Patient-derived Organoid Models Research Articles

Inhibition of the transmembrane transporter ABCB1 overcomes resistance to doxorubicin in patient-derived organoid models of HCC.

Transarterial chemoembolization is the first-line treatment for intermediate-stage HCC. However, the response rate to transarterial chemoembolization varies, and the molecular mechanisms underlying variable responses are poorly understood. Patient-derived hepatocellular carcinoma organoids (HCCOs) offer a novel platform to investigate the molecular mechanisms underlying doxorubicin resistance. We evaluated the effects of hypoxia and doxorubicin on cell viability and cell cycle distribution in 20 patient-derived HCCO lines. The determinants of doxorubicin response were identified by comparing the transcriptomes of sensitive to resistant HCCOs. Candidate genes were validated by pharmacological inhibition. Hypoxia reduced the proliferation of HCCOs and increased the number of cells in the G0/G1 phase of the cell cycle, while decreasing the number in the S phase. The IC50s of the doxorubicin response varied widely, from 29nM to >1µM. Doxorubicin and hypoxia did not exhibit synergistic effects but were additive in some HCCOs. Doxorubicin reduced the number of cells in the G0/G1 and S phases and increased the number in the G2 phase under both normoxia and hypoxia. Genes related to drug metabolism and export, most notably ABCB1, were differentially expressed between doxorubicin-resistant and doxorubicin-sensitive HCCOs. Small molecule inhibition of ABCB1 increased intracellular doxorubicin levels and decreased drug tolerance in resistant HCCOs. The inhibitory effects of doxorubicin treatment and hypoxia on HCCO proliferation are variable, suggesting an important role of tumor-cell intrinsic properties in doxorubicin resistance. ABCB1 is a determinant of doxorubicin response in HCCOs. Combination treatment of doxorubicin and ABCB1 inhibition may increase the response rate to transarterial chemoembolization.

Comparison of Surgical and Colonoscopy Tissue to Establish Colorectal Patient-derived Organoids.

Patient-derived organoids (PDOs) are ex vivo models that retain the functions and characteristics of individualized source tissues, including a simulated tumor microenvironment. However, the potential impact of undiscovered differences between tissue sources on PDO growth and progression remains unclear. This study aimed to compare the growth and condition of PDO models originating from surgical resection and colonoscopy and to provide practical insights for PDO studies. Tissue samples and relevant patient clinical information were collected to establish organoid models. PDOs were derived from both surgical and colonoscopy tissues. The growth of the organoids, including their state, size, and success rate of establishment, was recorded and analyzed. The activity of the organoids at the end stage of growth was detected using calcein-AM fluorescence staining. The results showed that the early growth phase of 2/3 colonoscopy-derived organoids was faster compared to surgical PDOs, with a growth difference observed within 11-13 days of establishment. However, colonoscopy-derived organoids exhibited a diminished growth trend after this time. There were no significant differences observed in the terminal area and quantity between the two types of tissue-derived organoids. Immunofluorescence assays of the PDOs revealed that the surgical PDOs possessed a denser cell mass with relatively higher viability than colonoscopy-derived PDOs. In the establishment of colorectal patient-derived organoids, surgically derived organoids require a slightly longer establishment period, while colonoscopy-derived organoids should be passaged prior to growth inhibition to preserve organoid viability.

Modelling esophageal adenocarcinoma and Barrett's esophagus with patient-derived organoids.

Currently, esophageal adenocarcinoma (EAC) research is hindered by a dearth of adequate models to study this disease. Traditional cell line and genetically engineered mouse models are lacking in biological and physiological significance, whilst the inefficiency of patient-derived xenografts limit their potential applications. This review describes the landscape of EAC research using patient-derived organoids (PDOs). Here, we detail the methods of establishment and optimization of EAC PDO cultures, as well as current and prospective applications of these models. We further highlight a crucial knowledge gap in the mechanisms of EAC transformation from its precursor lesion, Barrett's esophagus (BE). As such, we also describe the culture requirements of BE PDOs and attempts to model tumorigenesis using PDO models.

Combined spatial Keratin expression profiles at the invasive front represent a prognostic classifier for head and neck cancer

Keratin proteins are intermediate filaments that provide mechanical support and guide a plethora of important functions in epithelial cells. Although Keratins can be effectively used to characterize and differentiate cell lineages in many epithelia, it is still unclear if combinations of Keratin types and their spatial expression patterns can aid prognostication of head and neck squamous cell carcinoma (HNSCC). Here, we have assessed a panel of Keratin markers in four HPV negative HNSCC patient-derived organoid (PDO) models and their corresponding primary patient tissue to define spatial expression profiles of basal and supra-basal Keratins during invasion in Collagen-I matrices. Based on these profiles we have devised a scoring classification consisting of the following profiles: (i) no expression (Negative), (ii) uniform expression (Homogenous), (iii) heterogeneous expression (Mosaic), (iv) expression in the tumor core (Core), and (v) expression at the invasive front (Edge). Subsequently, this 5-tier system was validated in a well-documented tissue cohort of 157 HNSCC patients. We observe that patients with a Keratin 13 (K13) Edge profile have a significantly better prognosis than patients with K13 Core expression or K13 Negative tumors. Interestingly, we also find that K14 mosaicism strongly correlates with a favorable prognosis. We show that the absence of K13 from the tumor edge, in combination with K14 homogenous expression is a strong biomarker for poor prognosis in HNSCC. In short, our work indicates that defining the spatial expression patterns of basal and supra-basal markers in invasive HNSCC can benefit patient prognostication.

