Post-prandial hyperglycemia occurs despite meticulous carbohydrate counting and rapid acting insulin therapy. Furthermore, this occurs even in the setting of the closed loop system. Currently the algorithm used for calculating the glucose-responsive insulin delivery cannot respond in a timely fashion to the glucose absorption resulting from a meal. In diabetes, there is paradoxical immediate post-prandial hyperglucagonemia that results in immediate post-prandial hyperglycemia. Amylin deficiency and/or dysregulated GLP-1 seems to be the etiology. Pharmacologic replacement of these hormones alleviates immediate post-prandial hyperglycemia in diabetes. With this protocol, the investigators would like to optimize treatment of T1DM by physiologic replacement of hormones in addition to insulin and in the process also optimize the insulin algorithm.

This is a paired, randomized, and controlled comparison of pramlintide and insulin versus exenatide and insulin Vs insulin monotherapy using the ePID closed-loop system for insulin delivery.

The investigators will stratify the study subjects into the following sub-groups of 5 subjects of 22-30 years old, 4 subjects of 18-21 years old, 4 subjects of 16-18 years old. The investigators would also begin the study with the 21-25 year patient sub-group and then transition to the other sub-groups after evaluating all the safety issues. 22-30 year old ones would be considered as an adult subset, 18-21 year olds would be considered pediatric subset according to the guidelines of FDA's Center for Devices and Radiological Health (CDRH) and 16-18 year olds are considered typical pediatric population. At this time, Spanish-speaking subjects will not be recruited because Medtronic Minimed as yet does not have any literature in Spanish that may used in explaining the study to this group of patients. When in the future Medtronic is able to provide us with the appropriate Spanish literature, the investigators will at that time amend the protocol to include this group of subjects.

Diabetes complications can be minimized or forestalled with the use of intensive insulin management. In the post-Diabetes Control and Complications Trial (DCCT) era, diabetologists around the globe are engaged in developing better methods of improving glycemic control in patients with type 1 diabetes mellitus (T1DM). To achieve the DCCT-recommended glycemic and hemoglobin A1C (HbA1c) goals, in addition to physicians, engineering and pharmaceutical companies are augmenting our ability to improve control by innovative compounds and technologies such as continuous glucose monitoring. Despite these substantial advances, hyper- and hypoglycemia continue to be problematic in the management of T1DM, especially in children.

Current T1DM management involves checking blood glucose pre-prandially and once at bedtime (4 times/day), and making insulin dose adjustments based on these 4 blood glucose measurements. This approach is woefully inadequate. Post-meal hyperglycemia is missed, and low blood glucose is undetected until the patient has symptomatic hypoglycemia. Moreover, we depend on patients to make appropriate dose modifications to basal rates of insulin when on the pump, without the guidance of adequate blood glucose measurements. The continuous glucose monitoring system (CGMS) is increasingly studied as a tool to examine glucose trends, and is helping to combat inadequate self-monitoring. However, the CGMS still relies on the patient's ability to make accurate insulin dose adjustments based on glucose readings. To address this issue the closed loop system is now under investigation.

The closed loop system is developed to closely emulate physiologic insulin delivery. The algorithm was developed by modeling beta cells and validated by hyperglycemic clamp studies using the subcutaneous site for measurement of blood glucose and insulin delivery. The algorithm is known as the external Physiologic Insulin Delivery (ePID). Current studies suggest that basal requirements are detected accurately, and the algorithm responds effectively. However, meal related elevation in blood glucose remains a challenge, as when blood glucose levels begin to rise and are seen in the subcutaneous glucose sensor signal, it is already too late to administer insulin in a way that will effectively minimize the postprandial glucose peak. Improvements in the ePID algorithm have been studied to improve the postprandial response with some success, but require user intervention for best results, and despite that glucose excursions fail to normalize in the post-prandial period. It is likely that further improvements in postprandial glycemia will be limited with insulin monotherapy. Besides insulin there are other dysregulated hormones such as glucagon and amylin, which contribute to post-prandial hyperglycemia.

