Epidemiology, Clinical Features, and Outcome in a Cohort of Adolescents With Cerebral Venous Thrombosis

Abstract

Background and Objectives

We aimed to analyze the epidemiologic, clinical, and paraclinical features of adolescents with cerebral venous thrombosis (CVT) and its therapeutic management and outcome.

Methods

This multicenter retrospective cohort included patients 10 to 18 years of age hospitalized for a first episode of CVT in 2 French regions between 1999 and 2019. The number of cases was compared to the number recorded by the French health insurance system. The CVT registry of the Lariboisière hospital allowed comparisons with adults.

Results

One hundred two patients were included (52.9% female; median age 15.1 years). Estimated incidence was 0.37 to 0.38 per 100,000 adolescents per year; 45.5% of patients presented with focal deficits or seizures or in a coma. Male patients were younger than female patients (14.2 vs 15.6 years; p < 0.01) and more often admitted to intensive care (52.1% vs 24.1%; p = 0.0,035). The lateral sinus was the most common CVT location (72.3%), and 29.4% of adolescents had associated venous infarction or hematoma. Most patients (94.1%) received anticoagulation. Treatment also included an endovascular procedure (2.9%), decompressive craniectomy (4.9%), and CSF shunt (6.9%). The most frequently identified CVT-associated condition was local infection in male (18.6%) and systemic disease in female (14.8%; p < 0.001) patients. The proportion of CVTs in adolescents without an identified associated condition or risk factor was low (1.9% vs 11.4% in adults; p < 0.002). Adverse outcome at 1 year was more frequent than in adults (33.3% vs 11.8%; p = 0.0,001).

Discussion

CVT in adolescents is rare and complex with specific epidemiology, including differences in clinical presentation and associated conditions between sexes, and more severe outcomes than in adults. Careful specialized management and follow-up are therefore recommended.

Cerebral venous thrombosis (CVT) is a rare and serious disease. Its incidence is being revised upward, up to 1.57 per 100,000 per year, while its mortality rate is decreasing.^1-3 Overall, this condition predominates in women, young adults, and younger children, including neonates.⁴ The sex ratio differs by age, with an initial male predominance in neonates⁵ and children⁶ and a female predominance in young adults that does not persist among elderly adults.^4,7 Studies have highlighted important differences between the sexes in the causes, clinical presentation, and prognosis of CVT.^8,9

Adolescence is a very specific period of life that is more than just a transition from childhood to adulthood and, in terms of health issues, needs to be considered separately.¹⁰ Indeed, adolescence appears to be a pivotal age during which neurologic disease can have dramatic consequences for all aspects of a patient's life. For example, specific causes and courses have been shown in arterial strokes in adolescents.¹¹ For venous thromboembolic disease, age distribution shows a peak of incidence in adolescents.¹² Nevertheless, to the best of our knowledge, no study has yet been devoted to CVT during adolescence.

Our main objectives were to describe clinical and paraclinical features, associated conditions, risk factors, therapeutic management, and outcome of CVT in adolescents and to compare these features with those of CVT in adults to provide helpful data for clinicians managing these patients.

Methods

Study Design and Patients

We conducted a retrospective multicenter cohort study in 2 adjacent French regions representing a population of ≈10 million inhabitants. We postulated that patients with CVT who were ≤18 years of age were managed only in academic centers and focused our study on adult stroke units and neuropediatric units. Different neurology departments, including 11 adult stroke units and 4 neuropediatric units located in 14 university hospitals in the Paris-Ile-de-France region (Bicêtre, Bichat, Debré, Foch, Lariboisière, Mignot, Mondor, Necker, Pitié-Salpêtrière, Rothschild, Sainte-Anne, Sud-Francilien, Saint-Joseph, and Trousseau hospitals) and the only adult and pediatric stroke unit in the Burgundy region (Dijon), participated in the present study. All consecutive patients ≥10 to ≤18 years of age and diagnosed with a first-ever CVT hospitalized in one of the participating centers between December 1, 1999, and December 1, 2019, were included. Cutoff ages were chosen according to the World Health Organization and the American Psychological Association definitions of adolescence. Patients were screened via a search of the discharge databases of the participating hospitals (Programme de médicalisation des systèmes d'information, the French national hospital discharge database) according to the ICD-10 (thrombophlebitis, cerebral infarction due to CVT, CVT in pregnancy or in the puerperium).^2,13 CVT definition comprised evidence of recent thrombosis on venous sequences (time of flight/T1 gadolinium) on MRI or on CT venography.

