Not yet independently audited

This page was written from verified sources but has not yet passed the independent citation audit. Treat its claims as provisional.

Open questions — thoracic aortic aneurysm¶

Last curated: 2026-10-28. This is the repository's research frontier: concrete, answerable questions the assembled literature exposes, each with the evidence showing why it is open and what would close it. Questions carry stable IDs (OQ-n) so curation sessions can reference, sharpen, or close them; when one is answered, move it to a "Closed" section with the closing citation rather than deleting it. Supporting PMIDs/NCTs were verified during the sessions that wrote them (CONVENTIONS §1).

How to read priority¶

Tier 1 = answering it would change practice or reorder the field; a study design is conceivable today. Tier 2 = important, but blocked on tools, cohorts, or a Tier-1 answer. Tiers are curator judgment, not consensus — argue with them.

Dots not yet connected¶

A different lens on the same frontier: places where two bodies of evidence both exist in this knowledge base but no study has ever joined them. These are synthesis opportunities — often cheaper than new data collection, because the dots are already on the page.

#	Dot A	Dot B	The missing junction	Powers
D1	~130 GWAS loci for aortic diameter (PMID 34837083, 35902171)	21 GWAS loci for aortic dissection/aneurysm disease (PMID 37308786)	No formal reconciliation — are "genetics of size" and "genetics of failure" the same trait?	OQ-6
D2	Circulating biomarker discovery cohorts (cross-sectional; PMID 23312977, 30373522)	Surveillance cohorts with serial imaging and events	No study has ever banked baseline plasma in a surveillance cohort and read markers against subsequent growth	OQ-7
D3	Within-patient WSS→wall-degradation mapping in BAV (PMID 26293758)	BAV heritability ~89% (PMID 15234422)	No design separates the inherited wall defect from the flow-imposed one (e.g., discordant-flow longitudinal cohorts)	OQ-17, OQ-11
D4	Ex vivo biomechanical rupture indices from resected tissue (PMID 25979384, 35894942)	Prospective clinical outcomes in surveillance patients	Every rupture metric is derivation-only; none has met a prospective event	OQ-2
D5	Combined monogenic–polygenic TAAD association (PMID 42678067)	Imaged surveillance cohorts with known diameter	Incremental prospective event prediction and clinical utility remain to be established	OQ-18
D6	Mouse TGF-β/ARB rescue (PMID 16601194)	Human trial near-null (PMID 36049495)	No back-translation study explains the divergence (dose? timing? surrogate validity? species biology?)	OQ-4, OQ-19
D7	Sex-difference data — smaller-diameter dissection, faster growth, higher mortality (PMID 15197151, 30571545)	Threshold-setting analyses (all sex-neutral)	No sex-stratified threshold derivation or validation exists	OQ-3
D8	Patient-designed monitoring priorities (PMID 42790905)	TITAN:SvS already registers SF-36 quality of life (NCT03536312)	Validate a disease-specific instrument and test whether patient-designed monitoring improves outcomes; the earlier claim of zero patient-reported trial endpoints was incorrect	OQ-35, OQ-37, OQ-38
D9	IRAD reports type A in-hospital mortality falling 31%→22% across 28 centres, 1995–2013, among 2,952 type A patients, with surgical management rising 79%→90% (Pape 2015, PMID 26205591); the 2024 update reports 24.1%→16.7% hospital mortality across 61 centres in 15 countries to 2022, with surgical repair 89.1%→92.5%. Its 3-year postdischarge survival did not differ significantly across eras (p=0.296) (Bossone 2024, PMID 39145396)	The Danish nationwide cohort reports type A 30-day mortality of 22.0% with no significant change over time, 1996–2016, in 1,620 chart-validated cases drawn from hospital admissions with available medical records (Obel 2022, PMID 36321467)	Unresolved trend difference, with non-comparable measures. In-hospital versus 30-day mortality and centre-registry versus validated hospital-admission populations prevent a direct numerical comparison; they do not establish that the improving and flat trends are compatible. Bossone's unchanged postdischarge survival is conditional on surviving hospital care and does not resolve the 30-day trend. Denmark excludes prehospital deaths, and the extent of transfer selection and missing records remains relevant. A harmonised analysis would use the same time horizon and inclusion rules, then partition trend differences by treatment, case mix, transfers and risk-adjusted case fatality	OQ-27
D10	Diabetes inversely associated with aneurysm disease (PMID 30052821)	TAA mechanism/therapeutic pipeline	The protective mechanism has never been studied in thoracic disease; all interventional tests are abdominal-only	OQ-14
D11	Aortic length/tortuosity as dissection markers (PMID 31526537)	Guideline intervention criteria (diameter-only)	No multi-center prospective validation to carry length into criteria	OQ-24
D12	Guideline thresholds resting on Yale single-center natural history (PMID 11834007)	Population imaging biobanks now large enough to replicate	The replication has simply never been run	OQ-26

