Clinical trials landscape in thoracic aortic disease¶
TL;DR — A live ClinicalTrials.gov survey, first run 2026-08-27 and re-run 2026-08-28, returned 97 interventional trials with active status (recruiting, not-yet-recruiting, or active-not-recruiting) under thoracic aortic aneurysm or aortic dissection. The overwhelming majority test devices, surgical technique, or perioperative care. Exactly one randomised trial addresses the central clinical question of the field: TITAN:SvS (NCT03536312), randomising 610 patients with asymptomatic ascending aneurysms of 5.0–5.4 cm to early surgery versus surveillance, started 2018, primary completion 2035. Disease-modifying drug trials in the thoracic aorta are nearly absent: the drug pipeline that exists — metformin, colchicine, doxycycline — sits almost entirely in abdominal aortic aneurysm, a partly atherosclerotic disease that the genetics says is distinct (PMID 37308786). The two exceptions retrieved are a 50-patient spironolactone pilot with a pulse-wave-velocity endpoint (NCT07483177) and a 15-patient genotype-directed nicotinamide riboside study in ACTA2 R179 disease (NCT06280482). Trials are hard here because events are rare, horizons are decades, the accepted surrogate (aortic root Z-score) moves by ~0.07–0.13 per year (PMID 36049495), and the highest-risk patients are systematically excluded from the trials that would inform their care.
How to read trial stages on this page¶
- Drug trials carry FDA-defined phases, reported by ClinicalTrials.gov: phase 1 (safety and dose-finding, tens of patients), phase 2 (preliminary efficacy, usually on surrogates), phase 3 (confirmatory, endpoint-powered), phase 4 (post-marketing); "early phase 1" denotes exploratory testing before formal phase 1. Only a handful of trials on this page carry any drug phase at all (§3.1, §3.3).
- Device, surgical and strategy trials are phase-less by design: ClinicalTrials.gov assigns "NA" to interventional studies outside the drug-phase framework — which covers every endograft and surgical-technique trial here, and both strategy RCTs (TITAN:SvS, SUNDAY).
- Devices climb a different ladder under FDA investigational device exemption (IDE): an early feasibility study (EFS — small proof-of-principle cohorts under the FDA's EFS program, formalised 2013–2016; PMID 37778922) precedes a pivotal IDE trial whose results support premarket approval, followed by post-market surveillance. The ARISE lineage shows the rungs: ARISE (EFS, n=19) → ARISE II (pivotal, n=370) → ARISE III (pivotal extension to de novo type A dissection, n=112).
- The phase field therefore encodes regulatory pathway, not evidential weight: a "phase 1/2" label on a 50-patient pilot (NCT07483177) and "NA" on a 610-patient strategy RCT (NCT03536312) say nothing about which will change practice.
1. Shape of the field¶
From the searches run for this page (2026-08-27, re-run 2026-08-28 with identical counts), the interventional landscape stratifies as:
| Category | Representative volume | Comment |
|---|---|---|
| Endovascular / hybrid devices | Dominant share of active interventional trials | Single-arm, technical-success endpoints, 30-day and 5-year follow-up |
| Surgical technique and perioperative care | Large share | Cardiopulmonary bypass strategy, cerebral perfusion, haemostasis, ventilation, delirium |
| Disease-modifying pharmacotherapy for TAA | 2 trials retrieved (NCT07483177, NCT06280482) | Both small, both pilot/phase 1, neither event-powered |
| Threshold / management strategy | 1 randomised trial (NCT03536312) | The field's defining question |
| Observational registries and biorepositories | 142 observational studies returned under thoracic aortic aneurysm | Where most non-device knowledge is actually generated |
The imbalance is the finding. Devices get trials because they have sponsors and a regulatory pathway; the question of when to operate has neither.
2. The central trial: TITAN:SvS¶
| Field | Value |
|---|---|
| NCT | NCT03536312 |
| Name | Treatment in Thoracic Aortic Aneurysm: Surgery vs Surveillance (TITAN:SvS) |
| Sponsor | Ottawa Heart Institute Research Corporation, with 21 named collaborators across Canada and the US |
| Design | Prospective, multicentre, randomised, open (NA phase — surgical strategy) |
| Population | Age 18–79, asymptomatic ascending aortic aneurysm 5.0–5.4 cm maximal diameter on contrast CT. Patients at 4.5–4.9 cm are followed serially and enrolled on reaching 5.0 cm |
| Arms | Early elective ascending aortic surgery (within 10–12 weeks of randomisation) vs surveillance with annual CT; both arms receive maximal medical therapy (beta-blocker, statin, BP control) |
| N | 610 |
| Status | Recruiting (re-verified 2026-08-28); registry lists 29 sites — 28 recruiting, 1 not yet recruiting (St. Michael's, Toronto) |
| Dates | Start 2018-09-01; primary completion and completion 2035-05-31 |
| Primary endpoint | Composite of all-cause mortality and incidence of acute aortic events at 2 years |
| Key secondaries | 30-day operative mortality; incidence of crossover to surgery in the surveillance arm; aneurysm-related vs non-aneurysm-related death; stroke; prospectively measured annual growth rate in the surveillance arm; SF-36 quality of life at 1 and 2 years |
| Stated hypothesis | Early surgery will have significantly lower mortality and acute aortic events at 2 years |
Three design decisions determine what TITAN:SvS can and cannot answer:
- It excludes the genetically at-risk. Known connective tissue disease (Marfan, Loeys-Dietz, Turner), possible genetic aortopathy including any family history of aortic aneurysm or premature dissection, and inflammatory arteritis are all exclusions. The trial therefore answers the question for degenerative and bicuspid-valve-related aneurysm only — see syndromic aortopathies and bicuspid aortopathy.
