Type 1 diabetes — clinical trials landscape¶
TL;DR — T1D trials now span stage-1/2 prevention, new-onset immune preservation, glucose automation, hypoglycemia rescue, and renewable cell replacement. A registry entry proves a protocol exists, not efficacy; status can change and must be date-stamped. Teplizumab is the sole stage-2 program here with randomized delay evidence (HR 0.41, 95% CI 0.22–0.78), while zimislecel has early uncontrolled insulin-independence evidence and has entered phase 3 (Herold 2019, PMID 31180194; Reichman 2025, PMID 40544428; NCT01030861; NCT04786262). Combination trials reflect the field’s move from single transient perturbations toward sequenced immune and β-cell-directed therapy.
Selected registry-verified portfolio¶
ClinicalTrials.gov records below were re-fetched live on 2026-08-30.
| NCT | Program | Stage | Phase/status at query | Evidence boundary |
|---|---|---|---|---|
| NCT01030861 | Teplizumab TN-10 | Stage 2 | Phase 2; completed | Published delay RCT |
| NCT03875729 | Teplizumab PROTECT | New onset | Phase 3; completed | Published C-peptide benefit |
| NCT05757713 | Teplizumab pediatric stage 2 | Stage 2 | Phase 4; active, not recruiting | Post-approval evidence |
| NCT03929601 | Rituximab + abatacept | New onset | Phase 2; active, not recruiting | Sequence vs rituximab alone |
| NCT06455319 | Precision ATG ± verapamil | New onset | Phase 2; recruiting | Immune + β-cell protection |
| NCT07216391 | Teplizumab vs ATG platform | Stage 2 | Phase 2; not yet recruiting | Comparative prevention concept |
| NCT01773707 | Abatacept prevention | Stage 1 | Phase 2; completed | Published negative primary endpoint |
| NCT00279305 | Rituximab | New onset | Phase 2; completed | Published C-peptide signal |
| NCT00505375 | Abatacept | New onset | Phase 2; completed | Published transient C-peptide preservation |
| NCT02215200 | ATG-GCSF | New onset | Phase 2; completed | Published component findings |
| NCT04786262 | Zimislecel/VX-880 | Established severe T1D | Phase 3; recruiting | Early phase 1–2 report |
| NCT06832410 | Zimislecel after kidney transplant | Established T1D | Phase 3; recruiting | Distinct risk-benefit population |
| NCT05791201 | Encapsulated VX-264 | Established T1D | Phase 1/2; active, not recruiting | No efficacy conclusion |
| NCT06239636 | Gene-edited hypoimmune primary islets | Established T1D | Early phase 1; recruiting | Two-participant safety study; sponsor status last verified 2024-12 |
| NCT02940873 | HARPdoc | Problematic hypoglycemia | Completed | Published education RCT |
| NCT02536950 | Insulin-only bionic pancreas | Established T1D | Completed | Technology development |
Prevention and preservation¶
Teplizumab TN-10 randomized 76 relatives and delayed median diagnosis by about two years (Herold 2019, PMID 31180194). Stage-1 abatacept did not significantly reduce progression to dysglycemia/stage 3 (HR 0.702, 95% CI 0.452–1.09), despite immune and C-peptide effects (Russell 2023, PMID 36920087).
Oral insulin failed its primary prevention stratum, illustrating the difference between plausible antigen-specific tolerance and demonstrated clinical delay (TrialNet Oral Insulin Study Group 2017, PMID 29164254).
| Failure pattern | Example | Trial-design lesson |
|---|---|---|
| Primary endpoint negative | Oral insulin, stage-1 abatacept | Do not promote subgroup/biomarker results to efficacy |
| Surrogate positive, clinical endpoints neutral | PROTECT secondary outcomes | Extend follow-up and power patient outcomes |
| Effect wanes | Immune-cell reconstitution after rituximab | Test sequence/repeat strategies |
| Added agent fails to add | ATG + GCSF | Combinations need factorial/attribution logic |
| Selected population | Relative-based prevention | Validate general-population screen-detected cohorts |
Replacement pipeline¶
Donor-islet transplantation achieved the composite endpoint in 87.5% at one year among 48 selected adults but required immunosuppression (Hering 2016, PMID 27208344). Zimislecel generated C-peptide in all 14 early recipients and insulin independence in 10/12 full-dose recipients at one year, with neutropenia and deaths requiring careful longer follow-up (Reichman 2025, PMID 40544428).
