Hepatoblastoma (HB) is the most common type of liver cancer in children and accounts for less than 1% of all pediatric tumors. Despite improvements in the clinical management of HB patients over the last decades, there is still a significant portion of patients that face poor outcome. Mutations in the NFE2L2 or KEAP1 gene occur in approx. 5% of HB patients and have been associated with poor response to standard chemotherapy. In this study, we used CRISPR-Cas9 technology to create NFE2L2- and KEAP1-activated liver cancer models. The newly established NFE2L2- or KEAP1-mutated clones showed an increased NFE2L2 activity, increased cell growth and lower sensitivity towards cisplatin and doxorubicin treatment compared to the parental cell lines. RNA sequencing and integration of transcriptomic data into the drug prediction tool DrugSense allowed us to identify several potential drugs selectively targeting NFE2L2/KEAP1-activated liver cancer cells. Our study provides first insights into the molecular biology of NFE2L2 activated pediatric liver cancers and may lead to the development of novel treatment strategies for these patients.

Background and Aims: Management of hepatoblastoma (HB), the most frequent pediatric liver cancer, is based on surgical resection and perioperative chemotherapy regimens, commonly cisplatin. Here, we aimed to identify actionable targets in HB and assess the efficacy of molecular therapies in preclinical models of HB.Methods: Paired tumor and adjacent tissues from 31 resected HBs and a validation set of 50 HBs were analyzed at the transcriptomic, genomic and epigenomic level using RNAseq, SNP and methylation arrays, respectively. The main targetable driver in HB was identified by gene co-expression network analysis (GCN) and its overexpression was confirmed by qRT-PCR. The anti-tumor effect of driver inhibition with molecular therapies alone or in combination with cisplatin was assessed in cell lines, patient-derived HB organoids and in a HB xenograft murine model.Results: Seven network modules were significantly deregulated in tumors compared to non-tumoral samples, including IGF2 signaling pathway, cell cycle and survival and immune response. IGF2 overexpression (FC> 4 vs adjacent tissue) was identified as the top targetable HB driver (study cohort: 71%, 22/31; independent validation cohorts: 78% and 76%). IGF2high tumors displayed progenitor cell features and were significantly enriched in molecular classes with aggressive phenotypes and CTNNB1 mutations, while IGF2low tumors were enriched in inflammatory and TGF-β signaling. IGF2high tumors correlated with shorter recurrence-free survival after resection (median 34 months vs not reached for IGF2low; p = 0.02). IGF2 overexpression correlated in most cases (86%) with fetal promoter hypomethylation (50%), 11p15.5 loss of heterozygosity (LOH, 57%) or overexpression of miR483 (55%). Xentuzumab (anti-IGF1/2 mAb) alone or combined with cisplatin reduced proliferation and clonogenic capacity in IGF2high cell lines. The combination of xentuzumab and cisplatin exhibited synergistic effects in terms of cell viability in organoids derived from IGF2high human HBs. The combination treatment induced apoptosis and reduced IGF2 pathway activation in vitro. In mice (n = 13-14 per arm), this combination induced a significant decrease in the viable tumor volume (p < 0.01), extended survival compared to cisplatin alone (p < 0.05) and inhibited tumor angiogenesis (p < 0.05).Conclusion: IGF2 is an actionable driver in HB and its overexpression was associated with fetal promoter hypomethylation, LOH or miR483 overexpression. The combination of a mAb against IGF1/2 (xentuzumab) with cisplatin led to remarkable anti-tumoral effects in pre-clinical models, providing the rationale for exploring this regimen in IGF2high HB patients. Citation Format: Jordi Abril-Fornaguera, Laura Torrens, Juan Carrillo-Reixach, Alex Rialdi, Ugne Balaseviciute, Júlia Huguet-Pradell, Carla Montironi, Philipp Haber, Álvaro Del Río-Álvarez, Montserrat Domingo-Sàbat, Laura Royo, Nicholas Akers, Catherine E Willoughby, Judit Peix, Miguel Torres-Martin, Marc Puigvehi, Roser Pinyol, Stefano Cairo, Margaret Childs, Rudolf Maibach, Rita Alaggio, Piotr Czauderna, Bruce Morland, Bojan Losic, Vincenzo Mazzaferro, Ernesto Guccione, Daniela Sia, Carolina Armengol, Josep M Llovet. Identification of IGF2 as genomic driver and actionable therapeutic target in hepatoblastoma [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 PO009.