Targeting the MCP-GPX4/HMGB1 Axis for Effectively Triggering Immunogenic Ferroptosis in Pancreatic Ductal Adenocarcinoma.

Induction of ferroptosis can inhibit cancer cells in vitro, however, the role of ferroptosis in treatment in vivo is controversial. The immunosuppressive cells activated by the ferroptotic tumor cells can promote the growth of residual tumor cells, hindering the application of ferroptosis stimulation in tumor treatment. In this study, a new strategy is aimed to be identified for effectively triggering immunogenic ferroptosis in pancreatic ductal adenocarcinoma (PDAC) and simultaneously stimulating antitumor immune responses. Toward this, several molecular and biochemical experiments are performed using patient-derived organoid models and a KPC mouse model (LSL-KrasG12D /+, LSL-Trp53R172H/+,Pdx-1-Cre). It is observed that the inhibition of macrophage-capping protein (MCP) suppressed the ubiquitin fold modifier (UFM)ylation of pirin (PIR), a newly identified substrate of UFM1, thereby decreasing the transcription of GPX4, a marker of ferroptosis, and promoting the cytoplasmic transportation of HMGB1, a damage-associated molecular pattern. GPX4 deficiency triggered ferroptosis, and the pre-accumulated cytosolic HMGB1 is released rapidly. This altered release pattern of HMGB1 facilitated the pro-inflammatory M1-like polarization of macrophages. Thus, therapeutic inhibition of MCP yielded dual antitumor effects by stimulating ferroptosis and activating antitumor pro-inflammatory M1-like macrophages. The nanosystem developed for specifically silencing MCP is a promising tool for treating PDAC.

Abstract LB450: Mechanisms of acquired resistance to AKT inhibitor capivasertib in AKT1 mutant patient derived breast cancer models

Abstract Capivasertib, a pan-AKT inhibitor, has recently been approved in combination with fulvestrant to treat ER+ HER2- breast cancer patients with one or more of PI3KCA/AKT1/PTEN alterations. However, as with many targeted therapies acquired drug resistance remains a challenge. In this study, we investigated the mechanisms of acquired capivasertib resistance in AKT1 mutant breast cancer models. We established two estrogen receptor positive (ER+) AKT1 E17K mutant breast cancer patient derived organoids (PDOs), that were sensitive to capivasertib in vitro. PDOs were chronically exposed to 1µM capivasertib. PDOs were sequenced (DNA & RNA) at multiple timepoints of treatment (3 days, 4 and 9 weeks after development of resistance). Using CRISPR-Cas9 editing we also engineered CAMA1 (ER+) cells to express AKT1 E17K, and generated derived resistant AKT1 mutant and parental models in response to continuous capivasertib exposure. CAMA1 AKT1 mutant and parental resistant were screened with an siRNA Kinome library supplemented with genes from the PDO RNA-sequencing. Phosphoproteomics was performed on capivasertib resistant E17K models in the presence and absence of capivasertib. PDO models persisting after either 4 or 9 weeks of capivasertib treatment showed no acquisition of genetic driver events, or changes in AKT copy number. Resistant organoids at 9 weeks treatment showed increased expression of ERBB3 (HER3), EGFR3 and LPAR5 as well as increased expression of MAPK signalling components. In siRNA screens on CAMA1 derived resistant models, siRNA targeting ERBB2, CDK4, PLK4 and CHEK1, ESR1 and LPAR5 modulated sensitivity to capivasertib. Analysis of phosphoproteins in capivasertib resistant models revealed reactivation of mTORC1 signalling despite sustained AKT inhibition, with dramatic upregulation of PDK1 and mTORC2. Consistent with the siRNA screens, drug combinations identified fulvestrant as a synergistic agent with capivasertib in the resistant E17K models. Sensitivity to mTOR inhibition was maintained in capivasertib resistant models, with mTOR inhibition re-sensitizing cells to capivasertib. In these models, acquired resistance to capivasertib occurs through upregulation of ER signalling and rewiring of pathway signalling to bypass AKT inhibition and reactivate mTORC1. Targeting these mechanisms could prolong the use of capivasertib, with future work focusing on validating these mechanisms and re-sensitizing drug combinations in other AKT1 mutant models & PDOs. Citation Format: Sarah Mearns, Alex Pearson, Li-Xuan Sim, Rosalind Cutts, Prithika Sritharan, Heena Shah, Aditi Gulati, Stuart Williamson, Simon Barry, Elza De-Bruin, Nicholas C. Turner. Mechanisms of acquired resistance to AKT inhibitor capivasertib in AKT1 mutant patient derived breast cancer models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB450.

