Routine Clinical Genetic Testing in Patients with Autism Spectrum Disorder and Related Developmental Disorders Has a Diagnostic Yield up to 40%, Improves Medical Management, and Significantly Alters Recurrence Risk Counseling

Background: Genetic testing, via chromosomal microarray (CMA) for copy number variants (CNV) and whole exome sequencing (WES) for single gene sequence-level variants, has proven highly useful for identifying genetic etiologies of autism spectrum disorder (ASD) and other developmental brain disorders (DBD), such as intellectual disability and global developmental delay (ID/GDD). However, fewer than a third of U.S. children with DBD currently undergo recommended genetic testing (Kiely et al., 2016). In fact, a survey of child neurologists and developmental pediatricians found that 40% did not order genetic testing for children with ASD (Tchaconas et al., 2017).

Objectives: As part of our multidisciplinary autism and developmental medicine clinic, we have incorporated a genetics-first approach to routine care for all patients with ASD, ID/GDD, epilepsy, and/or congenital anomalies, including dysmorphic features. We report the diagnostic yield of clinical genetic testing in this patient population and the impact on medical management.

Methods: Patients were typically offered a testing cascade, starting with FMR1 analysis for fragile X followed by CMA and WES. Single gene tests/panels were offered alternatively if there was a strong clinical suspicion for a particular disorder. We compared the diagnostic yield in patients with ASD only, ASD+ID/GDD, and ID/GDD without ASD.

Results: We used genetic cascade testing to evaluate 764 patients and identified a genetic diagnosis to explain their DBD in 159 (21%): 2% had fragile X syndrome, 14% had a pathogenic CNV, and 24% had a pathogenic single gene sequence variant(s). Since not all patients have completed WES testing yet, we used our initial dataset to estimate a combined diagnostic yield of up to ~40% if WES had been completed for all eligible patients. In the 159 patients with a genetic etiology, 15 (9%) had ASD only, 31 (19%) had ASD+ID/GDD, and 80 (50%) had ID/GDD without ASD, with the remaining including other testing indications. Identifying a genetic etiology has had important implications for medical management. For example, identifying a patient’s 17q12 deletion informed the need to monitor for renal cysts/dysplasia and maturity-onset diabetes of the young. In addition, knowing a genetic etiology can significantly alter genetic counseling by providing more accurate recurrence risks. We identified a de novo pathogenic sequence variant in SCN2A which provided the family with a medical explanation for their child’s ASD+ID diagnoses and, importantly, lowered their recurrence risk estimate from the ~10% quoted risk with one affected child down to the background population risk of ~1.5%. In another family, a maternally inherited 15q13.3 deletion was identified, resulting in a recurrence risk of 50%.

Conclusions: This genotype-first testing approach in a pediatric developmental medicine cohort illustrates the diagnostic power of genomic testing, as we can now provide a genetic etiology in up to 40% of patients, including a significant proportion who only had an ASD diagnosis. Given recent advances in WES, a single test can now be used to identify single gene variants and CNVs, providing an efficient testing mechanism for use in routine clinical care to provide vital information for medical management and family counseling.