Pharmacogenomic landscape of patient-derived tumor cells informs precision oncology therapy

Disrupting glioblastoma networks with tumor treating fields (TTFields) in in vitro models

PURPOSE

This study investigates the biological effect of Tumor Treating Fields (TTFields) on key drivers of glioblastoma's malignancy-tumor microtube (TM) formation-and on the function and overall integrity of the tumor cell network.

METHOD

Using a two-dimensional monoculture GB cell network model (2DTM) of primary glioblastoma cell (GBC) cultures (S24, BG5 or T269), we evaluated the effects of TTFields on cell density, interconnectivity and structural integrity of the tumor network. We also analyzed calcium (Ca2+) transient dynamics and network morphology, validating findings in patient-derived tumoroids and brain tumor organoids.

RESULTS

In the 2DTM assay, TTFields reduced cell density by 85-88% and disrupted network interconnectivity, particularly in cells with multiple TMs. A "crooked TM" phenotype emerged in 5-6% of treated cells, rarely seen in controls. Ca2+ transients were significantly compromised, with global Ca2+ activity reduced by 51-83%, active and periodic cells by over 50%, and intercellular co-activity by 52% in S24, and almost completely in BG5 GBCs. The effects were more pronounced at 200 kHz compared to a 50 kHz TTFields. Similar reductions in Ca2+ activity were observed in patient-derived tumoroids. In brain tumor organoids, TTFields significantly reduced tumor cell proliferation and infiltration.

CONCLUSION

Our comprehensive study provides new insights into the multiple effects of Inovitro-modeled TTFields on glioma progression, morphology and network dynamics in vitro. Future in vivo studies to verify our in vitro findings may provide the basis for a deeper understanding and optimization of TTFields as a therapeutic modality in the treatment of GB.

Establishment of a patient-specific avatar organoid model derived from EUS-guided fine-needle biopsy for timely clinical application in pancreatic ductal adenocarcinoma (with video)

BACKGROUND AND AIMS

Pancreatic ductal adenocarcinoma (PDAC) has the worst survival rate among tumors. At the time of diagnosis, more than 80% of PDACs are considered to be surgically unresectable, and there is an unmet need for treatment options in these inoperable PDACs. This study aimed to establish a patient-derived organoid (PDO) platform from EUS-guided fine-needle biopsy (EUS-FNB) collected at diagnosis and to determine its clinical applicability for the timely treatment of unresectable PDAC.

METHODS

Patients with suspected PDAC were prospectively enrolled at the Samsung Medical Center from 2015 to 2019. PDAC tissues were acquired by means of EUS-FNB to establish PDAC PDOs, which were comprehensively analyzed for histology, genomic sequencing, and high-throughput screening (HTS) drug sensitivity test.

RESULTS

PDAC PDOs were established with a success rate of 83.2% (94/113). It took approximately 3 weeks from acquiring minimal EUS-FNB specimens to generating sufficient PDAC PDOs for the simultaneous HTS drug sensitivity test and genomic sequencing. The high concordance between PDAC tissues and matched PDOs was confirmed, and whole-exome sequencing revealed the increased detection of genetic alterations in PDOs compared with EUS-FNB tissues. The HTS drug sensitivity test showed clinical correlation between the ex vivo PDO response and the actual chemotherapeutic response of the study patients in the real world (13 out of 15 cases). In addition, whole-transcriptome sequencing identified candidate genes associated with nab-paclitaxel resistance, such as ITGB7, ANPEP, and ST3GAL1.

CONCLUSIONS

This PDAC PDO platform allows several therapeutic drugs to be tested within a short time window and opens the possibility for timely personalized medicine as a "patient avatar model" in clinical practice.

Spheroids Generated from Malignant Pleural Effusion as a Tool to Predict the Response of Non-Small Cell Lung Cancer to Treatment

BACKGROUND

Spheroids generated by tumor cells collected from malignant pleural effusion (MPE) were shown to retain the characteristics of the original tumors. This ex vivo model might be used to predict the response of non-small cell lung cancer (NSCLC) to anticancer treatments.

METHODS

The characteristics, epidermal growth factor receptor (EGFR) mutation status, and clinical response to EGFR-TKIs treatment of enrolled patients were recorded. The viability of the spheroids generated from MPE of enrolled patients were evaluated by visualization of the formazan product of the MTT assay.

RESULTS

Spheroids were generated from 14 patients with NSCLC-related MPE. Patients with EGFR L861Q, L858R, or Exon 19 deletion all received EGFR-TKIs, and five of these seven patients responded to treatment. The viability of the spheroids generated from MPE of these five patients who responded to EGFR-TKIs treatment was significantly reduced after gefitinib treatment. On the other hand, gefitinib treatment did not reduce the viability of the spheroids generated from MPE of patients with EGFR wild type, Exon 20 insertion, or patients with sensitive EGFR mutation but did not respond to EGFR-TKIs treatment.

CONCLUSION

Multicellular spheroids generated from NSCLC-related MPE might be used to predict the response of NSCLC to treatment.

CFSSynergy: Combining Feature-Based and Similarity-Based Methods for Drug Synergy Prediction

Drug synergy prediction plays a vital role in cancer treatment. Because experimental approaches are labor-intensive and expensive, computational-based approaches get more attention. There are two types of computational methods for drug synergy prediction: feature-based and similarity-based. In feature-based methods, the main focus is to extract more discriminative features from drug pairs and cell lines to pass to the task predictor. In similarity-based methods, the similarities among all drugs and cell lines are utilized as features and fed into the task predictor. In this work, a novel approach, called CFSSynergy, that combines these two viewpoints is proposed. First, a discriminative representation is extracted for paired drugs and cell lines as input. We have utilized transformer-based architecture for drugs. For cell lines, we have created a similarity matrix between proteins using the Node2Vec algorithm. Then, the new cell line representation is computed by multiplying the protein-protein similarity matrix and the initial cell line representation. Next, we compute the similarity between unique drugs and unique cells using the learned representation for paired drugs and cell lines. Then, we compute a new representation for paired drugs and cell lines based on the similarity-based features and the learned features. Finally, these features are fed to XGBoost as a task predictor. Two well-known data sets were used to evaluate the performance of our proposed method: DrugCombDB and OncologyScreen. The CFSSynergy approach consistently outperformed existing methods in comparative evaluations. This substantiates the efficacy of our approach in capturing complex synergistic interactions between drugs and cell lines, setting it apart from conventional similarity-based or feature-based methods.

Patient-derived tumor spheroid-induced angiogenesis preclinical platform for exploring therapeutic vulnerabilities in cancer

This study addresses the demand for research models that can support patient-treatment decisions and clarify the complexities of a tumor microenvironment by developing an advanced non-animal preclinical cancer model. Based on patient-derived tumor spheroids (PDTS), the proposed model reconstructs the tumor microenvironment with emphasis on tumor spheroid-driven angiogenesis. The resulting microfluidic chip system mirrors angiogenic responses elicited by PDTS, recapitulating patient-specific tumor conditions and providing robust, easily quantifiable outcomes. Vascularized PDTS exhibited marked angiogenesis and tumor proliferation on the microfluidic chip. Furthermore, a drug that targets the vascular endothelial growth factor receptor 2 (VEGFR2, ramucirumab) was deployed, which effectively inhibited angiogenesis and impeded tumor invasion. This innovative preclinical model was used for investigating distinct responses for various drug combinations, encompassing HER2 inhibitors and angiogenesis inhibitors, within the context of PDTS. This integrated platform could potentially advance precision medicine by harmonizing diverse data points within the tumor microenvironment with a focus on the interplay between cancer and the vascular system.

A personalized medicine approach identifies enasidenib as an efficient treatment for IDH2 mutant chondrosarcoma

BACKGROUND

Sarcomas represent an extensive group of malignant diseases affecting mesodermal tissues. Among sarcomas, the clinical management of chondrosarcomas remains a complex challenge, as high-grade tumours do not respond to current therapies. Mutations in the isocitrate dehydrogenase (IDH) 1 and 2 genes are among the most common mutations detected in chondrosarcomas and may represent a therapeutic opportunity. The presence of mutated IDH (mIDH) enzymes results in the accumulation of the oncometabolite 2-HG leading to molecular alterations that contribute to drive tumour growth.

METHODS

We developed a personalized medicine strategy based on the targeted NGS/Sanger sequencing of sarcoma samples (n = 6) and the use of matched patient-derived cell lines as a drug-testing platform. The anti-tumour potential of IDH mutations found in two chondrosarcoma cases was analysed in vitro, in vivo and molecularly (transcriptomic and DNA methylation analyses).