When a junction study appears in the literature, connect the dot: cite it here, and usually a question above sharpens or closes with it.

Tier 1 — highest-value directions¶

OQ-1. Does elective repair at 5.0–5.4 cm beat surveillance?¶

The central clinical question of the field. Guidelines permit 5.0 cm at experienced centers (Class 2a) on purely observational grounds; 59% of type A dissections occur below 5.5 cm, yet absolute risk in the 4.0–5.5 cm band is ~0.1%/patient-year — so earlier surgery could either save lives or mostly expose low-risk patients to operative mortality. First randomized evidence ever: TITAN:SvS (NCT03536312, n=610, readout ~2035). Until then, every threshold argument is denominator arithmetic. (PMID 36322642, 17709637, 25997607) → wiki: risk stratification, trials

OQ-2. Who below threshold dissects? (individualized risk beyond diameter)¶

Diameter is the only variable in guidelines, yet most dissections happen below its action point, and the population reservoir of 4.0–5.4 cm aortas is enormous. Candidate discriminators exist in silos — indexed size (ASI/AHI), aortic length (11 cm hinge), wall stress/strength metrics, 4D-flow WSS, polygenic scores — but none has prospective validation against events: every biomechanical rupture index derives from tissue resected at elective surgery, every biomarker cohort is cross-sectional, and no analysis shows a polygenic score adds prediction conditional on measured diameter. The unifying study (multi-modal baseline → longitudinal events) has never been run. (PMID 17709637, 16368358, 29395211, 31526537, 25979384, 35894942, 37662232) → wiki: risk stratification, hemodynamics, biomarkers

OQ-3. Should women have lower thresholds?¶

Women are diagnosed older, dissect and rupture at smaller absolute diameters, contribute 79% of ruptures in population data, grow faster in some series (2×), and carry higher adjusted dissection mortality (OR ~1.4) — yet every major guideline's primary threshold is sex-neutral (body-size indexing is a partial patch, not a validated sex-specific criterion). No trial or even large prospective cohort has tested sex-stratified action points. (PMID 15197151, 9851478, 30571545, 36322642) → wiki: risk stratification, epidemiology

OQ-4. Is elevated TGF-β signalling causal, compensatory, or epiphenomenal?¶

The field's oldest mechanistic dispute, and it underwrites drug strategy. Loss-of-function mutations in TGF-β receptors/ligands paradoxically increase tissue TGF-β signalling; TGF-β antagonism rescues mouse Marfan aortopathy, yet receptor deletion in SMCs worsens disease and Tgfb2 haploinsufficiency aggravates Fbn1 mice. No experiment has separated the cell type, compartment, and time window in which TGF-β harms from those in which it protects. Resolving this would explain why losartan underdelivered clinically and what to target instead. (PMID 12598898, 16601194, 24401272, 22772368, 28119285) → wiki: pathophysiology, medical therapy

OQ-5. Can any drug slow degenerative (non-syndromic) TAA — and is such a trial even powerable?¶

