Prediction Of Metastatic Risk Research Articles

Genetic Features of Uveal Melanoma.

Although it comprises only 5% of all melanomas, uveal melanoma (UM) is the most commonly observed primary intraocular cancer. Poor patient survival persists in spite of innovative systemic therapies. In fact, approximately fifty percent of UM patients develop metastases from micro-metastases that remain undetected at the exact time of diagnosis. The molecular understanding of UM is significantly enhanced by the recent identification of several mutations that are responsible for the metastasis, growth, and survival of UM. The crucial point is a more accurate genetic analysis for patient follow-up and metastatic risk prediction. This review provides a brief summary of the molecular features of UM that are recently discovered, as well as cytogenetic markers and biochemical pathways that are associated with the development of UM metastases.

Accuracy Goals in Predicting Preoperative Lymph Node Metastasis for T1 Colorectal Cancer Resected Endoscopically.

Submucosal invasive (T1) colorectal cancer is a significant clinical management challenge, with an estimated 10% of patients developing extraintestinal lymph node metastasis. This condition necessitates surgical resection along with lymph node dissection to achieve a curative outcome. Thus, the precise preoperative assessment of lymph node metastasis risk is crucial to guide treatment decisions after endoscopic resection. Contemporary clinical guidelines strive to identify a low-risk cohort for whom endoscopic resection will suffice, applying stringent criteria to maximize patient safety. Those failing to meet these criteria are often recommended for surgical resection, with its associated mortality risks although it may still include patients with a low risk of metastasis. In the quest to enhance the precision of preoperative lymph node metastasis risk prediction, innovative models leveraging artificial intelligence or nomograms are being developed. Nevertheless, the debate over the ideal sensitivity and specificity for such models persists, with no consensus on target metrics. This review puts forth postoperative mortality rates as a practical benchmark for the sensitivity of predictive models. We underscore the importance of this method and advocate for research to amass data on surgical mortality in T1 colorectal cancer. Establishing specific benchmarks for predictive accuracy in lymph node metastasis risk assessment will hopefully optimize the treatment of T1 colorectal cancer.

CCL5 promotes the epithelial-mesenchymal transition of circulating tumor cells in renal cancer

BackgroundCirculating tumor cells (CTCs) are pivotal in tumor metastasis across cancers, yet their specific role in renal cancer remains unclear.MethodsThis study investigated C–C motif chemokine ligand 5 (CCL5)'s tumorigenic impact on renal cancer cells and CTCs using bioinformatics, in vivo, and in vitro experiments. It also assessed renal cancer patients' CTCs prognostic value through Lasso regression and Kaplan–Meier survival curves.ResultsBioinformatics analysis revealed differential genes focusing on cellular adhesion and migration between CTCs and tumor cells. CCL5 exhibited high expression in various CTCs, correlating with poor prognosis in renal cancer. In 786-O-CTCs, CCL5 enhanced malignancy, while in renal cell carcinoma cell line CAKI-2 and 786-O, it promoted epithelial-mesenchymal transition (EMT) via smad2/3, influencing cellular characteristics. The nude mouse model suggested CCL5 increased CTCs and intensified EMT, enhancing lung metastasis. Clinical results shown varying prognostic values for different EMT-typed CTCs, with mesenchymal CTCs having the highest value.ConclusionsIn summary, CCL5 promoted EMT in renal cancer cells and CTCs through smad2/3, enhancing the malignant phenotype and facilitating lung metastasis. Mesenchymal-type CTC-related factors can construct a risk model for renal cancer patients, allowing personalized treatment based on metastatic risk prediction.

Brain metastasis risk prediction model in females with hormone receptor-positive breast cancer

BackgroundBreast cancer is a leading cause of cancer-related deaths in females, and the hormone receptor-positive subtype is the most frequent. Breast cancer is a common source of brain metastases; therefore, we aimed to generate a brain metastases prediction model in females with hormone receptor-positive breast cancer. MethodsThe primary cohort included 3,682 females with hormone receptor-positive breast cancer treated at a single center from May 2009 to May 2020. Patients were randomly divided into a training dataset (n = 2,455) and a validation dataset (n = 1,227). In the training dataset, simple logistic regression analyses were used to measure associations between variables and the diagnosis of brain metastases and to build multivariable models. The model with better calibration and discrimination capacity was tested in the validation dataset to measure its predictive performance. ResultsThe variables incorporated in the model included age, tumor size, axillary lymph node status, clinical stage at diagnosis, HER2 expression, Ki-67 proliferation index, and the modified Scarff-Bloom-Richardson grade. The area under the curve was 0.81 (95 % CI 0.75–0.86), p < 0.001 in the validation dataset. The study presents a guide for the clinical use of the model. ConclusionA brain metastases prediction model in females with hormone receptor-positive breast cancer helps assess the individual risk of brain metastases.

