Evidence for Instant Blood-Mediated Inflammatory Reaction in Clinical Autologous Islet Transplantation

Authors


Abstract

A nonspecific inflammatory and thrombotic reaction termed instant blood-mediated inflammatory reaction (IBMIR) has been reported when allogenic or xenogenic islets come into contact with blood. This reaction is known to cause significant loss of transplanted islets. We hypothesized that IBMIR occurs in patients undergoing total pancreatectomy followed by autologous islet transplantation (TP-AIT) and tested this hypothesis in 24 patients and in an in vitro model. Blood samples drawn during the peritransplant period showed a significant and rapid increase of thrombin–anti-thrombin III complex (TAT) and C-peptide during islet infusion, which persisted for up to 3 h, along with a decreased platelet count. A concomitant increase in levels of inflammatory proteins IL-6, IL-8 and interferon-inducible protein-10 was observed. An in vitro model composed of pure islets plus autologous blood also demonstrated significantly increased levels of TAT (p < 0.05), C-peptide (p < 0.05), tumor necrosis factor-alpha (p < 0.05) and MCP-1 (p < 0.05), as well as strong tissue factor expression in islets. Islet viability decreased significantly but was rescued by the presence of low-molecular-weight dextran sulfate. In conclusion, AIT-induced elevation of TAT and destruction of islets suggests that IBMIR might occur during AIT. Modulating this process may help improve islet engraftment and the insulin independence rate in TP-AIT patients.

Abbreviations
AIT

autologous islet transplantation

IBMIR

instant blood-mediated inflammatory reaction

IEQ

islet equivalents

IL

interleukin

IP-10

interferon-inducible protein-10

LMWDS

low-molecular-weight dextran sulfate

TAT

thrombin/anti-thrombin III complex

TF

tissue factor

TNF-α

tumor-necrosis factor-alpha

TP

total pancreatectomy

Introduction

Pancreatic islet cell transplantation is an effective beta cell replacement therapy for diabetic patients who lack the ability to produce insulin [1]. Allogeneic islet cell transplantation has met with greater success in recent years; however, pancreases from multiple donors are still required to achieve sufficient engraftment of transplanted islet mass for insulin independence [2]. Poor engraftment of transplanted islets has been attributed to several factors, including islet quality, transplant site, hypoxia and a nonspecific inflammatory and thrombotic reaction by the innate immune system known as instant blood-mediated inflammatory reaction (IBMIR) [3, 4]. Major characteristics of IBMIR include coagulation, complement activation and inflammatory cell infiltration [3]. Recent reports have shown that strong induction immunosuppressive therapy combined with the anti-inflammatory drug etanercept, a soluble tumor necrosis factor-alpha (TNF-α) blocker, are crucial to improve engraftment and the long-term function of allogenic islet transplants [5, 6]. IBMIR has also been demonstrated in xenogenic islet transplant models [7].

We hypothesized that incompatibility between isolated pancreatic islets and blood also exists in the setting of autologous islet transplantation (AIT) and that a significant mass of infused islets could be lost during the peritransplant period due to IBMIR. AIT is used in combination with total pancreatectomy (TP); TP is performed primarily to treat refractory chronic pancreatitis both in adults and in children [8], as well as to treat benign pancreatic tumors [9] and severe pancreatic trauma [10-12]. During the procedure, preservation of pancreatic endocrine function is important to prevent surgical diabetes. Previous reports have demonstrated that the insulin-independence rate after AIT is dependent on the mass of transplanted islets [13], and after intrahepatic infusion autologous islets survive better than allogenic islets [12]. Despite the improvements made in islet isolation techniques [14] and the absence of an autoimmune response against the islets, achievement of insulin independence among AIT patients is rather poor at about 40% [15].

