Clinical endocrinology can be defined in one sentence. It is the study of diseases that are characterized by too much or too little of a hormone or by over or under responsiveness to a hormone. That’s it. The rest is detail.

Type 1 diabetes is due to the lack of insulin secondary to autoimmune impairment of insulin production by the beta cells of the pancreas. Glucagon, made by the alpha cells, is stimulated by the lack of insulin in patients with type 1 DM. It is this combination that is responsible for the ketoacidosis characteristic of this disease. Type 2 DM is due to insulin resistance. Thus, insulin levels are high and glucagon is suppressed. This combination explains why ketoacidosis is not typically a feature of the disorder.

Another form is diabetes is that secondary to destruction of the pancreas. In his form of the disease both insulin and glucagon levels are low, thus hyperglycemia without ketoacidosis is the essential characteristic of the disease. Like type 1 DM, this disorder requires insulin therapy. Type 2 DM can often be treated without giving insulin.

Type 1 DM requires frequent injections of insulin and causes the afflicted patient to navigate between the Scylla of hypoglycemia and Charybdis of hyperglycemia. This never ending voyage often falls into the jaws of one or the other monster, typically both. For years the goal of optimum treatment has been the construction of an efficient artificial pancreas (AP). While Medtronic’s markets a device it calls an artificial pancreas, the instrument lacks many of the features need by an ideal AP. This machine monitors basal glucose levels and delivers insulin as needed. It requires the patient to input data about carbohydrate intake at meal times. While a big step towards a completely independent artificial pancreas, there is still room for significant improvement.

The ideal artificial pancreas would be completely self contained, like an artificial pacemaker. There would be no need for an external insulin pump as it would be implantable and part of the AP. A subcutaneous well would hold a day or more supply of insulin which could be refilled by a syringe. A glucose sensor would continuously monitor glucose levels and deliver insulin as needed on a continuous basis. It would be equally effective at both basal conditions and at mealtimes without any need for input from the patient. It would store both glucose blood concentrations and periodic measurements of HbA1C. These data would be uploaded daily to both the physician and patient. Errors in dosage would be self corrected, but could also be manually fixed. If the device worked as envisaged, the patient would have little more to do besides refilling the insulin reservoir.

Current technology seems sufficiently advanced for the construction of such a device in the immediate future. In fact, I’m surprised that it hasn’t already been done. Given that there are about 1.25 million type 1 diabetic patients in the US alone, there is obviously both a medical and financial incentive for the construction of such a machine. Obviously, an ideal AP would decrease the morbidity associated with this disease and markedly improve the quality of life of these patients. After a substantial initial investment the cost of caring for type 1 diabetic patients should likely fall.

The current advanced knowledge in artificial intelligence, miniaturization, and computer science are such that scientists and technologists who apply themselves to this project, ie an AP, should soon produce something close to the AP outlined above. There are a number of companies and labs that are actively pursuing this problem Even a pessimist would allow that a workable solution that would greatly ease the burden carried by patients afflicted with type 1 DM should be close.