Type 1 diabetes (T1D) is one of the most common chronic childhood diseases requiring lifelong insulin therapy. Children and adolescents with T1D need regular insulin injections or the continuous insulin delivery using an insulin pump in order to keep blood glucose levels normal. We know that keeping blood sugars in the normal range will help prevent longterm diabetes-related complications involving the eyes, kidneys and heart. However, achieving treatment goals can be very difficult as the tighter we try to control blood glucose levels, the greater the risk to develop symptoms and signs of low glucose levels (hypoglycaemia). This is a particular problem at night and one solution is to develop a system whereby the amount of insulin injected is controlled by a computer and is very closely matched to the blood sugar levels on a continuous basis. This can be achieved by what is known as a "closed-loop system" where a small glucose sensor placed under the skin communicates with a computer containing an algorithm that drives an insulin pump. We have been testing such a system in Cambridge over the last five years in children and have found that this system is effective at maintaining tight glucose control and preventing nocturnal hypoglycaemia. More recently the system has been tested in real life conditions in the home setting for three weeks during a pilot single-centre study.

The next step is to extend the evaluation of closed-loop over a prolonged period of three months. In the present study we are planning to study 24 young people aged 6-18 years on insulin pump therapy. During three months glucose will be controlled by the computer and during the other three months the subjects will make their own adjustments to the insulin therapy using real-time continuous glucose monitoring.

We aim to determine the effect of the computer algorithm in keeping glucose levels between 3.9 and 8 mmol/L (normal levels). Safety evaluation comprises assessment of the frequency of severe hypoglycaemic episodes. Participants' response to the use of the system in terms of lifestyle change, daily diabetes management and fear of hypoglycaemia will be assessed. We will also test for longer term glucose control by measuring glycated haemoglobin (HbA1c).

Purpose of the study

The purpose of this research is to evaluate the efficacy, safety and utility of the overnight automated closed-loop system in the home setting on consecutive nights over a 3-month period, as compared to the use of real-time continuous glucose monitoring (CGM) alone.

Design

The study is designed in the form of an open-label, cross-over study and it involves two 3-month-long home studies during which overnight glucose levels will be controlled either by a closed-loop (CL) system combined with a real-time continuous subcutaneous glucose monitoring (CGM) or by real-time continuous glucose monitoring alone, in random order.

The study will take place within the home setting where, initially, overnight supervision of the closed-loop system will be undertaken by a research nurse either at home or at an in-patient facility.

Three UK centres will be involved in the study: Addenbrooke's Hospital, Cambridge, University College London Hospital (UCLH), London, Leeds Teaching Hospitals, Leeds

Population

A total of 30 children and adolescents with T1D between 6 and 18 years of age on insulin pump therapy will be recruited from the Paediatric Diabetes Outpatient Clinics at the three investigation centres to allow for 24 evaluable subjects. Each investigation centre will recruit at least three subjects to achieve balanced representation of study population.

Potential participants will be identified by their treating clinicians and invited to contact the research team. They will be sent the study information leaflets and an invitation to join the study by the research team. They will have the opportunity to read and discuss the information. If they agree to participate in the study, informed consent and/or assent will be taken by a member of the research team who has a good working knowledge of the aims and practicalities of the study. The person taking consent will use language that is easy to understand and free from medical jargon.

Experimental Procedures

The study will consist of up to 10 visits, including two study periods (closed-loop vs. conventional pump therapy + continuous glucose monitoring). The study periods will last 3 months each. The order of the two interventions will be random.

Once the subjects have agreed to participate in the study, they will be invited for the recruitment visit, when participants' body weight and height, demographics, medical and diabetes history, and their previous days insulin therapy will be recorded. Urine pregnancy test will also be performed in females of children-bearing age.

Once consent has been obtained and if all the inclusion and exclusion criteria are met the subject will be included in the study and allocated a sequential subject number.

Blood samples will be taken for the baseline measurement of random C-peptide, glucose and HbA1c. Eligible subjects will also undergo a baseline evaluation where a blood sample for renal, liver, thyroid function, full blood count and anti-transglutaminase antibodies with IgA levels will be taken (if not done in the previous 3 months).

Validated questionnaires will be distributed to assess both the user friendliness of the devices used by study participants and fear of hypoglycaemia before entering the study. Study participants will also be invited to participate in a semi-structured qualitative interview conducted by trained staff, which will occur face-to-face or via telephone, within 2-4 weeks (during the training period), to determine their expectations of the diabetes treatments to be used in the study.

All eligible subjects who will have signed informed consent will be invited to a training session on the use of the devices which will be employed in the study.

• Two to four weeks prior to the first study night the subjects will be switched from their regular insulin pump to the study subcutaneous insulin infusion pump for the period of the study. Local diabetes clinic guidelines for switching insulin pumps will be followed. Subjects will be given training on the functionality of the pump by the research team and standard operational procedures, as well as the pump user manual, will be provided. The research and paediatric diabetes team will be available by telephone to assist with any concerns that the subjects may have regarding the pump.
• A training session on Continuous Glucose Monitoring (CGM) will take place about two to four weeks before the study starts. Subjects will be trained on the use of the study CGM system. Subjects will be trained at home/clinic by the research nurse to insert a subcutaneous glucose sensor and will be provided with the study CGM system to wear at home for at least three consecutive 5-day-sensor sessions (at least 12 days of use) or until subjects become confident with the device, namely with the performance of sensor insertion and calibrations. During the two to four week training period the CGM display may be masked so that subjects will not be able to see their glucose values. For subjects who use CGM as part of their normal diabetes care, the CGM display will not be masked during this training period. At the end of the training period the sensor data will be downloaded by the research team and will be used to optimise insulin therapy prior to the start of the study. This will be done in collaboration with the paediatrics diabetes team. The CGM readings recorded during this training period will used to assess baseline glucose control.

