Occult contrast retention post-thrombectomy on 24-h follow-up dual-energy CT: Associations and impact on imaging analysis

Abstract

Background Following reperfusion treatment in ischemic stroke, CT imaging at 24-hours is widely used to assess radiological outcomes. Even without visible hyperattenuation, occult angiographic contrast may persist in the brain and confound Hounsfield Unit-based imaging metrics such as Net Water Uptake (NWU). Aims We aimed to assess the presence and factors associated with retained contrast post-thrombectomy on 24-hour imaging using Dual Energy CT (DECT), and its impact on the accuracy of NWU as a measure of cerebral edema. Methods Consecutive patients with anterior circulation large vessel occlusion who had post-thrombectomy DECT performed 24-hours post-treatment from two thrombectomy stroke centers were retrospectively studied. NWU was calculated by interside comparison of Hounsfield Units of the infarct lesion and its mirror homolog. Retained contrast was quantified by the difference in NWU values with and without adjustment for iodine. Patients with visible hyperdensities from hemorrhagic transformation or visible contrast retention, and bilateral infarcts were excluded. Cerebral edema was measured by relative hemispheric volume (rHV) and midline shift (MLS). Results Of 125 patients analyzed (median age 71 [IQR 61-80], baseline NIHSS 16 [IQR 9.75-21]), reperfusion (defined as extended-Thrombolysis-In-Cerebral-Infarction 2b-3) was achieved in 113 patients (90.4%). Iodine-subtracted NWU was significantly higher than unadjusted NWU (17.1% vs 10.8%, p<0.001). In multivariable median regression analysis, increased age (p=0.024), number of passes (p=0.006), final infarct volume (p=0.023) and study site (p=0.021) were independently associated with amount of retained contrast. Iodine-subtracted NWU correlated with rHV (rho=0.154, p=0.043) and MLS (rho=0.165, p=0.033) but unadjusted NWU did not (rHV rho=-0.035, p=0.35; MLS rho=0.035, p=0.347). Conclusions Angiographic iodine contrast is retained in brain parenchyma 24-hours post-thrombectomy, even without visually obvious hyperdensities on CT, and significantly affects NWU measurements. Adjustment for retained iodine using DECT is required for accurate NWU measurements post-thrombectomy. Future quantitative studies analyzing CT after thrombectomy should consider occult contrast retention.

Novel advanced imaging techniques for cerebral oedema

“…Importantly, they found that NWU did not correlate with functional outcome at 90 days in the presence of hemorrhagic transformation or thrombectomy. These findings may be partially due to angiographic iodine retention, which may confound NWU measures which is dependent on Hounsfield-unit of the infarct tissue ( 32 ). Iodine extravasation may account for 27–84 percent of hyperdense lesions on CT post-thrombectomy at 24-h ( 53 , 54 ).…”

“…While no statistically significant difference were demonstrated in this small and likely underpowered cohort, notably 2 of 10 patients demonstrating more than 20% differences in NWU values between conventional CT and DECT, implying that the effect of iodine may affect NWU measures in a substantial proportion of patients. A subsequent larger multicenter study ( n = 125) aimed to provide a more definitive analysis of the confounding effects of post-thrombectomy iodine on follow-up CT demonstrated that contrast-adjusted NWU values were significantly higher than non-adjusted values, and correlated with measures of cerebral oedema, such as midline shift and rHV ( 32 ). They demonstrated that even without visible hyperdense lesions on follow up imaging at 24 h, angiographic iodine contrast is retained in brain parenchyma.…”

“…In contrast, there is a lack of correlation between these anatomical markers of cerebral oedema with NWU (10,31). Relative hemispheric volume is an effective measure of tissue volume and is shown to be superior to midline shift in predicting clinical outcome (10,(32)(33)(34). Currently, rHV has limited clinical application due to the…”

Does Reperfusion Benefit Patients Without Perfusion Mismatch?

“…Genuine inaccuracies in mismatch estimates exist but are often exaggerated by inaccuracies in assessment of the follow-up infarct volume reference standard due to early temporary diffusion lesion reversal 11 or the insensitivity of noncontrast computed tomography assessments that can be obscured by residual contrast staining. 15 How frequently these factors mean that patients assessed as having no mismatch do, in fact, have a clinically meaningful volume of salvageable penumbra and capacity to benefit from reperfusion is uncertain. The SELECT2 16 randomized trial (A Randomized Controlled Trial to Optimize Patient’s Selection for Endovascular Treatment in Acute Ischemic Stroke) found there was no statistically significant heterogeneity in EVT benefit between patients with and without mismatch, and the point estimates favored EVT, regardless of mismatch status, although very few patients had no mismatch using the mismatch ratio <1.2 and absolute mismatch <10 mL definition.…”