1. Diagnosis and Initial Clinical Assessment of Symptomatic Cerebral Venous Thrombosis (CVT)

Cerebral venous thrombosis (CVT) is a rare but potentially life-threatening type of stroke, representing 0.5–1.0% of all stroke admissions (Bousser and Ferro 2007). The reported rates of CVT vary from 8.7 (Zhou et al. 2023) to 20.3 per million,(Otite et al. 2020) and appear to be increasing over time. The risk of CVT is higher in women, where it is often associated with pregnancy and the puerperium, and with oral contraceptive use (Amoozegar et al. 2015). Other non-genetic factors associated with an increased risk of CVT include antiphospholipid antibodies (7.0-fold increase), autoimmune diseases (5.6-fold increase), anemia (4.0-fold increase) and malignancy (3.2-fold increase) (Green et al. 2018). The most common genetic factors associated with CVT are prothrombotic conditions, such as the presence of factor V Leiden and protein C deficiency (Green et al. 2018). Diagnosis of CVT is frequently delayed in part because it can mimic other acute neurological conditions (Bakradze et al. 2023).

Individuals with lived experience emphasized the importance of recognizing the common presenting signs and symptoms of CVT, as their experience mostly differed from typical presentations of stroke. It is important to readily recognize CVT signs and symptom so that diagnosis is not delayed. Their experiences highlighted the challenge of CVT diagnosis; in some cases, many tests were performed prior to their diagnosis to rule out other causes. People with lived experience also emphasized the importance of imaging in the diagnosis of CVT.

To ensure that people who experience CVT receive timely stroke assessments, interventions and management, interdisciplinary teams need to have the education, infrastructure and resources required. These may include the following components established at a systems level. 

1. Government funding and support for awareness initiatives to improve the recognition and recall of the signs of CVT, which may present differently from other types of strokes, as well as more typical stroke presentations (e.g., FAST, which is a global best practice) and the importance of contacting 9-1-1 immediately. Awareness and education campaigns should prioritize reaching communities who are less aware of the signs of stroke and most at risk of stroke and should be informed collaboratively through community engagement activities with those audiences. 
2. Enhanced collaboration among community organizations and healthcare professionals to ensure consistency in public education of the signs of stroke with a strong emphasis on the urgency of responding when the signs of stroke are recognized. Specific information related to other forms of stroke including CVT should be included as part of awareness and education efforts.
3. Training and education for EMS, emergency department and all in-hospital staff, medical and nursing students, physicians in primary and acute care as well as specialists, nurses, and allied health professionals to increase their ability to recognize potential individuals with stroke, including CVT, and provide rapid assessment and management. 
4. Comprehensive systems in place to ensure all people in Canada have access to timely and appropriate emergency medical services, including ambulatory services (e.g., outpatient services, emergency department, community health centres, nursing stations) without financial burden, and quality stroke care regardless of geographic location. 
5. Enhanced monitoring and awareness of stroke among all people in Canada. Healthcare systems and provincial/territorial and federal governments should generate linked health and social surveillance population-based and regional data and use it to drive quality improvement through better understanding of the health and social issues facing people in Canada.

System Indicators: 

1. Numbers of individuals presenting to the emergency department ED who receive a diagnosis of CVT.
2. Number of individuals with CVT who are admitted to hospital annually.
3. Number of individuals admitted with an alternate primary diagnosis who are subsequently diagnosed with CVT during their hospital stay (e.g. head trauma with late onset CVT, neurosurgical procedure with post-op CVT, etc.). 
4. Number of individuals diagnosed with CVT who receive vascular imaging (CTV or MRV) at time of diagnosis.

Process Indicators: 

5. Timing from presentation to medical attention to diagnosis.
6. Timing from first presentation to medical attention and implementation of definitive therapy (eg. antithrombotic therapy, typically with full anticoagulation).
7. Number of presentations to medical attention prior to diagnosis.

Patient-oriented outcome and experience indicators: 

8. Mortality rate from CVT at 30 days and one year following diagnosis.
9. Changes in quality of life index measure at 30 days, one year and 5 years following diagnosis with CVT.

Measurement Notes

1. For indicator 9, standardized quality of life measurement tools should be used, and the same measure used over time.

Resources and tools listed below that are external to Heart & Stroke and the Canadian Stroke Best Practice Recommendations may be useful resources for stroke care. However, their inclusion is not an actual or implied endorsement by the Canadian Stroke Best Practices team or Heart & Stroke. The reader is encouraged to review these resources and tools critically and implement them into practice at their discretion.