- It excludes fast growers. Known expansion >0.5 cm/yr over the preceding 5 years is an exclusion — reasonable ethically, but it removes the subgroup in whom early surgery is most plausibly beneficial, biasing toward the null.
- The surveillance arm is itself a dataset. Prospective annual CT with core-lab review in a 5.0–5.4 cm cohort will produce the best growth-rate estimates ever collected at this diameter, independent of the randomised comparison — directly feeding epidemiology and natural history and risk stratification and size thresholds.
The 17-year interval between first enrolment and completion is the single most eloquent statement about why this field lacks evidence.
3. Trials table¶
Every NCT ID below was re-verified against ClinicalTrials.gov on 2026-08-28; the trials in the active tables were pulled individually a second time the same day for this enrichment pass. No trial changed recruitment status, including the several sitting at "Unknown" well past their completion dates. Differences captured against what this page previously recorded: NCT06827990 now carries the acronym ARISE III (study completion 2031-09); TITAN:SvS lists 29 sites rather than 28; SUNDAY's site list extends beyond Scandinavia to seven Dutch centres; and study-completion dates are now recorded separately from primary-completion dates where the two differ.
3.1 Strategy and pharmacotherapy¶
| NCT | Name / acronym | Intervention | Design | Phase | N | Status | Readout |
|---|---|---|---|---|---|---|---|
| NCT03536312 | TITAN:SvS | Early ascending aortic surgery vs surveillance | RCT, 29 registered sites | NA (strategy) | 610 | Recruiting | Primary completion 2035-05 |
| NCT07483177 | HEART pilot RCT | Spironolactone vs placebo, degenerative TAA | RCT, Mayo Clinic | 1/2 | 50 | Recruiting (start 2026-05) | Primary: change in carotid-femoral PWV at 6 months; completion 2028-05 |
| NCT06280482 | NR in SMDS | Nicotinamide riboside, ACTA2 Arg179 | Single-arm, UT Health Houston | 1 | 15 | Recruiting (start 2024-03) | Registered primaries are mostly cerebrovascular and systemic — cerebral oxygenation, cognition, autonomic symptoms, headache impact, spirometry, systolic BP — plus echocardiographic aortic diameter, all at 8 weeks; blood NAD⁺/NR are secondary; completion 2027-07 |
| NCT04197648 | EXTRA — exercise in TAA | Structured exercise, Laval University | RCT, single centre | NA | 30 | Recruiting (start 2025-06) | Primary: aneurysm-diameter stability at 3 months; primary completion 2026-10, completion 2027-04 |
| NCT07495267 | Nutritional ketosis in Marfan | Ketogenic diet, chronic aortic dissection in Marfan | Single-arm pilot, WashU | NA | 15 | Not yet recruiting (start 2026-03) | Primary: freedom from aortic intervention at 12 months; completion 2028-03 |
| NCT07205250 | Aspirin after endovascular arch repair (type B / non-A-non-B dissection and TAA) | Enteric-coated aspirin 100 mg vs placebo, 3 vs 6 vs 12 months | RCT, multicentre, China | Early phase 1 | 224 | Not yet recruiting (start 2026-02) | Primary: postoperative stroke at 1 year; primary completion 2028-01, completion 2028-09 |
| NCT05215587 | SUNDAY — Scandinavian Trial of Uncomplicated Aortic Dissection Therapy | TEVAR + medical therapy vs medical therapy alone | RCT, 28 sites (Denmark, Sweden, Norway, Finland, Iceland + 7 Dutch centres), Aarhus | NA (strategy) | 554 | Recruiting | Primary: 5-year overall survival; primary completion 2026-12, completion 2030-12 |
Precedent for SUNDAY. The randomised evidence SUNDAY is built to supersede is INSTEAD/INSTEAD-XL (NCT01415804): 140 patients with stable type B dissection randomised to TEVAR plus optimal medical therapy versus medical therapy alone. At 5 years, all-cause mortality was 11.1% vs 19.3% (p=0.13) — the non-significant ~8% absolute difference that SUNDAY's registry record cites as motivation — while aorta-specific mortality (6.9% vs 19.3%, p=0.04) and disease progression (27.0% vs 46.1%, p=0.04) favoured TEVAR, and landmark analysis of years 2–5 showed all-cause mortality 0% vs 16.9% (p=0.0003) with stent-graft-induced false-lumen thrombosis in 90.6% (Nienaber 2013, PMID 23922146). SUNDAY is powered directly on the endpoint INSTEAD-XL could not settle.
