Pediatric and adolescent von Hippel-Lindau disease: tumor profiles, genotype-phenotype correlation and comparison with adults.

Purpose: Data on pediatric/adolescent von Hippel-Lindau (VHL) disease is sparse, and surveillance/management recommendations rely on expert opinions/extrapolations from adults. We aimed to characterize the childhood/adolescent VHL disease phenotype, compare it with adults, and identify genotype-phenotype correlations.

Methods: Retrospective review of children/adolescents (≤ 19 years) and adults with VHL disease from a single endocrine center (2000-2024). Only neoplasms diagnosed until age 19 years were included in the childhood/adolescent group and compared with the last follow-up of adults.

Results: Twenty-six children/adolescents (median age:15.5 years) were identified. By age 19 years, 81% had pheochromocytoma/paraganglioma (PPGL, of which 10% head-neck), 42% central nervous system hemangioblastoma (CNS-HB), 31% each retinal hemangioblastoma (RHB) and pancreatic neuroendocrine tumor (PNET), and none endolymphatic sac tumor/renal cell carcinoma. At diagnosis, all PPGLs were symptomatic (median size 4.5 cm). CNS-HBs showed female preponderance, with high disease burden (60% symptomatic, 50% synchronous) and surgical requirement by 19 years of age. 2/8 pediatric patients needed surgery for symptomatic PNET before recommended surveillance initiation age (15 years). Two children/adolescents developed polycythemia during follow-up. Compared to adults (n = 39), pediatric/adolescent PPGL patients had significantly higher plasma free-normetanephrine/bilateral/extra-adrenal disease and 8.3-fold higher operative-site recurrence. Neoplasm frequency and other characteristics by 19 years resembled adults. Childhood/adolescent PPGLs occurred predominantly (16/17), and PNETs exclusively with missense variants. Codon 167 missense variants were associated with synchronous bilateral pheochromocytomas.

Conclusion: In the largest Asian study describing children/adolescents with VHL disease, we report severe pediatric/adolescent phenotype comparable to adults, need for childhood/adolescent HNPGL, and earlier PNET surveillance, particularly with missense variants.