No RCT shows any drug slows growth or prevents dissection in the degenerative majority; recommendations reduce to blood-pressure control. Worse, the trial mathematics are hostile: typical growth (~0.1 cm/yr) sits below inter-scan measurement error (±2.4–5.2 mm), and the only accepted hard endpoints are rare. A validated growth/risk surrogate (OQ-2, OQ-7) is the unlock; until then the "medical therapy of TAA" literature is a Marfan literature. October 2026: a systematic review of observational statin and metformin studies reports a TAA statin-growth association of MD −0.32 mm/year (three studies, I² = 0%) versus a larger AAA association (MD −0.93; subgroup difference p = 0.008); GRADE low to very low, TAA underrepresented, overall prediction interval including the null, and no TAA RCT. Metformin evidence remains AAA-only; its one RCT (n=58 non-diabetic AAA) found no difference (Rodríguez-Palomares 2026, PMID 42942351). (PMID 11834007, 22864960, 36049495, 36322642, 42942351) → wiki: medical therapy

OQ-6. What explains the ~75% of familial TAAD with no identified gene?¶

Gene discovery by family exome sequencing is saturating — each large effort returns fewer new large-effect genes — implying the residual architecture is oligogenic, non-coding, or polygenic; meanwhile the two GWAS families (aortic diameter loci, ~100+, vs dissection/aneurysm disease loci, 21) have never been formally reconciled, so "genetics of size" and "genetics of failure" may be different questions. Both the missing-heritability hunt and the reconciliation are tractable with existing biobanks. October 2026: the biobank route is being taken — a five-biobank gene-based rare-variant burden meta-analysis (>10,000 cases, 880,000 controls) reports eight novel replicated candidate genes (FNDC3B, ROCK1, URM1, SLFN11, ENPP1, CLEC16A, CREM, VCAN), with follow-up variants found primarily in sporadic dissection or aortic-surgery cohorts (FNDC3B also in one HTAD family), and a companion analysis finds aortic phenotypes across de novo variants in syndromic genes outside HTAD panels (DNMT3A, NF1, PTPN11 and others) — both medRxiv preprints, not peer reviewed; family-based discovery still adds genes one at a time (ADAMTS6 loss of function, four unrelated individuals plus a mouse model). None yet has a ClinGen validity curation, so the "unexplained" fraction is shrinking in candidates, not in actionable diagnoses. (PMID 26838787, 34837083, 35902171, 37308786, 42818172, 42818859, 42829835) → wiki: genetics, omics

OQ-7. Can a circulating biomarker predict growth or dissection prospectively?¶

Every candidate (MMPs, TGF-β, D-dimer variants, miRNAs) was validated by discriminating operated aneurysms from normal aortas cross-sectionally — a design that cannot support surveillance use. No study has banked baseline plasma in a surveillance cohort and read markers against subsequent growth/events. This is a fixable design failure, not (yet) a demonstrated biology failure. (PMID 23312977, 30373522, 31409059) → wiki: biomarkers

OQ-8. Are gene-specific surgical thresholds correct, or only plausible?¶

Genotype-adjusted thresholds (e.g., TGFBR2 at 4.5 cm, lower with risk modifiers) rest on registry observations as thin as 6 dissections below 45 mm among 441 carriers; no prospective comparison of threshold strategies exists, so number-needed-to-operate per genotype is unknown. As panel testing spreads, miscalibrated thresholds scale their errors with it. (PMID 27879313, 36322642) → wiki: genetics, guidelines

