Cavernous Sinus Thrombosis in Children
Imaging Characteristics and Clinical Outcomes
Background and Purpose—Cavernous sinus thrombosis (CST) is a rare life-threatening cerebrovascular disease known to cause carotid artery narrowing (CAN) and arterial ischemic stroke. The imaging features of CST and related complications have been reported in adults, but rarely in children.
Methods—We performed a retrospective review of children with imaging confirmed CST from 2003 to 2014, describing presenting symptoms, imaging findings, and treatment.
Results—Ten patients with CST were identified. All had CAN and 6 of 10 developed infarcts. Of 8 patients treated with anticoagulation therapy, 3 developed new infarcts. None required discontinuation of anticoagulation therapy because of bleeding. Visual impairment secondary to infectious neuritis was common. Imaging characteristics include cavernous sinus expansion, filling defects, restricted diffusion, arterial wall enhancement, empyema, superior ophthalmic vein enlargement and thrombosis, orbital cellulitis, and pituitary inflammation. CAN resolved in 60% of cases. Outcomes were mostly good, with a modified Rankin Scale score of ≤1 for 7 of 10 patients at discharge and 1 death.
Conclusions—CAN and infarcts were common in this modest cohort of children with CST. Despite the high incidence of CAN and infarction, outcomes were often favorable. Although this is the largest cohort of childhood CST reported to date, large multicenter cohorts are needed to confirm our findings and determine the preferred therapeutic strategies for childhood CST.
Cavernous sinus thrombosis (CST) is a rare and life-threatening complication of infection involving the paranasal sinuses, facial, and oral cavities. The incidence has likely decreased significantly because of the use of antibiotics.1–3 CST has been reported to cause CAN and vasospasm,4,5 embolic infarcts, and large territory infarction from hypoperfusion.4,6 The standard acute treatment for childhood cerebral venous sinus thrombosis is typically anticoagulation therapy (ACT)7 although the optimal treatment for children with the subset of CST remains largely uncertain. The prevalence of neurological injury, risks of ACT, and range of outcomes is not known in childhood CST.
The purpose of our study is to describe the natural history of childhood CST in a retrospective cohort with CST seen at a single pediatric tertiary-care center >10 years. We report the clinical presentation, risk factor/s, imaging findings, treatment, incidence of infarct, complication/s of treatment, and outcome.
Standard Protocol Approvals
This study was approved by the local institutional review board.
We performed a single-center retrospective review of children with imaging confirmed CST from 2003 to 2014. Clinical data were extracted from the electronic medical record, and all available imaging was reviewed by 2 neuroradiologists. Data were extracted and stored in a Redcap database. Patients were identified via search of the electronic medical record, a local institutional review board–approved database for diagnosis of CSVT and subsequent chart review to identify the subset with CST.
Patients were included if aged between 1 month and 18 years, diagnosed with CST, and seen at Children’s Hospital Colorado at either presentation or follow-up with imaging available for review.
Internal Carotid Artery Narrowing Measurement
Experienced pediatric neuroradiologists (N.V.S. and D.M.M.), measured the degree of intracranial internal carotid artery stenosis according to the Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) trial method.8
Of the 121 patients with cerebral venous sinus thrombosis identified in the electronic medical record from January 2003 to July 2014, 7 patients had CST. An additional 3 were identified through local stroke databases. Ten patients had confirmed CST on imaging, ranging in age from 3 to 17 years (median, 11 years; Table; Table I in the online-only Data Supplement). Eight children had associated sinusitis. The most common presenting symptoms were headache (7/10), fever (6/10), and vomiting (6/10). The most common microbes were Streptococcus anginosus (6/10) and various staphylococcus/streptococcus species. All patients had a prothrombotic work-up revealing 1 patient as heterozygous for a factor V Leiden mutation.
The most common direct imaging findings of CST were CS expansion and filling defect/s (Figure 1; Table II in the online-only Data Supplement). Internal carotid artery narrowing (Figure 1; Figure I in the online-only Data Supplement) and arterial wall enhancement (Figure 1) were universal. There were 15 total cases of CAN between the 10 patients; all were associated with ipsilateral CST. The cavernous and petrous segments of the internal carotid artery were the most frequently narrowed (Table). CAN was >50% in 11 cases. In 6 of 10 patients, the CAN completely resolved, 2 had persistent >50% narrowing on their last available imaging study (24 months, 10 years), 1 had improvement of CAN (at 13 months), and 1 died 1 week after diagnosis of CST.
SOV enlargement and secondary thrombosis were frequently associated with CST. Draining and tributary veins of the CS were frequently thrombosed. The internal jugular veins and sigmoid sinuses were involved in 5 of 10 and 3 of 10 patients, respectively.
Restricted diffusion in the CS and in the SOV (when thrombosed) were common associated findings of CST (Figure I in the online-only Data Supplement). Subdural empyema (Figure 2) was universally present, whereas cerebritis was only noted in 2 patients. Restricted diffusion (Figure 2; Figure II in the online-only Data Supplement) was present in 6 of 10 patients in a pattern consistent with emboli or hypoperfusion. One patient had a large middle cerebral artery infarction. Pituitary involvement, evidenced by lack of normal pituitary enhancement or pituitary abscess, was present in 4 patients (Figure III in the online-only Data Supplement), only one of which was clinically significant. Orbital cellulitis was present in 5 patients and optic neuritis in 3 (Figure 2), all of which were clinically significant.
