Survival Probability Of Patients Research Articles

Multimodal deep learning for personalized renal cell carcinoma prognosis: Integrating CT imaging and clinical data

Background and Objective: Renal cell carcinoma represents a significant global health challenge with a low survival rate. The aim of this research was to devise a comprehensive deep-learning model capable of predicting survival probabilities in patients with renal cell carcinoma by integrating CT imaging and clinical data and addressing the limitations observed in prior studies. The aim is to facilitate the identification of patients requiring urgent treatment.Methods: The proposed framework comprises three modules: a 3D image feature extractor, clinical variable selection, and survival prediction. Based on the 3D CNN architecture, the feature extractor module predicts the ISUP grade of renal cell carcinoma tumors linked to mortality rates from CT images. Clinical variables are systematically selected using the Spearman score and random forest importance score as criteria. A deep learning-based network, trained with discrete LogisticHazard-based loss, performs the survival prediction. Nine distinct experiments are performed, with varying numbers of clinical variables determined by different thresholds of the Spearman and importance scores.Results: Our findings demonstrate that the proposed strategy surpasses the current literature on renal cancer prognosis based on CT scans and clinical factors. The best-performing experiment yielded a concordance index of 0.84 and an area under the curve value of 0.8 on the test cohort, which suggests strong predictive power.Conclusions: The multimodal deep-learning approach developed in this study shows promising results in estimating survival probabilities for renal cell carcinoma patients using CT imaging and clinical data. This may have potential implications in identifying patients who require urgent treatment, potentially improving patient outcomes. The code created for this project is available for the public on: GitHub

Confidence intervals of survival predictions with neural networks trained on molecular data

In medicine, an important objective is predicting patients’ survival based on their molecular and clinical characteristics. In this context, neural networks have recently been used for their ability to capture complex interactions in the data. Measuring the uncertainty associated with survival estimates obtained by neural networks is essential to enhance predictions’ reliability. We compared four methods adapted to multilayer perceptrons (MLPs) for building confidence intervals at the patient level. The methods were based either on bootstrap with Boot (Efron, 1979), ensembling with DeepEns (Lakshminarayanan et al., 2016), or Monte-Carlo Dropout with MCDrop and BMask (Gal and Ghahramani, 2016; Mancini et al., 2020). A comparison was made through MLP-based survival models: CoxCC and CoxTime (Kvamme et al., 2019) in a continuous time framework, DeepHit (Lee et al., 2018) and PLANN (Biganzoli et al., 1998) in a discrete time framework. We applied the methods to a simulation study, enabling us to estimate a coverage rate of the estimated confidence intervals. We also applied them to real-world datasets, and predicted the survival probability for patients with breast cancer and patients with lung cancer.In the simulation study, CoxCC and CoxTime obtained the mean C-indices numerically closest to those from the Oracle model (mean C-index of 0.723 for CoxCC, 0.726 for CoxTime, versus 0.743 for the Oracle model). Regarding the confidence intervals of survival probabilities, Boot with CoxCC obtained a coverage rate of 96.5%, the closest to the nominal value of 95%. MCDrop was slightly anticonservative and obtained a coverage rate of 89.8% with CoxTime. This method may represent a reasonable compromise in terms of coverage with regards of computational time. In the breast cancer cohort, MLPs had difficulty capturing additional prognostic information from the molecular data. In contrast, in the lung cancer cohort, the models led to substantially stronger discrimination values when adding molecular data to the clinical variables. In conclusion, we were able to represent uncertainty in the survival estimates at particular time points at the patient level using MLPs in the form of 95% confidence intervals. We recommend using CoxTime with either Boot or, for a less intensive computation time, MCDrop.

Ischemic Stroke Temporally Associated With New-Onset Atrial Fibrillation: A Population-Based Registry-Linkage Study.