Standard Protocol Approvals, Registrations, and Patient Consents

We submitted our study design to the local Institutional Review Board/Independent Ethics Committee (CPP VII). Because it is a retrospective observational study with anonymized data without any additional therapy or monitoring procedure, the Institutional Review Board/Independent Ethics Committee did not identify ethics issues and declared that, in accordance with French legislation, formal approval was not required.

Data Collection

We reviewed medical records of all included patients using a standardized approach to collect these clinical characteristics: demographic data, preexisting thrombotic risk factors and medical conditions, clinical presentation, and time between symptom onset and diagnosis. Initial management and treatment modalities, type and results of diagnostic workup (including cerebral imaging to determine involved sinus and parenchymal lesions, blood tests, CSF analysis, thrombophilia investigation, and any more specialized analyses), retained associated conditions and risk factors, clinical outcome, and potential recurrent venous thromboembolic events were also collected. We visualized and reanalyzed all diagnostic brain and venous sinus imaging data (J.D., M.K., and C.D.).

Clinical Features in the Acute Phase

Clinical presentation was classified into 3 categories: isolated headache (i.e., with or without nausea/vomiting), isolated intracranial hypertension (i.e. diplopia, CSF pressure ≥25 cm H₂O, or papilledema), or symptom of brain damage (i.e., focal deficit, coma, or seizures). Symptom onset was defined as acute (<48 hours), subacute (<1 month), or chronic (>1 month). Time before medical visit, number of visits before diagnosis, and time before diagnosis (defined as time between symptom onset and diagnosis confirmed by imaging) were noted when available.

Etiologic Investigations

CVT-causing factors were divided into 2 groups, associated conditions and risk factors, with the realization that for most of them, causality is not proved because they simply increase the risk of CVT. The first group of factors was called CVT-associated conditions if a condition/disease such as cancer or a local major infection was diagnosed, and the second group was called risk factors if only a common comorbidity was identified (such as obesity or use of combined oral contraception [COC]). CVT-associated conditions were classified into 7 categories: (1) hemopathy or cancer; (2) systemic inflammatory disease (including Behçet disease, systemic lupus erythematosus [SLE], etc); (3) nonacquired thrombophilia, including major genetic coagulation abnormalities (prothrombin heterozygous mutations, dysfibrinogenemia, antithrombin III or protein S deficiency, increased lipoprotein[a] level, cystathionine β-synthase deficient homocystinuria), and minor abnormalities (such as heterozygous factor 5 mutation, factor 5 deficiency, and homozygous MTHFR mutations); (4) general cause of hypercoagulability (including anemia, nephrotic syndrome, etc); (5) local infection; (6) noninfectious local cause (including CSF hypotension, head trauma, or vascular malformation); and (7) no established associated conditions. COC use, first-degree family history of venous thrombosis, and obesity were considered to be risk factors^14-16 rather than associated conditions. The number of patients with multiple CVT-associated conditions/risk factors was noted. Where several associated conditions were present, the most relevant was picked according to chronology and pathophysiology.

Outcome

Duration of hospital stay was noted. Clinical outcome was assessed with the modified Rankin Scale (mRS) 3 months and 1 year after CVT. Brain imaging follow-up results and time from initial CVT to venous recanalization were collected.

Incidence

In France, the national census regularly provides detailed population data according to age/region. We used these data to estimate the population of adolescents in the area covered by participating centers during the study period. Annual incidence of CVT in adolescents in the catchment area was then estimated from the number of patients included in our study compared with the population data.