FINDINGS

We treated several chondrosarcoma models with specific mIDH1/2 inhibitors. Among these treatments, only the mIDH2 inhibitor enasidenib was able to decrease 2-HG levels and efficiently reduce the viability of mIDH2 chondrosarcoma cells. Importantly, oral administration of enasidenib in xenografted mice resulted in a complete abrogation of tumour growth. Enasidenib induced a profound remodelling of the transcriptomic landscape not associated to changes in the 5 mC methylation levels and its anti-tumour effects were associated with the repression of proliferative pathways such as those controlled by E2F factors.

INTERPRETATION

Overall, this work provides preclinical evidence for the use of enasidenib to treat mIDH2 chondrosarcomas.

FUNDING

Supported by the Spanish Research Agency/FEDER (grants PID2022-142020OB-I00; PID2019-106666RB-I00), the ISC III/FEDER (PI20CIII/00020; DTS18CIII/00005; CB16/12/00390; CB06/07/1009; CB19/07/00057); the GEIS group (GEIS-62); and the PCTI (Asturias)/FEDER (IDI/2021/000027).

TheraVision: Engineering platform technology for the development of oncolytic viruses based on herpes simplex virus type 1

Viruses are able to efficiently penetrate cells, multiply, and eventually kill infected cells, release tumor antigens, and activate the immune system. Therefore, viruses are highly attractive novel agents for cancer therapy. Clinical trials with first generations of oncolytic viruses (OVs) are very promising but show significant need for optimization. The aim of TheraVision was to establish a broadly applicable engineering platform technology for combinatorial oncolytic virus and immunotherapy. Through genetic engineering, an attenuated herpes simplex virus type 1 (HSV1) was generated that showed increased safety compared to the wild-type strain. To demonstrate the modularity and the facilitated generation of new OVs, two transgenes encoding retargeting as well as immunomodulating single-chain variable fragments (scFvs) were integrated into the platform vector. The resulting virus selectively infected epidermal growth factor receptor (EGFR)-expressing cells and produced a functional immune checkpoint inhibitor against programmed cell death protein 1 (PD-1). Thus, both viral-mediated oncolysis and immune-cell-mediated therapy were combined into a single viral vector. Safety and functionality of the armed OVs have been shown in novel preclinical models ranging from patient-derived organoids and tissue-engineered human in vitro 3D tumor models to complex humanized mouse models. Consequently, a novel and proprietary engineering platform vector based on HSV1 is available for the facilitated preclinical development of oncolytic virotherapy.

Patient-Derived Microphysiological Systems for Precision Medicine

Patient-derived microphysiological systems (P-MPS) have emerged as powerful tools in precision medicine that provide valuable insight into individual patient characteristics. This review discusses the development of P-MPS as an integration of patient-derived samples, including patient-derived cells, organoids, and induced pluripotent stem cells, into well-defined MPSs. Emphasizing the necessity of P-MPS development, its significance as a nonclinical assessment approach that bridges the gap between traditional in vitro models and clinical outcomes is highlighted. Additionally, guidance is provided for engineering approaches to develop microfluidic devices and high-content analysis for P-MPSs, enabling high biological relevance and high-throughput experimentation. The practical implications of the P-MPS are further examined by exploring the clinically relevant outcomes obtained from various types of patient-derived samples. The construction and analysis of these diverse samples within the P-MPS have resulted in physiologically relevant data, paving the way for the development of personalized treatment strategies. This study describes the significance of the P-MPS in precision medicine, as well as its unique capacity to offer valuable insights into individual patient characteristics.

Drug Response of Patient-Derived Lung Cancer Cells Predicts Clinical Outcomes of Targeted Therapy

Intratumor heterogeneity leads to different responses to targeted therapies, even within patients whose tumors harbor identical driver oncogenes. This study examined clinical outcomes according to a patient-derived cell (PDC)-based drug sensitivity test in lung cancer patients treated with targeted therapies. From 487 lung cancers, 397 PDCs were established with a success rate of 82%. In 139 PDCs from advanced non-small-cell lung cancer (NSCLC) patients receiving targeted therapies, the standardized area under the curve (AUC) values for the drugs was significantly correlated with their tumor response (p = 0.002). Among 59 chemo-naive EGFR/ALK-positive NSCLC patients, the PDC non-responders showed a significantly inferior response rate (RR) and progression-free survival (PFS) for the targeted drugs than the PDC responders (RR, 25% vs. 78%, p = 0.011; median PFS, 3.4 months [95% confidence interval (CI), 2.8-4.1] vs. 11.8 months [95% CI, 6.5-17.0], p < 0.001). Of 25 EGFR-positive NSCLC patients re-challenged with EGFR inhibitors, the PDC responder showed a higher RR than the PDC non-responder (42% vs. 15%). Four patients with wild-type EGFR or uncommon EGFR-mutant NSCLC were treated with EGFR inhibitors based on their favorable PDC response to EGFR inhibitors, and two patients showed dramatic responses. Therefore, the PDC-based drug sensitivity test results were significantly associated with clinical outcomes in patients with EGFR- or ALK-positive NSCLC. It may be helpful for predicting individual heterogenous clinical outcomes beyond genomic alterations.

An unbiased approach of molecular characterization of the endometrium: toward defining endometrial-based infertility

Infertility is a complex condition affecting millions of couples worldwide. The current definition of infertility, based on clinical criteria, fails to account for the molecular and cellular changes that may occur during the development of infertility. Recent advancements in sequencing technology and single-cell analysis offer new opportunities to gain a deeper understanding of these changes. The endometrium has a potential role in infertility and has been extensively studied to identify gene expression profiles associated with (impaired) endometrial receptivity. However, limited overlap among studies hampers the identification of relevant downstream pathways that could play a role in the development of endometrial-related infertility. To address these challenges, we propose sequencing the endometrial transcriptome of healthy and infertile women at the single-cell level to consistently identify molecular signatures. Establishing consensus on physiological patterns in endometrial samples can aid in identifying deviations in infertile patients. A similar strategy has been used with great success in cancer research. However, large collaborative initiatives, international uniform protocols of sample collection and processing are crucial to ensure reliability and reproducibility. Overall, the proposed approach holds promise for an objective and accurate classification of endometrial-based infertility and has the potential to improve diagnosis and treatment outcomes.

PANCDR: precise medicine prediction using an adversarial network for cancer drug response

Pharmacogenomics aims to provide personalized therapy to patients based on their genetic variability. However, accurate prediction of cancer drug response (CDR) is challenging due to genetic heterogeneity. Since clinical data are limited, most studies predicting drug response use preclinical data to train models. However, such models might not be generalizable to external clinical data due to differences between the preclinical and clinical datasets. In this study, a Precision Medicine Prediction using an Adversarial Network for Cancer Drug Response (PANCDR) model is proposed. PANCDR consists of two sub-models, an adversarial model and a CDR prediction model. The adversarial model reduces the gap between the preclinical and clinical datasets, while the CDR prediction model extracts features and predicts responses. PANCDR was trained using both preclinical data and unlabeled clinical data. Subsequently, it was tested on external clinical data, including The Cancer Genome Atlas and brain tumor patients. PANCDR outperformed other machine learning models in predicting external test data. Our results demonstrate the robustness of PANCDR and its potential in precision medicine by recommending patient-specific drug candidates. The PANCDR codes and data are available at https://github.com/DMCB-GIST/PANCDR.

Construction of tumor organoids and their application to cancer research and therapy

Cancer remains a severe public health burden worldwide. One of the challenges hampering effective cancer therapy is that the existing cancer models hardly recapitulate the tumor microenvironment of human patients. Over the past decade, tumor organoids have emerged as an in vitro 3D tumor model to mimic the pathophysiological characteristics of parental tumors. Various techniques have been developed to construct tumor organoids, such as matrix-based methods, hanging drop, spinner or rotating flask, nonadhesive surface, organ-on-a-chip, 3D bioprinting, and genetic engineering. This review elaborated on cell components and fabrication methods for establishing tumor organoid models. Furthermore, we discussed the application of tumor organoids to cancer modeling, basic cancer research, and anticancer therapy. Finally, we discussed current limitations and future directions in employing tumor organoids for more extensive applications.

Single Cell Isolation from Surgically Resected Tissue Via Mechanical Dissociation Using TissueGrinder

Primary cells form the basis of modern-day in vitro research analysis tools. Many conventional procedures for generating single-cell suspensions from solid tissue are neither robust nor reproducible. Here we describe primary cells isolation from surgically resected tumor tissue via enzyme-free mechanical dissociation using TissueGrinder, a novel semi-automated benchtop device. The isolated cells can be used for any downstream biochemical or cell-based analytic assay.

