β-Cell Benchmarks: Defining Predictive Outcomes in Islet Transplantation Free

Islet transplantation, where islets are transplanted into the liver of selected adults with type 1 diabetes with problematic glycemic control, is an evidence-based procedure that prevents severe hypoglycemia (SH), stabilizes glycemic control, and improves quality of life (1–3). Following isolation, pancreatic islets, initially avascular, are transplanted via intraportal infusion, where they embed within the hepatic sinusoids and undergo gradual revascularization through the hepatic arterial system over several weeks. This process, known as engraftment, is critical for clinical success, with outcomes closely associated with the engrafted functional β-cell mass (4,5).

The functional β-cell mass can be assessed in a number of ways, including with a stimulated mixed-meal tolerance test (MMTT), with C-peptide and associated glycemic metrics, and with composite scores of β-cell function (6,7). However, the gold standard for assessment of β-cell secretory capacity is the glucose potentiation of arginine-induced insulin secretion (AIR_pot) determined during a hyperglycemic clamp. Attrition in islet graft function is seen over time, but few studies have conducted long-term metabolic assessment of AIR_pot to track the rate of decline in functional β-cell mass.

In this issue of Diabetes, Flatt et al. (8) extend our understanding of β-cell secretory capacity and its impact on metabolic outcomes over a 6-year period, providing valuable insights into optimizing clinical practices and guiding future interventions in islet transplantation (Fig. 1).

The study investigates 11 individuals (4 male, 7 female) with type 1 diabetes receiving one or two islet infusions of 10,090 ± 2,206 (mean ± SD) islet equivalents/kg, with peritransplant, a T-cell–depleting antibody, tumor necrosis factor-α inhibition (etanercept), heparinization, and intensive insulin therapy; a second islet infusion was given under the interleukin-2 receptor antagonist basiliximab plus etanercept if the individual was insulin dependent at day 75; maintenance immunosuppression included the calcineurin inhibitor tacrolimus at low dose and the mTOR inhibitor sirolimus (9). In carefully conducted studies, AIR_pot determined with a hyperglycemic (230 mg/dL) clamp, along with MMTT testing and blinded continuous glucose monitoring (CGM) monitoring for 3–7 days, were simultaneously assessed at day 75 posttransplant and then annually for a median (interquartile range) of 6 (5–7) years. An age-, sex-, and BMI-matched glucose-tolerant group underwent clamp studies to derive the “normal” 100% AIR_pot response in order to calculate the percentage of β-cell secretory capacity and β-cell function in islet transplant recipients. Receiver operating characteristic curves were generated to assess the ability of the functional β-cell mass to predict insulin independence, defined as >7 days off insulin with good glycemic control as well as MMTT and CGM threshold predictors.

All participants became insulin independent initially, but only three of five maintained this state after 6 years, with HbA_1c and insulin requirements increasing with time. Islet β-cell graft function assessed by the BETA-2 score (6), a validated composite score of islet function incorporating fasting glucose, C-peptide, HbA_1c, and insulin dose, also decreased over time, with a mean value <15, representing the published threshold associated with insulin independence, which was observed late in the study at 6–7 years. AIR_pot was lower in islet transplant recipients than controls with normal glucose tolerance and remained stable for the first 4 years posttransplant but subsequently declined. MMTT C-peptide and the MMTT C-peptide–to–glucose ratio decreased after 3 years. The proinsulin-to-insulin secretory ratio under glucose-potentiated conditions increased over time.

A functional β-cell mass of >40% of normal was strongly predictive of sustained insulin independence, with 100% specificity and 52% sensitivity. Good glycemic outcomes, including ≥90% time in range (70–180 mg/dL) and ≤1% time below 60 mg/dL, were achieved with a functional β-cell mass of >20%, although sustained insulin independence required higher reserves. Glycemic variability improved with higher β-cell mass, emphasizing the need for functional reserve. A 60-min MMTT C-peptide–to–glucose ratio threshold of 0.13 nmol/mmol and a BETA-2 score of 19 showed strong predictive power for β-cell mass of >40% and insulin independence. De novo donor-specific antibody was detected in two individuals with the codevelopment of autoantibodies against specific pancreatic β-cell autoantigens: in one case there was an increase in glutamic acid decarboxylase-65 (GAD-65) autoantibody, and in the other case there was an increase in islet cell autoantibody. One transplant recipient experienced de novo GAD-65 autoantibody formation with subsequent graft failure, and another recipient experienced islet cell autoantibody and GAD-65 autoantibody formation with diminished islet function.