Independent blinded validation of an AI-based digital histology classifier for prostate cancer recurrence and metastasis risk prediction.

Independent blinded validation of an AI-based digital histology classifier for prostate cancer recurrence and metastasis risk prediction.

The 15-gene expression profile test is independent from PRAME and 7-gene next-generation sequencing: Results from 3,267 clinically tested uveal melanomas.

9598 Background: The 15-gene expression profile (15-GEP) test is the gold standard in the US for metastatic risk prediction in uveal melanoma (UM). Studies have shown that the 15-GEP is superior to and independent of other clinicopathologic and molecular risk factors, including PRAME expression status and BAP1 mutation status. Supplemental tests for PRAME and the 7-gene NGS panel ( BAP1, SF3B1, EIF1AX, GNA11, GNAQ, CYSLTR2, PLCB4)are validated and can now be run clinically from the same biopsy specimen. However, the discordance between 15-GEP, PRAME and NGS in clinical specimens is unknown. This study aimed to determine whether PRAME status or DNA mutation status could be used to predict the 15-GEP result. Methods: Data from fine-needle aspiration biopsy or formalin-fixed paraffin-embedded UM specimens submitted for routine comprehensive clinical testing with all three commercial tests (15-GEP, PRAME, and NGS) between February 2018 and December 2023 were collected through an Institutional Review Board-approved protocol. Results: Overall, 3267 tumors with comprehensive molecular testing were analyzed. 15-GEP testing yielded 65.9% (n=2152/3267) Class 1 and 34.1% (n=1115) Class 2 tumors. While Class 2 tumors were more than twice as likely to be PRAME(+) vs Class 1, 54.3% (606/1115) of Class 2 cases were PRAME(–), and 20.9% (450/2152) of Class 1 cases were PRAME(+) resulting in a PRAME/15-GEP discordance rate of 32.3% (10-56/3267). Mutations in genes associated with increasing metastatic risk ( EIF1AX, SF3B1, and BAP1, respectively) were identified in 58.7% (1918/3267) of cases. Most EIF1AX mutations (569/623, 91.3%) occurred in Class 1 tumors that were PRAME(–), while most SF3B1mutations were found in Class 1, PRAME(+) tumors (244/404, 60.4%). Most BAP1 mutations (863/976, 88.4%) occurred in Class 2 (high risk) tumors, of which 53.9% (465/863) were PRAME(–). Importantly, 56.0% (1205/2152) of Class 1 tumors did not harbor detectable EIF1AX or SF3B1 mutations, and 22.6% (252/1115) of Class 2 tumors did not harbor a detectable BAP1 mutation, yielding an NGS/15-GEP discordance rate of 44.6% (1457/3267). Conclusions: The validated PRAME and NGS tests described here, which can be performed from the same initial biopsy sample taken for 15-GEP testing, provide results that are not systematically associated with 15-GEP class. Given that the 15-GEP has been shown to be superior to both PRAME and NGS in predicting metastatic outcomes, it is recommended that these supplemental tests only be considered in the context of a 15-GEP test result to prevent potentially over or under identifying high-risk biology. An ongoing prospective study with long term outcomes for over 1700 UM patients led by the Collaborative Ocular Oncology Group (COOG2) will provide insight into how best to effectively integrate these supplementary biomarkers with the 15-GEP class result.

Computational analysis of cancer cell adhesion in curved vessels affected by wall shear stress for prediction of metastatic spreading.

