High-yield points
- Pediatric traumatic leaks:
- Frequency: in about 28% of skull base fractures.
- Course: about two-thirds stop within 3 days, and only about 2% needed dural repair after drainage failed (3.6% had any surgery) .
- Prophylactic antibiotics: not supported (Cochrane, 5 RCTs) .
- Confirm with beta-2 transferrin or beta-trace protein, both about 95% sensitive .
- Localize:
- First: high-resolution CT plus heavily T2-weighted MR cisternography.
- Next: CT cisternography for active leaks.
- Least useful: radionuclide pledgets, which detect but poorly localize .
- Intrathecal fluorescein: about 89% localization; keep the dose below 50 mg (10–30 mg) .
- Repair:
- Endoscopic: about 94% primary success .
- Lumbar drain: cuts leaks after high-flow skull base repairs (8% vs. 21%) .
- Recurrent meningitis plus unilateral "rhinorrhea" in a child: think inner ear malformation (stapes footplate fistula) .
A CSF leak is a fistula between the subarachnoid space and the outside world:
- Rhinorrhea: through the anterior or central skull base.
- Otorrhea or otorhinorrhea: through the temporal bone.
- Spinal leaks: cause intracranial hypotension.
The danger is meningitis. The job is to:
- confirm that the fluid is CSF;
- localize the defect;
- close it, usually endoscopically, or let it heal with diversion.
Most traumatic leaks in children stop on their own. Spontaneous pediatric leaks usually have a structural cause, such as an encephalocele or inner ear malformation .
Etiology
- Traumatic (the majority in children):
- Basilar skull fracture review (21 studies) :
- Mechanisms: falls from height and traffic accidents.
- Complications: CSF leaks were reported in 10 of 21 studies, along with cranial nerve palsies (mostly facial) and meningitis.
- Largest pediatric series (196 children with skull base fractures) :
- Fracture sites: middle skull base in 112, frontal base in 62 and occipital base in 13.
- CSF leak: 28%.
- Temporal bone fractures (147 children): 37 had a CSF fistula .
- Iatrogenic: after endoscopic endonasal, skull base or ear surgery (see caveats).
- Spontaneous or congenital:
- Inner ear malformations: a fistula through the stapes footplate. In 13 children :
- Presentation: rhinorrhea (otorhinorrhea down the Eustachian tube) in 92%.
- Infection: meningitis in 69%.
- Congenital skull base encephaloceles: see the encephaloceles topic.
- Spinal CSF leaks (dural tears, CSF–venous fistulas) cause intracranial hypotension .
- Adult-type spontaneous leaks with raised ICP in obese adolescents (see caveats).
Signs and symptoms
- Rhinorrhea:
- Clear, often unilateral, worse leaning forward.
- Salty taste.
- "Halo" or "double-ring" sign on linen with blood-tinged fluid (see caveats).
- Otorrhea through a perforated eardrum, or hemotympanum with fluid behind an intact drum that drains down the Eustachian tube.
- Recurrent or unexplained meningitis in a child. This should prompt a search for a fistula .
- Pneumocephalus on CT.
- Delayed cranial nerve deficits after skull base fracture :
- In the 196-child series: delayed hearing loss in 12 and delayed facial palsy in 3 on long-term follow-up.
- Spinal leaks: orthostatic headache and imaging signs of intracranial hypotension .
Diagnosis
- Confirm the fluid is CSF:
- Beta-2 transferrin or beta-trace protein in the fluid.
- Meta-analysis (1,914 patients) :
- beta-2 transferrin: sensitivity 0.96, specificity 0.95;
- beta-trace protein: sensitivity 0.95, specificity 0.999.
- Glucose testing is unreliable (see caveats).
- Localize the defect:
- Accuracy by modality (meta-analysis, 16 studies, 472 patients) :
- high-resolution CT: sensitivity 0.74;
- CT cisternography: 0.89;
- MRI: 0.84;
- heavily T2-weighted MR cisternography: 0.86, with no intrathecal injection;
- intrathecal gadolinium MR cisternography: 0.93, the most accurate (off-label);
- radionuclide cisternography: 0.71.