3.2 Marfan / heritable aortopathy drug trials — status only¶
Results for these are covered on medical therapy; listed here for registry status and as design precedent.
| NCT | Intervention | N | Status |
|---|---|---|---|
| NCT00429364 | Atenolol vs losartan, Pediatric Heart Network | 608 | Completed (2007–2014), 26 sites |
| NCT00763893 | Losartan vs placebo, AP-HP (Marfan Sartan) | 303 | Terminated |
| NCT01949233 | Irbesartan ± doxycycline, Oxford, CMR endpoints | 56 | Unknown status |
| NCT00782327 | Losartan vs placebo on beta-blocker, Ghent | 22 | Completed |
| NCT00723801 | Losartan vs atenolol, aortic stiffness, Brigham | 40 | Completed |
| NCT01145612 | Losartan vs atenolol, Spain | 140 | Unknown status |
| NCT00683124 | Losartan vs nebivolol vs both, Pavia | 291 | Unknown status |
| NCT01715207 | Aliskiren vs control on atenolol, Samsung | 30 | Completed |
| NCT01295047 | Atenolol vs perindopril vs verapamil crossover, Cardiff | 18 | Completed |
| NCT00485368 | Perindopril, aortic wall properties | 17 | Completed |
| NCT00593710 | Losartan vs atenolol, PWV, UBC | 17 | Completed |
The pattern is unmistakable: a long tail of 17–40-patient mechanistic studies with surrogate endpoints, two adequately sized trials, one termination, and several trials whose registry status was never updated. It took an individual-patient-data meta-analysis of 1,442 patients across seven trials to extract a reliable answer (Pitcher 2022, PMID 36049495) — see §6.
Readouts and safety. Only two trials in this table were sized to answer anything alone, and both read out null on their primary endpoint:
- Pediatric Heart Network (NCT00429364) — atenolol vs losartan, 608 children and young adults, 21 centres, 3 years (Lacro 2014, PMID 25405392). Efficacy: aortic-root z-score change −0.139 vs −0.107 SD-units/yr (p=0.08); both slopes significantly below zero. The 3-year rates of aortic-root surgery, dissection, death, and their composite did not differ. Safety: total adverse events 408 vs 365 (p=0.10); serious adverse events 40 vs 50 (p=0.31); events attributed to study drug higher on atenolol (204 vs 163, p=0.03); withdrawal 11% in each arm.
- Marfan Sartan (NCT00763893) — losartan vs placebo on top of baseline therapy (86% on beta-blocker), 303 randomised, median follow-up 3.5 years (Milleron 2015, PMID 25935877). Efficacy: sinus-of-Valsalva growth 0.44 vs 0.51 mm/yr (p=0.36); z-score slopes −0.043 vs −0.01 z/yr (p=0.69). Losartan lowered systolic and diastolic pressure by ~5 mmHg. Safety/events: deaths 0 vs 3 (sudden death, suicide, oesophageal cancer — none aortic); aortic surgery in 15 vs 13 patients.
Neither trial found the ARB advantage the mouse genetics promised; the positive ARB-versus-control signal emerged only at IPD-meta-analysis scale (§4.2).
3.3 Metformin and anti-inflammatory agents — the AAA/TAA distinction¶
Every metformin aneurysm trial retrieved is in abdominal aortic aneurysm. This matters: TAAD is genetically non-atherosclerotic and distinct from other vascular disease (PMID 37308786), and diabetes is inversely associated with both AAA and TAA by mechanisms that remain unresolved (Raffort 2018, PMID 30052821). Extrapolating an AAA growth-inhibition result to the ascending aorta is not supported.
| NCT | Name | Intervention | Population | N | Status | Endpoint |
|---|---|---|---|---|---|---|
| NCT04500756 | LIMIT | Metformin XR vs placebo | AAA 35–<50 mm (men), 30–<45 mm (women); excludes known TAA >5 cm or prior dissection, excludes ACTA2 at-risk allele | 314 | Recruiting (phase 2), Stanford/NHLBI | Change in maximal orthogonal infrarenal diameter by CTA, baseline to 2 yr; completion 2028 |
| NCT04224051 | MAAAGI | Metformin vs standard care | AAA | 500 | Unknown (phase 2), Uppsala | Growth |
| NCT03507413 | MetAAA | Metformin vs placebo, non-diabetic | AAA | 170 | Unknown (phase 2/3), Vienna | Growth |
| NCT05361772 | Low-dose colchicine | Colchicine vs placebo | AAA | 230 | Unknown, Guangdong | Growth |
| NCT01756833 | N-TA³CT | Doxycycline vs placebo | AAA | 261 | Completed (2013–2019), 22 sites, phase 2 | Growth |
| NCT00538967 | Doxycycline and MMP activity | Doxycycline | AAA | 60 | Completed, Leiden | MMP expression/activity |
| NCT00126204 | Adjunctive doxycycline with EVAR | Doxycycline | AAA | 75 | Completed | — |
Note the LIMIT exclusion criteria explicitly exclude thoracic aneurysm and ACTA2 carriers (NCT04500756) — the trialists themselves treat these as different diseases.
