Josser E. Delgado Almandoz

Pamela Schaefer

Dural venous sinus thrombosis as a cause of intracerebral hemorrhage (ICH) is often a difficult diagnosis to make due to nonspecific presenting symptoms and complex patterns of hemorrhage and edema on CT and MRI.

Prior to our study, our standard work-up for ICH identified on noncontrast CT was to obtain first-pass CTA followed by a gadolinium-enhanced MRI and susceptibility-weighted images if a vascular cause for the hemorrhage was not identified on the CTA. The first-pass CTA allowed detailed visualization of the major arteries of the circle of Willis, but only allowed variable opacification of venous structures depending on exact bolus timing and cardiac output. When we started using a 64-section CT, we noted that there was less frequent venous opacification than we had seen with a 16-section CT. Due to the lack of opacification of venous structures, we were systematically making errors. We failed to make the diagnosis of venous sinus thrombosis, and in some cases, patients underwent unnecessary MRI and angiography. For other patients, the diagnosis was missed until they had a repeated bleed and underwent additional imaging. In other cases, due to poor venous opacification, we suggested the diagnosis of venous sinus thrombosis when the sinuses were normal.

In our study, we sought to determine exactly how often the major dural sinuses were opacified with both our 64-section and 16-section CTA first-pass techniques, and how often an inadequately opacified major intracranial venous structure could have potentially explained the hemorrhage when an arterial cause was not identified. We found that adequate venous opacification of noncavernous intracranial venous structures was only seen in approximately 33% of patients who had first-pass CTAs with a 64-section technique compared with 60% of patients with a 16-section technique. We also found that an inadequately opacified major intracranial venous structure could have potentially explained the ICH in 27% of our 170 patients, most of whom were examined with a 64-section technique.

Our study fundamentally changed the way we image patients with intraparenchymal hemorrhage. Our current routine is to image patients with a first-pass CTA, followed immediately by delayed images to obtain a CT venogram (CTV) through the head. This affords a 90-second delay and allows satisfactory opacification of all the major venous sinuses. While this increases the radiation dose, we feel that the delayed imaging is warranted in order to expedite patient diagnosis and management and potentially improve patient outcome. In some cases, additional radiation from unnecessary angiography or repeat CTA is avoided, and unnecessary MRI is also avoided. We have received very positive feedback; today, first-pass CTA followed by CTV is the standard of care for the work-up of intraparenchymal hemorrhage in many emergency departments.

Read this article at AJNR.org …