Previous reports from our center have identified patient characteristics and an islet isolation technique associated with improved islet transplant outcomes in AIT [16, 17]. Further, we have demonstrated the adverse impact of tissue volume on AIT outcome [18]. Importantly, analysis of serum samples collected during the peritransplant period after AIT showed increased release of high-mobility group box 1 protein, which is a marker for islet damage [19]. We sought to determine the occurrence of IBMIR after clinical AIT based on markers for coagulation, complement activation and proinflammatory proteins. We also developed a “miniature tube” model for in-depth analysis of events that occur during IBMIR against autologous islets.

Materials and Methods

Clinical study

A total of 24 consecutive patients with chronic pancreatitis underwent TP-AIT from October 2010 to February 2013 at Baylor University Medical Center (Dallas, TX). The patients were divided into two groups. First group included three patients who received <1000 islet equivalents (IEQ) per kg of body weight and the second group included 21 patients who received >1000 IEQ/kg of body weight. This study was approved by the institutional review board, and written informed consent was obtained from all patients included in this study.

After TP, the removed pancreases were preserved and processed for islet isolations as described earlier [15]. Liberase MTF with thermolysin (Roche Diagnostics, Indianapolis, IN) was infused into the main pancreatic duct. Pancreas digestion was performed using the modified Ricordi method [20, 21]. If the pellet volume was >20 mL, islets were purified with a COBE 2991 cell processor (CaridianBCT, Inc., Lakewood, CO) with continuous iodixanol-based density gradient centrifugation. The final preparation of islets was assessed for counts, purity, viability and sterility [18] prior to infusion without any culture.

Isolated islets were infused into the portal vein via the mesenteric vein with heparin (70 U/kg recipient body weight) over 15–60 min while the patients were under general anesthesia. After 6 h postsurgery, patients were continuously infused with 300 U/h heparin for 24 h. Subsequently patients were converted to enoxaparin (Lovenox®; Sanofi-aventis, Bridgewater, NJ) at 80 mg/day for 2 weeks. During islet infusion, portal vein pressure was monitored intermittently. If it exceeded 20 mmHg, islet infusion was stopped until portal vein pressure decreased. Among 24 patients analyzed for this study, 13 received etanercept alone and 11 patients received a combination of etanercept and anakinra during islet infusion as described earlier [22].

Full islet graft function was defined as the patient being insulin free; partial function, as needing once-daily long-acting insulin; and nonfunction, as needing more than once-daily long-acting insulin [15].

Measurements of coagulation factors and cytokines

Blood samples were obtained at admission, prior to islet infusion, in the middle of islet infusion, at the end of islet infusion, and 15 min, 1 h, 3 h, 6 h, 24 h, 3 days, 5 days and 7 days after islet infusion. After centrifugation (500 g, 15 min), plasma and serum samples were stored at −80°C until assayed.

The plasma levels of thrombin/anti-thrombin III complex (TAT), C-peptide, C3a, C4d and sC5b-9 were measured using Enzygnost® TAT micro (Marburg, Germany), a C-peptide enzyme-linked immunosorbent assay kit (ALPCO, Salem, NH) and enzyme immunoassay kits of C3a, C4d and sC5b-9 (QUIDEL, San Diego, CA), respectively, according to the manufacturers' instructions.

Concentrations of cytokines and chemokines in serum samples were measured by Luminex 200 using xMAP technology (Millipore, Billerica, MA). The serum concentrations of EGF, Eotaxin, G-CSF, GM-CSF, IFN-γ, IL-10, IL-12p40, IL-12p70, IL-15, IL-17, IL-1Rα, IL-6, IL-7, IL-8, IP-10, MCP-1, MIP-1β and VEGF in 21 patients were measured. The bead assay was performed according to the manufacturer's instructions.

In vitro model

Human pancreatic islets were isolated from research-grade organs procured from brain-dead donors with informed consent, using the modified Ricordi technique as previously described [22, 23]. The isolated pure islets were cultured in CMRL1066 (Mediatech, Inc., Manassas, VA) at 37°C in 95% air and 5% CO2 for 24 h.

