Screening and Prevention of Type 1 Diabetes: Where Are We?

Abstract

A diagnosis of type 1 diabetes (T1D) and the subsequent requirement for exogenous insulin treatment is associated with considerable acute and chronic morbidity and a substantial effect on patient quality of life. Importantly, a large body of work suggests that early identification of presymptomatic T1D can accurately predict clinical disease, and when paired with education and monitoring, can yield improved health outcomes. Furthermore, a growing cadre of effective disease-modifying therapies provides the potential to alter the natural history of early stages of T1D. In this mini review, we highlight prior work that has led to the current landscape of T1D screening and prevention, as well as challenges and next steps moving into the future of these rapidly evolving areas of patient care.

The incidence and prevalence of type 1 diabetes (T1D) is increasing over time in most countries globally (1-3). Notably, T1D is being diagnosed at increased rates in racial and ethnic minority populations in the United States and in resource-limited countries worldwide (1, 4). T1D can be difficult to identify clinically as it can masquerade as many common childhood illnesses or be latent in adults. Exogenous insulin replacement is now available in various formulations and methods of administration, but long-term insulin therapy in individuals with T1D is still associated with significant mortality, complications, and mental and financial burden (5-7). In contrast, disease-modifying therapies successfully addressing the underlying pathogenesis of T1D have potential to preserve endogenous insulin secretion and delay or reduce the need for insulin replacement. The clinical availability of the first US Food and Drug Administration (FDA)-approved treatment to delay T1D provides clear potential to intervene early and modify the course of disease (8). However, early identification and intervention requires screening and identification of presymptomatic T1D. In this mini review we will discuss the need for and current state of screening patients for early stages of T1D, as well as the current outlook for T1D prevention strategies, and potential challenges and areas for further study as screening and preventative therapies increase.

Type 1 Diabetes Screening: Where We’ve Been

Benefits of T1D screening include decreased diabetic ketoacidosis (DKA) rates at clinical T1D onset, a longer period for patients and families to adjust to the diagnosis, and identification of individuals who may benefit from therapeutic interventions. Without screening, T1D is often unrecognized until a person is critically ill with DKA. Rates of DKA at diagnosis range from 15% to 80% globally and are increasing over time in the United States (9-11). Beyond the acute risk of cerebral edema, individuals with DKA are at risk for adverse memory and neurocognitive outcomes (12-15). DKA at diagnosis has also been associated with repeated episodes of DKA, severe hypoglycemia, and worsened glycemic control over time, which translates to an increased risk of microvascular and macrovascular complications and health care costs (16-18). With early identification of presymptomatic T1D and regular follow-up, the rate of DKA drops significantly, often to less than 5% (19, 20). Beyond reducing DKA rates, identification of individuals in early stages of T1D before clinical onset affords patients, families, and health-care providers time to provide education and assess health and behaviors that may need to be proactively addressed to allow for optimal T1D management when insulin treatment is clinically necessary.

Type 1 Diabetes Screening Recommendations

Prospective longitudinal birth cohort studies that enrolled infants at high risk for T1D based on human leukocyte antigen (HLA) genotype or family history have shown that early-stage T1D can be accurately identified when individuals are typically presymptomatic by completing a screening blood test to measure major T1D-associated islet autoantibodies (IAs) (21-23). The 4 major biochemical IAs are glutamic decarboxylase 65 (GAD), protein tyrosine phosphatase islet antigen-2 (IA-2), insulin (mIAA), and zinc transporter 8 (ZnT8), which are all specific proteins that are found in the islet cells in the pancreas. Once 2 or more T1D-associated IAs appear, the risk of clinical T1D progression is 70% in 10 years and approaches 100% over a lifetime as indicated in Fig. 1 (24). Once islet autoimmunity is present, there is no evidence to support that progression to clinical T1D in the general population is different from that in people with a family history of T1D (25).

Figure 1.

Risk of progression to type 1 diabetes (T1D) after seroconversion to islet autoantibody positivity (24).

With the knowledge that T1D is predictable before symptoms develop, in 2015, a joint statement from the JDRF, the Endocrine Society, and the American Diabetes Association (ADA) introduced the idea of staging T1D, in which a person can be identified as having early-stage T1D (stage 1 or 2) if a blood sample is positive for 2 or more IAs. By assessing for impaired fasting glucose (100-125 mg/dL), impaired glucose tolerance (2-hour oral glucose tolerance test [OGTT] glucose 140-199 mg/dL), an abnormal glycated hemoglobin A_1c (HbA_1c) (5.7-6.4%/39-46 mmol/mol), or an increase in HbA_1c by 10% from the prior measure, an individual is then further classified as having stage 1 (normoglycemia) or stage 2 (dysglycemia) T1D.

In parallel to staging recommendations, the ADA recommended that relatives of patients with T1D be informed of the opportunity to be tested for IAs in the setting of a clinical research study, which has most often been accomplished through the TrialNet Pathway to Prevention Study (26, 27). In 2017, recommendations were expanded to include IA screening for anyone in the setting of a research study and screening outside research for first-degree family members of someone with T1D (28, 29). These recommendations reflected the initiation of general-population screening studies as well as the commercial availability of testing for all 4 major T1D IAs for health-care providers (30, 31). The 2022 International Society for Pediatric and Adolescent Diabetes (ISPAD) guidelines acknowledged the value of screening the general population and the importance of coupling IA screening with education and monitoring (32).