Readout. N-TA³CT is the one completed, adequately powered trial in this table, and it read out null with clean safety: 2-year change in maximum transverse aneurysm diameter 0.36 cm on doxycycline vs 0.36 cm on placebo (difference 0.0 cm, 95% CI −0.07 to 0.07, p=0.93); no patient was withdrawn for adverse effects, and joint pain was near-universal background noise in both arms (65% vs 63%) (Baxter 2020, PMID 32453369). MMP inhibition, the field's longest-running pharmacological hypothesis, failed cleanly — in the abdominal aorta.
3.4 Ascending endovascular devices¶
| NCT | Name / acronym | Device | Design | N | Status | Readout |
|---|---|---|---|---|---|---|
| NCT05800743 | ARISE II | GORE Ascending Stent Graft ± Thoracic Branch Endoprosthesis, vs surgical follow-up cohort | Prospective, multi-arm, 41 US sites | 370 | Enrolling by invitation | Co-primary: 30-day technical success + freedom from reintervention; 30-day composite safety (rupture, lesion-related mortality, disabling stroke, permanent paraplegia/paraparesis, new dialysis). Primary completion 2029-10, completion 2034 |
| NCT06827990 | ARISE III — GORE ASG in de novo type A dissection | GORE ASG | Prospective single-arm pivotal, 22 US sites | 112 | Recruiting (start 2025-09) | Primary safety (30-day death + disabling stroke) and effectiveness (30-day clinical success); primary completion 2030-09, completion 2031-09 |
| NCT02201589 | Ascending endovascular feasibility (single centre IDE) | Valiant PS-IDE stent graft | Feasibility | 20 | Recruiting since 2013 | Primary completion 2028-12, completion 2033-12 |
| NCT03322033 | Ascending endovascular feasibility (Baylor) | Valiant PS-IDE stent graft | Feasibility | 1 | Terminated (2018–2021) | Enrolled 1 patient |
| NCT05174767 | PERSEVERE | AMDS (Artivion) hybrid device, acute DeBakey I with malperfusion | Prospective single-arm IDE, 25–26 sites | 115 | Active, not recruiting | Co-primaries (30-day MAE composite; DANE) met — published, see readouts below; follow-up to 2029-12 |
| NCT06044259 | RADAR | Hemiarch ± AMDS stent | RCT, CUHK | 72 | Recruiting | Primary: DANE incidence within 12 months; primary completion 2026-09, completion 2027-09 |
| NCT02724072 | Thoraflex Hybrid IDE | Thoraflex Hybrid | Prospective, 12 US sites | 84 | Completed (2016–2021) | 1-year results published — see readouts below; no results posted on registry |
| NCT03214601 | Relay Branch early feasibility | Relay Branch thoracic stent graft, zone 0 | Early feasibility, 12 sites | 30 | Active, not recruiting | Primary completion 2023-09; completion 2028-08; no published readout retrieved |
| NCT06550986 / NCT06740721 | FLAGSHIP (NZ / France) — false-lumen embolisation | IMPEDE-FX RapidFill shape-memory polymer | First-in-human | 30 / 15 | Recruiting (start 2026-02) / not yet recruiting (start 2026-10) | Primary: technical success + 30-day major adverse events; primary completions 2027 |
ARISE lineage. The first-in-human ARISE early feasibility study enrolled 19 patients (mean age 75.7 yr, range 47–91; 57.9% female; 10 DeBakey I, 16 acute) at 7 of 9 US sites. Delivery and deployment succeeded in all cases; median procedure time 154 min (52–392); 30-day MACCE in 5 patients, comprising 3 deaths (15.8%), 1 disabling stroke (5.3%), 1 myocardial infarction (5.3%) (Roselli 2022, PMID 35587698). A mechanical substudy of 15 ARISE patients with ECG-gated CTA showed that ascending TEVAR damps pulsatile deformation of the stented segment (centreline curvature 0.224 ± 0.039 → 0.217 ± 0.039 cm systole–diastole, p<0.05) and mutes downstream native descending aortic pulsatility (Suh 2023, PMID 37144300) — a durability and remodelling concern that a 30-day technical-success endpoint would never surface. ARISE II (NCT05800743) is the pivotal successor — re-verified 2026-08-28: enrolling by invitation at 41 US sites, primary completion 2029-10, completion 2034-01 — and requires patients to be high surgical risk (age ≥75, previous sternotomy, or documented risk factors) and excludes known connective tissue disease. ARISE III (NCT06827990) extends the same platform to de novo type A dissection in patients high-risk for open repair. See surgical and endovascular repair.
Device readouts and safety numbers. Where the ascending/arch device trials have published, the numbers are:
- PERSEVERE, 30-day (NCT05174767) — 93 patients with acute DeBakey I dissection and malperfusion at 26 North American sites: 30-day mortality 9.7% (9/93), new disabling stroke 10.8%, new-onset renal failure requiring dialysis 19.4%, myocardial infarction 0; composite major-adverse-event rate 27% versus the 58% performance-goal reference cohort; zero distal anastomotic new entry (DANE) tears; technical success 99% — both co-primary endpoints met (Szeto 2025, PMID 39116932).