Progress in patient-derived liver cancer cell models: a step forward for precision medicine

The development of effective precision treatments for liver cancers has been hindered by the scarcity of preclinical models that accurately reflect the heterogeneity of this disease. Recent progress in developing patient-derived liver cancer cell lines and organoids has paved the way for precision medicine research. These expandable resources of liver cancer cell models enable a full spectrum of pharmacogenomic analysis for liver cancers. Moreover, patient-derived and short-term cultured two-dimensional tumor cells or three-dimensional organoids can serve as patient avatars, allowing for the prediction of patients' response to drugs and facilitating personalized treatment for liver cancer patients. Furthermore, the current novel techniques have expanded the scope of cancer research, including innovative organoid culture, gene editing and bioengineering. In this review, we provide an overview of the progress in patient-derived liver cancer cell models, focusing on their applications in precision and personalized medicine research. We also discuss the challenges and future perspectives in this field.

Identifying predictors of glioma evolution from longitudinal sequencing

Clonal evolution drives cancer progression and therapeutic resistance. Recent studies have revealed divergent longitudinal trajectories in gliomas, but early molecular features steering posttreatment cancer evolution remain unclear. Here, we collected sequencing and clinical data of initial-recurrent tumor pairs from 544 adult diffuse gliomas and performed multivariate analysis to identify early molecular predictors of tumor evolution in three diffuse glioma subtypes. We found that CDKN2A deletion at initial diagnosis preceded tumor necrosis and microvascular proliferation that occur at later stages of IDH-mutant glioma. Ki67 expression at diagnosis was positively correlated with acquiring hypermutation at recurrence in the IDH-wild-type glioma. In all glioma subtypes, MYC gain or MYC-target activation at diagnosis was associated with treatment-induced hypermutation at recurrence. To predict glioma evolution, we constructed CELLO2 (Cancer EvoLution for LOngitudinal data version 2), a machine learning model integrating features at diagnosis to forecast hypermutation and progression after treatment. CELLO2 successfully stratified patients into subgroups with distinct prognoses and identified a high-risk patient group featured by MYC gain with worse post-progression survival, from the low-grade IDH-mutant-noncodel subtype. We then performed chronic temozolomide-induction experiments in glioma cell lines and isogenic patient-derived gliomaspheres and demonstrated that MYC drives temozolomide resistance by promoting hypermutation. Mechanistically, we demonstrated that, by binding to open chromatin and transcriptionally active genomic regions, c-MYC increases the vulnerability of key mismatch repair genes to treatment-induced mutagenesis, thus triggering hypermutation. This study reveals early predictors of cancer evolution under therapy and provides a resource for precision oncology targeting cancer dynamics in diffuse gliomas.

Pan-Cancer Pharmacogenomic Analysis of Patient-Derived Tumor Cells Using Clinically Relevant Drug Exposures

As a result of tumor heterogeneity and solid cancers harboring multiple molecular defects, precision medicine platforms in oncology are most effective when both genetic and pharmacologic determinants of a tumor are evaluated. Expandable patient-derived xenograft (PDX) mouse tumor and corresponding PDX culture (PDXC) models recapitulate many of the biological and genetic characteristics of the original patient tumor, allowing for a comprehensive pharmacogenomic analysis. Here, the somatic mutations of 23 matched patient tumor and PDX samples encompassing four cancers were first evaluated using next-generation sequencing (NGS). 19 antitumor agents were evaluated across 78 patient-derived tumor cultures using clinically relevant drug exposures. A binarization threshold sensitivity classification determined in culture (PDXC) was used to identify tumors that best respond to drug in vivo (PDX). Using this sensitivity classification, logic models of DNA mutations were developed for 19 antitumor agents to predict drug response. We determined that the concordance of somatic mutations across patient and corresponding PDX samples increased as variant allele frequency increased. Notable individual PDXC responses to specific drugs, as well as lineage-specific drug responses were identified. Robust responses identified in PDXC were recapitulated in vivo in PDX-bearing mice and logic modeling determined somatic gene mutation(s) defining response to specific antitumor agents. In conclusion, combining NGS of primary patient tumors, high-throughput drug screen using clinically relevant doses, and logic modeling, can provide a platform for understanding response to therapeutic drugs targeting cancer.

High-Throughput Drug Screening of Primary Tumor Cells Identifies Therapeutic Strategies for Treating Children with High-Risk Cancer

For one-third of patients with pediatric cancer enrolled in precision medicine programs, molecular profiling does not result in a therapeutic recommendation. To identify potential strategies for treating these high-risk pediatric patients, we performed in vitro screening of 125 patient-derived samples against a library of 126 anticancer drugs. Tumor cell expansion did not influence drug responses, and 82% of the screens on expanded tumor cells were completed while the patients were still under clinical care. High-throughput drug screening (HTS) confirmed known associations between activating genomic alterations in NTRK, BRAF, and ALK and responses to matching targeted drugs. The in vitro results were further validated in patient-derived xenograft models in vivo and were consistent with clinical responses in treated patients. In addition, effective combinations could be predicted by correlating sensitivity profiles between drugs. Furthermore, molecular integration with HTS identified biomarkers of sensitivity to WEE1 and MEK inhibition. Incorporating HTS into precision medicine programs is a powerful tool to accelerate the improved identification of effective biomarker-driven therapeutic strategies for treating high-risk pediatric cancers.

SIGNIFICANCE

Integrating HTS with molecular profiling is a powerful tool for expanding precision medicine to support drug treatment recommendations and broaden the therapeutic options available to high-risk pediatric cancers.

DeepTraSynergy: drug combinations using multimodal deep learning with transformers

MOTIVATION

Screening bioactive compounds in cancer cell lines receive more attention. Multidisciplinary drugs or drug combinations have a more effective role in treatments and selectively inhibit the growth of cancer cells.

RESULTS

Hence, we propose a new deep learning-based approach for drug combination synergy prediction called DeepTraSynergy. Our proposed approach utilizes multimodal input including drug-target interaction, protein-protein interaction, and cell-target interaction to predict drug combination synergy. To learn the feature representation of drugs, we have utilized transformers. It is worth noting that our approach is a multitask approach that predicts three outputs including the drug-target interaction, its toxic effect, and drug combination synergy. In our approach, drug combination synergy is the main task and the two other ones are the auxiliary tasks that help the approach to learn a better model. In the proposed approach three loss functions are defined: synergy loss, toxic loss, and drug-protein interaction loss. The last two loss functions are designed as auxiliary losses to help learn a better solution. DeepTraSynergy outperforms the classic and state-of-the-art models in predicting synergistic drug combinations on the two latest drug combination datasets. The DeepTraSynergy algorithm achieves accuracy values of 0.7715 and 0.8052 (an improvement over other approaches) on the DrugCombDB and Oncology-Screen datasets, respectively. Also, we evaluate the contribution of each component of DeepTraSynergy to show its effectiveness in the proposed method. The introduction of the relation between proteins (PPI networks) and drug-protein interaction significantly improves the prediction of synergistic drug combinations.

AVAILABILITY AND IMPLEMENTATION

The source code and data are available at https://github.com/fatemeh-rafiei/DeepTraSynergy.

Artificial intelligence-assisted selection and efficacy prediction of antineoplastic strategies for precision cancer therapy

The rapid development of artificial intelligence (AI) technologies in the context of the vast amount of collectable data obtained from high-throughput sequencing has led to an unprecedented understanding of cancer and accelerated the advent of a new era of clinical oncology with a tone of precision treatment and personalized medicine. However, the gains achieved by a variety of AI models in clinical oncology practice are far from what one would expect, and in particular, there are still many uncertainties in the selection of clinical treatment options that pose significant challenges to the application of AI in clinical oncology. In this review, we summarize emerging approaches, relevant datasets and open-source software of AI and show how to integrate them to address problems from clinical oncology and cancer research. We focus on the principles and procedures for identifying different antitumor strategies with the assistance of AI, including targeted cancer therapy, conventional cancer therapy, and cancer immunotherapy. In addition, we also highlight the current challenges and directions of AI in clinical oncology translation. Overall, we hope this article will provide researchers and clinicians with a deeper understanding of the role and implications of AI in precision cancer therapy, and help AI move more quickly into accepted cancer guidelines.