Data Validation: Comparison With National Data From the French National Health Data System

To confirm the reliable identification of all potential cases despite the retrospective nature of our study, we compared the number and age distribution of first-ever CVT observed in our study to those observed in the total French population for the same age category over a corresponding period of time using data extracted from the French national health data system (acknowledging that CVT diagnosis within the French National Health Data System is not firmly validated because it is based only on the ICD coding within medical records). In France, the national health insurance information system contains individual, anonymized data concerning the beneficiaries of the various national health insurance schemes for the entire French population (64.7 million inhabitants on January 1, 2018, including 7.1 million individuals ≥10–≤18 year of age). Using these databases, we were able to compare the age-specific distribution in our study population with that in the total French population.

Subgroup Comparisons

Within our sample, we compared some characteristics according to sex or age (youngest adolescents, those <14.5 years; oldest, those >14.5 years).

Comparison With the Prospective Lariboisière Cohort of Adult CVT

Some characteristics were compared with those of this cohort of 498 adults with CVT, recruited from 1998 to 2018.¹⁷

Statistical Analysis

Statistical comparisons were made for different features between adolescents and adults and among adolescents between male and female patients and youngest and oldest patients. Descriptive data were reported associated with the interquartile range (IQR) or SD. Distributions were compared by use of a χ² test when the population in every group was ≥5 and with the Fisher exact test when the theoretical head count in at least 1 group was <5. When necessary, we used the Bonferroni method to correct the α value to avoid bias caused by multiple comparisons. When necessary, a multivariate analysis using a logistic regression model was realized, and odds ratios (ORs) were calculated.

Data Availability

Anonymized data on which this article is based will be shared on reasonable request with any appropriately qualified investigator.

Results

Population Characteristics and Demographic Data

Between 1999 and 2019, 102 patients ≥10 to ≤18 years of age with a first-ever CVT were admitted in 9 different centers of 15 participating hospitals (from 2–31 patients per center). A slight majority of patients (52.9%) were female. Median age at CVT occurrence was 15.1 years (IQR 12.7–17.0 years) in our study and 16.1 years in the national population (IQR 14.0–18.0 years). In the national data, CVT occurrence significantly increased with age (R = 0.83, p < 0.001). In our cohort, this increase was found only in female patients (Figure 1). Male patients were significantly younger than female patients at the time of CVT occurrence (p < 0.01), and the proportion of female patients tended to increase with age without reaching significance (48% in 10- to 14.5-year-old adolescents vs 58% in 14.5- to 18-year-old adolescents, p = NS) (Table 1). In line with this, 72% of female patients were pubescent at CVT occurrence vs 47.7% of male patients (p = 0.02). The estimated incidence of first-ever CVT in adolescents was 0.38 per 100,000 adolescents per year in our catchment population according to our analysis of medical records and 0.37 per 100,000 adolescents per year in the general French population according to ICD coding in medical records.

Figure 1. CVT Distribution According to Age and Sex.

Central venous thrombosis (CVT) distributions are presented according to age category and by sex (female patients in the left panel and male patients in the right panel). Our cohort is shown in black; the nationwide population based on the French national health system is shown in gray.

Table 1.

Initial Presentation of CVT in Adolescents: Comparison According to Sex

Presentation and Diagnosis

We found no difference in the pattern of symptom onset between male and female adolescents or between youngest and oldest adolescents; onset was equally acute or subacute and rarely chronic. The mean delay between symptom onset and diagnosis was >1 week. CVT was not suspected during the first medical consultation for 57.4% of the adolescents, and 24.5% of diagnoses required >2 medical visits (31.8% of male, 18.0% of female patients, p = NS). Delay between symptom onset and diagnosis was significantly shorter in the presence of a focal deficit (5.1 [ SD 8.3] vs 9.3 [SD 9.2] days; p < 0.01). Clinical presentation comprised brain damage symptoms in 46 patients (45.5%), of whom 28 (60.9%) had venous infarction, cerebral hemorrhage, or subarachnoid hemorrhage on imaging (this increased to 84% in the presence of epilepsy). The focal deficit was motor only in 12 cases, sensitive in 8 cases, and multimodal in 16 cases, including 8 cases with motor deficit and aphasia. The oldest adolescents presented with epilepsy or headaches more often than the youngest (11.4% vs 8.6%, p < 0.01, and 93.0% vs 78.6%, p = 0.04, respectively). Twenty-three patients had a sixth cranial nerve palsy. Funduscopic examination was performed in 57 patients (55.9%) and identified papilledema in 37 (64.9%; 86.6% of the patients with intracranial hypertension [n = 36]). Mean Glasgow Coma Scale score at admission was 13.6 (range 3–15). Male patients were admitted to intensive care more often than female patients (52.1% vs 24.1%; p < 0.01).

