Open questions — type 1 diabetes¶
Last curated: 2026-08-30
Questions are stable identifiers, not claims that no study exists. Tier 1 questions could change practice and are designable now; Tier 2 questions require enabling methods, longer horizons, or stronger preliminary evidence.
Tier 1 — practice-changing and designable today¶
OQ-1 — Can teplizumab delay become durable prevention?¶
TN-10 shifted median stage-3 diagnosis from 24.4 to 48.4 months (HR 0.41, 95% CI 0.22–0.78), but did not establish a treatment-free plateau. A repeat-course or adaptive sequence trial could test durable stage-3-free survival and cumulative toxicity (Herold 2019, PMID 31180194).
OQ-2 — Should intervention begin in stage 1 rather than stage 2?¶
Stage-1 abatacept changed immune and C-peptide measures without significantly reducing progression (HR 0.702, 95% CI 0.452–1.09), while stage-2 teplizumab succeeded. Head-to-head stage/timing strategies are needed (Russell 2023, PMID 36920087; Herold 2019, PMID 31180194).
OQ-3 — Which sequence of immune therapies is additive rather than merely toxic?¶
Rituximab, abatacept, and low-dose ATG each transiently preserved function; ATG plus GCSF did not improve on ATG alone. Factorial or platform trials should compare sequence, interval, and stopping rules (Pescovitz 2009, PMID 19940299; Orban 2011, PMID 21719096; Haller 2018, PMID 30012675).
OQ-4 — Does C-peptide preservation improve outcomes patients experience?¶
PROTECT improved stimulated C-peptide but not insulin dose, HbA1c, TIR, or clinically important hypoglycemia at 78 weeks. Trials need longer severe-event, burden, and complication follow-up (Ramos 2023, PMID 37861217).
OQ-5 — Is general-population screening cost-effective against optimized awareness care?¶
Fr1da found presymptomatic T1D in 0.31% of 90,632 children, while 41.9% of children worldwide present in DKA. Cluster comparisons should test screening, awareness, and combined pathways with cost and psychosocial endpoints (Ziegler 2020, PMID 31990315; Zhang 2026, PMID 42303108).
OQ-6 — What monitoring schedule minimizes both DKA and surveillance burden?¶
Consensus requires confirmation and periodic metabolic follow-up but acknowledges evidence gaps. A risk-adaptive trial can compare intervals using stage-3 condition, DKA, visits, anxiety, and costs (Phillip 2024, PMID 38912694; Weiss 2022, PMID 36028774).
OQ-7 — Can zimislecel’s insulin independence be sustained with acceptable net safety?¶
Ten of 12 full-dose recipients were insulin-independent at one year, but the cohort was small, immunosuppressed, and included neutropenia and two deaths. Comparative long-term follow-up is required (Reichman 2025, PMID 40544428).
OQ-8 — Can encapsulation remove systemic immunosuppression?¶
Stem-cell-derived endoderm and mature islets can function in humans. As of the 2026-08-30 PubMed and ClinicalTrials.gov recheck, immune-isolating and hypoimmune approaches have reached clinical investigation, including the two-participant NCT06239636 safety study, but the retrieved evidence does not establish clinically sufficient, durable function without systemic immunosuppression. Retrieval, oxygenation, fibrosis, immune surveillance, and graft function should be co-primary engineering outcomes (Ramzy 2021, PMID 34861146; Reichman 2025, PMID 40544428; Hassanein 2026, PMID 42626948; NCT06239636).
OQ-9 — Do AID systems prevent severe hypoglycemia and DKA at population scale?¶
RCTs show 8.7–11 percentage-point TIR gains, but rare acute events and trial support limit inference. Pragmatic registry-randomized studies should measure severe events, discontinuation, and supply interruption (Brown 2019, PMID 31618560; Ware 2022, PMID 35045227).
OQ-10 — Can meal-announcement-free AID work across exercise and high-fat meals?¶
Hybrid systems remain constrained by insulin kinetics; fully closed-loop qualitative evidence shows benefit and adaptation burden. Crossover trials should stratify meal composition, puberty, exercise, and baseline therapy (Lakshman 2024, PMID 38426909; Di Molfetta 2024, PMID 39298688).
OQ-11 — Which intervention restores hypoglycemia awareness in nonresponders?¶
Structured programs reduce severe events, but persistent impaired awareness remains. Trials should phenotype autonomic response, cognition, sleep, and technology exposure (Amiel 2022, PMID 35484106; Rickels 2019, PMID 31389033).
