Intrahepatic Islet Transplantation in Type 1 Diabetic Patients Does Not Restore Hypoglycemic Hormonal Counterregulation or Symptom Recognition After Insulin Independence

Seven islet allotransplant recipients, ages 30–54 years (mean age 43 ± 3 years) were examined at the University of Alberta Hospital between April and June 2001. The mean duration of insulin independence for all transplant recipients was 12.6 ± 0.6 months from the time of the final islet infusion. All recipients received transplants using the Edmonton Protocol for islet-alone transplantation (1). The primary indications for islet transplantation were the presence of type 1 diabetes for at least 5 years and either significant hypoglycemia unawareness, defined by the absence of adequate autonomic symptoms at plasma glucose levels <54 mg/dl (3.0 mmol/l), or severe metabolic instability, characterized by two or more episodes of severe hypoglycemia that required assistance from others over the 12 months before transplantation. Chronic immunosuppression in all patients consisted of combined low-dose sirolimus (rapamycin) and tacrolimus (FK-506) therapy in the absence of glucocorticoids. No subjects were taking glucose-lowering agents or β-blockers or had a history of epilepsy or active ischemic heart disease during the study. Results of the islet transplant group were compared to those of type 1 diabetic subjects and nondiabetic control subjects studied with the same stepped, hypoglycemic clamp technique. Four of the control subjects were studied in Edmonton at the same time as the islet transplant patients. The other three control subjects had been studied previously along with the type 1 diabetic subjects at the University of Minnesota. Each assay that included samples from an Edmonton control subject also contained samples from an Edmonton islet recipient.

The type 1 diabetic subjects and nondiabetic control subjects were matched for age and weight with the islet transplant subjects (Table 1). Fasting plasma glucose levels were normalized in all diabetic subjects using an overnight insulin infusion. All subjects gave informed, written consent before participation. The study was reviewed and approved by the Health Research Ethics Board of the University of Alberta.

All subjects underwent a 180-min stepped hypoglycemic clamp, as previously described (10). After a 12-h fast, intravenous access was obtained in a forearm vein of each subject for infusion of insulin and glucose. Venous blood was sampled from a separate intravenous catheter placed in the opposite arm. A blood sample was obtained every 5 min for the immediate determination of plasma glucose. After a 30-min equilibrium period, recombinant regular human insulin (Humulin; Lilly, Indianapolis, IN) was infused at 2 mU · kg⁻¹ · min⁻¹. A variable-rate 20% glucose infusion was used to maintain blood glucose concentrations at glucose targets of 70, 60, 50, and 40 mg/dl at sequential 45-min intervals. Blood samples were obtained during the basal period and at 15-min intervals for the determination of glucagon and epinephrine concentrations. Heart rate and blood pressure were monitored throughout the study. Hypoglycemic symptoms were assessed every 45 min using the symptom score of Hoeldtke et al. (11).

A 30-min arginine stimulation test was performed separately from the hypoglycemic clamp studies on four of the seven islet transplant recipients. After an overnight fast, intravenous access was obtained in a forearm vein of each subject and normal saline was infused at a rate of 25 ml/h to keep the vein open. After a basal blood sample was drawn, 5 g of arginine, diluted to 40 ml with normal saline, were injected over 30 s, followed by a 20-ml saline flush. Blood samples were then drawn at 2, 3, 4, 5, 7, 10, 15, 20, 25, and 30 min after arginine injection for determination of glucagon concentrations.

Plasma glucose was measured at the bedside using an automated glucose analyzer (YSI Life Sciences, Yellow Springs, OH). Blood samples were collected in test tubes containing EDTA and aprotinin (500 units/ml) (Trasylol; Bayer) for glucagon and EGTA (90 mg/ml) and glutathione (60 mg/ml) for catecholamines. Samples were cooled on ice and then rapidly centrifuged. Serum was separated and stored at −70° C for subsequent analysis. Glucagon levels were determined by radioimmunoassay using the method described by Harris et al. (12). Epinephrine and C-peptide levels were determined using single-isotope radioimmunoassay techniques (13,14). All assays were performed by the same technician using identical techniques in all islet transplant subjects, type 1 diabetic patients, and healthy control subjects.

All statistical comparisons were performed using one-way ANOVA using the Sigmastat program. Results are expressed as means ± SE. Statistical significance was set at P < 0.05.