Human patient derived organoids: an emerging precision medicine model for gastrointestinal cancer research.

Gastrointestinal cancers account for approximately one-third of the total global cancer incidence and mortality with a poor prognosis. It is one of the leading causes of cancer-related deaths worldwide. Most of these diseases lack effective treatment, occurring as a result of inappropriate models to develop safe and potent therapies. As a novel preclinical model, tumor patient-derived organoids (PDOs), can be established from patients' tumor tissue and cultured in the laboratory in 3D architectures. This 3D model can not only highly simulate and preserve key biological characteristics of the source tumor tissue in vitro but also reproduce the in vivo tumor microenvironment through co-culture. Our review provided an overview of the different in vitro models in current tumor research, the derivation of cells in PDO models, and the application of PDO model technology in gastrointestinal cancers, particularly the applications in combination with CRISPR/Cas9 gene editing technology, tumor microenvironment simulation, drug screening, drug development, and personalized medicine. It also elucidates the ethical status quo of organoid research and the current challenges encountered in clinical research, and offers a forward-looking assessment of the potential paths for clinical organoid research advancement.

Nanodrug modified with engineered cell membrane targets CDKs to activate aPD-L1 immunotherapy against liver metastasis of immune-desert colon cancer

Immunotherapy based on the PD-1/PD-L1 axis blockade has no benefit for patients diagnosed with colon cancer liver metastasis (CCLM) for the microsatellite stable/proficient mismatch repair (MSS/pMMR)) subtype, which is known as an immune-desert cancer featuring poor immunogenicity and insufficient CD8+ T cell infiltration in the tumor microenvironment. Here, a multifunctional nanodrug carrying a cyclin-dependent kinase (CDK)1/2/5/9 inhibitor and PD-L1 antibody is prepared to boost the immune checkpoint blockade (ICB)-based immunotherapy against MSS/pMMR CCLM via reversing the immunosuppressive tumor microenvironment. To enhance the MSS/pMMR CCLM-targeting efficacy, we modify the nanodrug with PD-L1 knockout cell membrane of this colon cancer subtype. First, CDKs inhibitor delivered by nanodrug down-regulates phosphorylated retinoblastoma and phosphorylated RNA polymerase II and meanwhile arrests the G2/M cell cycle in CCLM to promote immunogenic signal release, stimulate dendritic cell maturation, and enhance CD8+ T cell infiltration. Moreover, CDKi suppresses the secretion of immunosuppressive cytokines in tumor-associated myeloid cells sensitizing ICB therapy in CCLM. Notably, the great efficacy to activate immune responses is demonstrated in the patient-derived xenograft model and the patient-derived organoid model as well, revealing a clinical application potential. Overall, our study represents a promising therapeutic approach for targeting liver metastasis, remolding the tumor immune microenvironment (TIME), and enhancing the response of MSS/pMMR CCLM to boost ICB immunotherapy.

Stem cell dynamics and cellular heterogeneity across lineage subtypes of castrate-resistant prostate cancer.

To resist lineage-dependent therapies such as androgen receptor inhibition, prostate luminal epithelial adenocarcinoma cells often adopt a stem-like state resulting in lineage plasticity and phenotypic heterogeneity. Castrate-resistant prostate adenocarcinoma can transition to neuroendocrine (NE) and occasionally to amphicrine, co-expressed luminal and NE, phenotypes. We developed castrate-resistant prostate cancer (CRPC) patient-derived organoid models that preserve heterogeneity of the originating tumor, including an amphicrine model displaying a range of luminal and NE phenotypes. To gain biological insight and to identify potential treatment targets within heterogeneous tumor cell populations, we assessed the lineage hierarchy and molecular characteristics of various CRPC tumor subpopulations. Transcriptionally similar stem/progenitor (St/Pr) cells were identified for all lineage populations. Lineage tracing in amphicrine CRPC showed that heterogeneity originated from distinct subclones of infrequent St/Pr cells that produced mainly quiescent differentiated amphicrine progeny. By contrast, adenocarcinoma CRPC progeny originated from St/Pr cells and self-renewing differentiated luminal cells. Neuroendocrine prostate cancer (NEPC) was composed almost exclusively of self-renewing St/Pr cells. Amphicrine subpopulations were enriched for secretory luminal, mesenchymal, and enzalutamide treatment persistent signatures that characterize clinical progression. Finally, the amphicrine St/Pr subpopulation was specifically depleted with an AURKA inhibitor, which blocked tumor growth. These data illuminate distinct stem cell (SC) characteristics for subtype-specific CRPC in addition to demonstrating a context for targeting differentiation-competent prostate SCs.

