RT R E P O R T Feasibility of Intraperitoneal Insulin Therapy With Programmable Implantable Pumps in IPPM A multicenter study HELENE HANAIRE-BROUTIN, MD CHRISTIANE BROUSSOLLE, MD NATHALIE JEANDIDIER, MD ERIC RENARD, MD BRUNO GUERCI, MD MARIE-JOELLE HAARDT, MD VERONIQUE LASSMANN-VAGUE, MD THE EVADIAC STUDY GROUP OBJECTIVE — To report the overall French experience, obtained through the collab- oration of seven centers (EVADIAC [Evaluation dans le Diabete du Traitement par Implants Actifs] register), on the safety, feasibility, and efficacy of intraperitoneal insulin therapy by programmable implantable pumps, using three different devices. RESEARCH DESIGN AND METHODS— This is a multicenter prospective study involving 224 type I diabetic patients implanted with a programmable implantable pump (cumulative follow-up: 353 patient-years; mean duration: 1.5 ± 0.9 years [mean ± SD]). The Infusaid and the Promedos devices are equipped with a side port and refilled with U100 insulin (Hoechst 21 PH); the Minimed pump is not equipped with a side port and is refilled with U400 insulin (Hoechst 21 PH). Metabolic data and adverse events were recorded in a central register run by EVADIAC. RESULTS — A total of 29 local pump-pocket events (8/100 patient-years) and 9 pump failures (2.5/100 patient-years) occurred. The major technical problems were 1) pump flow rate reduction related to insulin aggregates, reversible after alkaline rinsing of the pump, and 2) 47 catheter obstructions requiring laparoscopic or conventional surgery. Pump therapy was abandoned in only 11 patients. HbA lc (7.4 ±1.8 vs. 6.8 ± 1.0%, P < 0.001), mean glycemia (8.7 ± 1.5 vs. 7.8 ± 1.0 mmol/1, P < 0.001), and blood glucose SDs (3.8 ± 0.8 vs. 3.3 ± 0.8 mol/1, P < 0.001) decreased significantly after 6 months and remained lower than baseline thereafter. CONCLUSIONS — Intraperitoneal insulin infusion using an implantable programma- ble pump is a feasible and relatively safe technique that may improve metabolic control and glycemic stability. Long-term studies, however, are needed to demonstrate whether or not the improvement in glycemic control could be sustained for several years. From the Hopital Central (N.J.), Strasbourg; Hopital Rangueil (H.H.-B.), Toulouse; Hopital Lapeyronie (E.R.), Montpellier; Hopital la Timone (V.L.-V.), Marseille; Hopital Jeanne d'Arc (B.G.), Nancy; Hotel Dieu (M.-J.H.), Paris; and Hopital Lyon-Sud (C.B.), Lyon, France. Address correspondence and reprint requests to Helene Hanaire-Broutin, MD, Service d'Endocrinologie- Diabetologie, CHU Rangueil, 31054 Toulouse Cedex, France. Received for publication 30 March 1994 and accepted in revised form 27 October 1994. BM1, body mass index; CSII, continuous subcutaneous insulin infusion; DCCT, Diabetes Control and Complications Trial; EVADIAC, Evaluation dans le Diabete du Traitement par Implants Actifs. I ntensive diabetes management by mul- tiple subcutaneous injections or con- tinuous subcutaneous insulin infusion (CSII) is efficient in preventing microvas- cular complications of type I diabetes, but it increases the risk of severe hypoglyce- mia (1) and does not completely restore metabolic homeostasis. Intraperitoneal insulin delivery by implantable pumps results in preferential insulin absorption by the portal system (2) and in lower pe- ripheral levels of insulinemia (3). In the future, implantable devices could include a glucose sensor, creating a closed-loop feedback insulin delivery system. Previ- ous studies have reported encouraging re- sults in regard to the safety and efficacy of intraperitoneal insulin infusion, but most of these involved a small number of pa- tients and focused on a single device (4- 7). This article reports the results of the overall experience from all French centers through a central register run by the Eval- uation dans le Diabete du Traitement par Implants Actifs (EVADIAC) Society. RESEARCH DESIGN AND METHODS — The study was ap- proved by a French ethics committee, and 224 patients recruited in seven centers gave written consent. The patients' char- acteristics were age 39 ± 9 years, sex ratio (men:women) 1.3:1, duration of diabetes 19 ± 8 years, and body mass index (BMI) 24 ± 3 kg/m 2 (means ± SD). Selection criteria included type I diabetes with neg- ative response of plasma C-peptide to 1 mg intravenous glucagon, no hypoglyce- mia unawareness, absence of unstable ret- inopathy or advanced nephropathy, good compliance with intensive diabetes man- agement (at least 4 home blood glucose tests daily [Glucometer M Memory Meter, Miles, Elkart, IN]), and adherence to treatment goals (i.e., near-normal glyce- mia, avoidance of severe hypoglycemia, and monthly clinic visits). A total of 111 patients had stable retinopathy, 24 had overt nephropathy, and 46 had symp- tomatic neuropathy. Before implantation, patients were treated by multiple injec- 388 DIABETES CARE, VOLUME 18, NUMBER 3 , MARCH 1995