With the recent approval of teplizumab as an intervention to offer to individuals who meet stage 2 T1D criteria, the World Health Organization's criteria for screening the general population have been satisfied for T1D (33). However, cost-effectiveness of IA screening requires further consideration as screening programs expand. Another consideration is optimal timing of screening. Most children who develop T1D develop IA at a young age. However, IA screening at only very young ages would yield negative results for the remainder of individuals that seroconvert later in life. A recent analysis that harmonized data from 5 prospective longitudinal birth cohorts found that screening children at age 2 and 6 years was 82% sensitive (95% CI, 79%-86%) and had a positive predictive value (PPV) of 79% (95% CI, 75%-80%) for identifying those who develop T1D by age 15 years (34). In the United States, ranges that include these ages overlap with the Centers for Disease Control Childhood Vaccination Schedule and the Bright Futures/American Academy of Pediatrics recommendations for preventive pediatric health care. Of note, IA should not be measured until after age 6 months because measuring them earlier may reflect maternal IA status rather than that of the infant. Additionally, even with this approach, counseling on the signs and symptoms of stage 3 T1D development are important guidance for individuals who screen IA negative, as a small percentage of individuals who screen IA negative could go on to develop disease later in life.

Approaches to Screening and Confirmation Testing

In a research setting, some studies use genetic risk factors such as HLA or a polygenic risk score to identify which children will benefit most from having T1D-associated IA measured (23, 35, 36). Potential benefits of this approach include the ability to combine testing with newborn screening to limit the scale and improve efficiency of IA screening. Clinically, the only recommended tests to diagnose early stages of T1D are IAs. IAs in the aforementioned birth cohort studies were all measured by radiobinding assays in research laboratories that regularly participate in assay optimization and standardization programs (37-39). It is unclear how the risk of progression and specificity for T1D are affected when interpreting IA measurements from commercial and consumer-based assays. The variable specificity of IA tests and sample collection may affect concordance if screening and confirmation tests are completed at different laboratories; however, for clinical purposes it is optimal to perform the second confirmatory antibody test with a separate serum sample in a College of American Pathologists– and Clinical Laboratory Improvement Amendment–approved laboratory that ideally participates in the Islet Autoantibody Standardization Program and has high performance metrics in this program when possible.

Disease Course and Clinical Care of Individuals Who Screen Positive

Understanding which screening-detected individuals are at highest risk for progression to clinical or stage 3 T1D is important for providing anticipatory guidance and for identifying the individuals who are most likely to benefit from a clinical or research-based intervention. Much of our knowledge about T1D pathogenesis stems from longitudinal cohort studies that have screened children and/or adults at high risk for T1D development due to the presence of high-risk HLA gene(s) or a family history of T1D. These studies show that higher numbers of measurable IAs at younger ages consistently predict increased progression to clinical T1D (40-44). The pattern of IAs also likely plays a role in disease phenotype and risk of progression, and there are efforts to more precisely predict progression at an individual level (45-49). Importantly, there are likely racial and ethnic differences in the risk and rate of progression to clinical T1D that need to be further studied (50). As expanded screening efforts evolve, effects of larger-scale testing in broader populations on the PPV of a confirmed positive IA test will need to be revisited.

Metabolic Data and Risk of Progression

Metabolic tests used to stage T1D per the ADA guidelines include fasting glucose, 2-hour OGTT glucose, and HbA_1c, optimally by gold-standard, high-performance liquid chromatography methods (51). Of these, HbA_1c measurement is possible in most health-care offices and 2 HbA_1c measurements greater than or equal to 5.9% (41 mmol/mol) over 3 to 12 months or a more than 10% in HbA_1c predicted progression to clinical T1D within approximately 1 year in a birth cohort (52, 53). The time to peak glucose and c-peptide levels at different OGTT timepoints may also be useful. A peak glucose after 30 minutes, and a peak C-peptide after 60 minutes is associated with an increased progression risk in IA-positive individuals (54). OGTT data can also be combined with other variables including age, body mass index, and/or IA data to calculate risk scores that help predict progression to clinical T1D (47, 55, 56). Efforts are underway to optimize tradeoffs of such scores in terms of ease of testing and prediction performance (57). Continuous glucose monitors are an emerging tool that may be useful in predicting progression (58). The Diabetes Autoimmunity Study in the Young (DAISY) showed that children with more than 18% of time greater than 140 mg/dL over a 2-week period progressed to clinical T1D with 75% sensitivity, 100% specificity, and a 100% PPV (59). In the Autoimmunity Screening for Kids (ASK) Study, 80% of children with more than 10% of glucose levels greater than 140 mg/dL over a 2-week period progressed to clinical T1D within the next year (60).

Type 1 Diabetes Screening: Where are We Going?

Larger-scale implementation of screening programs, particularly of individuals in the general population, faces multiple challenges moving forward. Ways in which T1D screening conforms to Wilson and Jungner's guidelines for screening tests, as well as challenges moving forward, have been recently outlined in detail (33, 61). Challenges include education and partnerships with primary care physicians; patient access to monitoring and treatment; improved strategies to optimize patient participation and engagement, particularly that of traditionally underrepresented groups; and strategies for broader feasibility and cost-effectiveness. A critical need is consensus guidelines for approaches to screening, and follow-up and monitoring of at-risk individuals by health-care providers. Next we highlight several challenges and questions in these areas.

