Highlight
Spontaneous spinal cerebrospinal fluid (CSF) leaks, previously linked to classic connective tissue disorders, are now associated with rare deleterious variants in the FBN2 gene.
Whole-exome sequencing in 42 patients with type 1b spontaneous spinal CSF leaks identified a significant enrichment of rare functional FBN2 variants compared to large control cohorts.
Functional assays demonstrated reduced adhesion of mutated fibrillin-2 fragments to human dural fibroblasts, while mouse models with patient-equivalent FBN2 mutations showed increased susceptibility to dural rupture.
These findings support integrating FBN2 genetic testing into clinical practice and pave the way for targeted therapeutic strategies in spontaneous spinal CSF leaks.
Study Background
Spontaneous spinal cerebrospinal fluid (CSF) leaks represent a clinical condition characterized by leakage of CSF through dural defects without preceding trauma or surgical history. These leaks cause debilitating symptoms including orthostatic headaches, nausea, and neurological deficits due to intracranial hypotension. While spontaneous CSF leaks are recognized in the context of connective tissue diseases such as Marfan and Loeys-Dietz syndromes—conditions caused by mutations in extracellular matrix (ECM) genes—many patients lack a definitive connective tissue diagnosis despite subtle manifestations suggestive of underlying ECM abnormalities. This gap prompted an investigation into whether genetic variants in ECM-related proteins contribute to more common forms of spontaneous spinal CSF leaks, potentially revealing novel pathophysiological mechanisms and therapeutic targets.
Study Design and Methods
The study retrospectively identified 42 individuals diagnosed with lateral (type 1b) spontaneous spinal CSF leaks at Cedars-Sinai Hospital, Los Angeles, from 2006 to 2019. Participants were predominantly female (83%) and of White ethnicity (90%). Whole-exome sequencing (WES) was performed on these patients, with variant data compared against three large independent control cohorts comprising 2244 unrelated unaffected adults from the USA and two Belgian cohorts totaling 1627 individuals.
Bioinformatic analyses in silico pinpointed enriched rare functional variants, particularly focusing on their localization within the tertiary structure of implicated proteins. For the top candidate gene, FBN2, in vitro functional assays tested the effect of wild-type and mutant fibrillin-2 fragments on integrin-mediated adhesion to human dural fibroblasts. Moreover, the creation of three distinct mouse models harboring CRISPR-Cas9–engineered Fbn2 mutations equivalent to patient variants allowed in vivo assessment of dura mater integrity and CSF leak susceptibility utilizing intrathecal infusion testing. Additionally, comparisons were made to an established Marfan syndrome mouse model (Fbn1C1039G/+) to contrast phenotypic features.
Key Findings
Whole-exome sequencing revealed that 9 out of 42 patients (21%) with spontaneous spinal CSF leaks carried rare functional variants in FBN2. These variants were significantly enriched compared to the Mendel discovery cohort (8% prevalence; p=0.041; OR 3.18 [95% CI 1.50–6.76]), and two Belgian validation cohorts (7% and 5% prevalence; p=0.004 and p=0.0003 respectively), indicating a robust association.
Interestingly, the identified FBN2 variants clustered non-randomly within the transforming growth factor-beta (TGF-β) binding protein-like (TB) domains of fibrillin-2, suggesting domain-specific pathogenicity. Functional testing underscored this by showing that two of three mutant fibrillin-2 fragments exhibited diminished adhesion to human dural fibroblasts, potentially weakening extracellular matrix-cell interactions critical for dural integrity.
In vivo, mice harboring patient-equivalent mutations (Fbn2A1052T/+, Fbn2D1581V/+, Fbn2M2387T/+) demonstrated increased susceptibility to dural rupture following controlled induction of CSF leaks, affirming the pathogenic role of these variants. Contrastingly, Marfan (Fbn1C1039G/+) mutant mice displayed increased meningeal compliance rather than direct rupture, highlighting differing biomechanical consequences of distinct ECM gene mutations.
These results collectively link deleterious FBN2 variants to spontaneous spinal CSF leaks by disrupting fibrillin-2 functions integral to extracellular matrix stability and cell adhesion.
Expert Commentary
This study provides pioneering genetic and functional evidence for the involvement of rare fibrillin-2 gene variants in spontaneous spinal CSF leaks—a previously underappreciated etiology. As Dr. H. Dietz and colleagues elucidate, this discovery extends the spectrum of connective tissue disorders beyond classical Marfan and Loeys-Dietz conditions to include more subtle, genetically mediated phenotypes.
The localization of mutations within TGF-β binding-like domains aligns with known roles of fibrillin proteins in modulating TGF-β signaling and ECM assembly, mechanisms crucial for dural membrane integrity. The in vitro and in vivo findings convincingly demonstrate that these variants compromise the biomechanical resilience of the dura, providing a plausible pathophysiological basis for spontaneous CSF leaks.
However, limitations include the sample size and demographic homogeneity, primarily reflecting individuals of White ethnicity from a single center. Expanded multicenter studies with diverse populations are needed to validate findings and assess penetrance and expressivity. Furthermore, future research should explore potential genotype-phenotype correlations and the utility of FBN2 screening in clinical algorithms for patients presenting with spinal CSF leaks.
Conclusion
Rare deleterious variants in the FBN2 gene represent a significant genetic contributor to type 1b spontaneous spinal CSF leaks, supporting the inclusion of FBN2 genetic testing in the evaluation of patients without overt connective tissue disease diagnoses. By affecting extracellular matrix–cell adhesion dynamics, these mutations elucidate novel molecular mechanisms underpinning dural fragility. The generation of precise mouse models recapitulating patient mutations offers valuable platforms for testing pharmacologic therapies aimed at reinforcing dural integrity and preventing CSF leaks.
Ultimately, this study advances precision medicine in neurology and rheumatology by clarifying genetic underpinnings and suggesting therapeutic avenues for a challenging clinical entity. Integrating genetic insights with clinical practice may substantially improve diagnosis, management, and outcomes for affected patients.
Funding and ClinicalTrials.gov
This work was supported by the Howard Hughes Medical Institute, the Marfan Foundation, the Pease-Scheeler Fund, and the Biotechnology and Biological Sciences Research Council. No relevant clinical trial registration was reported.
References
1. Parks CA, Singh M, Wohler E, et al. Spontaneous spinal CSF leaks: a rare variant exome sequencing study and functional analysis. Lancet Neurol. 2026 Jul;25(7):664-672. doi:10.1016/S1474-4422(26)00123-9.
2. Dietz HC. Marfan syndrome. In: Adam MP, Ardinger HH, Pagon RA, et al., editors. GeneReviews® [Internet]. University of Washington, Seattle; 1993–2024.
3. Loeys BL, Dietz HC, Braverman AC, et al. The revised Ghent nosology for the Marfan syndrome. J Med Genet. 2010 Jul;47(7):476-85.
4. Schievink WI. Spontaneous spinal cerebrospinal fluid leaks. J Neurosurg Spine. 2006 May;4(5):439-44.
5. Ramirez F, Rifkin DB. Cell signaling events: a view from the matrix. Matrix Biol. 2003;22(6):501-9.
6. Sakai LY, Keene DR, Renard M, De Backer J. FBN2: the gene for type B skeletal dysplasia and its role in the extracellular matrix. Matrix Biol. 2016;56:20-28.
[Note: URLs for references omitted for brevity]
