The success of the Edmonton Protocol for islet transplantation has provided new hope in the treatment of type 1 diabetes. This study reports on the assessment of 83 human islet grafts transplanted using the Edmonton Protocol since 1999. Cellular composition, as assessed by immunohistochemistry, showed a lower islet purity (∼40%) than has been reported in previous studies using dithizone staining to quantitate islet equivalents. Furthermore, grafts were found to contain substantial populations of exocrine and ductal tissue. Total cellular insulin transplanted was 8,097.6 ± 3,164.4 μg/patient, and was significantly lower in bottom gradient layer grafts than top gradient layer or whole/combined grafts (P < 0.0005). A static incubation test for islet function gave a stimulation index of 3–4, although this measure did not correlate with posttransplant metabolic outcome. Furthermore, we confirmed a previously reported trend in which donor age affects islet yield and purity. It is important to note that a significant positive correlation was observed between the number of islet progenitor (ductal-epithelial) cells transplanted and long-term metabolic success as assessed an by intravenous glucose tolerance test at ∼2 years posttransplant. In summary, careful assessment of islet graft composition is needed in a clinical transplantation program to accurately estimate islet purity and assess the contribution of other cell types present, such as islet progenitor cells.

The Edmonton Protocol (1) and follow-up studies (2,3) have shown that islet transplantation is now a therapeutic alternative for some patients with type 1 diabetes. Despite improvements in the isolation of human islets (1,4,5), donor recoveries continue to vary. The assessment of islet equivalents and the proportion of dithizone-positive aggregates have been the standard measures used to estimate yield and purity (6,7), respectively. However, these techniques are not necessarily quantitative largely due to observer subjectivity, and more accurate methods to assess human islet grafts are needed. Pipeleers and colleagues (8,9) have shown that islet graft composition with respect to endocrine as well as nonendocrine cell types can influence long-term metabolic function of rat islet grafts, and have also characterized human islet allografts to select the best preparations with respect to endocrine content and donor-recipient compatibility (10).

Graft function is another important factor when characterizing human clinical preparations. Methods to test human islet function include transplantation in immunodeficient mice (4), assessment of insulin biosynthesis (11), in vitro perifusion (12), and static incubation (13,14). The ability to correct hyperglycemia in an animal model is the most accurate assessment; however, this requires a large number of islets and several days to produce an outcome. Of the two in vitro techniques, static incubation is the fastest and easiest way to obtain an index of insulin secretory activity.

The assessment of clinical human islet grafts in a comprehensive manner would allow the refinement of isolation protocols and the selection of optimal cell preparations for transplant based on comparisons with patient outcome measures. We report here the characterization of 83 human islet grafts transplanted into 35 recipients since the inception of the Edmonton Protocol. Preparations were analyzed for endocrine and nonendocrine cellular composition, total cellular insulin content, total cell number, and β-cell function during static incubation. In addition, graft characteristics were compared with donor age and established posttransplant patient outcome measures (2,3).

Islet isolation and transplantation.

Pancreata were removed from cadaveric donors with prior informed written consent and stored in chilled University of Wisconsin solution (range 1–15.5 h) before islet isolation. Alternatively, several organs (n = 14) were preserved using a two-layer method (University of Wisconsin Solution/perfluorocarbon) (15) for 2–3 h before isolation. Islet isolation was performed as previously described for human islets (1,5). After gradient purification, tissue was collected from the topmost islet-rich gradient layers and transplanted as one preparation. Alternatively, in the less successful purifications, islets were collected separately from both the topmost layer (designated as “top”) as well as the more dense middle layers (designated as “bottom”); these preparations were transplanted sequentially during the same procedure. Immediately after islets were isolated, duplicate samples were taken for immunohistochemistry and dithizone staining as well as insulin and DNA content. Insulin and DNA content were measured using previously described methods (16). Quantification of islet equivalents and assessment of purity were performed using a standard islet diameter of 150 μm, as previously described (7). In addition, 100–400 μl of the isolated islet preparation were cultured in CMRL 1066 supplemented with 10% FCS (Sigma-Aldrich, Oakville, Canada) and 25 mmol/l HEPES buffer (Sigma-Aldrich) for 18–24 h before islet function was assessed.

Transplant recipients were selected based on the presence of labile diabetes, hypoglycemic unawareness, or progressive complications (1). Islets were transplanted immediately or up to 12 h after being isolated, according to the protocol developed by Shapiro et al. (1). A minimum of 10,000 islet equivalents/kg of recipient body weight was given over 2–3 procedures to 35 patients. Immunosuppression under the Edmonton Protocol (n = 26) was as previously described (1). Eight patients received a modification of this regimen, whereby infliximab was given before the other medications and one patient received a protocol involving Campath-1H for induction, followed by infliximab and maintenance immunosuppression with sirolimus and ultra-low-dose tacrolimus.

Islet graft characterization.

Samples taken at the time of isolation were analyzed for cellular composition using immunohistochemistry. Islets were enzymatically dissociated into single cell preparations, as previously described (16); cell suspensions were then placed on positively charged histobond slides (Marienfeld, Lauda-Koenigshofen, Germany) and fixed in a water-based Bouin fixative for 12 min before being stored at 4°C in 70% ethanol.

Immunostaining of single cell islet samples was performed using the ABC/DAB method (Avidin-Biotin complex visualized with the chromagen diaminobenzidine), as previously described (16). Primary antibody concentrations used were 1/100 rabbit anti-human amylase (Sigma-Aldrich) to identify exocrine cells, 1/1,000 guinea pig anti-porcine insulin (Dako, Carpinteria, CA) to identify β-cells, 1/50 mouse anti-human cytokeratin-19 (Dako) to identify ductal cells, 1/1,000 rabbit anti-human somatostatin (Dako) to identify δ-cells, 1/1,000 rabbit anti-human pancreatic polypeptide (Dako) to identify Peyer’s patch (PP) cells, and 1/1,000 mouse anti-human glucagon (Sigma-Aldrich) to identify α-cells. For quantification of specific cellular phenotypes, single cell counts were performed on a minimum of 500 cells per sample and the percent was calculated as the number of positive cells per 500. To calculate the cellular mass per graft, total DNA content was divided by 6.6 pg DNA per human pancreatic cell and then, using the proportions of each phenotype obtained by immunostaining, the absolute number of each cell type was determined. Total endocrine cell purity was calculated by summing the proportions of all four endocrine cell types (β, α, δ, and PP) in each preparation.