Introduction: The dynamics of circulating tumor cells (CTCs) within blood vessels play a pivotal role in predicting metastatic spreading of cancer within the body. However, the limited understanding and method to quantitatively investigate the influence of vascular architecture on CTC dynamics hinders our ability to predict metastatic process effectively. To address this limitation, the present study was conducted to investigate the influence of blood vessel tortuosity on the behaviour of CTCs, focusing specifically on establishing methods and examining the role of shear stress in CTC-vessel wall interactions and its subsequent impact on metastasis. Methods: We computationally simulated CTC behaviour under various shear stress conditions induced by vessel tortuosity. Our computational model, based on the lattice Boltzmann method (LBM) and a coarse-grained spectrin-link membrane model, efficiently simulates blood plasma dynamics and CTC deformability. The model incorporates fluid-structure interactions and receptor-ligand interactions crucial for CTC adhesion using the immersed boundary method (IBM). Results: Our findings reveal that uniform shear stress in straight vessels leads to predictable CTC-vessel interactions, whereas in curved vessels, asymmetrical flow patterns and altered shear stress create distinct adhesion dynamics, potentially influencing CTC extravasation. Quantitative analysis shows a 25% decrease in the wall shear stress in low-shear regions and a 58.5% increase in the high-shear region. We observed high-shear regions in curved vessels to be potential sites for increased CTC adhesion and extravasation, facilitated by elevated endothelial expression of adhesion molecules. This phenomenon correlates with the increased number of adhesion bonds, which rises to approximately 40 in high-shear regions, compared to around 12 for straight vessels and approximately 5-6 in low-shear regions. The findings also indicate an optimal cellular stiffness necessary for successful CTC extravasation in curved vessels. Discussion: By the quantitative assessment of the risk of CTC extravasation as a function of vessel tortuosity, our study offers a novel tool for the prediction of metastasis risk to support the development of personalized therapeutic interventions based on individual vascular characteristics and tumor cell properties.

The effect of surgery and metastatic risk prediction in patients with gastrointestinal tract signet-ring cell carcinoma

PurposeThis study aimed to assess the efficacy of surgery as a treatment option for patients with signet-ring cell carcinoma (SRCC) in the gastrointestinal tract (GI-SRCC). MethodsUsing the Surveillance, Epidemiology, and End Results database, patients with GI-SRCC who underwent surgery or received nonsurgical treatment were included. Propensity score matching (PSM) analysis was used to balance baseline characteristics and reduce bias. Overall survival (OS) was calculated in matching cohorts to estimate prognosis for patients with GI-SRCC. Nomogram was established to predict metastasis for patients with GI-SRCC. ResultsThe study enrolled a total of 9428 patients with GI-SRCC, with 1689 patients in the nonsurgery group and 7739 patients in the surgery group. After 1:1 PSM, we analyzed 743 patients from each group. Our survival analyses revealed that surgery independently correlated with improved OS for patients with GI-SRCC (hazard ratio, 0.37; 95% CI, 0.33–0.42; P < .001). Subgroup analysis further confirmed the positive impact of surgery on the prognosis of patients with nonmetastatic GI-SRCC. Notably, distinct subsets of patients with metastasis, particularly those originating from the upper GI (esophagus, proximal stomach, and distal stomach) and left colon, demonstrated a significant improvement in OS after surgery. However, no significant survival difference was observed for patients with metastatic right colon and rectum SRCC. Using nomogram, we quantitatively assessed the risk of metastasis in patients with right colon and rectum SRCC, which exhibited robust predictive accuracy, with area under the curve values of 0.829. ConclusionOur study highlighted surgery’s positive impact on prognosis for both patients with nonmetastatic and metastatic upper GI-SRCC and left colon SRCC. Hence, we recommend surgery as a treatment option for these groups. In addition, for patients with metastatic right colon and rectum SRCC ineligible for surgery, our predictive nomogram can offer a convenient tool to aid early intervention and improve prognosis.

Abstract 3791: A prognostic signature based on inherited genetic polymorphisms for patients with uveal melanoma

