Weight-corrected islet counts are predictive of outcome in the canine intrahepatic islet autograft model.

A combination of high-concentration collagenase digestion and density-gradient purification of canine pancreatic tissue made it possible to obtain relatively pure islets which could be quantified and the outcome of intrahepatic autotransplantation correlated with weight-corrected islet counts in the transplanted tissue. Of eleven dogs, seven achieved durable euglycaemia within 24 hours. All of these animals received islet doses of greater than 4,380 per kilogram of body weight. The remaining four animals became progressively hyperglycaemic necessitating sacrifice within one week; they all received islet doses of less than 4,380 per kilogram (p = 0.042). This model is therefore satisfactory for the investigation of preservation injury to islets and of allotransplantation because it identifies prospectively the transplants which ought to succeed, giving evidence for or against additional immune or ischaemic resistance to graft function.