Fresh autologous human blood, obtained from the same brain-dead donor as the research-grade pancreas, was collected in a surface-heparinized BD Vacutainer® blood collection tube (Becton Dickinson, Franklin Lakes, NJ). After 24 h of initial culture, purified human islets were mixed with donors' own autologous blood using an in vitro model [4]. Heparin-coated 1.5 mL Eppendorf tubes were used to mix islets (200 IEQ, corresponding to 2.0 µL tissue volume) with 500 µL blood (no additional anti-coagulant added). For the islet-only control, 200 IEQ in 500 µL culture medium without blood were used. For the blood-only control, 500 µL blood in 2.0 µL culture medium was used. The tubes were incubated on a rocking device from 15 min up to 6 h at 37°C. Quintuplicate samples from supernatant were collected at 0, 15 and 60 min and at 3 and 6 h. For each time point, one tube was used.

The “miniature tube” model was validated by measuring the partial oxygen pressure (pO2), pH and glucose levels in the blood with and without islets. pH and pO2 were determined using the I-Stat Blood gas analyzer instrument (Abbott Laboratories, Abbott Park, IL). The blood glucose was measured using the Accu-Chek Aviva glucometer (Roche Diagnostics). At the end of each time point, the blood with and without the islets was centrifuged and the plasma was collected for measuring TAT, C3a, C-peptide and inflammatory cytokine secretion.

Fluorescent staining

After mixing islets with autologous blood with 15 min or 3 h of agitation, islets were preserved in 10% formalin, embedded in paraffin and sectioned at 5 µm. Tissue sections were deparaffinized, and heat-mediated antigen retrieval was performed. The sections were stained immunohistochemically with mouse anti-tissue factor (TF) antibody (Abcam, Cambridge, UK), guinea pig anti-insulin antibody (Sigma-Aldrich, St. Louis, MO), CD11b (Abcam), CD41 (Abcam) or DAPI (Sigma-Aldrich).

Islet culture with low-molecular-weight dextran sulfate and evaluation of viability

Low-molecular-weight dextran sulfate (LMWDS) (MM 5000, Sigma Chemical Co., St. Louis, MO) was added in the miniature in vitro tube model as described above. After the plasma was collected from the blood samples, the islet cells were washed gently 3–5 times with 1× phosphate buffered saline to remove red blood cells and obtain only islets. The islets were then incubated for 10 min with Hoechst 33342 and propidium iodide. The islets were washed and placed on slides, and images were captured using the fluorescent microscope. The images were analyzed using Image J (National Institutes of Health, Bethesda, MD), and the corresponding viabilities were determined by measuring the ratio of the area of propidium iodide to Hoechst.

Statistical analysis

The significance of the maximum intraportal pressure during islet infusion was assessed with a paired t-test comparing results with the preinfusion level. The statistical significance of TAT, C-peptide and C3a levels, platelet counts, and secretion of cytokines and chemokines was determined by two-way repeated-measurement analysis of variance (ANOVA) followed by Holm–Sidak's multiple comparison test. All statistical analyses were performed using IBM SPSS Statistics Version 20 (IBM Corporation, Armonk, NY). Differences were considered significant when p-values were <0.05.

Results

Patient characteristics

The patient and islet characteristics of the 24 cases of AIT are shown in Table 1. In the first group, three patients received an average of 725 ± 103 (average ± SE) IEQ/kg and the mean packed cell volume was 1.0 ± 0 mL. In the second group, 21 patients received an average dose of 5201 ± 615 IEQ/kg, and the mean packed cell volume of islets was 11.1 ± 1.3 mL. There were significant differences in pretransplant HbA1c, islet dose, tissue volume and viability between the two groups. Intraportal pressures increased significantly only in the group receiving >1000 IEQ/kg during infusion from 7.4 ± 0.9 mmHg at preislet infusion to 15.9 ± 1.1 mmHg for maximum pressure during islet infusion (p with paired t-test < 0.001). After AIT, all three patients in first group had partial islet function and in second group 7 patients achieved insulin independence reflecting full islet function, 14 had partial islet function and no patients had nonfunction.

