I'm Francisco Milán Zamora. I'm an Associate Director of the Neurogenetics Program at GeneDx, and I'm excited to have this opportunity to discuss with you how genomic advances can help in the etiologic diagnosis and management of cerebral palsy. In this talk, we will review recent literature supporting the use of genetic testing in patients with cerebral palsy, examine the results of exome sequencing in cohorts of individuals with cerebral palsy, and consider the implications of establishing a genetic diagnosis, including discussing recurrence risk for families. So, let's get started. Because we have people in attendance from diverse genetic subspecialties, I wanted to briefly review some concepts, starting with what is currently called cerebral palsy. According to the formal definition, delineated by an international panel in the mid-2000s, cerebral palsy describes a group of permanent disorders of the development of movement and posture, causing activity limitation, that are attributed to non-progressive disturbances that occur in the developing fetal or infant brain. As you can see, this term refers not to a specific disease entity, but rather to a group of conditions with variable severity that has certain motor developmental features in common. In fact, this report on the definition and diagnosis of cerebral palsy states that cerebral palsy is not an etiologic diagnosis, but a clinical descriptive term. Cerebral palsy is a common clinical diagnosis, with an incidence of 1 in 500 births, and the most common cause of childhood-onset, lifelong physical disability in most countries. Although the motor disorder is the primary component, associated features such as intellectual disability, autism spectrum disorder, epilepsy, visual and hearing impairments, and secondary musculoskeletal problems are frequently noted. Regarding its causes, the etiology of cerebral palsy has been attributed to a variety of factors, such as prematurity, stroke, or hypoxic ischemic encephalopathy. However, the specific causal mechanism remains unknown in many cases. Historically, it has been attributed to pre- and perinatal complications, in particular birth asphyxia. However, large population-based studies have shown that birth asphyxia accounts for less than 10% of cerebral palsy cases. Still referring to the etiology of CP, there is emerging evidence suggesting that a significant proportion of cerebral palsy may be due to genetic causes. Recent studies performed in cohorts of patients with CP using chromosomal microarray or exome sequencing have found a diagnostic variant in 10 to 31% of patients using CMA, and in 14 to 53% of patients when using exome sequencing. Additional evidence comes from various case reports and theories that have identified a single gene abnormality in individuals with cerebral palsy, such as the examples provided here. So based on the exciting emerging literature, in 2018, we proceeded to examine our internal exome database to determine the diagnostic yield of exome sequencing in individuals with cerebral palsy. We partnered with our collaborators at Geisinger, who had examined the DISCOVER project database with the same objective, and we were able to delineate the largest cohort of individuals with cerebral palsy published to date that have undergone exome sequencing, accumulating 1,526 patients with a combined molecular diagnostic yield of 30.1%. If you're interested, this manuscript was published in JAMA in February this year, and I'm going to come back to this article to provide additional details of the analysis performed, but first, I wanted to mention that in 2004, the American Academy of Neurology, along with the Child Neurology Society, published guidelines for the diagnostic assessment of children with cerebral palsy, suggesting that genetic testing should not be routinely obtained. However, many of the newer testing techniques were not available at that time. These guidelines also recommended that future studies should investigate the utility of genetic testing for these patients, so we're hopeful that these guidelines will be updated in the future based on data from our and other studies. Now that you have reviewed some concepts related to cerebral palsy, I'd like to discuss a case study. This individual underwent testing at GeneTX. He's an 8-year-old boy with a history of cerebral palsy, primarily affecting the muscles of the upper extremities, face, and neck. He also had a history of central hypotonia with limb hypertonia in infancy. Global developmental delay and hearing loss were also reported. Behavioral concerns, including ADHD and frequent meltdowns, were noted. And finally, his subtle dysmorphic features were described as irregular eyebrows, a broad mouth, and abnormal philtrum. He's a middle child of non-consanguineous parents with no family history of genetic or neurological disease. His