Integrating molecular profiles into clinical frameworks through the Molecular Oncology Almanac to prospectively guide precision oncology

New horizons at the interface of artificial intelligence and translational cancer research

Artificial intelligence (AI) is increasingly being utilized in cancer research as a computational strategy for analyzing multiomics datasets. Advances in single-cell and spatial profiling technologies have contributed significantly to our understanding of tumor biology, and AI methodologies are now being applied to accelerate translational efforts, including target discovery, biomarker identification, patient stratification, and therapeutic response prediction. Despite these advancements, the integration of AI into clinical workflows remains limited, presenting both challenges and opportunities. This review discusses AI applications in multiomics analysis and translational oncology, emphasizing their role in advancing biological discoveries and informing clinical decision-making. Key areas of focus include cellular heterogeneity, tumor microenvironment interactions, and AI-aided diagnostics. Challenges such as reproducibility, interpretability of AI models, and clinical integration are explored, with attention to strategies for addressing these hurdles. Together, these developments underscore the potential of AI and multiomics to enhance precision oncology and contribute to advancements in cancer care.

A Cross-Sectional Study of Variant Interpretation and Reporting of NGS Data Using Tertiary Analysis Software: Navify® Mutation Profiler

INTRODUCTION

Personalized medicine has revolutionized the clinical management of patients with solid tumors. However, the large volumes of molecular data derived from next-generation sequencing (NGS) and the lack of harmonized bioinformatics pipelines drastically impact the clinical management of patients with solid tumors. A possible solution to streamline the molecular interpretation and reporting of NGS data would be to adopt automated data analysis software. In this study, we tested the clinical efficiency of the Navify Mutation Profiler (nMP) software in improving the interpretation of NGS data analysis in diagnostic routine samples from patients with solid tumors.

METHODS

This study included one coordinating institution (Federico II University of Naples) and five other Italian institutions. Variant call format (VCF) files from reference standard samples previously tested by the coordinating institution and from n = 8 diagnostic routine samples (n = 2 from colorectal carcinoma; n = 2 from non-small cell lung cancer; n = 2 from advanced melanoma; and n = 2 from patients with gastrointestinal stromal tumors) and previously analyzed by each participating institution (n = 5) with standardized internal analysis workflows were uploaded onto the Navify® Mutation Profiler (nMP) system (Roche Sequencing Solutions, Pleasanton, CA, USA) for automated analysis and interpretation of DNA and RNA molecular alterations analytical parameters, molecular profiling, and clinical interpretation were carried out by the nMP system and compared with the standard workflow data analyzed by the participating institutions.

RESULTS

Overall, all VCF files were successfully submitted and interpreted by the nMP system. A concordance agreement rate of 89.6% was observed between the automated and standard workflow systems. In particular, DNA and RNA molecular profiles obtained with the nMP system matched those obtained with standardized approaches in 44 out of 48 patients (91.7%) and in 11 out of 12 (91.7%) cases, respectively. In addition, the nMP system evidenced wild-type variants in 6 out of 7 (85.7%) cases.

CONCLUSIONS

The nMP system represents a valid, easily manageable, and clinically useful system to interpret NGS data on diagnostic routine samples from patients with solid tumors.

ESMO Recommendations on clinical reporting of genomic test results for solid cancers

BACKGROUND

Genomic tumour profiling has a crucial role in the management of patients with solid cancers, as it helps selecting and prioritising therapeutic interventions based on prognostic and predictive biomarkers, as well as identifying markers of hereditary cancers. Harmonised approaches to interpret the results of genomic testing are needed to support physicians in their decision making, prevent inequalities in precision medicine and maximise patient benefit from available cancer management options.

METHODS

The European Society for Medical Oncology (ESMO) Translational Research and Precision Medicine Working Group assembled a group of international experts to propose recommendations for preparing clinical genomic reports for solid cancers. These recommendations aim to foster best practices in integrating genomic testing within clinical settings. After review of available evidence, several rounds of surveys and focused discussions were conducted to reach consensus on the recommendation statements. Only consensus recommendations were reported. Recommendation statements were graded in two tiers based on their clinical importance: level A (required to maintain common standards in reporting) and level B (optional but necessary to achieve ideal practice).