After 18–24 h culture, islets were recombined in Hams F10 (Sigma-Aldrich) containing a low concentration of glucose (2.8 mmol/l) and washed three times by gravity sedimentation. Duplicate samples were then taken for cellular DNA and insulin content, as previously described (16), and the remaining tissue was subjected to a static incubation in Hams F10 containing 2.8 or 20 mmol/l glucose (16). Results are reported as the percent of cellular insulin content released, absolute release per 10 islet equivalents, and absolute release per individual β-cell.

Patient monitoring and metabolic testing.

Pre- and posttransplant metabolic testing was performed on islet recipients, as previously described (2,3). Intravenous glucose tolerance tests (IVGTTs) and arginine stimulation tests were performed between transplants, at 1 and 3 months posttransplant, and at 6-month intervals thereafter. The change in the acute insulin response to glucose (AIRg) from 1 month to 1 year posttransplant was obtained by calculating the difference between the values for AIRg at these two time points. Area under the curve (AUC) for insulin was calculated as previously described (3) and was the AUC above baseline over the 30-min period after infusion. Patient groups used for long-term clinical analysis were selected by the islet transplant program’s endocrinologist (Dr. Edmond A. Ryan, University of Alberta, Edmonton, Canada), who was blinded to graft function and composition data. For the success and failure groups, the mean follow-up time was 19.8 and 19.1 months, respectively. Patients in the failure-group had resumed exogenous insulin therapy at 16.9 ± 3.5 months posttransplant, whereas all patients in the success group remained insulin independent at the time of analysis.

Statistical analysis.

Data are expressed as means ± SD. The statistical significance of differences was calculated by a one-way ANOVA and the post hoc Scheffe’s test or unpaired Student’s t test. The analysis of correlations between graft function/composition and donor/isolation characteristics was by Pearson’s correlation coefficient and Fisher’s r to z test performed on Statview (SAS Institute, Cary, NC). Correlation analysis between graft composition and long-term metabolic outcome was performed using Sigma-Stat (Jandel Scientific, San Rafael, CA). Statistical significance for all tests was set at P < 0.05.

Phenotypic composition of clinical human islet grafts.

Table 1 illustrates the cellular composition and absolute cell number of 83 clinical human islet grafts. The majority of these preparations (n = 69) was evaluated and transplanted as whole preparations as multiple islet-rich gradient tissue layers were combined during the isolation procedure. Alternatively, during some isolations (n = 14), the top and bottom islet layers were collected from the Cobe processor and subsequently assessed and transplanted sequentially. In these cases, values are reported for the layers individually and as a weighted average for both combined, taking into account the total cell mass of each fraction.

Table 1 also shows that the proportion and mass of β-cells in each graft was variable, with 23.4 ± 11.9% insulin-positive cells and a mass of 147.4 ± 160.7 × 106 cells. The proportion of insulin-positive cells was slightly, but not significantly, lower than that found in top layer alone preparations (28.6 ± 15.3%), although the smaller overall cell mass of the top layer preparations resulted in a smaller absolute number of β-cells (94.2 ± 94.7 × 106). Separated bottom layer preparations, in contrast, showed a significantly lower proportion (7.1 ± 4.9%; P < 0.0001) and absolute number (21.7 ± 15.5 × 106; P < 0.05) of β-cells as compared with both top and whole preparations. When values were combined for top and bottom layers, the β-cell proportion and absolute number were not significantly different from that observed for whole preparations. Similar results were seen when comparing the remaining three islet endocrine cell types.

Grafts were found to contain a significant and variable amount of both exocrine (amylase-positive) and ductal (cytokeratin-19−positive) cells (Table 1). Bottom layer preparations were generally less pure than top layer and whole preparations, as evidenced by a significantly higher proportion of ductal (36.7 ± 15.5%; P < 0.05 vs. whole preparations) and exocrine (44.9 ± 9.4%; P < 0.005 vs. top layer preparations) cells. These differences, however, were not significant when considering the absolute numbers of ductal and exocrine cells present in the grafts due to differences in overall graft mass.

Immunostaining was also performed to examine the presence of other cell types in clinical islet grafts, such as immune cells and vascular endothelial cells. It was found that <1% of the cells expressed CD45 (n = 10) and smooth muscle cell α-actin (n = 10). For this reason, analysis of these markers was discontinued in the overall graft analysis protocol.

The total mass of each cell type implanted per patient was calculated by adding the individual values for all grafts transplanted. Patients received an average of 320.0 ± 206.5, 132.8 ± 96.5, 63.7 ± 46.0, and 76.2 ± 76.1 × 106 β-, α-, δ-, and PP-expressing cells, respectively. In comparison, a higher and more variable number of ductal and exocrine cells were received by these patients (ductal: 454.9 ± 306.7 × 106; exocrine: 754.0 ± 609.3 × 106).

Comparison of dithizone-based purity and endocrine purity as assessed by immunostaining.

Islet purity, as assessed subjectively at the time of isolation based on dithizone staining (4,6) in 82 clinical preparations, was compared with the results using immunostaining to quantitate total endocrine cellular composition (as described in research design and methods). The dithizone-based purity assessment suggested that these preparations were composed of 66.9 ± 15.8% islets, a number significantly higher than indicated by the endocrine immunostaining results for whole (41.1 ± 16.3%; P < 0.0001), top layer (46.7 ± 20.1%; P < 0.0005), and bottom layer (18.1 ± 11.0%; P < 0.0001) preparations. In addition, a significant difference was seen in total endocrine purity using immunostaining in bottom layer grafts versus both whole (P < 0.0001) and top layer (P < 0.0001) grafts.