- PERSEVERE, 1-year — through 1 year: mortality 20% (19/93), new disabling stroke 12%, dialysis 20%, MI 2%, with most events inside 30 days; still no DANE on any postoperative CTA; freedom from unanticipated aortic reoperation 96%; false-lumen thrombosis 88–92% across arch zones 1–3 (Szeto 2026, PMID 41238072). These are single-arm results judged against historical references — exactly the comparison the randomised RADAR trial (NCT06044259) will put to a concurrent control.
- Thoraflex Hybrid IDE (NCT02724072) — 1-year freedom from the MAE composite (permanent stroke, permanent paraplegia/paraparesis, unanticipated aortic reoperation, all-cause death) 81% (51/63) in the primary group (n=65 recruited; the registry's performance goal was 57.4%), with mortality 11% (7 deaths, 2 within 30 days/discharge), permanent stroke 5%, permanent paraplegia/paraparesis 5%; 41% underwent planned distal extension procedures at a median 122 days. In the rupture group (n=9): 1-year freedom from MAE 71%, mortality 14%, permanent stroke 29% (Coselli 2024, PMID 36253292). FDA approval followed in April 2022, making hybrid frozen-elephant-trunk devices commercially available in the US for the first time (PMID 40808788). The registry posts no results; the publication is the readout.
- RelayBranch EFS (NCT03214601) — primary completion passed 2023-09, but no trial-level publication was retrieved in this session's PubMed searches and no results are posted on ClinicalTrials.gov: three years on, the zone 0 branched-arch early feasibility study remains publicly unread.
3.5 Registries, biorepositories, and mechanism-facing observational studies¶
| Identifier | Name | Scope | N | Status |
|---|---|---|---|---|
| — (PMID 21982653) | GenTAC — National Registry of Genetically Triggered Thoracic Aortic Aneurysms and Cardiovascular Conditions | NHLBI registry established 2006; phase I 2007–2010 enrolled 2,046 patients (Marfan 576 [28.2%], BAV disease 504 [24.6%], aneurysm/dissection age <50 369 [18%]); DNA available in 97%, stored plasma in 60%; phase II added imaging and phenotyping core labs | 2,046 → +1,500 | Legacy resource; continues to generate analyses (e.g. PMID 37776192) |
| — (PMID 39424426) | Montalcino Aortic Consortium (MAC) | International heritable thoracic aortic disease registry; 2014–2023 cross-sectional analysis of 672 participants with pathogenic variants, genotype-stratified phenotyping | 672 in the cited analysis | Active; not registered as a ClinicalTrials.gov study in these searches |
| NCT03440697 | Pathogenetic Basis of Aortopathy and Aortic Valve Disease (Yale / NHLBI) | Paired blood and tissue biorepository with genome-wide sequencing, histopathology, transcriptomics, proteomics across 22 listed aortopathy conditions | 3,000 | Active, not recruiting; completion 2030 |
| NCT02256163 | Identification of genes and pathogenesis in familial TAA (AP-HP) | Familial TAA gene discovery | 258 | Completed (2011–2016) |
| NCT06783803 | Linkage analysis in familial aneurysms | Novel hereditary factors | 20 | Active, not recruiting; completion 2029 |
| NCT03142074 | Biomechanical and microstructural properties of the aortic wall (KU Leuven) | ATAA tissue mechanics + ECG-gated CT | 180 | Unknown; primary completion 2025-12 |
| NCT03410420 | Pimonidazole hypoxia detection in human thoracic aorta | Direct tissue hypoxia mapping, phase 1 | 9 | Completed |
| NCT03090763 | Micro AAA | Circulating microRNA in abdominal and thoracic aneurysm | 200 | Unknown |
| NCT05715203 | Aortic stiffness in genetic aortopathies | Pulsotonometry | 250 | Recruiting |
| NCT07034430 | Genetic risk stratification below surgical threshold | Retrospective single-centre | 102 | Active, not recruiting |
| NCT06377033 | Behavioural nudges for genomic medicine uptake (incl. TAA) | Implementation science, UPenn | 1,000 | Recruiting; completion 2027 |
Registries carry disproportionate weight in this field precisely because the randomised evidence is so thin. GenTAC's design — enrol, bank DNA and plasma, add imaging and phenotyping core labs in phase II (PMID 21982653) — is the template that Yale's NCT03440697 follows at 3,000 patients. MAC's contribution is genotype-stratified phenotyping at a granularity no single centre reaches: for example, resolving that mitral regurgitation and mitral valve prolapse, but not mitral annular disjunction, associate with adverse aortic and cardiac events, and that severe disjunction >10 mm occurs only in TGF-β-pathway variant carriers (PMID 39424426).
4. Why trials are hard in thoracic aortic disease¶
4.1 Event rates¶
Acute aortic events in patients under surveillance below the operative threshold occur at rates of order 1%/yr. A superiority trial on a hard composite therefore needs either very large numbers or very long follow-up. TITAN:SvS chose the latter: 610 patients, 2-year primary endpoint, and a 2035 completion date (NCT03536312). Compare the disease-modification pipeline in AAA, where growth is continuous and measurable and a 314-patient trial can be powered on 2-year diameter change (NCT04500756) — abdominal aneurysm has a usable surrogate; the thoracic aorta largely does not.
