Islet transplantation (IT) can restore normoglycemia and improve the quality of life of patients with unstable type 1 diabetes mellitus (T1DM) (1). We have previously reported an insulin independence rate of 79% at 1 year (2), similar to the experience by the Edmonton group (3). However, despite sustained C-peptide levels, only a small percentage of patients (<10%) maintain insulin independence at 5 years (4).
Several studies in islet allograft recipients have shown greatly reduced responses when compared with controls (2, 3, 5–8) suggesting a markedly reduced and dysfunctional islet mass. Many factors may play a role in leading to islet allograft dysfunction (IGD) including impaired engraftment (9), toxicity from the immunosuppressant agents (10, 11), allo-rejection (12, 13), recurrence of autoimmunity (14, 15), and β-cell metabolic exhaustion. The appearance of persistent hyperglycemia in a previously insulin independent normoglycemic islet allograft recipient is usually the outcome of an event(s) that led to dysfunction or loss of the islet mass rather than an indicator of an acute event. Any type of intervention at this point may prevent further dysfunction although it may be too late to recover whatever function was lost.
The detection of marker(s) that may predict IGD becomes crucial and could allow for early intervention(s) to preserve functional islet mass and maintain long-term insulin independence.
The aim of our study was to compare the metabolic responses between subjects who maintain insulin independence and those who eventually require reintroduction of insulin and to identify metabolic markers that may predict IGD.
PATIENTS AND METHODS
Sixteen subjects with long standing history of T1DM and hypoglycemia unawareness who underwent IT were recruited. The islet isolation, transplantation procedure, and immunosuppressive regimen have been described previously (2). Briefly, immunosuppression consisted of maintenance starting at postoperative day-1 with tacrolimus (Prograf, Fujisawa, Japan) target trough level 4 to 6 ng/mL and sirolimus (Rapamune, Wyeth Pharmaceuticals Inc., Madison, NJ) target trough level 12 to 15 ng/mL for 3 months and 10 to 12 ng/mL thereafter, and a five-dose induction course of daclizumab (Zenapax, Roche, Nutley, NJ) 1 mg/kg intravenously biweekly beginning the day of transplant. Two patients who received a single infusion and required cessation of immunosuppression because of immunosuppression related adverse events are not included in the analysis. The research protocol was approved by the University of Miami health research ethics board (Institutional Review Board) and each subject gave written informed consent.
C-peptide positive islet transplant recipients with capillary fasting and 2-hr postprandial glucose levels less than or equal to 140 mg/dL and less than or equal to180 mg/dL, respectively and an HbA1c less than or equal to 6.5% (16).
C-peptide positive islet transplant recipients with fasting capillary glucose levels more than 140 mg/dL and 2-hr postprandial capillary glucose levels more than 180 mg/dL in three or more occasions in 1 week or two consecutive HbA1c values more than 6.5% leading to reintroduction of insulin treatment.
Islet Transplant Completion
Attainment of insulin independence, as defined above, after one or up to two islet infusions.
Subjects underwent metabolic testing after an overnight fasting (8–12 hr) every 3 months postislet transplant completion up to 18 months and every 6 months thereafter as long as graft function was maintained (detectable fasting C-peptide or stimulated C-peptide >0.3 ng/mL). Tests were only conducted if capillary glucose was less than or equal to 140 mg/dL before the test. Plasma glucose concentrations were determined by the hexokinase method. Plasma C-peptide was measured by double antibody radioimmunoassay (detection limits: 0.3–5.0 ng/mL, inter- and intraassay variation coefficient ≤10%, and cross-reactivity with insulin and proinsulin: 20%). Serum insulin was measured by a solid phase 125I radioimmunoassay (DPC, Los Angeles, CA) in a gamma counter (detection limits: 3.0–400.0 μIU/mL; interassay variation coefficient 4.5%; intraassay variation coefficient 7.6%; and cross-reactivity with proinsulin: 20%).
