Duodenal-Jejunal Bypass Liner for the management of Type 2 Diabetes Mellitus and Obesity

Objective: The aim of this study was to examine the clinical efficacy and safety of the duodenal-jejunal bypass liner (DJBL) while in situ for 12 months and for 12 months after explantation. Summary Background Data: This is the largest randomized controlled trial (RCT) of the DJBL, a medical device used for the treatment of people with type 2 diabetes mellitus (T2DM) and obesity. Endoscopic interventions have been developed as potential alternatives to those not eligible or fearful of the risks of metabolic surgery. Methods: In this multicenter open-label RCT, 170 adults with inadequately controlled T2DM and obesity were randomized to intensive medical care with or without the DJBL. Primary outcome was the percentage of participants achieving a glycated hemoglobin reduction of ≥20% at 12 months. Secondary outcomes included weight loss and cardiometabolic risk factors at 12 and 24 months. Results: There were no significant differences in the percentage of patients achieving the primary outcome between both groups at 12 months [DJBL 54.6% (n = 30) vs control 55.2% (n = 32); odds ratio (OR) 0.93, 95% confidence interval (CI): 0.44–2.0; P = 0.85]. Twenty-four percent (n = 16) patients achieved ≥15% weight loss in the DJBL group compared to 4% (n = 2) in the controls at 12 months (OR 8.3, 95% CI: 1.8–39; P = .007). The DJBL group experienced superior reductions in systolic blood pressure, serum cholesterol, and alanine transaminase at 12 months. There were more adverse events in the DJBL group. Conclusions: The addition of the DJBL to intensive medical care was associated with superior weight loss, improvements in cardiometabolic risk factors, and fatty liver disease markers, but not glycemia, only while the device was in situ. The benefits of the devices need to be balanced against the higher rate of adverse events when making clinical decisions. Trial Registration: ISRCTN30845205. isrctn.org; Efficacy and Mechanism Evaluation Programme, a Medical Research Council and National Institute for Health Research (NIHR) partnership reference 12/10/04.


Primary Objectives
To compare the EndoBarrier with conventional medical therapy, diet and exercise for obesity related type II diabetes and their effectiveness on metabolic state as defined by the International Diabetes Federation (IDF) with an HbA1c reduced by 20%.

Secondary Objectives
To compare the EndoBarrier with conventional medical therapy, diet and exercise for obesity related type II diabetes and their effectiveness on:  Metabolic state as defined by the International Diabetes Federation (IDF) with an HbA1c < 6% ( or < 42 mmol/mol) To estimate the cost-effectiveness of the EndoBarrier device compared with conventional treatment over the trial period (within trial analysis).
To estimate the long-term cost-effectiveness (over 24 months) of the EndoBarrier device compared with conventional treatment and alternative surgical interventions.

Safety Objective
To evaluate the safety of the EndoBarrier device.

Primary Endpoint
 Subjects showing a reduction in HbA1c by 20%, to be measured 12 months from start of treatment.

Secondary Endpoints
Secondary efficacy endpoints for the Endobarrier trial are listed as below:  HbA1c of less than 6% (or 42 mmol/mol)  Microbiome  Biomarkers such as genetic markers

*A detailed breakdown of the above can be found in Appendix A
Additional secondary endpoints investigating the cost benefit of the EndoBarrier device will be assessed by a health economist and defined within a separate document. This will cover the following secondary objectives:  Cost of interventions and related health and social care  QALYs accrued (calculated from area under the EQ-5D curve)  Incremental cost per QALY within the trial period and over the long-term

Introduction
Obesity is a serious medical condition which is increasing in incidence worldwide. Obesity induces metabolic abnormalities which contribute to the development of diabetes mellitus and cardiovascular disease. The treatment modalities currently available for the treatment of obesity (i.e. lifestyle interventions, pharmacotherapy and surgery) have limited long-lasting success in producing major and sustained weight loss and are often associated with undesirable side effects, risks or complications. Therefore the need for new exits and effective strategies is necessary to prevent and reduce obesity and its complications such as T2DM. This study is a randomised, placebo-controlled trial which has been designed to further investigate the potential of the EndoBarrier device as an effective alternative treatment to surgery and existing medical therapies in T2DM.
If the EndoBarrier is effective at achieving long-lasting weight loss and glycaemic control, there is an obvious potential for savings on future health and social care; through the avoidance of diabetes and related complications. However, the overall cost-effectiveness of the EndoBarrier device depends on the balance of health benefits, harms, costs and savings compared with other surgical and medical treatment options. In addition to a 'within trial' economic analysis, modelling will be used to estimate lifetime impacts on morbidity, mortality and expenditure, and hence to evaluate whether the EndoBarrier offers the NHS good value for money.