4.2 The surrogate endpoint problem¶
Marfan trials solved feasibility by adopting body-surface-area-adjusted aortic root Z-score change as the primary endpoint. The individual-patient-data meta-analysis quantifies exactly how small the signal is: across four ARB-versus-control trials (676 participants, median 3 years), the annual increase in root Z-score was 0.07 (SE 0.02) on ARB versus 0.13 (SE 0.02) on control — an absolute difference of −0.07 (95% CI −0.12 to −0.01, p=0.012). ARBs and beta-blockers were indistinguishable (difference 0.03, −0.05 to 0.10, p=0.48), giving an indirect beta-blocker-versus-control estimate of −0.09 (−0.18 to 0.00, p=0.042) (Pitcher 2022, PMID 36049495).
Three consequences follow:
- Effect sizes on the surrogate are at the edge of measurement error. A difference of 0.07 Z-score units per year is smaller than typical inter-observer variability in root measurement — see imaging and surveillance. Only pooled IPD across seven trials made it detectable.
- The surrogate has never been formally validated against dissection. Whether halving Z-score progression halves dissection risk is assumed, not demonstrated. The meta-analysis authors phrase the clinical inference conditionally — combination therapy "would be expected to lead to a delay in the need for aortic surgery" (PMID 36049495).
- It rewards subgroup effects that then need replication. ARB effects were larger in patients with pathogenic FBN1 variants (heterogeneity p=0.0050) (PMID 36049495) — biologically plausible, statistically fragile, and a natural target for the next trial.
Device trials sidestep the problem by using technical success and 30-day composite safety (NCT05800743), which are achievable but say nothing about whether the strategy is better than surgery in the long run.
4.3 Structural obstacles¶
| Obstacle | Concrete evidence |
|---|---|
| Decade-scale horizons | TITAN:SvS 2018 → 2035 (NCT03536312); Yale biorepository → 2030 (NCT03440697); several device trials to 2034 |
| Highest-risk patients excluded | TITAN:SvS excludes connective tissue disease, family history of aneurysm/premature dissection, and growth >0.5 cm/yr; ARISE II and LIMIT both exclude known connective tissue disease (NCT05800743, NCT04500756) |
| Equipoise is site-specific | The randomised question ("operate at 5.0 cm?") is one many surgeons believe they already know the answer to, which slows enrolment across 28 sites |
| Registry status decay | Multiple Marfan trials sit at "Unknown" status years past their completion dates (NCT01145612, NCT00683124, NCT01949233) — the registry itself underestimates how many trials failed |
| Small single-centre pilots dominate | 17-, 18-, 22-, 30-, 40-patient trials recur throughout §3.2; almost none was powered for anything |
| Rarity and referral concentration | Heritable aortopathy is rare enough that international consortia (MAC, PMID 39424426; GenTAC, PMID 21982653) are the only route to adequate n |
5. The completed-trial graveyard, and what it taught¶
| Trial | Fate | Lesson |
|---|---|---|
| NCT01361087 — Circulating TGF-β in Marfan | Withdrawn, 0 enrolled | The most-cited candidate biomarker in aortopathy never got a dedicated prospective study off the ground; see biomarkers |
| NCT03322033 — Valiant PS-IDE ascending feasibility (Baylor) | Terminated after 1 patient (2018–2021) | Ascending endovascular repair is anatomically constrained: the ARISE II criteria (landing zones ≥2 cm, ≥2 cm from coronary ostia and brachiocephalic artery, diameter 27–48 mm, no heavy calcification or thrombus) exclude most candidates. Single-centre IDEs cannot find enough patients |
| NCT02538822 — Rupture-risk assessment from dynamic imaging (Saint-Étienne) | Terminated, 31 of intended enrolment (2014–2018) | Biomechanical rupture-prediction studies fail on the same arithmetic as everything else: ruptures do not accrue |
| NCT00763893 — Marfan Sartan (AP-HP) | Terminated, 303 enrolled | Even the larger ARB trials struggled to complete; the field's answer came from pooling, not from any single trial |
| NCT01756833 — N-TA³CT doxycycline in AAA | Completed 2019, 261 patients, 22 sites; null — 2-yr growth 0.36 cm in both arms (PMID 32453369) | MMP inhibition as a strategy was properly tested in AAA at adequate scale. That the concept was never carried into a thoracic trial reflects both the AAA/TAA distinction and the absence of a thoracic growth surrogate |
| NCT04381507 — CONFORM-TAA post-market device study | Terminated after 1 patient (2020–2024) | Post-market device evidence in descending TAA is fragile; registry-based surveillance has largely replaced it (e.g. NCT04545502, 2,000 patients) |
| NCT02299947 — FIBTEG fibrinogen after ascending surgery | Terminated, 27 of 200 (2014–2018) | Perioperative RCTs in aortic surgery also under-recruit |
The through-line: nothing fails here for lack of a hypothesis; it fails for lack of patients and events per unit time. Every solution the field has adopted — surrogate endpoints, IPD meta-analysis, international registries, single-arm device studies with technical-success endpoints — is a workaround for that one constraint.