The mean fasting plasma glucose levels of the islet transplant and control groups were 104 ± 0.2 and 90 ± 0.1 mg/dl, respectively (Fig. 1). The mean fasting plasma glucose level of the type 1 diabetic subjects after an overnight insulin infusion was 98 ± 0.9 mg/dl. There were no significant differences in the mean fasting and sequential 45-min glucose values among the three groups.

There were no statistical differences in the plasma C-peptide levels between the islet transplant recipients and control subjects during the clamp (Fig. 2). Type 1 diabetic subjects had undetectable C-peptide levels before the study (data not shown); therefore levels were not examined during the clamp.

The mean basal plasma glucagon level of the islet transplant group during the hypoglycemic clamp was higher than that seen in the type 1 diabetic subjects (210 ± 20 vs. 95 ± 19 pg/ml; P < 0.05), but was not significantly different from that of the control group (144 ± 23 pg/ml). A significant rise of plasma glucagon was observed in the control group during the clamp (180-min glucagon: 208 ± 36 pg/ml; P < 0.001 vs. basal). No significant rise in the mean plasma glucagon level was observed in the islet transplant group or the type 1 diabetic group during the clamp (180-min glucagon: 198 ± 14 and 91 ± 22 pg/ml, respectively) (Fig. 3A). The mean incremental glucagon response (basal to 180 min) of the islet transplant group was significantly less than that of the control group (−12 ± 12 vs. 64 ± 22 pg/ml; P < 0.05) and was not significantly different from that of the type 1 diabetic group (−17 ± 10 pg/ml) (Fig. 3B). An intact glucagon response to arginine challenge (basal to peak) was observed in the four islet transplant recipients studied (158 ± 34 pg/ml; P < 0.001) (Fig. 3A, inset).

The mean basal plasma epinephrine level of the islet transplant group was similar to that of the control group (18 ± 3 vs. 22 ± 3 pg/ml), but both were significantly less than that of subjects with long-standing type 1 diabetes (64 ± 5 pg/ml; P < 0.05). A significant rise in plasma epinephrine was observed in the control group during the clamp (180-min epinephrine: 655 ± 139 pg/ml; P < 0.001 vs. basal) (Fig. 4A). Two of the islet transplant recipients displayed an elevation of plasma epinephrine during the final 45 min of the clamp, but overall no significant rise in plasma epinephrine was observed in the islet transplant group. The mean incremental epinephrine response (basal to 180 min) of the islet transplant group was significantly less than that seen in the control group (195 ± 128 vs. 633 ± 139 pg/ml; P < 0.05), and not significantly different from that of subjects with long-standing type 1 diabetes (238 ± 73 pg/ml) (Fig. 4B).

The mean symptom scores of the islet transplant recipients did not differ significantly from those of subjects with long-standing type 1 diabetes at all glycemic levels (Fig. 5). None of the islet transplant recipients recognized any symptoms of hypoglycemia until they had reached a plasma glucose level ≤50 mg/dl. At a glucose level of 40 mg/dl, the mean symptom scores of the islet transplant recipients and type 1 diabetic subjects were significantly less than those observed in nondiabetic control subjects (symptom scores: 3.3 ± 0.9, 3.1 ± 1.1, and 6.7 ± 0.7, respectively; P < 0.05 vs. control).

The results of the current study indicate that despite prolonged stable insulin independence and near-normal glycemic control, intrahepatic islet allotransplantation did not restore hypoglycemic hormonal counterregulation in individuals with a history of type 1 diabetes and impaired counterregulatory responses. In particular, these results showed that glucagon and epinephrine responses and hypoglycemic symptom recognition were not improved by islet transplantation.

Currently, one of the primary indications for islet transplantation is severe hypoglycemia unawareness, which often occurs in patients with long-standing type 1 diabetes. This unawareness puts them at risk for frequent, severe hypoglycemic events, such as seizure and coma (1,3,4). Patients using intensive insulin regimens are at particularly high risk (15–17). Strict avoidance of low blood glucose levels has been shown to restore symptom recognition and epinephrine responses in most diabetic patients with hypoglycemia unawareness (18–20). However, individuals with long-standing diabetes (>15 years) appear to be less likely to show improvement in epinephrine response and symptom recognition after a period of hypoglycemia avoidance than individuals with diabetes of shorter duration (21).

