Journal Logo

Original Clinical Science—General

Beta Cell Death by Cell-free DNA and Outcome After Clinical Islet Transplantation

Gala-Lopez, Boris L. MD, MSc, PhD1,5; Neiman, Daniel2; Kin, Tatsuya MD, PhD1; O’Gorman, Doug1; Pepper, Andrew R. PhD1; Malcolm, Andrew J. PhD2; Pianzin, Sheina2; Senior, Peter A. MD, PhD3; Campbell, Patricia MD3; Glaser, Benjamin MD4; Dor, Yuval PhD2; Shemer, Ruth PhD2; Shapiro, A.M. James MD, PhD1,3,5

Author Information
doi: 10.1097/TP.0000000000002083

Despite the success of clinical islet transplantation (IT), outcomes may still be hampered by cell death occurring early after intraportal infusion. After transplantation, islets are hypoxic, and exposed to the instant blood-mediated inflammatory response (IBMIR) and alloimmunity.1 As a consequence, a significant portion of the graft is lost at this early stage often necessitating multiple subsequent donor islet infusions to achieve insulin independence.2,3 An inability to accurately determine cell death after intraportal IT remains a limiting factor in predicting early and long-term graft function, and in identification of modifiable factors that could enhance early engraftment. In most cases graft loss can only be measured by its functional consequence at a stage where a potential therapeutic window has been closed.4

Assessing engraftment and graft function is not a straightforward process and normally relies on complex metabolic tests of insulin secretory profiles, which are both time consuming and expensive.4 Multiple scoring systems have been implemented to measure posttransplant function more accurately including the beta 2 score recently described by our group,4 which can be obtained without using a stimulant to successfully discriminate between glucose intolerance and insulin independence after IT. A limitation to this functional approach is the lack of quantifiable evidence of graft loss and its corresponding clinical impact.

When cells die, fragments of their genomic DNA are released to circulation, where they travel shortly before being cleared by the liver. In a recent breakthrough, Akirav et al5 showed that hypomethylation in the insulin gene promoter, which is unique to beta cells, can be used to detect DNA derived specifically from dying beta cells. Indeed, studies from their group and other investigators showed that unmethylated insulin promoter circulating cell-free DNA (cfDNA) can be detected in the blood of patients recently diagnosed with type 1 diabetes mellitus (T1DM), raising the exciting possibility that this novel type of biomarker can be used to precisely monitor beta cell death.6-8 We have recently reported a new version of this technology, based on next generation sequencing, for the assessment of beta cell death as well as multiple other tissues based on tissue-specific methylation markers. We used the method to assess the levels of beta cell-specific cfDNA in serum and plasma of healthy individuals, in T1DM patients and in islet graft recipients.9 Although healthy individuals had extremely low concentrations of beta cell-derived cfDNA, a clear signal was observed in recently diagnosed T1DM patients. Specifically in islet graft recipients, we observed significant levels of beta cell-specific cfDNA as early as 1 hour after islet infusion, and a second less intense beta-cell cfDNA signal 24 hours after transplantation, which could be an objective expression of early graft loss.9 Since the clinical utility of this novel technology is not defined, we sought to characterize cfDNA measurements in a clinical islet allotransplantation setting and to correlate findings with clinical outcomes as a method to complement the initial assessment of islet engraftment and function. Here, we report the outcome of these studies, pointing to beta cell cfDNA as a promising prognostic biomarker for clinical IT.


Study Design

Clinical islet isolations and allotransplants were performed at the University of Alberta, Canada, over a 14-month period using defined standard of care. Only those subjects receiving a first intraportal allotransplant or a retransplant after a remote (>1 year) previous infusion were included in the analysis (n = 37) to avoid multiple confounding factors derived from previous transplants. This study was part of the ongoing review of islet transplant patients at the University of Alberta (protocol number Pro00001120) in collaboration with Juvenile Diabetes Research Foundation and the Hebrew University (study number RES-0024003). This investigation is approved by the University of Alberta Health Research Ethics Board and conducted in accordance with the principles endorsed by the Declaration of Helsinki.