Despite considerable progress in understanding the biology and genetics of cancer, the development of effective therapies is hampered by the lack of sufficient experimental models that recapitulate the genetic diversity of this disease. The recourse to patient-derived xenograft (PDX) for the evaluation of new candidate anticancer drugs is becoming the gold standard in preclinical oncology. The faithful reproduction of patients’ cancer features, and the possibility to generate a large number of models that recapitulate patient population genetic heterogeneity, confer PDXs a critical added value in the evaluation of new candidate drugs. These improved models will hopefully contribute to decrease the attrition rate observed in clinical trials, thus far unacceptably high. Over the last 15 years, we have generated and characterized a collection of 200+ PDXs from different solid tumors that accurately reproduce the histological and molecular heterogeneity of the tumors of origin. This panel has allowed for the preclinical validation of several anticancer drugs that are now used in the clinic. Although being an indispensable tool to complete preclinical studies, the use of PDX in vivo systems for large-scale screening during early drug discovery is hampered by ethical, economical and throughput burdens limiting the number of test articles being tested. To address this problem, we developed a panel of PDX-derived cell lines (PDXDCs) that we propose as a time and cost-effective medium-throughput screening tool to profile the anti-cancer activity of early test compounds. To date, 50+ PDXDCs from various indications such as breast, lung, prostate and many others have been generated and tested for their response in vitro towards standards of care and targeted anti-cancer agents matching patient clinical management. Differently from standard cell line establishment, which is obtained by expansion of a cell clone that survives in vitro plating, our cell line development technology allows for maintenance of tumor cell population heterogeneity. PDXDCs RNA and exome sequencing data faithfully match the parental PDX features, and by modulating experimental parameters, such as 2D or 3D growth conditions, drug exposure duration and endpoint read-outs, we could phenocopy in vitro the corresponding PDXs’ sensitivities to chemotherapies. These results show our PDXDCs panel is a valuable in vitro platform for drug screening to help selecting drug candidates for further validation in parental PDX models in vivo. Citation Format: Olivier Déas, Léa Sinayen, Emilie Indersie, Kathleen Flosseau, Sophie Banis, Enora Le Ven, Jean-Gabriel Judde, Stefano Cairo. PDX-derived cell line platform for pharmacological screening and functional studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1637.

Hepatoblastoma (HB) is a pediatric malignancy with prevalence of 1:1 000 000 in Europe. Median age at diagnosis is one year. Current treatment entails chemotherapy (platinum-based alone or combined with doxorubicin) followed by surgery (tumor resection or liver transplantation). Survival rate of HB patients has improved; however, high-risk HB tumors are still difficult to treat with 20-30% of HB patients responding poorly to current treatment modalities. Furthermore, while current treatment modalities improve the prognosis, side effects of the chemotherapeutics have a significant effect on the life quality of HB survivors.

Although survival of hepatoblastoma patients has dramatically increased by combining preoperative chemotherapy and surgical tumor resection, drug resistance remains a huge challenge in the clinical management. Here, we integrated gene expression data of five responders and two non-responders into the pharmacologic perturbation prediction tool Connectivity Map and identified the anthelmintic mebendazole as a putative drug to circumvent chemoresistance in hepatoblastoma. Mebendazole treatment of cell lines grown from patient-derived xenografts resulted in a potent reduction of tumor cell growth in a dose-dependent manner. Moreover, mebendazole treatment resulted in a reduced colony formation capability, induction of apoptosis, and cell cycle arrest in G2/M phase, which was associated with blockage of microtubule formation. Consequently, RNA sequencing analyses confirmed the transcriptional downregulation of tubulins. Using a subcutaneous patient-derived xenograft transplantation model we found that mebendazole significantly reduced tumor growth in vivo. In conclusion, our results strongly support the clinical use of mebendazole in the treatment of chemoresistant hepatoblastoma.