The pivotal distinction is exposed versus protected cells. Zimislecel uses systemic immunosuppression; VX-264 tests a device intended to shield cells. Neither phase/status nor corporate nomenclature substitutes for peer-reviewed outcomes.
Technology trials¶
Technology studies often use TIR, HbA1c, and time below range over weeks to months. iDCL’s 11-point TIR gain is clinically interpretable, but system-specific performance, onboarding support, and rapidly changing comparators limit cross-era comparison (Brown 2019, PMID 31618560).
Pediatric meta-analysis and network meta-analysis aggregate different algorithms and comparators; heterogeneity should be preserved rather than presenting a class effect as device interchangeability (Zeng 2023, PMID 38011519; Di Molfetta 2024, PMID 39298688).
Trial endpoints by domain¶
| Domain | Preferred endpoints | Required safety outcomes |
|---|---|---|
| Prevention | Time to stage 3; stage progression | Infection, cytokine effects, psychosocial burden |
| New onset | Stimulated C-peptide AUC | Severe events, immune toxicity |
| AID | TIR, HbA1c, time <54 | Severe hypoglycemia, DKA, device events |
| Hypoglycemia rescue | Recovery time/treatment success | Rebound, nausea, failure to administer |
| Cell replacement | Severe-event freedom + HbA1c, insulin independence | Death, infection, renal function, graft/procedure toxicity |
Why trials succeed or fail¶
- T1D is heterogeneous in age, tempo, immune phenotype, and residual function.
- Stage-1/2 trials require long surveillance and tolerate less toxicity.
- New-onset C-peptide decline creates a limited intervention window.
- Background CGM/AID reduces event rates and changes comparators.
- Small replacement cohorts can show function but not characterize rare harms.
- Registries can be outdated; publications and regulatory data must be reconciled.
- Commercial endpoints and mechanistic endpoints may answer different questions.
Effect sizes across trial domains¶
| Trial/domain | Sample | Primary quantitative result | Interpretation |
|---|---|---|---|
| TN-10 stage-2 teplizumab | 76 | HR 0.41 (95% CI 0.22–0.78); median 48.4 vs 24.4 months | Positive progression trial |
| Oral insulin primary stratum | 389 | HR 0.87; P=0.21 | Negative; secondary stratum exploratory |
| Stage-1 abatacept | 212 | HR 0.702 (0.452–1.09) | Negative primary progression endpoint |
| New-onset low-dose ATG | 89 | C-peptide 0.646 vs 0.406 nmol/L at one year | Positive preservation; GCSF non-additive |
| Pediatric verapamil | 88 | AUC difference 0.14 pmol/mL (0.01–0.27) | Partial β-cell preservation |
| Tight AID at diagnosis | 113 | C-peptide difference −0.01 (−0.11 to 0.10) | TIR improved, preservation absent |
| TRIMECO islets | 47 treated as allocated | β-score ≥6: 64% (43–82) vs 0% (0–15) | Metabolic efficacy with renal/procedure harm |
| Young-child closed loop | 102 | TIR +12.4 points (9.5–15.3) | Strong surrogate efficacy; rare-event uncertainty |
These estimates should not be ranked as if they answer one question. Prevention trials estimate time-to-diagnosis, preservation trials estimate stimulated secretion, technology trials estimate glucose exposure, and replacement trials combine graft function with severe-event outcomes.
Registry status versus published evidence¶
| Registry state on 2026-08-30 | What can be concluded | What cannot |
|---|---|---|
| Completed | Enrollment/intervention ended per sponsor record | Results are positive or peer reviewed |
| Recruiting | Site recruitment is reported open | Recruitment is global, rapid, or ultimately successful |
| Active, not recruiting | Follow-up or analysis continues | Product remains clinically available |
| Not yet recruiting | Protocol is registered | Enrollment date or feasibility is assured |
The live query found NCT04786262, NCT06239636, NCT06455319, and NCT06832410 recruiting; NCT03929601, NCT05757713, and NCT05791201 active but not recruiting; and NCT07216391 not yet recruiting. All other tabled records were completed. NCT06239636's sponsor verification date remains 2024-12 despite a recruiting label, illustrating why status is a dated administrative observation rather than an efficacy or operational-performance signal.
