Clinical islet transplantation in type 1 diabetes mellitus: results of Australia’s first trial

Patients eligible for selection for our study had had type 1 diabetes mellitus for more that 5 years and were aged between 18 and 65 years. Additionally, they had recurrent severe hypoglycaemia unawareness with coma that required constant monitoring or regular intervention by a third party. As a result, they had developed worsening diabetic control that could not be managed by intensive insulin therapy (which included trial of an insulin pump in two cases). The risk of recurrent hypoglycaemia had to be judged by at least two of the investigators (P J O and D J H) to be greater than the overall risk of transplantation and immunosuppression. Patients had to be free of significant diabetic nephropathy (proteinuria < 300 mg/day) and renal impairment (glomerular filtration rate [GFR] > 60 mL/min/1.73m2). All patients gave informed consent, and the protocol was approved by the Human Research Ethics Committee of the Western Sydney Area Health Service.

Islets were separated by the closed loop method described by Ricordi et al.9,10 Pancreases were removed from heart-beating deceased donors. The pancreas was disaggregated by infusing the ducts with cold Liberase enzyme (Liberase Human Islet, Roche Applied Science, Indianapolis, Ind, USA). Dissociated islet and acinar tissue was separated on a continuous Biocoll (Biochrom AG, Berlin) density gradient (polysucrose 400 and amidotrizoic acid) on a refrigerated apheresis system (Model 2991, COBE Laboratories, Lakewood, Colo, USA).

Purified islets were counted and islet number and mass were expressed in terms of islet equivalents (IEQ).11 Islet preparations underwent a pre-transplant quality assurance process, which included a gram stain, purity and viability assessment, packed cell volume measurement and evaluation of islet morphology to exclude excessive fragmentation. Aliquots of transplanted islet preparations underwent microbiological culture, endotoxin assay and calculation of a glucose stimulation index (based on an in-vitro test of response to a glucose stimulus).12

To be suitable for islet transplantation, a pancreas had to yield more than 3000 IEQ/kg (recipient weight) in a packed-tissue volume of less than 10 mL. The islets were resuspended in 120 mL of medium 199 (ThermoTrace, Melbourne) containing 5000 U heparin and 20% human albumin. Patients received a general anaesthetic, and minilaparotomy was performed to access a mesenteric vein. An arterial angiographic catheter was inserted into the main portal vein with the assistance of image intensification. The islets were infused over 15–20 minutes. Portal venous pressure was monitored throughout. A Doppler ultrasound scan of the portal vein and liver was performed on Day 1 and Day 3 after transplantation to confirm that the portal circulation remained patent after the islet infusion.

Immunosuppression was instituted at the time of transplantation. All patients received tacrolimus (Astellas), sirolimus (Wyeth) and daclizumab (Roche). Tacrolimus was commenced at a dose of 0.1 mg/kg per day in two equal divided doses, with the dose then adjusted to achieve a target blood concentration of 3–7 ng/mL. Sirolimus was commenced at 15 mg per day for the first day, followed by 7 mg per day, with the dose then adjusted to achieve a target blood concentration of 12–15 ng/mL for the first 12 weeks. Daclizumab was given as an infusion of 1 mg/kg every 14 days for a total of five doses commencing on the day of transplantation. In cases in which a second islet transplant was required more than 10 weeks after the first, a further five doses of daclizumab were given.

Cephazolin 1 g three times a day for 48 hours and a single dose of imipenem 500 mg were given perioperatively. Oral cotrimoxazole (sulfamethoxazole 400 mg and trimethoprim 80 mg) was given for the first 9 months after transplantation as prophylaxis for Pneumocystis carinii. Oral valaciclovir 1 g three times a day was given for 3 months to prevent cytomegalovirus and herpes virus infections. Vitamin E 800 IU/day, vitamin B6 100 mg/day and vitamin A 5000 IU on alternate days were given for the first 3 months as supplementary antioxidant therapy.

As prophylaxis for portal vein thrombosis, all patients received heparin 5000 U intravenously at the time of islet infusion. The original aim was for patients to receive full anticoagulation (ie, a heparin dose sufficient to double the activated partial thromboplastin time) for the first 48 hours. But this was changed to 15 000 U daily of subcutaneous prophylactic heparin after the first recipient suffered bleeding complications. Towards the latter half of the trial (after Patient 5 experienced portal vein thrombosis), full anticoagulation was again reinstituted.

Insulin therapy was ceased in the immediate postoperative period and was recommenced when blood sugar level rose above 10 mmol/L. Intensive insulin management was used to maintain normal glucose levels after transplantation (< 8 mmol/L postprandial), with the aim of reducing islet stress during engraftment. After 1 month, insulin use was gradually reduced and/or ceased, depending on glycaemic control (assessed by home blood glucose monitoring).