Human Islet Isolation, Purification, and Culture

Human islet preparations were isolated from deceased donor pancreata, as previously described using a collagenase/thermolysin enzyme mixture (Roche Diagnostics, Laval, QC, Canada), and the resulting digest suspension was purified using a modified COBE 2991 cell processor (Terumo BCT, Inc., Lakewood, CO) with continuous density gradients.2,10-12 The purified islet fraction(s) were assessed for clinical suitability and cultured in CMRL 1066-based medium at 22°C (5% CO2) for up to 72 hours before transplantation.

Islet Product Characterization

Islet dose was calculated using standard methods.12 Briefly, islets greater than 50 μm in diameter were enumerated by manual count with an inverted light microscope and classified into size ranges in increments of 50 μm. The number of islets particles in each size range was converted to islet equivalent (IE) to account for size difference. Furthermore, Islet Size Index (ISI) was calculated by dividing the total number of IE in a preparation by the total number of islet particles to reflect the average particle size in a preparation.12 After the culture period, islets were assessed for viability and functionality. Viability was assessed by fluorescent membrane integrity assay with counterstains using SYTO 13 green fluorescent nucleic acid stain (Life Technologies, Burlington, ON, Canada) and ethidium bromide (Sigma-Aldrich, ON, Canada).13-15 Samples were manually assessed using fluorescent microscopy and reported as a percentage of viable to all cells.

Hormonal islet secretory function was assessed by static glucose-stimulated insulin secretion (s-GSIS), sequentially performed at low (2.8 mmol/L) and high (28.0 mmol/L) glucose concentrations. The amount of insulin released was measured using an enzyme-linked immunosorbant assay (Mercodia Insulin ELISA, Mercodia, Uppsala, Sweden) and a stimulation index was calculated as the ratio of stimulated to basal insulin secretion.

Oxygen consumption rate (OCR) is a real-time, operator-independent method of assessing fractional cell viability. OCR was measured as described previously.16-19 OCR was normalized to the DNA content per chamber by collecting the islets and assessing for DNA by using a double-strand DNA fluorescent dye (Quant-iT PicoGreen dsDNA Assay Kit; Invitrogen, Life Technologies Corporation, Grand Island, NY) resulting in OCR/DNA (nmol O2/min·mg DNA). To further characterize the preparation, the mean OCR/DNA value of an islet preparation was normalized to ISI (OCR/DNA/ISI) and the islet dose (IE/kg), as previously reported.20

IT Procedure

Islet transplant procedures were performed using a previously reported procedure.21 Islets were suspended in 100 mL of CMRL-based transplant media supplemented with human serum albumin and HEPES buffer into an infusion bag. Patients received the infusion via a catheter placed intraportally, performed under local anesthetic and with combined ultrasonography and radiology guidance. Final vascular track ablation with Avitene paste (Davol, Inc., Warwick, RI) was done after the procedure.

Immunosuppression consisted of T depletional induction therapy with alemtuzumab (MabCampath, Genzyme Corp.) and anti-inflammatory therapy included etanercept (Enbrel; Amgen Canada Inc., Mississauga, ON) and anakinra (Kineret; Amgen Canada Inc., Mississauga, ON). Maintenance twice daily tacrolimus (Prograf; Astellas Pharma Canada Inc., Markham, ON) was adjusted to provide target trough levels of 10 to 12 μg/L, together with mycophenolate mofetil (CellCept; Hoffmann-La Roche Ltd., Mississauga, ON). Three subjects received basiliximab (Simulect; Novartis Pharmaceutical Canada Inc. Dorval, QC) induction instead of alemtuzumab. All subjects received peritransplant insulin-heparin infusions as per our standard protocol and after the procedure; their insulin requirements were adjusted according to functional indicators.22 All recipients and donors were HLA typed by low- to medium-resolution typing (LabType SSO One Lambda A Thermofisher Scientific at HLA A, B, C, DRB1, DRB345, DQA1, DQB1, DPA1, and DPB1). Pretransplant antibody testing was performed with single antigen bead testing (LabScreen One Lambda). Calculated panel-reactive antibodies (cPRA) was calculated using the Canadian cPRA calculator ( Flow crossmatches were performed on pretransplant serum.23