Study Design
This study is a randomised controlled trial of the EndoBarrier device compared with standard medical therapy for the management of obese subjects with T2DM. Subjects will be randomised to one of the two treatment arms (see Table 1 within section 3.3) of the study via the InForm system (the eCRF database for the study). Over 24 months, the study will be performed at two investigational sites, Imperial College Healthcare NHS Trust in London and University Hospital Southampton NHS Foundation Trust.
Individuals in both study arms will be invited regular medical check-ups including measurement of weight, blood pressure, and blood parameters (HbA1c, cholesterol, triglycerides, fasting blood glucose, insulin), quality of life (EQ-5D) and use of health services, as well as to record any adverse events and medications. Diabetes medication titrations will be conducted by the study diabetologists/endocrinologists in accordance with the guidelines of the American Diabetes Association (see section 5.5.1 for full details). They will also receive routine dietary and exercise counselling as well as telephone counselling from a specialist dietician. The control arm will be invited for routine follow-up visits to review their standard medical care. Both arms will also be invited to undergo a diversity of tests investigating the mechanism of the EndoBarrier. After 12 months, the EndoBarrier will be removed and both arms will be followed up for a remaining 12 months.
In order to investigate the mechanism of the effect of the EndoBarrier Device, the trial is divided into three sub-groups in each arm who will have the following additional assessments at visits 3, 5, 8 and 10, 14 (see section 5.4 and 5.6 for more detail):  Sub-group 1 (minimum n=24): functional MRI, eating behaviour, cognitive assessment, and post-meal gut hormones.
 Sub-group 2 (minimum n=18): insulin clamps  Sub-group 3 (minimum n=18): assessment of taste and food preference, eating behaviour assessment, and post-meal gut hormones Participation in the sub-groups is optional.

Study Population
The EndoBarrier study is recruiting patients between the ages of 18 and 65 that have had T2DM for at least 1 year (HbA1c 7.5-11.0% = 58-97 mmol/mol) and are on oral T2DM medications.
In order to be eligible for the study the patient must not have any of the following:  BMI outside the range of 30-50 kg/m 2  Females of childbearing potential who are pregnant, breast-feeding or intend to become pregnant or are not using adequate or reliable contraceptive methods  Evidence of absolute insulin deficiency as indicated by clinical assessment, a long duration of T2DM and a fasting plasma C-peptide of <333pmol/L  Current use of insulin At the screening visit, the patient will be examined to ensure that none of the exclusion criteria are met. Their general health will be assessed via a series of tests including; body weight, height, waist circumference, blood pressure, ECG, urine dipstick and pregnancy test, blood parameters. Female patients will also be asked to report the last day of their menstrual period, the length of their cycle and the length of their menstruation (bleeding) to ensure they are not pregnant and to monitor any changes in their menstrual cycle during the course of the study.
If the patient does not meet any of the exclusion criteria and no additional complications are encountered during the screening process then once written, informed consent has been received, the patient is eligible for randomisation.
Patients will be given the opportunity to consent for participation in one of the mechanistic subgroups (1-3) of this study. Following informed consent, patients interested in consenting for participation in Sub-group 1 of the study will also be screened for suitability by checking the patient does not self-report to be vegetarian, vegan, gluten or lactose intolerant (to be recorded in the CRF). They will also be checked for their fMRI suitability using these questionnaires: (i) Metal check form -to ensure safety for MRI scanning, as may preclude entry into fMRI study (ii) Handedness Inventory -inability to use a right-handed button pad will preclude entry into fMRI study using Handedness.