The autoimmune destruction of insulin-secreting beta cells of the pancreas is presumed to cause T1DM. In addition to insulin, other hormones such as glucagon and amylin play a role in normalizing glucose excursions. Failure of glucagon suppression results in immediate postprandial hyperglycemia, and a loss of glucagon response to hypoglycemia results in late postprandial hypoglycemia. Despite the increasing use of subcutaneous continuous insulin administration and newer insulin analogs, insulin replacement remains imperfect, and glucose excursions are inadequately controlled in diabetes. This suggests that other factors, in addition to diet and insulin management, may need to be addressed if postprandial glucose excursions are to be normalized in T1DM.

The recent discovery of the hormone amylin has enhanced our understanding of postprandial glucose homeostasis. Patients with T1DM have deficiencies of both insulin and amylin. Amylin, a 37-amino acid polypeptide hormone, is co-secreted from the pancreatic beta cells in conjunction with insulin in response to nutrient stimuli. Amylin, in the immediate postprandial period, may mediate part of its effect by suppressing glucagon secretion, resulting in the suppression of hepatic glucose production and the slowing of gastric emptying.

Pramlintide acetate is a synthetic analog of the naturally occurring human hormone amylin. It effectively reproduces amylin agonist activity in an equipotent fashion. Pramlintide acetate has been reported to improve glycemic control in adults with both type 1 and type 2 diabetes mellitus. Specifically, postprandial glucose excursions are improved with adjunctive pramlintide use compared with insulin monotherapy.

Exenatide is a synthetic analog of exendin-4. It is FDA approved for use in adults with type 2 diabetes mellitus (T2DM). Clinical studies indicate that exenatide is very effective in decreasing postprandial glucose excursions. The investigator's brochure for exenatide suggested that in adults with T2DM, exenatide was well tolerated at 0.1 mcg/ kg/ dose, and resulted in glucose lowering. Our preliminary data in children with T1DM suggests that 1.25 mcg and 2.5 mcg are good pre-prandial initial doses with a prandial insulin dose reduction of 30%. The principal mechanism of action in T1DM is suppression of glucagon, with a subsequent suppression of hepatic glucose production and decreased immediate post-prandial hyperglycemia. These effects result in further reduction of HbA1c beyond the reduction observed with insulin alone. Additionally, in T2DM, exenatide delays gastric emptying, and decreases food intake through centrally mediated mechanisms causing satiety and improved weight control. The most notable adverse effect of exenatide is nausea, which improves with the duration of treatment and steady dose escalation of exenatide. Hypoglycemia may occur with concurrent use of insulin.

In the current exploratory studies we are proposing a novel way to combat immediate postprandial hyperglycemia in subjects with T1DM by using the study medications through the experimental closed loop system. Subjects will not be on the closed loop system prior to eligibility, and would not have used the closed loop system at an earlier time. We will use prandial pramlintide/exenatide in the first protocol, and test the hypothesis that this intervention will be superior to using insulin monotherapy in the closed loop system. This would be a Biologic, phase I, single center study and a paired, randomized, and controlled comparison of pramlintide and insulin versus exenatide and insulin Vs insulin monotherapy using the ePID closed-loop system for insulin delivery.

We will stratify the study subjects into the following sub-groups of 5 subjects of 22-30 years old, 4 subjects of 18-21 years old, 4 subjects of 16-18 years old. We would also begin the study with the 22-30 year patient sub-group and then transition to the other sub-groups after evaluating all the safety issues. 22-30 year old ones would be considered as an adult subset, 18-21 year olds would be considered pediatric subset according to the guidelines of FDA's Center for Devices and Radiological Health (CDRH) and 16-18 year olds are considered typical pediatric population. At this time, Spanish-speaking subjects will not be recruited because Medtronic Minimed as yet does not have any literature in Spanish that may be used in explaining the study to this group of patients. When in the future Medtronic is able to provide us with the appropriate Spanish literature, we will at that time amend the protocol to include this group of subjects.