Mixed Meal Tolerance Test
Mixed meal tolerance test (MMTT) was conducted on the first day. Subjects ingested 12 oz of boost high protein (calories 360, total fat 9 g, total carbohydrate 49.5 g, and total protein 22.5 g) over a minute. Samples for glucose and C-peptide were obtained at 0, 15, 30, 60, 90, 120, 150, 180, 210, 240, 270, and 300 min after ingestion. We assessed the area under the curve for glucose (AUCglucose) and area under the curve for C-peptide (AUCC-peptide), mixed meal stimulation index (MMSI), 90-min glucose and C-peptide, peak glucose and C-peptide (defined as the highest glucose and C-peptide achieved during MMTT), and time-to-peak glucose and C-peptide. AUCglucose and AUCC-peptide were calculated using the trapezoidal rule and represent the area above baseline. The MMSI was calculated as follows:
Intravenous Glucose Tolerance Test
Intravenous glucose tolerance test (IVGTT) was performed the day after the MMTT. A solution of 50% dextrose (300 mg/kg) was infused over a minute and samples for glucose, C-peptide, and insulin were obtained at −10, 0, 3, 4, 5, 7, 10, 15, 20, 25, 30, and 60 min. We measured the acute C-peptide release to glucose (ACpRg) and acute insulin release to glucose (AIRg). ACpRg and AIRg were defined as the average of c-peptide or insulin values at 3, 4, and 5 min minus the average of baseline values (3, 18, 19).
Arginine Stimulation Test
Arginine stimulation test was performed 1 hr after the IVGTT. A dose of 5 g of arginine hydrochloride (R-gene 10, Pharmacia & Upjohn Company, New York, NY) was infused over 30 sec and samples were collected at −10, 0 (halfway through the arginine infusion), 2, 3, 4, 5, 7, and 10 min. We measured the acute C-peptide release to arginine (ACpRarg), and acute insulin release to arginine (AIRarg). ACpRarg and AIRarg were defined as the average of the three highest c-peptide or insulin values at 2, 3, 4, and 5 min minus the −10 min value (19, 20).
The C-peptide/glucose ratio (CPGR) was obtained from fasting C-peptide and glucose values at each of the testing time points and calculated as described previously (16).
Homeostatic Model Assessment
Insulin resistance was measured with the homeostatic model assessment (HOMA)-insulin resistance (IR) and calculated as follows:
Preliminary Analysis: Assessing Insulin Independent Subjects Versus Subjects Who Required Reintroduction of Insulin
As a preliminary analysis of the data, we first assessed differences in the metabolic responses between subjects who remained insulin independent at 18 months postcompletion (group 1) and those who required reintroduction of insulin within 18 months postcompletion (group 2). Five subjects (IA4, 5, 9, 12, and 16) comprise group 1 and nine subjects (IA1, 3, 6, 7, 8, 10, 13, 14, and 15) comprise group 2. Comparisons were made between groups at 3, 6, 9, 12, and 15 months to assess differences in each of the metabolic measures. Subjects in group 2 were not considered in comparisons at time points after they restarted insulin, so that the data analyzed would be an appropriate reflection of islet graft function without the effect of exogenous insulin. Comparisons between both groups were only possible up to 15 months postcompletion after which time all subjects in group 2 had restarted insulin therapy. At each time point under consideration, preliminary comparisons were made with two sample t-tests to assess differences in metabolic variables between groups. Values are expressed as mean±SE unless otherwise noted. Differences were considered statistically significant at a P value less than 0.05.
Linear Mixed Model Regression Analysis: Comparison of Measures Observed at Time Points Preceding Intervals With Dysfunction With Measures Observed at Time Points Preceding Intervals Without Dysfunction
Of primary interest in this investigation is to determine factors that may predict impending IGD. Thus, it is of interest to assess how metabolic measures may be different when observed at time points preceding intervals where dysfunction occurs in comparison with measures observed at time points preceding intervals without dysfunction. For each of the metabolic measures under consideration, we performed a repeated measures analysis of the data using linear mixed model regression. This method generalizes linear regression techniques to allow for repeated observations by taking into account the correlation that exists within observations on the same subject to more appropriately estimate variances used for the various tests of significance. In the regression model, a covariate was included for each observation which indexed if the outcome was measured at a time point preceding the interval where the patient experienced IGD or not. Using this approach, we were able to simultaneously estimate and test differences in each of the metabolic measures under consideration between measures taken at time points preceding intervals, where dysfunction occurred and time points preceding intervals where there was no dysfunction. This analysis considers data from every patient at every time point where the patient has not developed IGD. Analysis was performed using SAS 9.1 software (SAS Institute Inc., Cary, NC).
