By Scott King
A friend who worked at the FDA had a Harris cartoon of two medical researchers, one saying to the other, “This drug will bestow immortality. Unfortunately it will take forever to prove it.”
In medical research we are expected to prove our claims. We believe the Islet Sheet will be a functional cure of autoimmune diabetes. How can we prove it?
Simple answer: treat at least 7 humans with autoimmune diabetes, then establish that the implanted Islet Sheets are secreting enough insulin to produce normal blood sugars. But before we can do that we need permission from UCI (Institutional Review Board) and the government (Food & Drug Administration). And they require we show the Islet Sheet is safe and effective in animals. But what animals?
Easy, you say. Find animals that have developed autoimmune diabetes and fix them. Easily said. Not so easily done.
Pure, uncomplicated autoimmune diabetes (type 1; also called juvenile or ketosis-prone diabetes) in humans is associated with zero insulin production, a state caused by autoimmune destruction of the cells that make insulin. Untreated autoimmune diabetics display unmistakable symptoms including excessive eating, excessive drinking and excessive urination. Diagnosis is confirmed when blood sugar is over 200 mg/dl while fasting. Undetectably low insulin levels are the final confirmation that the symptoms have no other cause.
So we need an animal that naturally develops type 1 diabetes, and find individuals with the human symptoms of the disease.
Dogs with autoimmune diabetes would be ideal. I was told years ago that such dogs exist, are sometimes treated with insulin, and can be found through veterinary schools. Years ago, when the time came for canine studies of microencapsulated islets with diabetic dogs, I was told by the UC Davis veterinary school that autoimmune diabetes does not actually exist in dogs. The dogs receiving insulin injections are invariably type 2, or insulin resistance diabetic dogs. (I have visited several canine diabetes web sites that erroneously claim dogs get type 1 diabetes.)
For studies on diabetes transplantation generally, and islet transplantation in particular, the only sure way to make a dog diabetic is to do a total pancreatectomy, that is, surgical removal of the entire pancreas. In fact, this is the oldest model for juvenile diabetes because the role of the pancreas in preventing diabetes was discovered in 1889 using the pancreatectomized dog.
Another common model is the NOD mouse, usually described as an autoimmune IDDM mouse. This is not accurate; these mice suffer from a variety of autoimmune diseases. The genetics are very different from human diabetes. In humans particular genes are only weekly predictive of autoimmune disease; it is predominantly environmental. In NOD mice the disease is partly environmental, but only appears in mice with a certain genetic makeup. That is to say, it is a genetic disease. Thus, in my opinion, the NOD mouse is more of a genetic freak than a model for type 1 diabetes. I don’t think it is a good thing that much of what we know about autoimmune diabetes comes from the study of the NOD mouse. There are literally scores of interventions (over 150 at last count) that prevent or cure diabetes in NOD mice. None of these have been shown to work in humans.
In the case of the Islet Sheet we are limited to animals large enough to implant the device. We cannot make a sheet small enough for a mouse, so for the moment we study rats. As you know from my last report, it is difficult to induce type 1 diabetes in a rat using drugs. We are persevering and beginning to see good data.
There is a spontaneously diabetic rat, the BB rat. It has many points of similarity with the human disease. We may use this model once we have thoroughly investigated the streptozotocin diabetic rat.
I remember at the LCT science advisory board meeting three years ago Bob Elliott gave a presentation on animal results using LCT’s microencapsulated pig islets. He was asked what the best model is for encapsulated islet transplantation studies. He thought for a moment, and said “there are no good animal models” for autoimmune diabetes. I agree. The best model for transplantation studies remains a pancreatectomized large mammal. All animal models are approximations. All the “autoimmune” models differ significantly from human autoimmune diabetes. But at last metabolically pancretectomized large mammals are very similar to humans with autoimmune diabetes.
LCT was able to get permission to do clinical studies with the limited models available. We will do the same.