Clinical characterization and developmental follow-up of a child with Sotos syndrome carrying a novel NSD1 variant Authors: Huanqi Tang, Wenjin Feng, Ruyu Tan, Shihao Ren, Xiaohui Kong, Long Zhang, Xiufang Jiang, Huabin Wang Published: 2026 ABSTRACT BackgroundSotos syndrome is a rare overgrowth–intellectual disability(OGID) disorder typically associated with developmental delay that stabilizes or improves with early intervention.Case presentationWe report a 17-month-old boy with classic clinical features of Sotos syndrome who presented with severe speech and motor delay. Whole-exome sequencing(WES) revealed a novel de novo nonsense pathogenic variant in NSD1:c.3910 C > T (p.Gln1304Ter), which is predicted to truncate multiple functional domains, including the SET(Su(var)3–9, Enhancer-of-zeste, and Trithorax) catalytic domain. Despite early and continuous rehabilitation, his developmental quotient(DQ) decreased from 45.8 at 17 months to 39.9 at 28 months, indicating slow and limited developmental progress over time. Brain MRI (magnetic resonance imaging) revealed corpus callosum abnormalities and reduced white matter volume.ConclusionsThis is the first report of Sotos syndrome caused by the NSD1:c.3910 C > T (p.Gln1304Ter) pathogenic variant. Beyond expanding the spectrum of pathogenic variants, this case highlights a relatively severe neurodevelopmental profile with persistently limited progress, underscoring the importance of early genetic diagnosis, sustained multidisciplinary rehabilitation, and long-term follow-up to optimize outcomes in affected children.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12887-026-07136-3. Introduction Sotos syndrome (also known as cerebral gigantism; OMIM 117550) was first described by Juan Sotos and colleagues in 1964. It is classified as anOGID syndrome, with an estimated incidence of approximately 1 in 14,000 live births [1]. The cardinal clinical features include pre- and postnatal overgrowth with height and/or head circumference > + 2 SD, mild to moderate intellectual disability, and distinctive craniofacial features such as a broad, prominent forehead, a receding hairline, downslanting palpebral fissures, flushed cheeks, a long face, and a pointed chin [1]. Additional manifestations often include infantile hypotonia, speech delay, and behavioral abnormalities, particularly autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD), which affect nearly half of patients [2]. Other reported findings include advanced bone age, congenital heart defects, abnormal brain imaging, joint hyperlaxity with flatfoot, and neonatal complications such as jaundice and hypoglycemia, renal dysplasia, scoliosis, and seizures. In 2002, Kurotaki et al. identified heterozygous NSD1 deletions or pathogenic variants as the primary molecular cause of Sotos syndrome [3]. NSD1 encodes a histone H3 lysine 36 (H3K36) methyltransferase. Increasing evidence indicates that NSD1 haploinsufficiency causes broad dysregulation of the epigenetic machinery, leading to alterations in genome-wide DNA methylation that affect not only growth-related pathways but also genes involved in neuronal development and cellular morphogenesis and differentiation, as demonstrated by Choufani et al. [4]. More than 500 pathogenic NSD1 variants have been reported across different variant types [1, 5]. Here, we present a child with typical clinical features of Sotos syndrome in whom WES revealed a novel de novo nonsense pathogenic variant in NSD1:c.3910 C > T (p.Gln1304Ter), located at chr5:176662935 C > T based on the GRCh37/hg19 reference genome. To our knowledge, this is the first report of this variant. The patient displayed persistently limited neurodevelopmental progress despite early rehabilitation, indicating a relatively severe but clinically recognizable Sotos phenotype. Case presentation Initial presentation The patient was a 17-month-old boy referred to the child health clinic with the chief complaint of “no speech to date.” He was the second child of healthy, nonconsanguineous parents. The pregnancy was uneventful. The child was born by Caesarean section, with a birth weight of 4.4 kg. Perinatal hypoxia was uncertain. Since infancy, his head circumference had been noticeably enlarged, and he had a voracious appetite. Developmental milestones included head control at 4 months, rolling at 5 months, and independent sitting at 10 months, but he was unable to crawl. On physical examination, his weight was 16.55 kg (z score: +3.82), height was 94 cm (z score: +4.63), and head circumference was 55 cm ( z score: +5.76) (Fig. 1). Dysmorphic features included