Association of computed tomography-quantified low-attenuation plaque area with cerebral infarction beyond morphology-based atheroma grading in aortic arch repair: a retrospective cohort study.

Abstract

Objectives: To determine whether preoperative computed tomography quantification of low-attenuation plaque in the aortic arch improves perioperative cerebral infarction risk stratification beyond morphology-based atheroma grading.

Methods: This single-center retrospective cohort study included patients who underwent open or hybrid aortic arch repair with postoperative brain magnetic resonance imaging between January 2016 and December 2024. Preoperative computed tomography measured low-attenuation plaque area in the aortic arch (0 to 60 Hounsfield units) and arch atheroma grade. Low-attenuation plaque area was modeled as a continuous variable (per 10 mm2) in multivariable logistic regression with prespecified covariates, and discrimination was internally validated using 1,000 bootstrap resamples. Receiver operating characteristic and decision tree analyses were performed only as exploratory descriptive analyses; restricted cubic spline modeling was used to assess potential non-linearity.

Results: A total of 158 patients were analyzed. New cerebral infarction lesions occurred in 84 patients (53.2%), including 15 symptomatic strokes (9.5%). Low-attenuation plaque area was independently associated with new cerebral infarction lesions (adjusted odds ratio per 10 mm2, 1.69; 95% confidence interval, 1.34 to 2.22). Discrimination was acceptable (area under the curve, 0.82; 95% confidence interval, 0.75 to 0.88) with minimal optimism after bootstrap validation (optimism-corrected area under the curve, 0.79). In exploratory descriptive analyses stratified by atheroma grade and LAP burden, NCIL incidence was substantially higher in higher-LAP strata within atheroma grades III (78.6% vs. 25.0%) and IV (64.9% vs. 25.0%).

Conclusions: Computed tomography-quantified low-attenuation plaque area in the aortic arch is independently associated with perioperative magnetic resonance imaging-detected cerebral infarction and may provide complementary information beyond morphology-based atheroma grading, while exploratory stratified analyses suggested heterogeneity in embolic vulnerability within intermediate atheroma grades.