Measurement of cfDNA

Beta cell-specific cfDNA was measured in plasma from study subjects as reported by Lehmann-Werman and collaborators.9 Blood samples were collected from patients at 1 hour, 24 hours, 7 days, and 1 month after IT. cfDNA was extracted from 4 mL of plasma using the QIAsymphony liquid-handling robot (Qiagen) and was treated with bisulfite (Zymo Research). DNA concentration was measured using Qbit single-strand molecular probes (Invitrogen). Bisulfite-treated DNA was PCR-amplified, using insulin promoter primers specific for bisulfite-treated DNA but independent of methylation status at monitored CpG sites. Sequencing was performed on PCR products using MiSeq Reagent Kit v2 (MiSeq, Illumina method). Sequenced reads were separated by a barcode, aligned to the target sequence (insulin gene promoter), and analyzed using custom scripts written and implemented in Matlab. Reads were quality filtered based on Illumina quality scores. Reads were identified by having at least 80% similarity to target sequence and containing all the expected CpGs in the sequence. To calculate the concentration of cfDNA derived from beta cells, we multiplied the fraction of beta cell-specific cfDNA (as determined from the frequency of molecules carrying a beta cell-specific methylation pattern) by the concentration of cfDNA measured in each particular patient. The concentration was expressed in copies DNA/mL and a value of 50 copies/mL was considered as the minimum threshold for positive cell death.

Absolute beta cell loss was estimated multiplying the number of DNA copies by the patient’s total plasma volume (TPV). Calculation of TPV was done using the total body water and the extracellular fluid,24-26 using the formula:

TPV = Extracellular Fluid − Interstitial Fluid

With the following assumptions:

  • total body water = 60% of body weight (for male)
  • total body water = 50% of body weight (for female)
  • extracellular fluid = total body water/3
  • interstitial fluid = total body water × 0.25

Cell loss was estimated relative to the original cell count present in the islet preparation, assuming an average of 1140 beta cells/IE.27

Metabolic Studies After Transplantation

Posttransplant graft function was measured as previously reported.3,4,28 It consisted of sequential clinical and metabolic assessments including the results of several metabolic tests such as the record of hypoglycemia events, fasting blood glucose, basal and stimulated C-peptide levels, glycated hemoglobin, oral and intravenous glucose tolerance tests, mixed meal stimulation tests, glucagon and arginine, as well as requirements for exogenous insulin. Beta 2 scores were calculated 3 months after transplant as recently described by Forbes and collaborators,4 as a more accurate indicator of islet engraftment.

Statistical Analysis

To assess the significance of differences in isolation parameters and cell loss between groups with positive and negative cell death, we used a 2-tailed Mann-Whitney test. Two-tailed Pearson correlation was used to measure the linear dependence between cfDNA and immediate posttransplant function variables (1-month stimulated C-peptide and insulin requirements). Fisher exact test was used for comparison of proportions. Relationships between beta cell death and graft function (3-month insulin independence and beta 2 score) were examined using receiver operating characteristic (ROC) analysis. The area under the ROC curve (AUC) was calculated from ROC curves generated for each method. All comparisons between groups were performed with a 95% confidence interval, and a P value less than 0.05 was considered significant. Analysis was performed using GraphPad Prism (GraphPad Software version 6, La Jolla, CA).


We defined selection criteria for the study as subjects receiving a first transplant or a re-transplant after a remote (>1 year) previous infusion (see Materials and methods). During the study period (2014-2016), 100 islet allotransplants were performed in 83 patients. Only 37 (44.6%) subjects fulfilled these selection criteria and were included in the analysis. Baseline characteristics of these subjects are comparable to excluded patients during the same period (Table 1).

Baseline characteristics of subjects and islet preparations included in the study compared with other transplanted patients during the same study period (2014-2016)

Isolation Parameters Do Not Influence Posttransplant Beta Cell cfDNA Levels

Patient and isolation characteristics were comparable throughout the study cohort. The median level of cfDNA 1 hour after transplant was 852 copies/mL (range, 0-6647); assuming 3000-mL plasma per individual, an average islet transplant recipient had ~2.5 million beta cell genomes in the circulation at this stage. Positive beta cell-specific cfDNA was detected in 31 (83.8%) of 37 patients at 1 hour, whereas only 8 (21.6%) of 37 patients were positive at 24 hours posttransplant (P < 0.0001). Levels of beta cell cfDNA were particularly high 1 hour after transplant. Based on these cfDNA measurements, we estimated that 5.2 × 106 beta cells (range, 349 198-3.3 × 107) were lost before the infusion or during the first hour after transplant (Table S1 and Figure S1, SDC, These levels of 1 hour beta cell cfDNA were not associated with any of the islet preparation parameters including culture time, preparation purity and viability, islet size, dose and packed cell volume (PCV), as well as other functional markers such as sGSIS, pure OCR or OCR adjusted by islet size index or islet dose (Table 2). In principle, cfDNA shortly after transplantation could reflect material from beta cells that have died during islet isolation or culture, acute death of beta cells after transplantation, or a combination of both.

