Encapsulated stem cell-derived β cells exert glucose control in patients with type 1 diabetes

Type 1 diabetes (T1D) is characterized by a lack of insulin-producing β cells in the pancreas, leading to insufficient insulin regulation. While exogenous insulin mitigates hyperglycemia, it doesn't replicate the precise cellular control of insulin synthesis and release, posing risks of hypoglycemia and impacting quality of life. Intrahepatic islet transplantation from human donors has shown promise in resolving hypoglycemia and improving glucose control, but is limited by donor cell scarcity and the need for immunosuppression. Encapsulated β cells derived from human pluripotent stem cells (hPSCs) offer a potential solution, addressing both limitations. Methods for differentiating hPSCs into pancreatic endoderm (PE) and β cells have been developed, with successful transplantation in immunocompromised rodents. Cell encapsulation protects both recipient and donor cells from immune responses and allows for device retrieval and analysis. However, maintaining transplanted β cell viability within sealed encapsulation devices requires strategies like perforated membranes enabling capillary ingrowth. Previous studies using PEC-Direct (a combination product of pancreatic endoderm cells (PEC-01) in devices with perforated membranes) in immunosuppressed T1D patients demonstrated formation of functional β cells, but C-peptide levels remained insufficient for clinical benefit. This study investigated whether increasing cell dose and optimizing membrane perforation patterns could improve efficacy.

Extensive research has explored β-cell replacement therapy as a potential cure for T1D. Intrahepatic transplantation of islets from human donor pancreases has demonstrated significant improvements in glycemic control and quality of life, virtually normalizing glucose control and eliminating the need for exogenous insulin in a substantial percentage of patients over several years. However, the limited availability of donor cells and the need for continuous immunosuppression remain major obstacles. In vitro differentiation of hPSCs into pancreatic endoderm and subsequently into β cells has shown promising results in preclinical models. Cell encapsulation techniques have been developed to protect transplanted cells from immune rejection, offering a potential approach to overcome the challenges associated with islet transplantation. Early clinical trials using encapsulated hPSC-derived cells showed some evidence of β-cell formation but lacked the efficacy needed to significantly improve glycemic control. This study builds upon these prior efforts by addressing the limitations of the previous trials.

This phase 1/2, open-label, multicenter trial (NCT03163511) investigated the efficacy of subcutaneous PEC-Direct implants in T1D patients under immunosuppression (anti-thymocyte globulin induction, mycophenolate mofetil and tacrolimus maintenance). Ten patients with undetectable baseline C-peptide received twofold to threefold more devices than in previous studies, with optimized membrane perforation patterns. The primary endpoint was detectable plasma C-peptide at month 6. Secondary endpoints included C-peptide levels >0.07 nmol l⁻¹, improved CGM measures (TIR, TAR, TBR, GMI), reduced insulin dosing, and safety assessments. Mixed meal tolerance tests (MMTTs) were performed at 3-month intervals to measure C-peptide levels. Continuous glucose monitoring (CGM) data provided comprehensive glycemic control information. Sentinel devices were retrieved for analysis of cell composition. Immunosuppressive treatment was monitored and adjusted as needed. Histological analysis of retrieved devices utilized techniques such as RNAscope (to identify donor cells), hematoxylin and eosin staining, and immunohistochemistry (to identify various cell types and markers including insulin, glucagon, somatostatin, cytokeratin, chromogranin A, and CD34). Morphometric analysis was performed to quantify the volume of different cell types within the implants. The study assessed the relationship between C-peptide levels, CGM data, insulin dosing, and the composition of the retrieved implants.

Three of ten patients achieved metabolically significant C-peptide levels (≥0.1 nmol l⁻¹) from month 6, maintained until month 12. These patients also showed improvements in CGM measures (increased TIR, decreased TAR) and reduced insulin dosing. The patient with the highest C-peptide response (0.23 nmol l⁻¹) achieved a remarkable increase in TIR from 55% to 85% at month 12. Analysis of retrieved sentinel implants revealed that the β-cell mass in the best-performing patient (case 1) at month 6 represented only 4% of the initial cell mass and 3% of total cells within the device, significantly less than the α-cell mass (16%). A prolonged MMTT in this patient showed sustained elevation of proinsulin and C-peptide, indicating continued β-cell activation but insufficient hormone reserves. The six patients who did not achieve the primary endpoint (detectable C-peptide at month 6) showed no corresponding improvements in glucose control or reduction in insulin dose, except for one patient where improved control was attributed to increased insulin dosage and another where lifestyle changes influenced glycemic levels. Comparison between groups who achieved primary and secondary efficacy endpoints vs those who did not demonstrated significant differences in CGM metrics favoring the group with higher C-peptide levels. Histological analysis of retrieved implants revealed clusters of insulin-positive cells associated with capillaries, indicating functional secretory units. However, a substantial proportion of donor cells were lost during the process, with considerable infiltration of recipient cells. The relative proportion of α-cells to β-cells was substantially higher.

This study demonstrates the potential of device-delivered stem cell-derived β cells to improve glucose control in T1D patients. The results support further development of hPSC-derived PECs as a β-cell replacement therapy. The observed improvement in glucose control in three patients correlated with C-peptide levels ≥0.1 nmol l⁻¹, indicating a metabolically significant effect of the implanted β cells. The longitudinal analysis of individual patients and the absence of confounding factors strengthen the conclusion that the improvement is attributable to the implant function. While the glucose control achieved is less pronounced than in rodent models, the study highlights the feasibility of achieving a glucose-controlling effect with subcutaneously implanted stem cell-derived β cells in a retrievable device. The retrieval feature allows for in-situ analysis that provides valuable insights for future optimization. The low recovery of initial cell mass in the human setting is a key area for further investigation.

This study demonstrates the feasibility of achieving glucose control using subcutaneously implanted, stem cell-derived β cells in a retrievable device in patients with T1D. The findings reveal a correlation between C-peptide levels and glucose control improvements, suggesting areas for optimization. Further research should focus on increasing β-cell survival and differentiation while reducing α-cell mass, potentially through genetic modification and other approaches. The data gathered from retrieved devices can guide future efforts to improve efficacy.

The relatively small sample size (n=10) limits the generalizability of the findings. The use of immunosuppression could have influenced the results, and the introduction of liraglutide in one patient introduces a potential confounder. The study focused on a specific device design and cell dose; further research is needed to explore the effects of different device configurations and cell doses. The study's duration of 1 year may not reflect long-term effects.