Treatment of hepatoblastoma (HB) has drastically improved by refinements of surgical procedures and clinical risk stratification. However, identifying novel drug and molecular targets is still crucial due to the resistance towards conventional chemotherapy and its toxicity. Since the mutational frequency is extremely low in HB, targeting strategies based on genetic alterations is challenging, addressing the involvement of epigenetic modifications. In this project, we tested 11 compounds directed against epigenetic regulators on 10 HB cell lines and compared their efficacy to standard of care chemotherapeutics. Consequent viability assays revealed that the polycomb complex protein BMI-1 inhibitors PTC209 and PTC596 reduced HB cell growth in a dose-dependent manner. Moreover, reduction of colony formation capability, decrease of migration potential, induction of apoptotis, retardation of proliferation and 3-dimensional spheroid growth were observed in HB cells upon BMI-1 inhibition. Furthermore, the combination of BMI-1 inhibitors and cisplatin revealed a strong synergistic effect, suggesting BMI-1 inhibition as a potential target for HB therapy.

Background: In pediatric acute myeloid leukemia (AML) chemotherapy is the standard of care, but >25% of patients still relapse and after a disease recurrence the survival probability is extremely low (<50%). To ameliorate patients’ outcome there is an urgent need to discover new treatments. Nevertheless, pediatric drug development is extremely reduced by the need of a better understanding of the adverse event profile of adult cancer indications in children, by the lack of pediatric-specific formulations, and by the reduced number of pediatric AML patients that can be included in clinical trials. Thus, robust preclinical AML models to faithfully predict new drug efficacy is urgently needed to advance new drugs in clinical setting. Aims: This study aims to generate and characterize AML patient derived xenografts (PDXs) and accelerate the evaluation of innovative medicines for AML. Methods: We generated PDXs from primary AML samples by inoculating blasts in NSG mice and, when engrafted, in 3 consequent mice recipients (namely P0, P1 and P2-PDX). We characterized AML by immunophenotipic profile and by RNA and whole-exome sequencing (WES). According to somatic mutations and copy number alterations we determined AML clonal compositions. We selected drugs and performed drug testing in vivo, after the expansion of P3-PDXs. Results: We generated 22 AML-PDXs representing high-risk AML subtypes for genetic characterization harboring NUP98-NSD1, or NPM1-MLF1, or CBFA2T3-GLIS2, or FUS-ERG or KMT2A somatic translocations, or FLT3-ITD mutation. We monitored the AML associated immunophenotype in PDXs finding it was similar to that of the original AML. By WES we detected a consistent number of variants in each patients’ AML (ranging from 28 to 69), confirming an high AML intra-tumoral heterogeneity. Furthermore, we did not find any mutation recurrence among models, underlining an high inter-tumoral heterogeneity. In all models we tracked clonal evolution from patients’ AML to P2-PDX highlighting that most of the variants were maintained, with very few variants acquired during model development. Monitoring clonal dynamics we recognize a specific “founder” clone characterized by an average of 30 variants which are maintained up to P2 at the same allelic frequency, other small clones with average of 10 variants increasing the allelic frequencies in P2 and, in a restricted number of models, we observed that some clones were lost. By WES and transcriptome analysis we highlighted druggable mutations and pathways allowing the selection of novel targeted drugs. We screened their efficacy in vitro alone or combined with chemotherapic (Arabinoside) and biological agents (Venetoclax) by using AML ex vivo cells and mesenchymal stromal cells in a 3D co-culture system for exploring their synergy in reducing AML proliferation. Four selected drugs are under evaluation in AML-PDX models. Summary/Conclusion: We have created a series of paired AML and xenograft models for advancing pediatric AML therapeutics. Our models represent a concrete perspective for both, the identification of new variants and pathways involved in AML progression, and the possibility to perform novel drug screenings useful to increase AML drug portfolio.