RESULTS

Genomics reports should present key information in a front page(s) followed by supplementary information in one or more appendices. Reports should be structured into sections: (i) patient and sample details; (ii) assay and data analysis characteristics; (iii) sample-specific assay performance and quality control; (iv) genomic alterations and their functional annotation; (v) clinical actionability assessment and matching to potential therapy indications; and (vi) summary of the main findings. Specific recommendations to prepare each of these sections are made.

CONCLUSIONS

We present a set of recommendations aimed at structuring genomics reports to enhance physician comprehension of genomic profiling results for solid cancers. Communication between ordering physicians and professionals reporting genomic data is key to minimise uncertainties and to optimise the impact of genomic tests in patient care.

NECTIN4 Amplification Is Frequent in Solid Tumors and Predicts Enfortumab Vedotin Response in Metastatic Urothelial Cancer

PURPOSE

The anti-NECTIN4 antibody-drug conjugate enfortumab vedotin (EV) is approved for patients with metastatic urothelial cancer (mUC). However, durable benefit is only achieved in a small, yet uncharacterized patient subset. NECTIN4 is located on chromosome 1q23.3, and 1q23.3 gains represent frequent copy number variations (CNVs) in urothelial cancer. Here, we aimed to evaluate NECTIN4 amplifications as a genomic biomarker to predict EV response in patients with mUC.

MATERIALS AND METHODS

We established a NECTIN4-specific fluorescence in situ hybridization (FISH) assay to assess the predictive value of NECTIN4 CNVs in a multicenter EV-treated mUC patient cohort (mUC-EV, n = 108). CNVs were correlated with membranous NECTIN4 protein expression, EV treatment responses, and outcomes. We also assessed the prognostic value of NECTIN4 CNVs measured in metastatic biopsies of non-EV-treated mUC (mUC-non-EV, n = 103). Furthermore, we queried The Cancer Genome Atlas (TCGA) data sets (10,712 patients across 32 cancer types) for NECTIN4 CNVs.

RESULTS

NECTIN4 amplifications are frequent genomic events in muscle-invasive bladder cancer (TCGA bladder cancer data set: approximately 17%) and mUC (approximately 26% in our mUC cohorts). In mUC-EV, NECTIN4 amplification represents a stable genomic alteration during metastatic progression and associates with enhanced membranous NECTIN4 protein expression. Ninety-six percent (27 of 28) of patients with NECTIN4 amplifications demonstrated objective responses to EV compared with 32% (24 of 74) in the nonamplified subgroup (P < .001). In multivariable Cox analysis adjusted for age, sex, and Bellmunt risk factors, NECTIN4 amplifications led to a 92% risk reduction for death (hazard ratio, 0.08 [95% CI, 0.02 to 0.34]; P < .001). In the mUC-non-EV, NECTIN4 amplifications were not associated with outcomes. TCGA Pan-Cancer analysis demonstrated that NECTIN4 amplifications occur frequently in other cancers, for example, in 5%-10% of breast and lung cancers.

CONCLUSION

NECTIN4 amplifications are genomic predictors of EV responses and long-term survival in patients with mUC.

An artificial intelligence-assisted clinical framework to facilitate diagnostics and translational discovery in hematologic neoplasia

BACKGROUND

The increasing volume and intricacy of sequencing data, along with other clinical and diagnostic data, like drug responses and measurable residual disease, creates challenges for efficient clinical comprehension and interpretation. Using paediatric B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) as a use case, we present an artificial intelligence (AI)-assisted clinical framework clinALL that integrates genomic and clinical data into a user-friendly interface to support routine diagnostics and reveal translational insights for hematologic neoplasia.