Staging Challenges

Many types of screening programs, such as those applied as part of newborn screening in the United States (62), perform systematic testing for the presence of a condition. Testing for IAs with or without genetic testing identifies early-stage disease, but may also identify individuals along a risk spectrum who never develop clinical disease. As more people are screened and health-care providers determine follow-up plans and eligibility for interventions, many individuals will likely not fit cleanly into the currently defined stages of T1D. The variables that modify the risk for T1D development need to be better understood to stratify the frequency of required follow-up based on risk of progression. Similarly, there will be many individuals who meet criteria for stage 3 diagnosis per ADA diagnostic criteria but are not yet symptomatic and may not require insulin. Substages that consider the variability of symptoms and treatment requirements at the time of incident hyperglycemia will be useful in developing more personalized clinical care plans for at-risk individuals.

What Is the Risk for Someone With a Single Positive Islet Autoantibody, and How Should They Be Followed Clinically?

Overall risk for T1D progression in a person testing positive for a single IA is only 10% on average, with some individuals at very low risk of progression after 2 to 3 years of remaining single-IA positive (63). However, individual-level risk is also affected by IA type and other factors (64). Further, the predictive value increases when IAs are measured by high-affinity methods such as electrochemiluminescence detection (65), suggesting the need for guidelines surrounding risk and follow-up stratification in single-IA–positive individuals.

How Should Individuals Positive for Islet Autoantibody Be Followed Over Time?

The primary goal when caring for a person who screens positive for IA is medical safety, ensuring they are aware of and can recognize T1D symptoms so that they are clinically well at the time of clinical diagnosis. Many tools, including self-monitoring with fingerstick and home glucometer testing, HbA_1c, OGTT, and continuous glucose monitors, may be used to keep individuals safe and estimate the risk of progression to T1D, but there is limited evidence to guide which tools to use and at what frequency to use them. Fig. 2 displays the current TrialNet Pathway to Prevention research study protocol for follow-up and monitoring of individuals that screen positive for IA (26, 27). Because younger age indicates a risk for faster progression to clinical T1D, more frequent visits may be considered in young children with IA positivity (66). Beyond the accuracy of predicting T1D progression, factors such as feasibility and acceptability are important to consider in clinical practice, and consensus guidelines for monitoring of at-risk individuals in clinical practice is an important need moving forward.

Figure 2.

TrialNet pathway to prevention monitoring protocol (27).

Type 1 Diabetes Prevention: Where We’ve Been

A key consideration in screening and identification of early-stage T1D is the potential for intervention with disease-modifying therapies. Over the past 3 decades, researchers have made substantial efforts to identify therapeutic agents with approaches targeting: 1) primary prevention, before the onset of islet autoimmunity; 2) secondary prevention, after established autoimmunity but before the onset of clinical diabetes, and 3) tertiary prevention, after the onset of clinical T1D with a goal of maintaining endogenous insulin secretion, reducing the need for exogenous insulin, and prolonging the honeymoon period (32). Benefits of targeting stage 3 disease include clear-cut identification of patients that stand to benefit from β-cell preservation agents. On the other hand, a less-advanced disease process and larger residual β-cell mass could allow for larger potential effect of therapies targeting primary or secondary prevention. Given that not all individuals identified as genetically at risk will go on to develop T1D, primary prevention approaches require a more favorable safety profile compared to those in stage 1 or later T1D. Although comprehensive review is beyond the scope of this mini-review, positive and negative studies in each of these categories have recently been summarized (32). To date, larger-scale studies in primary prevention, including dietary interventions or antigen therapies, have not been demonstrated to affect clinical T1D progression (67-69) but several large randomized controlled trials (RCTs) have shown promise for tertiary or secondary interventions (Table 1). Next we highlight several approaches that have garnered substantial interest.

Table 1.

Highlighted large positive randomized controlled efficacy trials testing disease modifying therapies in type 1 diabetes

