Prospective study of epigenetic alterations responsible for isolated hemihyperplasia/hemihypoplasia and their association with leg length discrepancy

Molecular Basis and Diagnostic Approach to Isolated and Syndromic Lateralized Overgrowth in Childhood

OBJECTIVE

To demonstrate a high-yield molecular diagnostic workflow for lateralized overgrowth (LO), a congenital condition with abnormal enlargement of body parts, and to classify it by molecular genetics.

STUDY DESIGN

We categorized 186 retrospective cases of LO diagnosed between 2003 and 2023 into suspected Beckwith-Wiedemann spectrum, PIK3CA-related overgrowth spectrum (PROS), vascular overgrowth, or isolated LO, based on initial clinical assessments, to determine the appropriate first-tier molecular tests and tissue for analysis. Patients underwent testing for 11p15 epigenetic abnormalities or somatic variants in genes related to PI3K/AKT/mTOR, vascular proliferation, and RAS-MAPK cascades using blood or skin DNA. For cases with negative initial tests, a sequential cascade molecular approach was employed to improve diagnostic yield.

RESULTS

This approach led to a molecular diagnosis in 54% of cases, 89% of cases consistent with initial clinical suspicions, and 11% reclassified. Beckwith-Wiedemann spectrum was the most common cause, with 43% of cases exhibiting 11p15 abnormalities. PIK3CA-related overgrowth spectrum had the highest confirmation rate, with 74% of clinically diagnosed patients showing a PIK3CA variant. Vascular overgrowth demonstrated significant clinical overlap with other syndromes. A molecular diagnosis of isolated LO proved challenging, with only 21% of cases classifiable into a specific condition.

CONCLUSIONS

LO is underdiagnosed from a molecular viewpoint and to date has had no diagnostic guidelines, which is crucial for addressing potential cancer predisposition, enabling precision medicine treatments, and guiding management. This study sheds light on the molecular etiology of LO, highlighting the importance of a tailored diagnostic approach and of selecting appropriate testing to achieve the highest diagnostic yield.

Deciphering Epigenetic Backgrounds in a Korean Cohort with Beckwith-Wiedemann Syndrome

Background

Beckwith–Wiedemann syndrome (BWS) is a congenital overgrowth disorder caused by genetic or epigenetic alterations at two imprinting centers (ICs) in the 11p15.5 region. Delineation of the molecular defects is important for prognosis and predicting familial recurrence. We evaluated epigenetic alterations and potential epigenotype–phenotype correlations in Korean children with BWS.

Methods

Forty children with BWS with proven genetic or epigenetic defects in the 11p15.5 region were included. The phenotype was scored using the BWS consensus scoring system. Methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA), bisulfite pyrosequencing, a single-nucleotide polymorphism microarray, and CDKN1C sequencing were used for confirmative diagnosis.

Results

Patients met the criteria for genetic testing, with a mean clinical score of 5.4±2.0. Methylation alterations were consistent between MS-MLPA and bisulfite pyrosequencing in all patients. Twenty-six patients (65.0%) had IC2 loss of methylation (IC2-LoM), 11 (27.5%) had paternal uniparental disomy (patUPD), and one (2.5%) had IC1 gain of methylation. Macroglossia and external ear anomalies were more common in IC2-LoM than in patUPD, and lateralized overgrowth was more common in patUPD than in IC2-LoM (all P<0.05). Methylation levels at IC2 were inversely correlated with birth weight standard deviation score (r=–0.476, P=0.014) and clinical score (r=–0.520, P=0.006) in the IC2-LoM group.

Conclusions

Comprehensive molecular analysis of the 11p15.5 region revealed epigenotype–phenotype correlations in our BWS cohort. Bisulfite pyrosequencing can help clarify epigenotypes. Methylation levels were correlated with fetal growth and clinical severity in patients with BWS.

Development and evaluation of deep-learning measurement of leg length discrepancy: bilateral iliac crest height difference measurement

BACKGROUND

Leg length discrepancy (LLD) is a common problem that can cause long-term musculoskeletal problems. However, measuring LLD on radiography is time-consuming and labor intensive, despite being a simple task.

OBJECTIVE

To develop and evaluate a deep-learning algorithm for measurement of LLD on radiographs.

MATERIALS AND METHODS

In this Health Insurance Portability and Accountability Act (HIPAA)-compliant retrospective study, radiographs were obtained to develop a deep-learning algorithm. The algorithm developed with two U-Net models measures LLD using the difference between the bilateral iliac crest heights. For performance evaluation of the algorithm, 300 different radiographs were collected and LLD was measured by two radiologists, the algorithm alone and the model-assisting method. Statistical analysis was performed to compare the measurement differences with the measurement results of an experienced radiologist considered as the ground truth. The time spent on each measurement was then compared.

RESULTS

Of the 300 cases, the deep-learning model successfully delineated both iliac crests in 284. All human measurements, the deep-learning model and the model-assisting method, showed a significant correlation with ground truth measurements, while Pearson correlation coefficients and interclass correlations (ICCs) decreased in the order listed. (Pearson correlation coefficients ranged from 0.880 to 0.996 and ICCs ranged from 0.914 to 0.997.) The mean absolute errors of the human measurement, deep-learning-assisting model and deep-learning-alone model were 0.7 ± 0.6 mm, 1.1 ± 1.1 mm and 2.3 ± 5.2 mm, respectively. The reading time was 7 h and 12 min on average for human reading, while the deep-learning measurement took 7 min and 26 s. The radiologist took 74 min to complete measurements in the deep-learning mode.

CONCLUSION

A deep-learning U-Net model measuring the iliac crest height difference was possible on teleroentgenograms in children. LLD measurements assisted by the deep-learning algorithm saved time and labor while producing comparable results with human measurements.