PRMT3-mediated arginine methylation of IGF2BP1 promotes oxaliplatin resistance in liver cancer

Although oxaliplatin-based chemotherapy has been effective in the treatment of hepatocellular carcinoma (HCC), primary or acquired resistance to oxaliplatin remains a major challenge in the clinic. Through functional screening using CRISPR/Cas9 activation library, transcriptomic profiling of clinical samples, and functional validation in vitro and in vivo, we identify PRMT3 as a key driver of oxaliplatin resistance. Mechanistically, PRMT3-mediated oxaliplatin-resistance is in part dependent on the methylation of IGF2BP1 at R452, which is critical for the function of IGF2BP1 in stabilizing the mRNA of HEG1, an effector of PRMT3-IGF2BP1 axis. Also, PRMT3 overexpression may serve as a biomarker for oxaliplatin resistance in HCC patients. Collectively, our study defines the PRTM3-IGF2BP1-HEG1 axis as important regulators and therapeutic targets in oxaliplatin-resistance and suggests the potential to use PRMT3 expression level in pretreatment biopsy as a biomarker for oxaliplatin-resistance in HCC patients.

DROEG: a method for cancer drug response prediction based on omics and essential genes integration

Predicting therapeutic responses in cancer patients is a major challenge in the field of precision medicine due to high inter- and intra-tumor heterogeneity. Most drug response models need to be improved in terms of accuracy, and there is limited research to assess therapeutic responses of particular tumor types. Here, we developed a novel method DROEG (Drug Response based on Omics and Essential Genes) for prediction of drug response in tumor cell lines by integrating genomic, transcriptomic and methylomic data along with CRISPR essential genes, and revealed that the incorporation of tumor proliferation essential genes can improve drug sensitivity prediction. Concisely, DROEG integrates literature-based and statistics-based methods to select features and uses Support Vector Regression for model construction. We demonstrate that DROEG outperforms most state-of-the-art algorithms by both qualitative (prediction accuracy for drug-sensitive/resistant) and quantitative (Pearson correlation coefficient between the predicted and actual IC50) evaluation in Genomics of Drug Sensitivity in Cancer and Cancer Cell Line Encyclopedia datasets. In addition, DROEG is further applied to the pan-gastrointestinal tumor with high prevalence and mortality as a case study at both cell line and clinical levels to evaluate the model efficacy and discover potential prognostic biomarkers in Cisplatin and Epirubicin treatment. Interestingly, the CRISPR essential gene information is found to be the most important contributor to enhance the accuracy of the DROEG model. To our knowledge, this is the first study to integrate essential genes with multi-omics data to improve cancer drug response prediction and provide insights into personalized precision treatment.

Functional precision oncology using patient-derived assays: bridging genotype and phenotype

Genomics-based precision medicine has revolutionized oncology but also has inherent limitations. Functional precision oncology is emerging as a complementary approach that aims to bridge the gap between genotype and phenotype by modelling individual tumours in vitro. These patient-derived ex vivo models largely preserve several tumour characteristics that are not captured by genomics approaches and enable the functional dissection of tumour vulnerabilities in a personalized manner. In this Review, we discuss several examples of personalized functional assays involving tumour organoids, spheroids and explants and their potential to predict treatment responses and drug-induced toxicities in individual patients. These developments have opened exciting new avenues for precision oncology, with the potential for successful clinical applications in contexts in which genomic data alone are not informative. To implement these assays into clinical practice, we outline four key barriers that need to be overcome: assay success rates, turnaround times, the need for standardized conditions and the definition of in vitro responders. Furthermore, we discuss novel technological advances such as microfluidics that might reduce sample requirements, assay times and labour intensity and thereby enable functional precision oncology to be implemented in routine clinical practice.

Pharmacogenomic profiling reveals molecular features of chemotherapy resistance in IDH wild-type primary glioblastoma

BACKGROUND

Although temozolomide (TMZ) has been used as a standard adjuvant chemotherapeutic agent for primary glioblastoma (GBM), treating isocitrate dehydrogenase wild-type (IDH-wt) cases remains challenging due to intrinsic and acquired drug resistance. Therefore, elucidation of the molecular mechanisms of TMZ resistance is critical for its precision application.

METHODS

We stratified 69 primary IDH-wt GBM patients into TMZ-resistant (n = 29) and sensitive (n = 40) groups, using TMZ screening of the corresponding patient-derived glioma stem-like cells (GSCs). Genomic and transcriptomic features were then examined to identify TMZ-associated molecular alterations. Subsequently, we developed a machine learning (ML) model to predict TMZ response from combined signatures. Moreover, TMZ response in multisector samples (52 tumor sectors from 18 cases) was evaluated to validate findings and investigate the impact of intra-tumoral heterogeneity on TMZ efficacy.

RESULTS

In vitro TMZ sensitivity of patient-derived GSCs classified patients into groups with different survival outcomes (P = 1.12e-4 for progression-free survival (PFS) and 3.63e-4 for overall survival (OS)). Moreover, we found that elevated gene expression of EGR4, PAPPA, LRRC3, and ANXA3 was associated to intrinsic TMZ resistance. In addition, other features such as 5-aminolevulinic acid negative, mesenchymal/proneural expression subtypes, and hypermutation phenomena were prone to promote TMZ resistance. In contrast, concurrent copy-number-alteration in PTEN, EGFR, and CDKN2A/B was more frequent in TMZ-sensitive samples (Fisher's exact P = 0.0102), subsequently consolidated by multi-sector sequencing analyses. Integrating all features, we trained a ML tool to segregate TMZ-resistant and sensitive groups. Notably, our method segregated IDH-wt GBM patients from The Cancer Genome Atlas (TCGA) into two groups with divergent survival outcomes (P = 4.58e-4 for PFS and 3.66e-4 for OS). Furthermore, we showed a highly heterogeneous TMZ-response pattern within each GBM patient using in vitro TMZ screening and genomic characterization of multisector GSCs. Lastly, the prediction model that evaluates the TMZ efficacy for primary IDH-wt GBMs was developed into a webserver for public usage ( http://www.wang-lab-hkust.com:3838/TMZEP ).

CONCLUSIONS

We identified molecular characteristics associated to TMZ sensitivity, and illustrate the potential clinical value of a ML model trained from pharmacogenomic profiling of patient-derived GSC against IDH-wt GBMs.

High-throughput organo-on-pillar (high-TOP) array system for three-dimensional ex vivo drug testing

The development of organoid culture technologies has triggered industrial interest in ex vivo drug test-guided clinical response prediction for precision cancer therapy. The three-dimensional culture encapsulated with basement membrane (BM) components is extremely important in establishing ex vivo organoids and drug sensitivity tests because the BM components confer essential structures resembling tumor histopathology. Although numerous studies have demonstrated three-dimensional culture-based drug screening methods, establishing a large-scale drug-screening platform with matrix-encapsulated tumor cells is challenging because the arrangement of microspots of a matrix-cell droplet onto each well of a microwell plate is inconsistent and difficult to standardize. In addition, relatively low scales and lack of reproducibility discourage the application of three-dimensional organoid-based drug screening data for precision treatment or drug discovery. To overcome these limitations, we manufactured an automated organospotter-integrated high-throughput organo-on-pillar (high-TOP) drug-screening platform. Our system is compatible with various extracellular matrices, including BM extract, Matrigel, collagen, and hydrogel. In addition, it can be readily utilized for high-content analyses by simply exchanging the bottom plates without disrupting the domes. Our system demonstrated considerable robustness, consistency, reproducibility, and biological relevancy in three-dimensional drug sensitivity analyses using Matrigel-encapsulated ovarian cancer cell lines. We also demonstrated proof-of-concept cases representing the clinical feasibility of high-TOP-assisted ex vivo drug tests linked to clinical chemo-response in ovarian cancer patients. In conclusion, our platform provides an automated and standardized method for ex vivo drug-sensitivity-guided clinical response prediction, suggesting effective chemotherapy regimens for patients with cancer.

A fair experimental comparison of neural network architectures for latent representations of multi-omics for drug response prediction

BACKGROUND

Recent years have seen a surge of novel neural network architectures for the integration of multi-omics data for prediction. Most of the architectures include either encoders alone or encoders and decoders, i.e., autoencoders of various sorts, to transform multi-omics data into latent representations. One important parameter is the depth of integration: the point at which the latent representations are computed or merged, which can be either early, intermediate, or late. The literature on integration methods is growing steadily, however, close to nothing is known about the relative performance of these methods under fair experimental conditions and under consideration of different use cases.