Table 1. Patient and islet characteristics for 24 patients undergoing total pancreatectomy followed by autologous islet transplantation
VariableAIT <1000 IEQ/kg (n = 3)AIT ≥1000 IEQ/kg (n = 21)p-Value
  • AIT, autologous islet transplantation; IEQ, islet equivalents.
  • The data for age, body mass index, pretransplant hemoglobin A1c, total transplanted islets, transplanted islets/kg, tissue volume, purity and viability are expressed as mean ± SE. p-Values were calculated with Fisher's exact tests or unpaired t-test for categorical or continuous values.
  • 1Comparison between initial and maximum pressure.
Gender (female:male)1:214:70.53
Age (years)44.0 ± 1.542.2 ± 2.30.77
Body mass index (kg/m2)20.6 ± 1.326.7 ± 1.20.07
Etiology of pancreatitisIdiopathic, 1; pancreatic divisum, 1; alcoholic, 1Idiopathic, 9; hereditary, 5; pancreatic divisum, 3; alcoholic, 2; autoimmune, 21.0
Pretransplant hemoglobin A1c (%)7.6 ± 1.25.8 ± 0.20.006
Total transplanted islets (IEQ)43 717 ± 6453423 067 ± 41 7520.003
Islet dose (IEQ/kg)725 ± 1035201 ± 6150.002
Tissue volume (mL)1.0 ± 0.011.1 ± 1.30.01
Purity (%)26.7 ± 16.749.2 ± 5.60.18
Viability (%)94.1 ± 1.096.2 ± 0.30.02
Portal vein pressure (mmHg)
Initial6.0 ± 1.77.2 ± 0.80.53
Maximum, p-value18.3 ± 1.9, p = 0.1915.0 ± 1.1, p < 0.0010.03
Delta2.3 ± 1.27.8 ± 1.10.008

Coagulation factors, platelet count and C-peptide levels in AIT patients

Previous reports in allogenic islet transplantation [24, 25] have shown that major characteristics of IBMIR include activation coagulation, complement cascade and infiltration of inflammatory cells. We measured markers for these major events in plasma collected during the peritransplant period after AIT. There was no significant increase in TAT and C-peptide levels in patients who received <1000 IEQ/kg during islet infusion. In contrast, for the group of patients who received >1000 IEQ/kg, there was a rapid and significant (p < 0.05) increase in plasma TAT and C-peptide levels during islet infusion when compared to preinfusion levels. The elevation persisted for up to 3 h (Figure 1A and B). Comparison of TAT peak levels and tissue volume showed strong correlation (Spearman's r = 0.68, p = 0.01), but not with islet purity (Spearman's r = 0.11, p = 0.72). Platelet counts were significantly decreased after infusion until day 3 when compared to preinfusion counts (Figure 1F, p < 0.05).

Figure 1.

The levels of thrombin/anti-thrombin III complex (TAT), C-peptide, platelet count and complements during and after autologous islet transplantation. (A) A significant elevation in TAT levels was observed from the middle of the infusion to 15-min postinfusion when compared to the preinfusion level (*p with two-way repeated ANOVA < 0.05). (B) C-peptide levels from the middle of the infusion to day 7 were significantly higher than those at preinfusion (*p < 0.05); (C and D) however, no significant differences were found in C3a and C4d levels. (E) sC5b-9 levels had two peaks at islet infusion and day 5, but these changes were within the normal range, and no significant differences were observed during the peritransplant period up to day 7 when compared to preinfusion level. (F) Platelet counts significantly decreased 1 h to 3 days after infusion (*p < 0.05). The shadow area shows the normal range. Data are expressed as mean ± SE for 21 patients receiving > 1000 IEQ/kg (solid line) or 3 receiving < 1000 IEQ/kg.

C3a and C4d levels consistently remained within the normal range during the peritransplant period, with no significant differences from preinfusion levels (Figure 1C and D). Although sC5b-9 levels showed peaks at infusion and day 5, there was no significant change when compared to the prior islet infusion (Figure 1E).

