Marfan syndrome. Part 1: pathophysiology and diagnosis

Authors, Journal, Affiliations, Type, DOI

• Victoria Cañadas, Isidre Vilacosta, Isidoro Bruna, Valentin Fuster
• Nature Reviews Cardiology 2010; 7: 256–265
• Instituto Cardiovascular, Hospital Clínico San Carlos, Madrid, Spain (Cañadas, Vilacosta); Hospital Universitario Madrid Montepríncipe, Madrid (Bruna); Zena and Michael A. Wiener Cardiovascular Institute, Mount Sinai Hospital, New York (Fuster)
• Review article (Part 1 of 2; Part 2 covers treatment)
• DOI: https://doi.org/10.1038/nrcardio.2010.30

Overview

This 2010 Nature Reviews Cardiology review (Part 1) covers the pathophysiology, genetics, clinical manifestations, and diagnostic approach to Marfan syndrome. Published five years after the landmark Judge/Dietz 2005 Lancet Seminar, it incorporates new data: the mutation count in FBN1 has exceeded 1,000; the exon 24–32 cluster is linked to more severe/neonatal disease; and the Habashi 2006 Science paper demonstrates that losartan (AT1 blocker) prevents aortic aneurysm in fibrillin-1-deficient mice. Detailed echocardiographic measurement standards and quantitative dural ectasia criteria are provided. Loeys-Dietz syndrome is now subclassified into Type I (craniofacial features) and Type II (bifid uvula only). The paper still references Ghent 1996 criteria — published just before Ghent 2010 revision.

Keywords

Marfan syndrome, FBN1, fibrillin-1, TGF-beta, losartan, echocardiography, dural ectasia, Loeys-Dietz syndrome, Ghent criteria, connective tissue disorder

Key Takeaways

Epidemiology

• Prevalence 1:5,000–1:10,000 live newborns; equal sex distribution; worldwide
• ~75% familial; ~25% de novo mutations
• High penetrance (nearly all carriers develop the disease); marked phenotypic heterogeneity both between and within families

Genetics

• FBN1 on chromosome 15q21.1; >1,000 mutations described to date (updated from ~500 in 2005)
• Majority missense mutations altering a single amino acid of the 2,871 aa protein; mainly in EGF-like domains affecting cysteine residues or Ca²⁺-binding residues
• Exon 24–32 cluster: mutations in this region tend to predict more severe phenotype; most neonatal Marfan syndrome mutations (most severe form) located here
• Also: premature truncation codon mutations (severe skeletal/skin manifestations); exon skipping mutations
• No robust genotype–phenotype correlation overall, but the exon 24–32 exception is notable
• TGFBR2 controversy (Boileau 1991/1993): a large French family with Marfan-like syndrome mapped to chromosome 3p24.2-p25 = TGFBR2 gene ("Marfan syndrome type II"); controversy whether these patients truly met MFS criteria; aortic dissection at smaller dimensions suggests LDS phenotype, not classic MFS

Physiopathology — Historical and Current Models

• Early (structural) model: FBN1 mutations cause structural weakness via (a) dominant-negative effect on microfibrils or (b) haploinsufficiency; explains ectopia lentis, dural ectasia, joint laxity
• Limitation of structural model: cannot explain long-bone overgrowth, myxoid mitral valve changes — requires regulatory explanation
• TGF-β dysregulation model (current): fibrillin-1 microfibrils sequester TGF-β in ECM as large latent complexes (TGF-β + LAP + LTBP); mechanical/pH/cytokine stimuli → protease-mediated TGF-β release → receptor activation → downstream signalling; fibrillin-1 deficiency → failed ECM sequestration → excessive TGF-β activation
• Histological correlates: cystic medial necrosis (Erdheim 1920s) = fragmentation of elastic lamina + basophilic glucosaminoglycan material; nonspecific (seen in all thoracic aortic aneurysms); apoptosis of vascular smooth muscle cells (Nagashima 2001 — via angiotensin II type 2 receptor) contributes to remodelling
• MMP upregulation: excessive TGF-β → MMP-2 and MMP-9 overexpression → exaggerated elastolysis + increased hyaluronan content
• Mouse model evidence (key advances):
  • Neptune 2003 (Nat Genet): pulmonary alveolar septation failure rescued by TGF-β neutralising antibody
  • Ng 2004 (J Clin Invest): mitral valve prolapse prevented by TGF-β neutralising antibody
  • Habashi 2006 (Science): losartan (AT1 antagonist) prevents aortic aneurysm in fibrillin-1-deficient mice — reduced elastic fibre fragmentation + slower aortic root growth vs placebo; this is the preclinical basis for the COMPARE trial

Cardiovascular Manifestations

• Aortic root dilation at sinuses of Valsalva in 60–80% of adults; may begin in utero in severe cases
• Higher elastic fibre content + cyclic wall stress during ventricular ejection explains sinus of Valsalva predilection (vs other aortic aneurysm types)
• Extension of dilation beyond sinuses of Valsalva may predict worse outcomes (progressive dilation, dissection, severe AR) — not yet validated
• Pulmonary artery root dilation correlates strongly with degree of aortic dilation
• MVP in 50–80% (vs ~2% general population); bileaflet or anterior leaflet prolapse more common in MFS than isolated MVP; leaflets thicker but longer/thinner than in myxomatous disease
• AR is a late manifestation — secondary to aortic root expansion; leaflets usually normal
• Dilated cardiomyopathy/ventricular dysfunction: controversial — confirmed in some studies but not others; does not affect all patients

Echocardiographic Assessment (Detailed Methodology)