Study	Intervention(s) (vs Placebo)	Mechanism	Population	Outcome
TrialNet TN10 Teplizumab prevention study (70)	14-d course of IV teplizumab infusion	Anti-CD3 monoclonal antibody depletes T cells with recovery of partially exhausted CD8+ T cells	Stage 2 relatives of individuals with T1D, 8-45 y	32.5-mo delay in time to T1D (71)
Verapamil for Beta Cell Survival in T1D (72)	12 mo 120 mg-360 mg daily verapamil	Improves β-cell stress and survival via targeting of TXNIP	New-onset T1D (<3 mo since diagnosis), 18-44 y	Preserved 12-mo C-peptide AUC
CLVer (Hybrid Closed Loop Therapy and Verapamil) (73)	52 wk weight-based daily verapamil	Improves β-cell stress and survival via targeting of TXNIP	New-onset T1D (<31 d since diagnosis), 7-17 y	Preserved 52-wk C-peptide AUC
Imatinib treatment in recent-onset T1D (74)	400 mg imatinib given daily for 26 wk	Tyrosine kinase inhibition impacting immune cell pathways, β-cell IRE1α hyperactivity	New-onset T1D (<100 d since diagnosis), 12-45 y	Preserved 12-mo C-peptide AUC
TrialNet ATG-GCSF Study (75)	1) Low-dose (2.5 mg/kg) IV ATG or 2) 2.5 mg/kg ATG + 6 doses of subcutaneous GCSF	Lymphocyte depletion, followed by increased regulatory T-cell frequency and hematopoetic mobilization	New-onset T1D (<100 d), 12-45 y	Preserved 12-mo C-peptide AUC compared to placebo in those treated with ATG but not ATG/GCSF
T1GER (A Study of SIMPONI to Arrest Beta-cell Loss it T1D) (76)	Subcutaneous golimumab injections every 2 wk for 52 wk	Monoclonal antibody specific for tumor necrosis factor α (anti-inflammatory)	New-onset T1D (<100 d), 6-21 y	Preserved 52-wk 4-h C-peptide AUC
Anti-IL-21–liraglutide Study (77)	1) Anti-IL-21 antibody + liraglutide, 2) anti-IL-21 alone, or 3) liraglutide alone	Anti-IL-21: targeting IL-21 targeting of T-cell trafficking; GLP-1 receptor agonism enhance β-cell function; relieve β-cell stress and prevent apoptosis	Recent-onset T1D (within 20 wk), 18-45 y	Combination treatment, but not IL-21 alone preserved 54-wk change in C-peptide AUC compared to placebo
TrialNet TN09 Abatacept Study (78)	Chronic abatacept over 2 y	CTLA4-Ig-modulation of T-cell coactivation and stimulation	New-onset T1D (<100 d), 6-45 y	Preserved 2-y C-peptide AUC
Immune Tolerance Network T1DAL Study (Inducing Remission in T1D with Alefacept) (79, 80)	Two 12-wk courses of alefacept	Fusion protein binding CD2 interrupts CD2-mediated T-cell costimulation to deplete memory and effector T cells	New-onset T1D (<100 d), 12-35 y	Preserved C-peptide area under the curve at 24 mo, although no significant difference between groups at primary end point (12 mo; P = .065)
TrialNet TN05 Rituximab in New-Onset Diabetes Study (81)	4-dose course of IV rituximab vs placebo	Anti-CD20 monoclonal antibody yielding selective depletion of B lymphocytes	New-onset T1D (<3 mo), 8-40 y	Preserved 12-m C-peptide AUC
Protégé Study (Teplizumab for Treatment of T1D) (82)	14-d full dose, 14-d low dose, or 6-d full dose of teplizumab or placebo at baseline and 26 wk	Anti-CD3 monoclonal antibody depletes T cells with recovery of partially exhausted CD8+ T cells	New-onset T1D (<12 wk), 8-35 y	No difference in primary outcome (% participants with insulin use <0.5 U/kg/d and HbA1c < 6.5% (48 mmol/mol) at 1 y) between groups (82) but 14-d full-dose regimen preserved C-peptide AUC at 2-y follow-up (83)

Abbreviations: ATG, antithymocyte globulin; AUC, area under the curve; GCSF, granulocyte colony-stimulating factor; GLP-1, glucagon-like peptide 1; HbA_1c, glycated hemoglobin A_1c; IL, interleukin; IV, intravenous; T1D, type 1 diabetes.

Antigen Therapy

Antigen therapy, based on the idea that treatment with a potential autoantigen will induce immune tolerance, is appealing because of its minimal side-effect profile (67). Limited successes in identifying durable effects of antigen therapy in T1D may in part reflect outstanding questions regarding the optimal form, dose, adjuvant, and delivery methods (67). Several large trials have tested various forms and doses of GAD and insulin (reviewed in (84)). RCTs using GAD-based therapies have shown inconsistent efficacy, although a meta-analysis suggested a small but significant effect of GAD vaccine to preserve 12-month post-treatment C-peptide (85, 86). More recently, a meta-analysis suggested that the DR3-DQ2 HLA haplotype may identify individuals exhibiting a GAD treatment response (87). Consistent with this, a negative intervention study testing intralymphatic GAD-alum injection reported a treatment effect within the subset of DR3-DQ2–positive participants (88). This association is interesting because GAD positivity at IA seroconversion has been linked to HLA DR3 positivity (46). Based on these data, the DIAGNODE 3 study will prospectively test efficacy of the intralymphatic injection in the DR3-DQ2 population (NCT05018585).

The Diabetes Prevention Trial Type 1 (DPT-1) tested either parenteral or 7.5-mg oral insulin daily in individuals with early-stage T1D, neither of which affected incident clinical diabetes in the primary analysis (89, 90). Oral insulin reduced progression compared to placebo among individuals with elevated insulin IA titers (90), but a follow-up study of this population failed to show a treatment effect (91). A prespecified secondary stratum with low first-phase insulin secretion did exhibit delayed time to diabetes with oral insulin (91). A post hoc analysis supported this idea by showing that individuals with high baseline metabolic risk indices showed metabolic improvement compared to placebo (92). Interestingly, the Pre-POInT (Primary Oral Insulin trial) testing escalating doses of oral insulin in genetically at-risk IA-negative children suggested that immune responses to insulin were only increased with a much larger 67.5-mg daily dose (93). Unfortunately, the follow-up Pre-POInT-early study did not demonstrate immune efficacy with this regimen, although an increased treatment effect was detected in those with INS gene polymorphisms (94). The ongoing larger-scale GPPAD (Global Platform for the Prevention of Autoimmune Diabetes)-POInT trial will further address this dose-escalation oral insulin regimen in genetically at-risk infants (95), while the Fr1da Insulin Intervention study (NCT02620072) is testing dose escalation in children with stage 1 disease. Ongoing trials are also testing other dosages, administration forms, and mechanisms of effect ((NCT00336674, NCT0258087, NCT04279613).