METHODS

We performed targeted RNA sequencing in 1365 cases with haematological neoplasms, primarily paediatric B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) from the AIEOP-BFM ALL study. We carried out fluorescence in situ hybridization (FISH), karyotyping and arrayCGH as part of the routine diagnostics. The analysis results of these assays as well as additional clinical information were integrated into an interactive web interface using Bokeh, where the main graph is based on Uniform Manifold Approximation and Projection (UMAP) analysis of the gene expression data. At the backend of the clinALL, we built both shallow machine learning models and a deep neural network using Scikit-learn and PyTorch respectively.

FINDINGS

By applying clinALL, 78% of undetermined patients under the current diagnostic protocol were stratified, and ambiguous cases were investigated. Translational insights were discovered, including IKZF1plus status dependent subpopulations of BCR::ABL1 positive patients, and a subpopulation within ETV6::RUNX1 positive patients that has a high relapse frequency. Our best machine learning models, LDA and PASNET-like neural network models, achieve F1 scores above 97% in predicting patients' subgroups.

INTERPRETATION

An AI-assisted clinical framework that integrates both genomic and clinical data can take full advantage of the available data, improve point-of-care decision-making and reveal clinically relevant insights promptly. Such a lightweight and easily transferable framework works for both whole transcriptome data as well as the cost-effective targeted RNA-seq, enabling efficient and equitable delivery of personalized medicine in small clinics in developing countries.

FUNDING

German Ministry of Education and Research (BMBF), German Research Foundation (DFG) and Foundation for Polish Science.

Building Research Infrastructure to Develop Greater Learning Efficiencies (BRIDGE)

In this manuscript, we outline our developed version of a Learning Health System (LHS) in oncology implemented at the Department of Veterans Affairs (VA). Transferring healthcare into an LHS framework has been one of the spearpoints of VA's Central Office and given the general lack of evidence generated through randomized control clinical trials to guide medical decisions in oncology, this domain is one of the most suitable for this change. We describe our technical solution, which includes a large real-world data repository, a data science and algorithm development framework, and the mechanism by which results are brought back to the clinic and to the patient. Additionally, we propose the need for a bridging framework that requires collaboration between informatics specialists and medical professionals to integrate knowledge generation into the clinical workflow at the point of care.

Tri©DB: an integrated platform of knowledgebase and reporting system for cancer precision medicine

BACKGROUND

With the development of cancer precision medicine, a huge amount of high-dimensional cancer information has rapidly accumulated regarding gene alterations, diseases, therapeutic interventions and various annotations. The information is highly fragmented across multiple different sources, making it highly challenging to effectively utilize and exchange the information. Therefore, it is essential to create a resource platform containing well-aggregated, carefully mined, and easily accessible data for effective knowledge sharing.

METHODS

In this study, we have developed "Consensus Cancer Core" (Tri©DB), a new integrative cancer precision medicine knowledgebase and reporting system by mining and harmonizing multifaceted cancer data sources, and presenting them in a centralized platform with enhanced functionalities for accessibility, annotation and analysis.

RESULTS

The knowledgebase provides the currently most comprehensive information on cancer precision medicine covering more than 40 annotation entities, many of which are novel and have never been explored previously. Tri©DB offers several unique features: (i) harmonizing the cancer-related information from more than 30 data sources into one integrative platform for easy access; (ii) utilizing a variety of data analysis and graphical tools for enhanced user interaction with the high-dimensional data; (iii) containing a newly developed reporting system for automated annotation and therapy matching for external patient genomic data. Benchmark test indicated that Tri©DB is able to annotate 46% more treatments than two officially recognized resources, oncoKB and MCG. Tri©DB was further shown to have achieved 94.9% concordance with administered treatments in a real clinical trial.

CONCLUSIONS

The novel features and rich functionalities of the new platform will facilitate full access to cancer precision medicine data in one single platform and accommodate the needs of a broad range of researchers not only in translational medicine, but also in basic biomedical research. We believe that it will help to promote knowledge sharing in cancer precision medicine. Tri©DB is freely available at www.biomeddb.org , and is hosted on a cutting-edge technology architecture supporting all major browsers and mobile handsets.

