DOI: 10.1001/jamadermatol.2026.2933 ISSN: 2168-6068

Cherry Angiomas in Individuals With Neurofibromatosis Type 1

Laura Fertitta, Eric Pasmant, Christina Bergqvist, Houda Ayache, Sabine Moryousef, Salah Ferkal, Philippe Le Corvoisier, Ariane Lunati-Rozie, Benoit Funalot, Tu-Anh Duong, Caroline Barau, Sophie Kaltenbach, Patrick Villarese, Fanny Morice-Picard, Christine Labrèze, Sorilla Prey, Julien Masliah Planchon, Laurence Pacot, Dominique Vidaud, Eduard Serra, Eric Legius, Michael Frost, Herb Sarnoff, Léa Pechtimaldjian, Thomas Bettuzzi, Khaled Ezzedine, Cyrielle Robe, Piotr Topilko, François Moisan, Nicolas Ortonne, Pierre Wolkenstein
Show PDF Cite

Importance

Neurofibromatosis type 1 (NF1) is a multisystem, tumor predisposition syndrome in which vascular manifestations, including cutaneous lesions, remain poorly characterized. Cherry angiomas (CAs) have been sporadically reported in NF1, but their prevalence and biological basis are unknown.

Objective

To assess the association between NF1 and CAs and define the histopathologic, cellular, and genetic mechanisms underlying NF1-associated CAs.

Design, Setting, and Participants

This prospective, comparative, cross-sectional epidemiological study was conducted from October 2020 to March 2021 at a French national referral center for neurofibromatoses within the dermatology department at Henri-Mondor University Hospital, Créteil, France. It was followed by integrated histopathologic, genomic, and cell-specific molecular analyses. The epidemiological analysis included individuals 15 years or older with confirmed NF1 and controls without NF1. Pathophysiological analyses were conducted on CAs from individuals with NF1 and healthy volunteers following written consent. Data were analyzed in 2022.

Main Outcomes and Measures

Outcomes included the prevalence of CAs; age-adjusted and sex-adjusted odds ratios; histopathologic features; detection of somatic NF1 second-hit variants; identification of co-occurring oncogenic variants; localization of NF1 loss within vascular cell populations; and evidence of rat sarcoma–mitogen-activated protein kinase pathway activation.

Results

Of 259 participants (125 female individuals [48.3%]), 102 (39.4%) individuals had NF1 (median [range] age, 34 [18-70] years) and 157 (60.6%) were controls (median [range] age, 40 [15-91] years). CAs were more frequent in individuals with NF1 than controls (48% vs 18%; odds ratio, 4.26; 95% CI, 2.44-7.56) and occurred at a younger age. This association persisted after adjustment for age and sex and in propensity score–matched analyses. Somatic NF1 loss-of-function second hits were identified in 26 of 39 of NF1-associated CAs (67%) and none from controls, indicating biallelic NF1 inactivation. Comprehensive genomic profiling revealed frequent co-occurring activating variants, most commonly in GNAQ . Cell-specific sequencing showed that NF1 second hits arose predominantly in endothelial cells and telocytes, with higher variant allele frequencies in endothelial cells. Immunofluorescence demonstrated increased phosphoextracellular signal-regulated kinase signaling in these vascular cell populations.

Conclusions and Relevance

The results of this cross-sectional study suggest that CAs represent a frequent and previously unrecognized vascular manifestation of NF1, which is supported by epidemiological enrichment and mechanistic evidence of biallelic NF1 inactivation in vascular cells. These findings potentially expand the spectrum of NF1-associated neoplasms and establish CAs as a model for NF1-related vasculopathy.

More from our Archive