At the end of the training session, a competency assessment tool will be used to evaluate subjects' competency in the use of both devices. Subjects' compliance in diabetes management will also be evaluated in terms of time wearing the continuous glucose monitoring device. A minimum of 12 days of continuous glucose monitor readings should be recorded over the time of use of the device. Competency and compliance evaluations will determine participants' eligibility before randomisation.

Eligible subjects, who have completed the training session on the use of both the study subcutaneous insulin pump and the study CGM device and gained confidence in the use of both devices as assessed by the research team, will be randomised to one of the 2 intervention arms:

1. Real-time CGM + overnight CL
2. Real-time CGM alone Three to four weeks after the completion of each intervention arm, subjects will cross over to alternative intervention: those who were allocated to the CGM+CL intervention will be switched to CGM alone intervention, whereas those who completed the CGM alone intervention will be allocated to the CGM+CL intervention. During this three-to-four-weeks wash out period, real-time CGM may be used.

Additional training on the use of real-time CGM will be provided at the beginning of the first intervention arm. This involves training on how to interpret the data displayed on the real-time CGM device screen and how to perform insulin therapy adjustments according to glucose reports. Written guidelines on the use of both real-time and retrospective information will be provided to all subjects.

At the beginning of the closed-loop intervention, subjects will be provided with specific training on the use of the closed-loop system. Closed-loop training will be carried out in two consecutive stages:

1. Supervised Closed Loop: the closed-loop system will be started in the evening or at bedtime and will remain in operation until the following morning.

   Each subject will be supervised in starting the closed-loop system by the research nurse/doctor for one night at home or at an in-patient facility, with the option of a second supervised night at the start of the closed-loop intervention, until the subject is confident to use the equipment. For subjects who were supervised in using the closed-loop system at the in-patient facility, the research nurse/clinician may provide assistance and supervision in setting up the closed-loop system for the first night at home.

2. Unsupervised Closed Loop: Subjects will use the closed-loop system overnight at home for a total duration of 12 weeks. The subcutaneous CGM sensor will be replaced every 5 days by the subjects. Support and telephone advice will be provided by the research team for the duration of the study to deal with any concerns that may arise with the use of the closed-loop system components.

Cross-over procedures:

Having completed the first intervention arm, all subjects will cross over to the other intervention arm after a wash-out period of three to four weeks. Blood sampling for the determination of HbA1c will be taken to investigate any changes in metabolic control. Validated questionnaires will be distributed to evaluate the impact and satisfaction as well as the user friendliness of the devices employed in the completed arm of the study. A questionnaire evaluating fear of hypoglycaemia will also be distributed to both parents and children. Study participants will be invited to participate in a semi-structured qualitative interview conducted by trained staff within 0-4 weeks of completing the Closed Loop intervention arm, including their views on the diabetes treatment experienced. The second intervention treatment arm will start three to four weeks after the end of the first arm. In between the two intervention arms, subjects will resume their usual insulin regimen, using either the study insulin pump or usual patient's insulin pump. Real-time CGM may be used as part of their standard diabetes management.

End of study:

Having completed the second intervention arm of the study, subjects will revert to their conventional insulin pump therapy by switching to the insulin pump they were using before entering the study. Blood sampling for the determination of HbA1c will be taken to investigate any changes in metabolic control. Validated questionnaires will be distributed to evaluate the impact and satisfaction as well as the user friendliness of the devices employed in the completed arm of the study.

Inclusion Criteria:

• The subject is between 6 and 18 years of age (inclusive)
• The subject has type 1 diabetes, as defined by WHO for at least 1 year or is confirmed C-peptide negative
• The subject will have been an insulin pump user for at least 3 months, with good knowledge of insulin self-adjustment as judged by the investigator
• The subject is willing to perform regular finger-prick blood glucose monitoring, with at least 4 blood glucose measurements taken every day
• HbA1c ≤ 10 % based on analysis from central laboratory or equivalent
• The subject is literate in English

Exclusion Criteria:

• Non-type 1 diabetes mellitus including those secondary to chronic disease
• Untreated celiac disease
• Any other physical or psychological disease likely to interfere with the normal conduct of the study and interpretation of the study results as judged by the investigator
• Current treatment with drugs known to interfere with glucose metabolism, e.g. systemic corticosteroids, non-selective beta-blockers and MAO inhibitors etc.
• Known or suspected allergy against insulin
• Subjects with clinical significant nephropathy, neuropathy or proliferative retinopathy as judged by the investigator
• Total daily insulin dose ≤ 2 IU/kg/day
• Total daily insulin dose < 10 IU/day
• Pregnancy, planned pregnancy, or breast feeding
• Severe visual impairment
• Severe hearing impairment
• Subjects using implanted internal pace-maker