Study Procedure:

Screening: Screening evaluations will be performed 1-2 months prior to study enrollment, and will consist of an informed consent, medical history, physical examination (including height, weight, and vital signs), and blood samples for clinical laboratory tests. The clinical laboratory tests will include a CBC to check for anemia, hemoglobin A1C (within last month), serum electrolytes, serum amylase, Serum HCG pregnancy test for female subjects. The approximate volume of blood is expected to be less than 5 ml.

Following informed consent (and with appropriate subject assent) if subject is found eligible for the study, they will undergo a baseline open-loop evaluation using a Medtronic Data logger continuous glucose monitor (Medtronic Diabetes, Northridge, CA). The data on glucose excursions will be collected on their usual insulin dosages so as to appropriately configure the algorithm of the closed-loop system during the inpatient admission. Basal rates and insulin to carbohydrate ratios will be adjusted during the period between screening and the start of the 3 studies that they will undergo subsequently.

After screening, subjects will undergo a 3-period crossover design trial in random order.

Study A Subjects will be admitted to the general clinical research center (GCRC) on the evening prior to the study at 6 PM. For female subjects a stat serum HCG pregnancy test is done prior to the start of the study. Two sensors will be placed in the subcutaneous tissue. Insertion of IV line will occur between 9-10 PM. During the night hourly samples of blood glucose will be used to adjust insulin basal or boluses so as to maintain euglycemia. Around 6 a.m. subjects will be started on the closed-loop system for the automated delivery of insulin based on the glucose sensor measurements. An intravenous line will be inserted in the antecubital fossa to maintain blood glucose concentrations within the normal range.

Blood samples for basal glucose, insulin, and glucagon concentrations will be drawn at minus 30, minus 10, and 0 min prior to study start.

For study A (Control): Subjects will receive meals at 0 minutes (approximately 7 AM breakfast), 300 min (approximately Noon lunch) and dinner at 600 min (approximately 5 PM dinner). A partial bolus of insulin will be administered prior to breakfast in addition to closed loop system (since the system is not expected to have calibrated by the time of breakfast).

For study B: Subjects will receive 30 mcg of pramlintide just before 300 and 600 minutes. Pramlintide will be administered as a subcutaneous injection just prior to lunch and dinner. A partial bolus of insulin will be administered for breakfast in addition to closed loop system (since the system is not expected to have calibrated by the time of breakfast).

For study C: Subjects will receive 2.5 mcg of exenatide subcutaneously just before 300 and 600 min. Exenatide will be administered as a subcutaneous injection just prior to lunch and dinner. A partial bolus of insulin will be administered for breakfast in addition to closed loop system (since the system is not expected to have calibrated by the time of breakfast).

At 0 mins, subjects will drink a standard liquid meal of Boost High Protein Drink 9.6 oz (360 calories, 40 g of carbohydrate) over a period of 10 min. This meal will not be counted in the final analysis since the closed loop system will not have had enough time to adjust to the basal conditions of the meal. A partial bolus of insulin will be administered for breakfast in addition to closed loop system (since the system is not expected to have calibrated by the time of breakfast).

At 300 minutes, subjects will receive a solid meal based on Estimated Energy Requirements (EER) as reported by the Institute of Medicine Dietary Reference Intakes macronutrients report, 2002. The carbohydrate content of the meal will be 75 g. Pramlintide 30 mcg/ Exenatide 2.5 mcg will be administered subcutaneously before lunch and dinner. Insulin will be administered based on ePID algorithm by the closed loop system in response to glucose concentrations detected by subcutaneous sensing. Blood samples for glucose concentrations will be drawn every 15min-half hour (more frequently immediately after meals) after initiating closed loop system. Blood samples for hormone analysis (insulin, pramlintide/exenatide and glucagon) will be drawn at approximately every 15 min to half hour intervals after lunch which will be served at approximately around Noon in the first hour and then every half to one hour until dinner. Dinner will be served at approximately around 5 PM (the carbohydrate content of the meal will be similar to lunch). Blood glucose will be measured at the bedside using a YSI glucose analyzer every half hour. The study will end at 885 minutes.

Total blood volume drawn for studies A, B, and C will be approximately 97cc, 133cc and 133cc respectively.

