Neurodevelopmental Disorders and Array-Based Comparative Genomic Hybridization: Sensitivity and Specificity using a Criteria Checklist for Genetic Test Performance.

Abstract

Array-based comparative genomic hybridization (aCGH) is a molecular analysis method for identifying chromosomal anomalies or copy number variants (CNVs) correlating with clinical phenotypes. The aim of our study was to identify the most significant clinical variables associated with a positive outcome of aCGH analyses to develop a simple predictive clinical score. We conducted a cross-sectional study in a tertiary center comparing the genotype and phenotype of the cases. A score was developed using multivariate logistic regression. The best score cutoff point, sensitivity, specificity, positive and negative predictive values, and area under the curve were calculated with the receiver operating characteristic curve. aCGH identified structural chromosomal alterations responsible for the disorder in 13.7% (95% confidence interval [CI]: 10.9-16.5) of our sample (570 patients analyzed by aCGH). Based on the most frequent phenotypic characteristics among patients with a pathogenic CNV, we have created a checklist with the following items: alteration of the cranial perimeter, stature < percentile (p) 3, weight < p3, presence of brain malformations, ophthalmological malformations, two or more dysmorphic features in the same patient, and autism spectrum disorder diagnosis. Using a score ≥1.5 as the cutoff point for the test, we obtained a sensitivity of 82.4% (95% CI: 73.1-91.8) and a specificity of 54.2% (95% CI: 49.7-58.7). All individuals with a score of 1.5 or higher should be genetically screened by aCGH. This approach can improve clinical indications for aCGH in patients with neurodevelopmental disorders, but the scoring system should be validated in an external group.