Development and External Validation of Nomogram for Cerebral Infarction in Moyamoya Diseases

Abstract

More than 60% of moyamoya disease (MMD) patients suffers cerebral ischemia and preoperative cerebral infarction (CI) increases the risk of postoperative stroke and unfavorable outcome. We established a nomogram system for risk stratification of CI to help tailoring individualized management. We enrolled 380 patients including 680 hemispheres for the training cohort from our hospital and 183 patients including 348 hemispheres for the validation cohort from multicenter. A nomogram for CI was formulated based on the multivariable logistic regression analysis. The predictive accuracy and discriminative ability of nomogram were determined with concordance index (C-index) and calibration curve. For the training cohort, 246 hemispheres (36.2%) were found with CI. In multivariable logistic regression used generalized estimating equations approach, anterior choroidal artery (AchA) grade (grade 1, OR 0.214, 95%CI 0.124-0.372, P < 0.001; grade 2, OR 0.132, 95%CI 0.066-0.265, P < 0.001), cerebral perfusion (OR 4.796, 95%CI 2.922-7.872; P < 0.001), white matter hyperintensity (OR 3.652, 95%CI 1.933-6.902; P < 0.001), brush sign (OR 3.555, 95%CI 2.282-5.538; P < 0.001), and ivy sign (equivocal, OR 4.752, 95%CI 2.788-8.099, P < 0.001; present, OR 8.940, 95%CI 4.942-16.173, P < 0.001) were significant factors for CI. The C-index of the nomogram for predicting cerebral infarction was 0.890 (95%CI 0.866-0.915) in the training cohort and 0.847 (95%CI 0.805-0.889) in the validation cohort. The nomogram composed of AchA grade, cerebral perfusion, white matter hyperintensity, brush sign, and ivy sign could provide risk stratification of CI before surgery in patients with MMD. Active treatment might be recommended before CI, which could reduce the risk of stroke after surgery.

Keywords: Cerebral infarction; Moyamoya disease; Nomogram.