Abstract WP65: Clinical and Radiographic Characterization of Cerebral Hyperdensities on Post-thrombectomy Conventional Non-contrast Ct Brain: Differentiating Between Contrast Staining and Post-thrombectomy Hemorrhage

• Published in: Stroke (1970) Vol. 51; no. Suppl_1
• Main Authors: Kang, Hyung K, Ourfalian, Raffi, Nguyen, Emilie T, Torres, Fernando A, Feng, Lei T, Sangha, Navdeep S
• Format: Journal Article
• Language: English
• Published: 01-02-2020
• ISSN: 0039-2499; 1524-4628
• DOI: 10.1161/str.51.suppl_1.WP65

Abstract only
Background:
Post-interventional cerebral hyperdensities (PCHD) are present 31.2 to 87.5% of the time on post-thrombectomy (PT) CT. It can be difficult for radiologists to differentiate if PCHD represents intracerebral hemorrhage (ICH) or contrast staining using conventional CT. The ability to accurately determine the etiology of PCHD may be important for a patient’s outcome.
Methods:
We retrospectively investigated clinical risk factors, imaging findings, and interventional technique of patients who had a thrombectomy from 2011-2017 (n=238) at a Comprehensive Stroke Center. 112 patients with anterior circulation infarcts and immediate PT CTs as well as either a PT MRI or follow-up CT within 48 hours were included. Two experienced neuroradiologists interpreted all imaging. Baseline demographics and imaging characteristics were collected. The presence of ICH was determined by dephasing on gradient echo or by its persistence > 2 days on repeat CT. Chi-square and Fisher’s exact tests were used for statistical significance, and logistic regression to determine risk estimates.
Results:
84% of patients had PCHD, of which 56% had ICH. Patients had similar demographics (ICH: median age 69, 47.5% female; non-ICH: median age 71, 42.5% female). Higher pre-thrombectomy ASPECTS score (OR 0.62, p < 0.05) decreased the risk of ICH, and higher Hounsfield units (HU) of PCHD both on immediate (OR 1.04, p < 0.05) and intermediate (OR 1.1, p < 0.0005) PT CT increased the risk of ICH. PCHDs in the deep grey matter (OR 2.84, p < 0.05) and demonstrating a confluent pattern (OR 3.9 p < 0.05) increased the likelihood of ICH. 10% increase risk of ICH was seen for every HU increase in density on PT CT at 24 or 48 hours. A 50% decrease in ICH was observed for each unit increase of the pre-thrombectomy ASPECTS score. ROC revealed the optimal cut-off for predicted probability as 0.64, sensitivity 70.2% and specificity 69.2%. The AUC for the predicted model was 0.84 (95%CI= (0.75- 0.92, p <0.0001).
Conclusion:
Several imaging characteristics of PCHDs and ASPECTS score can help differentiate between the PT ICH and contrast staining. Application of these variables to patients in the immediate PT period may change clinical management.