Phenotypic predictors of dual genetic diagnosis in people with autism
Clinical Genetics and Therapeutics
-
Primary Categories:
- Clinical Genetics
-
Secondary Categories:
- Clinical Genetics
Introduction:
Introduction:
A major goal of SPARK (Simons Foundation Powering Autism Research for Knowledge), a cohort with sequencing data on 58,247 individuals with autism, is to return new genetic diagnoses related to autism. Previous analyses have shown that SPARK identified at least one likely pathogenic/pathogenic (LP/P) variant in 8.6% of autistic individuals, but the phenotypic presentations of people with more than one LP/P variant has not been investigated. The objective of this study was to characterize the phenotype of 102 individuals with two LP/P variants related to autism to identify phenotypic features that distinguish individuals with dual diagnoses. We compared individuals with one and two genetic diagnoses (regardless of variant type). We also compared individuals with a CNV and SNV/indel to individuals with only a CNV.
Methods:
Methods:
Exome sequencing was performed on 55,478 individuals with autism and genome sequencing on 2,769 individuals with autism, as well as 94,392 first degree family members. We identified genetic variants (CNVs, SNV/indels, and aneuploidies) that meet ACMG LP/P criteria in 2,734 individuals with autism. Of these, 102 individuals have dual genetic diagnoses. When available, we analyzed Vineland-3 Adaptive Behavior scales (VABS) overall score (Adaptive Behavior Composite, or ABC), the Social Communication Questionnaire (SCQ) summary score, the Repetitive Behavior Scale – Revised (RBS-R), and other phenotypic information, including autism diagnosis age, language abilities, developmental milestones, support requirements, and medical, and mental health diagnoses (congenital anomalies, seizures). A multivariate logistic regression was performed in R to determine the predictive features of having an additional SNV diagnosis in people with a CNV. A second regression was performed to determine the predictive features of having a second finding, regardless of variant type.
Results:
Results:
Of those with dual diagnoses, 51% had a CNV and a SNV/indel, 34% had two SNV/indels, 12% had two CNVs, and the remaining had a chromosomal aneuploidy and another variant type. A goal was to identify factors that could help clinicians identify individuals in whom additional testing may be warranted. We first analyzed phenotypic differences between people with one versus two LP/P variants. The odds of participants with any variant (n = 1,263) having any second variant (n = 102) are two times higher if they have a congenital anomaly (p= 0.04, 95% CI [1, 4.3]). We then refined the analysis by variant type. When comparing individuals with a CNV to individuals with a CNV and SNV, the odds of having a second genetic diagnosis were 7 times higher in individuals with a congenital anomaly (p = 0.004, 95% CI [2.0, 50]). Lower language abilities increased the odds of having a SNV in addition to a CNV 6-fold compared to having a CNV alone (p = 0.009, 95% CI [1.5, 23]).
Conclusion:
Conclusions:
Our analyses suggest that individuals with congenital anomalies and decreased language abilities among individuals with autism have an increased likelihood of having a second genetic diagnosis. The presence of either/both characteristics in patients with autism suggests that those who only had a chromosomal microarray analysis despite a diagnostic result may benefit from additional genetic sequencing to assess a possible second diagnosis.
Introduction:
A major goal of SPARK (Simons Foundation Powering Autism Research for Knowledge), a cohort with sequencing data on 58,247 individuals with autism, is to return new genetic diagnoses related to autism. Previous analyses have shown that SPARK identified at least one likely pathogenic/pathogenic (LP/P) variant in 8.6% of autistic individuals, but the phenotypic presentations of people with more than one LP/P variant has not been investigated. The objective of this study was to characterize the phenotype of 102 individuals with two LP/P variants related to autism to identify phenotypic features that distinguish individuals with dual diagnoses. We compared individuals with one and two genetic diagnoses (regardless of variant type). We also compared individuals with a CNV and SNV/indel to individuals with only a CNV.
Methods:
Methods:
Exome sequencing was performed on 55,478 individuals with autism and genome sequencing on 2,769 individuals with autism, as well as 94,392 first degree family members. We identified genetic variants (CNVs, SNV/indels, and aneuploidies) that meet ACMG LP/P criteria in 2,734 individuals with autism. Of these, 102 individuals have dual genetic diagnoses. When available, we analyzed Vineland-3 Adaptive Behavior scales (VABS) overall score (Adaptive Behavior Composite, or ABC), the Social Communication Questionnaire (SCQ) summary score, the Repetitive Behavior Scale – Revised (RBS-R), and other phenotypic information, including autism diagnosis age, language abilities, developmental milestones, support requirements, and medical, and mental health diagnoses (congenital anomalies, seizures). A multivariate logistic regression was performed in R to determine the predictive features of having an additional SNV diagnosis in people with a CNV. A second regression was performed to determine the predictive features of having a second finding, regardless of variant type.
Results:
Results:
Of those with dual diagnoses, 51% had a CNV and a SNV/indel, 34% had two SNV/indels, 12% had two CNVs, and the remaining had a chromosomal aneuploidy and another variant type. A goal was to identify factors that could help clinicians identify individuals in whom additional testing may be warranted. We first analyzed phenotypic differences between people with one versus two LP/P variants. The odds of participants with any variant (n = 1,263) having any second variant (n = 102) are two times higher if they have a congenital anomaly (p= 0.04, 95% CI [1, 4.3]). We then refined the analysis by variant type. When comparing individuals with a CNV to individuals with a CNV and SNV, the odds of having a second genetic diagnosis were 7 times higher in individuals with a congenital anomaly (p = 0.004, 95% CI [2.0, 50]). Lower language abilities increased the odds of having a SNV in addition to a CNV 6-fold compared to having a CNV alone (p = 0.009, 95% CI [1.5, 23]).
Conclusion:
Conclusions:
Our analyses suggest that individuals with congenital anomalies and decreased language abilities among individuals with autism have an increased likelihood of having a second genetic diagnosis. The presence of either/both characteristics in patients with autism suggests that those who only had a chromosomal microarray analysis despite a diagnostic result may benefit from additional genetic sequencing to assess a possible second diagnosis.