Perspectives on the revised Ghent criteria for the diagnosis of Marfan syndrome

Authors, Journal, Affiliations, Type, DOI

• Yskert von Kodolitsch, Julie De Backer, Helke Schüler, Peter Bannas, Cyrus Behzadi, Alexander M Bernhardt, Mathias Hillebrand, Bettina Fuisting, Sara Sheikhzadeh, Meike Rybczynski, Tilo Kölbel, Klaus Püschel, Stefan Blankenberg, Peter N Robinson
• The Application of Clinical Genetics 2015;8:137–155 (Dove Press)
• Centre of Cardiology, University Hospital Eppendorf, Hamburg, Germany; Centre for Medical Genetics, University Hospital Ghent, Ghent, Belgium; Diagnostic and Interventional Radiology, University Hospital Eppendorf; Institute of Medical Genetics and Human Genetics, Charité Universitätsmedizin Berlin
• Review article
• DOI: Not provided in source document

Overview

This 2015 review critically examines the three MFS nosologies (Berlin 1988, Ghent-1 1996, Ghent-2 2010) and their real-world comparative diagnostic performance. Ghent-1 and Ghent-2 yield nearly identical proportions of MFS diagnoses in patients with causative FBN1 mutations (90% vs 92%) and without FBN1 mutations (15% vs 13%), but Ghent-2 is significantly easier to use, more objective, and requires fewer investigations. A key conceptual finding is that no external diagnostic reference standard exists for MFS, meaning sensitivity, specificity, and accuracy of any nosology cannot formally be measured. The review adds important clinical context: actual population prevalence is 1.5–17.2 per 100,000 (far lower than the widely cited 1:5,000); diagnostic delays of 2 or more years are common and causally harmful; and the Lacro 2014 NEJM trial established atenolol and losartan as equally effective, with the opportunity to choose either.

Keywords

Marfan syndrome, Ghent nosology, diagnosis, FBN1, mutation, aorta, atenolol, losartan, screening

Key Takeaways

Prevalence — What Studies Actually Show

• Widely cited estimate of 1:5,000 (Milewicz 2005) is a commonly used but unsupported figure
• Published prevalence studies: 1.5–17.2 per 100,000, or 0.075–0.86 per 5,000 individuals; all far below 1:2,000 (European rare disease definition threshold); MFS is unambiguously a rare disease
• Studies: Northern Ireland 1.5/100,000 (Lynas); China schoolchildren 17.2/100,000 (Sun); Scotland 6.8/100,000 (Gray); Denmark 4.6/100,000 (Fuchs); Hamburg 7/100,000 adults (Rybczynski); Taiwan 10.2/100,000 (Chiu 2014)
• Wide range likely reflects different diagnostic criteria used AND under-diagnosis in the general population; MFS remains substantially under-diagnosed

Diagnostic Delay — A Major Clinical Problem

• EURORDIS survey of 682 MFS families from 18 European countries: median 2 years from first symptoms to diagnosis; 25% wait ≥4.5 years
• 38% of families consulted >5 physicians before diagnosis; 7% consulted >20 physicians
• 25% received an incorrect diagnosis before MFS; delay led to inappropriate treatment in 81%
• 72% of families reported the delay in diagnosis was responsible for deleterious consequences
• Roll (2012): higher physician density paradoxically associated with delayed diagnosis — structural deficiencies (organised irresponsibility, lack of long-term relationships, no networking) are key barriers, not lack of knowledge

Diagnostic Triggers — Which Features Actually Lead to Confirmation

• Series of 329 adults referred for suspected MFS: most common reasons for referral — family history (24%), aortic aneurysm/dissection (24%), skeletal features (22%), childhood MFS diagnosis (10%), eye manifestations (9%)
• Highest MFS confirmation rates: childhood MFS diagnosis (88%), eye manifestations (83%), family history of Marfan features (52%), aortic aneurysm/dissection (50%)
• Lowest confirmation rates: pneumothorax (43%), skeletal features (30%), tissue weakness/hypermobility (14%), nonaortic cardiovascular findings (0%)
• Key message: tall stature, scoliosis, and joint hypermobility are common reasons for referral but rarely confirm MFS