Proinflammatory cytokines following AIT

To understand the inflammatory process during AIT, we measured serum levels of 18 proinflammatory markers. Among these proteins, significant differences were observed only in IL-6, IL-8 and interferon-inducible protein-10 (IP-10) during the peri-islet infusion period (Figure 2). There was no significant increase in these three inflammatory proteins in the group of patients who received <1000 IEQ/kg. In contrast, among the patients who received >1000 IEQ/kg, IL-6 levels were elevated after admission and before islet infusion, indicating the period when TP and gastrointestinal reconstruction were performed, and further increased up to 6 h postinfusion (Figure 2A). Similarly, serum IL-8 levels were elevated up to 15 min postinfusion (p < 0.05 at 15 min postinfusion compared with preinfusion level, Figure 2B). IP-10 levels immediately increased during islet infusion (Figure 2C). IP-10 levels at the end of infusion and 15 min after infusion were significantly higher than before infusion (p < 0.05).

Figure 2.

Cytokine and chemokine levels during and after autologous islet transplantation. Levels of (A) IL-6, (B) IL-8 and (C) interferon-inducible protein-10 (IP-10) are shown. Significant elevation was observed in IL-8 levels at 15 min after infusion compared to the preinfusion level (*p with two-way repeated ANOVA < 0.05). IP-10 levels at the middle of infusion to 15 min after infusion were significantly higher than the preinfusion levels (*p < 0.05). Data are expressed as mean ± SE for 21 patients receiving > 1000 IEQ/kg (solid line) or 3 receiving < 1000 IEQ/kg (dotted line).

In vitro model with islets and autologous blood to analyze IBMIR

Currently the preferred technique for clinical allogenic and AIT is infusion via portal vein into the liver. Limited techniques are available to study the fate of islets after transplantation. Hence we employed the miniature tube model based on the original loop model described by Moberg et al [25]. In this model, the mixed blood and islet group had significantly increased TAT (p < 0.05) and C-peptide (p < 0.05) levels, but the blood-only sample and islet-only sample did not (Figure 3A and C). The blood-islet mixed group had baseline TAT levels of 91 ± 7 µg/L, which peaked at 3 h to 752 ± 92 µg/L and continue to remain elevated (738 ± 177 µg/L) (Figure 3A). A similar trend was seen in C-peptide levels: starting with baseline levels of 17 840 ± 1986 pmol/L, levels in the mixed blood-islet group significantly increased at 1 h (30 721 ± 1907 pmol/L) and then continued to rise through 6 h (37 983 ± 3164 pmol/L) (Figure 3C). C3a levels were elevated in both the mixed blood-islet group and the blood-only group, but not in the islet-only group (Figure 3E).

Figure 3.

Changes in thrombin/anti-thrombin III complex (TAT), C-peptide and C3a in the mixed blood and islets group compared with the blood-only and islet-only controls under in vitro conditions. The mixed blood and islet group (solid line) had significantly increased (A) TAT and (C) C-peptide levels (*p with repeated-measurement ANOVA < 0.05), although the blood-only sample (dotted line with squares) and islet-only sample (dotted line with triangles) did not. (E) C3a levels were elevated in both the mixed blood-islet and the blood-only samples, but not in the islet-only samples. Culture conditions on (B) pH, (D) pO2 and (F) glucose level were determined. The mixed blood and islet group showed pH and glucose levels similar to those of the blood-only group. Although pO2 levels were significantly lower in the mixed group than in the blood-only group (*p < 0.05), pO2 levels >60 mmHg were maintained for 6 h in the mixed group. Data are expressed as mean ± SE.

To validate this tube model, the culture conditions were monitored by measuring pH, pO2 and glucose levels (Figure 3B, D, and F). pH levels were stable in the model for 6 h, although pO2 and glucose levels were gradually decreased. The mixed blood and islet group showed significantly lower pO2 levels than the blood-only group; however, a minimum level of 60 mmHg of pO2, which corresponds to a 90% oxygen saturation level, was maintained throughout the in vitro model (Figure 3D). These results showed acceptable conditions of the miniature tube model for collecting short-term data.