Fourteen subjects (seven men and seven women) with a mean age of 43±2 years and duration of diabetes of 29±3 years who underwent IT at our Institution between February 2001 and July 2003 were evaluated (Table 1). Subjects received an average of 13,633±804 IEQ/kg and all achieved insulin independence, 13 after two islet infusions and one (IA 9) after a single infusion. Five subjects who remained insulin free at 18 months (IA 4, 5, 9, 12, and 16) were assigned to group 1 and the remaining nine subjects (IA 1, 3, 6, 7, 8, 10, 13, 14, and 15) were assigned to group 2. Subjects in group 2 were progressively removed from analysis as they restarted insulin. Diabetes duration was longer in group 2, 33±3 years vs. 20±4 years, P=0.023 (Table 1). Seven of nine subjects (78%) in group 2 had a history of microvascular complications as compared with two of five subjects (40%) in group 1. Group 2 received more IEQ/kg, 15,035±649 IEQ/kg vs. 11,108±1,376 IEQ/kg, P=0.012. Group 1 developed IGD at a median of 1,194 days postcompletion (range 679–1,397) and restarted insulin at a median of 1,197 days (range 828–1454), whereas group 2 developed IGD at a median of 263 days postcompletion (range 42–451) and restarted insulin at a median of 380 days (range 167–471).
Mixed Meal Tolerance Test
Group 1, Insulin Independent Subjects at 18 Months Postcompletion
At 3 months (baseline), basal glucose was 98±3 mg/dL with a C-peptide of 1.19±0.09 ng/mL (n=5). The 90-min glucose was 103±11 mg/dL with a C-peptide of 2.42±0.28 ng/mL (Table 2). Peak glucose was 139±6 mg/dL and peak C-peptide 2.98±0.18 ng/mL occurring at 42±7 min and 66±15 min, respectively. MMSI was 2.82±0.41 pmol/mg. Subjects achieved a 2.53±0.19 ng/mL C-peptide increase from basal (n=5) at 3 months, which persisted by 18 months (2.57±0.28 ng/mL, n=4).
At 18 months, 90-min glucose was significantly higher as compared with baseline, 148±8 mg/dL (n=5) vs. 103±11 mg/dL (n=4), P=0.016 (Table 2) and peak glucose, 162±7 (n=4) vs. 139±6 mg/dL (n=5), P=0.04 (Fig. 1). By contrast, fasting glucose was not significantly higher, 103±6 mg/dL (n=4) vs. 98±3 mg/dL (n=5), P=0.41. Time-to-peak glucose and C-peptide were both delayed to 75±9 min (n=4; P<0.05) and 98±14 min (n=4; P=NS), respectively. The MMSI, which had consistently remained above 2.0, declined to 1.83±0.66 (Table 2).
Group 2, Subjects Restarting Insulin Treatment Within 18 Months Postcompletion
Fasting glucose levels were consistently in the impaired fasting glucose range and significantly higher than group 1 at baseline, 3, and 6 months (P<0.01) (Table 2). Basal C-peptide levels were also higher at 3 months (1.85±0.24 ng/mL, n=9) although not significantly higher than group 1 (1.19±0.09 ng/mL, n=5; P=0.08) and remained comparable between groups thereafter (Fig. 1). Time-to-peak C-peptide was 110±14 min at baseline (n=9), remaining higher than group 1 and practically unchanged over time (Fig. 1). From 6 to 15 months, group 2 had significantly higher 90-min glucose, peak glucose, and AUCglucose than group 1 (Table 2). Similarly, 90-min C-peptide and peak C-peptide were significantly higher at baseline and remained higher than group 1 (Fig. 1). MMSI was higher than group 1 at baseline (6.33±2.46, n=9 vs. 2.82±0.41, n=5; P=NS) but decreased to 1.68±0.21 by 6 months and to less than 1.0 by 15 months (0.67±0.2, n=2).