Islet function was measured by assessing serum C-peptide levels. Patients also underwent glucose tolerance testing at least annually. Glycosylated haemoglobin and lipid levels were measured, and renal function was assessed at least every 3 months.

Of 50 patients referred for evaluation by an endocrinologist, six were selected for our trial. Patient characteristics and graft outcomes are shown in Box 1. All had undetectable C-peptide levels before transplant and all had evidence of severe hypoglycaemia unawareness that had failed to respond to intense management of their insulin therapy. In addition to hypoglycaemia unawareness, one patient had antibodies to exogenous insulin and required immunosuppression with azathioprine to maintain adequate glycaemic control. Apart from hypoglycaemia unawareness, the patients were remarkably free of end-organ diabetic complications. Mean GFR was 132 mL/min/1.73m2 (range, 104–153 mL/min/1.73m2). Two had had laserphotocoagulation treatment for diabetic retinopathy, two had mild subclinical neuropathy, and two had hypertension (with blood pressure well controlled by an angiotensin-converting enzyme [ACE] inhibitor) and microalbuminuria (maximum protein excretion when not taking an ACE inhibitor was 257 mg/day but normal when taking medication).

The six selected patients underwent islet transplantation between October 2002 and February 2005. The median follow-up time was 18 months (range, 3–31 months). The mean number of islet equivalents transplanted was 17 958 IEQ/kg (range, 6995–26 480 IEQ/kg). The absolute number of islets transplanted did not correlate with success in achieving insulin independence (Box 2). Five of the six patients received two islet infusions. One patient developed a portal vein thrombosis after the first infusion, had no measurable C-peptide (indicating there was no graft function), and was withdrawn from the study. Of the five remaining patients, all had evidence of islet function after the first graft. C-peptide levels and total insulin dose before and after transplant are summarised in Box 3. Three recipients were able to cease insulin treatment completely for a period of time. However, one of them withdrew from the study after 7.5 months because of intolerance to the immunosuppressive medication, suffering nausea and mouth ulcers. The remaining two had evidence of substantial islet function, with evidence of C-peptide secretion and major reductions in insulin dose.

All recipients with evidence of islet function showed marked improvements in blood glucose control and haemoglobin A1c (HbA1c) level (Box 3). In all patients, there was a reduction in insulin requirement and a fall in HbA1c level after the first islet infusion. There was a further fall in insulin requirement after the second infusion, even in recipients who did not completely stop taking insulin. Severe hypoglycaemic episodes were abolished. Three patients had episodes of mild hypoglycaemia (blood sugar level, 2.5 –3.5 mmol/L), but none required third-party intervention. These episodes became less frequent or ceased with time and were rarely seen 6 months after the second transplant. In the two patients (Patients 1 and 2) who received a second islet transplant and were insulin free for 2 years, there was evidence of loss of islet function with impairment of glucose tolerance towards the end of the second year (Box 4). Both patients have now returned to insulin therapy. Patient 1 has good control with 6–10 U insulin a day, and remains C-peptide positive with HbA1c level < 7%. (Outcomes were the same for Patient 2 until the graft ultimately failed.) Episodes of mild hypoglycaemia did occur in Patient 1 following the onset of graft dysfunction, but none of these required intervention by a third party. In Patient 3, who ceased immunosuppressive treatment after graft loss, and Patient 5, who never achieved any graft function, episodes of severe hypoglycaemia continued to occur.

Procedure-related side effects included a postoperative bleed in Patient 1 on Day 1 requiring transfusion and laparotomy for evacuation of haematoma. The patient had a rapid and full recovery. Patient 5 experienced right portal vein thrombosis, with eventual recovery, and Patient 6 had a partial left portal vein thrombosis that resolved completely within 9 days. In addition, all patients had a mild rise (a less than two-fold increase) in gamma-glutamyl transferase and alanine aminotransferase levels, and two patients had ultrasound evidence of fatty liver.

Side effects of immunosuppressive medication included mouth ulcers, raised cholesterol level, ankle swelling and, in one patient, a presumed recurrence of tuberculosis (symptoms were fevers, weight loss and ketonuria, which resolved with antituberculous therapy, although no organism was grown). Surprisingly, hypertension was not a major side effect, and no new patients commenced antihypertensive therapy. However, Patient 2 showed a significant fall in isotopic GFR at 18 months (from 143 mL/min/1.73m2 pretransplant to 63 mL/min/1.73m2) and required cessation of tacrolimus and substitution with mycophenolate mofetil. Patient 3 was diagnosed with a skin squamous cell carcinoma within 3 months of the transplant, but it is likely that it was present before commencement of immunosuppression.