Release of the proinflammatory cytokines MCP-1 and TNF-α in the in vitro model was also examined (Figure 4). Consistent with results for C-peptide levels, significantly higher levels of the proinflammatory cytokines were observed in the mixed blood and islet group compared to the blood-only group (p < 0.05).

Figure 4.

Proinflammatory cytokine expression under in vitro conditions. The mixed blood and islet group had significantly higher proinflammatory cytokine levels of (A) MCP-1 and (B) TNF-α in culture for 6 h when compared to the blood-only group (*p with repeated-measurement ANOVA < 0.05). Data are expressed as mean ± SE. TNF-α, tumor-necrosis factor-alpha.

TF expression in islets following autologous blood exposure

Using the miniature tube model, we also tested for TF expression on islets, since it is known to trigger an inflammatory reaction. TF expression was analyzed by immunohistochemical staining with mouse anti-TF antibody. Observation of the sections representing islets mixed with blood showed that islets were surrounded by erythrocytes at 0 and 15 min after agitation. The expression of TF (shown in green) at 15 min was remarkably higher than at 0 min (Figure 5).

Figure 5.

Tissue factor (TF) expression in islets following exposure to autologous blood under in vitro conditions. Control islets (0 min) and islets 15 min after mixing with autologous blood were stained with hematoxylin–eosin (H.E.) (left panel), insulin (red) and TF (green). TF expression as determined by the binding of anti-TF antibody was observed 15 min after mixing. Scale bars show 50 µm.

Inflammatory cell infiltration in islets following autologous blood exposure

Infiltration of inflammatory cells into islets is a hallmark of IBMIR in an allogenic model [25]. To analyze infiltration of granulocytes/macrophage and platelets, CD11b and CD41 staining was performed respectively with the mixed islet and blood group and compared to preculture results as well as results in islet-only culture samples (Figure 6). CD11b-positive cells infiltrated in islets after culture with autologous blood for 3 h, although CD41-positive cells were found surrounding islets (Figure 6B).

Figure 6.

CD11b and CD41 staining for the islet-only group and the mixed blood and islet group in an in vitro model. Triple staining for DAPI, insulin and CD11b or CD41 is shown for (A) the islet-only group and (B) the mixed blood and islet group. No infiltration of CD11b-positive monocytes and CD41-positive platelets was observed in the islet-only culture condition for both control (preculture) and 3-h culture. Although CD11b-positive cells were infiltrated in islets 3 h after culture with blood, CD41-positive cells were seen surrounding islets. Scale bars show 50 µm.

Effects of LMWDS on islet viability in autologous tube model

LMWDS is a heparin-like anti-coagulant and anti-complement agent. It has been shown to effectively inhibit IBMIR in allogenic and xenogenic models [26]. We tested the protective effect of LMWDS using an in vitro tube model and determined islet viability, as measured using Hoechst 33342 and propidium iodide staining (Figure 7). Mixing of blood with islets clearly decreased the viability of islets in 6 h. However, the islet-blood mix in the presence of LMWDS showed significantly better islet viability than the control group (p < 0.05).

Figure 7.

Effect of low-molecular-weight dextran sulfate (LMWDS) on islet viability in an autologous in vitro model. LWMDS was added in islet culture with autologous blood. The viability of islets cultured with LWMDS (circles) was significantly higher than that in the control group (triangle; *p with repeated-measurement ANOVA < 0.05).