Background:Diagnosis, treatment, minimal residual disease monitoring, and outcome of pediatric acute myeloid leukemia (AML) have made enormous progress during the past decade, although chemotherapy is still the pillar of pediatric treatment. Most of the emerged anti‐leukemic agents failed during experimentation, and one main limit in AML field is the inappropriateness of current pre‐clinical models used to study drug efficacy, reducing the advance of phase II and III clinical trials, especially for children.Aims:Set up and characterization of long term 3D‐AML cultures. Perform high throughput drug screening in vitro, and selection of best compounds to be then used in pre‐clinical AML‐PDX models. We would create a robust in vitro and in vivo pipeline to discover/reposition alternative treatments to improve AML children cure.Methods:The 3D structure is made up of engineered hydroxyapatite and collagen I to mimic endosteal bone marrow niche. We cultured mesenchymal stem cells derived from an AML patient (AML‐MSCs) together with its blasts. We studied 3D cultures also by using MSCs derived from healthy bone marrow donors (h‐MSCs), osteoblasts and endothelial cells. AML cells proliferation, immunophenotype and clonogenicity up to 21 days of 3D cultures have been analyzed. We characterized AML‐MSCs in comparison to h‐MSCs for proliferation rate, gene expression, secretome profile, osteogenic differentiation and anti‐inflammatory potential. We set up a drug targeting of the AML‐MSCs selecting agents among 480 compounds. We combined blasts and AML‐MSCs treatment in the 3D model. 3D‐AML cultures have been implanted in mice to generate robust pre‐clinical AML‐PDXs to study new combined treatments for AML.Results:We successfully set up 3D long‐term cultures of different primary AML (n = 20) and confirmed their proliferation up to 21 days. Clonogenic potential and immunophenotype of the original AML was also documented. We uncovered AML‐MSCs (n = 4) exhibiting an higher proliferation rate (p < 0.001) with respect to h‐MSCs (n = 2). AML‐MSCs (n = 8) were also found primed for osteogenic differentiation occurring after 7 days with respect to 21 days of h‐MSCs (n = 3, p < 0.01). Gene expression profile of AML‐MSCs (n = 15) showed 513 genes up‐regulated with respect to h‐MSCs (n = 6) mostly related to pathways involved in cell proliferation and chromosome instability. HUVEC tube formation assay suggested that AML‐MSCs did not exert anti‐inflammatory activity in vitro, this latter supported by a peculiar secretome defined by mass‐spectrometry. AML‐MSCs drug screening identified 17 out of 480 active compounds for reducing AML‐MSCs proliferation without toxicity over h‐MSCs and AML blasts. Pilot studies of the 3D‐combined targeting of AML‐MSCs and AML blasts (with targeted compounds for genetic lesions) showed this strategy being dramatically synergic (p < 0.05) in vitro. In vivo, we obtained 4 different AML‐PDXs with different genetic high‐risk characteristics in a period ranging from 3 to 7 months for the first engraftment, and a median time of 2 months from P1 to P3. Mice were shown to maintain the original AML immunophenotype and genetic of diagnosis.Summary/Conclusion:We confirm long‐term cultures of AML as a suitable model for more predictable drug testing in vitro. Our humanized 3D niche was also demonstrated to potentiate the original leukemia engraftment in NSG mice. We are confident this findings supporting a reliable tool for more robust pre‐clinical studies that will enhance novel strategies for pediatric AML treatment.