Initial Imaging and Workup

Overall, CVT diagnosis was established more frequently on a CT scan than on MRI except in the youngest group, in whom MRI was used more frequently (55.8% vs 22.8% in the oldest group; p < 0.001). Methods of CVT diagnosis did not change across the years of the study period, with diagnoses based on MRI in 30.6% and 41.5% of cases during the periods 1999 to 2010 and 2011 to 2019, respectively (p = 0.25). Thrombosis location was similar between male and female patients and between younger and older adolescents. Extended thrombosis was common (64.4% occurred in multiple locations). Cortical vein thrombosis existed in 26 patients (25.7% of cases), including in 5 patients (5%) without involvement of either dural sinuses or the deep venous system. These 5 patients with isolated cortical vein involvement presented with CVT revealed by seizures or focal neurologic deficit. Thirty adolescents (29.4%) had parenchymal damage, with venous infarction, cerebral hemorrhage, or both. Of these, only 3 patients did not have a neurologic deficit: 1 had a headache; the other 2 had signs of intracranial hypertension. Venous infarctions were significantly more frequent in female (27.8% vs 10.5%) while cerebral hemorrhages were more frequent in male (22.9% vs 14.8%; p = 0.03) patients. Of the 11 patients with subarachnoid hemorrhage, only 4 had recent head trauma. Thrombophilia workup was performed more often in female than in male patients (87% vs 66.7%; p = 0.01). The frequency of elevated C-reactive protein was similar in both sexes. Of the 42 patients who underwent lumbar puncture, 13 had elevated CSF pressure (>25 cmH2O), and 4 had aseptic meningitis. When analyzed, D-dimer levels were always high.

Therapeutic Management

Globally, medical CVT treatments included acute medications (anticoagulation, revascularization, chemotherapy, corticosteroids, or antibiotics) and long-term therapies (anticoagulants and/or antiepileptic drugs [AEDs]) and was combined in some patients with surgical procedures (craniectomy or CSF shunting). Fifteen adolescents (14.7%) were placed on mechanical ventilation, with no difference between male and female patients. Most patients received anticoagulation (male 91.7%, female 96.3%; p = NS). Six patients did not receive anticoagulation or any antithrombotic therapy. Two had a posttraumatic subarachnoid hemorrhage, and the other 4 underwent surgery for various reasons: 2 had a local infection; 1 had a posterior fossa tumor with postoperative empyema; and 1 had a dural fistula and had undergone a ventriculocisternostomy. Mean duration of anticoagulation was 12 months (no difference between male and female patients) and was significantly longer in the oldest adolescent group (mean 15.0 ± 10.8 months vs 7.7 ± 17.0 months in the youngest group; p < 0.001). A hemorrhagic event occurred during treatment in 4 patients. Two had epistaxis,; 1 underwent a splenectomy for spleen hemorrhage; and 1 patient with recent severe head trauma experienced cerebral hemorrhage. Heparin-induced thrombocytopenia occurred in 2 patients (1.9%). After initial parenteral anticoagulation, switching to oral anticoagulation occurred more frequently in female than in male patients (79.6% vs 43.8%; p < 0.001). Four patients received direct oral anticoagulants; none experienced bleeding complications. Four patients had revascularization treatment: 1 patient received IV thrombolysis; 2 patients underwent endovascular thrombectomy (associated with in situ thrombolysis in one patient); and 1 patient had surgical thrombectomy. Nine patients underwent decompressive craniectomy or CSF shunting. They all presented with intracranial hypertension, and 7 presented with either seizures or focal deficit or in a coma. Fifteen patients underwent drainage surgery for abscess, bone, or brain infection. Twenty-three patients received corticosteroids, mostly for systemic disease or in combination with antibiotics for ear/nose/throat infections. Thirty-one patients (30.4%) received AEDs, either therapeutic (n = 27) or prophylactic (n = 4). Twenty-five patients received monotherapy, and 6 received combination therapy. The most commonly used AED was levetiracetam (n = 18).