RESULTS

We developed a comparison framework that trains and optimizes multi-omics integration methods under equal conditions. We incorporated early integration, PCA and four recently published deep learning methods: MOLI, Super.FELT, OmiEmbed, and MOMA. Further, we devised a novel method, Omics Stacking, that combines the advantages of intermediate and late integration. Experiments were conducted on a public drug response data set with multiple omics data (somatic point mutations, somatic copy number profiles and gene expression profiles) that was obtained from cell lines, patient-derived xenografts, and patient samples. Our experiments confirmed that early integration has the lowest predictive performance. Overall, architectures that integrate triplet loss achieved the best results. Statistical differences can, overall, rarely be observed, however, in terms of the average ranks of methods, Super.FELT is consistently performing best in a cross-validation setting and Omics Stacking best in an external test set setting.

CONCLUSIONS

We recommend researchers to follow fair comparison protocols, as suggested in the paper. When faced with a new data set, Super.FELT is a good option in the cross-validation setting as well as Omics Stacking in the external test set setting. Statistical significances are hardly observable, despite trends in the algorithms' rankings. Future work on refined methods for transfer learning tailored for this domain may improve the situation for external test sets. The source code of all experiments is available under https://github.com/kramerlab/Multi-Omics_analysis.

DeepInsight-3D architecture for anti-cancer drug response prediction with deep-learning on multi-omics

Modern oncology offers a wide range of treatments and therefore choosing the best option for particular patient is very important for optimal outcome. Multi-omics profiling in combination with AI-based predictive models have great potential for streamlining these treatment decisions. However, these encouraging developments continue to be hampered by very high dimensionality of the datasets in combination with insufficiently large numbers of annotated samples. Here we proposed a novel deep learning-based method to predict patient-specific anticancer drug response from three types of multi-omics data. The proposed DeepInsight-3D approach relies on structured data-to-image conversion that then allows use of convolutional neural networks, which are particularly robust to high dimensionality of the inputs while retaining capabilities to model highly complex relationships between variables. Of particular note, we demonstrate that in this formalism additional channels of an image can be effectively used to accommodate data from different omics layers while implicitly encoding the connection between them. DeepInsight-3D was able to outperform other state-of-the-art methods applied to this task. The proposed improvements can facilitate the development of better personalized treatment strategies for different cancers in the future.

Dual Targeting of EGFR and MTOR Pathways Inhibits Glioblastoma Growth by Modulating the Tumor Microenvironment

Glioblastoma's (GBM) aggressive growth is driven by redundant activation of a myriad of signaling pathways and genomic alterations in tyrosine kinase receptors, such as epidermal growth factor receptor (EGFR), which is altered in over 50% of cases. Single agents targeting EGFR have not proven effective against GBM. In this study, we aimed to identify an effective anti-tumor regimen using pharmacogenomic testing of patient-derived GBM samples, in culture and in vivo. High-throughput pharmacological screens of ten EGFR-driven GBM samples identified the combination of erlotinib (EGFRi) and MLN0128 (a mammalian target of rapamycin inhibitor, or MTORi) as the most effective at inhibiting tumor cell viability. The anti-tumor activity of erlonitib+MLN0128 was synergistic and produced inhibition of the p-EGFR, mitogen-activated protein kinase (MAPK), and Phosphoinositide 3-kinase (PI3K) pathways in culture. Using an orthotopic murine model of GBM, we show that erlotinib+MLN0128 inhibited tumor growth in vivo and significantly prolonged the survival of tumor-bearing mice. Expression profiling of tumor tissues from treated mice revealed a unique gene signature induced by erlotinib+MLN0128, consisting of downregulation of immunosuppressive chemokines in the tumor microenvironment, including C-C motif chemokine ligand 2 (CCL2) and periostin. Lower periostin levels resulted in the inhibition of Iba1+ (tumor-promoting) macrophage infiltration of GBM xenografts. Taken together, our results demonstrate that pharmacological co-targeting of EGFR and MTOR using clinically available drugs represents an effective treatment paradigm for EGFR-driven GBMs, acting both by inhibiting tumor cell growth and modulating the immune tumor microenvironment.

Patient-derived cell-based pharmacogenomic assessment to unveil underlying resistance mechanisms and novel therapeutics for advanced lung cancer

BACKGROUND

A pharmacogenomic platform using patient-derived cells (PDCs) was established to identify the underlying resistance mechanisms and tailored treatment for patients with advanced or refractory lung cancer.

METHODS

Drug sensitivity screening and multi-omics datasets were acquired from lung cancer PDCs (n = 102). Integrative analysis was performed to explore drug candidates according to genetic variants, gene expression, and clinical profiles.

RESULTS

PDCs had genomic characteristics resembled with those of solid lung cancer tissues. PDC molecular subtyping classified patients into four groups: (1) inflammatory, (2) epithelial-to-mesenchymal transition (EMT)-like, (3) stemness, and (4) epithelial growth factor receptor (EGFR)-dominant. EGFR mutations of the EMT-like subtype were associated with a reduced response to EGFR-tyrosine kinase inhibitor therapy. Moreover, although RB1/TP53 mutations were significantly enriched in small-cell lung cancer (SCLC) PDCs, they were also present in non-SCLC PDCs. In contrast to its effect in the cell lines, alpelisib (a PI3K-AKT inhibitor) significantly inhibited both RB1/TP53 expression and SCLC cell growth in our PDC model. Furthermore, cell cycle inhibitors could effectively target SCLC cells. Finally, the upregulation of transforming growth factor-β expression and the YAP/TAZ pathway was observed in osimertinib-resistant PDCs, predisposing them to the EMT-like subtype. Our platform selected XAV939 (a WNT-TNKS-β-catenin inhibitor) for the treatment of osimertinib-resistant PDCs. Using an in vitro model, we further demonstrated that acquisition of osimertinib resistance enhances invasive characteristics and EMT, upregulates the YAP/TAZ-AXL axis, and increases the sensitivity of cancer cells to XAV939.

CONCLUSIONS

Our PDC models recapitulated the molecular characteristics of lung cancer, and pharmacogenomics analysis provided plausible therapeutic candidates.

Inhibition of histone deacetylase 6 destabilizes ERK phosphorylation and suppresses cancer proliferation via modulation of the tubulin acetylation-GRP78 interaction

BACKGROUND

The leading cause of cancer-related mortality worldwide is lung cancer, and its clinical outcome and prognosis are still unsatisfactory. The understanding of potential molecular targets is necessary for clinical implications in precision diagnostic and/or therapeutic purposes. Histone deacetylase 6 (HDAC6), a major deacetylase enzyme, is a promising target for cancer therapy; however, the molecular mechanism regulating cancer pathogenesis is largely unknown.

METHODS

The clinical relevance of HDAC6 expression levels and their correlation with the overall survival rate were analyzed based on the TCGA and GEO databases. HDAC6 expression in clinical samples obtained from lung cancer tissues and patient-derived primary lung cancer cells was evaluated using qRT-PCR and Western blot analysis. The potential regulatory mechanism of HDAC6 was identified by proteomic analysis and validated by immunoblotting, immunofluorescence, microtubule sedimentation, and immunoprecipitation-mass spectrometry (IP-MS) assays using a specific inhibitor of HDAC6, trichostatin A (TSA) and RNA interference to HDAC6 (siHDAC6). Lung cancer cell growth was assessed by an in vitro 2-dimensional (2D) cell proliferation assay and 3D tumor spheroid formation using patient-derived lung cancer cells.

RESULTS

HDAC6 was upregulated in lung cancer specimens and significantly correlated with poor prognosis. Inhibition of HDAC6 by TSA and siHDAC6 caused downregulation of phosphorylated extracellular signal-regulated kinase (p-ERK), which was dependent on the tubulin acetylation status. Tubulin acetylation induced by TSA and siHDAC6 mediated the dissociation of p-ERK on microtubules, causing p-ERK destabilization. The proteomic analysis demonstrated that the molecular chaperone glucose-regulated protein 78 (GRP78) was an important scaffolder required for p-ERK localization on microtubules, and this phenomenon was significantly inhibited by either TSA, siHDAC6, or siGRP78. In addition, suppression of HDAC6 strongly attenuated an in vitro 2D lung cancer cell growth and an in vitro 3D patient derived-lung cancer spheroid growth.

CONCLUSIONS

HDAC6 inhibition led to upregulate tubulin acetylation, causing GRP78-p-ERK dissociation from microtubules. As a result, p-ERK levels were decreased, and lung cancer cell growth was subsequently suppressed. This study reveals the intriguing role and molecular mechanism of HDAC6 as a tumor promoter, and its inhibition represents a promising approach for anticancer therapy.