Abstract Introduction: Uveal melanoma (UM) is the most frequent eye malignancy in adults. Two UM subtypes exist, defined by Gene Expression Profiling or by genomic status: GEP class 1/disomy 3 [D3] and GEP class 2/monosomy 3 [M3] defining low-risk or high-risk of developing metastasis, respectively. However, these prognostic signatures require tumor samples, which may be difficult to obtain. We previously identified by a genome-wide association study (GWAS) frequent single nucleotide polymorphisms (SNP) in IRF4 and HERC2 associated with risk for D3 or M3 UM, respectively (1). Here, we aim to design a prognostic signature for the D3 and M3 UM subtypes based on these constitutive polymorphisms, available from blood samples. Experimental Procedures: We used the GWAS cohort of 1476 UM patients genotyped on GSA or Omni5. Risk SNPs coupled with clinical variables were used on a cohort of 497 UM patients (191 D3, 306 M3) in a statistical machine learning process to define a predictive signature of UM subtype. A cohort of 852 UM patients was used for validation. Results: We retained two SNPs, rs12203592 (in IRF4) and rs12913832 (in HERC2), tumor diameter and thickness, age at diagnosis and gender. We validated the predictive model in the validation cohort. The predicted high-risk group showed a decrease in time to relapse (TR) (HR: 3.72 [2.57 - 5.41], logrank p-value = 1.35E-15), overall survival (OS) (HR: 3.85 [2.58 - 5.73], logrank p-value = 1.24E-14) and progression-free-survival (PFS) (HR: 3.61 [2.32 - 5.61], logrank p-value = 6.55E-11), compared with the predicted low-risk group. Conclusion: Common germline polymorphisms contribute to the prediction of metastatic risk in patients with localized uveal melanoma. This information could be especially useful when access to the tumor is problematic. 1. Mobuchon et al, J Natl Cancer Inst, 114, 2, (2022) Citation Format: Thibault Verrier, Anaïs Le Ven, Alexandre Houy, Agathe Garcia, Pierron Gaelle, Nathalie Cassoux, Manuel Rodrigues, Josselin Noirel, Marc-Henri Stern. A prognostic signature based on inherited genetic polymorphisms for patients with uveal melanoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3791.

Risk factors, prognostic factors, and nomograms for distant metastases in patients with gastroenteropancreatic neuroendocrine tumors: a population-based study.

Patients with gastroenteropancreatic neuroendocrine tumors (GEP-NETs) have a poor prognosis for distant metastasis. Currently, there are no studies on predictive models for the risk of distant metastasis in GEP-NETs. In this study, risk factors associated with metastasis in patients with GEP-NETs in the Surveillance, Epidemiology, and End Results (SEER) database were analyzed by univariate and multivariate logistic regression, and a nomogram model for metastasis risk prediction was constructed. Prognostic factors associated with distant metastasis in patients with GEP-NETs were analyzed by univariate and multivariate Cox, and a nomogram model for prognostic prediction was constructed. Finally, the performance of the nomogram model predictions is validated by internal validation set and external validation set. A total of 9145 patients with GEP-NETs were enrolled in this study. Univariate and multivariate logistic analysis demonstrated that T stage, N stage, tumor size, primary site, and histologic types independent risk factors associated with distant metastasis in GEP-NETs patients (p value < 0.05). Univariate and multivariate Cox analyses demonstrated that age, histologic type, tumor size, N stage, and primary site surgery were independent factors associated with the prognosis of patients with GEP-NETs (p value < 0.05). The nomogram model constructed based on metastasis risk factors and prognostic factors can predict the occurrence of metastasis and patient prognosis of GEP-NETs very effectively in the internal training and validation sets as well as in the external validation set. In conclusion, we constructed a new distant metastasis risk nomogram model and a new prognostic nomogram model for GEP-NETs patients, which provides a decision-making reference for individualized treatment of clinical patients.

Artificial intelligence applications in computed tomography in gastric cancer: a narrative review.

Artificial intelligence (AI) is a revolutionary technique which is deeply impacting and reshaping clinical practice in oncology. This review aims to summarize the current status of the clinical application of AI-based computed tomography (CT) for gastric cancer (GC), focusing on diagnosis, genetic status detection and risk prediction of metastasis, prognosis and treatment efficacy. The challenges and prospects for future research will also be discussed. We searched the PubMed/MEDLINE database to identify clinical studies published between 1990 and November 2022 that investigated AI applications in CT in GC. The major findings of the verified studies were summarized. AI applications in CT images have attracted considerable attention in various fields such as diagnosis, prediction of metastasis risk, survival, and treatment response. These emerging techniques have shown a high potential to outperform clinicians in diagnostic accuracy and time-saving. AI-powered tools showed great potential to increase diagnostic accuracy and reduce radiologists' workload. However, the goal of AI is not to replace human ability but to help oncologists make decisions in their practice. Therefore, radiologists should play a predominant role in AI applications and decide the best ways to integrate these complementary techniques within clinical practice.