Tier 2 — by theme¶

Mechanism¶

OQ-9. Which compartment drives disease — medial SMC or adventitial fibroblast — and does it differ by genotype? Marfan models show a modulated-SMC cluster; LDS tissue shows adventitial fibroblast activation without one. Model artifact or genotype-specific biology? (PMID 32698686, 40109260) → pathophysiology
OQ-10. Is the modulated-SMC state causal for wall failure or reactive to it? All human atlases are end-stage surgical snapshots; temporality is structurally unresolvable with current tissue access — needs longitudinal animal sampling or in-situ lineage/clock methods. (PMID 32698686, 36172868) → pathophysiology, omics
OQ-11. Is high or low wall shear stress the dangerous exposure? Elevated regional WSS co-localizes with elastin degradation within patients, yet low WSS correlates with degraded tissue properties elsewhere, and WSS falls as the aorta dilates. Stage-dependence vs opposing mechanisms is untested — needs serial 4D-flow with tissue endpoints. (PMID 26293758, 31331875) → hemodynamics
OQ-12. Does embryologic lineage (second heart field vs neural crest) confer differential vulnerability or only location? Disease localizes to SHF-derived outer media, but lineage markers are largely undetectable in adult human aorta, leaving the human relevance inferential. (PMID 35143327, 39697172) → pathophysiology
OQ-13. How much of the angiotensin-II mouse literature measured pseudoaneurysm rather than dissection? Careful re-phenotyping suggests endpoint misclassification in a foundational model family; a systematic re-read could reweight two decades of mechanism claims. (PMID 35132962) → animal models
OQ-14. Why does diabetes associate with less aneurysm disease? The inverse association is reproducible and mechanistically unexplained; all interventional tests (metformin) are abdominal-only and explicitly exclude TAA. A TAA-specific mechanism study would either export the protection or kill the analogy. October 2026: a narrative review still treats the diabetes paradox as primarily an AAA finding, with only limited TAA/dissection evidence, and a concurrent SRMA finds metformin growth associations in AAA observational cohorts only, with a null AAA RCT (Orchi 2026, PMID 42940178; Rodríguez-Palomares 2026, PMID 42942351). (PMID 30052821, 42940178, 42942351; NCT04500756, NCT04224051, NCT03507413) → medical therapy, trials

Genetics¶

OQ-15. How much penetrance variation can monogenic–polygenic models explain? Two biobanks now show higher TAAD prevalence in pathogenic-variant carriers with high versus low PRS (RR 2.32, 95% CI 1.29–4.17). Gene-specific, age-dependent absolute risks and prospective external validation remain needed before clinical threshold changes. (DePaolo 2026, PMID 42678067) → omics, genetics
OQ-16. How should MYLK-type disease be surveilled when it dissects without dilating? Diameter — the universal surveillance variable — is structurally uninformative here; no validated alternative (biomarker, imaging, fixed-age prophylaxis) has been proposed. (PMID 21055718) → genetics, imaging
OQ-17. Is BAV aortopathy hemodynamic, genetic, or sequential? Within-patient paired sampling shows shear-localized wall degradation (genotype held constant) while BAV heritability is ~89%; no design yet separates the intrinsic wall defect from the flow-imposed one — discordant-flow longitudinal cohorts could. (PMID 26293758, 15234422) → bicuspid aortopathy
OQ-18. Does a polygenic score add value conditional on known diameter? Combined monogenic–polygenic associations with EHR-defined TAAD now exist, but this does not demonstrate incremental prospective dissection prediction once diameter is known. Test calibration, discrimination and clinical utility in imaged surveillance cohorts. (DePaolo 2026, PMID 42678067) → omics

