Abstract PO011: Proteomic profiling of childhood liver cancer: identification of novel diagnostic and prognostic biomarkers

Abstract Childhood liver cancers hepatoblastoma (HB) and hepatocellular carcinoma (HCC) are rare diseases but with a rising incidence. HEMNOS (Hepatocellular malignant neoplasm, not otherwise specified) is a recent entity with histopathological features of HB and HCC. Current chemotherapy treatments are effective to shrink tumor before surgery, nonetheless can cause severe lifelong adverse effects and are not effective for patients with aggressive and metastatic HB or HCC (~20% die due to the disease). The HB and HCC diagnosis is key to assign therapeutic regimens in the ongoing Pediatric Hepatic International Tumor Trial (PHITT); however, differential diagnosis for some patients is challenging due to the lack of specific biomarkers. Nowadays, there is an urgent need to identify diagnostic and prognostic biomarkers to improve the clinical management of childhood liver cancer. Herein, we aimed to uncover the proteomic profiles of different types of childhood liver cancer and identify new biomarkers for improving tumor diagnosis and early detection of the most aggressive cases. For this, 99 frozen tissue samples from 71 patients (mean age: 54.5 months, 60.6% boys, 40.9% metastasis, 23.7% deaths) including 70 primary tumors (57 HB, 8 HCC and 5 HEM-NOS), 22 non-tumors and 7 metastases were analyzed by label free mass spectrometry analysis. A total of 5,417 proteins were sequenced in the different samples. Through supervised analysis, we identified a total of 1302 differential expressed proteins in tumors as compared to non-tumor livers (FC +2 and FDR <0.01). A total of 246 proteins of them (19%) were commonly differently expressed in all tumors types; having HEMNOS a mixture of HB and HCC protein profiling. A panel of potential diagnostic biomarkers including 179, 120 and 81 upregulated proteins (FC >2) was defined for HB, HCC and HEM-NOS, respectively. Hierarchical unsupervised clustering and principal component analysis showed two main cluster of tumors: one including HB and HEM-NOS and another one with a mixture of HB, HEM-NOS and HCC. Interestingly, this second cluster was associated with clinical and molecular features of aggressive tumors such as multifocality (p = 0.029), Epigenetic Epi-CB subclass (p = 0.006), high-risk Molecular Risk Stratification (p = 0.009) and poor survival (log rank test=0.012). A total of 2082 proteins were found to be differently expressed between the two prognostic clusters (FC +2 and FDR < 0.01); the GEO enrichment analysis revealed that these proteins were significantly enriched in post-transcriptional gene silencing and RNA splicing mechanisms (FDR <10−7). In addition, a 25-protein signature associated to liver cancer prognosis was defined (FC>6 and FDR < 10−14). In conclusion, we identified a list of potential diagnostic and prognostic protein biomarkers that after a validation in large patient cohort, could be used to improve the clinical management of childhood liver cancer. Citation Format: Álvaro Del Río-Álvarez, Juan Carrillo-Reixach, Laura Royo, Montse Domingo-Sàbat, Mikel Azkargorta, Roland Kapler, Stefano Cairo, Christian Vokuhl, Ronald de Krijger, Rita Alaggio, Marta Garrido, Gabriela Guillen, Constantino Sábado, Laura Guerra, Francisco Hernandez, Maria Elena Mateos, Manuel López-Satamaría, Barbara Torres, Maria Pilar Abad, Bajčiová Viera, Piotr Czauderna, Marie Annick Buendia, Felix Elortza, Keith Wheatley, Bruce Morland, Carolina Armengol. Proteomic profiling of childhood liver cancer: identification of novel diagnostic and prognostic biomarkers [abstract]. In: Proceedings of the AACR Special Conference: Advances in the Pathogenesis and Molecular Therapies of Liver Cancer; 2022 May 5-8; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2022;28(17_Suppl):Abstract nr PO011.