Luscan-Lumish syndrome (LLS, OMIM: 616831), as a rare condition characterized by overgrowth, macrocephaly, obesity, type Ⅰ Chiari malformation, and linguistic retardation, has been recognized over the recent years as an autosomal dominant genetic condition caused by SETD2 gene variant (van Rij et al., 2018; Suda et al., 2021). Locating in 3p21.31, the SETD2 gene consists of 23 exons and 22 introns, with a length of about 147 KB. The histone methyltransferases (HMTs) encoded by SETD2 gene, as a critical epigenetic regulator, plays a key role in multiple biological processes such as regulating histone methylation, gene transcription, maintaining gene stability and remodeling the epigenome and (epi) cytoskeleton (Park et al., 2016; Walker and Burggren, 2020; Koenning et al., 2021; Liu et al., 2021; Seervai et al., 2021). To date, few cases about this disease have been reported worldwide, and none has ever been recorded in China. In the present study, by next-generation sequencing on a proband with a major manifestation of delayed language development, recognition lag, hypomegasoma, and macrocephaly, we successfully identified the genetic etiology of this disease.

The male proband aged 5 years and 7 months old was admitted to our hospital in November 2020 due to “being a slow learner in kindergarten”. The child, once hospitalized for premature birth for half a month, was the third fetus and the third birth of the mother who showed no abnormalities during pregnancy but was admitted for cesarean section at 36 weeks owing to intrauterine distress. His motor and language development were delayed compared with children of the same ages, as demonstrated by not being able to walk and say some simple words till about 1.5 years of age. After a long period of stagnation in language progress, his vocabulary gradually expanded at about 5 years of age. When the boy was admitted, he could only speak four to five words in short sentences. In the middle class of his kindergarten, he was noticed to be a slow learner with a poor ability of language expression and have little communication with his classmates. At home, besides being capable of responding to his family and expressing his needs with simple words, the boy could also share attention with them. No obvious stereotyped behavior was observed except for his particular obsession with Altman, as well as the likings of shaping plasticine and painting. On examination, the child was measured as 124.5 cm (>+2SD) in height and 54 cm (>+2SD) in head circumference, without unusual facial features or any abnormalities observed in cardiopulmonary auscultation and other physical examinations. Behavior evaluation in the clinic revealed normal eyes contact, as well as the ability to call his mother and to ask and answer simple questions.

Neither history of major physical diseases nor family genetic history was reported in the family except the proband. The patient had two elder sisters, the normally-developed 16-year-old sister being a freshman in senior high school with a good academic performance, and the 7-year-old sister in grade one with an average school record.

Gesell development schedules (GDS) indicated a score of 47 in adaptability, 45 in large motor movement, 46 in fine motor movement, 33 in language, and 39 in personal-social development. No abnormalities were found in blood routine, thyroid function or biochemical tests. Hearing test also showed no anomalies. Craniocerebral magnetic resonance imaging (MRI) revealed a slightly shorter corpus callosum with a mildly thinner corpus callosum body (Supplementary Figure S1).

LLS, being an autosomal dominant syndrome as a result of SETD2 gene mutation, has only been recognized till 2012 (8-11). Since then, the syndrome has been officially named Luscan-Lumish syndrome (LLS), also known as “SETD2-related overgrowth syndrome” (van Rij et al., 2018). To date, LLS has manifested a diversity of clinical phenotypes, among which the commonly-observed phenotypes included postnatal overgrowth, macrocephaly, developmental delay and autism spectrum disorder (ASD) (Table1). Including our case, 56% of the cases were taller than children of the same age and sex in childhood, which has been speculated involving a mechanism relevant to GH-dependent enhancement of STAT5b signaling (Suda et al., 2021). However, according to the data reported in the literature, the patients who were followed up to adulthood did not finally have a high stature compared with their peers, which led to another speculation that the high stature in childhood might be related to the advanced bone age of these children (Luscan et al., 2014). As for our case, he was measured a height of 124.5 cm (>+2SD) at the age of 5 years and 7 months, which was higher than children of the same age, and the final height was also needed to monitored.