- Second systematic review (1,622 patients), median sensitivity/specificity :
- high-resolution CT: 0.93/0.50;
- MR cisternography: 0.94/0.77;
- CT cisternography: 0.95/1.00;
- radionuclide cisternography: 0.90/0.50;
- contrast-enhanced MR cisternography: 0.99/1.00;
- intrathecal fluorescein: 0.96/1.00.
- Its conclusion: MR cisternography beats high-resolution CT. CT cisternography, contrast MR cisternography and radionuclide cisternography are accurate but invasive.
- Practical sequence (see caveats):
- First: high-resolution CT (bone defect) plus heavily T2-weighted MR cisternography (CSF tract, encephalocele).
- If the leak is intermittent or the site unclear: CT cisternography, which needs an active leak. Radionuclide cisternography with nasal pledgets shows that a leak exists but localizes it poorly.
- Intrathecal fluorescein at surgery :
- Localization: 88.7% of 335 procedures.
- Complications: dose-related: 3.9% at doses below 50 mg vs. 80% at ≥50 mg.
- Recommended dose: 10–30 mg, diluted. The use is off-label.
- Spinal leaks: brain and spine MRI first, then dynamic myelography or digital subtraction myelography for localization .
Management strategy
- Traumatic leaks: conservative first.
- The 196-child series :
- Spontaneous resolution: 63% of leaks stopped within 3 days.
- Interventions:
- 3 needed lumbar CSF diversion;
- only 4 (2%) needed dural repair after failed continuous spinal drainage and acetazolamide;
- 3 had surgery on admission for an open wound.
- Temporal bone fistulas were nonsurgical in most cases; only 2 of 37 needed mastoid exploration or obliteration .
- Conservative measures:
- head elevation;
- avoid nose blowing and straining;
- acetazolamide;
- then lumbar drainage if the leak persists for several days.
- When to operate (see caveats):
- leaks persisting beyond 7–10 days despite diversion;
- large defects;
- brain herniation;
- meningitis;
- penetrating or open injuries.
- Antibiotic prophylaxis: not supported.
- Cochrane review (5 RCTs, 208 participants): no reduction in meningitis, with or without a leak .
- Pediatric series: meningitis occurred in only 2 of 196, supporting no routine antibiotics .
- Pneumococcal vaccination status should be updated (see caveats).
- Surgical repair:
- Endoscopic endonasal repair is the standard for anterior and central skull base leaks.
- Spontaneous CSF rhinorrhea (meta-analysis, 47 studies, 2,586 patients) :
- Success: 93.7% at the primary repair and 99.7% after revision.
- Technique: multilayer repair 94.6% vs. single-layer 87.6%, not significant.
- Location: sphenoid defects did worse.
- Technique (see caveats):
- Defects: small defects get an onlay or underlay graft; larger ones get multilayer inlay plus onlay with a vascularized nasoseptal flap.
- Young children: a small nasal cavity and an undeveloped sphenoid limit flap size.
- Temporal bone leaks:
- Inner ear malformations: transcanal endoscopic repair of the stapes footplate fistula succeeded in 13 of 14 .
- Tegmen defects: middle fossa or transmastoid repair (see caveats).
- CSF diversion:
- Lumbar drain after high-flow endoscopic skull base surgery (RCT, 170 adults) :
- Postoperative leak: 8.2% with vs. 21.2% without a drain (OR 3.0).
- Spontaneous leak repair: perioperative lumbar drains did not change success in the meta-analysis .
- Underlying hydrocephalus or raised ICP: treat it, with a shunt or ETV as appropriate, or repairs fail (see caveats).
Related topics
- Skull Fractures in Children: the usual cause of traumatic leaks.
- Encephaloceles: congenital skull base defects.
- Endoscopic Third Ventriculostomy (ETV): treating underlying hydrocephalus.