The subject's will be always monitored by nursing staff of the CRC, study staff and the PI is available by phone and on call during the study. Medtronic personnel will be present for the first four studies to ensure all the equipment is working optimally and to provide guidance related to the closed loop system.

During the study, if subject's blood glucose values are less than 70 mg/dl, oral glucose (5-15 g) will be administered for countering low blood glucose to achieve euglycemia (90-130 mg/dl). 1-2 doses of oral glucose should correct hypoglycemia. If more than 4 consecutive doses are required to achieve euglycemia the study will be terminated, and the subject will be offered a meal tray and blood sugar rechecked to ensure euglycemia. Additional 0.2 cc blood samples for glucose may be necessary to ensure safety during an episode of hypoglycemia and these may be done before the scheduled time for blood draws.

If blood glucose by Analox machine is greater than 300mg/dl, blood ketones will be measured and if ketones are above 1.5 mmol/L we will terminate the study.

At the end of the study the subject will be evaluated by the research staff for the need to receive a snack. A snack will be served before discharge only if BG of subject is less than 70mg/ dl at the end of the study. If the Principal Investigator or the attending physician feels that the subject needs to be observed for a longer period of time then it would be at his/her discretion.

The subjects will return two additional times (2-4 weeks apart) to complete the exact same study, with the difference that if the patient received pramlintide on the last occasion he/she will get either exenatide or just insulin via the closed loop system at the subsequent visits. (Total of 3 visits).

Female subjects would have a stat serum HCG pregnancy test is done prior to the start of the study for each visit. Subjects will have stat CBC prior to initiating closed loop study. Serum Electrolytes will be done at the onset of the study and then will be repeated if the patient had vomiting at the end of the study. Serum amylase would be done at each visit and the results would be tracked on the case report forms. Capillary blood glucose check would be done when necessary as an extra safety measure in addition to subject's blood glucose being monitored through YSI and interstitial glucose monitored through glucose sensor.

For uniformity sake, we would be using only using Insulin Aspart for the entire duration of the study.

If at least four subjects develop grade 3 nausea, with the 30 mcg dosage of pramlintide or with the 2.5 mcg of exenatide then the protocol would be modified so that the subjects receive only a 15 mcg dose of pramlintide and 1.25 mcg of exenatide.

TREATMENT FOR SEVERE NAUSEA: If severe nausea develops during the study, Zofran 0.15 mg/kg IV will be given. The maximum dose is 8 mg. If the subject still has emesis, the study subject would be stopped.

Study drugs and study laboratory tests will be provided at no cost to the subject. The subjects will receive $70 at screening and $210 for the three subsequent study visits. The total compensation involved for the study would be $ 700.

Inclusion Criteria:

1. Age greater than 18 years and less than 30 years.
2. Have had diabetes for at least 1 year, and in good control (HbA1C less than 8.5 %).
3. Be on continuous subcutaneous insulin infusion using an insulin pump
4. Subjects must be otherwise healthy except for T1DM, and treated for hypothyroidism if present
5. Menstruating women must have negative pregnancy test.
6. Hemoglobin (Hb) more than 12 g/dl
7. Weight must be equal to or greater than 50 Kg

Exclusion Criteria:

1. Any chronic disease (leukemia, asthma, inflammatory bowel disease, cystic fibrosis, juvenile rheumatoid arthritis, etc that directly, or as a result of treatment, directly or indirectly affect glucose homeostasis).
2. Hemoglobin less than 12 g/dl
3. Lack of a supportive family environment
4. Positive pregnancy test based on serum beta HCG in menstruating young women
5. Evidence or history of chemical abuse
6. HbA1c more than 8.5 %
7. Weight less than 50 Kg
8. History of gastroparesis and on medications that alter gastric emptying
9. History of Pancreatitis and impaired renal function
10. Hypoglycemic unawareness
11. History of sensitivity to 5-HT3 receptor antagonists
12. History of QT prolongation
13. Concomitant use of both Acetaminophen and vitamin C
14. Patients on glucocorticoid therapy
15. Known allergies to any of the study medication