Twenty‐year follow‐up of the facial phenotype of Brazilian patients with Sotos syndrome

Sotos syndrome is characterized by overgrowth starting before birth through childhood with intellectual disability and craniofacial anomalies. The majority of patients are large for gestational age with developmental delay or intellectual disability. The majority of cases are caused by pathogenic variants in NSD1. The most consistent physical features in this disorder are facial dysmorphisms including prominent forehead, downslanted palpebral fissures, prognathism with a pointed chin, and a long and narrow face. We present a follow‐up to a cohort of 11 individuals found to harbor heterozygous, pathogenic, or likely pathogenic variants in NSD1. We analyzed the facial dysmorphisms and the condition using retrospective over 20 years. Among these patients, followed in our medical genetics outpatient clinic for variable periods of time, all had a phenotype compatible with the characteristic Sotos syndrome facial features, which evolved with time and became superimposed with natural aging modifications. We present here a long‐term follow‐up of facial features of Brazilian patients with molecularly confirmed Sotos syndrome. In this largest Brazilian cohort of molecularly confirmed patients with Sotos syndrome to date, we provide a careful description of the facial phenotype, which becomes less pronounced with aging and possibly more difficult to recognize in adults. These results may have broad clinical implications for diagnosis and add to the global clinical delineation of this condition.


| INTRODUCTION
Sotos syndrome is an overgrowth condition with an autosomal dominant inheritance pattern, caused by haploinsufficiency of NSD1 or, less commonly, NFIX, and estimated to occur in 1:14,000 live births (Tatton-Brown et al., 2005).
The triad of cardinal clinical features is overgrowth, characteristic facial appearance, and learning disabilities, but other major features include orthopedic (advanced bone age, scoliosis, and joint laxity), cardiac (frequently septal defects), renal (vesicoureteral reflux), and neurologic abnormalities (ventricular dilatation and seizures) together with tumor predisposition (sacrococcygeal teratoma, neuroblastoma, presacral ganglioglioma, acute lymphoblastic leukemia, and small-cell lung cancer). Gestational and perinatal complications are common, including maternal preeclampsia, neonatal jaundice, hypotonia, and poor feeding (Tatton-Brown et al., 2020).
We present 11 individuals with molecularly confirmed Sotos syndrome who were followed for a period of up to 20 years , along with careful photographic documentation of the craniofacial phenotypes. Most literature focuses on patients of North American or European origins; however, to our knowledge, there is only one published cohort of Brazilian patients, which included only pediatric patients, with five molecularly confirmed individuals (Vieira et al., 2015). It is our hope that the photographic documentation of the aging craniofacial features we provide will be a valuable addition to the global delineation of Sotos syndrome.

| MATERIALS AND METHODS
We reviewed the medical records of 34 Brazilian patients with suspected Sotos syndrome from 1994 to 2019. These patients were examined by a clinical geneticist with experience in dysmorphology and each satisfied the clinical criteria for the diagnosis of Sotos syndrome. Furthermore, all photographs were reviewed by an outside clinical dysmorphologist to confirm the initial descriptions.
The majority of the patients were examined once a year for the duration of their enrolment in our outpatient clinic. All 34 patients were tested with Sanger sequencing of NSD1, and only patients with a confirmed molecular diagnosis (pathogenic or likely pathogenic variants in NSD1) were included in this study.
All experimental protocols were approved by the ethics committee of the institution and were in accordance with the Helsinki Declaration. Informed consent forms were signed by or their parents or guardians when participants were under 18 years of age.

| Molecular findings
Of the initial 34 patients with a clinical diagnosis of Sotos syndrome, 11/34 (32.3%) were found to harbor heterozygous pathogenic or likely pathogenic variants in NSD1 (Table 1) and, therefore, were included in this study. Of the 11 variants in NSD1, 2 were novel and 9 were previously reported pathogenic variants (Ha et al., 2016). Six subjects (S4-S9) were also evaluated by Multiplex Ligation-dependent Probe Amplification (MLPA) studies with negative test results. As of yet, none of the remaining patients have been submitted to more comprehensive genomic testing.

| The cohort
Among these 11 patients, four are male and seven are female. All were followed in the same medical genetics outpatient clinic for a var-   had respiratory problems and two (18.2%) had feeding difficulties.
None required tube feeding. These data as well as other major malformations found in the patients are summarized in Table 2.