6. What would change the landscape¶
- A validated surrogate for dissection risk, which would convert 15-year trials into 3-year trials. This is the strongest argument for the biomarker and imaging work on biomarkers and hemodynamics and biomechanics.
- Genotype-stratified randomisation. The FBN1-variant heterogeneity signal in the ARB meta-analysis (p=0.0050, PMID 36049495) and the existence of at least one genotype-directed trial (NCT06280482) suggest enrichment designs are feasible where whole-population designs are not.
- Registry-embedded randomisation. GenTAC (PMID 21982653), MAC (PMID 39424426) and institutional biorepositories such as NCT03440697 already hold the phenotyped populations; running randomisations inside them would sidestep the recruitment bottleneck that terminated NCT03322033, NCT02538822 and NCT02299947.
- A thoracic answer to the AAA drug question. Metformin, colchicine and doxycycline have all been tested at scale in AAA and never in TAA. Whether that reflects justified biological caution (PMID 37308786) or simple neglect is unresolved.
Open questions¶
- Will TITAN:SvS (NCT03536312) be adequately powered given that it excludes fast growers (>0.5 cm/yr) and all patients with known or suspected heritable aortopathy — the two groups in whom early surgery is most plausibly beneficial?
- Is the aortic root Z-score a valid surrogate for dissection? An annual difference of 0.07 Z units (PMID 36049495) is treated as clinically meaningful without any trial linking Z-score trajectory to hard events.
- Does the larger ARB effect in FBN1 pathogenic-variant carriers (heterogeneity p=0.0050, PMID 36049495) replicate prospectively, and should it govern who is randomised in the next Marfan trial?
- Should metformin be tested in thoracic aneurysm at all, given that all three registered trials are abdominal (NCT04500756, NCT04224051, NCT03507413), that LIMIT explicitly excludes TAA and ACTA2 carriers, and that the diabetes–aneurysm inverse association is mechanistically unexplained (PMID 30052821)?
- What is the durability consequence of ascending TEVAR damping pulsatile deformation in both the stented segment and the native descending aorta (PMID 37144300)? ARISE II's 30-day co-primary endpoints (NCT05800743) cannot detect it, and follow-up runs to 2034.
- Can registry-embedded randomisation within GenTAC-like (PMID 21982653) or MAC-like (PMID 39424426) cohorts solve the recruitment failures that terminated NCT03322033 (n=1), NCT02538822 and NCT02299947?
- Why has no interventional trial targeted the HDAC9/MALAT1 axis in human aortic disease despite convergent genetic and pharmacological validation (see omics and emerging science)?
- PERSEVERE met both 30-day co-primary endpoints against a historical performance goal (MAE 27% vs 58%, PMID 39116932), yet 1-year mortality was still 20% (PMID 41238072). Will the randomised RADAR comparison (NCT06044259) reproduce the DANE-suppression effect against a concurrent control?
- Why has the RelayBranch early feasibility study (NCT03214601) produced no publication retrievable by this page's searches, and no posted registry results, three years after its 2023-09 primary completion?
Related pages¶
- risk stratification and size thresholds — the thresholds TITAN:SvS is testing.
- medical therapy — results of the beta-blocker and ARB trials listed here by status only.
- surgical and endovascular repair — the operations and devices under trial.
- aortic dissection — the event that is the endpoint of nearly every trial on this page.
- epidemiology and natural history — the growth and event rates that determine trial feasibility.
- genetics of TAA — the genotypes used for enrichment and exclusion.
- syndromic aortopathies and bicuspid aortopathy — populations included in or excluded from the key trials.
- omics and emerging science — the target pipeline that has not yet reached trials.
- biomarkers — the missing surrogate endpoint.
- guidelines — the recommendations these trials are designed to change.
