12 | Integrated molecular and immunophenotypic profiling of uveal melanoma: prognostic insights and biological perspectives

Francesco Fortarezza¹, Giovanni Zarrilli², Giada Munari¹, Valentina Angerilli³, Mariangela Balistreri¹, Luisa Piccin⁴, Valentina Salizzato⁴, Jacopo Pigozzo⁴, Giulia Midena⁵, Raffaele Parrozzani⁶, Valentina Guarneri^4|7, Edoardo Midena^5|6, Marta Sbaraglia^1|8, Matteo Fassan^8|9, Angelo Paolo Dei Tos^1|8 | ¹Surgical Pathology and Cytopathology Unit, Department of Integrated Diagnostics. Azienda Ospedale-Università Padova, Padua; ²Surgical Pathology Unit, Azienda Ulss 3 Serenissima, Ospedale dell'Angelo, Venezia; ³Department of Surgical Pathology, Azienda ULSS2 Marca Trevigiana, Treviso; ⁴Medical Oncology 2, Veneto Institute of Oncology, IOV-IRCCS, Padua; ⁵IRCCS-Fondazione Bietti, Rome; ⁶Department of Ophthalmology, University of Padova, Padova; ⁷Department of Surgery, Oncology and Gastroenterology, University of Padua, Padua; ⁸Department of Medicine, University of Padua School of Medicine, Padua; ⁹Veneto Institute of Oncology, IOV-IRCCS, Padua, Italy.

Background: Uveal melanoma (UM) is the most common primary intraocular malignancy in adults, characterized by distinct histopathological and molecular features and often associated with poor prognosis due to its high metastatic potential. While histology, BAP1 status, and chromosomal changes are established prognostic markers, integration of morphological, immunophenotypic, and molecular data is evolving.

Methods: We retrospectively analysed 84 UM cases using an integrated approach combining histological classification, immunohistochemical profiling (including BAP1, p53, CD3, CD8, PD-L1, and mismatch repair proteins), and targeted next-generation sequencing with a 63-gene panel. Tissue microarrays were used for immunophenotyping, and mutation data were stratified by prognostic outcomes.

Results: Most tumours were localized to the choroid and predominantly exhibited spindle-cell morphology. Mutations in GNAQ or GNA11 were identified in 83% of sequenced cases. Loss of BAP1 expression correlated with epithelioid histology and denser T-cell infiltration, whereas tumours consistently lacked PD-L1 expression. Aberrant p53 staining was more frequent in spindle-cell tumours, though TP53 mutations were rare, suggesting functional inactivation through other mechanisms. Notably, mutations typically associated with cutaneous melanomas (e.g., BRAF, KIT, CDKN2A) were also detected, particularly in iris melanoma, suggesting site-specific molecular convergence. Additional recurrent alterations were found in NOTCH1, PTEN, PIK3CA, and KDR, implicating the mTOR and VEGF signalling pathways. A high mutational burden (p<0.001), along with mutations in genes such as H3F3A (p<0.001), IDH2 (p=0.01), and JAK3 (p=0.05), were more frequent in tumors with poorer prognosis, supporting their potential role in disease aggressiveness.

Conclusions: This study highlights the heterogeneous molecular landscape of UM and underscores the importance of integrating histopathological and molecular data for improved prognostic stratification. The identification of potential therapeutic targets and atypical mutations suggests avenues for future research and tailored therapeutic strategies. Prospective studies with extended molecular panels and clinical outcomes are warranted to validate and expand upon these findings.