Peak C-peptide increase from basal was 2.19±0.18 ng/mL (n=9) at 3 months and it was maintained throughout follow-up with a 2.76±0.28 ng/mL (n=3) and 2.55±0.18 ng/mL (n=2) increase at 12 and 15 months, respectively.
Intravenous Glucose Tolerance Test
Fasting glucose at baseline was similar in both groups, 103±5 mg/dL in group 1 (n=5) and 104±4 mg/dL in group 2 (n=9). Group 1 showed an initial AIRg of 27.70±2.83 μIU/mL and maintained stable responses for the first 15 months demonstrated by a fasting glucose of 92±5 mg/dL and AIRg of 22.62±4.52 μIU/mL (n=5). A significant reduction in AIRg as compared with baseline was seen at 18 months, 15.29±4.46 μIU/mL (n=5) vs. 27.70±2.83 μIU/mL (n=5) (P=0.047), whereas mean fasting glucose was almost unchanged at 104±6 mg/dL (n=5). By contrast, group 2 started with a markedly lower AIRg compared with group 1 (16.14±3.69, n=8 vs. 27.70±2.83 μIU/mL, n=5; P=0.048) (Fig. 2), despite having similar mean fasting glucose values, and the AIRg continuously declined with time. By 12 months, fasting glucose was 110±10 mg/dL and the AIRg had decreased to 5.62±1.21 μIU/mL (n=4), a 65% reduction as compared with baseline (Fig. 2). Data for a single patient in group 2 were available for analysis at 15 months and showed a fasting glucose of 131 mg/dL with a suppressed AIRg of 1.31 μIU/mL (Fig. 2).
Initial ACpRg was 0.99±0.14 ng/mL in group 1 (n=5), remaining unchanged until 15 months (0.84±0.20 ng/mL, n=4) with a slight reduction to 0.70±0.19 ng/mL (n=5) by 18 months. Group 2 had an ACpRg at baseline of 0.70±0.11 ng/mL (n=9), almost identical to the response in group 1 at 18 months. ACpRg in group 2 declined as a function of time reaching a value of 0.37±0.09 ng/mL (n=4) by 12 months.
Arginine Stimulation Test
Both groups showed similar responses at baseline, with an AIRarg of 24.55±6.73 μIU/mL and ACpRarg of 0.89±0.09 ng/mL (n=5) in group 1 and an AIRarg of 20.66±2.30 μIU/mL and ACpRarg of 0.82±0.08 ng/mL (n=9) in group 2 (Fig. 2). Basal glucose was higher in group 2 at all time points and significantly so at 6 (P=0.03) and at 9 months (P=0.01). However, AIRarg responses remained stable and were not significantly different between groups at any time point (Table 3).
CPGR in group 1 was 1.21±0.06 at 3 months (n=5). Responses remained consistently above 1.00 and practically unchanged at 18 months (1.14±0.08; n=5). In group 2, CPGR at baseline was 1.63±0.21 (n=9), slightly higher than in group 1 (P=NS). It declined by 12 months to 0.99±0.11 (n=5) and remained unchanged by 15 months (1.00±0.16; n=2).
HOMA-IR was comparable between both groups at 3 months, 1.88±0.12 (n=5) in group 1 vs. 1.59±0.16 (n=9) in group 2. Group 1 showed a significant reduction by 6 months to 1.0±0.15 (n=5; P=0.002) and this reduction was maintained by 15 months. By contrast, HOMA-IR in group 2 remained stable for the first 9 months but increased by 12 months and was significantly higher than in group 1, 1.99±0.47 (n=5) vs. 0.79±0.05 (n=5), P=0.03.