"Seven-Signs" Adult Screening Tool (Sheikhzadeh et al 2012)

• Simple pre-test probability tool for MFS in adults in the general population
• Points assigned: ectopia lentis = 4 pts; family history of MFS = 2 pts; 1 pt each for: previous thoracic aortic surgery, pectus excavatum, wrist and thumb sign, previous pneumothorax, skin striae
• Risk categories based on total score: Low (≤1 pt), Moderate (>1–3.5 pts), High (>3.5 pts)
• Practical screening tool for non-specialist settings prior to full diagnostic workup

Historical Nosology Comparison — Berlin → Ghent-1 → Ghent-2

• Berlin nosology (Beighton 1988): first international nosology; purely clinical criteria; later shown to overdiagnose
• Ghent-1 (De Paepe 1996): more stringent; first to incorporate genetic data; confirmed MFS in only 32–53% of patients previously diagnosed by Berlin criteria
• Ghent-2 (Loeys 2010): five major changes; simpler; de-emphasises weak criteria; gives greater weight to aortic root Z-score, ectopia lentis, and FBN1 status; explicitly addresses differential diagnosis

Ghent-1 vs Ghent-2 Comparative Diagnostic Performance

• Faivre et al (1,009 patients with FBN1 mutation): Ghent-1 MFS 89% vs Ghent-2 MFS 83% — marginal difference
• Sheikhzadeh et al (300 adults with suspected MFS): identical diagnosis in 245/300 (82%); FBN1+ MFS 90% vs 89%; FBN1- MFS 13% vs 12%
• Aalberts et al (343 adults): FBN1+ MFS 76% vs 80%; FBN1- MFS 11% vs 9%; Ghent-2 led to more non-MFS diagnoses (due to lower MVPS threshold)
• Yang et al (106 adults): FBN1+ MFS 100% both; FBN1- MFS 38% vs 31%
• Summary across studies (295 FBN1+ with suspected MFS): Ghent-1 90% vs Ghent-2 92% MFS diagnosis; FBN1- (287): 15% vs 13%
• Overall diagnostic yield (749 patients, 3 studies): MFS 34% vs 35%; other syndromes 26% vs 28%; no definitive diagnosis 40% vs 38% — essentially identical

Quality of Diagnostic Criteria — A Fundamental Limitation

• Quality criteria for diagnostic methods: objectivity → reliability → validity (each is prerequisite for the next)
• Objectivity: Ghent-2 likely higher than Ghent-1 — simpler rules, fewer subjective qualifiers, explicit FBN1 mutation causality criteria; but MVP and dural ectasia criteria still unstandardised in Ghent-2
• Reliability: No study comparing intrasubject, intraobserver, or interobserver variation for either nosology; similar overall diagnostic yields suggest similar reliability but this is not confirmed
• Validity (critical limitation): No external diagnostic reference standard for MFS exists — no single feature proves or excludes MFS with LR >10 or LR <0.1; therefore sensitivity, specificity, and accuracy cannot be formally measured for any Ghent nosology
• Conclusion: nosology-based MFS diagnosis lacks a reference standard; formal validity assessment is impossible

FBN1 Mutation Analysis — Role and Limitations

• 1,500 different disease-causative FBN1 mutations now in databases (updated from >1,000 in 2010)

• No single FBN1 genotype feature qualifies as a reliable predictor of disease evolution or severity
• An increasing number of FBN1 nucleotide variants have uncertain clinical significance (VUS) — a growing source of diagnostic uncertainty
• Small proportion of FBN1 mutations remain undetected due to technical limitations of screening
• FBN1 mutations can present with phenotypes outside classic MFS: purely skeletal phenotype in children, Shprintzen-Goldberg syndrome, Weill-Marchesani syndrome
• FBN1 analysis significantly reduces diagnostic uncertainty in all Ghent-2 scenarios (Table 8): with FBN1+ mutation, all routes to MFS diagnosis become straightforward; with FBN1 absent, differential diagnoses become more probable
• Practical message: FBN1 analysis should be performed whenever possible, especially when the MFS phenotype is incomplete