• Long-axis parasternal projection for aortic root
• Measurement technique: 2D or M-mode at end-diastole, leading-edge technique (ASE recommendations)
• Four measurement levels: (1) aortic annulus (hinge points), (2) sinuses of Valsalva, (3) sinotubular junction, (4) tubular ascending aorta
• Measurements must be strictly perpendicular to long axis — oblique views overestimate
• Compare to age- and BSA-indexed nomograms (Roman 1989); z-score >2 = dilation
• Caveat: nomograms poorly validated for individuals >95th percentile height (majority of MFS patients); relationship between anthropometrics and aortic dimensions is non-linear at extremes of height — assuming linearity overestimates upper normal limit
• Pediatric screening ratio: sinus of Valsalva:aortic annulus ratio ≥1.45 predicts aortic dilation with sensitivity 0.82 and specificity 1.00 in children; loses utility if aortic annulus is also dilated; can produce false positives with small aortic annulus
• CT or MRI if poor TTE window or for overall aortic assessment

Ophthalmological Assessment

• Slit-lamp examination with pupil dilation mandatory
• Ectopia lentis in 60%; usually superior quadrant displacement; may affect one or both eyes; corrective surgery not usually required
• Keratometry for corneal curvature; ocular biometry for anteroposterior globe diameter

Dural Ectasia Assessment (Quantitative Criteria)

• CT or MRI of lumbosacral column required for definitive assessment
• Qualitative: dural sac diameter at S1 > L4 = dural sac not tapering → highly suggestive
• Quantitative (Oosterhof criteria): dural sac ratio (dural sac diameter / anteroposterior vertebral body diameter) measured L1–S1
  • L3 ratio >0.47 OR S1 ratio >0.57 → diagnoses dural ectasia with 95% sensitivity, 98% specificity
  • Validated in pediatric/adolescent populations
  • Most useful criterion: S1 > L4 diameter AND abnormal dural sac ratio at L5 and S1
• Severe cases may show vertebral body erosion on plain X-ray; herniation of nerve roots; meningoceles

Diagnostic Criteria and Clinical Assessment

• Multidisciplinary assessment required: family history, physical examination, TTE, ophthalmological assessment
• X-ray spine + pelvis for scoliosis/protrusio acetabuli if insufficient criteria at initial assessment
• CT/MRI of lumbar spine for dural ectasia
• Children who do not fulfill criteria but have family history or Marfan-like features: reassess at ages 5, 10, and 15 years
• FBN1 mutations detectable in 90–95% of patients meeting Ghent 1996 criteria
• Best uses of genetic testing: (1) insufficient clinical criteria at initial exam; (2) presymptomatic diagnosis in families with known mutation; (3) atypical phenotype to rule out LDS or other CT disorders

Differential Diagnosis

• MASS phenotype: myopia + MVP + mild non-progressive aortic dilation + skin/skeletal; requires ≥2 systems; FBN1 mutation can be present; fibrillinopathy
• Loeys-Dietz syndrome (LDS): TGFBR1/TGFBR2 mutations; triad of arterial tortuosity + aneurysms, hypertelorism, bifid uvula/cleft palate
  • LDS Type I: craniofacial involvement (cleft palate, hypertelorism, craniosynostosis) — more severe arterial disease
  • LDS Type II: absent craniofacial involvement; isolated bifid uvula — relatively less severe but still aggressive
  • Arteriopathy not confined to ascending aorta (unlike MFS): abdominal aorta, pelvic vessels, intracranial vessels; dissection at earlier age and smaller diameter → surgery at smaller dimensions than MFS; strict follow-up essential
• Familial non-syndromic thoracic aortic aneurysm/dissection (TAAD): 11–19% of surgical aortic aneurysm patients have direct relatives with aneurysms; autosomal dominant, reduced penetrance; 7 genetic loci; 4 identified genes:
  • TGFBR2 (TAAD2)
  • TGFBR1 (TAAD5)
  • ACTA2 (TAAD4) — smooth muscle α2-actin; possibly the most common cause of familial TAAD
  • MYH11 (TAAD-patent ductus arteriosus) — smooth muscle βMHC
  • Family screening recommended if aneurysm patient has direct relatives, especially if young and no hypertension history

Limitations of the Document

• Predates revised Ghent 2010 nosology (published later in 2010 by Loeys et al.) — Ghent 1996 criteria used throughout
• Part 1 only — treatment/management covered in Part 2 (Cañadas et al. Nat Rev Cardiol 2010)
• Losartan mouse data (Habashi 2006) cited as promising but human clinical trial data not yet available (COMPARE not published until 2013)
• No Z-score anchoring for diagnosis — Ghent 2010 revision added this
• Dilated cardiomyopathy association described as controversial (some studies positive, some negative) — not definitively resolved

Key Concepts Mentioned

• concepts/Marfan-Syndrome — primary subject; updated with new data from this review
• entities/FBN1 — mutation count updated; exon 24–32 correlation added
• entities/Loeys-Dietz-Syndrome — subclassification into Type I vs II added

Key Entities Mentioned

• entities/FBN1 — updated mutation data; exon 24-32 cluster
• entities/Loeys-Dietz-Syndrome — LDS Type I vs Type II subclassification

Wiki Pages Updated

• wiki/sources/marfan-naturerc-2010.md — created (this file)
• wiki/concepts/Marfan-Syndrome.md — updated
• wiki/entities/FBN1.md — updated
• wiki/entities/Loeys-Dietz-Syndrome.md — updated
• wiki/sourceindex.md — updated
• wiki/wikiindex.md — updated (source_count on existing entries)