a broad forehead, a depressed nasal bridge, and bilateral ptosis of the upper eyelids (Fig. 2A). He could stand with support but was unstable, with bilateral valgus deformity of the feet (Fig. 2B). Radiographs demonstrated flatfoot-like changes (Fig. 3A and B). He was unable to imitate simple gestures such as waving goodbye. Fig. 1Growth parameters at 17, 18, and 28 months. (A) Head circumference > P97; (B) weight-for-length > P97; (C) weight-for-age > P97;(D) length-for-age > P97, based on World Health Organization (WHO) standards Growth parameters at 17, 18, and 28 months. (A) Head circumference > P97; (B) weight-for-length > P97; (C) weight-for-age > P97;(D) length-for-age > P97, based on World Health Organization (WHO) standards Fig. 2(A) Facial dysmorphism showing a broad forehead, a depressed nasal bridge, and bilateral upper eyelid ptosis. (B) Standing posture with bilateral foot valgus (A) Facial dysmorphism showing a broad forehead, a depressed nasal bridge, and bilateral upper eyelid ptosis. (B) Standing posture with bilateral foot valgus No clear autistic traits were observed during clinical evaluations, although subtle social-communication vulnerabilities could not be fully excluded at this age. Neuropsychological assessment revealed a DQ of 45.8, which is consistent with global developmental delay. Developmental assessment was conducted using the Children’s Developmental Scale (PEP), which evaluates gross motor, fine motor, language, social interaction, and adaptive behavior. The child demonstrated significant delays in expressive language and fine motor abilities, with relatively better but still delayed gross motor performance. Although the patient presented with speech delay, no audiometric evaluation was performed. This assessment established a baseline for future developmental follow-up. Brain MRI, particularly on axial T2-weighted images, demonstrated mild ventricular enlargement, reduced white matter volume, widened extracerebral spaces in the frontal and temporal lobes, and a small cystic lesion in the corpus callosum (Fig. 3C and D). These findings suggested impaired brain development. Fig. 3(A and B) Lateral foot radiographs showing flatfoot-like changes with marked eversion. (C) Sagittal T1-weighted MRI image showing a small cystic lesion in the corpus callosum. (D) Axial T2-weighted MRI image demonstrating mild ventricular enlargement, reduced white matter volume, and widened extracerebral spaces in the frontal-temporal regions, consistent with impaired brain development. Arrows indicate the cyst in the corpus callosum (A and B) Lateral foot radiographs showing flatfoot-like changes with marked eversion. (C) Sagittal T1-weighted MRI image showing a small cystic lesion in the corpus callosum. (D) Axial T2-weighted MRI image demonstrating mild ventricular enlargement, reduced white matter volume, and widened extracerebral spaces in the frontal-temporal regions, consistent with impaired brain development. Arrows indicate the cyst in the corpus callosum Treatment included mouse nerve growth factor injection (18 µg intramuscularly every other day for a total of 10 doses), rehabilitation training, and custom orthopedic shoes. Rehabilitation therapy was continued, and WES was subsequently arranged. Follow-up (18 to 28 months) At 18 months, the patient weighed 18 kg ( z score: +4.50), measured 94 cm in height (z score: +4.18), and had a head circumference of 55 cm (z score: +5.61) (Fig. 1). Developmental assessment showed a DQ of 44.2. He was able to crawl and stand with support but could not walk independently. Fine motor skills remained poor, and he could not perform a pincer grasp. He responded to his name but produced no active speech. Echocardiography revealed mild mitral and tricuspid regurgitation. Rehabilitation therapy was continued. At 28 months, the patient weighed 21 kg (z score: +4.11), measured 105 cm in height (z score: +4.13), and had a head circumference of 57 cm ( z score: +5.89). (Fig. 1). He was able to walk short distances but exhibited an abnormal waddling gait with left foot valgus and poor stability. Fine motor skills remained impaired, and he was still unable to perform a pincer grasp. DQ declined to 39.9. Bone age was advanced by 1.2 years (Fig. 4). No new structural abnormalities were identified. Rehabilitation therapy continued. The summary of the patient’s clinical features and developmental milestones is presented in Table 1. Fig. 4Anteroposterior X-ray of the left hand and wrist at 28 months, showing a bone age of ~ 3 years 6 months, which is advanced compared with chronological age Anteroposterior X-ray