Echocardiographic Z-Score Nomograms — Calibration Issue

• Radonic et al (Ghent-2 validation study) found that Z-scores calculated using Roman 1989 normograms underestimate aortic root dilatation in patients with large body surface area — a relevant issue since MFS patients tend to be tall with large BSA
• This problem was subsequently addressed by publication of improved normograms (Devereux and others)
• Practical implication: clinicians should verify which normograms are used and whether they are validated for tall patients

Medical Management — Atenolol vs Losartan (Lacro 2014, NEJM)

• Randomised double-blind trial (Pediatric Heart Network; Lacro RV et al, NEJM 2014): atenolol versus losartan in children and young adults with MFS
• Results: equally beneficial effects with atenolol and losartan; no significant differences in aortic root growth rate or side effects
• Practical implication: either drug may be used — choice is based on individual patient factors and tolerability
• This result resolves the prior question of "losartan superiority" suggested by mouse model data and earlier small trials; no clinically significant advantage of losartan over atenolol

Utility of Diagnostic Criteria — A Framework

• Medical utility: diagnosis should be congruent with history, inform about etiology, pathogenesis, diagnostic triggers, prognostic triggers, and potential complications
• Social utility: acceptance by patients, organisations, clinicians; practicability and cost (Ghent-1 outpatient workup ~€389/year in Germany; full FBN1 sequencing >€4,000 — costs may not be outweighed by reduction in clinical workup)
• Psychological utility: reduction of diagnostic uncertainty and anxiety; identification with a disease community
• Utility is graded and context-dependent (unlike validity which is invariant); diagnostic rules should be applied "with common sense and flexibility" (Faivre et al)
• Fair and transparent nosology development process recommended (akin to guideline development rules)

Prognostic Features and Complications

• Aortic root dilatation is the primary prognostic trigger in MFS (risk rises with diameter)
• Sleep-disordered breathing: prognostic trigger for aortic dissection/rupture
• Myocardial dysfunction (elevated NT-proBNP): prognostic trigger for HF and SCD
• Main pulmonary artery rupture/dissection: rare complication
• Dural ectasia complications: lumbar/sacral radiculopathy, atlantoaxial/lumbar subluxation, intrapelvic meningocele, dural leak with postural headache
• Osteopenia and skeletal malformations: fractures and cardiorespiratory compromise
• Still 7.8% of MFS patients die from aortic dissection outside hospital (autopsy study, Prakash 2011)

Limitations of the document

• Review article based on literature synthesis; no new original data
• Comparison of Ghent-1 vs Ghent-2 limited to 5 studies; most are relatively small cohorts; none prospective RCTs of the nosologies themselves
• No standardised formal assessment of intraobserver/interobserver reliability for either nosology
• FBN1 testing costs are country-specific and have fallen significantly since 2015 — €4,000 estimate may be outdated
• The fundamental limitation (no external reference standard) is acknowledged but cannot be resolved within the current paradigm

Key Concepts Mentioned

• concepts/Marfan-Syndrome — primary topic; nosology comparison, prevalence, diagnostic delay, screening tool, Lacro 2014 result
• concepts/Ghent-Nosology — comparative analysis of Ghent-1 vs Ghent-2; quality framework

Key Entities Mentioned

• entities/FBN1 — >1,500 mutations; VUS burden; role in diagnostic certainty
• entities/Loeys-Dietz-Syndrome — key alternative diagnosis requiring TGFBR1/2 testing

Wiki Pages Updated

• wiki/sources/marfan-ghent-aclingene-2015.md — created
• wiki/concepts/Marfan-Syndrome.md — added: precise prevalence data, diagnostic delay, seven-signs tool, Lacro 2014 atenolol=losartan result, Z-score BSA issue, no external reference standard; source_count 4→5
• wiki/entities/FBN1.md — updated mutation count to >1,500; added VUS issue; source_count 3→4
• wiki/sourceindex.md — added marfan-ghent-aclingene-2015 entry
• wiki/wikiindex.md — updated Marfan-Syndrome entry
• log.md — updated