Healthcare provider information

• CSBPR Cerebral Venous Thrombosis Module: Table 1 Common Clinical Features at the Time of Presentation with CVT
• CSBPR Cerebral Venous Thrombosis Module: Figure 1 Patient Characteristics, Risk Factors, and Conditions Associated with Cerebral Venous Thrombosis
• CSBPR Cerebral Venous Thrombosis Module: Appendix Three Recommended Laboratory Investigations for Individuals with Cerebral Venous Thrombosis
• CSBPR Cerebral Venous Thrombosis Module: Appendix Four Antiphospholipid Antibody Testing Flowsheet
• Heart & Stroke: Signs of stroke
• Heart & Stroke: FAST Signs of Stroke...what are the other signs? 
• Stroke Engine
• CVT Consortium

Information for individuals with lived experience of stroke, including family, friends and caregivers

• Heart & Stroke: Cerebral Venous Thrombosis Infographic
• CVT Consortium
• Heart & Stroke: Signs of stroke
• Heart & Stroke: FAST Signs of Stroke...what are the other signs? 
• Heart & Stroke: What is stroke? 
• Heart & Stroke: Your stroke journey
• Heart & Stroke: A family guide to pediatric stroke
• Heart & Stroke: Stroke in young adults: a resource for patients and families
• Heart & Stroke: Online and peer support
• Stroke Engine

Evidence Table and Reference List

Cerebral venous thrombosis is distinct from other stroke types. It is relatively uncommon within the general population, presenting symptoms can be gradual and non-focal, and it most commonly affects younger individuals, particularly women (Coutinho et al. 2012; Duman et al. 2017; Ferro et al. 2004; Yaghi et al. 2022b). This combination of factors makes it critical for front-line clinicians to be aware of the disease and its presenting symptoms and risk factors. In general, both younger adults and women with stroke are at increased risk for initial misdiagnosis and/or diagnostic delay, which is common in CVT (Newman-Toker et al. 2014; Yu et al. 2021). With respect to CVT specifically, one retrospective study found that of 53 individuals with CVT, 20.8% had experienced an initial error in diagnosis (Liberman et al. 2019). Another retrospective study using a combination of administrative claims data from the United States and single-centre chart review found that of 5966 individuals with a diagnosis of CVT, 3.6% had been seen in the emergency room in the 14 days prior to diagnosis and discharged with a diagnosis of headache or seizure. Those who were seen and sent home from the emergency department were younger (mean age 38.5 years) than in the remainder of the cohort (44.4 years) (Liberman et al. 2018). Non-focal presentations, such as isolated symptoms and signs of intracranial hypertension, are associated with longer times (>10 days) from symptom onset to diagnosis (Bakradze et al. 2023; Ferro et al. 2009).

Incidence of CVT is approximately 10-20 per million in the general population, (Zhou et al. 2023) although the incidence of CVT in association with pregnancy, approximately 9/100,000, is similar to that of pregnancy-associated ischemic stroke and intracerebral hemorrhage (Swartz et al. 2017). Many population-based series have reported increased incidence of CVT over time (Devasagayam et al. 2016; Otite et al. 2020; Zhou et al. 2023). A recent US-based health services study found increasing rates between 2005 and 2017 in men and older individuals, with stable rates in younger women. There were also increased rates over time of comorbidity codes for malignancy, trauma and inflammatory disease alongside those for CVT (Otite et al. 2020). Thus, increased rates might be due to improved overall ascertainment with more frequent use of vascular neuroimaging, better survival of medically complex individuals who go on to develop CVT, and/or better ascertainment in medically complex individuals.

Sex-specific risk factors for CVT are discussed in detail below (See Sex, gender and other equity-related considerations). Risk factors are summarized in Figure 1 and have been explored in detail in a recent meta-analysis of genetic and non-genetic risk factors (Green et al. 2018). A recent large prospective cohort study found that adults with an identified risk factor had an earlier age of onset of CVT than those without (Ranjan et al. 2023). Malignancy in particular was associated with older age of onset of CVT.

The limited available literature examining cancer types associated with CVT specifically include a high representation of some hematologic malignancies (specifically acute lymphoblastic leukemia (ALL), Janus-Kinase-2 (V617F) mutation-associated myeloproliferative neoplasms and Waldenstrom’s macroglobulinemia) and some solid organ cancers, including breast, gastrointestinal, lung and CNS cancers (Silvis et al. 2018). Some therapies, such as L-asparaginase for ALL, and steroids, are also known risk factors for CVT. Head and neck infection is a well-established risk factor for CVT; COVID-19 infection has been associated with increased risk of CVT in both community- and hospital-based series (McCullough-Hicks et al. 2022; Taquet et al. 2021). CVT, with and without other venous and arterial thromboembolic events, was a common presentation of Vaccine-induced thrombosis with thrombocytopenia (VITT), a rare (1/26500 to 1/1273000) autoimmune reaction to non-replicant adenovirus vector-based COVID-19 vaccines (ChAdOx1, AstraZeneca/COVISHIELD and Ad26COV2.S, Janssen, Johnson & Johnson) characterized by anti-platelet factor-4 antibodies (Klok et al. 2022).