Normalization of drug and therapeutic concepts with Thera-Py

Objective

The diversity of nomenclature and naming strategies makes therapeutic terminology difficult to manage and harmonize. As the number and complexity of available therapeutic ontologies continues to increase, the need for harmonized cross-resource mappings is becoming increasingly apparent. This study creates harmonized concept mappings that enable the linking together of like-concepts despite source-dependent differences in data structure or semantic representation.

Materials and Methods

For this study, we created Thera-Py, a Python package and web API that constructs searchable concepts for drugs and therapeutic terminologies using 9 public resources and thesauri. By using a directed graph approach, Thera-Py captures commonly used aliases, trade names, annotations, and associations for any given therapeutic and combines them under a single concept record.

Results

We highlight the creation of 16 069 unique merged therapeutic concepts from 9 distinct sources using Thera-Py and observe an increase in overlap of therapeutic concepts in 2 or more knowledge bases after harmonization using Thera-Py (9.8%-41.8%).

Conclusion

We observe that Thera-Py tends to normalize therapeutic concepts to their underlying active ingredients (excluding nondrug therapeutics, eg, radiation therapy, biologics), and unifies all available descriptors regardless of ontological origin.

Molecular tumour boards - current and future considerations for precision oncology

Over the past 15 years, rapid progress has been made in developmental therapeutics, especially regarding the use of matched targeted therapies against specific oncogenic molecular alterations across cancer types. Molecular tumour boards (MTBs) are panels of expert physicians, scientists, health-care providers and patient advocates who review and interpret molecular-profiling results for individual patients with cancer and match each patient to available therapies, which can include investigational drugs. Interpretation of the molecular alterations found in each patient is a complicated task that requires an understanding of their contextual functional effects and their correlations with sensitivity or resistance to specific treatments. The criteria for determining the actionability of molecular alterations and selecting matched treatments are constantly evolving. Therefore, MTBs have an increasingly necessary role in optimizing the allocation of biomarker-directed therapies and the implementation of precision oncology. Ultimately, increased MTB availability, accessibility and performance are likely to improve patient care. The challenges faced by MTBs are increasing, owing to the plethora of identifiable molecular alterations and immune markers in tumours of individual patients and their evolving clinical significance as more and more data on patient outcomes and results from clinical trials become available. Beyond next-generation sequencing, broader biomarker analyses can provide useful information. However, greater funding, resources and expertise are needed to ensure the sustainability of MTBs and expand their outreach to underserved populations. Harmonization between practice and policy will be required to optimally implement precision oncology. Herein, we discuss the evolving role of MTBs and current and future considerations for their use in precision oncology.

Genome-guided discovery of cancer therapeutic targets

The success of precision oncology-which aims to match the right therapies to the right patients based on molecular status-is predicated on a robust pipeline of molecular targets against which therapies can be developed. Recent advances in genomics and functional genetics have enabled the unbiased discovery of novel molecular targets at scale. We summarize the promise and challenges in integrating genomic and functional genetic landscapes of cancer to establish the next generation of cancer targets.

myCMIE: My cancer molecular information exchange

Molecular profiling reports (MPRs) are critical for determining treatment options for cancer patients. They include several pages of information on genomic findings, drugs, and trial options that are challenging to synthesize for effectively and expeditiously informing on treatment. Xu and Kowalski present a web application, myCMIE, that synthesizes MPR content to define a patient-centric, information system in which molecular profiles are exchanged between a query case(s) and public resources or user-input case series for context-informed treatment and conjecture with therapeutic implication. myCMIE offers an interactive build of coordinately connected digital-twin communities to expand our understanding of treatment context with multiple visuals to stimulate discussions among diverse stakeholders in care.