Mass and insulin content of human islet grafts.

Clinical islet grafts were also assessed for cellular insulin and DNA content and number of islet equivalents (Table 2). Bottom layers contained a significantly lower amount of cellular insulin (P < 0.0005) than all other preparations, despite the fact that the total DNA content of top and bottom layers was similar. Thus, there was a significantly lower insulin-to-DNA ratio in the bottom layer as opposed to in the top layer preparations (P < 0.01). Furthermore, the number of islet equivalents was significantly lower in bottom layer preparations (63.6 ± 43.5 × 103; P < 0.0001) versus whole (344.2 ± 103.3 × 103) and top layer (383.5 ± 158.6 × 103) preparations, whereas top and bottom grafts combined (447.1 ± 169.3 × 103) contained a significantly higher number (P < 0.005) of islet equivalents than the whole preparations. The average total cellular insulin, DNA, and islet equivalents summed over all grafts received by a patient was 8,097.6 ± 3,164.4 μg, 12,064.1 ± 6,438.9 μg (1,827.9 ± 975.6 × 106 cells), and 815.9 ± 195.2 × 103 islet equivalents, respectively.

In vitro insulin secretory activity.

Insulin secretion from islet preparations was assessed during static incubation (14,16). The insulin secretory activity was not significantly different when comparing whole, top layer, and bottom layer preparations (Table 3). Thus, the calculated stimulation indexes were also similar. The only significant difference observed was that of insulin release per 10 islet equivalents at high glucose in bottom layer grafts (P < 0.005 vs. whole and top layer); however, because of a higher basal release at 2.8 mmol/l glucose, this did not make the stimulation index significantly different.

Effect of donor age on islet graft composition/function.

Isolations from donors age <25 years yielded fewer islets than those from donors age 25–50 or >50 years. (Table 4), although this difference did not reach statistical significance. However, fewer islet preparations from young donors (<25 years) were considered suitable for transplantation (20%) than those from donors age 25–50 (38%) or >50 (39%) years. No significant differences were found in endocrine, ductal, or exocrine mass or cellular proportion regardless of the donor’s age-group (Table 4). Similarly, although the islet stimulation index was lower in preparations from older donors and the insulin content per β-cell was lower in younger donors, neither of these differences reached statistical significance due to the large degree of variation within these age-groups.

Effect of graft composition on long-term clinical outcome.

We compared several established patient outcome measures, including insulin independence and insulin response to arginine or glucose challenge, to islet graft composition and glucose-stimulated insulin secretion in an attempt to define a predictor(s) of clinical islet transplant success.

We assessed whether the total number of β-cells transplanted could provide a prediction of short-term metabolic outcome. Figure 1 shows a comparison of β-cell number transplanted and both the acute insulin response to arginine (AIRarg) and AUC for insulin (AUCins) for the same patient group as was assessed in a previous report with respect to the number of islet equivalents transplanted (3). Although β-cell number versus AIRarg showed a positive correlation approaching statistical significance (r = 0.609; P = 0.08) (Fig. 1A), neither metabolic measure was strongly associated with absolute β-cell number transplanted.

We also attempted to define a correlation between graft cellular composition and long-term transplant success. Two groups of patients transplanted >19 months earlier were selected based on whether they remained insulin independent (n = 6) or required subsequent insulin injections posttransplant (n = 5). These groups were chosen by an endocrinologist blinded to any graft assessment data; the average time posttransplant did not differ between these groups (success group 19.8 months and failure group 19.1 months). One difference observed between these groups was the absolute number of ductal-epithelial cells (cytokeratin-19 positive) transplanted. Patients in the success group were implanted with a total of 500.5 ± 247.5 × 106 (median 435.3 × 106) ductal cells, whereas those in the failure group received 277.3 ± 224.3 × 106 (median 232.3 × 106) cells (P = 0.093). When assessed as success and failure groups, no difference was seen between the mean number of β-cells transplanted between the groups (276.0 ± 58.2 × 106 in success group, 421.2 ± 400.2 × 106 in failure group; P = 0.40).

In addition to this analysis, when examining a larger group of patients individually (n = 19), a statistically significant positive correlation was observed between the number of ductal cells transplanted and AIRg at ∼2 years posttransplant (r = 0.490, P < 0.05) (Fig. 2D). In this same group of patients, no significant correlation was observed between the number of ductal cells transplanted and AIRg at only 3 months posttransplant (r = 0.421, P = 0.118) (Fig. 2C) or between the number of β-cells (Fig. 2A and B) or exocrine cells (Fig. 2E and F) transplanted and AIRg at 3 months and 2 years posttransplant. Furthermore, no significant correlation was observed between the total islet equivalents or total pack cell volume transplanted and AIRg at 2 years posttransplant (data not shown). A significant positive correlation was seen between the number of ductal cells transplanted and the change in AIRg from 1 month to 1 year posttransplant (r = 0.513, P < 0.05) (Fig. 3).

The posttransplant metabolic outcome was also compared between patients receiving immunosuppression under the Edmonton Protocol and those receiving a modified infliximab protocol. No significant difference in AIRg at 3 months or 2 years posttransplant was seen between these patient groups.

Although methods exist to subjectively assess islet mass (i.e., islet equivalents) and graft purity, few researchers are rigorously assessing the cellular composition and function of these grafts. We have reported here data from 83 human islet preparations transplanted into 35 patients beginning with the inception of the Edmonton Protocol. These data were compared with donor age and long-term posttransplant metabolic function in recipients to define graft-dependent predictor(s) of subsequent isolation and transplant success.