Immunomodulatory Therapies in Secondary and Tertiary Prevention

Given the autoimmune nature of T1D, immunomodulation is a natural choice for therapies aimed at underlying disease pathology. Because of the potential for side effects, these therapies are often initially trialed in stage 3 populations, with extension to secondary prevention based on efficacy. Although trials testing immunomodulation have also met with mixed success over the years, recent positive outcomes have identified agents with enormous potential for ultimate implementation for delay of T1D in at-risk individuals (see Table 1).

Immunomodulatory approaches have included cellular therapies and drugs that promote induction of regulatory T cells, reduction of inflammation, and a decreased number of effector immune cells or their activity. Immunomodulation with cell-based approaches has been targeted by several groups, although, to date, outside the context of immunosuppression, durable efficacy of stem cell, regulatory T cell, or tolerogenic dendritic cell infusions remains to be demonstrated in large-scale RCTs (96). Although anti-inflammatory approaches have met with mixed success overall (97), tumor necrosis factor α blockade exhibited a promising effect on C-peptide preservation in the new-onset period (76).

Various approaches have targeted effector immune cells in the new-onset period. Treatment with mycophenolate mofetil (MMF) to inhibit T- and B-lymphocyte proliferation, with and without a monoclonal antibody binding the α subunit of the interleukin-2 (IL-2) receptor (daclizumab), was ineffective at C-peptide stabilization (98). Infusions of the CTLA-4 immunoglobulin abatacept to modulate costimulation and activation had an early effect to stabilize C-peptide compared to placebo in new-onset disease, and also improved C-peptide but failed to show a significant effect to delay progression to stage 2 or stage 3 disease in individuals with stage 1 T1D (78, 99). Alefacept, a fusion protein targeting CD2-mediated T-cell costimulation and depletion, did not reach a statistically significant effect on C-peptide at 12 months (79), but yielded higher C-peptide 24 months after initiating treatment (80).

Several studies have achieved successful effects on C-peptide in the new-onset period via transient effector cell depletion. Rituximab, an anti–CD-20 monoclonal antibody selectively targeting B cells, increased C-peptide 12 months after treatment, but levels were indistinguishable from the placebo group by 2 years (81, 100). Low-dose antithymocyte globulin (ATG) and monoclonal antibodies to CD3 have both been used to induce temporary T-cell depletion and demonstrated benefits on C-peptide in recent-onset T1D (70, 75). Several doses of ATG with and without administration of granulocyte colony–stimulating factor (GCSF) have been tested, with low-dose ATG in isolation performing the best to preserve C-peptide (75, 101). The TrialNet network is now testing this agent in stage 2 disease (NCT04291703).

The biggest success in T1D disease modification to date has surrounded treatment with monoclonal antibody to CD3, which yields temporary T-cell depletion, with treatment response associated with a partially exhausted state of CD8+ T cells (102). In addition to C-peptide stabilization after stage 3 onset (83), the Fc receptor-nonbinding anti-CD3 monoclonal antibody teplizumab has shown success in T1D prevention, where a 14-day infusion delayed diabetes onset by 32.5 months and improved C-peptide values in individuals with stage 2 disease (70, 71, 103). Based on these data, on November 17, 2022, teplizumab became the first FDA-approved therapy for delay of T1D (8).

Agents Targeting β-Cell Stress and Function

The idea of targeting β-cell stress in T1D is not new, but studies in this area have seen a recent renewed interest. In the Diabetes Control and Complications Trial, the intensive glycemic control group exhibited improved endogenous insulin secretion, suggesting benefit of relief of glucose toxicity or β-cell rest (104). This could be consistent with the improvement in β-cell function that often occurs in the honeymoon period after therapeutic insulin administration (105). Multiple studies have tested β-cell rest in recent-onset T1D. Diazoxide administration yielded variable results (106, 107), while a small study testing 2 weeks of optimized insulin dosing using a Biostator yielded increased 12-month stimulated C-peptide (108). Recent hybrid closed-loop (HCL) systems have also been tested as a form of beta-cell rest, first with 72 hours of inpatient HCL, which had no effect on 12-month C-peptide (109). Longer-term treatment with more modern HCL technology has also been tested by the CLOuD (Closed Loop from Onset in Type 1 Diabetes) and CLVer (Hybrid Closed Loop Therapy and Verapamil) studies (110, 111). Neither trial showed an effect of HCL therapy on endogenous insulin secretion.

Several agents initially purposed for other indications have shown potential to affect insulin secretion in recent-onset T1D via presumed effects on the β cell. The antihypertensive calcium channel blocker verapamil, which improves β-cell function and survival via transcriptional inhibition of the cellular redox regulator thioredoxin-interacting protein (TXNIP) (112), showed a durable effect of ongoing verapamil treatment on insulin secretion in a small study of adults (72, 113). This beneficial effect of verapamil was recently replicated in a large pediatric population in the CLVR trial (73) and is being followed up in a larger group of adult participants in the INNODIA Ver-A-T1D (Verapamil SR in Adults with Type 1 Diabetes) trial (NCT04545151). The tyrosine kinase inhibitor imatinib, an established treatment for chronic myelogenous leukemia, significantly stabilized C-peptide secretion compared to placebo after 1 year of treatment in adults with recent-onset T1D (74). This drug is postulated to work in part via effects on the β-cell endoplasmic reticulum stress signaling cascade (114). The JAK-STAT pathway acts downstream of cytokine receptors in immune cells and β cells (115, 116). Ongoing studies are testing drugs affecting the JAK-STAT and polyamine biosynthesis pathways, both of which are involved in the response of β cells to proinflammatory cytokines (117), NCT05594563.

