaemia (defined as capillary blood glucose <4 mmol/L). Reported hypoglycaemic episodes in the pre-

| D ISCUSS I ON
In this first study with lixisenatide in patients with T1D, 4-week exposure to lixisenatide did not make significant difference to the pre- In the first trial with liraglutide, 4-week treatment did not make any difference to glucose levels, although reduced insulin requirement was observed. In longer trials, liraglutide was found to result in small but significant improvement in HbA1C level after 12, 26 and 52 weeks treatment. [13][14][15] In our trial, similar to liraglutide, 12 the dose of insulin was significantly reduced at the start of the treatment to avoid risk of hypoglycaemia.
This study comprehensively investigates the effect of lixisenatide on post-prandial blood glucose in a real life setting, as well as experimental conditions.
In summary, our study raises the possibility that in selected patients, a short-acting GLP-1 receptor agonist could be a useful adjunctive treatment in T1D to limit post-prandial glucose rise.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

AUTH O R CO NTR I B UTI O N S
Chitrabhanu Ballav conceived and designed the study under su-

E TH I C A L A PPROVA L
The study was performed in accordance with good clinical practice and was approved by the clinical ethics committee in UK (ISRCTN No. 00290196). Patients were recruited from outpatient clinics at the Oxford Centre for Diabetes, Endocrinology and Metabolism.
Informed consent was obtained prior to participation, and the study was monitored by an external monitor.

DATA AVA I L A B I L I T Y S TAT E M E N T
Individual participant data will not be available.