Primary-endpoint discipline¶
Oral insulin failed its prespecified primary stratum (58/203 vs 62/186 diagnoses; HR 0.87, P=0.21), although a 55-person secondary stratum yielded HR 0.45 (Krischer 2017, PMID 29164254). Stage-1 abatacept likewise missed progression significance even though later modeling identified a high-secretor subgroup with HR 0.46 (Galderisi 2026, PMID 41237315). Both examples require prospective confirmation of subgroup hypotheses.
Low-dose ATG shows why factorial attribution matters: ATG alone preserved C-peptide at one and two years, but adding GCSF did not meet the prespecified C-peptide criterion (Haller 2018, PMID 30012675; Haller 2019, PMID 30967424). A combination arm does not prove synergy merely because both agents are biologically plausible.
Comparator drift¶
The comparator for severe unstable T1D has changed. TRIMECO showed islet efficacy versus intensive insulin therapy, but contemporary real-world AID synthesis estimates TIR +11.61 points (95% CI 10.47–12.76) and HbA1c −0.42% (−0.47 to −0.37) versus prior therapy (Lablanche 2018, PMID 29776895; Yang 2024, PMID 38888056). Replacement programs should therefore measure incremental severe-event freedom, burden, renal harm, and durability against optimized AID.
CGM itself has changed control arms. Across 22 trials, CGM versus SMBG reduced HbA1c −0.23% (95% CI −0.35 to −0.10), with larger benefit at baseline HbA1c >8%, while severe hypoglycemia and DKA remained too rare for firm conclusions (Teo 2022, PMID 35141761).
Trial representativeness¶
| Selection | Efficiency gained | Generalizability lost |
|---|---|---|
| Relatives with multiple antibodies | Higher stage-3 event rate | Most population cases lack family history |
| Recent-onset residual C-peptide | Detectable preservation window | Excludes later presenters and rapid loss |
| Severe hypoglycemia despite expert care | Larger replacement benefit | Not routine T1D population |
| High baseline HbA1c | Larger technology signal | Smaller incremental effect in well-controlled users |
| Technology-capable trial sites | Reliable data and support | Underrepresents supply, literacy, and connectivity barriers |
Response heterogeneity can be biological or operational. A routine-measure Beta2 score predicted immunotherapy response with AUC 0.765 in validation datasets, but this remains an adaptive-design tool until prospectively validated (So 2025, PMID 40424079). For AID, qualitative evidence documents meal, exercise, trust, alarm, and supply burdens that TIR alone does not capture (Garza 2025, PMID 39212346).
Minimum endpoint sets¶
| Domain | Efficacy core | Patient-experienced core | Long-term safety core |
|---|---|---|---|
| Stage 1/2 | Stage-free survival; metabolic trajectory | Anxiety, monitoring burden, days without insulin | Infection, malignancy, immune sequelae |
| New onset | Standardized stimulated C-peptide | Insulin dose, TIR, severe events, burden | Immune toxicity and off-treatment durability |
| AID | TIR, HbA1c, time <54 | Sleep, alarms, discontinuation, trust | DKA, severe hypoglycemia, device failures |
| Replacement | Graft function and insulin independence | Severe-event freedom and treatment burden | Death, renal function, infection, neoplasm, sensitization |
Cell-replacement trials especially need denominators that distinguish all enrolled, transplanted, full-dose, evaluable, and insulin-independent participants. Early function in every recipient can coexist with incomplete full-dose follow-up and serious adverse events (Reichman 2025, PMID 40544428).