Glucagon-like peptide 1 (GLP-1) agonism has been proposed as an alternative way to enhance β-cell function and promote β-cell health in T1D (118). Although GLP-1–based monotherapy has not shown a significant effect in recent-onset T1D (119), combination treatment with liraglutide and anti–IL-21 treatment improved C-peptide secretion, which rapidly fell to levels comparable to placebo after stopping the drug (77). Longer-term follow-up after treatment will be important to understand long-term effects of agents directly targeting β-cell function on β-cell health.

Type 1 Diabetes Prevention: Where Are We Going?

While a huge stride in the direction of moving disease-modifying therapies into clinical practice, the FDA approval of teplizumab to delay T1D onset is still an early step on this journey, with many questions and challenges remaining to be addressed. Table 2 outlines currently ongoing trials of disease-modifying therapies.

Table 2.

Ongoing or newly completed/still under analysis recent studies in primary or secondary prevention or tertiary prevention/intervention trials in new or recent-onset type 1 diabetes

Study	Intervention	Mechanism	Population	Primary end point
Primary Prevention
GPPAD-POInt (Primary Oral Insulin Trial) study (95)	Daily oral insulin at escalating dose up to 67.5 mg vs placebo until age 36 mo	Antigen therapy	Infants at genetic risk, 4-7 mo	Development of persistent confirmed multiple β-cell autoantibodies or diabetes
GPPAD SINT1A Study (Supplementation with B infantis for Mitigation of T1D Autoimmunity) (120)	Daily oral B infantis EVC001 probiotic or placebo for 12 mo	Influence on intestinal flora and peripheral tolerance	Infants at genetic risk, 7 d-6 wk	Persistent confirmed multiple β-cell autoantibodies through study completion (up to 6.5 y)
Secondary prevention
Fr1da Insulin Intervention study (NCT02620072)	Daily oral insulin at escalating dose up to 67.5 mg vs placebo for 12 mo	Antigen therapy	Stage 1, 2-12 y	Immune response to insulin and dysglycemia or diabetes development
INIT-II (Trial of Intranasal Insulin in Children and Young Adults at Risk of T1D) (NCT00336674)	Intranasal insulin for 12 mo vs placebo	Antigen therapy	Stage 1, 4-30 y with family history; also with FPIR criteria	Diagnosis of diabetes
TrialNet TN20 Immune Effects of Oral Insulin in Relatives at Risk for T1D (NCT02580877)	Oral insulin 67.5 mg daily compared to 500 mg every other wk for 6 mo	Antigen therapy	IAA + stage 1 children and adults 3-45 y; also 3-7 y with AGT	Change in immune function compared to baseline
TrialNet TN22 Hydroxychloroquine prevention study (NCT03428945)	Oral hydroxychloroquine daily vs placebo	Anti-inflammatory, may affect Ag presentation, insulin sensitivity	Stage 1, ≥3 y	Time to development of AGT or diabetes
TN28 TrialNet ATG prevention study (NCT04291703)	Low-dose ATG infusion ×2 (2.5 mg/kg total) vs placebo	Lymphocyte depletion	Stage 2, 12-45 y	Time to diabetes
Tertiary Prevention/Intervention
DIAGNODE-3 (Diamyd Administered Into Lymph Nodes in Individuals Recently Diagnosed with Type 1 Diabetes, Carrying the HLA DR3-DQ2 Haplotype, NCT05018585)	3 intralymphatic injections of recombinant GAD Alum (Diamyd) + cholecalciferol for 4 mo vs placebo injections + cholecalciferol	Antigen therapy	12-28 y with T1D (≤6 mo since diagnosis) and HLA DR3-DQ2 haplotype	Change in C-peptide AUC and HbA1c at 24 mo
USTEK1D (Phase II Multicentre, Double-Blind, Randomised Trial of Ustekinumab in Adolescents With New-Onset T1D) (121)	Ustekinemab dosed every 4-8 wk for 44 mo vs placebo	Inhibits p40 molecular subunits of IL-12 and IL-23 (to inhibit IL-17 and IFN-γ autoimmune cytokine pathways)	Stage 3 (≤100 d of diagnosis), 12-18 y	C-peptide AUC at wk 52
ITAD (Interleukin-2 Therapy of Autoimmunity in Diabetes) (122)	Subcutaneous ultra-low dose IL-2 (aldesleukin), twice-weekly, or placebo, for 6 mo	Selective enhancement of Treg response	Stage 3 (<6 wk from diagnosis), 6-18 y	Change in dried blood spot C-peptide (slopes) during 6-mo treatment period
INNODIA Ver-A-T1D (NCT04545151)	Oral verapamil SR 120-360 mg vs placebo for 12 mo	Reduces β-cell stress and death via targeting of TXNIP	Stage 3 (<6 wk since diagnosis), 18-45 y	C-peptide AUC at 12 mo
INNODIA MELD-ATG (Phase II Dose Ranging Efficacy Study of ATG Within 6 Weeks of Diagnosis of T1D, NCT04509791)	Test minimal effective dose of ATG vs placebo (down to 0.1 mg/kg)	T-cell depletion	Stage 3 (≤3-9 wk of diagnosis), 5-25 y	C-peptide AUC at 12 mo
CFZ533X2207 (iscalamib) Trial (NCT04129528)	12 mo of CFZ533X2207 (infusion then subcutaneous) vs placebo	Anti-CD40 monoclonal antibody	Stage 3 (<100 d since diagnosis), 6-21 y	C-peptide AUC at 12 mo
TADPOL Study (Targeting T1D Through Polyamines) (NCT05594563)	6 mo of 1000 mg/m2/d DFMO vs placebo	Inhibition of polyamine biosynthesis pathway	Stage 3 (<100 d of diagnosis), 6-45 y	C-peptide AUC at 6 mo
PROTECT (NCT03875729)	2 courses of 12-d teplizumab infusions 6 mo apart vs placebo	Anti-CD3 monoclonal antibody; renders CD8+ T cells partially exhausted	Stage 3 (<6 wk since diagnosis), 8-17 y	C-peptide AUC at 78 wk
BANDIT (Baricitinib in New-Onset T1D) (117)	Baricitinib 4 mg/d or placebo for 48 wk	JAK pathway inhibition	Stage 3 (<100 d since diagnosis), 10-30 y	C-peptide AUC at 48 wk

