A new method for quantification of false lumen thrombosis in aortic dissection using magnetic resonance imaging and a blood pool contrast agent

Abstract

Background: False lumen thrombosis after aortic dissection is a major predictor of prognosis. First pass computed tomography (CT) and magnetic resonance (MR) imaging are used routinely, where the image acquisition is timed to the arrival of contrast in the proximal unaffected aorta. Delayed phase imaging is difficult to refine because flow rates in the false lumen are often very slow and highly variable between patients. Blood pool contrast agents bind to albumin and become homogenously distributed in the intravascular circulation, allowing accurate imaging of areas where flow is low. We compared first pass MR and CT with a delayed phase MR acquisition using a blood pool agent to assess whether this more accurately quantified false lumen thrombosis.

Methods: Patients with medically treated chronic type B aortic dissection and evidence of false lumen thrombosis on previous CT imaging underwent first pass CT, first pass MR, and delayed phase MR with blood pool agent. Absence of false lumen contrast enhancement was quantified to assess the apparent differences in thrombosis. Phase-contrast MR data were also obtained to assess the affect of flow velocity on false lumen contrast enhancement, and direct thrombus MR images were used to confirm the presence of thrombus.

Results: Twelve patients were recruited. No difference was seen in apparent thrombus volume between first pass CT and first pass MR imaging (146.9 cm(3) [SD, 88.6] vs 187.6 cm(3) [SD, 136.1], P = .1119; R(2) = .67 [95% confidence interval (CI), r = .46-.95], P = .0012). In all patients, the volume of thrombus derived from first pass acquisitions was greater than the volume derived from delayed phase MR imaging with blood pool agent: first pass CT (paired t test, P = .0007; mean difference = 83.4 cm(3) [95% CI, 44.1-122.6]) and first pass MR (paired t test, P = .0009; mean difference = 124.0 cm(3) [95% CI, 63.2-184.9]). The difference in thrombus volume between first pass and delayed phase MR imaging with blood pool agent correlated significantly with the mean velocity of flow in the false lumen, with lower flow related to a greater difference (R(2) = .61, P = .0028 [95% CI, r = -.94--.37]). Direct thrombus MR images were able to correctly discriminate between thrombus and blood and matched the area of contrast absence on delayed phase MR with blood pool agent images.

Conclusion: First pass techniques to assess false lumen thrombosis in aortic dissection consistently overestimate the apparent thrombus volume by five to six times. This has implications for interpretation of cohort studies and clinical trials that use false lumen thrombosis as an outcome measure. We recommend delayed phase MR imaging with a blood pool agent when accurate assessment of false lumen thrombosis is required.