In this Section we start with the data that is probably of greatest interest, the ongoing large animal studies, or preclinical studies. The next section surveys data on sheets mostly from laboratory studies. Finally we describe our preliminary large animal studies that made the preclinical studies possible.
Large Animal Studies (preclinical studies)
When we set out in 1998 to show that the Islet Sheet could control blood sugar without immune suppression we knew we were up against terrible odds. So we published on this web site an analysis of the bioartificial pancreas and why we believed all prior efforts had failed to produce a clinically useful device. This analysis led to the conclusion that there were certain minimal criteria to make a working bioartificial pancreas. Of these, the most pernicious was biocompatibility: most islet implant devices provoke the host's foreign body response and the islets in the devices died of starvation following fibrosis. Researchers did not recognize the severity of the problem. Other criteria were equally as important but more straightforward in solution. For example, permitting sufficient nutrient diffusion was simply a matter of making the sheet very thin, something easily measured in the laboratory.
Starting in 1998, we did animal studies in dogs and rats that allowed us to prove that the Islet Sheet met the criteria we had established. These studies culminated in an effort to cure a pancratectomized dog by attaching six Islet Sheets to the dog's omentum, a flap of vascularized tissue found in the abdomen. We had chosen the omentum because of its accessibility and because of experiments showing that conventional islet transplants work when the islets are secured in an omental purse or "taco."
The initial results were encouraging: the dog's blood sugars were near normal despite the number of islets used being marginally sufficient. When we saw indications of trouble, we decided to remove the omentum and sheets to establish that the dog really was diabetic, and within a week of omentectomy blood sugars rose, establishing that the Islet Sheets were producing enough insulin to cure the diabetes.
But the condition of the sheets after ten weeks on the dog's omentum was discouraging. On the left below is a photograph taken as a sheet is being sutured onto the canine omentum. The right shows the omentum ten weeks later, with the Islet Sheets crumpled and cracked inside the fibrotic masses.
The masses were examined and found to contain functioning islets which was consistent with the dog's metabolism. They also contained fragments of sheets and the mesh used to reenforce the sheet. We believe the foreign body reaction visible in the right photograph is the result of the polyester mesh becoming exposed in cracks where the sheet folded. Sections of the sheets that remained flat did not provoke a fibrotic response.
The sheet was being attacked by the same systems that had caused the failure of almost all of the devices tested before the Islet Sheet. It was encouraging that islets performed well in the sheets - well enough to cure diabetes for weeks - but such a foreign body reaction would made the Islet Sheet clinically worthless.
This development caused us to step back and reexamine our reasons for believing that the Islet Sheet does not provoke such a foreign body reaction. As postulated in our criteria, the requirement is twofold: the material must be inherently nonfibrogenic and the surface must be very smooth. Most materials always provoke a fibrotic response. From work in collaboration with researchers at the University of California in early 1990's we had come to believe that highly purified alginate does not cause fibrosis. Subsequently other researchers have reported similar results, including recent work at the Joslin Clinic at Harvard University. Our sheet takes advantage of alginate's biocompatibility as well as its phase change from to hydrogel in the presence of calcium and back to liquid in the presence of citric acid. The exterior surface of the Islet Sheet is highly purified alginate.
The other concern was that the surface of the sheet be so smooth that there would not be any pits or cracks to attract fibroblasts. In rats cracked sheets would accumulate fibroblasts, and we observed beginnings of fibrosis in the cracks. So our methods were refined so that the sheet was smooth on a microscopic scale. Sheets placed in the highly fibrogenic rat subcutaneous space produced minimal fibrosis.
We had observed fibrosis in rats when the sheets crumpled. It appeared that the alginate was not flexible enough to survive the sheet being creased or folded. This suggested that, if the sheets remained substantially flat, they would not induce fibrosis. We decided to test that hypothesis experimentally.
At Edmonton in late October, 2000, four dogs were implanted with six sheets each, on the omentum, mesentery, pancreas, liver, diaphragm and under the skin. Normal, non-diabetic dogs received sheets which contained no islets. Two dogs were examined at 2 ¸ weeks and two at 9 weeks. The experiment was a success and showed what we hoped to find. In general if the sheet remained flat, it produced minimal or no foreign body reaction. All four sheets in the omentum produced masses like the one in the photograph above. Other sites that performed well were less flexible.
On the left is a sheet that has just been sutured onto the dog's liver. You can see a bluish suture on the upper right corner. At 2 1/2 weeks the sheet on the right remains transparent - you can still see the color of the liver.
The results of the pancreas site are shown next. The purplish tube at the top is the intestine. The pancreas is much lighter than surrounding tissue and looks pinkish-white. In the dog most of the pancreas is found associated with the intestine and the structures attaching it to the rest of the dog. The four sutures can be clearly seen. The sheet is so transparent that the pancreatic structure can be clearly seen.
9 weeks later
After 9 weeks the sheet is so clean that it is hard to make out. The sutures mark the corners. If a fibrotic response occurs, it starts within two weeks and is well advanced in two months. So these results considerably boost our confidence that the Islet Sheet will function for a long time if we keep it flat.
The following shows sheets on the dog's diaphragm.
This result confirms that a relatively rigid site like the diaphragm is all that is required to prevent fibrosis of the Islet Sheet.
Based on these results we plan soon to start experiments that attach Islet Sheets with islets to the liver, diaphragm and pancreas of dogs with diabetes. We believe that these are the most likely sites to achieve long term euglycemia.
To take a look at the previous studies click here.