Abbreviations: AGT, abnormal glucose tolerance; ATG, antithymocyte globulin; AUC, area under the curve; DFMO, difluoromethylornithine; FPIR, first-phase insulin secretion; GCSF, granulocyte colony-stimulating factor; GLP-1, glucagon-like peptide 1; HbA_1c, glycated hemoglobin A_1c; IAA, insulin islet autoantibody; IFN, interferon; IL, interleukin; JAK, Janus kinase; T1D, type 1 diabetes; Treg, regulatory T cell; TXNIP, thioredoxin-interacting protein.

Intervention Regimens

The ultimate optimal regimen for T1D prevention has yet to be determined. Many trials in stage 3 have used monotherapies or single dosing, but durable success may require repeated drug dosing, combination regimens using an induction/maintenance approach, and/or multiple agents targeting different mechanisms of action (123). Unfortunately, the ability to test different regimens in asymptomatic individuals is limited by the cost and effort required to implement time-to-event prevention studies. Thus, an approach that has been frequently used is to test the effect of treatment in the new-onset population. Because clinical outcomes like remission or insulin dose are often affected by many variables beyond β-cell function (124), and given the known benefits of residual β-cell function on diabetes complications (125), these trials typically target C-peptide secretion as a primary end point, with plans to then test therapies that show efficacy in the pre-symptomatic stages of T1D.

Timing of Disease-modifying Interventions

The optimal timing of treatment could play a key role in the success of disease-modifying therapies. For example, as noted, treatment at clinical T1D onset is appealing based on ease of identification of patients that may stand to benefit. However, limited successes of treatments in this period may reflect irreversible effects of the relatively advanced disease process on β-cell mass, health, and function (126). This line of thinking suggests that treatment in earlier stages of disease, before clinical onset, may have the most effect. Another population to consider for earlier intervention could be individuals who meet ADA criteria for a T1D diagnosis based on OGTT, and so are technically stage 3, but are asymptomatic and may not yet require insulin dosing (32).

Optimal timing of dosing likely also varies depending on the mechanism of action of individual therapies. For example, because β cells may be important early in the pathogenesis of T1D (127), agents targeting β cells could theoretically perform optimally in early-stage disease. On the other hand, for certain immunotherapies, an ideal therapeutic window may exist during periods of “active” autoimmune attack. For example, preclinical studies treating nonobese diabetic mice with anti-CD3 effectively reversed diabetes, but showed no effect on diabetes development when given at earlier time points (4 or 8 weeks) (128). This could be consistent with a reduced effect of teplizumab on C-peptide preservation in individuals with a more remote T1D diagnosis compared to very recently after disease onset, and the prevention study observation that lower C-peptide at study entry was associated with a greater response to the drug (70, 129). Along these lines, cyclosporin treatment in individuals with recent-onset T1D yielded greater remission rates compared to placebo in individuals with elevated proinsulin:C-peptide ratios, a marker of β-cell stress that may serve as a readout of active disease (130). Moving forward, well-validated markers of β-cell function, stress, and health may better inform the validity and feasibility of this idea of an optimal therapeutic window.

Prevention Trials Are Long and Expensive

Given variable time to progression of T1D, a major challenge in terms of cost and feasibility is the duration of prevention trials, which use time to T1D as a primary end point (131). As an example, the CT.gov website lists a range of 8 to 9 years from start date to primary completion date for the DPT-1, TrialNet oral insulin, and teplizumab prevention studies. Long study duration is a barrier to implementation of subsequent studies that could test for alternative agents, combination therapies, or alternative dosing regimens of therapies that have previously demonstrated an effect.

The issue of feasibility is further complicated by the effort required to identify at-risk individuals who stand to benefit from therapies aimed at delaying the onset of T1D. Given the increased risk of islet autoimmunity and T1D in family members of individuals with disease (132), one approach to improve screening yield, undertaken by the DPT-1 and Type 1 Diabetes TrialNet groups, has been to screen family members of individuals with T1D (89, 133). However, the large majority of individuals presenting with new-onset T1D do not have a family history (134, 135). Thus, larger-scale testing and implementation of disease-modifying therapies in those who stand to benefit will ultimately require screening efforts in the general population, potentially implemented as part of routine preventive health care.