Sialyl-Tn serves as a potential therapeutic target for ovarian cancer

BackgroundOvarian cancer remains the deadliest of the gynecologic cancers in the United States. There have been limited advances in treatment strategies that have seen marked increases in overall survival. Thus, it is essential to continue developing and validating new treatment strategies and markers to identify patients who would benefit from the new strategy. In this report, we sought to further validate applications for a novel humanized anti-Sialyl Tn antibody-drug conjugate (anti-STn-ADC) in ovarian cancer.MethodsWe aimed to further test a humanized anti-STn-ADC in sialyl-Tn (STn) positive and negative ovarian cancer cell line, patient-derived organoid (PDO), and patient-derived xenograft (PDX) models. Furthermore, we sought to determine whether serum STn levels would reflect STn positivity in the tumor samples enabling us to identify patients that an anti-STn-ADC strategy would best serve. We developed a custom ELISA with high specificity and sensitivity, that was used to assess whether circulating STn levels would correlate with stage, progression-free survival, overall survival, and its value in augmenting CA-125 as a diagnostic. Lastly, we assessed whether the serum levels reflected what was observed via immunohistochemical analysis in a subset of tumor samples.ResultsOur in vitro experiments further define the specificity of the anti-STn-ADC. The ovarian cancer PDO, and PDX models provide additional support for an anti-STn-ADC-based strategy for targeting ovarian cancer. The custom serum ELISA was informative in potential triaging of patients with elevated levels of STn. However, it was not sensitive enough to add value to existing CA-125 levels for a diagnostic. While the ELISA identified non-serous ovarian tumors with low CA-125 levels, the sample numbers were too small to provide any confidence the STn ELISA would meaningfully add to CA-125 for diagnosis.ConclusionsOur preclinical data support the concept that an anti-STn-ADC may be a viable option for treating patients with elevated STn levels. Moreover, our STn-based ELISA could complement IHC in identifying patients with whom an anti-STn-based strategy might be more effective.

A novel dihydroacridine derivative targets epidermal growth factor receptor-expressing cancer cells in vitro and in vivo.

Small molecules are considered a source of novel medicines targeting carcinogenic intracellular pathways including epidermal growth factor receptor (EGFR) signaling. The main goal of the study is to assess whether LHT-17-19 could be considered an effective target molecule against EGFR-expressing tumor cells in silico, in vitro, and in vivo. This was an in vivo, ex vivo, and in vivo experimental study. LHT-17-19 affinity to EGFR's kinase domain was assessed by the ligand's molecular docking. EGFR-expressing Hs746T human gastric cancer cell culture and patient-derived organoid (PDO) model of EGFR-positive breast cancer (BC) were used for in vitro assessment of the molecule anticancer property. IC50 and GI50 indexes were estimated using MTT- and MTS-based tests, respectively. Anticancer activity of LHT-17-19 against EGFR-expressing mutant lung carcinoma was studied on patient-derived xenograft (PDX) model established in 10 humanized BALB/c male mice. Continuous variables were presented as a mean ± standard deviation. Intergroup differences were assessed by two-way t-test. Kaplan-Meier's curves were used for survival analysis. High affinity of LHT-17-19 for the EGFR kinase domain with dG score -7.9 kcal/mol, EDoc-5.45 kcal/mol, and Ki 101.24 uM was due to intermolecular π-σ bonds formation and the ligand intramolecular transformation. LHT-17-19 induced anti-EGFR-expressing gastric cancer cells cytotoxicity with IC50 0.32 µM (95% confidence interval [CI] 0.11-0.54 µM). The derivative inhibited growth of EGFR-expressing BC PDO with GI50 16.25 µM (95% CI 4.44-28.04 µM). 2 mg/kg LHT-17-19 intravenously daily during 7 days inhibited PDX tumor growth and metastatic activity, prolonged animals' survival, and eliminated EGFR-mutant lung cancer cells from residual tumor's node. LHT-17-19 may be considered a molecular platform for further search of promising molecules, EGFR-expressing cancer cell inhibitors.

Research autopsy programmes in oncology: shared experience from 14 centres across the world.

While there is a great clinical need to understand the biology of metastatic cancer in order to treat it more effectively, research is hampered by limited sample availability. Research autopsy programmes can crucially advance the field through synchronous, extensive, and high-volume sample collection. However, it remains an underused strategy in translational research. Via an extensive questionnaire, we collected information on the study design, enrolment strategy, study conduct, sample and data management, and challenges and opportunities of research autopsy programmes in oncology worldwide. Fourteen programmes participated in this study. Eight programmes operated 24 h/7 days, resulting in a lower median postmortem interval (time between death and start of the autopsy, 4 h) compared with those operating during working hours (9 h). Most programmes (n = 10) succeeded in collecting all samples within a median of 12 h after death. A large number of tumour sites were sampled during each autopsy (median 15.5 per patient). The median number of samples collected per patient was 58, including different processing methods for tumour samples but also non-tumour tissues and liquid biopsies. Unique biological insights derived from these samples included metastatic progression, treatment resistance, disease heterogeneity, tumour dormancy, interactions with the tumour micro-environment, and tumour representation in liquid biopsies. Tumour patient-derived xenograft (PDX) or organoid (PDO) models were additionally established, allowing for drug discovery and treatment sensitivity assays. Apart from the opportunities and achievements, we also present the challenges related with postmortem sample collections and strategies to overcome them, based on the shared experience of these 14 programmes. Through this work, we hope to increase the transparency of postmortem tissue donation, to encourage and aid the creation of new programmes, and to foster collaborations on these unique sample collections. © 2024 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Establishing conditions for the generation and maintenance of estrogen receptor-positive organoid models of breast cancer