In our analysis, dissociated cell preparations were used as opposed to staining of whole fixed cellular aggregates (17,18) to more accurately assess the proportion of each cell type. Although, based on dithizone staining, human islet preparations are routinely estimated to be >50% in purity (19), and often approach 80–90% (4) purity, our results demonstrated this to be an overestimation. In fact, the difference in overall endocrine purity of our preparations as indicated by immunostaining versus that indicated by dithizone-based estimations was highly significant. Our results indicated 50% endocrine purity to be an exceptional isolation outcome, even for separated top layer−only preparations. Furthermore, our findings indicated an approximate endocrine composition of 60% β-cells, 23% α-cells, 10% δ-cells, and 10% PP-cells. Although PP cells were more abundant than α- or δ-cells in bottom layer preparations, overall they were less common (this result was considered to be an anomaly of the specific bottom layer preparations used in this study).

Analysis of cellular composition showed that islet grafts were composed of mostly nonendocrine cells. Using cytokeratin-19 as a ductal cell marker (21,22), it was observed that a significant proportion of ductal cells (20–30%) were present. Furthermore, a large proportion (20–50%) of these preparations consisted of amylase-positive exocrine cells. As expected, the less pure bottom layer preparations contained a significantly higher proportion of both ductal and exocrine cells. Correspondingly, these preparations also had a significantly lower proportion of endocrine cells than the purer top layer preparations. These findings underscore the need to determine the positive and negative effects of having nonislet tissue present in clinical islet grafts, including the possible presence of islet progenitor cells that could contribute to new β-cell mass through differentiation posttransplant.

The measurement of total cellular insulin content and total cell number in islet grafts can also give an estimate of purity. Previous studies have suggested the total cellular insulin content of the adult pancreas to be ∼140 μg/g tissue (23). Assuming an average pancreas weight of 90 g, this translates to ∼12,600 μg total insulin. We obtained an average of 3,000–4,000 μg insulin in each islet preparation. Therefore, we either recovered only 25–30% of the β-cell mass or β-cell degranulation occurred during the isolation process. Our patients received a total of ∼8,000 μg of insulin, suggesting that even after undergoing 2–3 transplant procedures, islet recipients receive only ∼65% of the cellular insulin content present in a normal pancreas. Furthermore, each patient received an average of 320 × 106 β-cells over multiple transplants, with an average of 815,000 islet equivalents, thereby suggesting an average β-cell number of 400 per islet equivalent.

Glucose-stimulated insulin secretion is an important predictor of islet function. Calculating insulin secretory activity as the percent of the total cellular insulin content released, absolute insulin release per 10 islet equivalents and absolute release per individual β-cell gave similar stimulation indexes in the range of 3–4. This compares with a fivefold average increase reported by Ricordi et al. (4) using the perifusion method and is in agreement with an earlier study by Grant et al. (24), who also reported a fivefold stimulation index for human islets using a static incubation. Our observations, however, suggest that static incubation data are not necessarily a good predictor of clinical outcome, as no correlations were seen with posttransplant metabolic success. The fact that some successful patients receive islets with poor in vitro insulin release (i.e., stimulation index <1) suggests that functional recovery occurs posttransplant or that static incubation is not a representative test of islet function in vivo. Other tests of function should be evaluated in the future, including perifusion to show a more dynamic biphasic insulin release profile, assessment of insulin biosynthesis, and animal transplantation to test actual in vivo function of grafted islets.

It has been proposed that islets derived from younger donors exhibit metabolic advantages for clinical transplantation (23,25). A younger pancreas contains more insulin per gram of tissue as well as a lower concentration of exocrine digestive enzymes, thus limiting islet destruction by enzymes released from necrotic exocrine tissue during and after isolation (23). However, isolations from donors age <30 years are technically difficult and yield fewer islets than isolations from older donors (25,26). We have shown a similar trend to that previously reported (25,26), where isolations from donors age <25 years yielded fewer islet equivalents, although in contrast to the other studies, our results did not reach statistical significance. Islet β-cell mass was also examined with respect to donor age and a trend was observed for younger donors to exhibit a lower insulin content per β-cell, lower number of islet equivalents recovered per gram of pancreas, and a higher stimulation index than donors age >50 years. It should be noted that a smaller percentage of islet isolations from young donors were considered suitable for transplantation (20 vs. 38% of donors age 25–50 years and 39% of donors age >50 years). For this reason, the results of this study may be skewed toward indicating less of a difference among age groups, as only the best preparations from younger donor isolations were included in the clinical dataset.

Variability in graft cellular composition may be useful as a predictor of clinical transplant outcome. Our group has previously reported (2,3) that the AIRg and AIRarg indicate both graft mass and function posttransplant. Furthermore, Ryan et al. (3) have reported a correlation between the number of islet equivalents transplanted and the metabolic measures AIRg and AUCins. We examined whether the same metabolic measures correlated with the total number of β-cells transplanted in the same patients as the Ryan et al. study (3). A correlation was found between AIRg and β-cell number transplanted that approached statistical significance (r = 0.609; P = 0.08); however, there was no correlation between AUCins and number of β-cells transplanted. The reason for this discrepancy with the previous study is likely the inaccuracy of the method for counting islet equivalents as compared with immunostaining and quantification of dissociated cell samples.

Several patients under the Edmonton Protocol have remained insulin independent for >3 years (unpublished data); however, in some cases, graft failure not associated with immunorejection has necessitated a return to insulin therapy. Finegood et al. (27) used a mathematical model to predict an average β-cell life of 1–3 months and proposed a dynamic system of β-cell turnover in the adult pancreas. This suggests that islet cell turnover may also occur in clinical grafts to sustain long-term function after transplant. We examined whether the number of ductal-epithelial cells (widely believed to be islet progenitor cells) (28,29) transplanted influences long-term clinical outcome. A difference nearing statistical significance was found between success and failure groups with regard to the total number of cytokeratin-positive ductal cells received and AIRg at ∼2 years posttransplant. Furthermore, a significant positive correlation was observed in a larger group of patients between the number of ductal cells received and both AIRg at 2 years posttransplant and the change in AIRg from 1 month to 1 year posttransplant. This suggests that a progenitor cell population in human islet grafts may play a role in β-cell turnover and increase long-term graft survival. However, given recent data from Dor et al. (30) that indicate that β-cell renewal in adults may not originate from islet progenitors, it is possible that ductal cells influence graft survival through another mechanism. For example, ductal cells may secrete factors that exert a protective effect on existing islet cells in the graft or facilitate improved re-vascularization posttransplant. At this time, however, significant evidence still suggests that ductal cells are islet progenitors (28,29,31,32) and are the most likely mechanism for the effect on long-term transplant outcome.