Another potential solution to limited capacity for prevention trials could be increased utilization of alternative end points providing more rapid readouts of treatment effect. For example, in the teplizumab prevention study between-group differences in OGTT area under the curve C-peptide could be detected as early as 6 months after randomization (71). Utilization of end points combining changes in OGTT C-peptide and glucose data identified a treatment effect as early as 3 months following treatment (103). Although not equivalent to time to diabetes, in certain situations, such measures could be useful rapid readouts for early stopping rules, or pilot studies to gauge potential for an effect on diabetes onset. Promising results could then be followed up by a longer trial with traditional end points. Other proposed alternative trial designs have included adaptive designs, dose-ascending studies, or factorial studies that could allow for testing of more agents using fewer total participants (131).

Heterogeneity in Disease

A growing movement has recognized heterogeneous phenotypes in the progression of IA positivity to T1D and clinical presentation of disease (136). Indeed, analysis of clinical biomarkers and cadaveric tissue sections has linked differences in pathologies associated with disease to patient characteristics, most frequently age of T1D onset (136, 137). A natural question is if these differences are also associated with differences in response to disease-modifying therapies. Several studies have identified individuals exhibiting improved treatment response, or “responders” based on characteristics such as age, baseline C-peptide and measures of β-cell function, and immune cell profiles (70, 102, 130, 138, 139). To validate these intriguing findings, an important consideration is the need for prospective/prespecified mechanistic studies linking treatment response to differences in disease pathology. This is especially key given the need to attract further pharmaceutical investment in this therapeutic area. On the other hand, identification of true endotypes is an attractive goal as it may allow for a more individualized approach to improve outcomes.

Diversity of Intervention Study Participants

In contrast to heterogeneity in disease phenotypes, an important challenge is the homogeneity in ethnic and racial populations in T1D natural history and intervention studies (140). This finding has persisted despite reports that, at least in the United States, T1D incidence may be increasing the most in traditionally underrepresented populations (3). In part this could be related to differences in screening and monitoring for T1D in these groups (61). Trials that include diverse representation of race and ethnicity are critical to truly understand heterogeneity in disease and treatment response and must be an emphasized priority moving forward. Additionally, now that a disease-modifying treatment is approved, outreach to improve screening and monitoring, as well as ensure access to interventions, is an important goal.

Effect of Recent Food and Drug Administration Approval of Teplizumab for Type 1 Diabetes Delay

The recent FDA approval of teplizumab represents a landmark success for this field, opening the doors for future approvals of effective therapies or regimens in this area. Important next considerations will be implementation and access to this treatment for patients that may benefit, including advocating for payer coverage, and education of key stakeholders, such as patients, family members, and primary care providers. Additionally, as noted earlier, expansion of current screening strategies will allow for increased identification of individuals who stand to benefit. The existence of teplizumab as an available treatment option may affect enthusiasm for trials that involve long-term dosing with placebo for some participants, and so this should be considered in designing prevention trials moving forward.

Conclusions

Built on decades of dedication and perseverance of the T1D patient and research communities, major strides have transformed the fields of T1D screening and prevention in recent years. However, as outlined here, challenges remain to effectively move screening and disease-modifying interventions into routine clinical care. Additionally, further refinement of T1D risk stratification with individual contributors to heterogeneity in risk of T1D progression and treatment response may ultimately allow for more optimized T1D prediction and prevention.

Abbreviations

ADA

American Diabetes Association

AGT

abnormal glucose tolerance

ATG

antithymocyte globulin

AUC

area under the curve

DFMO

difluoromethylornithine

DKA

diabetic ketoacidosis

DPT-1

Diabetes Prevention Trial Type 1

FDA

US Food and Drug Administration

GAD

glutamic decarboxylase 65

GCSF

granulocyte colony-stimulating factor

GLP-1

glucagon-like peptide 1

HbA_1c

glycated hemoglobin A_1c

HCL

hybrid closed-loop

HLA

human leukocyte antigen

IA

islet autoantibody

IFN

interferon

IL

interleukin

JAK

Janus kinase

OGTT

oral glucose tolerance test

PPV

positive predictive value

RCT

randomized controlled trial

T1D

type 1 diabetes

Treg

regulatory T cell

TXNIP

thioredoxin-interacting protein

Funding

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases (grant Nos. R01DK121929A1 and R01DK133881 to E.K.S. as well as U01 DK127382 and U01DK127786), and by the Doris Duke Charitable Foundation through the COVID-19 Fund to Retain Clinical Scientists collaborative grant program (grant No. 2021258) and was made possible by support from the John Templeton Foundation (grant No. 62288).

Disclosures

E.K.S. has received compensation for educational lectures on T1D screening from Medscape and the American Diabetes Association. K.M.S. has received compensation for educational lectures on T1D screening from Medscape and MJH Life Sciences; has consulting agreements with DexCom and Provention Bio and has served on advisory boards for Provention Bio; and receives research funding from Novartis, Provention Bio, JDRF, Helmsley Charitable Trust, and the National Institutes of Health.

Data Availability

No original data were generated as part of this manuscript.