Patient-derived organoid models of estrogen receptor-positive (ER+) breast cancer would provide a much-needed tool to understand drug resistance and disease progression better. However, the establishment and long-term maintenance of ER expression, function, and response in vitro remains a significant challenge. Here, we report the development of an ER+ breast tumor organoid medium (BTOM-ER) that conserves ER expression, estrogen responsiveness, and dependence, as well as sensitivity to endocrine therapy of ER+ patient-derived xenograft organoids (PDXO). Our findings demonstrate the utility of subtype-specific culture conditions that better mimic the characteristics of the breast epithelial biology and microenvironment, providing a powerful platform for investigating therapy response and disease progression of ER+ breast cancer.

Establishing mesothelioma patient-derived organoid models from malignant pleural effusions

ObjectivesPleural mesothelioma is a cancer arising in the cells that line the lungs and chest wall with poor survival and poor response to first-line therapy. Organoid models of cancer can faithfully recapitulate the genetic and histopathological characteristics of individualized tumors and have potential to be used for precision medicine, however methods of establishing patient-derived mesothelioma organoids have not been well established in the published literature. Materials and methodsLong-term mesothelioma patient-derived organoids were established from ten malignant pleural effusion fluids. Mesothelioma patient-derived organoids were compared to the corresponding biopsy tissue specimens using immunohistochemistry labelling for select diagnostic markers and the TruSight Oncology-500 sequencing assay. Cell viability in response to the chemotherapeutic drug cisplatin was assessed. ResultsWe established five mesothelioma patient-derived organoid cultures from ten malignant pleural effusion fluids collected from nine individuals with pleural mesothelioma. Mesothelioma patient-derived organoids typically reflected the histopathological and genomic features of patients’ matched biopsy specimens and displayed cytotoxic sensitivity to cisplatin in vitro. ConclusionThis is the first study of its kind to establish long-term mesothelioma organoid cultures from malignant pleural effusions and report on their utility to test individuals’ chemotherapeutic sensitivities ex vivo.

Abstract 535: Resistant PDO models: Unraveling drug resistance mechanisms for discovering new drug targets

Abstract Resistance to cancer treatment remains a significant obstacle in curing patients with cancer. Despite recent advancements in novel drug targets for cancer therapeutics, patients often develop treatment resistance even after initial disease remission, leading to tumor relapse. To investigate the mechanisms responsible for developing resistance to standard-of-care therapies, it is essential to gain insights into the stepwise progression leading to complete resistance. Recent research has revealed that cancer cells can evade the effects of chemotherapy and targeted therapy by entering a reversible and slow proliferation state known as the drug-tolerant persister (DTP) state. This DTP state allows cancer cells to survive drug therapy long enough for additional mechanisms of acquired drug resistance to develop. HUB Organoids® or patient-derived organoid (PDO) models are derived from patient tissues, making them more physiologically relevant than traditional cell lines. Thus, studying drug persistence and resistance in PDO models is more likely to reflect in vivo conditions. This study demonstrates the successful development of persistent and resistant CRC PDO models to standard-of-care therapies. All models were characterized using whole exome sequencing and RNA sequencing to confirm their persistent or resistant state and to identify possible pathways involved in developing therapy resistance, as well as potential new drug targets. In conclusion, developing persistent and resistant PDO models is crucial for understanding drug resistance mechanisms, enabling the development of more effective and personalized cancer treatments, and advancing our overall knowledge of how resistance develops in cancer patients. Citation Format: Yasmine Abouleila, Mayke Doorn, Roel Verkerk, Joris Maas, Carla Verissimo, Sylvia Boj. Resistant PDO models: Unraveling drug resistance mechanisms for discovering new drug targets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 535.