A decision support system based on artificial intelligence and systems biology for the simulation of pancreatic cancer patient status

Oncology treatments require continuous individual adjustment based on the measurement of multiple clinical parameters. Prediction tools exploiting the patterns present in the clinical data could be used to assist decision making and ease the burden associated to the interpretation of all these parameters. The goal of this study was to predict the evolution of patients with pancreatic cancer at their next visit using information routinely recorded in health records, providing a decision-support system for clinicians. We selected hematological variables as the visit's clinical outcomes, under the assumption that they can be predictive of the evolution of the patient. Multivariate models based on regression trees were generated to predict next-visit values for each of the clinical outcomes selected, based on the longitudinal clinical data as well as on molecular data sets streaming from in silico simulations of individual patient status at each visit. The models predict, with a mean prediction score (balanced accuracy) of 0.79, the evolution trends of eosinophils, leukocytes, monocytes, and platelets. Time span between visits and neutropenia were among the most common factors contributing to the predicted evolution. The inclusion of molecular variables from the systems-biology in silico simulations provided a molecular background for the observed variations in the selected outcome variables, mostly in relation to the regulation of hematopoiesis. In spite of its limitations, this study serves as a proof of concept for the application of next-visit prediction tools in real-world settings, even when available data sets are small.

Challenging gold standard hematology diagnostics through the introduction of whole genome sequencing and artificial intelligence

The diagnosis of hematological malignancies is rather complex and requires the application of a plethora of different assays, techniques and methodologies. Some of the methods, like cytomorphology, have been in use for decades, while other methods, such as next-generation sequencing or even whole genome sequencing (WGS), are relatively new. The application of the methods and the evaluation of the results require distinct skills and knowledge and place different demands on the practitioner. However, even with experienced hematologists, diagnostic ambiguity remains a regular occurrence and the comprehensive analysis of high-dimensional WGS data soon exceeds any human's capacity. Hence, in order to reduce inter-observer variability and to improve the timeliness and accuracy of diagnoses, machine learning based approaches have been developed to assist in the decision making process. Moreover, to achieve the goal of precision oncology, comprehensive genomic profiling is increasingly being incorporated into routine standard of care.

Artificial intelligence in hematological diagnostics: Game changer or gadget?

The future of clinical diagnosis and treatment of hematologic diseases will inevitably involve the integration of artificial intelligence (AI)-based systems into routine practice to support the hematologists' decision making. Several studies have shown that AI-based models can already be used to automatically differentiate cells, reliably detect malignant cell populations, support chromosome banding analysis, and interpret clinical variants, contributing to early disease detection and prognosis. However, even the best tool can become useless if it is misapplied or the results are misinterpreted. Therefore, in order to comprehensively judge and correctly apply newly developed AI-based systems, the hematologist must have a basic understanding of the general concepts of machine learning. In this review, we provide the hematologist with a comprehensive overview of various machine learning techniques, their current implementations and approaches in different diagnostic subfields (e.g., cytogenetics, molecular genetics), and the limitations and unresolved challenges of the systems.

Addition of Germline Testing to Tumor-Only Sequencing Improves Detection of Pathogenic Germline Variants in Men With Advanced Prostate Cancer

PURPOSE

Guidelines recommend somatic and germline testing for men with advanced prostate cancer (PCa). Barriers to widespread implementation result in underutilization of germline testing. Somatic testing alone risks missing pathogenic germline variants (PGVs). We sought to determine whether the addition of germline testing to tumor-only sequencing improves detection of PGVs in men with advanced PCa. Secondarily, we sought to define the added value of combining somatic and germline testing to optimize detection of clinically actionable alterations.

PATIENTS AND METHODS

We analyzed results of independent germline testing and tumor-only sequencing from 100 men with advanced PCa from a prospective clinical trial (ClinicalTrials.gov identifier: NCT03328091). The primary outcome was the proportion of PGVs not reported with tumor-only sequencing. The secondary outcome was the association of locus-specific loss of heterozygosity for PGVs in homologous recombination genes with clinical-genomic features.

RESULTS

In the 100 men who underwent germline testing and tumor-only sequencing, 24 PGVs were identified, 17 of which were clinically actionable, in 23 patients. Tumor-only sequencing failed to report four (17%) of the PGVs. One additional PGV (4.2%) had variant allele frequency on tumor-sequencing below the threshold for follow-up germline testing. When integrating tumor-only sequencing with germling testing results, 33% of patients harbored clinically actionable alterations. Rates of locus-specific loss of heterozygosity were higher for BRCA2 PGVs in castration-resistant PCa than PGVs in other homologous recombination genes in hormone-sensitive PCa (P = .029).

