Study design and intervention
The INSPIRE study was a single-center, single-arm, prospective, open-label clinical study that evaluated the effect of a single DMR procedure combined with GLP-1RA (liraglutide), in patients with T2DM, treated with insulin therapy. The study protocol was approved by the medical ethics committee of the Amsterdam University Medical Center. The study was conducted in accordance with ICH Good Clinical Practice Guidelines and the Declaration of Helsinki. The study is registered under EudraCT number 2017-00,349-30 at Clinicaltrialsregister.eu. The primary endpoints of this study have been reported elsewhere [14]. This report presents the results of a pre-defined sub-study of the INSPIRE study investigating the changes in parameters of cardiovascular health after replacing insulin by DMR with GLP-1RA.
Clinical study summary
We included 16 patients with T2DM using basal insulin, aged 28–75 years, with a body mass index of 24–40 kg/m2, a hemoglobin A1c (HbA1c) ≤ 8.0% (64 mmol/mol), and an adequate β-cell reserve (defined as fasting C-peptide > 0.5 nmol/l) [14]. Baseline characteristics can be found in Additional file 1: Table S1. The endoscopic DMR procedure was performed under deep sedation with propofol by a single endoscopist (JB) with experience in endoscopic DMR procedures. The DMR procedure involved circumferential hydrothermal ablation of the duodenal mucosa using an over-the-guidewire catheter, as described previously [12, 13]. Exogenous insulin administration was discontinued immediately after the DMR procedure. Patients were instructed to adhere to a 2 week post-procedural diet (i.e. gradual transition from liquid to solid food to allow adequate regeneration of the duodenal mucosa). After finishing the 2 week post-procedural diet, patients began with self-administration of subcutaneous GLP-1RA, liraglutide (Victoza®, Novo Nordisk A/S). Standard mild nutritional counselling and lifestyle education were provided before DMR and during follow-up [14]. All 16 enrolled patients underwent a successful DMR procedure defined as ≥ 5 sequential ablations of 2 axial centimeters each.
Cardiovascular assessments
At baseline and 6 months after DMR multiple measurements were conducted to assess cardiovascular health and the risk of cardiovascular events. These assessments are listed below.
Visceral and subcutaneous fat volume measurements
During the clinical study, MRI images (MRI; model clinical 3 Tesla scanner, Achieva, Philips) were made to measure liver fat content at baseline and 6 months after DMR. We decided to use these available MRI images to measure abdominal visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT). Body composition measures are usually estimated at the level of lumbar vertebra 3 (L3) or 4 (L4) [15]. As these levels were not consistently available on the MRI examinations of the upper abdomen, measurements at one transversal slide closest to mid L2 were used. VAT and SAT segmentation was performed using manual outlining (mDixon fat images) and semi-automated thresholding using Radiant DICOM viewer (Medixant, Poznan, Poland) by an experienced radiologist. This method is considered the reference standard for the quantitative assessment of intra-abdominal adipose tissue.
Ambulatory blood pressure monitoring
Systolic and diastolic blood pressure, and heart rate were measured during 24 h using an ambulatory blood pressure monitoring (ABPM) device (IEM, Mobil-O-Graph NG ABPM Monitor) at baseline and at 6 months follow-up. The ABPM cuff was placed on the non-dominant arm unless there was a 20/10 mmHg difference between arms, in which case the arm with the higher reading was used. Patients were instructed to maintain normal activity during ABPM and to hold the arm still and at heart level during recording. Recordings were programmed for every 30 min during the day (07.00 to 22.00) and every 60 min during the night (22.00 to 07.00). After 24 h, the monitor was detached and returned to the hospital.
Lipid panel
At baseline and 6 months, blood was drawn to measure fasting total cholesterol, high density lipoprotein (HDL), low density lipoprotein (LDL) and triglycerides. Blood for total cholesterol, HDL, and triglyceride quantification was collected in a heparin tube at room temperature. Samples were brought to the laboratory for processing and analysis. Quantification was done using enzymatic colorimetric test performed on a Cobas c502 machine.
Postprandial insulin and triglycerides
All patients underwent a mixed meal tolerance test (MMTT) to measure postprandial plasma insulin and triglyceride concentrations at baseline and 6 months. All patients were using exogenous insulin at baseline and none were using exogenous insulin at the 6 month follow-up MMTT. If insulin was restarted in patients with an HbA1c > 7.5%, it was restarted after the 6 months follow-up visit. Patients were asked to ingest a liquid meal (Fresubin 200 ml, 2.0 kcal/ml) within 10 min. During the MMTT, blood samples were drawn at 0 min (fasting) and at 15, 30, 45, 60, 90, 120, 180, and 240 min following the start of the meal. Triglyceride quantification was performed with serum heparin tubes at room temperature with use of the enzymatic colorimetric test on a Cobas c702 machine. Insulin quantification was performed on SST tubes with gel and clot activator with use of the ILMA method on an Atellica (Siemens) machine.
Urine microalbumin
All patients delivered a urine sample at baseline and 6 months. Microalbumin quantification was performed with urine monovette tubes at room temperature with use of the immunoturbidimetry on a Cobas c502 machine.
10-years cardiovascular risk score
Ten year risk of cardiovascular events was estimated with use of an online calculator of the ASCVD risk score (ACC/AHA ASCVD Risk Calculator (cvriskcalculator.com). The calculator has been validated and is based on the algorithm published by Goff et al. 2013 in the ACC/AHA Cardiovascular Risk Assessment Guidelines [16]. Calculation of the 10-year risk estimate for ASCVD risk can best be described as a series of steps, in which multiple calculated risks variables, based on pooled data, result in a combined risk estimate. The risk variables include age, gender, race, plasma cholesterol levels, blood pressure values, diabetes and smoking status, and the use of blood pressure-lowering medications.
Estimated life-years free of cardiovascular disease
Life-years free of cardiovascular disease were estimated by using the diabetes lifetime-perspective prediction (DIAL) model, consisting of two complementary competing risk adjusted Cox proportional hazards functions using data from people with T2DM registered in the Swedish National Diabetes Registry (n = 389,366) [17]. The risk variables include age, gender, geographic region, smoking status, history of ASCVD, duration of diabetes, insulin use, systolic blood pressure, body mass index (BMI), plasma cholesterol levels, HbA1c level, glomerular filtration rate (eGFR) and albuminuria values, cholesterol-lowering drug and anticoagulant use.