Comparison of baseline islet preparation characteristics between patients with positive versus negative beta cell-specific free circulating DNA, 1 h after clinical transplantation

The levels of beta cell cfDNA measured at 24 hours were not associated with any isolation parameter but we observed increased beta cell cfDNA at this time point when islets were infused in a larger PCV. Patients with positive cfDNA were transplanted with a median PCV of 4.0 mL (range, 3.5-7.5) versus 3.0 mL (range, 2.0-5.0) in patients with negative cfDNA (P = 0.002, Table 3). These 24-hour cfDNA measurements resulted in a significantly lower estimation of 427 991 cells (range, 297 815-1.6 × 106) lost after transplant, corresponding to a significantly lower graft loss from the initial islet preparation compared to the earlier time point (24 hours, 0.09% cell loss; range, 0.06-0.3% vs 1 hour, 1.2%; range, 0.06-8.9%; P < 0.0001; Table S1 and Figure S1, SDC,

Comparison of baseline islet preparation characteristics between patients with detectable or undetectable beta cell-specific free circulating DNA, 24 hours after clinical transplantation

Beta Cell-Specific cfDNA Measured at 24 Hours Is Associated With Immediate Posttransplant Graft Function

Subjects with a signal of beta cell cfDNA at 1 hour had similar initial graft function compared with those with no 1-hour beta cell cfDNA, expressed as comparative 1-month exogenous insulin requirements (0.14 U/kg per day vs 0.11 U/kg per day, P = 0.55) and 1 month-stimulated C-peptide levels (0.92 nmol/L vs 1.3 nmol/L, P = 0.22). These represent the most relevant indicators of immediate graft function after IT. Significant differences, however, were found at the 24 hours time point. Patients with 24 hours positive cfDNA had significantly higher 1-month insulin requirements (0.26 U/kg per day vs 0.13 U/kg per day, P = 0.04), higher 1 month-absolute insulin usage (15 U/d vs 8 U/d, P = 0.04) and significantly lower 1 month-stimulated C-peptide (0.79 nmol/L vs 1.4 nmol/L, P = 0.01). These findings were also supported by positive 24-hour correlations between cfDNA concentration and 1-month exogenous insulin requirements (r2 = 0.26, P = 0.001), and between cfDNA concentration and 1-month stimulated C-peptide levels (r2 = 0.15, P = 0.02). These correlations indicate a possible significant association between these variables despite a low r2 value, which may be a consequence of an inherently higher amount of unexplainable variability between subjects. It is possible, however, that additional predictors can increase the true explanatory power of this particular model (Figures 1 and 2).

A 24-hour cfDNA is better correlated with immediate posttransplant exogenous insulin demand. A, 1-month insulin requirements in subjects with positive (n = 31) and negative cfDNA (n = 6) measured 1 hour posttransplant (P = 0.55). B, 1-month insulin requirements in subjects with positive (n = 8) and negative cfDNA (n = 29) measured 24 hours posttransplant (P = 0.04*). C, Linear regression of cfDNA measured 24 hours posttransplant and 1-month insulin requirements (n = 37, r2 = 0.26, P = 0.001). D, 1-month absolute insulin used in subjects with positive and negative cfDNA measured 24 hours posttransplant (P = 0.04*). A greater value of cfDNA suggests greater beta cell death and lower insulin requirement reflect better posttransplant function. Summary data are reported as median (interquartile range), 2-tailed Mann-Whitney, 95% confidence interval.
A 24-hour cfDNA is better correlated with immediate posttransplant C-peptide levels. A, 1-month stimulated C-peptide blood levels in subjects with positive (n = 31) and negative cfDNA (n = 6) measured 1 hour posttransplant (P = 0.22). B, 1-month stimulated C-peptide blood levels in subjects with positive (n = 8) and negative cfDNA (n = 29) measured 24 hours posttransplant (P = 0.01). C, Linear regression of cfDNA measured 1 hour posttransplant and 1-month stimulated C-peptide blood levels (n = 37, r2 = 0.0002, P = 0.94). D, Linear regression of cfDNA measured 24 hours posttransplant and stimulated C-peptide blood levels (n = 37, r 2 = 0.15, P = 0.02). A greater value of cfDNA suggests greater beta cell death and higher stimulated c-peptide levels reflect better posttransplant function. Summary data are reported as median (interquartile range), 2-tailed Mann-Whitney, 95% confidence interval.