Correlation Between IEQ/kg and Metabolic Markers
All measured stimulated MMTT C-peptide responses at baseline correlated with total IEQ/Kg with the highest correlation coefficient given by the 90-min C-peptide (r=0.643, P<0.001; n=14). There was a positive correlation with the basal glucose from the MMTT at baseline and the total IEQ/kg transplanted (r=0.627, P<0.01; n=14). Neither the AIRg nor the AIRarg correlated with the transplanted IEQ/kg when measured at 3 months.
Linear Mixed Model Regression Analysis
The AIRg, ACpRg, and MMSI are significantly decreased when measured at posttransplant time points preceding intervals where dysfunction occurred compared with intervals where there was no dysfunction. Similarly, 90-min glucose, time-to-peak C-peptide, and AUCglucose from the MMTT are significantly increased when measured at posttransplant time points preceding intervals where dysfunction occurred compared with posttransplant time points preceding intervals where there was no dysfunction (Table 3).
Additionally, analysis reveals a significant association between basal glucose and AIRg. A 10 mg/dL increase in basal glucose is significantly associated with a 3.38 unit decrease in AIRg (P<0.0001) Table 4.
Despite normalization of glycemic control and achievement of insulin independence, β-cell secretory responses to glucose and nonglucose secretagogues are significantly impaired in islet transplant recipients (2, 3, 5, 7, 8). Recipients of intrahepatic islet autotransplants also display markedly reduced metabolic responses with the difference that in this setting insulin independence with near-normoglycemia can be maintained in many cases for over 10 years (22). However, as differs to islet autotransplantation, islet allografts are exposed to several factors that can lead to dysfunction or loss of the β-cell mass, including allorejection, recurrence of autoimmunity, and toxicity from the immunosuppressive drugs, which likely account for the reduced long-term insulin independence rates (4, 23).
The subjects reported herein demonstrated markedly reduced responses when compared with control data available in the literature (3, 5, 7, 19). Clear differences were detected at baseline between the metabolic responses of both groups. Group 2 showed higher basal glucose, basal C-peptide, and peak C-peptide during the MMTT. These findings may suggest an inadequate islet mass in group 2 leading to a compensatory increase in β-cell function to maintain euglycemia. Insulin resistance may have also explained these results but we did not find significant differences in HOMA-IR values between groups at baseline. In addition, Rickels et al. (24) demonstrated that IT on sirolimus-tacrolimus maintenance immunosuppression led to an improvement on insulin sensitivity as compared with subjects with T1DM.
Group 2 also demonstrated a progressive increase in the 90-min glucose with a concomitant increase in AUCglucose and MMSI during a MMTT. All these variables proved to be significantly associated with development of IGD. The AUCglucose during oral glucose tolerance testing has been described as significantly more predictive of T1DM than the 2-hr glucose in islet cell autoantibodies (ICA)-positive relatives of type 1 diabetic patients (25).
Despite a progressive deterioration in glucose levels during the MMTT and a slight decline with time in basal C-peptide in group 2, both groups were able to maintain an almost identical increase, consistently more than two fold, in stimulated C-peptide during the MMTT suggesting that despite alterations in C-peptide production in the basal state, transplanted β-cells retain the ability to double their secretory capacity in response to ambient glucose.
Interestingly, group 2 developed IGD earlier than group 1 despite receiving more IEQ/kg. One possible explanation for this observation could be errors in the islet count because islet enumeration currently remains subjective. Another possibility is changes in the hepatic microvascularity. Group 2 had a longer duration of diabetes and a higher incidence of microvascular complications as compared with group 1. The role of chronic diabetes in liver disease has been described and diabetic microangiopathy has been associated with liver sinusoidal abnormalities (26, 27). The longer duration of diabetes in group 2 may have contributed to alterations in the hepatic vascular bed leading to problems with islet engraftment.