Abstract Pediatric patients with recurrent and refractory cancers are in most need for new treatments. This study developed patient-derived-xenograft (PDX) models within the European MAPPYACTS cancer precision medicine trial (NCT02613962). To date, 131 PDX models were established following heterotopical and/or orthotopical implantation in immunocompromised mice: 76 sarcomas, 25 other solid tumors, 12 central nervous system tumors, 15 acute leukemias, and 3 lymphomas. PDX establishment rate was 43%. Histology, whole exome and RNA sequencing revealed a high concordance with the primary patient’s tumor profile, human leukocyte-antigen characteristics and specific metabolic pathway signatures. A detailed patient molecular characterization, including specific mutations prioritized in the clinical molecular tumor boards are provided. Ninety models were shared with the IMI2 ITCC Paediatric Preclinical Proof-of-concept Platform (IMI2 ITCC-P4) for further exploitation. This new PDX biobank of unique recurrent childhood cancers provides an essential support for basic and translational research and new treatments development in advanced pediatric malignancies.

Background & AimsMicroRNAs are important genetic regulators of physiological and pathophysiological processes including cancer initiation and progression of hepatoblastoma, the most common liver tumour in childhood. We aimed to identify malignant and metastasis promoting effects of miR-492, a miRNA, previously reported to be overexpressed in metastatic hepatoblastoma. Furthermore, we intended to evaluate its diagnostic and prognostic potential. MethodsStable and transient overexpression of miR-492 in two liver tumour cell lines HepT1 and HUH7 was used to analyse features of metastatic tumour progression such as proliferation, anchorage-independent growth, migration and invasion. Via a mass spectrometry based proteomic screen, we investigated miRNA-492-dependent effects on proteome level and explored the underlying biology. One of the predicted target genes, CD44, was experimentally validated via luciferase assays. Diagnostic and prognostic properties of miR-492 were studied in hepatoblastoma tumour samples. ResultsWe show that miR-492 significantly enhances cell proliferation, anchorage-independent growth, migration and invasion of hepatoblastoma cells. We also identified and validated CD44, a transmembrane adhesion receptor for hyaluronan, as direct and functional target of miR-492. This miRNA has a strong direct impact on two CD44 isoforms (standard and v10). High miR-492 expression correlates with high-risk or aggressive tumours and further bears potential for predicting reduced event-free survival. ConclusionsWe identified miR-492 and its target CD44 as regulators of a number of biological features important for malignancy and metastasis. Furthermore, we demonstrated the diagnostic and prognostic potential of miR-492, a promising novel therapeutic target and biomarker for hepatoblastoma.

: 36 Purpose: FGFR1 amplification occurs in ~15% of ER+ human breast cancers.We investigated mechanisms by which FGFR1 amplification confers antiestrogen resistance to ER+ breast cancer. 38 Methods: ER+ tumors from patients treated with letrozole before surgery were subjected to Ki67 39 immunohistochemistry, FGFR1 FISH, and RNA-sequencing. ER+/ FGFR1 amplified breast cancer 40 cells and patient-derived xenografts (PDXs) were treated with FGFR1 siRNA or the FGFR tyrosine 41 kinase inhibitor lucitanib. Endpoints were cell/xenograft growth, FGFR1/ER  association by co- 42 immunoprecipitation and proximity ligation, ER genomic activity by ChIP-sequencing, and gene 43 expression by RT-PCR. 44 Results: ER+/FGFR1 amplified tumors in patients treated with letrozole maintained cell 45 proliferation (Ki67). Estrogen deprivation increased total and nuclear FGFR1 and FGF ligands 46 expression in ER+/ FGFR1- amplified primary tumors and breast cancer cells. In estrogen-free 47 conditions, FGFR1 associated with ER  in tumor cell nuclei and regulated the transcription of ER- 48 dependent genes. This association was inhibited by a kinase-dead FGFR1 mutant and by treatment 49 with lucitanib. ChIP-seq analysis of estrogen-deprived ER+/FGFR1 amplified cells showed binding 50 of FGFR1 and ERα to DNA. Treatment with fulvestrant and/or lucitanib reduced FGFR1 and ERα 51 binding to DNA. RNA-seq data from FGFR1 -amplified patients’ tumors treated with letrozole 52 showed enrichment of estrogen response and E2F target genes. Finally, growth of ER+/ FGFR1- 53 amplified cells and PDXs was more potently inhibited by fulvestrant and lucitanib combined than each drug alone.