Beta Cell-specific cfDNA Measured at 24 Hours May Be a Predictor for Islet Engraftment

All patients were closely monitored for graft function and exogenous insulin requirements before receiving subsequent complementary islet infusions. We examined beta cell cfDNA levels at 1 hour and 24 hours after the procedure to correlate beta cell death rate with graft function at a later time point. Three months after transplant 20 (54.1%) of 37 patients were insulin independent and a favorable Beta 2 score (>15 points) was also calculated in these same subjects. These 2 indicators provide a more objective assessment of islet engraftment and may predict the need for a supplementary transplant to facilitate long term benefits.4 The 1-hour beta cell-specific cfDNA failed to correlate with both, insulin independence (P = 0.45) and with Beta 2 score (P = 0.10), whereas 24-hour positive beta cell cfDNA was inversely associated with both outcomes. Only 4 (20%) of 20 subjects with positive 24 hours beta cell cfDNA were insulin independent at 3 months (P = 0.04) and only 1 of those 3 patients (1/20, 5%) with detectable 24 hours beta cell cfDNA achieved a 3 month-beta 2 score greater than 15 points (P = 0.008). The ROC analysis supported using the 24 hours beta cell cfDNA as an optimal model for insulin independence (ROC:AUC, 0.70; P = 0.03; sensitivity, 75%; specificity, 58.8%), and beta 2 score (ROC:AUC, 0.77; P = 0.006; sensitivity, 88.9%; specificity, 52.6%) at 3 months, using a discrimination threshold of 50 copies/mL (Figures 3 and 4). PCV of the islet preparation was not correlated with these 3-month outcomes.

A 24-hour cfDNA can predict 3-months insulin independence after intraportal islet allotransplantation. A and B, Correlation of cfDNA measured 1 hour posttransplant with clinical outcome expressed as 3-month insulin independence and its corresponding ROC curve. C and D, Correlation of cfDNA measured 24 hours posttransplant with clinical outcome expressed as 3-month insulin independence and its corresponding ROC curve. The AUC has been calculated and displayed for each group. A greater value of cfDNA suggests greater beta cell death.
A 24-hour cfDNA can predict 3-months engraftment after intraportal islet allotransplantation. A and B, Correlation of cfDNA measured 1 hour posttransplant with clinical outcome expressed as 3-month Beta 2 Score and its corresponding ROC curve. C and D, Correlation of cfDNA measured 24 hours posttransplant with clinical outcome expressed as 3-month beta 2 score and its corresponding ROC curve. The AUC has been calculated and displayed for each group. A greater value of cfDNA suggests greater beta cell. Beta 2 score is a composite measure of beta cell function after transplant.4 A beta 2 score > 15 points reflects a functioning graft.

A positive beta cell cfDNA signal was only observed in 2 patients at 7 days postprocedure and in 2 different subjects, 1 month after transplant. This late beta cell mortality was not associated with unfavorable outcomes (Table S2, SDC, Moreover, 24 hours positive beta cell cfDNA failed to predict the time to second supplementary transplant, and there was no correlation between the cPRA or the presence of donor-specific antibodies (DSA) or the levels of cfDNA at any time point. Similarly, there was no relationship between cPRA/DSA and the 3-month insulin independence or beta-2 score (data not shown).