CVT-Associated Conditions and Risk Factors

A definite associated condition was found in 84 patients (82%). In 31 patients (30.4%), CVT led to the diagnosis of an underlying disease. The distribution of CVT-associated conditions differed between the male and female patients in the cohort (p < 0.001). CVT-associated conditions were the following (see Table 2). (1) Hemopathy or cancer (n = 8) was significantly more frequent in male than in female patients (p = 0.02). With the exception of dysgerminoma in a female patient, all systemic cancers were lymphoma or leukemia in male patients; 4 patients had received l-asparaginase, a drug known to be prothrombogenic.¹⁸ Brain tumors were proliferative angiopathy or posterior fossa tumors. (2) Systemic disease was seen in 16 patients. Behçet disease was the most frequent diagnosis (n = 10, 7 male and 3 female patients); all diagnosed patients originated from the Mediterranean area (9 had North African and 1 had Italian heritage). The other systemic diseases were SLE (n = 3; all female), inflammatory bowel disease (n = 1), and antiphospholipid antibody syndrome (n = 6: isolated [n = 2] or associated with SLE [n = 2] or Behçet disease [n = 2]). (3) Thrombophilia, including major genetic coagulation abnormalities, was noted in 10 patients. Prothrombin heterozygous mutation was seen in 5 patients. Five other patients had dysfibrinogenemia (n = 1), antithrombin III deficiency (n = 1), protein S deficiency (n = 1), increased lipoprotein(a) (n = 1), and homocystinuria caused by cystathionine β-synthase deficiency (n = 1) (thrombophilia workup also found minor abnormalities in another 5 patients: heterozygous factor 5 mutation [n = 3], factor 5 deficiency [n = 1], and homozygous MTHFR mutation [n = 1], but these abnormalities were not considered major causes of CVT). (4) Fourteen patients had a general cause of hypercoagulability. All 6 cases of severe anemia occurred in female patients (p = 0.03). (5) Local infection (n = 19), mainly sinusitis and otitis, was found in 18.6% of patients, more frequently in male patients (29.2% vs 9.3% in female patients; p = 0.01). (6) Noninfectious local cause was noted in 17 patients. CSF hypotension (n = 2) was related to lumbar puncture in 1 patient and was spontaneous in another. (7) No identified CVT-associated condition was found in 18 patients. Of these 18 patients with a CVT without an established associated condition, only 2 did not present with any risk factor. The others had a first-degree familial history of venous thrombosis, were obese, or were taking COC (Table 2). If only postpubertal female adolescents were considered, 23 of 37 were taking COC (62.2%), similar to the proportion observed in the adult group (64.7%) (Table 2). In the whole cohort, 41 patients (40.2%) presented with >1 CVT-associated condition or risk factor. There was an inbreeding background for 8 patients (7.8%). Seventy-two percent of female patients were pubescent compared to 47.7% of the male patients (p = 0.02). All in all, CVT-associated conditions and risk factors represented transient/acquired causes in 71.6% of cases (hemopathy and cancers, acquired causes of hypercoagulability, infections, noninfectious local causes, COC, etc) and permanent conditions in 28.4% of cases (thrombophilia, constitutional hemopathy, etc).

Table 2.