Radiomic signature accurately predicts the risk of metastatic dissemination in late-stage non-small cell lung cancer.

Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, and the median overall survival (OS) is approximately 2-3 years among patients with stage III disease. Furthermore, it is one of the deadliest types of cancer globally due to non-specific symptoms and the lack of a biomarker for early detection. The most important decision that clinicians need to make after a lung cancer diagnosis is the selection of a treatment schedule. This decision is based on, among others factors, the risk of developing metastasis. A cohort of 115 NSCLC patients treated using chemotherapy and radiotherapy (RT) with curative intent was retrospectively collated and included patients for whom positron emission tomography/computed tomography (PET/CT) images, acquired before RT, were available. The PET/CT images were used to compute radiomic features extracted from a region of interest (ROI), the primary tumor. Radiomic and clinical features were then classified to stratify the patients into short and long time to metastasis, and regression analysis was used to predict the risk of metastasis. Classification based on binarized metastasis-free survival (MFS) was applied with moderate success. Indeed, an accuracy of 0.73 was obtained for the selection of features based on the Wilcoxon test and logistic regression model. However, the Cox regression model for metastasis risk prediction performed very well, with a concordance index (C-index) score equal to 0.84. It is possible to accurately predict the risk of metastasis in NSCLC patients based on radiomic features. The results demonstrate the potential use of features extracted from cancer imaging in predicting the risk of metastasis.

Long-term outcomes in a population-based cohort of 2,967 uveal melanoma patients clinically tested with the 15-gene expression profile: A collaborative study with the National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) Program Registries.

e21574 Background: Metastatic risk prediction in uveal melanoma is commonly performed using a commercially available 15-gene expression profile (15-GEP) test. A recent collaborative effort with the NCI’s SEER program registries allowed for serial linkage of UM patients captured by SEER registries with clinical 15-GEP test results. The objective of this study was to confirm the 15-GEP test’s performance in clinically tested uveal melanoma patients with long-term outcomes at the population level. Methods: Following established operating procedures, SEER registries linked UM cases diagnosed from 2010–2019 to 15-GEP test results provided by Castle Biosciences using a trusted third-party (honest broker). The patients were clinically tested as recommended by guidelines (testing was performed either on fine needle aspiration biopsy tissue or formalin-fixed, paraffin-embedded tissue from enucleations). De-identified data were provided to NCI and Castle Biosciences scientists for this study. Kaplan-Meier analysis with log-rank test was used to evaluate the available endpoints of melanoma-specific survival (MSS) and overall survival (OS). Multivariable analysis was performed with Cox proportional hazards modeling. Results: A total of 2,967 UM patients ≥ 18 years of age with localized disease (excluding Stage IV) and 15-GEP test result were included in this study (Class 1, 1904; Class 2, 1063). The cohort was 51.7% male, with a median age of 63 years. Median follow-up time was 2.2 years. Patients with a low-risk (Class 1) vs. high-risk (Class 2) result had significantly higher 7-year survival outcomes (OS, 78.9 vs. 31.8%, p &lt; 0.001; MSS, 90.4 vs. 41.6%, p &lt; 0.001, respectively). Patients with a Class 2 result were over six times more likely to experience a UM-related death than those with a Class 1 result (31.8 [338/1063] vs. 5.0% [96/1904]; p &lt; 0.001). Multivariable analysis showed that the 15-GEP was the strongest independent predictor of MSS when compared with clinicopathologic features (tumor diameter or thickness) and age, with a Class 2 hazard ratio of 6.2 (95% CI, 3.9–9.7; p &lt; 0.001). Notably, there was no association between GEP class call distribution and socioeconomic status as measured by the Yost index (p = 0.79). Conclusions: In this analysis, the 15-GEP accurately stratified risk of death from UM in a clinically tested, population-based cohort of patients, confirming previous studies. Socioeconomic status did not impact the likelihood of getting a high-risk result, suggesting equitable access to care for patients who receive 15-GEP testing. Ongoing analyses of SEER-linked UM data will include patients who received a PRAME test result, allowing for assessment of this adjunctive biomarker in combination with the 15-GEP result.