Prediction of drug sensitivity based on multi-omics data using deep learning and similarity network fusion approaches

With the rapid development of multi-omics technologies and accumulation of large-scale bio-datasets, many studies have conducted a more comprehensive understanding of human diseases and drug sensitivity from multiple biomolecules, such as DNA, RNA, proteins and metabolites. Using single omics data is difficult to systematically and comprehensively analyze the complex disease pathology and drug pharmacology. The molecularly targeted therapy-based approaches face some challenges, such as insufficient target gene labeling ability, and no clear targets for non-specific chemotherapeutic drugs. Consequently, the integrated analysis of multi-omics data has become a new direction for scientists to explore the mechanism of disease and drug. However, the available drug sensitivity prediction models based on multi-omics data still have problems such as overfitting, lack of interpretability, difficulties in integrating heterogeneous data, and the prediction accuracy needs to be improved. In this paper, we proposed a novel drug sensitivity prediction (NDSP) model based on deep learning and similarity network fusion approaches, which extracts drug targets using an improved sparse principal component analysis (SPCA) method for each omics data, and construct sample similarity networks based on the sparse feature matrices. Furthermore, the fused similarity networks are put into a deep neural network for training, which greatly reduces the data dimensionality and weakens the risk of overfitting problem. We use three omics of data, RNA sequence, copy number aberration and methylation, and select 35 drugs from Genomics of Drug Sensitivity in Cancer (GDSC) for experiments, including Food and Drug Administration (FDA)-approved targeted drugs, FDA-unapproved targeted drugs and non-specific therapies. Compared with some current deep learning methods, our proposed method can extract highly interpretable biological features to achieve highly accurate sensitivity prediction of targeted and non-specific cancer drugs, which is beneficial for the development of precision oncology beyond targeted therapy.

Evolution of precision oncology-guided treatment paradigms

Cancer treatment is gradually evolving from the classical use of nonspecific cytotoxic drugs targeting generic mechanisms of cell growth and proliferation. Instead, new "patient-specific treatment paradigms" that are based on an individual patient's tumor-specific molecular features are emerging, and these include "druggable" genomic alterations such as oncogenic driver mutations, downstream activities of cancer-signaling pathways, and the expression of specific genes involved in tumorigenesis and cancer progression. This evolving landscape of making evidence-based treatment decisions forms the foundation of precision oncology, which aims to deliver "the right drug, to the right patient and at the right time". The long-term vision for this approach is to maximize the treatment efficacy while minimizing exposure to ineffective therapy and reducing co-morbidity-related side effects. Successful clinical translation and implementation of this vision have the potential to revolutionize treatment paradigms from predominantly reactive, to more evidence-based, proactive and predictive care. In this article, we review the past and current approaches in precision oncology, and describe their remarkable power and limitations. We also speculate on the evolution of newly emerging methodologies of the future that can be used to address some of the key challenges associated with the existing paradigms. This article is categorized under: Cancer > Genetics/Genomics/Epigenetics Cancer > Molecular and Cellular Physiology Cancer > Computational Models.

Pharmacotypes across the genomic landscape of pediatric acute lymphoblastic leukemia and impact on treatment response

Contemporary chemotherapy for childhood acute lymphoblastic leukemia (ALL) is risk-adapted based on clinical features, leukemia genomics and minimal residual disease (MRD); however, the pharmacological basis of these prognostic variables remains unclear. Analyzing samples from 805 children with newly diagnosed ALL from three consecutive clinical trials, we determined the ex vivo sensitivity of primary leukemia cells to 18 therapeutic agents across 23 molecular subtypes defined by leukemia genomics. There was wide variability in drug response, with favorable ALL subtypes exhibiting the greatest sensitivity to L-asparaginase and glucocorticoids. Leukemia sensitivity to these two agents was highly associated with MRD although with distinct patterns and only in B cell ALL. We identified six patient clusters based on ALL pharmacotypes, which were associated with event-free survival, even after adjusting for MRD. Pharmacotyping identified a T cell ALL subset with a poor prognosis that was sensitive to targeted agents, pointing to alternative therapeutic strategies. Our study comprehensively described the pharmacological heterogeneity of ALL, highlighting opportunities for further individualizing therapy for this most common childhood cancer.

Personalized functional profiling using ex-vivo patient-derived spheroids points out the potential of an antiangiogenic treatment in a patient with a metastatic lung atypical carcinoid

Lung carcinoids are neuroendocrine tumors representing 1 to 2% of lung cancers. This study outlines the case of a patient with a metastatic lung atypical carcinoid who presented with a pleural effusion and progression of liver metastases after developing resistance to conventional treatments. Personalized functional profiling (PFP), i.e. drug screening, was performed in ex-vivo spheroids obtained from the patient's liver metastasis to identify potential therapeutic options. The drug screening results revealed cediranib, an antiangiogenic drug, as a hit drug for this patient, from a library of 66 Food and Drug Administration (FDA)-approved and investigational drugs. Based on the PFP results and the reported evidence of clinical efficacy of bevacizumab and capecitabine combination in gastro-intestinal neuroendocrine tumors, this combination was given to the patient. Four months later, the pleural effusion and pleura carcinosis regressed and the liver metastasis did not progress. The patient experienced 2 years of a stable disease under the PFP-guided personalized treatment.

Gene-centric multi-omics integration with convolutional encoders for cancer drug response prediction

MOTIVATION

Tumor heterogeneity, including genetic and transcriptomic characteristics, can reduce the efficacy of anticancer pharmacological therapy, resulting in clinical variability in patient response to therapeutic medications. Multi-omics integration can allow in silico models to provide an additional perspective on a biological system.

METHODS

In this study, we propose a gene-centric multi-channel (GCMC) architecture to integrate multi-omics for predicting cancer drug response. GCMC transformed multi-omics profiles into a three-dimensional tensor with an additional dimension for omics types. GCMC's convolutional encoders captures multi-omics profiles for each gene and yields gene-centric features to predict drug responses.

RESULTS

We evaluated GCMC on various datasets, including The Cancer Genome Atlas (TCGA) patients, patient-derived xenografts (PDX) mice models, and the Genomics of Drug Sensitivity in Cancer (GDSC) cell line datasets. GCMC achieved better performance than baseline models, including single-omics models, in more than 75% of 265 drugs from GDSC cell line datasets. Furthermore, as for the clinical applicability of GCMC, it achieved the best performance on TCGA and PDX datasets in terms of both AUPR and AUC. We also analyzed models' capability of integrating multi-omics profiles by measuring the contribution ratio of omics types. GCMC can incorporate multi-omics profiles in various manners to enhance performance for each drug type. These results suggested that GCMC can improve performance and feature extraction capability by integrating multi-omics profiles in a gene-centric manner.

EGFR, the Lazarus target for precision oncology in glioblastoma

The Lazarus effect is a rare condition that happens when someone seemingly dead shows signs of life. The epidermal growth factor receptor (EGFR) represents a target in the fatal neoplasm glioblastoma (GBM) that through a series of negative clinical trials has prompted a vocal subset of the neuro-oncology community to declare this target dead. However, an argument can be made that the core tenets of precision oncology were overlooked in the initial clinical enthusiasm over EGFR as a therapeutic target in GBM. Namely, the wrong drugs were tested on the wrong patients at the wrong time. Furthermore, new insights into the biology of EGFR in GBM vis-à-vis other EGFR-driven neoplasms, such as non-small cell lung cancer, and development of novel GBM-specific EGFR therapeutics resurrects this target for future studies. Here, we will examine the distinct EGFR biology in GBM, how it exacerbates the challenge of treating a CNS neoplasm, how these unique challenges have influenced past and present EGFR-targeted therapeutic design and clinical trials, and what adjustments are needed to therapeutically exploit EGFR in this devastating disease.

Deep Morphology Learning Enhances Ex Vivo Drug Profiling-Based Precision Medicine

Drug testing in patient biopsy-derived cells can identify potent treatments for patients suffering from relapsed or refractory hematologic cancers. Here we investigate the use of weakly supervised deep learning on cell morphologies (DML) to complement diagnostic marker-based identification of malignant and nonmalignant cells in drug testing. Across 390 biopsies from 289 patients with diverse blood cancers, DML-based drug responses show improved reproducibility and clustering of drugs with the same mode of action. DML does so by adapting to batch effects and by autonomously recognizing disease-associated cell morphologies. In a post hoc analysis of 66 patients, DML-recommended treatments led to improved progression-free survival compared with marker-based recommendations and physician's choice-based treatments. Treatments recommended by both immunofluorescence and DML doubled the fraction of patients achieving exceptional clinical responses. Thus, DML-enhanced ex vivo drug screening is a promising tool in the identification of effective personalized treatments.

SIGNIFICANCE

We have recently demonstrated that image-based drug screening in patient samples identifies effective treatment options for patients with advanced blood cancers. Here we show that using deep learning to identify malignant and nonmalignant cells by morphology improves such screens. The presented workflow is robust, automatable, and compatible with clinical routine. This article is highlighted in the In This Issue feature, p. 476.