Assessing graft loss is a particularly challenging aspect of IT due to the many mechanisms eliciting islet injury and the lack of tools to measure beta cell mass in vivo.1 We have described a 14% islet mass loss during culture associated with cell fragmentation and disintegration.29 Another important islet loss event occurs immediately after infusion, largely due to cell hypoxia and inflammatory responses after transplantation. In particular, IBMIR accounts for significant graft attrition at this early stage and is reported to reach up to 70% of the initial preparation within the first 24 hours, when using the intraportal route.30,31 The sum of all these events paired with sustained immune-related cell death, is the rationale for using more than 1 infusion per patient to achieve durable normoglycemia.32 Nonetheless, implementation of new protocols in selected centers can result in high rates of single-donor insulin independence.33

To date, clinical IT lacks an accurate estimator for cell loss to support clinical and metabolic indicators from the early posttransplant phase. Beta cell-specific cfDNA has been identified as a novel biomarker to detect islet loss.9,34,35 We have recently reported the use of a novel cfDNA measurement technology, based on next generation sequencing, to identify beta cell death in patients with new-onset T1DM and recent clinical IT patients.9 Moreover, a recent publication used a similar technique to detect beta cell mortality after islet autotransplantation with a full characterization of islet loss throughout the process of pancreatitis, total pancreatectomy, and transplant.35 We herein report the first observations of a correlation between beta cell-specific cfDNA after clinical allotransplantation and patient outcomes.

After measuring cfDNA at 1 hour, 24 hours, 7 days, and 1 month, we observed 2 distinct signal peaks in our study population. One, at 1 hour after infusion, very intense (934 copies/mL) and generalized (83.8% of patients) and another, less intense (93 copies/mL) and less frequent (21.6% of patients) at 24 hours. Although the 1-hour signal is strong, it is transient, and given the limited number of sample time points and the rapid clearance of cfDNA, it likely represents a combination of beta-cell death carried from the islet isolation and culture procedures and acute cell loss early after infusion. These observations are consistent with the recent report on autotransplantation.35 In contrast, cfDNA signal at 24 hours was more informative of future graft function. In our study, this was not related to any isolation parameter, including well-established viability indicators like OCR/DNA, OCR/DNA/ISI, or OCR dose.16,19,20 Surprisingly, only the PCV showed a significant association with 24 hours cfDNA. PCV is directly dependent on the islet fraction purity and increasing PCV (>5.5 mL) has been associated with high portal venous pressure, increased risk for portal thrombosis21,36 and self-limited hepatocellular damage.37,38 Our findings indicate that increased PCV may also be associated with early beta cell death. Despite this 24-hour cell mortality being relatively small compared with the signal measured at 1 hour, it correlated with less favorable posttransplant outcomes. The 24-hour beta cell-specific mortality was also a good predictor of islet engraftment as subjects with no beta cell mortality at this time point were more likely to be insulin independent and have a favorable Beta 2 score at 3 months posttransplant.

A clear limitation of this study is that blood for early cfDNA measurements was only drawn at 1 and 24 hours posttransplant. As a consequence, our 24 hours beta cell cfDNA values represent a snapshot of graft cell death within that particular hour. Clearly, the interpretation of our cell loss estimates is limited by the lack of more frequent time points within the first 24 hours and may be further influenced by unknown factors such as the islet death rate and the half-life of cfDNA, which is currently estimated between 15 and 120 minutes.39,40 Additional studies with more frequent sampling may clarify the dynamics of graft loss in the critical hours and days after transplantation. Moreover, sampling for cfDNA at different stages of the islet isolation process could also provide more accurate information on the number of beta cells lost to digestion, purification and culture.

Islet culture performed at 22°C has been the standard approach at our site for more than 15 years to minimize islet loss before transplantation as oppose as a 37°C. With the primary goal of clinical islet isolation being the recovery of the highest islet mass possible after purification, we often choose to compromise purity to maximize transplantable islet mass. This particular cfDNA assay may also serve as an evaluation to our current approach to islet isolation and culture leading to refinement of standard operating protocols.

Further clinical studies using cfDNA measurement paired with beta cell DNA kinetic may help estimate islet cell death rate before and after transplantation as a consequence of the process of islet isolation and posttransplant hypoxia, IBMIR, or inflammatory response. All patients in this study received anti-inflammatory treatments (etanercept + anakinra) posttransplant. We have not carried out subanalyses to evaluate cfDNA levels in the absence of these anti-inflammatory treatments, but this would be insightful if data were available.

Only 10% of patients achieve and maintain insulin independence with single donor islet infusions across our entire experience. None of the study subjects fulfilled these criteria at 1 year. However, in a larger study group, it would be interesting to look for associations between cfDNA levels in a subset that did maintain single-donor islet transplant full function beyond 1 year versus others that did not.