MP41-13 QUANTITATIVE ASSESSMENT OF MESENCHYMAL DIFFERENTIATION FOR METASTATIC RISK PREDICTION OF CLEAR CELL RENAL CARCINOMA (CCRCC)

You have accessJournal of UrologyCME1 Apr 2023MP41-13 QUANTITATIVE ASSESSMENT OF MESENCHYMAL DIFFERENTIATION FOR METASTATIC RISK PREDICTION OF CLEAR CELL RENAL CARCINOMA (CCRCC) Daniel Vilchez, Bodong Zhang, Alejandro Sanchez, Deepika Sirohi, Tolga Tasdizen, and Beatrice Knudsen Daniel VilchezDaniel Vilchez More articles by this author , Bodong ZhangBodong Zhang More articles by this author , Alejandro SanchezAlejandro Sanchez More articles by this author , Deepika SirohiDeepika Sirohi More articles by this author , Tolga TasdizenTolga Tasdizen More articles by this author , and Beatrice KnudsenBeatrice Knudsen More articles by this author View All Author Informationhttps://doi.org/10.1097/JU.0000000000003279.13AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Epithelial-to-mesenchymal transition (EMT) results in poor prognosis sarcomatoid and rhabdoid histologic growth patterns (HGP) of ccRCC. However, only ∼15% of metastatic cases reveal areas of sarcomatoid/rhabdoid differentiation in the primary tumor. We questioned whether a computer can learn salient features of EMT that are associated with risk of metastatic progression across all HGPs of ccRCC. METHODS: The project uses 410 digital H&E slides of ccRCC. We incorporated a novel co-training approach into a machine learning model (MLM) and trained the model to distinguish epithelial (nested) and mesenchymal (sarcomatoid, rhabdoid) HGPs in a Utah cohort. The trained MLM is applied to the TCGA KIRC cohort (n=360) with annotated regions of nested, tubulo-papillary, low grade-spindle, sarcomatoid and rhabdoid HGPs for which the hazard ration of metastatic progression was published. For each case, the MLM outputs a mesenchymal-differentiation (DP-MD) score on a scale of 0 to 1 where closer to 1 is higher likelihood of EMT. RESULTS: A unique machine learning / AI model was developed to quantify sarcomatoid differentiation in ccRCC with no obvious histologic signs of mesenchymal morphology. H&E stains were used as separate inputs of a contrastive co-training model (H/E-CCTM) to force the learning of mesenchymal features. The inclusion of a semi-supervised learning component improved the accuracy of predicting mesenchymal differentiation by >4% over other state-of-the-art co-training algorithms and performed equally well when trained on 10% versus 100% of labeled cancer data. Sarcomatoid, low grade spindle and rhabdoid cases received average DP-MD scores of 0.91, 0.97 and 0.83, respectively in a held-out test set, while cases with nested morphology, including expansile, and high nuclear to cytoplasmic ratio averaged DP-MD scores of 0.08. Tubulo-papillary HGPs without obvious mesenchymal features obtained an average DP-MD score of 0.49. CONCLUSIONS: Computer-assisted morphometric analysis can be used to quantify mesenchymal differentiation and has the potential to improve the prediction of metastatic risk in ccRCC and the selection of patients for adjuvant treatment after nephrectomy. Source of Funding: Huntsman Cancer Foundation, NIH P30CA042014 © 2023 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 209Issue Supplement 4April 2023Page: e560 Advertisement Copyright & Permissions© 2023 by American Urological Association Education and Research, Inc.MetricsAuthor Information Daniel Vilchez More articles by this author Bodong Zhang More articles by this author Alejandro Sanchez More articles by this author Deepika Sirohi More articles by this author Tolga Tasdizen More articles by this author Beatrice Knudsen More articles by this author Expand All Advertisement PDF downloadLoading ...

Clinicomics-guided distant metastasis prediction in breast cancer via artificial intelligence

BackgroundBreast cancer has become the most common malignant tumour worldwide. Distant metastasis is one of the leading causes of breast cancer-related death. To verify the performance of clinicomics-guided distant metastasis risk prediction for breast cancer via artificial intelligence and to investigate the accuracy of the created prediction models for metachronous distant metastasis, bone metastasis and visceral metastasis.MethodsWe retrospectively enrolled 6703 breast cancer patients from 2011 to 2016 in our hospital. The figures of magnetic resonance imaging scanning and ultrasound were collected, and the figures features of distant metastasis in breast cancer were detected. Clinicomics-guided nomogram was proven to be with significant better ability on distant metastasis prediction than the nomogram constructed by only clinical or radiographic data.ResultsThree clinicomics-guided prediction nomograms on distant metastasis, bone metastasis and visceral metastasis were created and validated. These models can potentially guide metachronous distant metastasis screening and lead to the implementation of individualized prophylactic therapy for breast cancer patients.ConclusionOur study is the first study to make cliniomics a reality. Such cliniomics strategy possesses the development potential in artificial intelligence medicine.

EMLI-ICC: an ensemble machine learning-based integration algorithm for metastasis prediction and risk stratification in intrahepatic cholangiocarcinoma.

Robust strategies to identify patients at high risk for tumor metastasis, such as those frequently observed in intrahepatic cholangiocarcinoma (ICC), remain limited. While gene/protein expression profiling holds great potential as an approach to cancer diagnosis and prognosis, previously developed protocols using multiple diagnostic signatures for expression-based metastasis prediction have not been widely applied successfully because batch effects and different data types greatly decreased the predictive performance of gene/protein expression profile-based signatures in interlaboratory and data type dependent validation. To address this problem and assist in more precise diagnosis, we performed a genome-wide integrative proteome and transcriptome analysis and developed an ensemble machine learning-based integration algorithm for metastasis prediction (EMLI-Metastasis) and risk stratification (EMLI-Prognosis) in ICC. Based on massive proteome (216) and transcriptome (244) data sets, 132 feature (biomarker) genes were selected and used to train the EMLI-Metastasis algorithm. To accurately detect the metastasis of ICC patients, we developed a weighted ensemble machine learning method based on k-Top Scoring Pairs (k-TSP) method. This approach generates a metastasis classifier for each bootstrap aggregating training data set. Ten binary expression rank-based classifiers were generated for detection of metastasis separately. To further improve the accuracy of the method, the 10 binary metastasis classifiers were combined by weighted voting based on the score from the prediction results of each classifier. The prediction accuracy of the EMLI-Metastasis algorithm achieved 97.1% and 85.0% in proteome and transcriptome datasets, respectively. Among the 132 feature genes, 21 gene-pair signatures were developed to establish a metastasis-related prognosis risk-stratification model in ICC (EMLI-Prognosis). Based on EMLI-Prognosis algorithm, patients in the high-risk group had significantly dismal overall survival relative to the low-risk group in the clinical cohort (P-value < 0.05). Taken together, the EMLI-ICC algorithm provides a powerful and robust means for accurate metastasis prediction and risk stratification across proteome and transcriptome data types that is superior to currently used clinicopathological features in patients with ICC. Our developed algorithm could have profound implications not just in improved clinical care in cancer metastasis risk prediction, but also more broadly in machine-learning-based multi-cohort diagnosis method development. To make the EMLI-ICC algorithm easily accessible for clinical application, we established a web-based server for metastasis risk prediction (http://ibi.zju.edu.cn/EMLI/).

201. PREDICTION OF LYMPH NODE METASTASIS IN PATIENTS WITH ESOPHAGEAL SQUAMOUS CELL CARCINOMA USING GENETIC SIGNATURES FROM LARGE-SCALE GENOMIC REPERTOIRES

Abstract The presence of lymph node metastasis puts patients with resectable esophageal squamous cell carcinoma (ESCC) at a high risk of disease recurrence, and these patients would benefit from neoadjuvant treatment. However, current clinical staging is not enough to capture all patients with high-risk disease, and novel diagnostic tools are required to improve the accuracy of treatment planning. We tried to explore a better prediction using genomic features from large-scale gene mutation data. The major public data repositories of whole-exome sequencing for ESCC were involved in this study. The absolute differences in gene mutation frequency were calculated using Fisher’s exact test in nodal negative versus nodal positive patients. The classification of gene mutation status was dichotomized as wild-type and mutant type. To find a reliable gene panel that could provide sufficient population coverage, we set the cut-off level above 85% in both nodal phenotypes. The genomic classifier for nodal metastasis risk prediction was further generated by LASSO logistic regression analysis with 10-fold cross-validation based on the topmost significantly different genes. Above the threshold values, a total of 120 differentially mutated genes were identified in the 561 ESCC genomic sequencing data. The population coverage rates were 85.97% in the nodal positive group (n = 335) and 88.05% in the nodal negative group (n = 226), respectively. According to the LASSO logistic regression approach, a panel of 112 topmost relevant genes (termed LNM112) was selected. The receiver operating characteristic curve of the LNM112 predictive model was plotted with an area under the curve value of 0.95 (95% CI: 0.93–0.96). The sensitivity and specificity of the LNM112 model were 96.72% (324/335) and 71.24% (161/226), respectively. We identified a genomic signature (112 genes) from the large-scale genomic repertoires capable of predicting lymph node metastasis in ESCC. It can be incorporated into the current staging modalities to help inform treatment decisions, and external validation is warranted in different populations.