delivery was unremarkable at term, and although prior genetic testing had not been performed, subsequent brain MRI was unremarkable. This child underwent exome sequencing as a trio with both parents. As depicted in the images, the results revealed a de novo nonsense variant in the anchoring D11 gene. Based on the ACM GIMP joint guidelines for variant classification, this variant is classified as pathogenic. Supporting evidence of that classification includes that it is a nonsense variant predicted to result in protein truncation or nonsense mediated decay in a gene for which loss of function is a mechanism of disease. It was found to be de novo with confirmed parentage in multiple patients tested at GNDX, and it was not observed in large population cohorts of unaffected individuals. Moving on to the phenotype, pathogenic variants in the anchoring D11 gene cause KBG syndrome. This condition is characterized by a number of neurologic issues, including developmental delay, intellectual disability, and seizures. Hearing loss and skeletal abnormalities are also reported. And there is a distinct craniofacial appearance as illustrated in these photos taken from gene reviews. You can appreciate the triangular face, brachycephaly, synophrys with full eyebrows, widely spaced eyes, and macrodontia. Looking back to the patients in our case study, the delays, hearing loss, behavioral concerns, and subtle dysmorphic features appear consistent with the diagnosis. So this raises an interesting question. Is it still cerebral palsy despite the genetic diagnosis? There has always been discussion in genetics and medical community as to whether a clinical diagnosis of cerebral palsy should be revised given genetic findings, and whether they are more appropriately referred to as CP mimics or masquerades rather than just cerebral palsy. This discussion was also brought up within the International Cerebral Palsy Genomics Consortium, which brings together world experts on this topic. Because cerebral palsy is a neurodevelopmental disorder diagnosed based on clinical science, not etiology, the clear consensus was that the clinical diagnosis of cerebral palsy should not change despite the identification of a genetic or non-genetic cause if the person exhibits a non-progressive permanent disorder of movement and posture. Instead, the clinical diagnosis of cerebral palsy should remain, should prompt appropriate genetic studies, and can subsequently be classified by etiology. In essence, cerebral palsy is a diagnosis, a genetic finding is an etiology. So while cerebral palsy is not a common feature of KBG syndrome, it has been previously reported. At least two other patients in the published literature carry a diagnosis of cerebral palsy with the underlying etiology of KBG syndrome. The motor limitations diagnosed as cerebral palsy remain even when the underlying etiology of KBG syndrome is established. Fernandez et al. reported a patient with cerebral palsy characterized by generalized muscular hypertonia, multiple contractures, and hip dysplasia, among other features. This individual had a pathogenic frameshift variant resulting in a premature stop cotton. Novara et al. reported on a patient who had a diagnosis of cerebral palsy at a young age, as well as a number of other findings. This individual had a de novo 16q24 microdeletion that includes the entire anchoring D11 gene. I want to briefly bring to your attention this study that revealed variability in diagnosing cerebral palsy in current clinical practice. These studies surveyed physicians who treat people with cerebral palsy across different specialties. It was based on the 2007 consensus definition and presented four hypothetical scenarios. One of these scenarios described a child with a permanent and non-progressive disorder of the development of movement and posture with associated pyramidal features in the lower extremities. The family history is positive for a similarly affected older brother. And genetic testing revealed biallelic pathogenic variants in ADD3 in this individual, a gene known to be associated with non-progressive spasticity and intellectual disability. Only 67% of the participants would diagnose cerebral palsy in this scenario in spite of clinical criteria being met, revealing that there is still work to do in raising awareness about this concept within the medical community and a need for training on genetic etiologies of CP and their implications for clinical practice in order to avoid revisionist diagnosis for children who fit clinical criteria for CP. Okay, so as promised before, I will now review with you some aspects of the study we published with our collaborators. As you can see in this diagram, we evaluated two independent cohorts of patients with cerebral palsy. On the left side of the diagram is the clinical laboratory referral cohort from