Sample Size
The primary end-point of a 20% reduction in HbA1c has been chosen as the International Diabetes Federation produced in June 2011 new guidelines for the conduct of studies in diabetes using bariatric surgery or devices aiming to produce standardisation allowing comparison between studies. To date there are thus no published large patient group studies using this end-point, so using this new endpoint in a well-designed and conducted study will be of scientific value in itself.
Conservatively, it was estimated that 15% of patients in the control arm will achieve the target but believe this to be an overestimate. The Steno study is the best quality randomised study (80 patients in each arm) into the effect of best medical therapy published to date and demonstrated over an average 7.8 years significant improvements in HbA1c amongst those having intensive medical therapy from 8.4+/-1.6 to 7.7+/-1.2, but no change in HbA1c amongst those continuing with standard medical therapy. This study defines the very best that could realistically be achieved in the control arm, but expect there to be very little if any change in this group. The reporting of HbA1c as an outcome measure was not in accordance with the newly defined IDF criteria, but considering the small average reduction achieved in the Steno study, it will be assumed that a target of 15% of patients reaching the endpoint is a conservative estimate. Company data on the small number of patients who have reached a year with the device in place suggest that 40% will achieve this target.
According to our own experience with the device in the commercially sponsored study, up to 30% of patients in the treatment group may have the device removed early. Nevertheless other commercially sponsored (unpublished) studies of this device have achieved lower explant rates (J Tetreault -GI Dynamics). We have therefore diluted the treatment effect from 40% vs. 15% to 35% vs. 15% achieving the target of 20% reduction in HbA1c for treatment arm vs. standard arm. 73 patients per group will give 80% power to detect a significant effect. Adding 10% loss of followup increases the sample size to 80 per group.
The dilution was calculated starting from the assumption that 40% of patients with the device will reach the target (this estimate is based on company data based on diabetic patients in the same range of BMI as in the present proposal). If 30% of patients in the treatment group need to remove the device early but remain available for follow-up, in the worst case scenario, the proportion reaching the target is the same as in the control group, bringing the estimate for the treatment group to 32.5%. However most of them will keep the device for some time, having some benefit, so it is plausible to assume that the estimate is higher than 32.5%.
Dividing the main effect 15% vs. 40% in three parts we assume that in the 30% of patients with removal, for 1/3 the same effect will be achieved as in the control group (15% reaching the target), for 1/3 it will be increased (23% reach the target) and for 1/3 more increased (31% reach the target). Overall, this would give an estimate of 35% for the treatment group.
For that reason, 2 arms of 80 patients will be sufficient to ensure demonstration of a significant effect (if one exists) and very conservatively allows for explant rates of up to 30%, a higher, level of benefit in the control arm than is likely to be achieved, and drop-out rates of 10%.
Furthermore the landmark Steno study which in some ways may be considered similar to this study, that had in all likelihood a less effective intervention arm was sufficiently powered with 80 patients in each arm.

Legend:
A -Weight, waist, blood pressure, routine bloods, adverse events, changes in medication/medical history

Randomisation
After the screening visit, all eligible patients for the trial will be randomised into one of the two arms of the study via the InForm system (the eCRF database for the study) which will be programmed with a randomisation schedule provided by an independent statistician. This will protect against bias in the randomisation process as patients are allocated automatically. The randomisation will be stratified by site and two BMI groups, 30-40 and 40-50 kg/m2. The subjects will be informed about their allocated treatment arm on visit 2.
Only the subject number and subject initials will be recorded in the case report form, and if the subjects name appears on any other document (e.g. pathologist report) it will be completely anonymised. The investigator will maintain a personal subject identification list (subject numbers with the corresponding subject names) to enable records to be identified.

Analysis Set
All summary tables and listings produced for the trial report will show the allocated randomisation group and will be performed according to the intention to treat principle.
All randomised subjects will be considered part of the analysis population. A sensitivity analysis will also be performed excluding patients that have withdrawn from the study during the run-in phase between randomisation (Visit 2) and start of treatment (Visit 4). Any subjects withdrawn during this period will still be required to attend the end-of-study follow-up visit.

6.
Variables of Analysis

Primary Endpoint Variable
In order to satisfy our primary objective we will be looking for a reduction in glycosylated haemoglobin levels (HbA1c) at 12 months from start of treatment by 20%.

Efficacy Variables
In order to compare the EndoBarrier with conventional medical therapy, diet and exercise for obesity related type II diabetes we will be studying the following at 12 months from start of treatment:  HbA1c levels; values of less than 6% (or 42 mmol/mol)  Blood pressure; values below 135/85 mmHg  Weight; to investigate an absolute weight loss greater than 15%, body fat content.

