Cerebral venous sinus thrombosis

Central venous sinus thrombosis (CVT) and venous cerebral infarction are important differentials for the causes of headache or seizure in the young person, particularly in the young woman who is/has recently been pregnant. It is probably the most rare form of stroke, but - when it happens - the patient usually will end up in ICU, and the critical care trainee will probably see more of this than any other specialty.

This has come up in Question 26 from the second paper of 2015. Dural sinus thrombosis in that question was associated with pregnancy (as typically they are) and the question could easily be shoved into the O&G section, but the specific information requested from the candidates ("outline the management priorities", etc) was more neurological/neurosurgical in nature.

If one needed to limit one's reading to a single article, that sole article should probably be the AHA/ASA CVT management algorithm (2011).

Central venous sinus thrombosis (CVT) and venous cerebral infarction are important differentials for the causes of headache or seizure in the young person, particularly in the young woman who is/has recently been pregnant. It is probably the most rare form of stroke, but - when it happens - the patient usually will end up in ICU, and the critical care trainee will probably see more of this than any other specialty.

This has come up in Question 26 from the second paper of 2015. Dural sinus thrombosis in that question was associated with pregnancy (as typically they are) and the question could easily be shoved into the O&G section, but the specific information requested from the candidates ("outline the management priorities", etc) was more neurological/neurosurgical in nature.

If one needed to limit one's reading to a single article, that sole article should probably be the AHA/ASA CVT management algorithm (2011).

Risk factors for cerebral venous thrombosis

The following lisk of risk factors was compiled using UpToDate and Ferro et al (2004). The latter was a multinational multicentre observational study which identified 624 patients with confirmed CVT. Of this group, the following risk factors were observed:

• Genetic thrombophilia (22.4% of cases)
• Pregnancy during peripartum period or puerperium (13.8%)
• Infection, be it CNS or ENT (12.3%)
• Drugs (7.5%)
• Malignancy (7.4%)
• Pregnancy at any stage (6.3%)
• Antiphospholipid syndrome (5.9%)
• Trauma, neurosrugery or lumbar puncture (4.5%)
• Vasculitis (3%)
• Polycythaemia or thrombocytosis (2.8%)
• Surgery (2.7%)
• Vascular anomaly in the CNS like dural fistula, AVM, etc (1.9%)
• Dehydration (1.9%)

If all the possible risk factors were piled into one big table, it would look like this:

Local inflammation or injury: Meningitis Sinusitis Otitis media Neck infections, eg. Lemierre's syndrome Head injury, especially base of skull fracture Lumbar puncture Neurosurgery Dural fistula Internal jugular CVC-associated thrombosis Drugs and diet: Oral contraceptive Hormone replacement Androgens Asparaginase Tamoxifen Corticosteroids

Acquired procoagulant states: HONK Dehydration of any other aetiology Pregnancy Polycythaemia Malignancy Nephrotic syndrome Post-splenectomy thrombocytosis Autoimmune disease: Sarcoidosis SLE (antiphospholipid) Wegener's granulomatosis Behçet disease Thromboangiitis obliterans Inflammatory bowel disease Paroxysmal nocturnal haemoglobinuria Congenital procoagulant states: Protein C, S deficiency Antithrombin gene mutation Factor V Leiden mutation Hyperhomocysteinaemia

Clues from history and examination

Ferro et al (2004) found the following common presentation features:

• Headache was present in 88.8%
  • Gradual onset, usually
  • Worse with Valsalva and recumbency
  • Vidsual symptoms may occur, resembling migraine aura
• Seizures in 39.3%
• Focal neurological signs (eg. hemiparesis) in 37.2%
• Decreased level of consciousness was seen in 13.9%
• Visual loss in 13.2%

So, highly nonspecific. That young person with seizures and headache could just as easily have had an SAH, be full of cocaine, have a brain tumour, abscess, or any damn thing. Specific features are sadly lacking.

Specific diagnostic imaging

Imaging may include:

• MRI venogram (gold standard)
• CT venogram (used to be the gold standard)
• Non-contrast CT (often pointless)

A contrast CT venogram may reveal a characteristic feature known as "dense triangle sign" or "empty delta sign" A picture tells a thousand words:

The arrow points at  "a triangular area of enhancement or high attenuation with a relatively low-attenuating center on multiple contiguous transverse CT images obtained in the region of the superior sagittal sinus" (the picture is shamelessly stolen from the same article by Lee, 2002). Another variant of this is the "Cord sign", essentially the same "empty" appearance in a deeper cortical vein. You really need to do a venogram to see these, as they are apparently only spotted on 25% of non-contrast CTs of known CVT. This often-quoted statistic is probably wrong, because it comes from a 1987 study (Virapongse et al)- back in the day when one slice of CT data took three weeks to reconstruct on the Commodore Amiga.

Other CT features (on a non-contrast CT) may include the following:

• Haemorrhagic infarction
• Cerebral oedema
• Mass effect, herniation
• Hydrocephalus
• Subdural effusion (isodense with CSF)

Unenhanced MRI is more sensitive in picking up CVT, and is recommended as the modality of choice by the AHA/ASA CVT management algorithm (2011).

Management

The major approaches to fixing this problem are anticoagulation, thrombolysis, endovascular thrombectomy and open clot retrieval.

• Anticoagulation has been the mainstay of treatment. Coutinho et al (2012) performed a systematic review of the practice and found it to be very helpful, with risk of death reduced by two thirds (ARR of 13%) and substantially improved neurological outcomes. As the patients usually have some sort of procoagulant state, you're going to end up anticoagulating them anyway.
• Thrombolysis for CVT is a logical extension of the use of thrombolysis for ischaemic stroke (i.e. "if its good enough for arterial clots, why not try it with venous clots"). This is the topic of the TO-ACT trial (Coutinho et al, 2013) which appears to still be recruiting. Apart from this undertaking, evidence for the use of thrombolysis in CVT consists of case series only (Ciccone et al, 2004, and Canhão et al, 2003, with about 200 patients in total). Discouraging features included a high rate of extracranial (21%) and intracranial (17%) haemorrhage, which occured in spite of the fact that in the majority of cases (88%) the thrombolytic agent was administered directly into the thrombosed sinus. 
• Endovascular clot retrieval is typically reserved for situations where there is significant thrombus burden, contraindication to anticoagulation or a lack of clinical improvement or worsening of symptoms despite systemic heparin. It is the new exciting thing on this scene, and also only known from case series.  As a representative example, Mokin et al (2015) report on the American experience. Thirteen patients were identified, of whom five had a favorable clinical outcome (defined as modified Rankin Scale score of 0–2) and three died.  A systematic review of such case series (Siddiqui et al, 2015) found a good outcome in 84% and a mortality of 12% among a total of 185 patients, even though they represented the more severe end of the CVT spectrum (60% were in a coma).
• Open clot retrieval seems to be an end-of-the-line manoeuvre. It is known only from heroic case reports (eg. Persson et al, 1990 and Ekseth et al, 1998). In each case, the surgeons followed thrombectomy by an infusion of a thrombolytic agent.
• Decompressive craniectomy for this condition is discussed elsewhere. In brief, the results have been encouraging and full systemic anticoagulation within 24 hours of surgery seems to be safe. A 2009 review from Stroke courageously asserted that the therapy is "lifesaving" and that the outcome can be "excellent", on the basis of three cases. The AHA/ASA gudelines recommend this as a viable option.