CONCLUSION

Tumor-only sequencing failed to report more than 20% of PGVs in men with advanced PCa. These findings strongly support guideline recommendations for universal germline and somatic testing in this population. Combining tumor and germline sequencing doubled the chance of detecting a clinically actionable alteration.

Detection of oncogenic and clinically actionable mutations in cancer genomes critically depends on variant calling tools

Motivation

The analysis of cancer genomes provides fundamental information about its etiology, the processes driving cell transformation or potential treatments. While researchers and clinicians are often only interested in the identification of oncogenic mutations, actionable variants or mutational signatures, the first crucial step in the analysis of any tumor genome is the identification of somatic variants in cancer cells (i.e. those that have been acquired during their evolution). For that purpose, a wide range of computational tools have been developed in recent years to detect somatic mutations in sequencing data from tumor samples. While there have been some efforts to benchmark somatic variant calling tools and strategies, the extent to which variant calling decisions impact the results of downstream analyses of tumor genomes remains unknown.

Results

Here, we quantify the impact of variant calling decisions by comparing the results obtained in three important analyses of cancer genomics data (identification of cancer driver genes, quantification of mutational signatures and detection of clinically actionable variants) when changing the somatic variant caller (MuSE, MuTect2, SomaticSniper and VarScan2) or the strategy to combine them (Consensus of two, Consensus of three and Union) across all 33 cancer types from The Cancer Genome Atlas. Our results show that variant calling decisions have a significant impact on these analyses, creating important differences that could even impact treatment decisions for some patients. Moreover, the Consensus of three calling strategy to combine the output of multiple variant calling tools, a very widely used strategy by the research community, can lead to the loss of some cancer driver genes and actionable mutations. Overall, our results highlight the limitations of widespread practices within the cancer genomics community and point to important differences in critical analyses of tumor sequencing data depending on variant calling, affecting even the identification of clinically actionable variants.

Availability and implementation

Code is available at https://github.com/carlosgarciaprieto/VariantCallingClinicalBenchmark.

Supplementary information

Supplementary data are available at Bioinformatics online.

Prostate cancer research in the 21st century; report from the 2021 Coffey-Holden prostate cancer academy meeting

INTRODUCTION

The 2021 Coffey-Holden Prostate Cancer Academy (CHPCA) Meeting, "Prostate Cancer Research in the 21st Century," was held virtually, from June 24-25, 2021.

METHODS

The CHPCA Meeting is organized by the Prostate Cancer Foundation as a unique discussion-oriented meeting focusing on critical topics in prostate cancer research envisioned to bridge the next major advances in prostate cancer biology and treatment. The 2021 CHPCA Meeting was virtually attended by 89 investigators and included 31 talks over nine sessions.

RESULTS

Major topic areas discussed at the meeting included: cancer genomics and sequencing, functional genomic approaches to studying mediators of plasticity, emerging signaling pathways in metastatic castration resistant prostate cancer, Wnt signaling biology and the challenges of targeted therapy, clonal hematopoiesis, neuroendocrine cell plasticity and antitumor immunity, cancer immunotherapy and its synergizers, and imaging the tumor microenvironment and metabolism.

DISCUSSION

This meeting report summarizes the research presented at the 2021 CHPCA Meeting. We hope that publication of this knowledge will accelerate new understandings and the development of new biomarkers and treatments for prostate cancer.

Interdisciplinary team science to understand and intercept rare cancers

For most rare cancers, precision oncology approaches are not established because these entities are poorly understood and their investigation requires the collaboration of many centers. The MASTER precision oncology network demonstrates that clinical whole-genome/exome and RNA sequencing yield molecular mechanism-aware treatments that benefit a substantial proportion of patients with advanced rare cancers and will prepare the ground for future clinical trials.