Therapy & trials¶

OQ-19. Is aortic root Z-score a valid surrogate for dissection? The entire Marfan drug literature rests on Z-trajectory differences (e.g., 0.07/yr) never linked to hard events; a registry-based surrogate-validation study is feasible and overdue. (PMID 36049495) → medical therapy
OQ-20. Does medical therapy prevent hard endpoints in Marfan? All randomized evidence is surrogate growth; the only event-level signal is non-randomized follow-up. (PMID 36049495, 32548624) → medical therapy
OQ-21. Is the fluoroquinolone–aorta association causal? Strong cohort signals attenuate toward null under active-comparator designs with baseline imaging; regulatory warnings may be steering therapy on confounding. A target-trial-emulation with imaging covariates would settle it. (PMID 29519881, 32897358, 32897307) → medical therapy
OQ-22. Does PEARS change outcomes versus waiting for valve-sparing root replacement? ~700 implants, zero randomized comparison. (PMID 30165980, 36094493) → surgery
OQ-23. What is the long-term cost of ascending TEVAR damping aortic pulsatility? Stent grafts stiffen the system (with effects propagating to native segments); trial endpoints (30-day) cannot see late ventricular/vascular consequences. Needs a decade of device-cohort follow-up with functional endpoints. (PMID 37144300; NCT05800743) → surgery, trials
OQ-41. Are calcium-channel blockers harmful to the human thoracic aorta? Mouse mechanism and a confounded human heritable-TAA cohort (Doyle 2015) are now joined by a CACNA1C aortic-expression Mendelian-randomisation preprint pointing the same direction. The question is not confined to Marfan: the MR analysis used general-population TAA/dissection and ascending-diameter genetic data. An active-comparator target-trial emulation in imaged TAA cohorts (CCB vs other antihypertensives, growth and events) would test it without a trial. (PMID 26506064; Horjus 2026, medRxiv doi:10.64898/2026.10.02.26364570, preprint) → medical therapy
OQ-24. Can ascending aortic length graduate from marker to intervention criterion? The 11-cm hinge and length-based risk are single-program observations; prospective multi-center validation against events is the gate to guideline entry. (PMID 31526537) → risk stratification

Epidemiology, statistics & data infrastructure¶

OQ-25. What is the true prevalence of silent TAA? Clinical detection (~0.16%) vs autopsy series (~0.76%) implies a large hidden reservoir; no modern population imaging-screening study has sized it — which also blocks rational screening policy. (PMID 33705940, 25609416) → epidemiology
OQ-26. Do the Yale hinge points and size-band rates replicate at population level? The numbers anchoring every guideline derive from one referral center with censoring at operation; population replication (now feasible in imaging biobanks) has never been done. (PMID 11834007, 16368358, 10391339) → epidemiology, risk stratification
OQ-27. Is type A dissection mortality improving across all hospital-admitted patients, or chiefly within selected centres and treatment groups? IRAD reports in-hospital mortality falling (31%→22% to 2013; 24.1%→16.7% to 2022), while the Danish chart-validated hospital-admission cohort reports 30-day mortality of 22.0% with no significant change from 1996 to 2016. These are different endpoints, populations and partly different eras; the observed trend difference remains substantively open, although the figures cannot be compared directly. Bossone's unchanged 3-year postdischarge survival (p=0.296) describes a further conditional endpoint, not a harmonised counterestimate. Denmark excludes prehospital deaths and cases without retrievable records, so its national coverage does not eliminate selection concerns. A common 30-day estimand including operated and non-operated admissions, with explicit accounting for transfers and missing records, could test how much of the trend difference reflects treatment rates, case mix, referral selection or risk-adjusted case fatality. (PMID 26205591, 36321467, 39145396) → dissection, statistics
OQ-28. How much do ICD-coded databases overstate dissection? Only 60.2% of coded dissections survived validation in Denmark; no US validation figure exists, yet NIS/WONDER studies keep publishing. (PMID 36321467) → statistics
OQ-29. What is the TAA-specific share of global aortic-aneurysm burden? The September 2026 GBD analysis still pools abdominal and thoracic aneurysms. Its 153,930 estimated deaths in 2021 must not be presented as TAA mortality; segment-specific ascertainment remains the missing denominator. (Jung 2026, PMID 42708483) → statistics
OQ-30. Is family screening cost-effective, and with what protocol? Screening first-degree relatives finds new disease in ~1/3, but predictive accuracy, interval, and cost-effectiveness are unmeasured — the evidence base is yield-only. Delivery is a second gap: in one US academic centre complete cascade screening of all at-risk first-degree relatives occurred in only 33% of 184 HTAAD families (78% in a combined cardiovascular-genetics clinic vs 10–12% in single-specialty clinics), and screened relatives had fewer unexpected aortic events — a retrospective association, not an effectiveness estimate. (PMID 30371227, 42862301) → epidemiology, genetics
OQ-31. What supports "1–2% per hour" untreated type A mortality? The canonical claim traces to 1958/1972 series whose internals predate modern abstracting; no contemporary estimate exists (and ethically cannot be generated directly) — deserves formal evidence-archaeology and modeling. (PMID 13577293, 4557973, 10807810) → dissection, statistics