References
- Jung G, Xavier J, Reisert H, et al. Clinical features and management of skull base fractures in the pediatric population: a systematic review. Children (Basel) 2024;11(5):564. doi:10.3390/children11050564. PMID 38790559. verified
- Leibu S, Rosenthal G, Shoshan Y, et al. Clinical significance of long-term follow-up of children with posttraumatic skull base fracture. World Neurosurg 2017;103:315–321. doi:10.1016/j.wneu.2017.04.068. PMID 28433849. verified
- Deng W, Liu J, Pang F, et al. Diagnosis and management of pediatric cerebrospinal fluid leakage secondary to inner ear malformations: a report of 13 cases. Int J Pediatr Otorhinolaryngol 2020;135:110049. doi:10.1016/j.ijporl.2020.110049. PMID 32497907. verified
- McGuirt WF Jr, Stool SE. Cerebrospinal fluid fistula: the identification and management in pediatric temporal bone fractures. Laryngoscope 1995;105(4 Pt 1):359–364. doi:10.1288/00005537-199504000-00003. PMID 7715377. verified
- Mamlouk MD. Detecting the drip: pediatric neuroradiologist guide to spinal CSF leaks. Pediatr Radiol 2026;56(12):2613–2625. doi:10.1007/s00247-026-06764-2. PMID 42684458. verified
- Bhat SN, Thunga G, Kamath A, et al. CSF biomarkers in leak detection: a systematic review & meta-analysis of diagnostic test accuracy (DTA) studies. Clin Chim Acta 2025;577:120458. doi:10.1016/j.cca.2025.120458. PMID 40602704. verified
- Kim DH, Kim SW, Kim SH, et al. Usefulness of imaging studies for diagnosing and localizing cerebrospinal fluid rhinorrhea: a systematic review and meta-analysis. Int Forum Allergy Rhinol 2022;12(6):828–837. doi:10.1002/alr.22932. PMID 34889068. verified
- Xie M, Zhou K, Kachra S, et al. Diagnosis and localization of cerebrospinal fluid rhinorrhea: a systematic review. Am J Rhinol Allergy 2022;36(3):397–406. doi:10.1177/19458924211060918. PMID 34846218. verified
- Jolly K, Gupta KK, Muzaffar J, et al. The efficacy and safety of intrathecal fluorescein in endoscopic cerebrospinal fluid leak repair: a systematic review. Auris Nasus Larynx 2022;49(6):912–920. doi:10.1016/j.anl.2022.03.014. PMID 35393153. verified
- Ratilal BO, Costa J, Pappamikail L, et al. Antibiotic prophylaxis for preventing meningitis in patients with basilar skull fractures. Cochrane Database Syst Rev 2015;2015(4):CD004884. doi:10.1002/14651858.CD004884.pub4. PMID 25918919. verified
- Mughal Z, Osuji J, Jonah S, et al. Outcomes of endoscopic management of spontaneous cerebrospinal fluid rhinorrhea: a meta-analysis. Laryngoscope 2026;136(1):36–49. doi:10.1002/lary.32428. PMID 40650638. verified
- Zwagerman NT, Wang EW, Shin SS, et al. Does lumbar drainage reduce postoperative cerebrospinal fluid leak after endoscopic endonasal skull base surgery? A prospective, randomized controlled trial. J Neurosurg 2019;131(4):1172–1178. doi:10.3171/2018.4.JNS172447. PMID 30485224. verified
Verification (2026-11-17):
- Each reference's authors, title, journal, year, volume, pages, DOI and PMID were checked on PubMed. Numbers were checked against the PubMed abstracts.
- Reference 1's article number is taken from its DOI.
Caveats:
- The following are standard teaching, not drawn from the cited abstracts:
- iatrogenic causes;
- spontaneous leaks with raised ICP in obese adolescents;
- the halo sign;
- unreliability of glucose testing;
- the practical imaging sequence;
- conservative measures and timing of surgery;
- pneumococcal vaccination;
- repair techniques and pediatric nasoseptal flap limits;
- tegmen repair approaches;
- orthostatic headache and the specific spinal leak localization techniques (dynamic or digital subtraction myelography), which reference 5 covers only as "advanced imaging techniques" in its abstract;
- treating hydrocephalus or raised ICP before or with repair.
- The diagnostic accuracy studies (6–8), the spontaneous leak meta-analysis (11), the Cochrane review (10) and the lumbar drain RCT (12) are mainly or entirely adult.
- Intrathecal fluorescein and intrathecal gadolinium are off-label.
- The figure is an original schematic.