Artificial Pancreas – Reality Or Fantasy

Written by Dr. Mathews J. Alapatt

In the year 1922, Leonard Thompson’s life was saved using painful intramuscular injections of crude insulin preparation. Fortunately, the treatment of insulin has moved on with better insulin and vastly improved insulin delivery systems. Our understanding of diabetes has also improved. Landmark clinical trials like the DCCT and UKPDS have demonstrated that good control of diabetes with intensive insulin treatment results in a significant decrease in complications related to diabetes. Interestingly the benefits of this period of intensive control last for another 5 to 10 years. Preventing the complications of diabetes is important as it heralds an increase in mortality and morbidity and, as a corollary, an increase in health care costs.

In current clinical practice, intensive insulin treatment is administered using multiple daily doses of insulin or an insulin pump. This then has to be combined with self-monitoring of blood sugars or continuous glucose monitoring. Alongside this, the patient needs to pay constant attention to parameters that affect blood glucose, such as food intake and exercise, and match insulin doses to this. All this requires a high degree of motivation, commitment, and competence from the patient and caregivers to ensure treatment is safe and effective. Insulin pumps resolve some of these issues; however, they still require regular calibration and vigilance. It is not surprising that many patients do not achieve optimal control. An Artificial Pancreas device that can check blood sugars and deliver insulin without human intervention is the ultimate solution to these problems. If perfected, such a device will liberate patients with diabetes needing insulin.

There are two broad approaches mechanical and biological. The biological or Bio Artificial Pancreas refers to the transplantation of encapsulated islet cells. The capsule protects them from immune attacks while allowing the diffusion of nutrients, oxygen, and insulin. The use of nature’s own sensor, pump, and algorithm is very attractive; however, the supply of islets is limited, and there are many technical issues to be overcome. It, therefore, remains a work in progress with very few clinical trials to date. The recent announcement by Harvard stem cell researchers that they have been able to produce human insulin-producing beta cells in large quantities from human embryonic stem cells may revitalize research in this field.

The Mechanical Artificial Pancreas consists of three components 1) a glucose sensing device or continuous glucose monitor, 2) a computational system or controller to calculate the dose of insulin to be delivered 3) an insulin delivery system or pump. Continuous glucose monitors have been around for some time and have, over time, become smaller, more accurate, and provide data in real-time. The majority of the ones in use currently are minimally invasive. The major problem is that there is a lag of up to 20 minutes when compared to blood glucose, which is expensive and has to be changed every 3 to 7 days. There is a major research effort underway to develop an accurate noninvasive glucose monitor. Insulin pumps have been in use since the 1990s. Significant improvements in technology have resulted in smaller pumps and allow for the fine adjustments of the insulin dose. The key issue with the pumps is that they use the subcutaneous route, which results in delayed insulin action as the absorption takes time. The third component is the computer algorithm which calculates the dose of insulin. This algorithm needs to deal with the inaccuracies of the glucose sensing and delivery mechanisms. Newer algorithms called the Model Predictive controllers to tend to overcome some of these problems and can also ‘learn” the patient.

Considering the complex requirements of developing such a system, it will only be achieved in slow incremental steps. The initial steps have been taken with the development of the “Low glucose suspend systems,” which shut off if the blood glucose falls below preset levels. The next step would the control blood glucose overnight. Bihormonal pumps that can deliver two hormones would be helpful with the delivery of other hormones like Glucagon. Integrating the different components would be the next major hurdle. There are several initiatives around the world to hasten the development and optimize the research and development efforts. These include the ones by the Juvenile Diabetes Research Foundation in America and the AP@home by the EU in Europe.

Once the Artificial Pancreas has been engineered, it will face rigorous scrutiny from regulatory agencies concerned with efficacy and safety. Insurers and payers will be interested in evidence that the initial high cost will be offset by the health cost benefits later on. The final hurdle may be the legal system, which may be the arena for potentially devastating lawsuits against the manufacturer and caregivers.

The road to the development of the Artificial Pancreas may seem long and potholed. However, we live in times of rapid change in technology and communications. What seems inconceivable a few years ago becomes a reality in a few years’ time. So if I were to ask myself, what will the future Artificial Pancreas look like? The answer would be that it would be a combination of a discrete insulin pump, a wearable watch, and a smartphone. The watch will be a noninvasive glucose monitor and a display unit. The computational system would be housed in the smartphone. This, in turn, will be connected to the health provider to provide additional backup. When will all this reality is a difficult question to answer but what is certain is that the Artificial Pancreas is moving from the realms of fantasy to becoming a reality?