The authors are to be congratulated on their careful longitudinal analysis to evaluate β-cell function using gold standard methods with glucose-potentiated arginine testing alongside other functional assessments to 6 years posttransplant, with the important inclusion of a glucose-tolerant group to benchmark normal β-cell function. The inclusion of functional assessments is important, as they can be carried out in all centers, whereas AIR_pot determination needs considerable resources and is generally only performed in research centers. The study underscores the pivotal role that the functional β-cell mass plays in predicting sustained insulin independence and glycemic control and gives clear thresholds to aim for: >40% is associated with sustained insulin independence, and >20% is associated with excellent glycemic control over the longer term. Achieving appropriate numbers of transplanted islets is important for achievement and maintenance of glycemic control, with most programs aiming to transplant >10,000 islet equivalents per kg of body weight (4), but with this protocol ∼25% of islets may fail to engraft secondary to inflammation, alloimmunity, and autoimmunity, and insulin independence is not consistently achieved with this dose, highlighting that the interplay between donor and recipient factors and immunosuppression is of central importance. Factors in addition to numbers of islets transplanted that are known to be associated with favorable outcomes include recipient age ≥35 years, induction immunosuppression with T-cell depletion, and/or tumor necrosis factor-α inhibition and maintenance with both mTOR and calcineurin inhibitors (4). Limitations in current analytical techniques hinder detection of transient in vivo changes as well as objective assessments of islet viability, so what drives good engraftment is not fully elucidated. The impact of the purity of the islet preparation (10), donor and recipient sex (11), alternative immunosuppressive regimens, age-related adiposity, and donor HLA susceptibility antigens on graft survival (12) all require further investigation.

The functional β-cell mass, representing the engrafted islet mass, is a more personalized approach that accounts for multiple donor and recipient factors. The study supports and extends findings from a recent Collaborative Islet Transplant Registry (CITR) study that showed that the BETA-2 score at 28 days posttransplant was independently associated with 5-year metabolic outcomes (5). The BETA-2 score of 19, which predicted insulin independence, was higher than that reported in previous studies (1,10), but the resulting insulin independence was more durable (1,13); the 60-min stimulated MMTT C-peptide–to–glucose threshold of 0.13 nmol/mmol for predicting insulin independence was the same as the 90-min value reported in a recent CITR study (7).

The current study demonstrates diminished islet graft function in recipients at 3 years with this induction-immunosuppression regimen (4). Integrating data on immunological markers, the study highlights autoimmunity as the predominant factor in graft failure occurring >12 months posttransplant. GAD-65 autoantibodies arise as a consequence of β-cell stress or death, leading to antigen release and presentation to B cells (14,15), with increased titers associated with reduced graft survival (16). It is of interest that alloimmunity and autoimmunity coexisted, consistent with the concept that autoimmunity may trigger alloimmunity (17). Monitoring of cellular autoimmunity was not undertaken but has shown potential in predicting clinical outcomes of islet and pancreas transplantation, with strong associations between the presence of CD4⁺ and/or CD8⁺ T-cell autoreactivity and graft failure (18), and could usefully be included in future studies, although such monitoring is difficult to perform and interpret over years in the context of adequacy of immunosuppression and intercurrent infection. Nevertheless, there remains a critical need for sensitive biomarkers for early detection of graft dysfunction, which would enable early intervention with tailored therapies that could mitigate long-term losses.

Induction with T-cell–depleting agents and maintenance with tacrolimus and sirolimus were largely effective in controlling alloimmunity but come with trade-offs. Tacrolimus induces insulin resistance, exacerbating β-cell stress: the increased proinsulin secretory ratios seen here suggest metabolic stress and graft exhaustion in at least two of the transplant recipients (19). Tacrolimus may cause nephrotoxicity, and here it was associated with a modest decline in estimated glomerular filtration rate (20). Immunosuppression is associated with an increased risk of infection as well as a number of different cancers, notably skin cancers (4). Optimizing immunosuppression regimens to balance graft protection while minimizing risks is paramount. There is an urgent need to explore alternative immunosuppression protocols and adjunctive therapies that suppress recurrent autoimmunity and invoke tolerance (12,21) without reliance on calcineurin inhibitors, which could substantially improve metabolic outcomes and decrease risks from immunosuppression.

The weakness of the study is the small cohort size, with 11 participants, and attrition in follow-up further narrows the data set, potentially introducing bias. Furthermore, the findings may not be generalizable to immunosuppression protocols seen across other centers.

The implications of this work are translatable to stem cell–derived islet transplantation (22). These therapies promise unlimited access to insulin-producing cells with potentially reduced immunogenicity. Identifying a functional β-cell mass of >40% posttransplant as critical for sustained insulin independence offers a practical target with predictive value for clinical stem cell islet transplantation programs.

In the future, incorporating functional β-cell mass benchmarks will help predict outcomes and guide intrahepatic retransplantation strategies in our patients. Equally important is the development of biomarkers to detect recurrence of autoimmunity early and strategies to mitigate this to enhance graft durability, alongside safer, efficacious immunosuppressive regimens. These advancements hold the potential to overcome long-standing challenges and significantly improve outcomes in the field of islet transplantation.