Mechanistic Sub-groups
In order to investigate the mechanism of the effect of the EndoBarrier we will be analysing the following:  Bio-electrical Impedance: %age body fat, total fat and fat-free mass at Visits 3, 5, 8, 10 & 14.
In addition to the above, the following sub-teams will be investigating additional variables: A detailed breakdown of the above can be found in Appendix A

Metabonomic Analysis
Urine, Plasma and Stool samples will be taken at Visits 3,5,8,10 and analysed for metabolite levels.

Cost Effectiveness
To estimate the cost-effectiveness of the EndoBarrier device over the trail period and longterm we will be studying the following: 1. Cost of interventions and related health and social care 2. QALYs accrued (calculated from area under the EQ-5D curve) 3. Incremental cost per QALY within the trial period and over the long-term Final analysis of health economics will be performed by a health economist and will be defined in a document external to this SAP.

Safety Variables
To evaluate the safety of the EndoBarrier we will be investigating the following:

Exploratory Efficacy Analysis
In addition to the secondary efficacy analysis we will also investigate the effect of the EndoBarrier device over multiple time points. Analysis will be run in comparison with conventional medical therapy, diet and exercise for the following: * Blood Glucose will be dichotomised into low/high groups based on the cut-off point of 82 mg/dl

Baseline Demographics
Patient characteristics will be summarized. Summaries of continuous variables will be presented as means and standard deviations if normally distributed, and as medians and inter-quartile ranges for skewed data, whilst categorical variables will be presented as frequencies and percentages.

Primary Analysis
The difference between the two study groups in the proportion of patients achieving substantial improvement in the metabolic syndrome at 12 months will be analysed using logistic regression adjusting for the stratification variables (BMI groups and sites).
All statistical tests will be two-tailed with a 5% significance level.
A sensitivity analysis to take missing data due to subject withdrawal into account will be carried out. Details will be provided in Section 7.6.

Secondary Analysis
Analysis of secondary outcomes will be conducted using standard statistical procedures applicable to categorical or continuous data as deemed appropriate within the following sections:

Secondary Efficacy Analysis:
All statistical tests will be carried out as described below; two-tailed (where appropriate) and with a 5% significance level.
The difference between the two study groups in achieving the following secondary outcome variables will be analysed using logistic regression adjusting for the stratification variables (BMI groups and sites). Tables detailing summary statistics of the above, including change from baseline, will also be provided.

Mechanistic Sub-Group Analysis
Mechanistic Analytes defined in section 6.2.2 and Appendix A will be assessed using a mixed model: Yijk = µ + πj + τi + CV1 +…+ CVr + (πj* τi) + bj(k) + eijk In this model: • µ is the intercept of the model; • τi is the ith fixed treatment, i = 0 (standard therapy) or = 1 (EndoBarrier); • πj is the fixed visit effect at j months where j = 1,...,12 ;  CV1 +…+ CVr is the fixed effect of Covariates* 1 to r • bj(k) is the random visit effect at the jth visit month of the kth subject; • eijk is the random error associated with the kth subject receiving treatment i at visit j ; • bj(k) and eijk are independent for i = 2,1 , j =1,...,5 , and k =1,..., n .

*Covariates will include: Age, Gender
Where repeated measure variables are nested (for example, where multiple blood samples are taken per visit), an additional fixed effect (and subsequent interaction terms) will be included to account for the nested variable. Due to the small size of the sub-groups, an additional random effect will not be included in order to keep the nested model as simple as possible. An unstructured covariate structure will be used. In the even where convergence has not been met, an alternative structure will be used as appropriate.
Analysis will be presented in the form of test results of fixed effects and estimates of model parameters. Post-hoc testing via least square means may also be performed on any model parameters with a p-value of p<0.05.
Additional testing will be performed within the following mechanistic sub-group as detailed below:  Food preference: Regressions will be performed with clinical outcomes (i.e. BMI, glucose control) to identify predictive markers and generate mechanistic hypotheses.

 fMRI Tests:
fMRI studies analysis will use region of interest analyses (e.g. for food pictures orbitofrontal cortex (OFC), amygdala, caudate, nucleus accumbens and anterior insula; for MID task nucleus accumbens, dorsal striatum and OFC; for Go-NoGo task pre-supplementary motor area). ANCOVA may be used in place of the mixed model, adjusting for the relevant covariates as data is only collected at one timepoint post baseline.