Studies of counterregulation in whole pancreas transplant recipients have shown both early (1 month post-transplant) and sustained (up to 19 years post-transplant) improvements in glucagon and epinephrine responses to hypoglycemia, as well as improvements in hypoglycemic symptom recognition (6–10). However, because pancreas transplantation carries the risks and morbidities of major surgery as well as long-term immunosuppression, its application outside of kidney transplantation remains limited (22). For type 1 diabetic patients who are at high risk for severe hypoglycemic events, but are not candidates for whole pancreas transplantation, intrahepatic islet transplantation has been proposed as a means of not only stabilizing blood glucose, but also allowing patients to potentially avoid severe episodes of hypoglycemia by improving hormonal counterregulation and symptom recognition. However, the current study suggests that this is not the case.

This is the first study to systematically examine counterregulatory responses at severe hypoglycemic levels (40 mg/dl) in a group of stable islet allotransplant recipients with prolonged insulin independence (mean post-transplantation duration >12 months). One previous study examined hormonal counterregulation in three islet allotransplant recipients before and 1 month after transplantation (23). Glucagon responses were not improved in any of the subjects, although epinephrine, norepinephrine, and cortisol responses were somewhat improved. However, two of the three subjects were not insulin independent at the time of the study, and the third became insulin independent over a period of 2 weeks, suggesting tenuous glycemic control. Also, the glycated hemoglobin values of all three subjects after transplantation were >8.0%, suggesting that post-transplant glycemic control in these individuals was suboptimal, thereby limiting the conclusions that could be drawn from the study.

Another study of two islet allotransplant and four islet autotransplant recipients with prolonged insulin independence (23–107 months) found that glucagon responses of both auto- and allotransplant recipients did not improve after transplantation (24). However, the number of study subjects was small and the glucose nadir reached was only 55 mg/dl. Therefore, as with the previously discussed study, it is difficult to draw firm conclusions regarding the absence of α-cell responsiveness to hypoglycemia based on these results. In addition, the study did not examine epinephrine responses or hypoglycemic symptom recognition, both of which are important factors in protecting individuals with labile diabetes from severe hypoglycemia. An accompanying study suggested that impaired glucagon responsiveness might be related to the site of transplantation (25). This study found that the glucagon response was absent in pancreatectomized dogs with intrahepatic autotransplanted islets, but was intact in dogs with intraperitoneally transplanted islets. Thus it appears in this study that the failure of glucagon responsiveness was specific to the intrahepatic site.

The physiological explanation for such a defect remains unclear. Because the liver is the site of endogenous glucose production, higher intrahepatic glucose levels could potentially prevent transplanted islets from sensing peripheral hypoglycemia, thus causing an impairment of glucagon secretion. However, this explanation is not entirely satisfactory, as transplanted islets are presumably located on the portal venous side of the hepatic circulation, which tends to have a slightly lower ambient glucose level (25) and where it appears that an important glucosensor for the detection of hypoglycemia is located (26,27). More study is needed, therefore, if this apparent contradiction is to be resolved.

In the current study, the gradual reduction of the C-peptide levels observed in the islet transplant group during the clamp indicated an appropriate reduction of insulin secretion and verified that intrahepatic β-cells were able to respond to progressive hypoglycemia. Although there were no statistical differences in the C-peptide levels between the islet transplant and control groups, the levels appeared to be more reduced in the control group at lower plasma glucose values (i.e., <60 mg/dl). This finding raises the possibility that glucagon secretion in the islet transplant group may have been relatively more suppressed by insulin at glucose levels <60 mg/dl. However, as mentioned previously, a similar defect in glucagon secretion was found to exist in intrahepatic autoislet recipients who had C-peptide disappearance curves identical to those of the control group (24).

The robust glucagon response to arginine stimulation (Fig. 3A, inset) indicated the presence of functioning α-cells in the transplant recipients; however, this result does not differentiate between glucagon secreted by transplanted α-cells versus that secreted by the native pancreas. Clearly, glucagon secretion from native islets remained absent despite restoration of euglycemia.

Another potential cause of impaired α-cell function is toxicity of immunosuppressive drugs on transplanted islets. Tacrolimus has been shown to impair β-cell function and reduce insulin secretion in vitro and in vivo, but there is no clear evidence to show it impairs α-cell function in a similar fashion (28–31). Few studies have specifically examined the effect of sirolimus on α-cell function. However, there appears to be little effect of sirolimus on β-cell function in the setting of islet transplantation (32,33). More studies are needed before a firm conclusion can be reached regarding whether immunosuppressive toxicity plays a significant role in impaired α-cell function after islet transplantation.