Acute Management, Associated Conditions, and Risk Factors of CVT in Adolescents: Comparison According to Sex

Outcome

Median follow-up was 24 months (IQR 10–69 months). Two patients of the cohort had a fatal outcome: 1 early death related to a complication of endovascular treatment of a dural fistula at CVT diagnosis and 1 later death related to relapse of meningeal leukemia 20 months after CVT. Ninety-four patients (93.1%) had follow-up imaging (CT scan only in 12%, MRI only in 72%, and both in 16%). Venous recanalization was more frequent in female than in male patients (79.6% vs 60.0%; p = 0.04) at an average of 3.9 ± 3.8 months after CVT diagnosis. Absence of recanalization was not a predictor of thrombosis recurrence. Recurrent venous thrombosis was more frequent in male (n = 6) than in female (n = 1) patients: 1 patients had Behçet disease, and 2 had congenital thrombophilia. One male adolescent with leukemia had a second CVT 1 month after the first event while on effective anticoagulation. The other recurrences were in the postpartum period for 1 patient and in the setting of cerebral arteriovenous malformation for the remaining patient.

At the 3-month follow-up, female patients had better outcomes than male patients (mRS score 0–1 prevalence in female vs male patients; p = 0.02, Table 3). This difference was no longer significant after 1 year (Figure 2). Among those with an adverse outcome (mRS score ≥2) after 1 year of follow-up, significantly more patients had had a severe initial presentation compared to patients with an mRS score of 0 or 1: coma (OR 4.45, 95% confidence interval [CI] 1.21–16.35; p = 0.03), focal deficit (OR 3.09, 95% CI 1.16–8.23; p = 0.02), epilepsy (OR 5.18, 95% CI 1.85–14.49; p = 0.02), or initial brain hemorrhage (OR 6.72, 95% CI 1.92–23.48; p < 0.01). A higher proportion of female than male adolescents had chronic secondary headaches (46.0% vs 14.6%; p < 0.01).

Table 3.

Outcome After CVT in Adolescents: Comparison According to Sex

Figure 2. Clinical Outcome and Autonomy (According to mRS Score at 1 Year).

Comparisons of the distribution of modified Rankin Scale (mRS) score at 1 year are CVT are shown according to sex and age (all adolescents [n = 102] vs adults [n = 498]). Deaths occurring during the acute phase are excluded (mRS score 0 = no symptoms; 1 = no significant disability despite symptoms; able to perform all usual duties and activities; 2 = slight disability; unable to perform all previous activities but able to look after own affairs without assistance; 3 = moderate disability; requires some help but able to walk without assistance; 4 = moderately severe disability; unable to walk without assistance and unable to attend to own bodily needs without assistance; 5 = severe disability; bedridden, incontinent, and requires constant nursing care and attention; and 6 = dead).

Comparison With CVT in Adults

CVT characteristics in our cohort of adolescents differed from those in the adult registry in terms of clinical presentation (mainly less frequent isolated headaches [p < 0.01] and subacute or chronic onset [p < 0.0001]) and the distribution of associated conditions and risk factors (hemopathies/cancers, infections, or hypercoagulability overrepresented in adolescents; p < 0.0001); we also observed a lower rate of CVT without identified associated condition or risk factor in adolescents (2.0% vs 11.4%; p < 0.01) (Table 4). Poor outcome (mRS score ≥2) at 1 year was more frequent in adolescents than in adults (p < 0.001) (Table 4 and Figure 2).

Table 4.

CVT Characteristics in Adolescents: Comparisons With Adult CVT

Discussion

To the best of our knowledge, the present CVT cohort is the first specifically comprising adolescents. Calculated CVT incidence was 0.37 to 0.38 per 100,000 adolescents per year in our study. CVT remains a rare condition in adolescents. The incidence seems to be lower than in adults (1.3–1.5 per 100,000 adults per year in high-income countries^1,2) or newborns (2.6–3.4 per 100,000 neonates per year^19,20) and quite similar to that in children (0.3–0.6 per 100,000 children per year^20,21). This rarity may be why misdiagnosis during the first medical consultation was common (57.4%), possibly more so than in adults.²²