In conclusion, we have provided a summary of human islet grafts with respect to cellular composition, mass, and function. A discrepancy has been shown between purity estimated by dithizone staining and the more quantitative immunostaining of representative dissociated cell samples. In addition, we have shown that islet graft assessment may provide information about subsequent transplant outcome in that the presence of ductal cells may improve long-term metabolic outcome. It is important to note, however, that significant variability exists in islet isolation success, graft characteristics, and patient outcome. Thus, it remains difficult to accurately define characteristics of donors or graft composition that affect clinical outcome. However, with the improvement and standardization of islet isolation techniques, detailed information collected on human islet grafts in the future will allow the refinement of clinical islet transplantation procedures to build on the success of the Edmonton Protocol.

FIG. 1.

Comparison of total β-cell number transplanted and two posttransplant metabolic outcome measures. A: Relation of total number of β-cells transplanted and AIRarg as derived from intravenous arginine infusion test at midtransplant (n = 6) and 3 months posttransplant (n = 3). B: Relation of AUCins and the number of β-cells transplanted as derived from IVGTT at midtransplant (n = 13) and 3 months posttransplant (n = 13).

FIG. 1.

Comparison of total β-cell number transplanted and two posttransplant metabolic outcome measures. A: Relation of total number of β-cells transplanted and AIRarg as derived from intravenous arginine infusion test at midtransplant (n = 6) and 3 months posttransplant (n = 3). B: Relation of AUCins and the number of β-cells transplanted as derived from IVGTT at midtransplant (n = 13) and 3 months posttransplant (n = 13).

FIG. 2.

Comparison of graft composition to clinical transplant outcome. Relation of β-cell (A and B), ductal cell (C and D), and exocrine cell (E and F) number transplanted to AIRg at 3 months and 2 years posttransplant.

FIG. 2.

Comparison of graft composition to clinical transplant outcome. Relation of β-cell (A and B), ductal cell (C and D), and exocrine cell (E and F) number transplanted to AIRg at 3 months and 2 years posttransplant.

FIG. 3.

Relation of change in AIRg from 1 month to 1 year posttransplant (n = 17) with total number of ductal cells transplanted in patients receiving islet grafts using the Edmonton Protocol.

FIG. 3.

Relation of change in AIRg from 1 month to 1 year posttransplant (n = 17) with total number of ductal cells transplanted in patients receiving islet grafts using the Edmonton Protocol.

TABLE 1

Cellular composition of clinically transplanted human islet grafts

nPhenotype
βαδPPDuctalAcinar
Composition (%)        
    Whole preparations 69 23.4 ± 11.9 9.2 ± 6.2 4.0 ± 1.9 4.2 ± 2.9 23.8 ± 11.9 34.7 ± 16.1 
    Separated preparations 14 — — — — — — 
        Top — 28.6 ± 15.3 10.4 ± 5.8 3.9 ± 1.6 3.8 ± 2.9 29.9 ± 14.8 23.2 ± 13.9 
        Bottom — 7.1 ± 4.9* 3.9 ± 3.5 2.1 ± 2.5 5.4 ± 3.5 36.7 ± 15.5 44.9 ± 9.4§ 
        Top/bottom combined — 17.7 ± 9.9 7.2 ± 4.5 3.2 ± 1.7 4.5 ± 2.4 33.5 ± 14.1 33.6 ± 9.7 
Cell mass/graft (×106       
    Whole preparations 69 147.4 ± 160.7 61.5 ± 69.6 29.9 ± 33.3 32.8 ± 41.4 192.6 ± 217.9 346.3 ± 492.1 
    Separated preparations 14 — — — — — — 
        Top — 94.2 ± 94.7 36.2 ± 50.7 14.0 ± 14.9 13.8 ± 14.9 107.9 ± 97.4 92.7 ± 92.4 
        Bottom — 21.7 ± 15.5 11.8 ± 13.7 5.2 ± 3.0 17.4 ± 14.8 129.1 ± 103.7 161.0 ± 111.5 
        Top/bottom combined — 115.9 ± 104.7 48.0 ± 62.1 20.4 ± 17.2 31.2 ± 23.5 237.0 ± 179.9 253.6 ± 196.6 
Cell mass/patient (×10635 320.0 ± 206.5 132.8 ± 96.5 63.7 ± 46.0 76.2 ± 76.1 454.9 ± 306.7 754.0 ± 609.3 
nPhenotype
βαδPPDuctalAcinar
Composition (%)        
    Whole preparations 69 23.4 ± 11.9 9.2 ± 6.2 4.0 ± 1.9 4.2 ± 2.9 23.8 ± 11.9 34.7 ± 16.1 
    Separated preparations 14 — — — — — — 
        Top — 28.6 ± 15.3 10.4 ± 5.8 3.9 ± 1.6 3.8 ± 2.9 29.9 ± 14.8 23.2 ± 13.9 
        Bottom — 7.1 ± 4.9* 3.9 ± 3.5 2.1 ± 2.5 5.4 ± 3.5 36.7 ± 15.5 44.9 ± 9.4§ 
        Top/bottom combined — 17.7 ± 9.9 7.2 ± 4.5 3.2 ± 1.7 4.5 ± 2.4 33.5 ± 14.1 33.6 ± 9.7 
Cell mass/graft (×106       
    Whole preparations 69 147.4 ± 160.7 61.5 ± 69.6 29.9 ± 33.3 32.8 ± 41.4 192.6 ± 217.9 346.3 ± 492.1 
    Separated preparations 14 — — — — — — 
        Top — 94.2 ± 94.7 36.2 ± 50.7 14.0 ± 14.9 13.8 ± 14.9 107.9 ± 97.4 92.7 ± 92.4 
        Bottom — 21.7 ± 15.5 11.8 ± 13.7 5.2 ± 3.0 17.4 ± 14.8 129.1 ± 103.7 161.0 ± 111.5 
        Top/bottom combined — 115.9 ± 104.7 48.0 ± 62.1 20.4 ± 17.2 31.2 ± 23.5 237.0 ± 179.9 253.6 ± 196.6 
Cell mass/patient (×10635 320.0 ± 206.5 132.8 ± 96.5 63.7 ± 46.0 76.2 ± 76.1 454.9 ± 306.7 754.0 ± 609.3 