Abnormalities in the first phase of insulin secretion have been well documented in the prediabetic stage of T1DM (28, 29) and a loss of first phase is the classical finding at diagnosis of type 2 diabetes mellitus (30, 31). This leads to an abnormal regulation of postprandial glucoses resulting in postprandial hyperglycemia and it is usually the result of a qualitative defect of the β cell, a markedly reduced islet mass, or a combination of both. We observed markedly reduced AIRg responses at baseline in both groups with significantly lower responses in group 2 suggesting an already dysfunctional and likely lower engrafted mass in these subjects. Brunzell et al. (32) evaluated the insulin secretion and glucose disappearance rate in 66 subjects with a wide range of fasting plasma glucose and reported that the AIRg is consistently absent when fasting glucose levels exceed 115 mg/dL. However, similar data are not available for islet transplant subjects who are insulin independent by islet transplant criteria (16). We observed that in the few subjects with fasting glucose levels between 116 and 140 mg/dL, a reduced AIRg was still present. Monitoring of these AIRg responses proved to be a significant predictor of IGD.
In the setting of type 2 diabetes mellitus, despite a loss of the first phase of insulin release to glucose, β cells are still able to respond to nonglucose secretagogues such as arginine suggesting primarily a qualitative defect in the glucose sensing ability of the β cell (31).
Both groups maintained comparable AIRarg responses, demonstrating that transplanted islets also tend to first lose their glucose sensing properties but maintain responses to nonglucose secretagogues. It is possible that in the setting of a markedly reduced first phase of insulin release and a stable preserved AIRarg, a qualitative defect may be the primary problem rather than a reduced islet mass but this remains to be demonstrated.
All MMTT stimulated C-peptide responses at baseline correlated with total IEQ/kg transplanted but we found no correlation with the AIRg or AIRarg. Paradoxically, there was a positive correlation with the basal glucose from the MMTT and the total IEQ/kg transplanted (r=0.627, P<0.01; n=14) which further raises the possibility of errors in islet enumeration and stresses the need for objective islet counting techniques.
Marked metabolic abnormalities are evident early post-IT in subjects who eventually will develop IGD. Monitoring of the 90-min glucose, time-to-peak C-peptide, AUCglucose, and MMSI from a MMTT and the AIRg and ACpRg may predict the onset of IGD.
Simple tests, such as the IVGTT and MMTT, provide adequate graft function information and may be useful in the prediction of IGD making them essential components of the metabolic testing of islet transplant recipients.
The authors thank the staff of the Clinical Islet Transplant Program for their continued support.
1.Poggioli R, Faradji RN, Ponte G, et al. Quality of life after islet transplantation. Am J Transplant
2006; 6: 371.
2.Froud T, Ricordi C, Baidal DA, et al. Islet transplantation in type 1 diabetes mellitus using cultured islets and steroid-free immunosuppression: Miami experience. Am J Transplant
2005; 5: 2037.
3.Ryan EA, Lakey JR, Paty BW, et al. Successful islet transplantation: Continued insulin reserve provides long-term glycemic control. Diabetes
2002; 51: 2148.
4.Ryan EA, Paty BW, Senior PA, et al. Five-year follow-up after clinical islet transplantation. Diabetes
2005; 54: 2060.
5.Rickels MR, Schutta MH, Markmann JF, et al. β-cell function following human islet transplantation for type 1 diabetes. Diabetes
2005; 54: 100.
6.Alejandro R, Lehmann R, Ricordi C, et al. Long-term function (6 years) of islet allografts in type 1 diabetes. Diabetes
1997; 46: 1983.
7.Ryan EA, Lakey JR, Rajotte RV, et al. Clinical outcomes and insulin secretion after islet transplantation with the Edmonton protocol. Diabetes
2001; 50: 710.
8.Hering BJ, Kandaswamy R, Ansite JD, et al. Single-donor, marginal-dose islet transplantation in patients with type 1 diabetes. JAMA
2005; 293: 830.
9.Johansson H, Lukinius A, Moberg L, et al. Tissue factor produced by the endocrine cells of the islets of Langerhans is associated with a negative outcome of clinical islet transplantation. Diabetes
2005; 54: 1755.
10.Desai NM, Goss JA, Deng S, et al. Elevated portal vein drug levels of sirolimus and tacrolimus in islet transplant recipients: Local immunosuppression or islet toxicity? Transplantation
2003; 76: 1623.
11.Shapiro AM, Gallant HL, Hao EG, et al. The portal immunosuppressive storm: Relevance to islet transplantation? Ther Drug Monit
2005; 27: 35.