GeneDx, encompassing 1,345 individuals with cerebral palsy that underwent routine clinical exome sequencing analysis and who were predominantly pediatric patients. And on the right side of the diagram is a healthcare-based cohort consisting of 181 individuals with cerebral palsy that underwent research-based exome sequencing. These patients were recruited from the DISCOVER project, a collaboration between Geisinger and Regeneron, and they were predominantly adult patients. Variants identified in both cohorts were interpreted based on the ACMG-AMP joint guidelines for sequence variance interpretation. The molecular diagnostic yield was the prevalence of pathogenic and likely pathogenic variants, and based on the diagnostic variants identified, genomic variation was analyzed across both cohorts. In this table, I'm showing demographic and comorbidity information for both cohorts. We were able to identify 1,556 individuals with cerebral palsy overall. As mentioned earlier, the lab-based cohort consisted predominantly of pediatric patients, whereas the healthcare-based cohort consisted predominantly of adult patients. Only proband analysis was possible for the adult cohort, whereas trio analysis, that is, concurrent testing and analysis of proband in both parents was possible in a large fraction of the cohort coming from GeneDx, three-fourths of it. Gender distribution was almost even in both cohorts, and we also determined the presence of three neurodevelopmental comorbidities in these patients, intellectual disability, epilepsy, and autism spectrum disorder. A positive diagnostic result was defined as the identification of one or two clinically relevant pathogenic or likely pathogenic variants in a single gene, depending on the mode of inheritance. Based on these, we were able to determine that the molecular diagnostic yield was 32.7% in the clinical laboratory referral cohort, as you can see on the middle column of the table. Among all the patients with a positive result in this subset, 94.3% had single nucleotide variants, 4.3% had copy number variants, and 1.4% had both single nucleotide variants and CNVs. On the third column of this table, you can see that pathogenic and likely pathogenic variants were identified in 10.5% of patients in the healthcare-based cohort, of which 73.7% had SNVs and 26.3% had copy number variants. Okay, so as we believe it would be clinically useful to know whether there is any particular comorbidity that could increase the likelihood of obtaining a diagnostic answer from genetic testing, we reviewed the likelihood of obtaining a positive result based on the presence or absence of these comorbidities. There is a lot of information on this table adapted from our publication, so let me walk you through it for a moment. Complete comorbidity data on the presence or absence of intellectual disability, epilepsy, or autism spectrum disorder was available for all patients from both cohorts. Each comorbidity and their possible combinations are listed in the first column. In the next two columns, you can see the diagnostic yield in the subgroup of patients with or without the comorbidities listed on the first column. The odds ratio on the fourth column were estimated from a logistic regression model that controlled for cohort and each individual neurodevelopmental comorbidity. This analysis assesses the likelihood of detection of a positive finding in a patient with CP and the neurodevelopmental comorbidity listed in the first column versus patients with cerebral palsy but without the comorbidity in the first column. Finally, the fifth column on the left presents a statistical significance calculated using a key squared test. So what do these results mean? For instance, for patients with cerebral palsy and intellectual disability versus patients with cerebral palsy but without intellectual disability, the adjusted odds ratio was 2.59 for the association between neurodevelopmental comorbidities and the likelihood of having a positive result. This difference is statistically significant. The odds ratio was 1.01 for patients with cerebral palsy and epilepsy versus patients with cerebral palsy but without epilepsy. And it was 1.04 for patients with cerebral palsy and autism spectrum disorder versus patients with cerebral palsy but without autism spectrum disorder. We can also see that the odds ratio increases whenever intellectual disability is included in the group, reaching statistical significance in all of them, and increases to up to 2.73 for patients with cerebral palsy and all three neurodevelopmental comorbidities versus patients with cerebral palsy but without the three neurodevelopmental comorbidities. Additional details on the analysis are available on the published manuscript. Now I'm going to show you some features of the positive results identified in these cohorts. When grouping the 459 positive findings by inheritance pattern, it is clear that the inheritance of the majority of the disorders identified in this cohort was autosomal dominant, encompassing almost 65% of them. Among the 459 positive tests, 229 distinct genes were identified, illustrating the vast genetic heterogeneity associated with cerebral palsy. 