Guidelines & care delivery¶

OQ-32. Are the three overlapping European guidelines (ESC 2024, EACTS/STS 2024, ESVS 2026) concordant with ACC/AHA 2022 — and where not, why? Documented historical conflicts (BAV thresholds needed a formal 2016 clarification; EACTS/STS defines aneurysm at >45 mm vs the 1.5× convention) make a systematic concordance audit valuable and cheap. (PMID 26637530, 38408364, 39210722, 41448425, 36322642) → guidelines, registry

2026-10-21 refinement of OQ-32: root-measurement convention itself needs comparison, alongside intervention thresholds. A CMR comparison documents category changes when switching conventions, but excludes BAV and has no outcome follow-up. Test both conventions against prospective events before preferring one on prognostic grounds (Zelis 2026, PMID 42890866). See imaging.

OQ-33. Which dissection classification should trials and registries speak? SVS/STS 2020 vs EACTS/STS 2024 TEM ("non-A non-B") bifurcates the literature's denominators; no harmonization document exists. (PMID 32001058, 38408364) → dissection
OQ-34. Do the updated sport and occupational recommendations improve aortic outcomes and quality of life? The 2025 AHA/ACC competitive-sports and 2026 ACC/AHA tactical-athlete statements replace the premise that US guidance stopped in 2015. Prospective validation of condition-specific exercise decisions remains needed. (Kim 2025, PMID 39976316; Dineen 2026, PMID 42691142) → guidelines, patient experience

Patient-centered research¶

OQ-35. What should a validated TAA/dissection-specific patient-reported outcome measure contain? None exists, so trials cannot register patient-relevant endpoints even if they wanted to; qualitative groundwork exists to build from. (PMID 37451607, 35701761, 39789589) → patient experience
OQ-36. Which surveillance patients benefit from psychological screening or intervention? A small population-based study found no significant HRQoL/anxiety/depression differences between mild-to-moderate ascending-dilation cases and controls (67 vs 50). Separate asymptomatic surveillance from post-dissection and syndromic trajectories; determine whether targeted screening improves patient outcomes. (Nilsson 2026, PMID 42768306) → patient experience

2026-10-21 refinement of OQ-36: distinguish early perceived post-traumatic growth from durable recovery. New qualitative interviews were all conducted before discharge; longitudinal follow-up and intervention studies are the next step, rather than treating these themes as a treatment effect (Zhang 2026, PMID 42879453).

OQ-37. Can a shared decision-making tool serve patients at threshold sizes? AAA has a randomized decision-aid trial; TAA — where the surgery-vs-surveillance dilemma is sharper (OQ-1) — has nothing. (PMID 24913683, 35501043) → patient experience, risk stratification
OQ-38. What exercise prescription is safe and beneficial after TAA diagnosis or repair? Postoperative rehabilitation is associated with better outcomes in retrospective matched data, but causality and applicability to unrepaired aneurysms remain unresolved. EXTRA is recruiting with no posted results, estimated primary completion October 2026; track its results rather than asserting that no exercise trial exists. (Yamazaki 2026, PMID 42283670; NCT04197648) → patient experience, trials
OQ-39. Do awareness campaigns (THINK AORTA and kin) measurably reduce the ~34% dissection misdiagnosis rate? Campaigns scale internationally on face validity; no before/after or regional-comparison evaluation exists. (PMID 34968970, 34838743) → patient experience, dissection
OQ-40. Does organized peer support improve surveillance retention and adherence? Patient organizations run large support programs; the effect on measurable care outcomes is unstudied. (PMID 42491301, 42285644) → patient experience

Closed¶

(None yet. When a question closes, move it here with the closing citation and date.)