of the left hand and wrist at 28 months, showing a bone age of ~ 3 years 6 months, which is advanced compared with chronological age Table 1Clinical features, developmental milestones, and corresponding interventions in the patientAgeClinical Features/EvaluationInterventionAt birthWeight 4.4 kg; macrocephalyNone4 monthsLifted headNone5 monthsRolled overNone10 monthsSat unsupported; unable to crawlNone17 monthsSpeech delay; macrocephaly; facial dysmorphism; flatfoot; DQ: 45.8; abnormal brain MRIRehabilitation; nerve growth factor; orthopedic shoes; genetic testing18 monthsCrawled; stood with support; poor fine motor; no speech; DQ: 44.2; mild cardiac regurgitation; NSD1 pathogenic variant (Sotos syndrome) confirmedContinued rehabilitation28 monthsShort-distance walking; abnormal gait; poor fine motor and speech; DQ: 39.9; advanced bone age (3 year 6 months)Continued rehabilitation Table 1Clinical features, developmental milestones, and corresponding interventions in the patientAgeClinical Features/EvaluationInterventionAt birthWeight 4.4 kg; macrocephalyNone4 monthsLifted headNone5 monthsRolled overNone10 monthsSat unsupported; unable to crawlNone17 monthsSpeech delay; macrocephaly; facial dysmorphism; flatfoot; DQ: 45.8; abnormal brain MRIRehabilitation; nerve growth factor; orthopedic shoes; genetic testing18 monthsCrawled; stood with support; poor fine motor; no speech; DQ: 44.2; mild cardiac regurgitation; NSD1 pathogenic variant (Sotos syndrome) confirmedContinued rehabilitation28 monthsShort-distance walking; abnormal gait; poor fine motor and speech; DQ: 39.9; advanced bone age (3 year 6 months)Continued rehabilitation Clinical features, developmental milestones, and corresponding interventions in the patient Molecular diagnosis Genomic DNA extracted from the proband’s peripheral blood was submitted to a certified commercial genetic testing laboratory (Kangso Medical Laboratory, Beijing, China) for WES. The average sequencing depth of the nuclear exome was approximately 100× and that of the mitochondrial genome was 3000×. Library preparation was performed using the Agilent SureSelect V6 hybridization-based exome capture kit, followed by Sequencing was performed on the BGI T7 platform. Raw sequencing data were processed using a validated bioinformatic pipeline, including Consensus Assessment of Sequence and Annotation Variation (CASSAVA) for initial data processing, Genome Analysis Toolkit (GATK) for variant calling [6], and Annotator of Genetic Variants (ANNOVAR) for variant annotation [7]. Variant annotation incorporated multiple population and disease databases, including gnomAD, Exome Aggregation Consortium (ExAC) [8], 1000 Genomes, Human Gene Mutation Database (HGMD) [9], Online Mendelian Inheritance in Man (OMIM) [10], and Human Phenotype Ontology (HPO) [11]. Copy number variation (CNV) analysis was performed using eXome Hidden Markov Model (XHMM) [12], ExomeDepth, and ClinCNV, and no clinically significant CNVs were identified. Variant interpretation was conducted in accordance with the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) 2015 guidelines. A heterozygous nonsense variant in NSD1, NM_022455.4:c.3910 C > T (p.Gln1304Ter) (Fig. 5), was identified and classified as pathogenic on the basis of PVS1 (very strong evidence for loss of function), PS2 (confirmed de novo occurrence), and PM2 (absence from population databases), and PP4, because the patient’s phenotype was highly specific for Sotos syndrome. Sanger sequencing confirmed the presence of the variant in the proband and its absence in both parents, establishing its de novo origin. The variant is located in exon 6 of 23 exons in NSD1 and is predicted to truncate approximately half of the protein, resulting in the loss of multiple C-terminal functional domains, including several PHD (Plant Homeodomain) motifs and the catalytic SET domain. This truncation is consistent with the known pathogenic mechanism of Sotos syndrome. On the basis of these molecular findings, a definitive diagnosis of Sotos syndrome due to a de novo NSD1 nonsense pathogenic variant was established. This novel NSD1 variant has been submitted to ClinVar under submission accession number SCV007126705. Fig. 5High-throughput sequencing showing the heterozygous NSD1:c.3910C>T (p.Gln1304Ter) variant in the proband (arrow), which was absent from both parents (Sanger sequencing) High-throughput sequencing showing the heterozygous NSD1:c.3910C>T (p.Gln1304Ter) variant in the proband (arrow), which was absent from both parents (Sanger sequencing) Discussion We confirmed the diagnosis of Sotos syndrome in this patient