References¶
- Pitcher A, Spata E, Emberson J, et al. Angiotensin receptor blockers and β blockers in Marfan syndrome: an individual patient data meta-analysis of randomised trials. Lancet. 2022;400:822-831. PMID 36049495
- Roselli EE, Atkins MD, Brinkman W, et al. ARISE: First-In-Human Evaluation of a Novel Stent Graft to Treat Ascending Aortic Dissection. J Endovasc Ther. 2023;30:550-560. PMID 35587698
- Suh GK, Bondesson J, Zhu YD, et al. Ascending Aortic Endograft and Thoracic Aortic Deformation After Ascending Thoracic Endovascular Aortic Repair. J Endovasc Ther. 2023;32:7-17. PMID 37144300
- Kroner BL, Tolunay HE, Basson CT, et al. The National Registry of Genetically Triggered Thoracic Aortic Aneurysms and Cardiovascular Conditions (GenTAC): results from phase I and scientific opportunities in phase II. Am Heart J. 2011;162:627-632.e1. PMID 21982653
- Asokan KL, Landes JR, Renders W, et al. Mitral Annular Disjunction in Heritable Thoracic Aortic Disease: Insights From the Montalcino Aortic Consortium. J Am Heart Assoc. 2024;13:e036274. PMID 39424426
- Stephens SB, Shalhub S, Dodd N, et al. Vertebral Tortuosity Is Associated With Increased Rate of Cardiovascular Events in Vascular Ehlers-Danlos Syndrome. J Am Heart Assoc. 2023;12:e029518. PMID 37776192
- Klarin D, Devineni P, Sendamarai AK, et al. Genome-wide association study of thoracic aortic aneurysm and dissection in the Million Veteran Program. Nat Genet. 2023;55:1106-1115. PMID 37308786
- Raffort J, Lareyre F, Clément M, et al. Diabetes and aortic aneurysm: current state of the art. Cardiovasc Res. 2018;114:1702-1713. PMID 30052821
- Lacro RV, Dietz HC, Sleeper LA, et al. Atenolol versus losartan in children and young adults with Marfan's syndrome. N Engl J Med. 2014;371:2061-2071. PMID 25405392
- Milleron O, Arnoult F, Ropers J, et al. Marfan Sartan: a randomized, double-blind, placebo-controlled trial. Eur Heart J. 2015;36:2160-2166. PMID 25935877
- Nienaber CA, Kische S, Rousseau H, et al. Endovascular repair of type B aortic dissection: long-term results of the randomized investigation of stent grafts in aortic dissection trial. Circ Cardiovasc Interv. 2013;6:407-416. PMID 23922146
- Baxter BT, Matsumura J, Curci JA, et al. Effect of doxycycline on aneurysm growth among patients with small infrarenal abdominal aortic aneurysms: a randomized clinical trial. JAMA. 2020;323:2029-2038. PMID 32453369
- Szeto WY, Fukuhara S, Fleischman F, et al. A novel hybrid prosthesis for open repair of acute DeBakey type I dissection with malperfusion: early results from the PERSEVERE trial. J Thorac Cardiovasc Surg. 2025;170:114-123.e3. PMID 39116932
- Szeto WY, Fukuhara S, Fleischman F, et al. One-year results of novel aortic arch hybrid prosthesis for repair of acute DeBakey type I dissection with malperfusion: PERSEVERE study. Ann Thorac Surg. 2026;121:1069-1079. PMID 41238072
- Coselli JS, Roselli EE, Preventza O, et al. Total aortic arch replacement using a frozen elephant trunk device: results of a 1-year US multicenter trial. J Thorac Cardiovasc Surg. 2024;167:1680-1692.e2. PMID 36253292
- Orozco-Sevilla V, Coselli JS, Green SY, et al. Total aortic arch replacement using the Thoraflex Hybrid device: evolution from investigational to federally approved use in the United States. Ann Cardiothorac Surg. 2025;14:279-290. PMID 40808788
- Holmes DR, Farb A, Dib N, et al. The medical device development ecosystem: current regulatory state and challenges for future development: a review. Cardiovasc Revasc Med. 2023;60:95-101. PMID 37778922
Trial registrations cited¶
TITAN:SvS NCT03536312 · HEART spironolactone NCT07483177 · Nicotinamide riboside in SMDS NCT06280482 · Exercise in TAA NCT04197648 · Ketogenic diet in Marfan NCT07495267 · Aspirin in TBAD/TAA NCT07205250 · Scandinavian uncomplicated TBAD NCT05215587 · Pediatric Heart Network atenolol/losartan NCT00429364 · Marfan Sartan NCT00763893 · Oxford irbesartan/doxycycline NCT01949233 · Ghent losartan NCT00782327 · Brigham losartan/atenolol NCT00723801 · Spain losartan/atenolol NCT01145612 · Pavia losartan/nebivolol NCT00683124 · Aliskiren NCT01715207 · Cardiff crossover NCT01295047 · Perindopril NCT00485368 · UBC losartan/atenolol NCT00593710 · Circulating TGF-β in Marfan (withdrawn) NCT01361087 · LIMIT NCT04500756 · MAAAGI NCT04224051 · MetAAA NCT03507413 · Colchicine AAA NCT05361772 · N-TA³CT NCT01756833 · Leiden doxycycline NCT00538967 · Adjunctive doxycycline NCT00126204 · ARISE II NCT05800743 · GORE ASG type A NCT06827990 · Valiant PS-IDE NCT02201589 · Valiant PS-IDE Baylor (terminated) NCT03322033 · PERSEVERE NCT05174767 · RADAR NCT06044259 · Thoraflex Hybrid NCT02724072 · Relay Branch NCT03214601 · IMPEDE-FX RapidFill NCT06550986, NCT06740721 · Yale aortopathy biorepository NCT03440697 · AP-HP familial TAA genes NCT02256163 · Linkage analysis familial aneurysms NCT06783803 · KU Leuven wall biomechanics NCT03142074 · Pimonidazole hypoxia NCT03410420 · Micro AAA NCT03090763 · Aortic stiffness in genetic aortopathies NCT05715203 · Genetic risk below threshold NCT07034430 · Genomic medicine nudges NCT06377033 · Saint-Étienne rupture risk (terminated) NCT02538822 · CONFORM-TAA (terminated) NCT04381507 · FIBTEG (terminated) NCT02299947 · Terumo graft registry NCT04545502 · INSTEAD/INSTEAD-XL NCT01415804