In summary, we present a validation of our recently described method to detect beta cell death based on beta cell-specific methylation patterns in circulating DNA. Our results indicate that 24 hours estimation of beta cell death correlates with clinical allotransplantation outcomes and could predict islet engraftment at 3 months. This technique may represent a useful tool to accurately estimate the rate of cell loss after transplantation of islets and other organs, and potentially a means to monitor graft rejection and to optimize immunosuppression. Together with existing clinical and metabolic indicators of islet graft performance, it may contribute to secure long-term graft survival by allowing adequate timely interventions and judicious planning of subsequent islet infusions.


1. Pepper AR, Gala-Lopez B, Ziff O, et al. Revascularization of transplanted pancreatic islets and role of the transplantation site. Clin Dev Immunol. 2013;2013:352315.
2. Ricordi C, Goldstein JS, Balamurugan AN, et al. NIH-sponsored clinical islet transplantation consortium phase 3 trial: manufacture of a complex cellular product at eight processing facilities. Diabetes. 2016;db160234.
3. Al-Adra DP, Gill RS, Imes S, et al. Single-donor islet transplantation and long-term insulin independence in select patients with type 1 diabetes mellitus. Transplantation. 2014;98:1007–1012.
4. Forbes S, Oram RA, Smith A, et al. Validation of the BETA-2 Score: an improved tool to estimate beta cell function after clinical islet transplantation using a single fasting blood sample. Am J Transplant. 2016;16:2704–2713.
5. Akirav EM, Lebastchi J, Galvan EM, et al. Detection of β cell death in diabetes using differentially methylated circulating DNA. Proc Natl Acad Sci U S A. 2011;108:19018–19023.
6. Fisher MM, Watkins RA, Blum J, et al. Elevations in circulating methylated and unmethylated preproinsulin DNA in new-onset type 1 diabetes. Diabetes. 2015;64:3867–3872.
7. Lebastchi J, Deng S, Lebastchi AH, et al. Immune therapy and β-cell death in type 1 diabetes. Diabetes. 2013;62:1676–1680.
8. Husseiny MI, Kaye A, Zebadua E, et al. Tissue-specific methylation of human insulin gene and PCR assay for monitoring beta cell death. PLoS One. 2014;9:e94591.
9. Lehmann-Werman R, Neiman D, Zemmour H, et al. Identification of tissue-specific cell death using methylation patterns of circulating DNA. Proc Natl Acad Sci U S A. 2016;113:E1826–E1834.
10. Wang LJ, Kin T, O'Gorman D, et al. A multicenter study: North American islet donor score in donor pancreas selection for human islet isolation for transplantation. Cell Transplant. 2016;25:1515–1523.
11. Andres A, Kin T, O'Gorman D, et al. Clinical islet isolation and transplantation outcomes with deceased cardiac death donors are similar to neurological determination of death donors. Transpl Int. 2016;29:34–40.
12. Balamurugan AN, Naziruddin B, Lockridge A, et al. Islet product characteristics and factors related to successful human islet transplantation from the Collaborative Islet Transplant Registry (CITR) 1999–2010. Am J Transplant. 2014;14:2595–2606.
13. Kin T. Islet isolation for clinical transplantation. Adv Exp Med Biol. 2010;654:683–710.
14. Ricordi C, Gray DW, Hering BJ, et al. Islet isolation assessment in man and large animals. Acta Diabetol Lat. 1990;27:185–195.
15. Barnett MJ, McGhee-Wilson D, Shapiro AM, et al. Variation in human islet viability based on different membrane integrity stains. Cell Transplant. 2004;13:481–488.
16. Papas KK, Bellin MD, Sutherland DE, et al. Islet oxygen consumption rate (OCR) dose predicts insulin independence in clinical islet autotransplantation. PLoS One. 2015;10:e0134428.
17. Kitzmann JP, O'Gorman D, Kin T, et al. Islet oxygen consumption rate dose predicts insulin independence for first clinical islet allotransplants. Transplant Proc. 2014;46:1985–1988.
18. Papas KK, Pisania A, Wu H, et al. A stirred microchamber for oxygen consumption rate measurements with pancreatic islets. Biotechnol Bioeng. 2007;98:1071–1082.