Data are means ± SD of the given number of different human islet preparations and represent the percent of the total cell composition or the absolute number of cells present in each graft. Whole preparations identify grafts where all gradient tissue layers were collected, transplanted, and evaluated as a single preparation. Top and bottom layer grafts identify preparations where top and bottom gradient layers from one islet isolation were collected, transplanted, and evaluated separately. Percentages of combined preparations were calculated by weighting the contribution (in total cell number) of top and bottom from the same donor.

*

P < 0.0001 vs. top and whole preparations;

P < 0.05 vs. top and whole preparations;

P < 0.05 vs. whole preparations;

§

P < 0.005 vs. top preparations.

TABLE 2

Composition of clinical human islet grafts

nInsulin (μg)DNA (μg)Insulin/DNATotal cells (×106)Islet equivalents (×103)
Whole preparations 65 3,325.8 ± 1,915.6 5,426.5 ± 5,608.7 1.3 ± 1.3 822.2 ± 849.8 344.2 ± 103.3 
Separated preparations 14 — — — — — 
    Top — 2,930.7 ± 1,740.7 2,370.1 ± 1,988.2 2.0 ± 1.5 359.1 ± 301.2 383.5 ± 158.6 
    Bottom — 452.2 ± 230.0* 2,298.6 ± 1,384.8 0.5 ± 0.5 348.3 ± 209.8 63.6 ± 43.5 
    Top/bottom combined — 3,382.9 ± 1,843.1 4,668.7 ± 3,166.6 1.3 ± 1.0 707.4 ± 479.8 447.1 ± 169.3§ 
Total per patient 35 8,097.6 ± 3,164.4 12,064.1 ± 6,438.9 0.9 ± 0.5 1,827.9 ± 975.6 815.9 ± 195.2 
nInsulin (μg)DNA (μg)Insulin/DNATotal cells (×106)Islet equivalents (×103)
Whole preparations 65 3,325.8 ± 1,915.6 5,426.5 ± 5,608.7 1.3 ± 1.3 822.2 ± 849.8 344.2 ± 103.3 
Separated preparations 14 — — — — — 
    Top — 2,930.7 ± 1,740.7 2,370.1 ± 1,988.2 2.0 ± 1.5 359.1 ± 301.2 383.5 ± 158.6 
    Bottom — 452.2 ± 230.0* 2,298.6 ± 1,384.8 0.5 ± 0.5 348.3 ± 209.8 63.6 ± 43.5 
    Top/bottom combined — 3,382.9 ± 1,843.1 4,668.7 ± 3,166.6 1.3 ± 1.0 707.4 ± 479.8 447.1 ± 169.3§ 
Total per patient 35 8,097.6 ± 3,164.4 12,064.1 ± 6,438.9 0.9 ± 0.5 1,827.9 ± 975.6 815.9 ± 195.2 

Data are means ± SD of the given number of different human islet preparations or patients and are expressed as absolute numbers or relative ratios. Whole preparations identify grafts where all gradient tissue layers were collected, transplanted, and evaluated as a single preparation. Top and bottom layer grafts identify preparations where top and bottom gradient layers from one islet isolation were collected, transplanted, and evaluated separately.

*

P < 0.0005 vs. whole, top layer, and top/bottom combined preparations;

P < 0.01 vs. top layer preparations;

P < 0.0001 vs. whole, top layer, and top/bottom combined preparations;

§

P < 0.005 vs. whole preparations.

TABLE 3

Insulin secretory activity of clinical islet preparations as assessed by static incubation

n2.8 mmol/l glucose20 mmol/l glucoseStimulation index
Cellular insulin content (%)     
    Whole preparations 66 4.3 ± 5.1 11.2 ± 7.7 4.0 ± 3.9 
    Top preparations 14 4.5 ± 3.2 12.2 ± 6.9 3.1 ± 1.2 
    Bottom preparations 14 3.1 ± 3.1 10.5 ± 6.7 4.1 ± 2.3 
Insulin release per 10 islet equivalents (ng)     
    Whole preparations 59 2.5 ± 2.5 5.8 ± 4.2 3.7 ± 4.0 
    Top preparations 14 2.6 ± 2.1 7.4 ± 5.1 3.1 ± 1.2 
    Bottom preparations 14 3.2 ± 2.9 10.9 ± 7.1* 4.1 ± 2.3 
Insulin release per β-cell (pg)     
    Whole preparations 31 1.5 ± 1.3 3.6 ± 2.6 3.2 ± 3.4 
    Top preparations 12 1.2 ± 0.9 3.2 ± 2.5 3.1 ± 1.3 
    Bottom preparations 12 0.8 ± 0.3 3.0 ± 1.5 4.1 ± 2.3 
n2.8 mmol/l glucose20 mmol/l glucoseStimulation index
Cellular insulin content (%)     
    Whole preparations 66 4.3 ± 5.1 11.2 ± 7.7 4.0 ± 3.9 
    Top preparations 14 4.5 ± 3.2 12.2 ± 6.9 3.1 ± 1.2 
    Bottom preparations 14 3.1 ± 3.1 10.5 ± 6.7 4.1 ± 2.3 
Insulin release per 10 islet equivalents (ng)     
    Whole preparations 59 2.5 ± 2.5 5.8 ± 4.2 3.7 ± 4.0 
    Top preparations 14 2.6 ± 2.1 7.4 ± 5.1 3.1 ± 1.2 
    Bottom preparations 14 3.2 ± 2.9 10.9 ± 7.1* 4.1 ± 2.3 
Insulin release per β-cell (pg)     
    Whole preparations 31 1.5 ± 1.3 3.6 ± 2.6 3.2 ± 3.4 
    Top preparations 12 1.2 ± 0.9 3.2 ± 2.5 3.1 ± 1.3 
    Bottom preparations 12 0.8 ± 0.3 3.0 ± 1.5 4.1 ± 2.3 

Data are means ± SD. Values for insulin release are expressed as the percent of total cellular insulin content released, nanograms per islet equivalent, or picograms per individual β-cell. The stimulation index is expressed as the ratio of insulin release at 20 mmol/l glucose to insulin release at 2.8 mmol/l glucose.