 Insulin Clamps:
Overall and tissue specific insulin sensitivity will be quantified for each patient and compared using the mixed model defined above. Regressions may be performed with clinical outcomes (i.e. BMI, glucose control) to identify predictive markers and generate mechanistic hypotheses.

Metabonomic Analysis
Metabolic datasets will be analysed using principal component analysis (PCA) and orthogonal partial least-squares analysis (O-PLS). The metabolic and microbial data will also be analysed in relation to response measurements such as BMI, gut hormone levels and etc. using O-PLS regression analysis and Bayesian approaches. A range of statistical methods will be optimised and applied to the data to identify weight loss and T2DM-associated microbiota and metabolites.

Exploratory Efficacy Analysis
The absolute change from baseline over time for variables listed in Section 6.2.1.2 will be will be analysed using a linear mixed model based upon a repeated-measures design.

*Baseline covariates will include: Age, Gender, BMI Group
A plot showing mean change (± 95% CI) over time will also be produced for all continuous variables.
Analysis will be presented in the form of test results of fixed effects and estimates of model parameters. Post-hoc testing via least square means may also be performed on any model parameters with a p-value of p<0.05.

Cost Effectiveness
A health economist is to analyse selected data to determine the overall economic benefit of the EndoBarrier device. Details of this analysis will be covered in a separate document, external from the SAP.

Safety Analysis
Adverse events will be summarised by treatment group with a separate summary table showing severity and relation to study. Serious adverse events will be listed and summarised by site and category with a separate summary table showing severity and relation to study.
Blood and urine analyte levels will be summarised by absolute value by change from baseline. Any clinically significant values or changes will be reported and investigated using the mixed model and least square means testing as presented in section 7.3.4.

Interim Analysis
Interim reports detailing baseline demographics, safety endpoints and primary endpoints (if requested) will be produced for assessment and discussion at DMEC meetings, the framework of which will be provided within the DMEC charter.

Missing Data and Withdrawn Subjects
The primary analysis is likelihood based and will naturally handle missing outcome data (substantial improvement in the metabolic syndrome at 12 months) under the assumption that data are Missing-At-Random (MAR), conditional on treatment group and stratification factors (BMI groups and sites). Based on the pattern and extent of missingness, if appropriate we will explore the effect of alternative missing data assumptions on the primary analysis within sensitivity analysis using multiple imputation by chained equations (MICE). Where applicable, we will assume that the probability of missing data is not dependant on the values of the unobserved data and that the data is Missing-At-Random (MAR), conditional on treatment group and stratification factors (BMI groups and sites) as well as on HBa1c values at timepoints M3, M6, M9 and M12. A total of 20 Imputed data sets will be drawn separately for each randomised group, replace missing outcome values with simulated values from a set of imputation models containing BMI group, sites, HBa1c values at M3, M6, M9 and M12. Using MICE, missing values for the binary outcome will be imputed using a binary logistic model, including all other covariates. Missing values for any of the continuous interim HBa1c included in the imputation model will be imputed using linear regression models.
Parameter estimates across the iterations will be combined using Rubin's rules and subsequently analysed using the same methodology as the primary analysis described in Section 7.2.
Alongside MICE, we will also examine the difference in proportion of substantial improvement amongst those missing and those observed to obtain an alternative result from that concluded from the complete case analysis difference between the two arms. Four scenarios are to be considered: 1. Missing participants within the EndoBarrier arm to have an increased rate of substantial improvement. Missing participants within the standardised treatment arm to have the same rate of substantial improvement. 2. Missing participants within the standardised treatment arm to have an increased rate of substantial improvement. Missing participants within the EndoBarrier arm to have the same rate of substantial improvement. 3. Missing participants within the EndoBarrier arm to have a lower rate of substantial improvement. Missing participants within the standardised treatment arm to have the same rate of substantial improvement. 4. Missing participants within the standardised treatment arm to have a lower rate of substantial improvement. Missing participants within the EndoBarrier arm to have the same rate of substantial improvement.

Fisher's Exact Test
Diffusion tensor imaging White matter tract fractional anisotropy, mean diffusivity fMRI 3,12 (for all of above applies to both whole brain and region of interest analyses) fMRI 3,13