throughWES, which revealed a de novo heterozygous nonsense pathogenic variant in NSD1:c.3910 C > T, p.Gln1304Ter. This variant lies in exon 6 and is predicted to truncate multiple C-terminal domains, including PHD zinc fingers and the SET catalytic domain, resulting in profound loss of protein function [3]. On the basis of the ACMG/AMP criteria described earlier, this variant meets the requirements for pathogenic classification [13]. Separately, large cohort analyses have demonstrated that the phenotypic variability of Sotos syndrome is broad and largely independent of genotype. Accordingly, the severity observed in our patient does not imply a direct genotype–phenotype correlation [14]. At the time of genetic analysis, this variant was not present in ClinVar, gnomAD, or other population databases, supporting its novelty [15–16]. The variant has since been submitted to ClinVar under submission accession number SCV007126705. Before the molecular diagnosis was established, several pediatric overgrowth syndromes were considered in the differential diagnosis. Weaver syndrome was initially considered due to postnatal overgrowth and macrocephaly; however, the absence of its characteristic craniofacial features and the lack of EZH2 variants made this diagnosis unlikely [17]. Beckwith–Wiedemann syndrome was also evaluated, but key features such as macroglossia, abdominal wall defects, visceromegaly, or lateralized overgrowth were not observed [18]. Other overgrowth syndromes, including Malan syndrome and DNMT3A-related overgrowth syndrome (Tatton–Brown–Rahman syndrome), were considered but were not supported by the clinical presentation or genetic findings [19]. Clinically, our patient exhibited the typical features of Sotos syndrome, including macrocephaly, overgrowth, a characteristic craniofacial appearance, and developmental delay. Notably, despite receiving early rehabilitation, the child’s DQremained persistently low during follow-up. These findings highlight the slow and variable progression of the patient’s neurodevelopmental condition, emphasizing the need for ongoing monitoring and individualized intervention. Although early intervention usually helps to stabilize or improve neurodevelopmental outcomes [14], the patient’s progress remained slower than expected despite rehabilitation therapy. The decline in DQ reflects a slower pace of development rather than true neuroregression, emphasizing the variability in neurodevelopmental trajectories and the need for ongoing monitoring. This highlights the variability in neurodevelopmental trajectories and underscores the importance of long-term monitoring and individualized management.In the early stages of diagnosis, the patient was empirically treated with nerve growth factor injections due to developmental delays. However, there is currently insufficient evidence to support its effectiveness in improving developmental outcomes in Sotos syndrome. Therefore, this intervention is only reported as part of the clinical history, not as a recommended treatment. This highlights the uncertainty in early management of Sotos syndrome, particularly when a definitive diagnosis has not yet been established. The decline in DQ reflects that the child’s developmental age increased more slowly than his chronological age did, rather than any loss of previously acquired milestones, a distinction that has been emphasized in prior work demonstrating that early developmental scores may fluctuate within the measurement error and confidence intervals of standardized tests [20]. This pattern is consistent with the developmental principle described by Fisch (2015), in which declining standard scores reflect increasing divergence from age-matched peers rather than true regression [21]. The cognitive and behavioral difficulties observed in our patient are not unique to Sotos syndrome and may overlap with those reported in other neurodevelopmental conditions. In toddlers, early social-communication vulnerabilities, attentional capacity, and test engagement can influence the validity of developmental testing, underscoring the need for cautious interpretation of early developmental assessments [20]. More broadly, vulnerabilities in attention and executive processing may limit a child’s ability to cope with increasingly complex environmental demands as they age. Consequently, widening gaps in standardized scores may reflect an accumulated developmental lag rather than any loss of previously acquired abilities. Although truncating variants and microdeletions have been reported more frequently in individuals with pronounced developmental