Abstract 6006: Modeling antibody drug conjugate sensitivity using urothelial carcinoma patient-derived models

Abstract Introduction: Enfortumab Vedotin (EV), an antibody-drug conjugate (ADC) targeting Nectin-4, has emerged as a new standard-of-care for patients with metastatic urothelial cancer (UC). Promising results have also been observed in bladder cancer (BLCA) patients with ADCs targeting HER2. To identify predictive biomarkers of ADC sensitivity, we characterized a large and expanding biobank of patient-derived organoid (PDO) and xenograft (PDX) models for Nectin-4 and HER2 expression, and Nectin-4 and HER2-targeted ADC sensitivity. Methods: We generated 19 PDOs and 34 PDXs that reflect the genomic and biologic heterogeneity of human UC. A subset of these models was further characterized using a multiplatform approach, including bulk and single-cell DNA and RNA sequencing and immunoblot for Nectin-4 and HER2 expression. We leveraged these models to study the biologic determinants of sensitivity to EV and the HER2-targeted ADC trastuzumab deruxtecan (T-DXd). PDX models were treated with EV (5 mg/kg IV on day 1, 3, 6) or T-DXd (10 mg/kg IV on day 1, 21, 42) or vehicle only as control. We also studied sensitivity to T-DXd and EV, as well as to their cytotoxic payloads (exatecan and vedotin) in vitro using paired PDO models. Results: We observed high degree of variability of HER2 and Nectin-4 expression across UC PDOs and PDXs. In PDO models, sensitivity to EV correlated strongly with Nectin-4 expression, whereas HER2 expression was not strongly correlated with sensitivity to T-DXd. We also noted a poor correlation between Nectin-4 expression in PDOs and PDXs derived from the same patient. A subset of models, for example, SMBO-104 (ERBB2 wildtype with low HER2 expression), exhibited significantly greater sensitivity to EV than to T-DXd. Conversely, T-DXd was significantly more active than EV in SMBO-170, a model with ERBB2 amplification (~40 copies) and HER2 3+ expression. Sensitivity to exatecan, an analogue of the T-DXd payload deruxtecan, as measured by inhibition of cell proliferation and induction of apoptosis, was more predictive of T-DXd sensitivity than HER2 expression level. Conclusion: In this study, we generated and biologically characterized UC patient-derived models with the goal of identifying predictive biomarkers of sensitivity to Nectin-4 and HER2-targeted ADCs. We observed significant heterogeneity of HER2 and Nectin-4 expression across models. Models were identified in which T-DXd was superior to EV, and vice versa. Sensitivity to the T-DXd cytotoxic payload was more predictive of PDX response than HER2 expression levels, with significant tumor regression observed with T-DXd in BLCA models with low HER2 expression. Our data suggest that clinical trials of HER2-targeted ADCs such as T-DXd should not be restricted to patients with HER2 overexpression and that novel biomarkers will be needed to identify those UC patients most likely to benefit from FDA-approved and investigational ADCs. Citation Format: Xinran Tang, Ziyu Chen, Jasmine Thomas, Karan Nagar, John Christin, Naryan Rustgi, Sizhi Gao, Elisa de Stanchina, Jonathan Coleman, Michael Shen, Hikmat Al-Ahmadie, Gopa Iyer, Kwanghee Kim, David Solit. Modeling antibody drug conjugate sensitivity using urothelial carcinoma patient-derived models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6006.

Abstract 4248: Patient-derived MTAP-deleted bladder cancer organoid model: A unique platform for drug development

Abstract Introduction: Approximately 30% of advanced bladder cancer (BCa) tumors exhibit MTAP gene deletion, which has been associated with luminal subtype BCa, poor response to chemotherapy, and shorter progression to metastatic disease (De Souza 2023). MTAP deletion has been proposed as a potential biomarker for response to pemetrexed and an emerging class of PRMT5 and MAT2A inhibitors in development for MTAP-deleted tumors. Here, we describe the first patient-derived 3D model of BCa with MTAP deletion and its translational potential as a drug sensitivity platform. Methods: A patient (pt) with metastatic BCa harboring MTAP deletion underwent a pelvic lymph node biopsy. A fresh tumor sample was processed under collagenase/dispase solution for 30 minutes and filtered through a 70-um cell strainer. Cells were cultures in low attachment plates with MammoCult™ Basal Medium (Stem Cell Technologies). After 5 passages and 3D spheroid formation, immunostaining for cytokeratin (CK) 5 and 6, GATA3, CK20, and Ki67 staining were performed to confirm the urothelial origin. The MTAP-del was confirmed by a western blot of the organoid specimen and compared to other BLCa cell lines (UMUC, RT4, J82, and 5637). The viability assays were performed using the CellTiter-Glo® 3D Cell kit. Next-generation sequencing (NGS) of primary BCa tumor revealed deletion of MTAP, CDKN2A, CDKN2B, and mutations in p53, TERT, and KDM6A. Results: The patient organoid sample was positive for the luminal CK5/6 markers and negative for GATA3. Ki67 and CK20 were primarily expressed in the nucleus of the same sample. The organoid retained morphologic and immunohistochemical features of the patient’s primary bladder tumor, showing the same squamous differentiation and similar Ki-67 proliferative index. The organoid displayed no evidence of the MTAP protein by western blot (similar findings with the MTAP-del cell lines UMUC and RT4). The organoid showed sensitivity to pemetrexed [pemetrexed IC50=0.12 uM (other MTAP-del BLCa IC50: 0.1-0.23 uM)] and resistance to gemcitabine and cisplatin [Gem IC50=55 nM (other BLCa IC50: 1-29 nM), Cis IC50=90.5 uM (other BLCa IC50: 2-17 uM). Conclusions: We describe the first patient-derived MTAP-deleted BCa organoid recapitulating the features of the primary tumor. The organoid displayed sensitivity to pemetrexed as described with MTAP-deleted BCa. The organoid represents a unique platform for testing novel therapeutic strategies for MTAP-deleted BCa. Citation Format: Maximilian Schwermann, Benedito Carneiro, Shaolei Lu, Andre De Souza, Matthew Hadfield, Anthony Mega, Galina Lagos, Sari Khaleel, Dragan Golijanin, Sheldon Holder, Praveen Srinivasan, William MacDonald, Andrew George, Lanlan Zhou, Leiqing Zhang, Wafik El-Deiry. Patient-derived MTAP-deleted bladder cancer organoid model: A unique platform for drug development [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4248.

