Executive summary
PRISE-hATG plans to recruit about 72 participants prospectively, with 36 more borrowed from SAFEGUARD. Four U.S. institutions appear in the public record for both studies: UCSF Benioff Children’s Hospital, the Barbara Davis Center, the University of Florida, and Riley Hospital for Children.
- What the evidence supports. All four have documented T1D research infrastructure and are relevant to the study.
- What it does not. No site’s eligible-patient volume, comparable enrollment rate, or operational readiness has been verified, so the public record cannot rank them.
- Main risks. Overlapping recent-onset trials at several sites, unquantified C-peptide and other eligibility losses, and the gap between study leadership and local recruitment capacity.
- Confidence. Low to moderate. Documented infrastructure, estimated or assumed enrollment.
- Recommendation. Send a standardized site questionnaire to all four before building a site-level forecast, and keep the longlist open until responses arrive.
A note on “strong.” Here it means relevant public research evidence exists. It is not a measured comparison of enrollment performance.
Study requirements
Site selection starts from what the protocol demands. This matrix records the current understanding and what must be verified against the protocol.
| Requirement | Current understanding | Verification needed |
|---|---|---|
| Design | Phase 3, multicenter, randomized, double-blind, placebo-controlled; SAB-142 | Schedule and dosing |
| Population | Stage 3 T1D, ages 5–40 | Exact inclusion and exclusion criteria |
| Prospective cohort | About 72 participants, randomized 2:1 | Target and cohort allocation |
| External cohort | About 36 from SAFEGUARD (NCT07187531) | Eligibility, data availability, planned analysis |
| Disease duration | Over 100 days to two years (prospective); within 100 days (external) | How diagnosis and enrollment dates are defined |
| Primary endpoint | Change in log C-peptide AUC+1, two-hour MMTT | Assessment schedule and analysis population |
| Treatment and follow-up | Initial dose, another at Month 6, follow-up to Month 12 | Visit schedule and extra procedures |
| Screening | C-peptide and other clinical, laboratory, safety, and participation criteria | Exact thresholds and procedures |
| Geography | Participating sites and status | Current site list and activation |
Site overview
| Site | Relevant public evidence | Preliminary interpretation |
|---|---|---|
| UCSF Benioff Children’s Hospital San Francisco, CA | PRISE-hATG and SAFEGUARD site; TrialNet participation | Direct program experience; test its ability to recruit beyond the first 100 days |
| Barbara Davis Center Aurora, CO | PRISE-hATG and SAFEGUARD site; large T1D center | Specialized research center; check recent-onset and established-onset volumes |
| University of Florida Gainesville, FL | Lead institution; central contacts; SAFEGUARD site | Strongest study-level link; separate leadership from local recruitment |
| Riley Hospital for Children Indianapolis, IN | PRISE-hATG and SAFEGUARD site; pediatric research setting | Check how its age distribution fits the 5–40 range |
These are evidence profiles, not a validated ranking. Public data support the relevance of all four sites but not their comparative enrollment capacity.
Site evidence profiles
Facts below come from public registry records and institutional web pages, and should be verified before relying on them.
UCSF Benioff Children’s Hospital · San Francisco, CA
Public evidence. The site’s pediatric diabetes program reports caring for more than 800 children, adolescents, and young adults with diabetes, and participates in TrialNet with a focus on beta-cell preservation. UCSF is listed for PRISE-hATG and SAFEGUARD, and for other recent-onset T1D studies including a Phase 2 rezpegaldesleukin trial and WAVE T1D.
Feasibility interpretation. Established T1D research infrastructure is a strong positive. The other studies create possible overlap, particularly within 100 days of diagnosis. PRISE’s prospective cohorts start after that window, so real competition depends on each study’s eligibility and recruitment windows.
Questions to resolve
- How many patients aged 5–40 are diagnosed or referred each year?
- How many fall in the two PRISE duration cohorts, and how many exceed the C-peptide threshold?
- Which competing studies are enrolling eligible patients now?
- What recruitment rate and yield has the site achieved in comparable studies?
Preliminary rating. Strong documented disease-area expertise; enrollment capacity unverified.
Barbara Davis Center for Diabetes · Aurora, CO
Public evidence. A specialized T1D center for children and adults. Institutional materials report more than 7,500 active patients with diabetes (about 4,600 children and 2,900 adults), a broad Rocky Mountain referral base, and more than 20,000 outpatient encounters a year. These figures cover diabetes care generally, not PRISE-eligible patients. The center is involved in TrialNet, TEDDY, DAISY, and other networks, and is listed for PRISE-hATG and SAFEGUARD.
Feasibility interpretation. A compelling site to investigate for its specialized patient base, research operations, and referral reach. Total patient count indicates infrastructure but cannot serve as the denominator for eligible participants.
Questions to resolve
- How many patients aged 5–40 are actively followed?
- Annual newly diagnosed patients, and patients diagnosed 100 days to two years earlier?
- What share receive care at the center rather than affiliated practices?
- Is there coordinator capacity for the treatment and follow-up schedule?
Preliminary rating. Strong public evidence of T1D clinical and research infrastructure; duration-specific volume and capacity unknown.