challenges in some cohort studies [22–23], these findings represent population-level associations rather than predictors of individual outcomes. Large cohort analyses, including those by Tatton-Brown et al. [15], have shown that the phenotypic variability of Sotos syndrome is broad and largely independent of NSD1 variant type. Therefore, the severity observed in this patient cannot be attributed to the specific truncating variant identified. At the mechanistic level, NSD1 encodes a histone H3K36 methyltransferase essential for chromatin organization and transcriptional regulation during neural development. Loss-of-function variants disrupt H3K36me2/3 deposition and lead to secondary alterations in epigenetic marks such as H3K27me3 and DNA methylation patterns, thereby perturbing neurodevelopmental pathways [24]. The truncating variant identified in our patient removes the SET catalytic domain, making a substantial reduction in enzymatic function highly plausible. Although in vitro functional assays were not performed, the convergence of clinical observations, genetic evidence, and established molecular mechanisms supports the pathogenicity of this novel NSD1 variant while underscoring the wide phenotypic spectrum associated with Sotos syndrome [24–26]. From a management perspective, this case highlights both the benefits and the limitations of early intervention. Rehabilitation was initiated at 18 months, but limited continuity and intensity due to socioeconomic factors likely contributed to the limited developmental trajectory. Outcomes in Sotos syndrome depend not only on the timing of therapy but also on its intensity, sustainability, and family support. Evidence suggests that early, intensive, and multidisciplinary rehabilitation—including physical, occupational, speech, and behavioral therapy—can improve long-term outcomes [2, 27]. In summary, this case confirmed Sotos syndrome caused by a de novo NSD1:c.3910 C > T nonsense pathogenic variant. Unlike most patients with Sotos syndrome who typically show stable or improved developmental trajectories after early intervention, our patient demonstrated persistently limited developmental progress, which does not indicate true neuroregression but instead reflects the variability in developmental trajectories observed in Sotos syndrome. Although truncating variants and microdeletions have been reported more frequently among individuals with more pronounced developmental challenges [24–25], the considerable genotype–phenotype variability in Sotos syndrome prevents the attribution of clinical severity to a specific variant type at the individual level. This unusual clinical course further illustrates the broad phenotypic spectrum of Sotos syndrome and provides valuable insights for clinical practice. From a practical perspective, this case emphasizes three important points. First, in children with unexplained overgrowth and developmental delay, early genetic testing should be considered to establish a timely diagnosis. Second, sustained and multidisciplinary rehabilitation is essential for optimizing developmental outcomes [2, 27]. Third, even with early intervention, clinicians should remain aware that some children may show persistently limited developmental gains and a widening developmental gap over time, underscoring the need for individualized long-term follow-up and management strategies [28]. Fourth, the genetic diagnosis also informed genetic counseling for the family. Because the NSD1 variant was confirmed to be de novo and absent in both parents, the recurrence risk for future siblings is considered low, although gonadal mosaicism cannot be completely excluded. In addition, affected individuals would have a 50% risk of transmitting the variant to their offspring. Family perspective The parents reported that the genetic diagnosis helped them better understand the child’s condition and reduced some uncertainty. They also expressed willingness to continue long-term rehabilitation and regular follow-up. Conclusions This report describes the first case of Sotos syndrome caused by a de novo NSD1:c.3910 C > T nonsense pathogenic variant, characterized by a variable neurodevelopmental course with persistently limited developmental progress. Beyond expanding the spectrum of pathogenic variants in NSD1, this case highlights the clinical heterogeneity of neurodevelopmental trajectories in Sotos syndrome and underscores the importance of early genetic diagnosis, sustained multidisciplinary rehabilitation, and long-term follow-up. Supplementary Information Supplementary Material 1. Supplementary Material 1.