Macrocephaly has been recognized as another major phenotype of LLS, as demonstrated by the available data from literature review showing that 13 out of the 15 cases (87%) had a head circumference greater than 97% of children of the same age and sex, which was also a feature consistent with our case (Table1). Among these patients, abnormal craniocerebral MRI findings, though with a great heterogeneity between individuals, were spotted in 7 cases (Table1). In our case, the craniocerebral MRI image displayed a slightly shorter corpus callosum accompanied by a mildly thinner corpus callosum body, which was a finding not reported in any of the previous cases of LLS. Otherwise, about 14%–18% of individuals with autism have head circumferences above the 97th percentile (Stevenson et al., 1997). Thus, it is necessary to distinguish between Luscan-Lumish syndrome and autism, and the next-generation sequencing technology is a good way to detect these diseases, especially for LLS. ASD and ID are also considered as major phenotypes of LLS syndrome. Among all the cases that we reviewed, eight children were diagnosed with ASD and 11 children with ID. It has been speculated that SETD2 gene variant is likely to be a candidate gene causing ASD or ID (O'Roak et al., 2012; Tlemsani et al., 2016; Lumish et al., 2015). Also, our case was presented to the hospital because of learning difficulties, and he did not correspond to the diagnostic criteria for ASD. However, the excessive height and head circumference of the child failed to draw much attention from the parents, which suggested the necessity and significance for clinical practitioners to carry out systematic physical examinations for children with developmental delay, especially genetics testing aiming at etiological analysis, for those with abnormal symptoms and signs.

Among the 15 reported LLS, the mutations showed a uniform distribution throughout the whole SETD2 gene, including 5 missense mutations, 5 frameshift mutations and 5 non-sense mutations (Figure 1). As a result, these variants led to malfunction of the proteins encoded by the SETD2 gene, the p. Arg1625Cys and p. Tyr1666Cys were located in the catalytic domain, the other three missense variants were distributed between the functional domains; The p. Leu1815Trp variant was spotted in a patient similar to Sotos syndrome (Luscan et al., 2014); The p. Ile41Phe variant was identified to be associated with isolated ASD (O'Roak et al., 2012); the p. Leu79His variant was detected in a 20-year-old male patient with LLS (Suda et al., 2021); Furthermore, SETD2 and NSD1 proteins, as two methyltransferases of histone H3 lysine 36 (H3K36) in nucleosomes, are involved in some specific biological processes such as transcription elongation, RNA processing and DNA repair. Functional studies on p. Arg1625Cys revealed that complete loss of H3K36me3 was associated with the loss of protein function (Hacker et al., 2016; Marzin et al., 2019). However, due to the limited number of patients having been investigated, definitive conclusions on the genotype-phenotype correlation cannot be drawn at this time.

A number of studies have reported that variants in somatic SETD2 gene are related to the occurrence of a variety of tumors, especially renal cell carcinoma, pulmonary adenocarcinoma, enteropathy-associated T cell lymphoma, chronic lymphocytic leukemia, and gastrointestinal stromal tumors (D'Avella et al., 2020; Tessema et al., 2018; Moffitt et al., 2017; Parker et al., 2016; Huang et al., 2016). Compared with the general population, patients with overgrowth syndrome tend to register a slightly increased prevalence of tumors, with studies reporting a 7% risk of developing tumors in Beckwith-Wiedemann syndrome and a 3%-5% elevation in the risk of a wide range of tumors in Sotos and Weaver syndromes (Tatton-Brown and Rahman, 2013; Villani et al., 2017; Brioude et al., 2019). So far, only a solitary case with meningioma was detected by craniocerebral MRI among all the reported patients with LLS syndrome. Given that most of the patients were children or young adults, long-term follow-ups should be performed to assess the potential risk of neoplasia.

The proband in our study presented with developmental delay, tall stature and macrocephaly, in combination with an abnormal craniocerebral MRI imaging, which was considered to be consistent with the main phenotypes of LLS. Subsequently, a heterozygous c.2514_2516delTAG (p.Ser838del) mutation in the SETD2 gene was identified, which, as a codon mutation, consequently resulting in the malfunction of the amino acid encoded by the gene. The present study, as the first report of LLS syndrome in China where the mutation of SETD2 gene has never been previously delineated, is believed to be a contribution to the enrichment of the mutation spectrum of SETD2 gene as well as the improvement of clinicians’ further understanding on this disease.