Presenting symptoms of CVT may also differ from those of arterial strokes. The onset of symptoms is generally more insidious. Overall, in multiple recent large series, less than half of patients present within 48 hours of symptom onset, although more acute presentations can occur with thunderclap headache or stroke-like sudden focal symptom onset in addition to seizures (Duman et al. 2017; Lindgren et al. 2022; Yaghi et al. 2022b). Symptoms may result from increased intracranial pressure, focal parenchymal injury, and/or mass effect. Headache is the most common symptom, reported in approximately 90%, although it may be a less common presenting features in older individuals presenting with CVT (Coutinho et al. 2015). The exact prevalence varies depending on cohort, although the other most common presenting symptoms include focal deficits, seizures, vision loss, encephalopathy or depressed level of consciousness or cranial neuropathies. (Table 1) Headache types, onset patterns and locations at presentation are variable. One series of 200 consecutive patients with CVT found that headache at presentation was not associated with neuroimaging evidence of hemorrhage or hydrocephalus. There was no association between lateralization of pain and site of thrombosis, and none between thrombus location and headache apart from occipital and neck pain association with transverse and/or sigmoid involvement (Wasay et al. 2010). Characteristic symptoms of headaches due to increased intracranial pressure (ICP), which is associated with papilledema on fundoscopic examination, may include supine or nocturnal headache, associated nausea and/or vomiting, and blurred vision, transient visual obscurations, and/or diplopia. 

Several small studies and meta-analyses have examined diagnostic imaging modalities for CVT. Large high-quality studies comparing diagnostic imaging modalities for CVT, particularly current ones comparing contemporary contrast-enhanced CT venography (CTV) and/or contrast-enhanced MRI venography (MRV) against gold-standard digital subtraction angiography, are lacking. A 2020 critical review of English and Dutch neuroimaging studies examining performance of CT/CT venography and MRI for the diagnosis of CVT concluded that studies were observational, mostly small, outdated and with a high risk of bias (van Dam et al. 2020). The accuracy of parenchymal CT and MRI in the differential diagnosis of cerebral venous thrombosis and cerebral venous sinus thrombosis (i.e. CVT with sinus involvement only) was examined in a systematic review using any of MR venography, CT venography, or digital subtraction angiography (DSA) as the standard reference (Xu et al. 2018). Among 2,822 cases, the pooled sensitivity and specificity for the identification of CVT using CT was 0.79 (95% CI 0.76- 0.82), and 0.90 (95% CI 0.89- 0.91), respectively. The corresponding values for the use of MRI were 0.82 (95% CI 0.78- 0.85) and 0.92 (95% CI 0.91-0.94) (Xu et al. 2018). The 2020 critical review found that, using DSA as the reference standard, small observational studies comparing CT venography to DSA have reported sensitivity and specificity of both 100% (van Dam et al. 2020). Other small studies using other imaging modalities and final clinical outcome as reference standard have demonstrated a sensitivity of 100% (95% CI 88-100%) and specificity of 100% (95-100%) for cases of sinus thrombosis, but lower sensitivity for cortical vein thrombosis. Non-contrast-enhanced time-of-flight (TOF) MR venography, compared with digital subtraction angiography also was not sensitive in the assessment of small veins but accurate for larger veins and sinuses. When compared against contrast-enhanced MRI, TOF MRV and non-contrast phase contrast (PC) MRV had a sensitivity of 64-100% and 48-100%, respectively, with wide confidence intervals, and lower accuracy for identifying cortical vein thrombosis (van Dam et al. 2020). Studies comparing contrast-enhanced MRI to DSA reported sensitivities of 86-97% and specificities of 55-97% for diagnosis of CVT. MRI with gradient-echo (GRE) or susceptibility-weighted imaging (SWI) had the most consistently reported adequate sensitivity and specificity for cortical vein thrombosis (97-98% and 100%, respectively) (Altinkaya et al. 2015; Idbaih et al. 2006; Linn et al. 2010).

In summary, larger and contemporary higher-quality studies are needed. Both contrast-enhanced CTV and MRV are acceptable modalities, although additional imaging with MRI, including gradient-echo MRI, may need to be considered if isolated cortical vein thrombosis is suspected. CT venography may be more quickly performed and more easily accessible, and with fewer contraindications to MRI, while MR venography does not expose patients to ionizing radiation. However, MR contrast should be avoided in pregnancy and some patients may have additional contraindications to MR imaging, such as pacemaker or retained ferromagnetic material. Although some groups have reported alternative native-contrast thrombus imaging MRI sequences, such as black-blood thrombus imaging, with high reported diagnostic accuracy for CVT, protocols tend to be site-specific and can be lengthy (Yang et al. 2016). Given these limitations, native-contrast sequences are not recommended for routine use at this time. There is no extensive literature comparing neuroimaging approaches (e.g. time-of-flight versus contrast-enhanced MR venography) to assess venous recanalization in individuals with known CVT. (Section 3.2).

D-dimer has been explored as a screening tool to decide who should have neuroimaging to exclude CVT where index of suspicion for CVT might be lower, such as isolated headache. A 2012 systematic review and meta-analysis including 14 studies with 1,134 individuals evaluated for suspected CVT, 363 had a confirmed diagnosis. The weighted mean sensitivity for elevated D-dimer in those with confirmed CVT was 89.1% (95% CI 84.8–92.8). Sensitivities varied and were lower in those with longer duration of symptoms, isolated headache and thrombosis of a single venous sinus. The pooled positive and negative likelihood ratios were 9.1 (95% CI, 6.8–12.2) and 0.07 (95% CI, 0–0.14), respectively (Dentali et al. 2012b). Another small retrospective study found that presentations with focal neurologic deficits were associated with higher D-dimer levels at baseline (Juli et al. 2020). A more recent prospective study of 359 individuals with suspected CVT, 94 of whom had a subsequent confirmed diagnosis, found the sensitivity and specificity of a D-dimer cut-off of 500 μg/L or above was 89.4% and 66.4%, respectively (Heldner et al. 2020). Thus, given that D-dimer alone cannot reliably identify almost all individuals with the disease, particularly those who may have a less classic presentation for CVT, D-dimer is not recommended at this time as a screening tool. Further studies will determine whether D-dimer in combination with other diagnostic modalities, such as non-contrast CT, might be a suitable approach in lower-resource environments.

Sex, gender and other equity-related considerations

CVT is more common in females, although the greatest disparities in incidence between males and females are in patients aged 50 years and under (Zhou et al. 2023). The most common risk factors for CVT include oral contraceptive use, pregnancy and the puerperium and hormone replacement therapy (Green et al. 2018; Silvis et al. 2016). One large prospective cohort study found that women experienced CVT 9 years earlier than men on average, and that gender-specific risk factors (pregnancy, puerperium or oral contraceptive use) were associated with earlier age of onset in women (Ranjan et al. 2023). Oral contraceptive use may increase the risk of CVT up to 8-fold, and there may possibly be a synergistic risk between oral contraceptive use and obesity (Amoozegar et al. 2015; Zuurbier et al. 2016). Although the general stroke (Bushnell et al. 2014)  and venous thromboembolism (Dragoman et al. 2018; Oedingen et al. 2018) literature has compared formulations of hormonal contraceptives and has identified consistent increased risk with estrogen-containing, but not progesterone-only formulations, this has not been examined in detail in CVT specifically. The risk of pregnancy associated CVT is highest in the first six weeks post-partum (Silvis et al. 2019).

Women are at higher risk than men for misdiagnosis of stroke, and the risk may be even higher with CVT, as individuals may be younger (another risk factor for misdiagnosis) and the indolent and often non-focal onset of symptoms may be mistaken for a benign cause of headache, such as migraine, which affects approximately 20% of women (Frederick et al. 2014; Stewart et al. 1995).

There is a dearth of information related to considerations of race-ethnicity and social determinants of health in CVT. One study using US administrative data found that Black individuals had the highest risk of CVT, followed by white and Asian individuals (Otite et al. 2020). This mirrors population-based data related to peripheral VTE (i.e. DVT and PE) (Goldhaber 2014). Another recent study reported that Black race-ethnicity was associated with worse outcomes after CVT, although contributors related to health inequities, structural racism and social determinants of health as opposed to any genetic or other biological factors is not known (Goldhaber 2014; Klein et al. 2022). Data from lower- and middle-income countries, particularly from Africa and South America, is under-represented, and to date, the only genome-wide association study data is from European patients (Baduro and Ferro 2021; Ken-Dror et al. 2021; Zhou et al. 2023).