19. Papas KK, Colton CK, Nelson RA, et al. Human islet oxygen consumption rate and DNA measurements predict diabetes reversal in nude mice. Am J Transplant. 2007;7:707–713.
20. Pepper AR, Hasilo CP, Melling CW, et al. The islet size to oxygen consumption ratio reliably predicts reversal of diabetes posttransplant. Cell Transplant. 2012;21:2797–2804.
21. Kawahara T, Kin T, Kashkoush S, et al. Portal vein thrombosis is a potentially preventable complication in clinical islet transplantation. Am J Transplant. 2011;11:2700–2707.
22. Koh A, Senior P, Salam A, et al. Insulin-heparin infusions peritransplant substantially improve single-donor clinical islet transplant success. Transplantation. 2010;89:465–471.
23. Liwski R, Pochinco D, Tinckam K, et al. 30-OR: Canada-wide evaluation of rapid optimized flow crossmatch (ROFCXM) protocol. Hum Immunol. 2012;73(Suppl):26.
24. Finsterbusch L, Hartmann H. Fluid balance of the calf. 1. Simultaneous determination of the volumes for total body water and blood plasma using isotope dilution methods and results with body water excretion. Arch Exp Veterinarmed. 1982;36:851–862.
25. Johnson HL, Virk SP, Mayclin P, et al. Predicting total body water and extracellular fluid volumes from bioelectrical measurements of the human body. J Am Coll Nutr. 1992;11:539–547.
26. Watson PE, Watson ID, Batt RD. Total body water volumes for adult males and females estimated from simple anthropometric measurements. Am J Clin Nutr. 1980;33:27–39.
27. Pisania A, Weir GC, O'Neil JJ, et al. Quantitative analysis of cell composition and purity of human pancreatic islet preparations. Lab Invest. 2010;90:1661–1675.
28. Paty BW, Senior PA, Lakey JR, et al. Assessment of glycemic control after islet transplantation using the continuous glucose monitor in insulin-independent versus insulin-requiring type 1 diabetes subjects. Diabetes Technol Ther. 2006;8:165–173.
29. Kin T, Senior P, O'Gorman D, et al. Risk factors for islet loss during culture prior to transplantation. Transpl Int. 2008;21:1029–1035.
30. Johansson H, Goto M, Dufrane D, et al. Low molecular weight dextran sulfate: a strong candidate drug to block IBMIR in clinical islet transplantation. Am J Transplant. 2006;6:305–312.
31. Pawelec K, Juszczak MT, Kumar A, et al. Time course of islet loss after intraportal transplantation. Ann N Y Acad Sci. 2008;1150:230–233.
32. Shapiro AM. Strategies toward single-donor islets of Langerhans transplantation. Curr Opin Organ Transplant. 2011;16:627–631.
33. Rickels MR, Liu C, Shlansky-Goldberg RD, et al. Improvement in beta-cell secretory capacity after human islet transplantation according to the CIT07 protocol. Diabetes. 2013;62:2890–2897.
34. Ritz-Laser B, Oberholzer J, Toso C, et al. Molecular detection of circulating beta-cells after islet transplantation. Diabetes. 2002;51:557–561.
35. Bellin MD, Clark P, Usmani-Brown S, et al. Unmethylated insulin DNA is elevated after total pancreatectomy with islet autotransplantation: assessment of a novel beta cell marker. Am J Transplant. 2016:14054.
36. Kawahara T, Kin T, Shapiro AM. A comparison of islet autotransplantation with allotransplantation and factors elevating acute portal pressure in clinical islet transplantation. J Hepatobiliary Pancreat Sci. 2012;19:281–288.
37. Rafael E, Ryan EA, Paty BW, et al. Changes in liver enzymes after clinical islet transplantation. Transplantation. 2003;76:1280–1284.
38. Barshes NR, Lee TC, Goodpastor SE, et al. Transaminitis after pancreatic islet transplantation. J Am Coll Surg. 2005;200:353–361.
39. Diehl F, Schmidt K, Choti MA, et al. Circulating mutant DNA to assess tumor dynamics. Nat Med. 2008;14:985–990.
40. Lo YM, Zhang J, Leung TN, et al. Rapid clearance of fetal DNA from maternal plasma. Am J Hum Genet. 1999;64:218–224.

Supplemental Digital Content

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.