Organic Small Molecule Contrast Agent for Targeted Photoacoustic Imaging of Patient-Derived Brain Tumors

Traditional glioblastoma (GBM) cell lines do not maintain the heterogeneity of the original tumor, cell interactions, and therapy response, thus limiting their investigation in GBM theranostics. Herein, a kind of GBM tumor-targeting nanoparticles (NPs) TCFNP@iRGD are designed and constructed, which are generated by photoacoustic (PA) contrast agent 2-(3-cyano-4,5,5-trimethylfuran-2(5H)-ylidene) malononitrile (TCF)-OH through facile nanoprecipitation and decorated with an active targeting ligand iRGD. Their potential in GBM detection via PA imaging on glioma patient-derived cells intracranial xenograft models is evaluated for the first time. Excellent tumor-specific PA mapping performance of GBM is realized by TCFNP@iRGD, demonstrating its promising potential in the clinical diagnosis of GBM.

An ex vivo platform to guide drug combination treatment in relapsed/refractory lymphoma

Although combination therapy is the standard of care for relapsed/refractory non-Hodgkin's lymphoma (RR-NHL), combination treatment chosen for an individual patient is empirical, and response rates remain poor in individuals with chemotherapy-resistant disease. Here, we evaluate an experimental-analytic method, quadratic phenotypic optimization platform (QPOP), for prediction of patient-specific drug combination efficacy from a limited quantity of biopsied tumor samples. In this prospective study, we enrolled 71 patients with RR-NHL (39 B cell NHL and 32 NK/T cell NHL) with a median of two prior lines of treatment, at two academic hospitals in Singapore from November 2017 to August 2021. Fresh biopsies underwent ex vivo testing using a panel of 12 drugs with known efficacy against NHL to identify effective single and combination treatments. Individualized QPOP reports were generated for 67 of 75 patient samples, with a median turnaround time of 6 days from sample collection to report generation. Doublet drug combinations containing copanlisib or romidepsin were most effective against B cell NHL and NK/T cell NHL samples, respectively. Off-label QPOP-guided therapy offered at physician discretion in the absence of standard options (n = 17) resulted in five complete responses. Among patients with more than two prior lines of therapy, the rates of progressive disease were lower with QPOP-guided treatments than with conventional chemotherapy. Overall, this study shows that the identification of patient-specific drug combinations through ex vivo analysis was achievable for RR-NHL in a clinically applicable time frame. These data provide the basis for a prospective clinical trial evaluating ex vivo-guided combination therapy in RR-NHL.

Towards precision oncology with patient-derived xenografts

Under the selective pressure of therapy, tumours dynamically evolve multiple adaptive mechanisms that make static interrogation of genomic alterations insufficient to guide treatment decisions. Clinical research does not enable the assessment of how various regulatory circuits in tumours are affected by therapeutic insults over time and space. Likewise, testing different precision oncology approaches informed by composite and ever-changing molecular information is hard to achieve in patients. Therefore, preclinical models that incorporate the biology and genetics of human cancers, facilitate analyses of complex variables and enable adequate population throughput are needed to pinpoint randomly distributed response predictors. Patient-derived xenograft (PDX) models are dynamic entities in which cancer evolution can be monitored through serial propagation in mice. PDX models can also recapitulate interpatient diversity, thus enabling the identification of response biomarkers and therapeutic targets for molecularly defined tumour subgroups. In this Review, we discuss examples from the past decade of the use of PDX models for precision oncology, from translational research to drug discovery. We elaborate on how and to what extent preclinical observations in PDX models have confirmed and/or anticipated findings in patients. Finally, we illustrate emerging methodological efforts that could broaden the application of PDX models by honing their predictive accuracy or improving their versatility.

Pharmacogenomic landscape of head and neck squamous cell carcinoma informs precision oncology therapy

Head and neck squamous cell carcinoma (HNSCC) is a common and frequently lethal cancer with few therapeutic options. In particular, there are few effective targeted therapies. Development of highly effective therapeutic strategies tailored to patients with HNSCC remains a pressing challenge. To address this, we present a pharmacogenomic study to facilitate precision treatments for patients with HNSCC. We established a large collection of 56 HNSCC patient-derived cells (PDCs), which recapitulated the molecular features of the original tumors. Pharmacological assessment of HNSCCs was conducted using a three-tiered high-throughput drug screening using 2248 compounds across these PDC models and an additional 18 immortalized cell lines. We integrated genomic, transcriptomic, and pharmacological analysis to predict biomarkers, gene-drug associations, and validated biomarkers. These results supported drug repurposing for multiple HNSCC subtypes, including the JAK2 inhibitor fedratinib, for low KRT18-expressing HNSCC cases, and the topoisomerase inhibitor mitoxantrone, for IL6R-activated HNSCC cases. Our results demonstrated concordance between susceptibility predictions from the PDCs and the matched patients' responses to standard clinical medication. Moreover, we identified and experimentally confirmed that high expression of ITGB1 elicited therapeutic resistance to docetaxel and high SOD1 expression conferred resistance to afatinib. We further validated ITGB1 as a predictive biomarker for the efficacy of docetaxel therapy in a phase 2 clinical trial. In summary, our study shows that this HNSCC cell resource, as well as the resulting pharmacogenomic profiles, is effective for biomarker discovery and for guiding precision oncology therapies in HNSCCs.

Patient-Derived Tumor Organoids for Guidance of Personalized Drug Therapies in Recurrent Glioblastoma

An obstacle to effective uniform treatment of glioblastoma, especially at recurrence, is genetic and cellular intertumoral heterogeneity. Hence, personalized strategies are necessary, as are means to stratify potential targeted therapies in a clinically relevant timeframe. Functional profiling of drug candidates against patient-derived glioblastoma organoids (PD-GBO) holds promise as an empirical method to preclinically discover potentially effective treatments of individual tumors. Here, we describe our establishment of a PD-GBO-based functional profiling platform and the results of its application to four patient tumors. We show that our PD-GBO model system preserves key features of individual patient glioblastomas in vivo. As proof of concept, we tested a panel of 41 FDA-approved drugs and were able to identify potential treatment options for three out of four patients; the turnaround from tumor resection to discovery of treatment option was 13, 14, and 15 days, respectively. These results demonstrate that this approach is a complement and, potentially, an alternative to current molecular profiling efforts in the pursuit of effective personalized treatment discovery in a clinically relevant time period. Furthermore, these results warrant the use of PD-GBO platforms for preclinical identification of new drugs against defined morphological glioblastoma features.

Proof of concept for the use of trained sniffer dogs to detect osteosarcoma

Sarcomas are mesenchymal cancers which often show an aggressive behavior and patient survival largely depends on an early detection. In last years, much attention has been given to the fact that cancer patients release specific odorous volatile organic compounds (VOCs) that can be efficiently detected by properly trained sniffer dogs. Here, we have evaluated for the first time the ability of sniffer dogs (n = 2) to detect osteosarcoma cell cultures and patient samples. One of the two dogs was successfully trained to discriminate osteosarcoma patient-derived primary cells from mesenchymal stem/stromal cells (MSCs) obtained from healthy individuals. After the training phase, the dog was able to detect osteosarcoma specific odor cues in a different panel of 6 osteosarcoma cell lines with sensitivity and specificity rates between 95 and 100%. Moreover, the same VOCs were also detected by the sniffer dog in saliva samples from osteosarcoma patients (n = 2) and discriminated from samples from healthy individuals with a similar efficacy. Altogether, these results indicate that there are common odor profiles shared by cultures of osteosarcoma cells and body fluid samples from patients and provide a first proof of concept about the potential of canine odor detection as a non-invasive screening method to detect osteosarcomas.

The Patient-Derived Cancer Organoids: Promises and Challenges as Platforms for Cancer Discovery

The cancer burden is rapidly increasing in most countries, and thus, new anticancer drugs for effective cancer therapy must be developed. Cancer model systems that recapitulate the biological processes of human cancers are one of the cores of the drug development process. PDCO has emerged as a unique model that preserves the genetic, physiological, and histologic characteristics of original cancer, including inter- and intratumoral heterogeneities. Due to these advantages, the PCDO model is increasingly investigated for anticancer drug screening and efficacy testing, preclinical patient stratification, and precision medicine for selecting the most effective anticancer therapy for patients. Here, we review the prospects and limitations of PDCO compared to the conventional cancer models. With advances in culture success rates, co-culture systems with the tumor microenvironment, organoid-on-a-chip technology, and automation technology, PDCO will become the most promising model to develop anticancer drugs and precision medicine.

Leveraging Deep Learning Techniques and Integrated Omics Data for Tailored Treatment of Breast Cancer

Multiomics data of cancer patients and cell lines, in synergy with deep learning techniques, have aided in unravelling predictive problems related to cancer research and treatment. However, there is still room for improvement in the performance of the existing models based on the aforementioned combination. In this work, we propose two models that complement the treatment of breast cancer patients. First, we discuss our deep learning-based model for breast cancer subtype classification. Second, we propose DCNN-DR, a deep convolute.ion neural network-drug response method for predicting the effectiveness of drugs on in vitro and in vivo breast cancer datasets. Finally, we applied DCNN-DR for predicting effective drugs for the basal-like breast cancer subtype and validated the results with the information available in the literature. The models proposed use late integration methods and have fairly better predictive performance compared to the existing methods. We use the Pearson correlation coefficient and accuracy as the performance measures for the regression and classification models, respectively.

Zebrafish Patient-Derived Avatars from Digestive Cancers for Anti-cancer Therapy Screening

Patient-derived xenografts (PDXs), also called "avatars," are generated by the implantation of human primary tumor cells or tissues into a host animal. Given the complexity and unique characteristics of each tumor, PDXs are models of choice in cancer research and precision medicine. In this context, the zebrafish PDX model (zPDX or zAvatar) has been recognized as a promising in vivo model to directly challenge patient cells with anti-cancer therapies in a personalized manner. The assay relies on the injection of tumor cells from patients into zebrafish embryos to then test and identify the best available drug combination for a particular patient. Compared to mouse PDXs, zAvatar assays take less time and do not require in vitro or in vivo cell expansion. The present article describes how to generate zAvatars from resected digestive cancer from surgeries and how to then use them for anti-cancer therapy screening. We describe the steps for tumor sample collection and cryopreservation, sample preparation and fluorescent labeling for microinjection into zebrafish embryos, drug administration, and analysis of tumor behavior by single-cell confocal imaging. We provide detailed protocols and helpful tips for performing this assay, and we address the technical challenges associated with the workflow. © 2022 Wiley Periodicals LLC. Basic Protocol 1: Patient tumor sample collection and cryopreservation Basic Protocol 2: Generation of zAvatars and anti-cancer treatment Basic Protocol 3: Whole-mount immunofluorescence Basic Protocol 4: Confocal imaging and analysis.

A clinically compatible drug-screening platform based on organotypic cultures identifies vulnerabilities to prevent and treat brain metastasis

We report a medium-throughput drug-screening platform (METPlatform) based on organotypic cultures that allows to evaluate inhibitors against metastases growing in situ. By applying this approach to the unmet clinical need of brain metastasis, we identified several vulnerabilities. Among them, a blood-brain barrier permeable HSP90 inhibitor showed high potency against mouse and human brain metastases at clinically relevant stages of the disease, including a novel model of local relapse after neurosurgery. Furthermore, in situ proteomic analysis applied to metastases treated with the chaperone inhibitor uncovered a novel molecular program in brain metastasis, which includes biomarkers of poor prognosis and actionable mechanisms of resistance. Our work validates METPlatform as a potent resource for metastasis research integrating drug-screening and unbiased omic approaches that is compatible with human samples. Thus, this clinically relevant strategy is aimed to personalize the management of metastatic disease in the brain and elsewhere.

Functional Therapeutic Target Validation Using Pediatric Zebrafish Xenograft Models

Simple Summary

Despite the major progress of precision and personalized oncology, a significant therapeutic benefit is only achieved in cases with directly druggable genetic alterations. This highlights the need for additional methods that reliably predict each individual patient’s response in a clinically meaningful time, e.g., through ex vivo functional drug screen profiling. Moreover, patient-derived xenograft (PDX) models are more predictive than cell culture studies, as they reconstruct cell–cell and cell–extracellular matrix (ECM) interactions and consider the tumor microenvironment, drug metabolism and toxicities. Zebrafish PDXs (zPDX) are nowadays emerging as a fast model allowing for multiple drugs to be tested at the same time in a clinically relevant time window. Here, we show that functional drug response profiling of zPDX from primary material obtained through the INdividualized Therapy FOr Relapsed Malignancies in Childhood (INFORM) pediatric precision oncology pipeline provides additional key information for personalized precision oncology.

Abstract

The survival rate among children with relapsed tumors remains poor, due to tumor heterogeneity, lack of directly actionable tumor drivers and multidrug resistance. Novel personalized medicine approaches tailored to each tumor are urgently needed to improve cancer treatment. Current pediatric precision oncology platforms, such as the INFORM (INdividualized Therapy FOr Relapsed Malignancies in Childhood) study, reveal that molecular profiling of tumor tissue identifies targets associated with clinical benefit in a subgroup of patients only and should be complemented with functional drug testing. In such an approach, patient-derived tumor cells are exposed to a library of approved oncological drugs in a physiological setting, e.g., in the form of animal avatars injected with patient tumor cells. We used molecularly fully characterized tumor samples from the INFORM study to compare drug screen results of individual patient-derived cell models in functional assays: (i) patient-derived spheroid cultures within a few days after tumor dissociation; (ii) tumor cells reisolated from the corresponding mouse PDX; (iii) corresponding long-term organoid-like cultures and (iv) drug evaluation with the corresponding zebrafish PDX (zPDX) model. Each model had its advantage and complemented the others for drug hit and drug combination selection. Our results provide evidence that in vivo zPDX drug screening is a promising add-on to current functional drug screening in precision medicine platforms.

Diffusion kernel-based predictive modeling of KRAS dependency in KRAS wild type cancer cell lines

Recent progress in clinical development of KRAS inhibitors has raised interest in predicting the tumor dependency on frequently mutated RAS-pathway oncogenes. However, even without such activating mutations, RAS proteins represent core components in signal integration of several membrane-bound kinases. This raises the question of applications of specific inhibitors independent from the mutational status. Here, we examined CRISPR/RNAi data from over 700 cancer cell lines and identified a subset of cell lines without KRAS gain-of-function mutations (KRAS^wt) which are dependent on KRAS expression. Combining machine learning-based modeling and whole transcriptome data with prior variable selection through protein-protein interaction network analysis by a diffusion kernel successfully predicted KRAS dependency in the KRAS^wt subgroup and in all investigated cancer cell lines. In contrast, modeling by RAS activating events (RAE) or previously published RAS RNA-signatures did not provide reliable results, highlighting the heterogeneous distribution of RAE in KRAS^wt cell lines and the importance of methodological references for expression signature modeling. Furthermore, we show that predictors of KRAS^wt models contain non-substitutable information signals, indicating a KRAS dependency phenotype in the KRAS^wt subgroup. Our data suggest that KRAS dependent cancers harboring KRAS wild type status could be targeted by directed therapeutic approaches. RNA-based machine learning models could help in identifying responsive and non-responsive tumors.

An overview of kinase downregulators and recent advances in discovery approaches

Since the clinical approval of imatinib, the discovery of protein kinase downregulators entered a prosperous age. However, challenges still exist in the discovery of kinase downregulator drugs, such as the high failure rate during development, side effects, and drug-resistance problems. With the progress made through multidisciplinary efforts, an increasing number of new approaches have been applied to solve the above problems during the discovery process of kinase downregulators. In terms of in vitro and in vivo drug evaluation, progress was also made in cellular and animal model platforms for better and more clinically relevant drug assessment. Here, we review the advances in drug design strategies, drug property evaluation technologies, and efficacy evaluation models and technologies. Finally, we discuss the challenges and perspectives in the development of kinase downregulator drugs.

Predicting heterogeneity in clone-specific therapeutic vulnerabilities using single-cell transcriptomic signatures

While understanding molecular heterogeneity across patients underpins precision oncology, there is increasing appreciation for taking intra-tumor heterogeneity into account. Based on large-scale analysis of cancer omics datasets, we highlight the importance of intra-tumor transcriptomic heterogeneity (ITTH) for predicting clinical outcomes. Leveraging single-cell RNA-seq (scRNA-seq) with a recommender system (CaDRReS-Sc), we show that heterogeneous gene-expression signatures can predict drug response with high accuracy (80%). Using patient-proximal cell lines, we established the validity of CaDRReS-Sc's monotherapy (Pearson r>0.6) and combinatorial predictions targeting clone-specific vulnerabilities (>10% improvement). Applying CaDRReS-Sc to rapidly expanding scRNA-seq compendiums can serve as in silico screen to accelerate drug-repurposing studies. Availability: https://github.com/CSB5/CaDRReS-Sc .