OQ-12 — Which transition bundle prevents care gaps and DKA?¶
Transition programs are heterogeneous and evidence quality is limited. A health-system trial can randomize tracked appointments, supply bridging, navigation, and mental-health integration (Campbell 2016, PMID 27128768; Garvey 2016, PMID 26681724).
OQ-13 — How can technology rollout close racial and socioeconomic gaps?¶
Young-adult and safety-net cohorts show disparities in pump/CGM use. Default offering, benefit design, navigator support, and clinician-bias feedback are testable implementation strategies (Agarwal 2021, PMID 33155826; Fantasia 2021, PMID 33719610).
OQ-14 — Which intervention reduces recurrent DKA in established T1D?¶
Recurrent DKA may reflect insulin access, pump failure, disordered eating, distress, cognition, or fragmented care. Multicomponent pathways should use causal phenotyping rather than one educational intervention (Cengiz 2013, PMID 23469984; Peterson 2015, PMID 25502449).
Tier 2 — enabling science or long-horizon questions¶
OQ-15 — What initiates autoimmunity in a genetically susceptible child?¶
Prospective viral associations and meta-analysis support enterovirus involvement without establishing sufficiency; microbiome differences may be cause, consequence, or correlate (Lönnrot 2017, PMID 28770319; Isaacs 2023, PMID 37390839; Vatanen 2018, PMID 30356183).
OQ-16 — Why are some β cells resistant to immune attack?¶
Human pancreatic pathology shows patchy insulitis and β-cell heterogeneity. Single-cell spatial studies must connect stress phenotype to clonotypic immune recognition and survival (Pugliese 2016, PMID 27411434; Atkinson 2023, PMID 37478842).
OQ-17 — Can direct β-cell-death markers become clinically quantitative?¶
Unmethylated insulin DNA is detectable by droplet digital PCR, but low abundance and uncertain calibration limit clinical use (Usmani-Brown 2014, PMID 25004096).
OQ-18 — How should progression models handle adult-onset T1D?¶
Most prospective staging evidence comes from childhood cohorts even though the modeled median onset age is 39. Adult longitudinal antibody, C-peptide, and genetic calibration cohorts are needed (Gregory 2022, PMID 36113507; Evans-Molina 2025, PMID 40230204).
OQ-19 — What causes residual cardiovascular risk at target HbA1c?¶
Excess risk persists across glycemic strata and is extreme with early onset. Mechanistic cohorts should integrate cumulative TIR, kidney markers, lipids, inflammation, hypoglycemia, and social exposure (Rawshani 2017, PMID 28402770; Rawshani 2018, PMID 30129464).
OQ-20 — Can hypoimmune cells remain safe under lifelong surveillance?¶
Immune-evasive editing could avoid alloimmunity but may impair tumor or infection surveillance. Reversible safety switches, genomic stability, and retrieval must be demonstrated before broad efficacy trials.
OQ-21 — Which antibody features should set monitoring intensity?¶
Antibody count alone discards identity and titer: type-specific thresholds stratified five-year risk from 6% to 75%, and IA-2A increased progression risk within each formal stage. A prospective comparison should test whether titer/identity-adaptive monitoring reduces DKA and visits without delaying diagnosis (Ng 2022, PMID 34758977; Sims 2025, PMID 40016443).
OQ-22 — What is the right adult staging calibration?¶
At the same HbA1c ≥5.7% threshold, one-year risk among single-antibody-positive TrialNet relatives was 38% in children but 13% in adults. Adult cohorts should compare age-adjusted HbA1c, OGTT, CGM trajectories, C-peptide, antibody phenotype, and hard stage-3 outcomes (Templeman 2026, PMID 42090204).
OQ-23 — Can screening scale without creating unequal surveillance?¶
Specialist cohorts report low DKA with retained monitoring, but cost models depend on assumed DKA reduction and small durable HbA1c effects. Cluster implementation studies should measure loss to follow-up by race, income, geography, language, insurance, and family structure—not only assay uptake (McQueen 2020, PMID 32327420; Martinenghi 2025, PMID 40439773).
OQ-24 — Which people have net psychosocial harm or benefit from early diagnosis and AID?¶
Average AID distress improves and screening anxiety may decline, yet both averages obscure discontinuers, surveillance burden, and identity effects. Trials need prespecified heterogeneity analyses and repeated child/caregiver measures linked to access and clinical outcomes (Canha 2025, PMID 39726162; O'Donnell 2023, PMID 37673098).
