Cerebral venous thrombosis (CVT) is a rare cause of stroke that affects mainly young adults and children, in contrast to arterial stroke and venous thromboembolism.1 CVT has an incidence of 0.4–0.7 cases per infant per year.2 The incidence has increased over time, probably due to increased clinical awareness as well as improvement in neuroimaging techniques that enable the diagnosis of less severe cases.1 Indeed, the diagnosis of CVT needs to be confirmed by neuroimaging, including computed tomography (CT) and/or magnetic resonance imaging (MRI) with venography and/or catheter angiography.

CVT represents a serious and life-threatening disease with significant morbidity and mortality, usually due to hemorrhagic or non-hemorrhagic venous infarction.3

Affected patients usually present with non-focal neurological signs and symptoms, including headache, seizures and vomiting. Other possible presentations include focal neurologic deficits (hemiparesis and aphasia), isolated intracranial hypertension, diffuse encephalopathy or cavernous sinus syndrome.1

CVT is a multicausal disease triggered by the interaction of several risk factors, such as inherited or acquired thrombophilia, pregnancy and the most common, infection.4

When assessing the thrombotic risk, it is important to assess genetic and environmental risk factors.5 Hereditary thrombophilia, with a prevalence of <1% in the general population, may be caused primarily by deficiencies of natural anticoagulants including protein C and protein S or by genetic mutation of prothrombin or factor V.6 Several conditions, such as antiphospholipid syndrome (APS), are associated with acquired thrombophilia.7 Anticoagulation is the primary treatment regardless of the cause. Although intracerebral thrombosis secondary to thrombophilia is not very rare, it may be susceptible to misdiagnosis.

As this is a rare disorder in pediatric patients, there are only a few published case series of cases of CVT focused in children and we aimed to provide a description of all our pediatric diagnosed cases of CVT, including their clinical presentation, laboratory and neuroimaging studies and outcomes.

This is a retrospective single-center observational and analytical study of all consecutive pediatric patients admitted to Centro Hospitalar Vila Nova de Gaia/Espinho, Portugal from 2003 to 2021 with the diagnosis of cerebral venous thrombosis. The diagnosis was made based on the clinical presentation and confirmed by neuroimaging (CT and/or MRI with venography (CTV/MRV)) at the time of the initial diagnosis. Demographic and clinical information, as well as imaging findings, predictor and prothrombotic factors, treatment strategies and outcome of CVT, were documented for each patient. No patients were excluded.

Herein we present the clinical and imaging findings at presentations of our 12 pediatric patients (including two newborns) with CVT, as well as prothrombotic factors, treatment strategies, clinical outcomes and recanalization.

In children, CVT is relatively equally distributed between genders in all age groups.8 However, we found a higher proportion of females, against to previous articles2,9 eventually due to our small sample size. Some studies exclude newborns since the clinical presentation is different, but in this study we included two newborns.

Some reports describe seizures, decreased consciousness and headaches as the most common initial symptoms,10 with seizures being common in newborns.11 Focal and diffuse neurologic signs are likely to occur3 and other possible presenting symptoms may vary according to the cause/trigger of the CVT. As previously described, neurologic signs as vomiting and headaches were the most common presenting symptoms in our study.

There are three possible neuroimaging techniques to diagnose CVT: CT/CTV (computed tomography-venography), MRI/MRV (magnetic resonance imaging with venography) and catheter angiography.1 MRI/MRV is the imaging technique of choice for the diagnosis of CVT in children, allowing visualization of the thrombus within the vessel combined with absent flow on MRV as well as presence of associated venous infarction and/or subjacent causes, without the use of ionizing radiation.4 However, CT/CTV is frequently the initial diagnostic imaging method in the acute setting due to the easy access in the emergency room and the speed of acquisition, while diagnosis with catheter angiography is exceptional. Indeed, except for one newborn diagnosed at presentation by MRI/MRV, all our patients were initially investigated with CT/CTV, although majority of them also performed subsequently an MRI/MRV. Nevertheless, sedation techniques may be required for both techniques, especially MRI due to the longer acquisition time.

Multiple sinuses thrombosis was the most frequent scenario in our population, which is concordant with previous pediatric studies.8–10 Transverse and sigmoid sinuses thrombosis were the most frequently affected sinuses in our study, followed by the superior sagittal sinus, also as previously reported in children.11

The most common trigger for cerebral venous thrombosis in our cohort was central nervous system (CNS) or head and neck infection, as reported in most previous studies.2,3,11 Most patients with CNS or head and neck infections had multiple sinuses involvement, as well as the two cases of nephrotic syndrome. Oral contraceptive may have contributed for the CVT in two female adolescents, one with Down Syndrome, obesity and hypothyroidism. No trigger nor prothrombic factor was discovered in the first event of the patient with recurrent CVT. In the subsequent CVT episode, she was on oral contraceptives for more than three years against medical advice. In newborns, there is an increased risk of cerebral venous thrombosis that may be related with alternative mechanisms such as mechanical compression.12 The patient presenting with seizures in the neonatal period had no other risk factor and the cesarean section delivery with ventouse probably contributed to the CVT, in a multifactorial process.

It remains unclear whether all individuals with thrombosis should be screened for thrombophilia. Screening is generally not recommended for the general population, but there is consensus for some specific indications that include: thrombosis during perinatal period, following the use of oral contraceptives or estrogen replacement therapy and venous thromboembolism in uncommon sites such as the brain.13 Therefore, there is a formal recommendation for screening prothrombotic factors in all the patients included in this study. At our center, the screening for thrombophilia is mainly performed after the recommended anticoagulation period. All patients, except the two lost to follow-up, were investigated for prothrombotic factors. No major genetic prothrombotic factor was found. Nonetheless it is always important to exclude these since there may be an indication for lifetime anticoagulation.