Discussion

Transplantation of isolated pancreatic islets represents a promising minimally invasive method of treating not only “brittle” type 1 diabetes, but also other debilitating conditions involving the pancreas such as chronic pancreatitis, benign tumors and trauma. Both the short-term and long-term successes of the islet transplant procedure are critically dependent on the intrahepatic engraftment of islet tissue. IBMIR has been shown to be a major hurdle in the successful engraftment of allogenic and xenogenic islets. In this report we have demonstrated, for the first time, the occurrence of IBMIR in AIT based on a comprehensive study. Our analysis of 24 AIT patients has shown early damage to transplanted islets based on a rapid increase in C-peptide in plasma with a concomitant rise in TAT. Furthermore, a significant rise in the proinflammatory cytokines IL-6, IL-8 and IP-10 was observed, indicating the activation of the innate immune system. In the group of patients who received <1000 IEQ/kg islet dose, there were minimal changes in C-peptide, TAT and inflammatory cytokines, further supporting a role for the autologous islets in triggering IBMIR. Interestingly, no significant activation of complement was observed, as analyzed by C3a, C4d and sC5b-9 levels. Due to limitations in further in vivo analysis of intrahepatic islets, we developed and validated an in vitro model, and data obtained from this model clearly supported the in vivo findings. Islets mixed with autologous blood showed increased expression of TF, a key activator of inflammation and infiltration of CD11b- and CD41-positive cells. Taken together, the data presented here indicate a strong role for coagulation and inflammation in the response against autologous islets.

Moberg et al [25] successfully demonstrated IBMIR in allogeneic islet transplantation, as shown by an increase in the plasma levels of TAT and C-peptide immediately after islet infusion, and showed that the increased C-peptide level was considered a result of leaking insulin from damaged transplanted islets. Using an in vitro “loop model,” they further demonstrated that TF was expressed strongly after mixing islets and allogenic blood. Enhancement of TF expression in islets was proposed as the trigger of IBMIR, and blockade of TF activity abrogated IBMIR [25]. Based on this report, we developed an in vitro miniature tube model to study IBMIR in vitro. Validation of the in vitro tube model was performed by measuring pH, pO2 and glucose, which supported good survival conditions for cells. Our in vivo and in vitro analyses showed rapid increases of TAT and C-peptide concentrations after mixing islets and autologous blood as well. Kinetics of the TAT and C-peptide increase observed in the present study matched the observations reported for clinical allogenic islet transplants [24]. Therefore, coagulation activation is clearly evidenced after islet infusion in AIT, despite the presence of heparin in the islet preparation at 70 U/kg recipient body weight.

Eriksson et al [24] have characterized IBMIR in clinical allogenic islet transplantation using dynamic positron emission tomography/computed tomography imaging with [18F]FDG-labeled islets that allowed qualitative and quantitative assessment. Two of their major observations included the heterogeneous distribution of islets in the liver after intraportal infusion and the damage to islets occurring within minutes after infusion. Islet damage could also be caused to a smaller extent by hemodynamic shear stress during infusion; however, the increase in coagulation and inflammatory markers points to a major role of IBMIR mediated by the innate immune system.

In the allogenic islet transplant model, deposition of significant levels of IgG, IgM, C4, C3 and C9 on the surface of islets was observed, suggesting activation of complement through the classical pathway to the formation of the membrane attack complex [27]. Islet cell surfaces that are not exposed to blood under normal conditions are infused into the recipient circulation. IBMIR has also been reported in clinical allogenic hepatocyte transplantation [28], suggesting that the IBMIR reaction is common to all cell therapies in which cells not intended for contact with blood are infused into the circulation. Cells such as islets and hepatocytes could be recognized as foreign by the innate immune system. Recognition by natural antibodies in the absence of membrane regulators such as DAF, MCP or CD59 could lead to complement activation. The characteristics of islet surface antigens recognized by antibodies are unknown. Potential candidates include extracellular matrix proteins such as collagen and laminin, which are present on both endocrine and exocrine tissue after digestion with collagenase during the isolation procedure. Characteristics of IBMIR in a nonhuman primate model include strong activation of coagulation and complement cascades coupled with major infiltration of platelets, macrophages, neutrophilic granulocytes and T cells [7]. Strong deposition of xenoreactive antibody combined with C1q, C3 and C4 was observed. LMWDS was more effective than heparin in minimizing IBMIR [29]. In allogenic and xenogenic islet studies, we observed mild elevation of sC5b-9 levels during islet infusion and at days 3–7; however, little C3a and C4d response during the peritransplant period was seen. Thus complement activation may play a minor role in autologous IBMIR.

