Summary
Until recently the overall clinical success of islet transplantation in humans was low. Most clinical trials were undertaken in type 1 diabetic patients requiring a kidney transplant and were therefore subject to immune suppressive therapy. In a recently published study [1], seven type 1 diabetic patients received an islet graft without a kidney transplant; all became insulin-independent. The use of a new immune suppressive regimen without glucocorticoids was considered to be one element in this success [1]. In the present paper the same investigators report on the side effects of their protocol as applied in the first series of seven patients and in five subsequent cases; the median follow-up was 10 months.
All 12 recipients were C-peptide-negative type 1 diabetic patients with brittle diabetes, reduced hypoglycaemia awareness or progressive complications, and a diabetes duration of 29 ± 3.2 years. They received, on two or three occasions, a freshly isolated human islet preparation and were immune-suppressed with an anti-interleukin-2 receptor antibody (during the first 10 weeks) and with a combination of sirolimus (rapamycin) plus tacrolimus (continuous) [1]. All patients maintained insulin production throughout the observation period and stabilized their glycaemia at significantly lower levels without hypoglycaemic episodes and with reduced HbA1c levels (Fig. 1).

Fig. 1: (A) HbA1c at 3-month intervals post-islet transplant. All patients showed an improvement in HbA1c. (B) Capillary glucose values obtained from the memory glucose meter in patients undergoing islet transplantation over time. (C) Mean amplitude of glycaemic excursion (MAGE) values derived from frequent capillary glucose sampling (seven values/day for 2 consecutive days) in patients undergoing islet transplantation over time. Values are means ± SE.

At the end of this period, four patients had normal glucose tolerance, five were glucose-intolerant and three had diabetic glucose values for which they were treated with oral hypoglycaemic agents and low doses of insulin. Graft recipients had gained an insulin secretory response to an intravenous glucose bolus but this capacity was only 20% that of non-diabetic controls. These data confirm that islet allografts survive in all recipients treated with the Edmonton protocol and achieve a sustained but variable metabolic benefit for at least 3–20 months. This benefit certainly needs to be assessed again at a later point but can already be weighed against the side effects of this form of therapy.
The intraportal implantation caused transient side effects in three patients, one presenting a thrombosis in a peripheral branch of the portal vein and a further two with bleeding at the puncture site; the technique was subsequently adjusted to avoid these problems. Two patients had a vitreous haemorrhage, perhaps resulting from the fall in blood glucose levels.
The immune suppressive therapy was held responsible for the appearance of the following symptoms: the administration of tacrolimus is thought to have induced a rise in serum creatinine in two patients and in proteinuria in one patient. The use of sirolimus may have caused a rise in blood pressure in four patients and in cholesterol in five patients, and diarrhoea in five patients. Haemoglobin levels were in general decreased by 20%. One patient developed pneumonia. Most of these side effects were treated, either by drugs or by reducing the dose of tacrolimus. None of the recipients have so far developed cytomegalovirus disease, malignancies or serious infection.
The authors conclude that their protocol for islet transplantation offers a favourable risk-benefit ratio in patients with labile type 1 diabetes but consider elevated creatinine levels to be a contraindication.

Comment
Islet transplantation has long been considered a promising method for curing diabetes [2]. Compared with pancreas transplantation, the technique is simple and associated with lower morbidity; moreover, it offers the possibility of overcoming current obstacles to organ grafting such as the lifelong need for immune suppressive therapy and the shortage of donor tissue. Over the years, these expectations have been challenged by growing doubts arising from the low success of islet transplantation in achieving and maintaining insulin independence in humans [3]. Scepticism has, however, sharply declined since Shapiro et al. [1] reported seven successive cases of insulin independence following islet transplantation in C-peptide-negative type 1 diabetic patients. These results have attracted wide attention and have raised hopes among both the medical and patient community. They have also raised the question whether this technique can now be offered as a treatment or whether it should be further explored in clinical trials. Whichever is chosen, it will be necessary to define the inclusion and exclusion criteria for patient recruitment. These issues require consideration of several elements, one being the balance between clinical benefit on the one hand and side effects on the other. While such analyses will need long-term follow-up in larger patient groups, it is nevertheless useful to examine the initial group in this perspective, even if the number of patients is low (n = 12) and the observation period short (maximally 20 months).
The paper of Ryan et al. reports that no serious side effects were noted during this relatively short period except for a worsening kidney function in two patients who already had increased creatinine levels before the start of immune therapy. For this reason, increased creatinine levels in candidate recipients are viewed as a contraindication to the present protocol. Longer follow-up in the other patients will indicate whether their creatinine levels remain within normal control limits or whether they will eventually increase. The present protocol did not select patients on the basis of progressive kidney disease but recruited patients with labile diabetes. Since few patients can be included in current trials, it will be necessary to define more precisely the inclusion criteria.
Although several of the reported side effects were mild, they required medical attention and administration of additional drugs. The authors correctly emphasize that islet transplant protocols do carry a risk even if it remains lower than that of organ transplantation. The weight of this risk may vary according to the care of the transplant procedure and follow-up. It will probably be lower in an experienced team such as the Edmonton group. The weight of the graft benefit is dependent on the same care and experience.
The metabolic data demonstrate that all recipients showed a clear benefit in terms of stabilization of glycaemia at lower levels, which was normal in four of the 12 patients and near-normal in a further five. In three recipients, the grafts could not prevent their diabetes. It is unclear whether these differences in metabolic effect are related to differences in the donor tissue or in the recipient. The possibility exists that the implanted b-cell mass is still marginal in comparison with that in normal pancreases. Insulin supplements may subsequently be required in some recipients under certain conditions. A marginal b-cell mass may also explain the reduction in graft function which other centres have witnessed during the first post-transplantation years [4]. It remains to be established whether this interferes with the long-term metabolic control that is needed to prevent the development of chronic diabetic lesions [5].
If the authors had listed the metabolic data and side effects per patient, it would have indicated whether there was any relationship between the patients’ initial insulin needs and metabolic control on the one hand and the graft effect on the other. It would also have allowed an assessment of the risk-benefit ratio per patient. The paper does not tell the reader whether the risks and benefits were evenly distributed. Inclusion of a table to illustrate this would have been a helpful addition.

References
1. Shapiro AMJ, Lakey JRT, Ryan EA et al. Islet transplantation in seven patients with Type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N Engl J Med 2000; 343: 230–8.
2. Lacy PE. Treating diabetes with transplanted cells. Sci Am 1995; 273: 54–8.
3. International Islet Transplant Registry. Newsletter no. 9. www.med.uni-giessen.de/itr.
4. Davalli AM, Maffi P, Socci C et al. Insights from a successful case of intrahepatic islet transplantation into a type 1 diabetic patient. J Clin Endocrinol Metab 2000; 85(10): 3847–52.
5. Fioretto P, Steffes MW, Sutherland DER et al. Reversal of lesions of diabetic nephropathy after pancreas transplantation. N Engl J Med 1998; 339: 69–75.

Summary and Comment:
Daniel Pipeleers, Brussels, Belgium