OQ-25 — What contemporary comparator justifies cell replacement?¶
Islet transplantation can eliminate severe hypoglycemia for selected candidates but adds portal-procedure, renal, infection, and immunosuppression harm; historical comparators predate modern AID. A pragmatic comparison should use severe-event-free survival, treatment burden, renal function, quality-adjusted life-years, and patient preference rather than insulin independence alone (Lablanche 2018, PMID 29776895; Vantyghem 2019, PMID 31615852).
OQ-26 — How should recurrent DKA be causally phenotyped?¶
Repeated admissions predict mortality, but identical biochemical presentations can arise from insulin insecurity, delivery failure, eating pathology, suicidality, cognitive barriers, or classification error. A rapid-response cohort should prespecify these causes, test adjudication reliability, and match intervention to phenotype (Gibb 2016, PMID 27397023; Santos 2023, PMID 37106409).
Dots not yet connected¶
| # | Dot A | Dot B | The missing junction | Powers |
|---|---|---|---|---|
| D1 | Teplizumab delays stage 3 (PMID 31180194) | Fr1da finds stage 1/2 in the public (PMID 31990315) | Effectiveness and equity of screen-to-infusion pathways outside relatives | OQ-1, OQ-5, OQ-13 |
| D2 | Verapamil preserves C-peptide (PMID 36826844) | Tight AID control does not (PMID 36826834) | β-cell stress signatures predicting metabolic-rest-independent response | OQ-3, OQ-17 |
| D3 | AID reduces low-glucose exposure (PMID 31618560) | Awareness is plastic but sometimes persistent (PMID 31389033) | Whether early AID prevents autonomic failure rather than only episodes | OQ-9, OQ-11 |
| D4 | Stem-cell islets restore secretion (PMID 40544428) | Immune therapies transiently modify autoimmunity (PMID 19940299) | Minimal targeted immune regimen for replacement grafts | OQ-3, OQ-7, OQ-8 |
| D5 | Adult onset is common (PMID 36113507) | Genetic scores improve classification (PMID 30655379) | Prospective ancestry-calibrated adult case-finding pathway | OQ-18 |
| D6 | Early onset drives cardiovascular risk (PMID 30129464) | AID raises TIR (PMID 31618560) | Lifetime vascular effect of childhood AID-era exposure | OQ-9, OQ-19 |
| D7 | Distress changes over time (PMID 27118163) | AID reduces distress on average (PMID 39726162) | Predictive phenotype for benefit versus device burden | OQ-10, OQ-13 |
| D8 | DKA varies 15.6–78.5% by country (PMID 42303108) | Pediatric service access varies globally (PMID 37362165) | Decomposition of awareness, insulin supply, diagnostic, and referral effects | OQ-5, OQ-14 |
| D9 | Enterovirus associates with autoimmunity (PMID 37390839) | HLA/genetic risk stratifies susceptibility (PMID 22315720) | Genotype-by-viral-timing prevention trial | OQ-15 |
| D10 | Insulin omission is diabetes-specific eating pathology (PMID 24022608) | Recurrent DKA is an acute systems endpoint (PMID 23469984) | Nonpunitive real-time pathway linking ketosis patterns to integrated care | OQ-14 |
| D11 | Antibody titers stratify five-year risk 6–75% (PMID 34758977) | Monitoring schedules remain largely consensus-based (PMID 38912694) | Prospective utility trial of titer-adaptive surveillance | OQ-6, OQ-21 |
| D12 | Adult HbA1c thresholds overstate pediatric-equivalent risk (PMID 42090204) | Adult-onset T1D is common and misclassified (PMID 34670785) | Adult-specific staging and case-finding calibration | OQ-18, OQ-22 |
| D13 | Specialist follow-up can nearly eliminate DKA at progression (PMID 40439773) | Population screening cost-effectiveness is assumption-sensitive (PMID 32327420) | Equity-adjusted implementation effectiveness across routine systems | OQ-5, OQ-23 |
| D14 | Transplanted islets can remove severe hypoglycemia (PMID 29776895) | Contemporary AID improves TIR in real-world cohorts (PMID 38888056) | Head-to-head net-benefit threshold for invasive replacement | OQ-7, OQ-25 |
| D15 | Recurrent DKA predicts mortality (PMID 37106409) | Insulin omission, access, and pump failure are mechanistically distinct | Valid causal phenotype with matched intervention | OQ-14, OQ-26 |