This report demonstrates biallelic NSD1 inactivation leading to pineoblastoma in a child with Sotos syndrome, indicating tumor suppressor implications.
To the Editor: Sotos syndrome (MIM #117550) is an overgrowth disorder characterized by macrocephaly, tall stature, distinctive craniofacial features, and neurodevelopmental impairments caused by a monoallelic loss-of-function variants in NSD1, which encodes a histone H3 lysine-36 methyltransferase that regulates promoter methylation through PRC2-mediated silencing [1, 2]. Histone lysine methyltransferases have been broadly implicated in carcinogenesis [3]. Somatic NSD1 promoter hypermethylation occurs in neuroblastomas and gliomas, while loss-of-function mutations are seen in squamous cell carcinomas, and associated with a DNA hypomethylation signature [4, 5]. Individuals with Sotos syndrome have a 3%–5% lifetime risk of childhood malignancy, including Wilms tumor, neuroblastoma, teratoma, hepatoblastoma, and leukemia [6, 7]. Pineoblastoma, a rare, aggressive embryonal neoplasm of the pineal parenchyma, has only recently been reported in association with Sotos syndrome [8-10]. Here, we describe a child with Sotos syndrome and a pineoblastoma with biallelic NSD1 inactivation and identify a second patient with Sotos syndrome and neuroblastoma with loss of heterozygosity, supporting the role of NSD1 as a tumor suppressor. The proband is a 4-year-old female, firstborn to a non-consanguineous couple of Haitian and Puerto Rican ancestry. Prenatal imaging revealed mild ventriculomegaly; however, repeated postnatal imaging demonstrated stability, and her early development was normal before being lost to follow-up. At 2.5 years old, she presented with persistent vomiting. Brain MRI demonstrated obstructive hydrocephalus and a pineal mass (Figure 1A). Biopsy revealed a poorly differentiated primary neoplasm with embryonal features, consistent with pineoblastoma, CNS WHO Grade 4. DNA methylation profiling classified the tumor as MYC/FOXR2-activated pineoblastoma [11]. She was treated according to Children's Oncology Group protocol ACNS0034, with induction chemotherapy with vincristine, methotrexate, etoposide, cyclophosphamide, and cisplatin. This was followed by a second-look surgery achieving a gross total resection, high-dose chemotherapy with autologous stem cell rescue, and 54 Gy intensity-modulated proton therapy. She remains without evidence of disease for 7 months posttreatment. Physical examination demonstrated marked macrocephaly, with a head circumference of 55 cm (z = +5.0), and generalized overgrowth, with a height of 107 cm (z = +4.1), and a weight of 18.4 kg (z = +2.4). She had a prominent forehead, long face, hypertelorism, downslanting palpebral fissures, and broad nasal bridge suggestive of Sotos syndrome. Trio exome sequencing through GeneDx (Gaithersburg, MD) identified a de novo variant NSD1 NM_022455.5:c.3584dup (p.Val1196Cysfs*7), which was classified as pathogenic based on ACMG criteria (Figure 1B) [12, 13]. Exome sequencing of the resected pineoblastoma at Caris Life Sciences (Irving, TX) revealed a second truncating variant in NSD1 NM_022455.5:c.4591dup (p.Met1531Asnfs*4) at an allele frequency of 39%, indicating biallelic inactivation (Figure 1C) [14]. No pathogenic variants were detected in other genes commonly associated with pineoblastoma, including RB1 or DICER1. To further interrogate the role of NSD1 in pediatric carcinogenesis, we analyzed 6602 tumors in the St. Jude Cloud [15]. Six germline variants in NSD1 were identified among the patients, of which five were missense variants whose impact on protein function was uncertain (Table 1). The remaining germline variant NSD1 c.5356_5360del (p.Lys1786Glufs*3) causes a frameshift, meeting criteria for pathogenicity. Tumor sequencing of the patient's neuroblastoma revealed that the variant was present at an 82% allele frequency, consistent with loss of heterozygosity. Eight somatic truncating variants in NSD1 were documented in the St. Jude Cloud, occurring in medulloblastoma (4), neuroblastoma (2), diffuse high-grade glioma, and acute myeloid leukemia (Figure S1). No somatic variants in NSD1 were observed among 11 cases of pineal tumors (Figure S2). Moreover, RNA sequencing performed in 14 patient samples with pineoblastoma showed uniform NSD1 expression, further suggesting that genetic or epigenetic alterations in NSD1 are rare drivers of tumorigenesis in pineoblastoma (Figure S3). This study provides molecular evidence supporting NSD1 as a tumor suppressor in Sotos syndrome through demonstration of biallelic inactivation in a child with pineoblastoma and corroborating loss of heterozygosity in a second case with neuroblastoma, aligning with a classical two-hit model of tumorigenesis. NSD1 plays a critical role in epigenetic regulation of growth and neurodevelopmental pathways through modulation of PRC2 activity [2]. Other overgrowth disorders including Weaver syndrome and Tatton–Brown–Rahman syndrome are also associated with PRC2 dysregulation and an increased risk of pediatric cancer [7, 16]. Although previous molecular analyses of pineoblastoma have not identified recurrent NSD1 alterations, the findings implicate NSD1 loss-of-function mutations as a rare but biologically meaningful driver in the context of Sotos syndrome [17, 18]. Notably, both our proband and the previously reported patient with Sotos syndrome and pineoblastoma developed the malignancy in early childhood [10]. While our cohort analysis did not find additional cases of pineoblastoma with NSD1 alterations, future large-scale sequencing and methylation studies may clarify whether disruption of NSD1 or related histone methyltransferases contributes to sporadic disease [19]. Given the variable expressivity of Sotos syndrome, with many patients exhibiting mild manifestations that may go unrecognized in early life, our findings also support the inclusion of NSD1 into multigene panel testing for pediatric cancer susceptibility [20]. Although routine cancer surveillance is not currently recommended, early recognition may facilitate prompt evaluation of concerning symptoms [7]. Prospective longitudinal studies will be essential to more precisely define the continuum of pediatric malignancy risks and inform surveillance strategies for Sotos syndrome patients. This study complies with the 1964 Helsinki Declaration and its later amendments and was approved by the University of Miami institutional research board (IRB protocol #20221338). The authors declare no conflicts of interest. Supplementary Figure 1. Lollipop plot of all somatic NSD1 loss-of-function variants identified through tumor exome/genome sequencing in the St. Jude Cloud cohort (n = 6,602), as seen with the ProteinPaint view. Variants are annotated according to protein change based on the NSD1 NM_172349 transcript. Supplementary Figure 2. Oncoplot of somatic driver mutations identified by tumor exome/genome sequencing among pediatric patients with pineoblastoma (n = 9) within the St. Jude Cloud cohort, as seen with the Oncoprint view. These samples were separately queried for somatic variants in NSD1, and no variants in the gene were found to be present. Supplementary Figure 3. Violin plot with overlaid boxplot and individual data points representing NSD1 gene expression levels across pediatric patients with pineoblastoma (n = 14) recorded within the St. Jude Cloud cohort. The expression levels are normalized by dividing sample counts by the geometric mean of counts across all samples for that gene, with the median of these ratios (MoR) per sample being used as a size factor to account for variants in sequencing depth or library size. The MoR is presented on a log10-transformed scale for visualization purposes. Values here are clustered tightly around a median value of 3.9 and ranging from 3.79 to 4.02, indicating a consistent level of NSD1 expression among the samples analyzed with no significant outliers observed. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
No takes yet. Share an insight, caveat, or question.
Borja et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: