Machine learning for genetics-based classification and treatment response prediction in cancer of unknown primary

Deep learning using histological images for gene mutation prediction in lung cancer: a multicentre retrospective study

BACKGROUND

Accurate detection of driver gene mutations is crucial for treatment planning and predicting prognosis for patients with lung cancer. Conventional genomic testing requires high-quality tissue samples and is time-consuming and resource-consuming, and as a result, is not available for most patients, especially those in low-resource settings. We aimed to develop an annotation-free Deep learning-enabled artificial intelligence method to predict GEne Mutations (DeepGEM) from routinely acquired histological slides.

METHODS

In this multicentre retrospective study, we collected data for patients with lung cancer who had a biopsy and multigene next-generation sequencing done at 16 hospitals in China (with no restrictions on age, sex, or histology type), to form a large multicentre dataset comprising paired pathological image and multiple gene mutation information. We also included patients from The Cancer Genome Atlas (TCGA) publicly available dataset. Our developed model is an instance-level and bag-level co-supervised multiple instance learning method with label disambiguation design. We trained and initially tested the DeepGEM model on the internal dataset (patients from the First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China), and further evaluated it on the external dataset (patients from the remaining 15 centres) and the public TCGA dataset. Additionally, a dataset of patients from the same medical centre as the internal dataset, but without overlap, was used to evaluate the model's generalisation ability to biopsy samples from lymph node metastases. The primary objective was the performance of the DeepGEM model in predicting gene mutations (area under the curve [AUC] and accuracy) in the four prespecified groups (ie, the hold-out internal test set, multicentre external test set, TCGA set, and lymph node metastases set).

FINDINGS

Assessable pathological images and multigene testing information were available for 3697 patients who had biopsy and multigene next-generation sequencing done between Jan 1, 2018, and March 31, 2022, at the 16 centres. We excluded 60 patients with low-quality images. We included 3767 images from 3637 consecutive patients (1978 [54·4%] men, 1514 [41·6%] women, 145 [4·0%] unknown; median age 60 years [IQR 52-67]), with 1716 patients in the internal dataset, 1718 patients in the external dataset, and 203 patients in the lymph node metastases dataset. The DeepGEM model showed robust performance in the internal dataset: for excisional biopsy samples, AUC values for gene mutation prediction ranged from 0·90 (95% CI 0·77-1·00) to 0·97 (0·93-1·00) and accuracy values ranged from 0·91 (0·85-0·98) to 0·97 (0·93-1·00); for aspiration biopsy samples, AUC values ranged from 0·85 (0·80-0·91) to 0·95 (0·86-1·00) and accuracy values ranged from 0·79 (0·74-0·85) to 0·99 (0·98-1·00). In the multicentre external dataset, for excisional biopsy samples, AUC values ranged from 0·80 (95% CI 0·75-0·85) to 0·91 (0·88-1·00) and accuracy values ranged from 0·79 (0·76-0·82) to 0·95 (0·93-0·96); for aspiration biopsy samples, AUC values ranged from 0·76 (0·70-0·83) to 0·87 (0·80-0·94) and accuracy values ranged from 0·76 (0·74-0·79) to 0·97 (0·96-0·98). The model also showed strong performance on the TCGA dataset (473 patients; 535 slides; AUC values ranged from 0·82 [95% CI 0·71-0·93] to 0·96 [0·91-1·00], accuracy values ranged from 0·79 [0·70-0·88] to 0·95 [0·90-1·00]). The DeepGEM model, trained on primary region biopsy samples, could be generalised to biopsy samples from lymph node metastases, with AUC values of 0·91 (95% CI 0·88-0·94) for EGFR and 0·88 (0·82-0·93) for KRAS and accuracy values of 0·85 (0·80-0·88) for EGFR and 0·95 (0·92-0·96) for KRAS and showed potential for prognostic prediction of targeted therapy. The model generated spatial gene mutation maps, indicating gene mutation spatial distribution.

INTERPRETATION

We developed an AI-based method that can provide an accurate, timely, and economical prediction of gene mutation and mutation spatial distribution. The method showed substantial potential as an assistive tool for guiding the clinical treatment of patients with lung cancer.

FUNDING

National Natural Science Foundation of China, the Science and Technology Planning Project of Guangzhou, and the National Key Research and Development Program of China.

TRANSLATION

For the Chinese translation of the abstract see Supplementary Materials section.

Interpretable Machine Learning Algorithms Identify Inetetamab-Mediated Metabolic Signatures and Biomarkers in Treating Breast Cancer

BACKGROUND

HER2-positive breast cancer (BC), a highly aggressive malignancy, has been treated with the targeted therapy inetetamab for metastatic cases. Inetetamab (Cipterbin) is a recently approved targeted therapy for HER2-positive metastatic BC, significantly prolonging patients' survival. Currently, there is no established biomarker to reliably predict or assess the therapeutic efficacy of inetetamab in BC patients.

METHODS

This study harnesses the power of metabolomics and machine learning to uncover biomarkers for inetetamab in BC therapy. A total of 23 plasma samples from inetetamab-treated BC patients were collected and stratified into responders and nonresponders. Ultra-high-performance liquid chromatography-quadrupole time-of-flight mass spectrometry was utilized to analyze the metabolites in blood samples. A combination of univariate and multivariate statistical analyses was employed to identify these metabolites, and their biological functions were then ascertained by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. Finally, machine learning algorithms were employed to screen responsive biomarkers from all differentially expressed metabolites.

RESULTS

Our finding revealed 6889 unique metabolites that were detected. Pathways like retinol metabolism, fatty acid biosynthesis, and steroid hormone biosynthesis were enriched for differentially expressed metabolites. Notably, two key metabolites associated with inetetamab response in BC were identified: FAPy-adenine and 2-Pyrocatechuic acid. There was some negative correlation between progress-free survival (PFS) and their kurtosis content.

CONCLUSIONS

In summary, the identification of these two significant differential metabolites holds promise as potential biomarkers for evaluating and predicting inetetamab treatment outcomes in BC, ultimately contributing to the diagnosis of the disease and the discovery of prognostic markers.

Artificial intelligence-aided data mining of medical records for cancer detection and screening

The application of artificial intelligence methods to electronic patient records paves the way for large-scale analysis of multimodal data. Such population-wide data describing deep phenotypes composed of thousands of features are now being leveraged to create data-driven algorithms, which in turn has led to improved methods for early cancer detection and screening. Remaining challenges include establishment of infrastructures for prospective testing of such methods, ways to assess biases given the data, and gathering of sufficiently large and diverse datasets that reflect disease heterogeneities across populations. This Review provides an overview of artificial intelligence methods designed to detect cancer early, including key aspects of concern (eg, the problem of data drift-when the underlying health-care data change over time), ethical aspects, and discrepancies between access to cancer screening in high-income countries versus low-income and middle-income countries.

Harnessing the power of artificial intelligence for human living organoid research

As a powerful paradigm, artificial intelligence (AI) is rapidly impacting every aspect of our day-to-day life and scientific research through interdisciplinary transformations. Living human organoids (LOs) have a great potential for in vitro reshaping many aspects of in vivo true human organs, including organ development, disease occurrence, and drug responses. To date, AI has driven the revolutionary advances of human organoids in life science, precision medicine and pharmaceutical science in an unprecedented way. Herein, we provide a forward-looking review, the frontiers of LOs, covering the engineered construction strategies and multidisciplinary technologies for developing LOs, highlighting the cutting-edge achievements and the prospective applications of AI in LOs, particularly in biological study, disease occurrence, disease diagnosis and prediction and drug screening in preclinical assay. Moreover, we shed light on the new research trends harnessing the power of AI for LO research in the context of multidisciplinary technologies. The aim of this paper is to motivate researchers to explore organ function throughout the human life cycle, narrow the gap between in vitro microphysiological models and the real human body, accurately predict human-related responses to external stimuli (cues and drugs), accelerate the preclinical-to-clinical transformation, and ultimately enhance the health and well-being of patients.

NACNet: A histology context-aware transformer graph convolution network for predicting treatment response to neoadjuvant chemotherapy in Triple Negative Breast Cancer

Neoadjuvant chemotherapy (NAC) response prediction for triple negative breast cancer (TNBC) patients is a challenging task clinically as it requires understanding complex histology interactions within the tumor microenvironment (TME). Digital whole slide images (WSIs) capture detailed tissue information, but their giga-pixel size necessitates computational methods based on multiple instance learning, which typically analyze small, isolated image tiles without the spatial context of the TME. To address this limitation and incorporate TME spatial histology interactions in predicting NAC response for TNBC patients, we developed a histology context-aware transformer graph convolution network (NACNet). Our deep learning method identifies the histopathological labels on individual image tiles from WSIs, constructs a spatial TME graph, and represents each node with features derived from tissue texture and social network analysis. It predicts NAC response using a transformer graph convolution network model enhanced with graph isomorphism network layers. We evaluate our method with WSIs of a cohort of TNBC patient (N=105) and compared its performance with multiple state-of-the-art machine learning and deep learning models, including both graph and non-graph approaches. Our NACNet achieves 90.0% accuracy, 96.0% sensitivity, 88.0% specificity, and an AUC of 0.82, through eight-fold cross-validation, outperforming baseline models. These comprehensive experimental results suggest that NACNet holds strong potential for stratifying TNBC patients by NAC response, thereby helping to prevent overtreatment, improve patient quality of life, reduce treatment cost, and enhance clinical outcomes, marking an important advancement toward personalized breast cancer treatment.

A review on trends in development and translation of omics signatures in cancer

The field of cancer genomics and transcriptomics has evolved from targeted profiling to swift sequencing of individual tumor genome and transcriptome. The steady growth in genome, epigenome, and transcriptome datasets on a genome-wide scale has significantly increased our capability in capturing signatures that represent both the intrinsic and extrinsic biological features of tumors. These biological differences can help in precise molecular subtyping of cancer, predicting tumor progression, metastatic potential, and resistance to therapeutic agents. In this review, we summarized the current development of genomic, methylomic, transcriptomic, proteomic and metabolic signatures in the field of cancer research and highlighted their potentials in clinical applications to improve diagnosis, prognosis, and treatment decision in cancer patients.

Advances in machine learning for tumour classification in cancer of unknown primary: A mini-review

Cancers of unknown primary (CUP) are a heterogeneous group of aggressive metastatic cancers where standardised diagnostic techniques fail to identify the organ where it originated, resulting in a poor prognosis and resistance to treatment. Recent advances in large-scale sequencing techniques have enabled the identification of mutational signatures specific to particular tumour subtypes, even from liquid biopsy samples such as blood. This breakthrough paves the way for the development of new cost-effective diagnostic strategies. This mini-review explores recent advancements in Machine Learning (ML) and its application to tumour classification methods for CUP patients, identifying its weaknesses and strengths when classifying the tumour type. In the era of multi-omics, integrating several sources of information (e.g., imaging, molecular biomarkers, and family history) requires important theoretical advancements: increasing the dimensionality of the problem can result in lowering the predictive accuracy and robustness when data is scarce. Here, we review and discuss different architectures and strategies for incorporating cutting-edge machine learning into CUP diagnosis, aiming to bridge the gap between theory and clinical practice.

Integration of Machine Learning and Experimental Validation to Identify Anoikis-Related Prognostic Signature for Predicting the Breast Cancer Tumor Microenvironment and Treatment Response

Background/Objectives: Anoikis-related genes (ANRGs) are crucial in the invasion and metastasis of breast cancer (BC). The underlying role of ANRGs in the prognosis of breast cancer patients warrants further study. Methods: The anoikis-related prognostic signature (ANRS) was generated using a variety of machine learning methods, and the correlation between the ANRS and the tumor microenvironment (TME), drug sensitivity, and immunotherapy was investigated. Moreover, single-cell analysis and spatial transcriptome studies were conducted to investigate the expression of prognostic ANRGs across various cell types. Finally, the expression of ANRGs was verified by RT-PCR and Western blot analysis (WB), and the expression level of PLK1 in the blood was measured by the enzyme-linked immunosorbent assay (ELISA). Results: The ANRS, consisting of five ANRGs, was established. BC patients within the high-ANRS group exhibited poorer prognoses, characterized by elevated levels of immune suppression and stromal scores. The low-ANRS group had a better response to chemotherapy and immunotherapy. Single-cell analysis and spatial transcriptomics revealed variations in ANRGs across cells. The results of RT-PCR and WB were consistent with the differential expression analyses from databases. NU.1025 and imatinib were identified as potential inhibitors for SPIB and PLK1, respectively. Additionally, findings from ELISA demonstrated increased expression levels of PLK1 in the blood of BC patients. Conclusions: The ANRS can act as an independent prognostic indicator for BC patients, providing significant guidance for the implementation of chemotherapy and immunotherapy in these patients. Additionally, PLK1 has emerged as a potential blood-based diagnostic marker for breast cancer patients.

Archetypal Analysis of Kidney Allograft Biopsies Using Next-generation Sequencing Technology

BACKGROUND

In kidney transplantation, molecular diagnostics may be a valuable approach to improve the precision of the diagnosis. Using next-generation sequencing (NGS), we aimed to identify clinically relevant archetypes.

METHODS

We conducted an Illumina bulk RNA sequencing on 770 kidney biopsies (540 kidney recipients) collected between 2006 and 2021 from 11 European centers. Differentially expressed genes were determined for 11 Banff lesions. An ElasticNet model was used for feature selection, and 4 machine learning classifiers were trained to predict the probability of presence of the lesions. NGS-based classifiers were used in an unsupervised archetypal analysis to different archetypes. The association of the archetypes with allograft survival was assessed using the iBox risk prediction score.

RESULTS

The ElasticNet feature selection reduced the number of the genes from a range of 859-10 830 to a range of 52-867 genes. NGS-based classifiers demonstrated robust performances (precision-recall area under the curves 0.708-0.980) in predicting the Banff lesions. Archetypal analysis revealed 8 distinct phenotypes, each characterized by distinct clinical, immunological, and histological features. Although the archetypes confirmed the well-defined Banff rejection phenotypes for T cell-mediated rejection and antibody-mediated rejection, equivocal histologic antibody-mediated rejection, and borderline diagnoses were reclassified into different archetypes based on their molecular signatures. The 8 NGS-based archetypes displayed distinct allograft survival profiles with incremental graft loss rates between archetypes, ranging from 90% to 56% rates 7 y after evaluation (P < 0.0001).

CONCLUSIONS

Using molecular phenotyping, 8 archetypes were identified. These NGS-based archetypes might improve disease characterization, reclassify ambiguous Banff diagnoses, and enable patient-specific risk stratification.

Unraveling the role of low-density lipoprotein-related genes in lung adenocarcinoma: Insights into tumor microenvironment and clinical prognosis

BACKGROUND

The hypothesized link between low-density lipoprotein (LDL) and oncogenesis has garnered significant interest, yet its explicit impact on lung adenocarcinoma (LUAD) remains to be elucidated. This investigation aims to demystify the function of LDL-related genes (LRGs) within LUAD, endeavoring to shed light on the complex interplay between LDL and carcinogenesis.

METHODS

Leveraging single-cell transcriptomics, we examined the role of LRGs within the tumor microenvironment (TME). The expression patterns of LRGs across diverse cellular phenotypes were delineated using an array of computational methodologies, including AUCell, UCell, singscore, ssGSEA, and AddModuleScore. CellChat facilitated the exploration of distinct cellular interactions within LDL_low and LDL_high groups. The findmarker utility, coupled with Pearson correlation analysis, facilitated the identification of pivotal genes correlated with LDL indices. An integrative approach to transcriptomic data analysis was adopted, utilizing a machine learning framework to devise an LDL-associated signature (LAS). This enabled the delineation of genomic disparities, pathway enrichments, immune cell dynamics, and pharmacological sensitivities between LAS stratifications.

RESULTS

Enhanced cellular crosstalk was observed in the LDL_high group, with the CoxBoost+Ridge algorithm achieving the apex c-index for LAS formulation. Benchmarking against 144 extant LUAD models underscored the superior prognostic acuity of LAS. Elevated LAS indices were synonymous with adverse outcomes, diminished immune surveillance, and an upsurge in pathways conducive to neoplastic proliferation. Notably, a pronounced susceptibility to paclitaxel and gemcitabine was discerned within the high-LAS cohort, delineating prospective therapeutic corridors.

CONCLUSION

This study elucidates the significance of LRGs within the TME and introduces an LAS for prognostication in LUAD patients. Our findings accentuate putative therapeutic targets and elucidate the clinical ramifications of LAS deployment.

Progress and opportunities of foundation models in bioinformatics

Bioinformatics has undergone a paradigm shift in artificial intelligence (AI), particularly through foundation models (FMs), which address longstanding challenges in bioinformatics such as limited annotated data and data noise. These AI techniques have demonstrated remarkable efficacy across various downstream validation tasks, effectively representing diverse biological entities and heralding a new era in computational biology. The primary goal of this survey is to conduct a general investigation and summary of FMs in bioinformatics, tracing their evolutionary trajectory, current research landscape, and methodological frameworks. Our primary focus is on elucidating the application of FMs to specific biological problems, offering insights to guide the research community in choosing appropriate FMs for tasks like sequence analysis, structure prediction, and function annotation. Each section delves into the intricacies of the targeted challenges, contrasting the architectures and advancements of FMs with conventional methods and showcasing their utility across different biological domains. Further, this review scrutinizes the hurdles and constraints encountered by FMs in biology, including issues of data noise, model interpretability, and potential biases. This analysis provides a theoretical groundwork for understanding the circumstances under which certain FMs may exhibit suboptimal performance. Lastly, we outline prospective pathways and methodologies for the future development of FMs in biological research, facilitating ongoing innovation in the field. This comprehensive examination not only serves as an academic reference but also as a roadmap for forthcoming explorations and applications of FMs in biology.

Site-specific therapy guided by a 90-gene expression assay versus empirical chemotherapy in patients with cancer of unknown primary (Fudan CUP-001): a randomised controlled trial

BACKGROUND

Empirical chemotherapy remains the standard of care in patients with unfavourable cancer of unknown primary (CUP). Gene-expression profiling assays have been developed to identify the tissue of origin in patients with CUP; however, their clinical benefit has not yet been demonstrated. We aimed to evaluate the efficacy and safety of site-specific therapy directed by a 90-gene expression assay compared with empirical chemotherapy in patients with CUP.

METHODS

This randomised controlled trial was conducted at Fudan University Shanghai Cancer Center (Shanghai, China). We enrolled patients aged 18-75 years, with previously untreated CUP (histologically confirmed metastatic adenocarcinoma, squamous cell carcinoma, poorly differentiated carcinoma, or poorly differentiated neoplasms) and an Eastern Cooperative Oncology Group (ECOG) performance status of 0-2, who were not amenable to local radical treatment. Patients were randomly assigned (1:1) by the Pocock and Simon minimisation method to receive either site-specific therapy or empirical chemotherapy (taxane [175 mg/m2 by intravenous infusion on day 1] plus platinum [cisplatin 75 mg/m2 or carboplatin area under the curve 5 by intravenous infusion on day 1], or gemcitabine [1000 mg/m2 by intravenous infusion on days 1 and 8] plus platinum [same as above]). The minimisation factors were ECOG performance status and the extent of the disease. Clinicians and patients were not masked to interventions. The tumour origin in the site-specific therapy group was predicted by the 90-gene expression assay and treatments were administered accordingly. The primary endpoint was progression-free survival in the intention-to-treat population. The trial has been completed and the analysis is final. This study is registered with ClinicalTrials.gov (NCT03278600).

FINDINGS

Between Sept 18, 2017, and March 18, 2021, 182 patients (105 [58%] male, 77 [42%] female) were randomly assigned to receive site-specific therapy (n=91) or empirical chemotherapy (n=91). The five most commonly predicted tissues of origin in the site-specific therapy group were gastro-oesophagus (14 [15%]), lung (12 [13%]), ovary (11 [12%]), cervix (11 [12%]), and breast (nine [10%]). At the data cutoff date (April 30, 2023), median follow-up was 33·3 months (IQR 30·4-51·0) for the site-specific therapy group and 30·9 months (27·6-35·5) for the empirical chemotherapy group. Median progression-free survival was significantly longer with site-specific therapy than with empirical chemotherapy (9·6 months [95% CI 8·4-11·9] vs 6·6 months [5·5-7·9]; unadjusted hazard ratio 0·68 [95% CI 0·49-0·93]; p=0·017). Among the 167 patients who started planned treatment, 46 (56%) of 82 patients in the site-specific therapy group and 52 (61%) of 85 patients in the empirical chemotherapy group had grade 3 or worse treatment-related adverse events; the most frequent of these in the site-specific therapy and empirical chemotherapy groups were decreased neutrophil count (36 [44%] vs 42 [49%]), decreased white blood cell count (17 [21%] vs 26 [31%]), and anaemia (ten [12%] vs nine [11%]). Treatment-related serious adverse events were reported in five (6%) patients in the site-specific therapy group and two (2%) in the empirical chemotherapy group. No treatment-related deaths were observed.

INTERPRETATION

This single-centre randomised trial showed that site-specific therapy guided by the 90-gene expression assay could improve progression-free survival compared with empirical chemotherapy among patients with previously untreated CUP. Site-specific prediction by the 90-gene expression assay might provide more disease information and expand the therapeutic armamentarium in these patients.

FUNDING

Clinical Research Plan of Shanghai Hospital Development Center, Program for Shanghai Outstanding Academic Leader, and Shanghai Anticancer Association SOAR PROJECT.

TRANSLATION

For the Chinese translation of the abstract see Supplementary Materials section.

Metastatic renal cell carcinoma with occult primary: a multicenter prospective cohort

Metastatic carcinoma of presumed renal origin (rCUP) has recently emerged as a new entity within the heterogeneous entity of Cancers of Unknown Primary (CUP) but their biological features and optimal therapeutic management remain unknown. We report the molecular characteristics and clinical outcome of a series of 25 rCUP prospectively identified within the French National Multidisciplinary Tumor Board for CUP. This cohort strongly suggests that rCUP share similarities with common RCC subtypes and benefit from renal-tailored systemic treatment. This study highlights the importance of integrating clinical and molecular data for optimal diagnosis and management of CUP.

Multi-omics based artificial intelligence for cancer research

With significant advancements of next generation sequencing technologies, large amounts of multi-omics data, including genomics, epigenomics, transcriptomics, proteomics, and metabolomics, have been accumulated, offering an unprecedented opportunity to explore the heterogeneity and complexity of cancer across various molecular levels and scales. One of the promising aspects of multi-omics lies in its capacity to offer a holistic view of the biological networks and pathways underpinning cancer, facilitating a deeper understanding of its development, progression, and response to treatment. However, the exponential growth of data generated by multi-omics studies present significant analytical challenges. Processing, analyzing, integrating, and interpreting these multi-omics datasets to extract meaningful insights is an ambitious task that stands at the forefront of current cancer research. The application of artificial intelligence (AI) has emerged as a powerful solution to these challenges, demonstrating exceptional capabilities in deciphering complex patterns and extracting valuable information from large-scale, intricate omics datasets. This review delves into the synergy of AI and multi-omics, highlighting its revolutionary impact on oncology. We dissect how this confluence is reshaping the landscape of cancer research and clinical practice, particularly in the realms of early detection, diagnosis, prognosis, treatment and pathology. Additionally, we elaborate the latest AI methods for multi-omics integration to provide a comprehensive insight of the complex biological mechanisms and inherent heterogeneity of cancer. Finally, we discuss the current challenges of data harmonization, algorithm interpretability, and ethical considerations. Addressing these challenges necessitates a multidisciplinary collaboration, paving the promising way for more precise, personalized, and effective treatments for cancer patients.

Application of Transcriptome-Based Gene Set Featurization for Machine Learning Model to Predict the Origin of Metastatic Cancer

Identifying the primary site of origin of metastatic cancer is vital for guiding treatment decisions, especially for patients with cancer of unknown primary (CUP). Despite advanced diagnostic techniques, CUP remains difficult to pinpoint and is responsible for a considerable number of cancer-related fatalities. Understanding its origin is crucial for effective management and potentially improving patient outcomes. This study introduces a machine learning framework, ONCOfind-AI, that leverages transcriptome-based gene set features to enhance the accuracy of predicting the origin of metastatic cancers. We demonstrate its potential to facilitate the integration of RNA sequencing and microarray data by using gene set scores for characterization of transcriptome profiles generated from different platforms. Integrating data from different platforms resulted in improved accuracy of machine learning models for predicting cancer origins. We validated our method using external data from clinical samples collected through the Kangbuk Samsung Medical Center and Gene Expression Omnibus. The external validation results demonstrate a top-1 accuracy ranging from 0.80 to 0.86, with a top-2 accuracy of 0.90. This study highlights that incorporating biological knowledge through curated gene sets can help to merge gene expression data from different platforms, thereby enhancing the compatibility needed to develop more effective machine learning prediction models.

Deep-Learning Model for Tumor-Type Prediction Using Targeted Clinical Genomic Sequencing Data

Tumor type guides clinical treatment decisions in cancer, but histology-based diagnosis remains challenging. Genomic alterations are highly diagnostic of tumor type, and tumor-type classifiers trained on genomic features have been explored, but the most accurate methods are not clinically feasible, relying on features derived from whole-genome sequencing (WGS), or predicting across limited cancer types. We use genomic features from a data set of 39,787 solid tumors sequenced using a clinically targeted cancer gene panel to develop Genome-Derived-Diagnosis Ensemble (GDD-ENS): a hyperparameter ensemble for classifying tumor type using deep neural networks. GDD-ENS achieves 93% accuracy for high-confidence predictions across 38 cancer types, rivaling the performance of WGS-based methods. GDD-ENS can also guide diagnoses of rare type and cancers of unknown primary and incorporate patient-specific clinical information for improved predictions. Overall, integrating GDD-ENS into prospective clinical sequencing workflows could provide clinically relevant tumor-type predictions to guide treatment decisions in real time.

SIGNIFICANCE

We describe a highly accurate tumor-type prediction model, designed specifically for clinical implementation. Our model relies only on widely used cancer gene panel sequencing data, predicts across 38 distinct cancer types, and supports integration of patient-specific nongenomic information for enhanced decision support in challenging diagnostic situations. See related commentary by Garg, p. 906. This article is featured in Selected Articles from This Issue, p. 897.

Artificial Intelligence in Oncology: Current Landscape, Challenges, and Future Directions

Artificial intelligence (AI) in oncology is advancing beyond algorithm development to integration into clinical practice. This review describes the current state of the field, with a specific focus on clinical integration. AI applications are structured according to cancer type and clinical domain, focusing on the four most common cancers and tasks of detection, diagnosis, and treatment. These applications encompass various data modalities, including imaging, genomics, and medical records. We conclude with a summary of existing challenges, evolving solutions, and potential future directions for the field.

SIGNIFICANCE

AI is increasingly being applied to all aspects of oncology, where several applications are maturing beyond research and development to direct clinical integration. This review summarizes the current state of the field through the lens of clinical translation along the clinical care continuum. Emerging areas are also highlighted, along with common challenges, evolving solutions, and potential future directions for the field.

A cfDNA methylation-based tissue-of-origin classifier for cancers of unknown primary

Cancers of Unknown Primary (CUP) remains a diagnostic and therapeutic challenge due to biological heterogeneity and poor responses to standard chemotherapy. Predicting tissue-of-origin (TOO) molecularly could help refine this diagnosis, with tissue acquisition barriers mitigated via liquid biopsies. However, TOO liquid biopsies are unexplored in CUP cohorts. Here we describe CUPiD, a machine learning classifier for accurate TOO predictions across 29 tumour classes using circulating cell-free DNA (cfDNA) methylation patterns. We tested CUPiD on 143 cfDNA samples from patients with 13 cancer types alongside 27 non-cancer controls, with overall sensitivity of 84.6% and TOO accuracy of 96.8%. In an additional cohort of 41 patients with CUP CUPiD predictions were made in 32/41 (78.0%) cases, with 88.5% of the predictions clinically consistent with a subsequent or suspected primary tumour diagnosis, when available (23/26 patients). Combining CUPiD with cfDNA mutation data demonstrated potential diagnosis re-classification and/or treatment change in this hard-to-treat cancer group.

Automatic origin prediction of liver metastases via hierarchical artificial-intelligence system trained on multiphasic CT data: a retrospective, multicentre study

Background

Currently, the diagnostic testing for the primary origin of liver metastases (LMs) can be laborious, complicating clinical decision-making. Directly classifying the primary origin of LMs at computed tomography (CT) images has proven to be challenging, despite its potential to streamline the entire diagnostic workflow.

Methods

We developed ALMSS, an artificial intelligence (AI)-based LMs screening system, to provide automated liver contrast-enhanced CT analysis for distinguishing LMs from hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC), as well as subtyping primary origin of LMs as six organ systems. We processed a CECT dataset between January 1, 2013 and June 30, 2022 (n = 3105: 840 HCC, 354 ICC, and 1911 LMs) for training and internally testing ALMSS, and two additional cohorts (n = 622) for external validation of its diagnostic performance. The performance of radiologists with and without the assistance of ALMSS in diagnosing and subtyping LMs was assessed.

Findings

ALMSS achieved average area under the curve (AUC) of 0.917 (95% confidence interval [CI]: 0.899-0.931) and 0.923 (95% [CI]: 0.905-0.937) for differentiating LMs, HCC and ICC on both the internal testing set and external testing set, outperformed that of two radiologists. Moreover, ALMSS yielded average AUC of 0.815 (95% [CI]: 0.794-0.836) and 0.818 (95% [CI]: 0.790-0.842) for predicting six primary origins on both two testing sets. Interestingly, ALMSS assigned origin diagnoses for LMs with pathological phenotypes beyond the training categories with average AUC of 0.761 (95% [CI]: 0.657-0.842), which verify the model's diagnostic expandability.

Interpretation

Our study established an AI-based diagnostic system that effectively identifies and characterizes LMs directly from multiphasic CT images.

Funding

National Natural Science Foundation of China, Guangdong Provincial Key Laboratory of Medical Image Processing.

The Rise of Hypothesis-Driven Artificial Intelligence in Oncology

Cancer is a complex disease involving the deregulation of intricate cellular systems beyond genetic aberrations and, as such, requires sophisticated computational approaches and high-dimensional data for optimal interpretation. While conventional artificial intelligence (AI) models excel in many prediction tasks, they often lack interpretability and are blind to the scientific hypotheses generated by researchers to enable cancer discoveries. Here we propose that hypothesis-driven AI, a new emerging class of AI algorithm, is an innovative approach to uncovering the complex etiology of cancer from big omics data. This review exemplifies how hypothesis-driven AI is different from conventional AI by citing its application in various areas of oncology including tumor classification, patient stratification, cancer gene discovery, drug response prediction, and tumor spatial organization. Our aim is to stress the feasibility of incorporating domain knowledge and scientific hypotheses to craft the design of new AI algorithms. We showcase the power of hypothesis-driven AI in making novel cancer discoveries that can be overlooked by conventional AI methods. Since hypothesis-driven AI is still in its infancy, open questions such as how to better incorporate new knowledge and biological perspectives to ameliorate bias and improve interpretability in the design of AI algorithms still need to be addressed. In conclusion, hypothesis-driven AI holds great promise in the discovery of new mechanistic and functional insights that explain the complexity of cancer etiology and potentially chart a new roadmap to improve treatment regimens for individual patients.

Advances in Cancer Research: Current and Future Diagnostic and Therapeutic Strategies

Cancers of unknown primary (CUP) exhibit significant cellular heterogeneity and malignancy, which poses significant challenges for diagnosis and treatment. Recent years have seen deeper insights into the imaging, pathology, and genetic characteristics of CUP, driven by interdisciplinary collaboration and the evolution of diagnostic and therapeutic strategies. However, due to their insidious onset, lack of evidence-based medicine, and limited clinical understanding, diagnosing and treating CUP remain a significant challenge. To inspire more creative and fantastic research, herein, we report and highlight recent advances in the diagnosis and therapeutic strategies of CUP. Specifically, we discuss advanced diagnostic technologies, including 12-deoxy-2-[fluorine-18]fluoro-D-glucose integrated with computed tomography (18F-FDG PET/CT) or 68Ga-FAPI (fibroblast activation protein inhibitor) PET/CT, liquid biopsy, molecular diagnostics, self-assembling nanotechnology, and artificial intelligence (AI). In particular, the discussion will extend to the effective treatment techniques currently available, such as targeted therapies, immunotherapies, and bio-nanotechnology-based therapeutics. Finally, a novel perspective on the challenges and directions for future CUP diagnostic and therapeutic strategies is discussed.

Measuring Progress in Precision Oncology

SUMMARY

In this issue of Cancer Discovery, Suehnholz and colleagues describe their efforts to quantify the gradual yet steady progress of precision oncology by surveying the regulatory approvals of targeted cancer therapies, and thus the actionability of corresponding molecular alterations in clinical practice, over more than 20 years. Their work also suggests a relationship between the discovery of candidate therapeutic targets through comprehensive tumor profiling and molecularly guided cancer drug development. See related article by Suehnholz et al., p. 49 (5).

Machine learning in onco-pharmacogenomics: a path to precision medicine with many challenges

Over the past two decades, Next-Generation Sequencing (NGS) has revolutionized the approach to cancer research. Applications of NGS include the identification of tumor specific alterations that can influence tumor pathobiology and also impact diagnosis, prognosis and therapeutic options. Pharmacogenomics (PGx) studies the role of inheritance of individual genetic patterns in drug response and has taken advantage of NGS technology as it provides access to high-throughput data that can, however, be difficult to manage. Machine learning (ML) has recently been used in the life sciences to discover hidden patterns from complex NGS data and to solve various PGx problems. In this review, we provide a comprehensive overview of the NGS approaches that can be employed and the different PGx studies implicating the use of NGS data. We also provide an excursus of the ML algorithms that can exert a role as fundamental strategies in the PGx field to improve personalized medicine in cancer.

The leading global health challenges in the artificial intelligence era

Millions of people's health is at risk because of several factors and multiple overlapping crises, all of which hit the vulnerable the most. These challenges are dynamic and evolve in response to emerging health challenges and concerns, which need effective collaboration among countries working toward achieving Sustainable Development Goals (SDGs) and securing global health. Mental Health, the Impact of climate change, cardiovascular diseases (CVDs), diabetes, Infectious diseases, health system, and population aging are examples of challenges known to pose a vast burden worldwide. We are at a point known as the "digital revolution," characterized by the expansion of artificial intelligence (AI) and a fusion of technology types. AI has emerged as a powerful tool for addressing various health challenges, and the last ten years have been influential due to the rapid expansion in the production and accessibility of health-related data. The computational models and algorithms can understand complicated health and medical data to perform various functions and deep-learning strategies. This narrative mini-review summarizes the most current AI applications to address the leading global health challenges. Harnessing its capabilities can ultimately mitigate the Impact of these challenges and revolutionize the field. It has the ability to strengthen global health through personalized health care and improved preparedness and response to future challenges. However, ethical and legal concerns about individual or community privacy and autonomy must be addressed for effective implementation.

Integrating somatic mutation profiles with structural deep clustering network for metabolic stratification in pancreatic cancer: a comprehensive analysis of prognostic and genomic landscapes

Pancreatic cancer is a globally recognized highly aggressive malignancy, posing a significant threat to human health and characterized by pronounced heterogeneity. In recent years, researchers have uncovered that the development and progression of cancer are often attributed to the accumulation of somatic mutations within cells. However, cancer somatic mutation data exhibit characteristics such as high dimensionality and sparsity, which pose new challenges in utilizing these data effectively. In this study, we propagated the discrete somatic mutation data of pancreatic cancer through a network propagation model based on protein-protein interaction networks. This resulted in smoothed somatic mutation profile data that incorporate protein network information. Based on this smoothed mutation profile data, we obtained the activity levels of different metabolic pathways in pancreatic cancer patients. Subsequently, using the activity levels of various metabolic pathways in cancer patients, we employed a deep clustering algorithm to establish biologically and clinically relevant metabolic subtypes of pancreatic cancer. Our study holds scientific significance in classifying pancreatic cancer based on somatic mutation data and may provide a crucial theoretical basis for the diagnosis and immunotherapy of pancreatic cancer patients.