For some, adolescence is seen as a continuous transition process between childhood and adulthood. An age effect was observed in our cohort, delineating a CVT profile more comparable to that of pediatric CVT in younger adolescents and to that of adult CVT in older adolescents. Our adolescents exhibited clinical presentation features that fit into a clinical continuum from childhood to adulthood, with a typical increase in headache frequency^20,23 and decrease in epilepsy, focal deficit (occurring as frequently as in adults but less than in young children²¹), and acute mode of onset²⁴ according to age. An important effect of sex was also observed, probably related to puberty and hormonal changes. This effect is well known in adults but much less pronounced in pediatric CVT. Schematically, 2 typical profiles of adolescent patients with CVT emerged in our study: a young male patient with local infection, Behçet disease, or cancer who has a significant risk of early neurologic sequelae (typically resembling a late pediatric CVT profile) and an older female patient with anemia or COC use who is overweight and has a slightly better prognosis except for chronic headaches (typically resembling an adult CVT profile). This effect of sex was observed not only in clinical presentation, management, and type of workup but also in outcomes. Male adolescents presented with more severe clinical and radiologic features. Intracranial hemorrhage was observed; intensive care unit management was required more frequently; and underlying cancer disease was more common. They also had more frequent thrombosis recurrence and poorer outcomes at 3 months. In contrast, female adolescents more frequently had a thrombophilic workup and received anticoagulation. They also had a better rate of venous recanalization and better outcomes at 3 months.

Anticoagulation was widely used in our cohort. The rate of anticoagulation treatment was again between reported rates in pediatric and adult cohorts. Children with CVT less frequently receive anticoagulation (60%–80%, depending on the series^21,23,25), while adults almost systematically receive anticoagulation.^5,8,11 The duration of treatment was longer in older adolescents, closer to the adult recommendations of 6 to 12 months²⁶ than to pediatric recommendations of <6 months^27,28 The proportion of adolescents with CVT in whom no associated conditions or risk factor was found is low (1.9% vs 11.4% in adults). In children, this proportion varies from 2% to 39%.^21,25,29 Risk factors found in our cohort such as overweight, use of COC, and family history are those also found in deep vein thrombosis in adolescents.³⁰ The number of adolescent patients with multiple CVT-associated conditions or risk factors is high in our cohort, suggesting that broad and systematic etiologic investigations including thrombophilia workup are required so as not to neglect underlying diseases requiring specific treatments and factors that may favor recurrence. Last, deaths were very rare and not directly related to CVT in our study, but the rate of sequelae was high (≈30%), and sequelae were predominantly cognitive deficits associated with school difficulties. This should alert physicians to the risk of disability for the future lives of these adolescents and raises the need for appropriate follow-up, especially developmental, with cognitive therapy if necessary. In addition, recurrent secondary headaches were common, observed in nearly half of the female adolescents.

The main limitation of our study is its retrospective nature, with the intrinsic risk of missing cases (mild, spontaneous resolution and undiagnosed cases) and poor-quality data. However, the good consistency between the number of cases collected in our cohort and the number of expected cases from the national database, together with comparable epidemiologic characteristics, makes selection bias unlikely. Furthermore, most of our patients were hospitalized in a limited number of centers of expertise, allowing homogeneity of practice and a high quality of collected medical data.

CVT in adolescents is a rare and complex disease with a wide spectrum of clinical presentations and its own epidemiology, differing from those in both adults and children. CVT-associated conditions and risk factors are numerous, often multiple, with important differences between sexes, and their distribution appears to be specific to this age group. Anticoagulation is the key to treatment, together with the specific management of any underlying associated pathology. Overall survival rates are excellent, but the risk of cognitive disability and impaired quality of life requires prolonged follow-up and special attention.

Glossary

AED

antiepileptic drug

CI

confidence interval

COC

combined oral contraception

CVT

cerebral venous thrombosis

ICD-10

International Classification of Diseases, 10th revision

IQR

interquartile range

mRS

modified Rankin Scale

OR

odds ratio

SLE

systemic lupus erythematosus

Appendix. Authors

Study Funding

The authors report no targeted funding.

Disclosure

The authors report no disclosures relevant to the manuscript. Go to Neurology.org/N for full disclosures.