Patients were managed with a combination of antibiotics, ACT, antiplatelets, nimodipine, and steroids (Table; Table I in the online-only Data Supplement). All patients were treated with appropriate broad-spectrum antibiotics and then adjusted based on microbial sensitivities. Eight of 10 patients were treated with ACT. Three patients developed new infarcts on ACT. Four of the treated patients had bleeding complications during ACT. Of the 3 cases of intracerebral hemorrhage, 1 occurred spontaneously and 2 occurred post procedurally. There was no infratentorial extension. All 3 had minimal intraventricular extension and 1 had minimal subarachnoid extension. Two had minimal intraparenchmal hemorrhage. The volume of hemorrhage in each case was <20 mL, the lowest cutoff used in the most commonly used intracerebral hemorrhage grading scales.9 Nimodipine was given to 2 patients attempting to treat potential vasospasm. In both cases, nimodipine was initiated when infarction was noted during ACT. One of the 2 developed further infarcts while being administered nimodipine.
The majority of patients were managed operatively with functional endoscopic sinus surgery. Three patients underwent ventriculostomy shunt placement: 2 had orbital decompression, mastoidectomy, and craniotomy, and 1 underwent orbital exenteration and cerebral debridement (Table I in the online-only Data Supplement).
Four patients became blind in the affected eye (one bilateral) secondary to optic nerve injury or orbital exenteration. Seven patients were discharged with a modified Rankin Scale score of ≤1 (Table). At the last follow-up, the median modified Rankin Scale score was 1 (range, 0–3). One patient had care withdrawn because of stroke-related neurological injury.
In this cohort of childhood CST, many patients had a favorable outcome, despite the high incidence of CAN and infarct. There was little long-term impairment related to infarction in the majority of our series. This may be related to aggressive medical management and theoretical prevention of larger more debilitating infarcts. The most common neurological deficit was loss of vision in 1 or both eyes. The single death was related to a large infarction before ACT or antiplatelet therapy. Four of 8 patients were treated with ACT without complications, and none of the complications required discontinuation of ACT. In addition, nimodipine was used in 2 patients who had severe bilateral internal carotid artery narrowing and infarction despite ACT.
CST most commonly arose from sphenoid and ethmoid sinusitis. As in adult CST,2,4,6 common imaging findings of CST are CS expansion, CS filling defects, SOV enlargement, and thrombosis,2,4,6,10 as well as additional tributary/draining vein and dural venous sinus thrombosis.2,4 As additional venous thromboses were observed in 70% of our patients, these structures should be evaluated in patients with CST. Additional findings included restricted diffusion in the CS and SOV.11 These associated findings could be used to identify CST. Three of 5 patients with orbital cellulitis developed optic neuritis with vision loss, and 1 patient required bilateral orbital exenteration. Given the serious complications associated with orbital cellulitis in the setting of CST, it is important to carefully assess for changes in the orbital fat. CAN and arterial wall enhancement were universal in our cohort, findings frequently reported in adults.2,4,5 The arterial wall inflammation and CAN are a likely source of embolic and hypoperfusion-related infarcts.
In aggregate, previous pediatric series of CST reported 2 of 16 children with vision loss.2,3 Infarction was less prevalent than in our cohort, with only 1 child demonstrating infarct. In an adult series, 4 of 7 patients developed neurological impairment, including 1 with vision loss.12 Our experience suggests that infarcts and visual loss may be more frequent than previously reported. A detection bias may be present because 9 of 10 of our patients were imaged with magnetic resonance imaging. The proportion of patients with magnetic resonance imaging in the previously reported series is unclear. Paranasal sinus disease was an uncommon cause of CST in previous pediatric reports though common in our cohort. The most common pathogen identified in our cohort was S anginosus, which was less common in previous reports. Treatment with ACT, antibiotics, and surgical management was standard in the adult cases, whereas the pediatric cases were not treated with ACT in previous reports.
In our modest cohort of childhood CST, CAN and infarcts were common. Although therapeutic interventions to prevent clot propagation or vasospasm did not prevent all new infarcts, they may have contributed to the prevention of progressive infarction. Despite the high incidence of CAN and infarction, outcomes were often favorable. The variation in treatment, incidence of infarcts, and variable outcomes between our cohort and previous series precludes treatment recommendations without further multicenter studies for childhood CST.
Sources of Funding
This study was supported by American Stroke Association/Bugher Foundation Stroke Collaborative Research Center 14BFSC17540000.
↵* Drs Press and Lindsay contributed equally.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.009437/-/DC1.
- Received March 20, 2015.
- Revision received July 12, 2015.
- Accepted July 14, 2015.
- © 2015 American Heart Association, Inc.
- Schuknecht B,
- Simmen D,
- Yüksel C,
- Valavanis A.
- Ahmadi J,
- Keane JR,
- Segall HD,
- Zee CS.
- Roach ES,
- Golomb MR,
- Adams R,
- Biller J,
- Daniels S,
- Deveber G,
- et al
- Samuels OB,
- Joseph GJ,
- Lynn MJ,
- Smith HA,
- Chimowitz MI.