IBMIR in allogenic islet transplantation is characterized by consumption of platelets and activation of the coagulation and complement system [3, 25]. The platelet count was significantly decreased after autologous islet infusion, consistent with results from a previous study using the islet loop model with ABO-compatible blood [3]. That study suggested platelet consumption due to activated coagulation reaction. It is important to point out that unlike allogenic transplants, AIT involves major surgery, and hence a drop in platelet counts is expected during the procedure.

We found elevated secretions of IL-6, IL-8 and IP-10 during TP-AIT, suggesting active participation of macrophages/granulocytes in the response against islets. Previously it was shown that the secretion of cytokines and chemokines increased after allogeneic islet transplantation [30, 31]. In addition, other reports mentioned that secretion of cytokines and chemokines was related to TAT elevation [32, 33]. Interestingly, we found two peaks of IL-6 and IL-8 elevation; the first peak was before islet infusion, and the second was after islet infusion. Since secretion of IL-6 and IL-8 is increased by major abdominal surgery [34, 35], it seems reasonable to consider the first peak due to TP and the second peak due to AIT. IL-6 and IP-10 are major inflammatory cytokines, and IL-8 is a major chemokine; therefore, increases in those cytokines by both TP and AIT can deteriorate the outcome of AIT. Recently, it was shown that potent anti-inflammatory therapy improved engraftment of transplanted islets in an allogeneic model [23, 36]. Since TP-AIT strongly induced inflammation by both TP and AIT, anti-inflammatory therapy might improve the engraftment and function of transplanted islets.

Data compiled from 409 TP-AIT cases from the University of Minnesota showed that at 3 years posttransplant, 30% of adult recipients remained insulin independent and 33% had partial islet function [8]. Among patients who received <2500 IEQ/kg, 12% remained insulin independent whereas 22% were insulin independent after an islet dose of 2500–5000 IEQ/kg and 72% after a dose of >5000 IEQ/kg [8]. These data clearly demonstrate the importance of the mass of islets transplanted for the posttransplant outcome. Improving islet engraftment by minimizing the damaging effects of IBMIR is a key approach to improve outcomes. Several approaches are being made to mitigate the effects of IBMIR. Recently, it was shown that surface modification with poly(ethylene glycol)-lipid and urokinase effectively protected islets from IBMIR [37]. LMWDS (MW5000) could inhibit the activation of both the coagulation and complement system [38, 39]. LMWDS is a heparin-like anti-coagulant and anti-complement agent that has been proposed for attenuation of IBMIR. The safety of LMWDS administration and its mechanism of action by the induction of hepatocyte growth factor has been demonstrated in a recent phase I clinical trial [40]. Furthermore, recombinant human-activated protein C and the platelet inhibitor tirofiban have successfully inhibited IBMIR [41]. Those modifications and/or drugs might also inhibit IBMIR during TP-AIT. Alternative sites to implant islets other than the liver are under consideration. In a series of four patients, islets were infused in bone marrow with encouraging outcomes in terms of survival and function [42]. A case study involving intramuscular autotransplantation of pancreatic islets in a child has also been reported [43].

In conclusion, this is the first study of IBMIR during TP-AIT. Modulating this process may help improve islet engraftment and the insulin-independence rate in TP-AIT patients.

Acknowledgments

The authors thank Yoshiko Tamura, Ana M. Rahman and Anne-Marie Brun for technical support and Ms. Cynthia Orticio for professional editing. This study was supported in part by Baylor Health Care System Foundation and grants from NIDDK (1R21DK090513-019 to MFL) and JDRF (5-2011-372 to BN and 3-2011-447 to MT).

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Ancillary