EXPERIENCE WITH INTRAPERITONEAL INSULIN INFUSION FROM IMPLANTABLE PROGRAMMABLE SYSTEMS IN TYPE-1 (INSULIN-DEPENDENT) DIABETES-MELLITUS PREVIOUSLY TREATED BY EXTERNAL PUMPS

The feasibility of the continuous intraperitoneal insulin infusion fro m implantable programmable systems was evaluated in 17 Type 1 (insulin -dependent) diabetic patients currently treated by external pumps. Eig ht subjects were using a continuous subcutaneous insulin infusion (cum ulated experience: 7.4 patients x years) and nine subjects had a cumul ated experience of 61.4 patients x years of continuous intra peritonea l insulin delivery by external pumps. The two studied implantable syst ems (A and B) were delivering the same pH neutral surfactant-stabilize d semi-synthetic human insulin, but at the respective concentrations o f 400 units per mi in groupe A (n = 12) and of 100 units per mi in gro upe B (n = 5), via a peritoneal catheter. The pump was implanted in a subcutaneous pocket created in the abdominal wall under general anesth esia. The reported trial of implantable programmable systems is 21 pat ient x years with an individual experience of 1.2 +/- 0.4 years (mean +/- SD). A sustained and stable glycaemic control was obtained, as ind icated by the haemoglobin Alc: 7.4 +/- 1.3 % (normal range: 4.3-6.1 %) , the mean blood glucose: 7.4 +/- 2.1 mM.1(-1), the standard deviation of blood glucose values: 3.4 +/- 0.4 mM.1(-l) and the percentages of 13.1 +/- 4.5 % and 6.1 +/- 6.1 % of blood glucose values respectively over 11 mM.1(-1) and under 2.8 mM.1(-l) after one year. The only metab olic event to occur was one episode of ketoacidosis secondary to a cat heter obstruction. No hypoglycaemic coma occurred contrasting with inc idences of 0.54 and 0.05 per patient x year during subcutaneous extern al and intraperitoneal pumps respectively (p < 0.02). Minor problems w ere one post-surgical regressive haematoma of the pump pocket in group A, five reversible slowdowns of the pump flow-rate in groupe B and oc casional non complicated defects of the pump communicator in both grou ps. Five patients, all under previous external intraperitoneal pumps, experienced irreversible obstructions of the peritoneal catheter, incl uding a double recurrence in one subject leading to the removal of the implanted system. The incidence of catheter obstructions was thus nul l for the patients under previous subcutaneous insulin infusion, but 0 .58 per patient x year for those under previous external intraperitone al systems (p < 0.03). We conclude that, although perfectible, the imp lantable programmable system is a reliable technique which provides in creased safety for Type 1 diabetic patients who are candidates for int ensive continuous insulin therapy. A peritoneal history of insulin del ivery appears however to significantly affect the life expectancy of t he catheter.