In the current study, the epinephrine response of the islet transplant group also remained impaired after transplantation. This result was surprising, as a partial restoration of epinephrine response to hypoglycemia has been demonstrated in whole pancreas transplant recipients (9,10). In a study of pancreas transplant recipients who had autonomic insufficiency and poor symptom recognition, Kendall et al. (9) observed restoration of hormonal counterregulation and symptom recognition after transplantation.

As mentioned previously, strict avoidance of hypoglycemia in diabetic subjects has been shown to restore epinephrine responsiveness; however, individuals with long-standing diabetes appear to have a more modest recovery than individuals with diabetes of shorter duration (21). This may be the result of progressive diabetic autonomic neuropathy, which tends to be more significant in patients with long-standing diabetes (34,35). Because the mean duration of diabetes of the islet transplant recipients in the current study was 27 years, it is possible that underlying autonomic neuropathy was at least partially responsible for the lack of epinephrine response observed. However, because the current study was cross-sectional, comparisons of pre- and post-transplant autonomic function were not possible, so no firm conclusions regarding autonomic neuropathy can be drawn from these data.

Notably, two of the islet transplant recipients had significant rises in epinephrine during the study. One subject demonstrated a robust rise, whereas another had a more modest response. Because this was not a prospective study, it is impossible to know whether the epinephrine responses in these two subjects improved after transplantation. These two subjects also had the highest hypoglycemic symptom scores of the transplant group, including significant “adrenergic” symptoms such as tachycardia and perspiration, corresponding to an increase in plasma epinephrine. This correlation between symptoms and plasma hormone levels supports the proposition that the absence of epinephrine responses in the majority of islet transplant recipients is an accurate reflection of the counterregulatory state in these patients rather than an error in sample handling or assay technique.

Finally, symptom recognition did not improve after islet transplantation, even at very low plasma glucose levels (40 mg/dl). This result was also surprising, as pancreas transplantation and avoidance of hypoglycemia have both been shown to restore symptom recognition in diabetic subjects (9,10). As with epinephrine, the long duration of diabetes in this patient group may have made it less likely that they would recover symptom recognition after transplantation (21). Another explanation that may be considered is that despite apparent euglycemia, islet transplant patients could be subject to episodes of asymptomatic hypoglycemia after transplantation. As noted previously, the apparent lack of full suppression of C-peptide secretion at plasma glucose levels <60 mg/dl in the islet transplant group suggests that insulin secretion may persist at these levels, which could put these patients at risk for unrecognized hypoglycemia, especially after meals. If insulin secretion persisted inappropriately in the islet transplant patients, then a significant recovery of epinephrine or symptom response would not necessarily be expected.

In summary, the current study found that despite prolonged insulin independence and near-normal glycemic control in islet transplant subjects, intrahepatic islet transplantation does not appear to restore hypoglycemic counterregulation or symptom recognition in subjects with long-standing diabetes and loss of symptom awareness. The absence of glucagon responsiveness suggests either an intrahepatic factor or an acquired defect in the ability of transplanted α-cells to sense or appropriately respond to hypoglycemia. Alternatively, an inadequate reduction of insulin secretion from transplanted islets could result in the suppression of glucagon responsiveness, possibly related to the site of implantation. If this is the case, alternative sites might conceivably improve physiological counterregulation. Also, the absence of the recovery of epinephrine and symptom recognition may reflect an underlying autonomic neuropathy of long-standing diabetes or repeated subclinical hypoglycemic episodes after transplantation. Finally, even though we have not observed clinically significant hypoglycemic episodes in insulin-independent islet recipients, the failure to restore counterregulatory responsiveness after islet transplantation may put transplant recipients who resume using exogenous insulin at risk for recurrent episodes of hypoglycemia.

This work was supported by National Institutes of Health grants T32-EK-07247 (B.W.P.) and R01-DK-39994 (R.P.R.) and by the AB Heritage Foundation, the AB Foundation for Diabetes Research, the Medical Research Council of Canada, and the Juvenile Diabetes Research Foundation.

The authors wish to thank Sharleen Imes and the staff of the Clinical Investigation Unit at the University of Alberta Hospital for their assistance in performing the hypoglycemic clamps, and Elizabeth Oseid for her technical assistance in performing the glucagon and epinephrine assays. The assistance of the members of the clinical islet transplant team and the technical staff of the human islet isolation laboratory is also gratefully acknowledged.