143 genes were identified only once, which is almost two-thirds of the positive genes. We did identify some usual suspects, but not too many. Only seven genes were identified in seven or more cases. Within this group, CTNNB1 particularly stood out, accounting for 4% of the positive tests. Other genes frequently identified were KIF1A, COL4A1, GNA01, KCNQ2, MEGP2, and STXBP1. In addition, there were 10 genes in which pathogenic and likely pathogenic variants were independently identified in both cohorts. As I told you before, tree analysis, defined as the concurrent sequencing and review of the probe and parents, was possible for a large fraction of the laboratory-based cohort. In this chart, we illustrate the positive results obtained in the laboratory cohort based on the type of analysis performed. Our data showed that the positive diagnostic rate was significantly higher with a tree analysis at 35% as compared to probing alone at 23%. Tree analysis allows for improved variant interpretation and segregation analysis, which likely influenced this outcome. The chart on the left side of this slide shows the breakdown of the analysis approach within the lab-based cohort. You can see, as I mentioned before, that tree analysis was possible for almost three-fourths of this cohort. When looking at all the positive cases within the lab-based cohort included on this pie chart, we were able to determine that 30% of them, almost one-third, had inherited causative variants. For 12% of the results, infirmity family members were not available to determine inheritance. And more remarkable, overall, 58% of these positive variants occurred de novo. These de novo variants represented 72% of the positive results obtained with tree analysis. And of note, these de novo variants included five mosaic cases, or 1% of the overall positive results in the lab-based cohort. Now, breaking down the group of inherited positive variants, the pink slice of the pie, 21% were associated with autosomal recessive disorders, 6% were in genes associated with autosomal dominant disorders, and 3% were associated with X-linked disorders. Within this group of inherited pathogenic variants, we can see a small fraction of them were coming from mosaic parents. So, as you know, there are a number of important reasons to identify an underlying genetic cause of a disorder in an individual. In particular, determining a molecular diagnosis aids in genetic counseling with increased accuracy for recurrence risk assessment. So let's take a look at our data. Having all this information on inheritance of pathogenic variants in this cohort, we then proceeded to adjust the recurrence risk for these families, starting from the empiric recurrence risk, which for cerebral palsy is 1% to 2%. We based this adjustment and information from the over 1,000 trios available, since we can accurately determine the inheritance in these cases. Overall, we were able to provide an updated recurrence risk for 35% of our trio cases. For the majority of the cases, 65%, because of a negative or non-diagnostic result, the recurrence risk remains the empiric risk of 1% to 2%. Cases with de novo variants, 25% of all trios, allow us to provide families with a recurrence risk of less than the empiric risk. And finally, 10% of our trio cases have a recurrence risk of greater than 1% to 2%, due to inherited variants. In breaking down the families with a recurrence risk greater than the empiric risk, the majority of these families, 70%, had a recurrence risk of 25% due to the finding of a positive result in a gene associated with an autosomal recessive disorder. The remaining 30% of families have a recurrence risk of up to 50% due to a positive result in a gene associated with an autosomal dominant or X-linked disorder. So going back to the utility of genetic testing, we've already reviewed recurrence risk assessment, but it can also provide prognostic information and guidance for further clinical evaluation. Evaluation for the presence of other associated medical findings may be critical, and identifying a genetic diagnosis may also further enable precision medicine. Knowledge of the specific cause can help guide medical care and treatment, prevent other costly or inbasic tests, and provide eligibility for supported services. Let's delve into the second case study to illustrate these points. This individual is a 14-year-old boy with a history of asbastic hemiplegic CP and several CNS anomalies. Intellectual disability, plagiocephaly with accompanying facial asymmetry, and