*

P < 0.005 vs. release per 10 islet equivalents in whole preparations. Three preparations did not contain a sufficient tissue volume after 18–24 h culture to accurately measure function and thus were not included in this assessment.

TABLE 4

Effect of donor age on human islet graft composition and function

Age <25 yearsAge 25−50 yearsAge >50 years
n 10 31 23 
Islet equivalents (×103290.0 ± 58.9 339.6 ± 104.9 362.6 ± 99.5 
Endocrine cell mass (×106251.3 ± 197.4 267.2 ± 219.5 289.7 ± 368.8 
Endocrine cell proportion (%) 40.3 ± 15.8 42.8 ± 15.4 40.3 ± 17.1 
Ductal cell mass (×106185.8 ± 188.1 193.8 ± 255.0 184.7 ± 181.6 
Ductal cell proportion (%) 21.6 ± 7.6 23.3 ± 13.2 25.2 ± 12.2 
Exocrine cell mass (×106452.4 ± 578.9 379.5 ± 555.0 245.4 ± 357.3 
Exocrine cell proportion (%) 38.7 ± 18.3 35.3 ± 17.5 31.7 ± 13.2 
Stimulation index 4.4 ± 6.7 4.8 ± 4.1 2.8 ± 1.4 
Insulin content per β-cell (pg) 23.9 ± 15.9 34.8 ± 25.3 35.1 ± 25.4 
Age <25 yearsAge 25−50 yearsAge >50 years
n 10 31 23 
Islet equivalents (×103290.0 ± 58.9 339.6 ± 104.9 362.6 ± 99.5 
Endocrine cell mass (×106251.3 ± 197.4 267.2 ± 219.5 289.7 ± 368.8 
Endocrine cell proportion (%) 40.3 ± 15.8 42.8 ± 15.4 40.3 ± 17.1 
Ductal cell mass (×106185.8 ± 188.1 193.8 ± 255.0 184.7 ± 181.6 
Ductal cell proportion (%) 21.6 ± 7.6 23.3 ± 13.2 25.2 ± 12.2 
Exocrine cell mass (×106452.4 ± 578.9 379.5 ± 555.0 245.4 ± 357.3 
Exocrine cell proportion (%) 38.7 ± 18.3 35.3 ± 17.5 31.7 ± 13.2 
Stimulation index 4.4 ± 6.7 4.8 ± 4.1 2.8 ± 1.4 
Insulin content per β-cell (pg) 23.9 ± 15.9 34.8 ± 25.3 35.1 ± 25.4 

Data are means ± SD and represent percent, absolute cell numbers, or ratios for the given number of human islet grafts derived from donors aged <25, 25–50, and >50 years.

C.N.S. is supported in part by a grant from the Stem Cell Network (Canadian Centres of Excellence). J.R.T.L. and G.S.K. are recipients of scholarships from the Canadian Diabetes Association as well as the Alberta Heritage Foundation for Medical Research. G.S.K. also has a Career Development Award from the Juvenile Diabetes Research Foundation International. A.M.J.S. is a clinical investigator of the Alberta Heritage Foundation for Medical Research and holds the Clinical Research Chair in Transplantation (CIHR/Wyeth). Funding for this project was provided by grants from the Alberta Foundation for Diabetes Research, Juvenile Diabetes Research Foundation, and the Canadian Institute for Health Research.

The authors thank Dawne Colwell for assistance with figures and formatting as well as Jen O’Hara, Lynette Elder, Tanya Binette, Monique Tourand, and Crystal Harris for technical assistance and data collection. They also thank the staff of the Clinical Islet Laboratory (University of Alberta) for the isolation of pancreatic islets as well as the H.O.P.E. program for the identification/procurement of organs from cadaveric donors.