12.Olack BJ, Swanson CJ, Flavin KS, et al. Sensitization to HLA antigens in islet recipients with failing transplants. Transplant Proc
1997; 29: 2268.
13.Campbell PM, Salam A, Ryan EA, et al. Pretransplant HLA antibodies are associated with reduced graft survival after clinical islet transplantation. Am J Transplant
2007; 7: 1242.
14.Stegall MD, Lafferty KJ, Kam I, et al. Evidence of recurrent autoimmunity in human allogeneic islet transplantation. Transplantation
1996; 61: 1272.
15.Worcester Human Islet Transplantation Group. Autoimmunity after islet-cell allotransplantation. N Engl J Med
2006; 355: 1397.
16.Faradji RN, Monroy K, Messinger S, et al. Simple measures to monitor beta-cell mass and assess islet graft dysfunction
. Am J Transplant
2007; 7: 303.
17.Scharp DW, Lacy PE, Weide LG, et al. Intraportal islet allografts: The use of a stimulation index to represent functional results. Transplant Proc
1991; 23(1 Pt 1): 796.
18.Chen M, Porte D Jr. The effect of rate and dose of glucose infusion on the acute insulin response in man. J Clin Endocrinol Metab
1976; 42: 1168.
19.Teuscher AU, Kendall DM, Smets YF, et al. Successful islet autotransplantation in humans: Functional insulin secretory reserve as an estimate of surviving islet cell mass. Diabetes
1998; 47: 324.
20.Ward WK, Bolgiano DC, McKnight B, et al. Diminished B cell secretory capacity in patients with noninsulin-dependent diabetes mellitus. J Clin Invest
1984; 74: 1318.
21.Levy JC, Matthews DR, Hermans MP. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care
1998; 21: 2191.
22.Robertson RP, Lanz KJ, Sutherland DE, et al. Prevention of diabetes for up to 13 years by autoislet transplantation after pancreatectomy for chronic pancreatitis. Diabetes
2001; 50: 47.
23.Close N, Alejandro R, Hering B, et al. Second annual analysis of the collaborative islet transplant registry. Transplant Proc
2007; 39: 179.
24.Rickels MR, Naji A, Teff KL. Insulin sensitivity, glucose effectiveness, and free fatty acid dynamics after human islet transplantation for type 1 diabetes. J Clin Endocrinol Metab
2006; 91: 2138.
25.Sosenko JM, Palmer JP, Greenbaum CJ, et al. Increasing the accuracy of oral glucose tolerance testing and extending its application to individuals with normal glucose tolerance for the prediction of type 1 diabetes: The diabetes prevention trial-type 1. Diabetes Care
2007; 30: 38.
26.Bernuau D, Guillot R, Durand AM, et al. Ultrastructural aspects of the liver perisinusoidal space in diabetic patients with and without microangiopathy. Diabetes
1982; 31: 1061.
27.Latry P, Bioulac-Sage P, Echinard E, et al. Perisinusoidal fibrosis and basement membrane-like material in the livers of diabetic patients. Hum Pathol
1987; 18: 775.
28.Srikanta S, Ganda OP, Gleason RE, et al. Pre-type I diabetes. Linear loss of beta cell response to intravenous glucose. Diabetes
1984; 33: 717.
29.Bardet S, Rohmer V, Maugendre D, et al. Acute insulin response to intravenous glucose, glucagon and arginine in some subjects at risk for type 1 (insulin-dependent) diabetes mellitus. Diabetologia
1991; 34: 648.
30.Del Prato S, Tiengo A. The importance of first-phase insulin secretion: Implications for the therapy of type 2 diabetes mellitus. Diabetes Metab Res Rev
2001; 17: 164.
31.Poitout, V, Robertson RP. An integrated view of beta-cell dysfunction in type-II diabetes. Annu Rev Med
1996; 47: 69.
32.Brunzell JD, Robertson RP, Lerner RL, et al. Relationships between fasting plasma glucose levels and insulin secretion during intravenous glucose tolerance tests. J Clin Endocrinol Metab
1976; 42: 222.