| Craniofacial features
Serial photographs of nine patients were available. All these individuals had a phenotype compatible with the characteristic facial features of Sotos syndrome, including prominent forehead, prognathism with a pointed chin, downslanted palpebral fissures, and widely spaced eyes; these features evolved with time and became superimposed with natural aging modifications to the facial phenotype (Figure 1).

| Growth
The data on growth varied by participant. Eight of the 11 patients had at least seven or more sets of anthropometric measurements at different appointments. One had two sets of measurements and the remaining two had written documentation of excessive growth velocity. Seven patients (63.6%) had advanced bone age. Data on growth are summarized in Table 2. None of our patients developed benign or malignant tumors of any type during the duration of the study.

| Cognitive development
All patients had developmental delay or intellectual disability, with speech delay present in 10/11 patients (92.3%). Among 7 patients for whom data about school performance after the age of 10 was available, 5 (71.4%) were enrolled in special schools or inclusion projects for children with learning disabilities. One of the two remaining patients was enrolled in regular school; however, she had a history of learning difficulties and needed extra credits to achieve passing marks. One of the patients discontinued follow-up at the age of 2 years, before we could accurately assess school performance. One of the patients, at 9 years old, is enrolled in regular school; however, she has problems with mathematics. She is also reported to have juvenile behavior and difficulties socializing with other children.
Among the seven adult patients, two did not complete school, although one of them was later enrolled in an adult education program. Only one patient in our cohort finished high school and went on to pursue further studies. This patient is currently enrolled in a mechatronics technical course at a private university; however, he did have a long history of learning difficulties when he was younger. Data on performance and development are summarized in Table 2.

| DISCUSSION
Our patients followed the expected clinical courses and natural history of Sotos syndrome in other populations (Cole & Hughes, 1994).
The majority (90.9%) were clinically diagnosed with Sotos at their first appointment in our Medical Genetics outpatient clinic, which suggests how highly sensitive the clinical diagnosis by an experienced dysmorphologist can be for this condition.
Facial gestalt has been reported as the most consistent criteria for a clinical diagnosis of Sotos syndrome (Tatton-Brown et al., 2005).
While most of the published literature describes patients of Caucasian ethnicity, either North American or European in origin, we provide a detailed study of the facial phenotype of patients from Brazil, a country whose population is highly mixed.
We have made a careful documentation of the facial phenotype (  Clinical follow-up, however, evidenced that some facial characteristics change with time, becoming superimposed with normal aging features, as evidenced in Figure 1. The pointed chin, present at infancy and childhood, becomes square, consistent with a finding that was previously reported in adult individuals with Sotos syndrome (Foster et al., 2019). Besides that, the forehead becomes less prominent, the widely spaced eyes less evident and the jaw line less narrow.
This suggests a powerful compensatory mechanism that eventually obscures the facial dysmorphism with aging. Therefore, the facial findings typical of Sotos syndrome seem to become less pronounced with aging, which could possibly make the clinical recognition more difficult even for experienced dysmorphologists.  -Mulla et al., 2004;Deardorff et al., 2004;Kato et al., 2009;Lapunzina, 2005), and it differs from other overgrowth syndromes. Interestingly, none of our patients presented with or developed any benign or malignant tumors during the duration of the study. We acknowledge, however, that this is a small cohort, and the value of this information is limited.
There is only one established genotype-phenotype correlation in Sotos syndrome: patients with copy number variations in NSD1 have less overgrowth and more pronounced learning disabilities (de Boer et al., 2004;Nagai et al., 2003). All our patients had point mutations; therefore, we could not observe this correlation. Moreover, patients with negative results from NSD1 sequencing were not evaluated through other techniques, such as exome sequencing or chromosomal microarray, which could lead to the diagnosis of other conditions with a clinical overlap to Sotos syndrome.

| CONCLUSION
To our knowledge, this is the largest Brazilian cohort of molecularly confirmed patients with Sotos syndrome so far. By a careful examination of the craniofacial phenotype, we documented the gradual and moderate change of facial phenotype, which becomes less pronounced with aging and possibly more difficult to recognize in adults. Thus, physicians must have a higher level of suspicion for this condition in adults.
We hope this will be a valuable contribution to the global clinical delineation of Sotos syndrome, especially considering that Brazil has a diverse and mixed population of different ethnic origins, in a context of few descriptions of individuals with Sotos syndrome from diverse populations in the literature.

ACKNOWLEDGMENT
We would like to thank the patients and their families who participated in this study. We also thank the support of funding provided by Caroline Jones-Carrick and Collin Carrick. Open Access funding was provided by the Qatar National Library.