Real-World Evidence Shows Clinicians Appropriately Use the Prognostic 40-Gene Expression Profile (40-GEP) Test for High-Risk Cutaneous Squamous Cell Carcinoma (cSCC) Patients

Treatment decisions for patients with cutaneous squamous cell carcinoma (cSCC) are traditionally based upon clinicopathologic risk factors and staging systems. Due to the accuracy limitations of these resources in predicting poor outcomes, there is a clinically significant need for more accurate methods of risk assessment. The 40-gene expression profile (40-GEP) test was developed to augment metastatic risk prediction of high-risk cSCC patients and has been validated in two independent, multi-center studies involving over 1,000 patients. This study substantiates that the 40-GEP is appropriately utilized by clinicians and that the personalized risk-stratification results are impactful in guiding risk-aligned patient management.

Genetic Basis and Molecular Mechanisms of Uveal Melanoma Metastasis: A Focus on Prognosis.

Uveal melanoma (UM) is the most frequently found primary intraocular tumor, although it accounts for only 5% of all melanomas. Despite novel systemic therapies, patient survival has remained poor. Indeed, almost half of UM patients develop metastases from micro-metastases which were undetectable at diagnosis. Genetic analysis is crucial for metastatic risk prediction, as well as for patient management and follow-up. Several prognostic parameters have been explored, including tumor location, basal dimension and thickness, histopathologic cell type, vascular mimicry patterns, and infiltrating lymphocytes. Herein, the Authors review the available literature concerning cytogenetic prognostic markers and biochemical pathways correlated to UM metastasis development.

Establishment and Validation of a Novel MRI Radiomics Feature-Based Prognostic Model to Predict Distant Metastasis in Endemic Nasopharyngeal Carcinoma

PurposeWe aimed to establish a prognostic model based on magnetic resonance imaging (MRI) radiomics features for individual distant metastasis risk prediction in patients with nasopharyngeal carcinoma (NPC).MethodsRegression analysis was applied to select radiomics features from T1-weighted (T1-w), contrast-enhanced T1-weighted (T1C-w), and T2-weighted (T2-w) MRI scans. All prognostic models were established using a primary cohort of 518 patients with NPC. The prognostic ability of the radiomics, clinical (based on clinical factors), and merged prognostic models (integrating clinical factors with radiomics) were identified using a concordance index (C-index). Models were tested using a validation cohort of 260 NPC patients. Distant metastasis-free survival (DMFS) were calculated by using the Kaplan-Meier method and compared by using the log-rank test.ResultsIn the primary cohort, seven radiomics prognostic models showed similar discrimination ability for DMFS to the clinical prognostic model (P=0.070-0.708), while seven merged prognostic models displayed better discrimination ability than the clinical prognostic model or corresponding radiomics prognostic models (all P<0.001). In the validation cohort, the C-indices of seven radiomics prognostic models (0.645-0.722) for DMFS prediction were higher than in the clinical prognostic model (0.552) (P=0.016 or <0.001) or in corresponding merged prognostic models (0.605-0.678) (P=0.297 to 0.857), with T1+T1C prognostic model (based on Radscore combinations of T1 and T1C Radiomics models) showing the highest C-index (0.722). In the decision curve analysis of the validation cohort for all prognostic models, the T1+T1C prognostic model displayed the best performance.ConclusionsRadiomics models, especially the T1+T1C prognostic model, provided better prognostic ability for DMFS in patients with NPC.