Open questions¶
- What adaptive platform design can compare sequences while preserving interpretable safety? (NCT07216391)
- Which biomarker should enrich stage-2 trials without excluding slow progressors who benefit? (Weiss 2022, PMID 36028774)
- What minimum follow-up establishes durable cell-replacement net benefit? (Reichman 2025, PMID 40544428)
- How should AID trials measure burden, discontinuation, and access alongside TIR? (Canha 2025, PMID 39726162)
Related pages¶
- disease-modifying immunotherapy — intervention interpretation.
- beta-cell replacement — cell products and immune barrier.
- insulin therapy and technology — device evidence.
- biomarkers — trial endpoints.
References¶
- Hering BJ, et al. Phase 3 Trial of Human Islets. Diabetes Care. 2016;39:1230-1240. PMID 27208344
- TrialNet Oral Insulin Study Group. Oral Insulin Prevention Trial. JAMA. 2017;318:1891-1902. PMID 29164254
- Herold KC, et al. Teplizumab in Relatives at Risk. N Engl J Med. 2019;381:603-613. PMID 31180194
- Brown SA, et al. Six-Month Closed-Loop Trial. N Engl J Med. 2019;381:1707-1717. PMID 31618560
- Russell WE, et al. Abatacept for Stage-1 Progression. Diabetes Care. 2023;46:1005-1013. PMID 36920087
- Zeng B, et al. AID in Children and Adolescents. Diabetes Care. 2023. PMID 38011519
- Di Molfetta S, et al. Hybrid Closed Loop Network Meta-analysis. Diabetes Metab Res Rev. 2024. PMID 39298688
- Reichman TW, et al. Stem Cell-Derived Fully Differentiated Islets. N Engl J Med. 2025;393:858-868. PMID 40544428
- Weiss A, et al. Progression likelihood score. Diabetologia. 2022;65:2121-2131. PMID 36028774
- Canha D, et al. AID and diabetes distress. Diabet Med. 2025. PMID 39726162
- Haller MJ, et al. Low-Dose ATG Preserves β-Cell Function and Improves HbA1c. Diabetes Care. 2018;41:1917-1925. PMID 30012675
- Haller MJ, et al. Low-Dose ATG: Two-Year Clinical Trial Data. Diabetes. 2019;68:1267-1276. PMID 30967424
- Forlenza GP, et al. Effect of Verapamil on Pancreatic Beta Cell Function. JAMA. 2023;329:990-999. PMID 36826844
- McVean J, et al. Tight Glycemic Control and Pancreatic Beta Cell Function. JAMA. 2023;329:980-989. PMID 36826834
- Lablanche S, et al. Islet transplantation versus insulin therapy: TRIMECO. Lancet Diabetes Endocrinol. 2018;6:527-537. PMID 29776895
- Wadwa RP, et al. Trial of Hybrid Closed-Loop Control in Young Children. N Engl J Med. 2023;388:991-1001. PMID 36920756
- Galderisi A, et al. Baseline Insulin Secretion Determines Response to Abatacept. Diabetes. 2026;75:229-240. PMID 41237315
- Yang Q, et al. Real-world glycaemic outcomes of automated insulin delivery. Diabetes Obes Metab. 2024;26:3753-3763. PMID 38888056
- Teo E, et al. Effectiveness of continuous glucose monitoring in type 1 diabetes. Diabetologia. 2022;65:604-619. PMID 35141761
- So M, et al. β-Cell Function Derived From Routine Measures Predicts Immunotherapy Response. Diabetes Care. 2025;48:1370-1376. PMID 40424079
- Garza KP, et al. Adult's Lived Experience Using the Insulin-Only Bionic Pancreas. J Diabetes Sci Technol. 2025;19:11-17. PMID 39212346
- Barton FB, et al. Improvement in outcomes of clinical islet transplantation. Diabetes Care. 2012;35:1436-1445. PMID 22723582
- Vantyghem MC, et al. Ten-Year Outcome of Islet Transplantation. Diabetes Care. 2019;42:2042-2049. PMID 31615852
- Perdigoto AL, et al. Teplizumab follow-up after 7 years. Diabetologia. 2019;62:655-664. PMID 30569273
- Gubitosi-Klug RA, et al. Residual β cell function in long-term type 1 diabetes. J Clin Invest. 2021;131. PMID 33529168