constipation were also reported. He's the youngest child of non-consanguineous parents with no family history of genetic or neurological disease. His delivery was unremarkable at term, and prior genetic testing through karyotype and microarray analysis were normal. This child underwent exome sequencing as a trio with both of his parents, as depicted in the image. Results revealed a maternally inherited intragenic deletion of the Based on the ACMG AMP joint guidelines for parent classification, this parent is classified as pathogenic. Supporting evidence of that classification include that it's a deletion involving multiple coding exons of the gene for which loss of function is a known mechanism of disease. It is not observed that a significant frequency in the general population or in the hemizygous state in large population cohorts. An interesting study was conducted in the United States, in large population cohorts, and interestingly, the majority of affected individuals have a sequence variant in the L1-CAM gene. Copy number variants are significantly less common, but exon level deletions have been reported in the human gene mutation database. As you may know, pathogenic variants in the L1-CAM gene cause a phenotypic spectrum of neurological disorders collectively referred to as L1 syndrome. There's a wide range of features associated with L1 syndrome, including hydrocephalus and other CNS anomalies, intellectual disability, abducted thumbs, and spastic paraplegia. This patient's phenotype more closely resembles the X-linked complicated corpus callosum agenesis phenotype. Based on our testing, this individual's mother was determined to be a carrier. Research shows that approximately 5% of heterozygous females manifest symptoms, but they're typically on the milder end of the clinical spectrum. This case is a good example of how an underlying enteology identified by genetic testing resulted in actionable findings for this family. The recurrence risk was able to be revised based on the results of genetic testing. The prior empiric recurrence risk of 1 to 2% is now revised to 50% for a future pregnancy to inherited pathogenic variants, with a 25% risk for an affected son. In addition, other at-risk family members will have the option of carrier and prenatal testing. Given the L1 syndrome diagnosis, this child would clearly benefit from a multidisciplinary team approach, including genetic counseling services. Head imaging to assess for hydrocephalus and other CNS abnormalities will guide appropriate treatment options. Interestingly, this child has a history of chronic constipation. Tearspring disease is associated with L1 syndrome, which highlights the relevance of his history of constipation. Clearly, a comprehensive workup for this condition is warranted, and treatment options are available if affected. For patients with adducted thumbs, surgical correction is not typically performed. A more conservative approach may be warranted based on the diagnosis and child severity. Fortunately, this family was able to obtain results, which directly impacts this child's care. Additionally, these results allow for reproductive planning options for the parents and potentially extended family members. So, to summarize the most important points of this review, and hopefully what will be your takeaway of this session, there is emerging and accumulating evidence that supports an expanding role for genomics in the etiology of cerebral palsy. It is important to keep in mind that a genetic finding may provide an etiology in a patient, but it does not exclude the clinical diagnosis of cerebral palsy. Given the appreciable diagnostic rate, genetic testing has to be considered in the diagnostic workup of patients with cerebral palsy. And when using exome sequencing, a trio-testing approach further improves the likelihood of getting a positive result. Establishing a genetic diagnosis is critical, as it may dramatically change the diagnosis is critical, as it may dramatically change the recurrence risk for a family. In addition, it can aid in clinical management and in delineating the prognosis. Finally, I'd like to acknowledge consented patients and family, healthcare providers, our collaborators at Geisinger, and the Neurogenetics, Clinical Genomics, and Data Science programs at GeneDx. If you have any questions, you can contact us at the email address provided here. For ordering tests, please visit our website at genedx.com. If you have questions about this presentation, please come join us for a live Q&A session on Thursday, April 15th at 12 p.m. Eastern time during the ACMG meeting. For details on how to join, please visit our booth. Hope you enjoyed the ACMG meeting, and thank you very much for joining us today.

cerebral palsy

genomic advances

genetic testing techniques

diagnostic variants

clinical management

trio analysis