The American Society of Hematology (ASH) recommends and suggests the use of anticoagulants in pediatric CVT with and without hemorrhagic venous infarction, respectively.14 Except for the two newborn patients, all cases were anticoagulated with low molecular weight heparin (LMWH). Three patients were latter switched to warfarin. ASH recommendations do not prefer LMWH to warfarin or the opposite and determine the decision depends on patients values and preferences, underlying conditions and comorbidities and health resources.14 Most patients were treated for a period of 3–6 months. One of them was treated for almost 18 months but he missed several appointments and continued anticoagulated longer than recommended by the specialist. Prothrombotic factors were investigated in 83% (10/12) of cases. A thrombophilia panel was performed including Protein C, Protein S and antithrombin activity, Activated protein C resistance, and the genetic mutations of the factor V (Leiden), prothrombin, Methylenetetrahydrofolate reductase and Plasminogen activator inhibitor 1 (PAI-1) gene. Five had heterozygous mutations in one or two genes and one was homozygous for the 4G/4G polymorphism of the PAI-1 gene. All had normal homocysteine levels, so no major genetic prothrombotic factor was found in our series.

Pediatric antiphospholipid syndrome was also excluded discarding the presence of lupus anticoagulant in plasma, anticardiolipin antibodies of IgG or IgM isoforms and anti-beta-2 glycoprotein I (anti-β2GPI) antibodies of IgG or IgM isoforms.

As the most common trigger of CVT in our cohort was CNS or neck/head infection, 75% of patients was treated with antibiotics. The used antibiotic varied depending on the primary infection focus. Two patients were treated with antivirals because of the presenting symptoms: one presented with right hemiparesis and the other was a 16-year-old female with seizures at day 2 of hospital admission. Two cases of severe infection (deep neck abscess and acute mastoiditis) and the two cases of nephrotic syndrome were treated with corticosteroids. Our two cases with nephrotic syndrome had hypoalbuminemia upon presentation, and low albumin levels has been considered the major risk factor associated with thromboembolic events in this syndrome.15

No deaths occurred in our study during the follow-up time and most patients evolved without any major sequelae. Only one case of recurrence after 18 years of the primal event, both without any neurological sequelae. Our data is similar to other pediatric studies, with overall favorable prognosis being described in this age group,1 and mainly dependent on the etiological factor for thrombosis11 and the occurrence (or not) of venous infarction. Indeed, cerebral thromboembolism is the most common neurological complication of nephrotic syndrome with favorable outcome in approximately 90% of patients.15 Accordingly, our two patients with CVT associated with nephrotic syndrome are now adults without any neurological sequelae. Ritchey et al. described a worse outcome for patients with cortical vein thrombosis without involvement of the dural sinuses,16 we have found similar results. Indeed, our patient with isolated cortical vein occlusion evolved with severe global development delay, most likely due to the bacterial meningitis. The other two patients with neurological sequelae had severe central nervous system infections, one other case of bacterial meningitis with empyema and a case of meningoradiculitis due to herpes-7 and this girl maintained right hemiparesis and learning difficulties. The injury related with CNS infection is more likely the cause of the long-term neurological sequelae than the CVT itself.

There are currently no guidelines requiring a systematic venographic follow-up evaluation in children for assessment of recanalization nor establishing the ideal imaging method to confirm recanalization.17 Indeed, long term recanalization is so far not considered a significant outcome factor of CVT11 as persistent intracranial venous occlusion does not appear to mean worse outcome in both children and adults.8 Therefore, imagiologic resolution of CVT is not always documented and depends on clinical judgement at an individual basis. However, clinical symptomatic control should be performed.

Although all our cases performed at least one neuroimaging study after the initial event, in most cases they were acquired in the emergency room in the setting of other acute, unrelated clinical reasons. Nevertheless, recanalization could be assessed in 10/12 cases, and considered complete in 8 of them. Concerning the two cases with only partial recanalization, it should be noted that their last imaging study was performed less than a year after the initial CVT event, and therefore subsequent evolution towards complete recanalization cannot be fully excluded.18

Finally, a recent study demonstrated a very low yield of routine follow-up neuroimaging aiming to detect new dural arteriovenous fistulae 6 months after CVT19 and accordingly we have neither detected development of this type of brain vascular malformations in any of our patients during imaging follow-up.

The results of this cohort were overall similar to those currently found in current literature. The clinical presentation of CVT in pediatric patients is not specific, most often with headache and seizures, therefore a high level of suspicion should be present. MRI/MRV is the preferred imaging method for diagnosis in children, despite not always being available in an emergency setting. Multiple sinuses involvement was the most common pattern, with the transverse and sigmoid sinuses being the most affected vessels. Anticoagulation should be started at diagnosis and the presence of hereditary or acquired thrombophilia evaluated after the minimum recommended time of anticoagulation. Anticoagulation treatment with LMWH is recommended and important to reduce the mortality rate, as seen in our case series. CVT is frequently associated with CNS and/or head and neck infections in children, and the former can also lead to parenchymal injury and lead to high morbidity and poor outcomes as demonstrated in a few of our cases. Recanalization does not appear to be a significant factor for prognosis and predictor of neurological outcome in both children and adults and there are currently no studies or guidelines concerning the expected time of recanalization and if follow-up neuroimaging control MRI/MRV should be routinely performed and at what time point.

The authors declare that there is no conflict of interest.