University of Florida · Gainesville, FL
Public evidence. Listed as the lead institution for PRISE-hATG, with central study contacts named in the registry. Its Diabetes Institute lists T1D studies across clinical, translational, and longitudinal research, and its TrialNet program reports screening nearly 3,500 participants in Pathway to Prevention. That shows infrastructure, not a count of PRISE-eligible patients. It is also a SAFEGUARD site.
Feasibility interpretation. The strongest documented organizational connection, because it leads the study. Study leadership should be kept separate from local enrollment: leading PRISE does not show that Gainesville will recruit faster than other centers.
Questions to resolve
- What share of the 72 prospective participants is expected from Gainesville?
- How many potentially eligible patients are in its pediatric and adult practices?
- What is the history in recent-onset T1D intervention trials?
- How much enrollment will come from internal versus external referrals?
Preliminary rating. Strong study-level connection and research infrastructure; site-specific targets and capacity unverified.
Riley Hospital for Children · Indianapolis, IN
Public evidence. The site’s pediatric endocrinology division focuses on T1D prevention and treatment, with research staff including nurses, coordinators, and technicians. Public study listings include WAVE T1D (an ATG-based trial for ages 9 to under 21) and a recent-onset CNP-103 trial, indicating both research activity and possible competition. Riley is listed for PRISE-hATG and SAFEGUARD.
Feasibility interpretation. Relevant pediatric T1D expertise and infrastructure. Its contribution depends on how many eligible adolescents and young adults fall inside the duration windows, and on whether competing trials draw from the same clinic population.
Questions to resolve
- What is the age distribution of patients in the clinic and referral network?
- Are WAVE T1D or other trials recruiting the same patients?
- What are historical screening and randomization numbers in comparable trials?
- Are there referral pathways to adult services for participants near age 40?
Preliminary rating. Strong documented pediatric expertise; age-aligned pool and competition need investigation.
Cross-site findings
| Dimension | UCSF | Barbara Davis | Florida | Riley |
|---|---|---|---|---|
| Listed PRISE-hATG site | Yes | Yes | Yes | Yes |
| Listed SAFEGUARD site | Yes | Yes | Yes | Yes |
| Public T1D research infrastructure | Strong | Strong | Strong | Strong |
| Eligible-patient volume verified | No | No | No | No |
| Comparable enrollment rate verified | No | No | No | No |
| Operational readiness verified | No | No | No | No |
How we rate confidence
Every claim and estimate carries an evidence level, so a sponsor can see how much weight it bears.
| Level | Meaning | Example |
|---|---|---|
| AConfirmed | Verified with the site or a reliable primary source | Coordinator confirms current capacity |
| BDocumented | Supported by a public primary source | Registry shows prior trial participation |
| CEstimated | Derived from public data by an explicit method | Eligible population in a catchment |
| DAssumed | Scenario input without direct support | Assumed monthly recruitment rate |
Evidence quality is not site quality. An excellent site that is poorly documented publicly gets a low confidence rating until the gap is filled.
Our approach to site feasibility
A full engagement extends this sample in nine steps: extract protocol requirements, build a site longlist (separating registry-listed sites from confirmed, activated ones), research each site, estimate the eligible patient pool, map competing studies, build the enrollment model, score risks and evidence gaps, validate with sites, and quality-check the report.
Patient-pool funnel
The pool narrows through five stages: people with the condition, relevant age and stage, meeting key clinical criteria, reachable by the site, and potentially enrollable. Each reduction needs recorded evidence. Unknowns are shown as ranges, never invented percentages, and a state’s population is never simply divided across its sites, because patients are unevenly distributed and catchments overlap.
Enrollment model
Randomizations (site i) = screened candidates (site i) × screening yield (site i)
Screening yield means the share of screened candidates who are ultimately randomized, and the denominator (prescreened, consented, or fully screened) is kept consistent. Sites are not assumed to activate together or recruit equally. Output includes conservative, base, and optimistic timelines and the assumptions driving the spread.
Risk dimensions
| Dimension | Question |
|---|---|
| Patient availability | Is there evidence of an adequate eligible population? |
| Eligibility uncertainty | Are key protocol restrictions hard to quantify? |
| Recruitment competition | Are similar studies drawing from the same population? |
| Investigator experience | Is relevant experience documented? |
| Operational readiness | Are activation and staffing confirmed? |
| Forecast uncertainty | How much does the estimate depend on assumptions? |
Each site is rated low, moderate, or high on each dimension with the evidence stated. A site can have strong scientific expertise and still carry high recruitment uncertainty.
Recommended next steps
- Send a standardized questionnaire to all four sites: eligible patient volume by age and duration, referral sources, competing trials, historical screening and enrollment, staffing, and activation timelines.
- Verify the status and locations of overlapping studies at each institution.
- Replace assumed inputs with confirmed or documented ones, then build the site-level scenarios.
- Expand the longlist only if responses show the four sites cannot meet the target.
Conclusion. Public evidence shows four relevant, experienced institutions. It cannot show which will deliver, so the next step is site-confirmed data, not a more elaborate model.