1.
Shapiro AMJ, Lakey JRT, Ryan EA, Korbutt GS, Toth EL, Warnock GL, Kneteman NM, Rajotte RV: Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid free immunosuppressive regimen.
N Engl J Med
343
:
230
–238,
2000
2.
Ryan EA, Lakey JRT, Rajotte RV, Korbutt GS, Kin T, Imes S, Rabinovitch A, Elliot JF, Bigam D, Kneteman NM, Warnock GL, Larsen I, Shapiro AMJ: Clinical outcomes and insulin secretion after islet transplantation with the Edmonton Protocol.
Diabetes
50
:
710
–719,
2001
3.
Ryan EA, Lakey JRT, Paty BW, Imes S, Korbutt GS, Kneteman NM, Bigam D, Rajotte RV, Shapiro AMJ: Successful islet transplantation: continued insulin reserve provides long-term glycemic control.
Diabetes
51
:
2148
–2157,
2002
4.
Ricordi C, Lacy PE, Scharp DW: Automated islet isolation from human pancreas.
Diabetes
38 (Suppl. 1)
:
140
–142,
1989
5.
Lakey JRT, Warnock GL, Shapiro AMJ, Korbutt GS, Ao Z, Kneteman NM, Rajotte RV: Intraductal collagenase delivery into the human pancreas using syringe loading or controlled perfusion.
Cell Transplant
8
:
285
–292,
1999
6.
Ricordi C: Quantitative and qualitative standards for islet isolation assessment in humans and large mammals.
Pancreas
6
:
242
–244,
1991
7.
Ricordi C, Gray DW, Hering BJ, Kaufman DB, Warnock GL, Kneteman NM, Lake SP, London NJ, Socci C, Alejandro R, Zeng Y, Scharp DW, Viviani G, Falqui L, Tzakis A, Bretzel RG, Federlin K, Pozza G, James RF, Rajotte RV, Di Carlo V, Morris PJ, Sutherland DE, Starzl TE, Mintz DH, Lacy PE: Islet isolation assessment in man and large animals.
Acta Diabetol
27
:
185
–195,
1990
8.
Keymeulen B, Anselmo J, Pipeleers D: The length of metabolic normalization after rat islet cell transplantation depends on endocrine cell composition of graft and on donor age.
Diabetologia
40
:
1152
–1158,
1997
9.
Keymeulen B, Korbutt GS, De Paepe M, Gorus F, Kloppel G, Pipeleers DG: Long-term metabolic control by rat islet grafts depends on the composition of the implant.
Diabetes
45
:
1814
–1821,
1996
10.
Keymeulen B, Ling Z, Gorus FK, Delvaux G, Bouwens L, Grupping A, Hendrieckz C, Pipeleers-Marichal M, Van Schravendijk C, Salmela K, Pipeleers DG: Implantation of standardized beta-cell grafts in a liver segment of IDDM patients: graft and recipient characteristics in two cases of insulin independence under maintenance immunosuppression for prior kidney graft.
Diabetologia
41
:
452
–459,
1998
11.
Pipeleers DG: Composition of islet grafts.
Diabetes Nutr Metab
5 (Suppl. 1)
:
29
–32,
1992
12.
Lacy PE, Finke EH, Conant S, Naber S: Long-term perifusion of isolated rat islets in vitro.
Diabetes
25
:
484
–493,
1976
13.
Andersson A, Borg H, Groth CG, Gunnarsson R, Hellerstrom C, Lundgren G, Westman J, Ostman J: Survival of isolated human islets of Langerhans maintained in tissue culture.
J Clin Invest
57
:
1295
–1301,
1976
14.
Gray DWR, McShane P, Grant A, Morris PJ: A method for isolation of islets of Langerhans from the human pancreas.
Diabetes
33
:
1055
–1061,
1984
15.
Lakey JRT, Tsujimura T, Shapiro AMJ, Kuroda Y: Preservation of the human pancreas before islet isolation using a two layer (UW solution-perfluorochemical) cold storage method.
Transplantation
74
:
1809
–1811,
2002
16.
Korbutt GS, Elliot JF, Ao Z, Smith DK, Warnock GL, Rajotte RV: Large scale isolation, growth, and function of neonatal porcine islets.
J Clin Invest
97
:
2119
–2129,
1996
17.
Latif ZA, Noel J, Alejandro R: A simple method of staining fresh and cultured islets.
Transplantation
45
:
827
–830,
1988
18.
Morgan CR, Lazarow A: Immunoassay of insulin: antibody system. Plasma insulin levels of normal, subdiabetic, and diabetic rats.
Diabetes
12
:
115
–126,
1963
19.
Lakey JRT, Rajotte RV, Warnock GL, Kneteman NM: Human pancreas preservation prior to islet isolation.
Transplantation
59
:
689
–694,
1995
21.
Bouwens L, Braet B, Heimberg H: Identification of rat pancreatic duct cells by their expression of cytokeratins 7, 19, and 20 in vivo and after isolation and culture.
J Histochem Cytochem
43
:
245
–253,
1995
22.
Bouwens L: Cytokeratins and cell differentiation in the pancreas.
J Pathol
184
:
234
–239,
1998
23.
Sutherland DER, Matas AJ, Steffes MW, Najarian AS: Infant human pancreas: a potential source of islet tissue for transplantation.
Diabetes
25
:
1123
–1128,
1976
24.
Grant AM, Christie MR, Ashcroft SJH: Insulin release from human pancreatic islets in vitro.
Diabetologia
19
:
114
–117,
1980
25.
Socci C, Davalli AM, Vignali A, Pontiroli AE, Maffi P, Magistretti P, Gavazzi F, De Nittis P, Di Carlo V, Pozza G: A significant increase of islet yield by early injection of collagenase into the pancreatic duct of young donors.
Transplantation
55
:
661
–663,
1993
26.
Lakey JRT, Warnock GL, Rajotte RV, Suarez-Almazor ME, Ao Z, Shapiro AMJ, Kneteman NM: Variables in organ donors that affect the recovery of human islets of Langerhans.
Transplantation
61
:
1047
–1053,
1996
27.
Finegood DT, Scaglia L, Bonner-Weir S: Dynamics of β-cell mass in the growing rat pancreas: estimation with a simple mathematical model.
Diabetes
44
:
249
–256,
1995
28.
Bonner-Weir S, Taneja M, Weir GC, Tatarkiewicz K, Song K, Sharma A, O’Neil JJ: In vitro cultivation of human islets from expanded ductal tissue.
Proc Natl Acad Sci U S A
97
:
7999
–8004,
2000
29.
Bouwens L, Pipeleers DG: Extra-insular beta cells associated with ductules are frequent in adult human pancreas.
Diabetologia
41
:
629
–633,
1998
30.
Dor Y, Brown J, Martinez OI, Melton DA: Adult pancreatic β-cells are formed by self-duplication rather than stem cell differentiation.
Nature
429
:
41
–46,
2004
31.
Heremans Y, Van den Casteele M, Veld P, Gradwohl G, Serup P, Madsen O, Pipeleers D, Heimberg H: Recapitulation of embryonic neuroendocrine differentiation in adult human pancreatic duct cells expressing neurogenin 3.
J Cell Biol
159
:
303
–311,
2002
32.
Dudek RW, Lawrence IE, Hill RS, Johnson RC: Induction of islet cytodifferentiation by fetal mesenchyme in adult pancreatic ductal epithelium.
Diabetes
40
:
1041
–1048,
1991