Abstract 907: ScreenDL: A transfer learning framework integrating tumor omics and functional drug screening for personalized clinical drug response prediction

Abstract Precision oncology hinges on accurate prediction of patient-specific treatment response from tumor molecular inputs. While deep learning (DL) models achieve state-of-the-art response prediction in cell lines, existing methods do not readily translate to the clinic where training data is limited. Here, we have developed and validated ScreenDL, a novel DL framework designed explicitly for use in clinical precision oncology applications. The underlying architecture of ScreenDL consists of two fully connected branches dedicated to learning rich embeddings of a drug’s chemical structure and a tumor’s transcriptomic profile respectively. These tumor omic and drug molecular embeddings are fed into a shared response prediction subnetwork which outputs predicted ln(IC50) values. The ScreenDL training schema incorporates three phases, each designed to improve clinical response prediction: (1) an initial pretraining phase leverages large-scale cell line pharmacogenomic datasets to extract patterns/relationships linking tumor omics and drug response; (2) subsequent transfer learning integrates pharmacogenomic data from patient-derived models of cancer, adapting the pretrained model to a more patient-relevant context; (3) patient-level omics and drug screening data (e.g., from matched patient-derived organoid (PDO) models) is integrated through patient-specific fine-tuning, generating a patient-specific response prediction model. Importantly, biomaterial from surgical tumor resection is often available for organoid establishment and patient-specific drug screening in practical clinical scenarios. To assess the utility of our ScreenDL framework for clinical response prediction, we applied ScreenDL to predict treatment response in 50 advanced/metastatic triple negative breast cancer (TNBC) patient-derived xenograft models (PDX). We leveraged tumor omics profiles for each PDX and drug screening data from matched PDX-derived organoids (PDxOs), mirroring the combination of patient-level tumor omic characterization and functional drug screening in matched PDOs, a protocol currently being piloted in Functional Precision Oncology trials at the University of Utah Huntsman Cancer Institute. After cell line pretraining, ScreenDL achieves a modest median Pearson correlation per drug of 0.15 relative to 0.03 for the leading compared model. However, transfer learning and subsequent PDX-specific fine-tuning significantly improve performance, producing median Pearson correlations per drug of 0.39 and 0.51, respectively. These results demonstrate the ability of our ScreenDL framework to dramatically improve response prediction in clinically relevant domains, bringing DL-based precision treatment selection closer to clinical application. Citation Format: Casey Sederman, Tonya Di Sera, Yi Qiao, Xiaomeng Huang, Bryan E. Welm, Alana L. Welm, Gabor Marth. ScreenDL: A transfer learning framework integrating tumor omics and functional drug screening for personalized clinical drug response prediction [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 907.

Abstract 228: Patient-derived gbm organoids reflect tumour heterogeneity and treatment sensitivity in patients

Abstract Glioblastoma (GBM) is the most common and incurable adult primary brain tumour. Treatment resistance and relapse is the primary cause of death in GBM. Tumour heterogeneity is one of the significant determinants of treatment failure. Patient-derived cancer organoids have been shown to maintain genetic and phenotypic tumour features and recapitulate intratumoral heterogeneity. Here, we report the development and molecular characterisation of patient-derived GBM organoid (PGO) models to study treatment response to existing and new treatment combinations. GBM biopsies were dissociated, suspended in Cultrex, and cultured in a brain stem cell medium. PGOs were characterised by whole-exome (WES) single-cell karyotype (scKaryoSeq) and RNA sequencing. The IC-50 and viability of PGO upon temozolomide (TMZ) and small molecule inhibitors were determined. Gene expression profiling was performed before and after TMZ treatment. Single-cell spatial heterogeneity was determined using Multiplex Immunostaining of PGO and matched patient samples. PGOs were established with a success rate of 80% (n=31). WES analysis between PGOs from the same patient at multiple time points (1-3 months) showed high maintenance of Single Nucleotide Variants (92.3%-97.7%). In PGOs key somatic oncogene variants and copy number variations were retained compared to the primary tumour (58%-90%). scKaryoseq (N=7) of different clinical GBM subtypes showed a high degree of genetic heterogeneity, with some PGO having up to 10 subclones. Single-cell phenotypic landscaping of patient biopsies showed mixed cellular states maintained after long-term culture in the matched PGO. PGOs showed a differential response to TMZ, corresponding with MGMT methylation and patient mutation status. RNAseq of PGO revealed TMZ-responsive genes, including upregulation of the prognostic JUN kinase pathway. Combination treatment of a JUN kinase inhibitor and TMZ showed a synergistic effect. In addition, PGOs were screened for their sensitivity to clinically approved brain-penetrable small molecule inhibitors targeting key GBM driver pathways. Combining small molecular inhibitors of MEK/PI-3/CDK4/EGFR kinase strongly reduced cell viability and activation of cell death pathways. phosphoKinome screens detected compensatory pathway activation upon kinase inhibition (MEK/CDK4 inhibitor) in several PGO. Data will show novel kinase interaction pathways that increase drug sensitivity in PGO. These data show that PGOs can be used as a stable phenotypic and genetic representative of the patient's tumour and have the potential to function as patient avatars for adaptive treatment selection and actionable drug target discovery. Citation Format: Marc A. Vooijs, Linde Hoosemans, Jolanda Piepers, Maikel Verduin, Maxime Vanmechelen, Mike Van Heumen, Frederik De Smet, Ann Hoeben. Patient-derived gbm organoids reflect tumour heterogeneity and treatment sensitivity in patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 228.

Abstract 3280: ASCL2+ tumor cells modulate the response of colorectal cancer to MAPK targeting therapy

Abstract Colorectal cancer (CRC) is commonly treated with MAPK pathway inhibitors (MAPKi) including EGFRi, BRAFi, and most recently KRAS G12C inhibitors. However, clinical trials repeatedly demonstrate that durability of benefit is short-lived in many patients. While genomic mechanisms of acquired resistance have been described, this explains a small minority of patients and further research is needed to determine the underlying biology limiting MAPKi therapy durability in CRC. Potential responsible mechanisms were explored using transcriptomic data from patients treated with MAPKi combination therapy, showing upregulation of stem-related signatures in nonresponders (FDR<0.001), therefore CRC stem markers with the greatest expression increase were identified. The top markers (ASCL2, LGR5, LRIG1) were evaluated using a metacohort of single cell RNAseq (scRNAseq) data from patients with CRC (n=72), and ASCL2 was selected for further study due to its high marker coexpression and tumor cell specificity (OR=8.27, p<0.0001). Additionally, ASCL2 expression increased in multiple patient-derived organoid and xenograft (PDX) models that were treated with a spectrum of MAPKi therapies (BRAFi, EGFRi, G12Ci, MEKi, SOSi). As such, we postulated that ASCL2+ tumor cells play an important role in the limited response durability of MAPKi therapy in CRC. To investigate, ASCL2 reporter models were treated with MAPKi therapy, resulting in the significant enrichment of ASCL2+ tumor cells, independent of driver mutation or treatment regimen used. Using Single Cell BarcodeSeq, by combining molecular barcoded PDXs with single cell transcriptomics, we found that MAPKi therapy engaged intrinsic cellular plasticity resulting in cells transitioning into the ASCL2+ phenotype. Additional patient-level data confirmed enrichment of the ASCL2+ transcriptomic signature at progression while on MAPKi therapy. Characterization of ASCL2+ cells showed higher stem cell frequency (p=9.7 × 10−62), clonogenic capacity (p<0.0001), expression of stem-related genes, and retention of phenotypic plasticity, indicating ASCL2+ tumor cells feature stem-like abilities, as previously described, which include therapy resistance. Lastly, depletion of ASCL2+ tumor cells resulted in improved initial response to MAPKi therapy (p<0.0001) and, more importantly, prolonged cell growth inhibition. Overall, we observe that MAPKi therapy enriches stemness programs in CRC and demonstrate MAPKi drives cells into a transient ASCL2+ stem-like state with limited sensitivity to MAPK inhibitors. Further studies are underway to identify vulnerabilities that will enable targeting of ASCL2+ tumor cells. Ultimately these findings will help guide future optimization of MAPKi regimens to prolong response durability and improve outcomes for patients with CRC. Citation Format: Oscar E. Villarreal, Hey Min Lee, Ha Tran, Annette Machado, Jumanah Alshenaifi, Chi Wut Wong, Olu Coker, Rosalba Minelli, Michael Peoples, Alejandro Hernandez Martinez, Natalie Fowlkes, Preeti Kanikarla, Kangyu Lin, Christopher Bristow, Andrea Viale, Joseph R. Marszalek, John Paul Shen, Scott Kopetz. ASCL2+ tumor cells modulate the response of colorectal cancer to MAPK targeting therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3280.