Limited data exist on the temporal relationship between new-onset atrial fibrillation (AF) and ischemic stroke and its impact on patients' clinical characteristics and mortality. A population-based registry-linkage database includes all patients with new-onset AF in Finland from 2007 to 2018. Ischemic stroke temporally associated with AF (ISTAF) was defined as an ischemic stroke occurring within ±30 days from the first AF diagnosis. Clinical factors associated with ISTAF were studied with logistic regression and 90-day survival with Cox proportional hazards analysis. Among 229 565 patients with new-onset AF (mean age, 72.7 years; 50% female), 204 774 (89.2%) experienced no ischemic stroke, 12 209 (5.3%) had past ischemic stroke >30 days before AF, and 12 582 (5.8%) had ISTAF. The annual proportion of ISTAF among patients with AF decreased from 6.0% to 4.8% from 2007 to 2018. Factors associated positively with ISTAF were higher age, lower education level, and alcohol use disorder, whereas vascular disease, heart failure, chronic kidney disease cancer, and psychiatric disorders were less probable with ISTAF. Compared with patients without ischemic stroke and those with past ischemic stroke, ISTAF was associated with ≈3-fold and 1.5-fold risks of death (adjusted hazard ratios, 2.90 [95% CI, 2.76-3.04] and 1.47 [95% CI, 1.39-1.57], respectively). The 90-day survival probability of patients with ISTAF increased from 0.79 (95% CI, 0.76-0.81) in 2007 to 0.89 (95% CI, 0.87-0.91) in 2018. ISTAF depicts the prominent temporal clustering of ischemic strokes surrounding AF diagnosis. Despite having fewer comorbidities, patients with ISTAF had worse, albeit improving, survival than patients with a history of or no ischemic stroke. URL: https://www.clinicaltrials.gov; Unique identifier: NCT04645537. URL: https://www.encepp.eu; Unique identifier: EUPAS29845.

Real-time optimization for relocation and dispatching of Emergency Medical Services with balanced workload and outsourced ride-hailing services

Real-time optimization of Emergency medical services (EMS) leads to excessive and unbalanced workload unless balance is directly modeled. Online redeployment policies aim to reduce response times, compensating for the coverage gap left by busy ambulances, however, it is important to account for the additional effort required to relocate to a new position. Additionally, irregular demand patterns result in unequal distributions of workload among the personnel. In this regard, we propose a new survival-based real-time optimization framework that maximizes the survival probability of patients suffering from life-threatening emergencies while maintaining a balanced workload among the crew. Our approach implements an optimization model inside a discrete event simulation model that feeds the former with real-time vehicle utilization metrics to maintain an optimal overall relocation policy. Moreover, we evaluate the impact on the system’s congestion of a mixed dispatched policy that includes external ride-hailing services for minor emergencies. Our case study in New York City indicates that the proposed model increases the average percentage of life-threatening emergencies responded in less than four minutes in more than 10% through the use of survival functions compared to a coverage model from the literature. Furthermore, even though the fleet’s aggregated driving distances increased compared to the coverage approach, our online model effectively balanced the additional workload among the crew. Additionally, ride-hailing services’ introduction presents an important opportunity for reducing personnel workload by 13.5%, on average, and operational costs in $141,220 per day of operation for the entire system. Finally, when ambulance availability is severely reduced, outsourced services significantly help maintaining the response level for low severity emergencies.

Comprehensive analysis of immune cell landscapes revealed that immune cell ratio eosinophil/B.cell.memory is predictive of survival in sepsis

BackgroundImmune dysregulation is a feature of sepsis. However, a comprehensive analysis of the immune landscapes in septic patients has not been conducted.ObjectivesThis study aims to explore the abundance ratios of immune cells in sepsis and investigate their clinical value.MethodsSepsis transcriptome data sets were downloaded from the NCBI GEO database. The immunedeconv R package was employed to analyze the abundance of immune cells in sepsis patients and calculate the ratios of different immune cell types. Differential analysis of immune cell ratios was performed using the t test. The Spearman rank correlation coefficient was utilized to find the relationships between immune cell abundance and pathways. The prognostic significance of immune cell ratios for patient survival probability was assessed using the log-rank test. In addition, differential gene expression was performed using the limma package, and gene co-expression analysis was executed using the WGCNA package.ResultsWe found significant changes in immune cell ratios between sepsis patients and healthy controls. Some of these ratios were associated with 28-day survival. Certain pathways showed significant correlations with immune cell ratios. Notably, six immune cell ratios demonstrated discriminative ability for patients with systemic inflammatory response syndrome (SIRS), bacterial sepsis, and viral sepsis, with an Area Under the Curve (AUC) larger than 0.84. Patients with a high eosinophil/B.cell.memory ratio exhibited poor survival outcomes. A total of 774 differential genes were identified in sepsis patients with a high eosinophil/B.cell.memory ratio compared to those with a low ratio. These genes were organized into seven co-expression modules associated with relevant pathways, including interferon signaling, T-cell receptor signaling, and specific granule pathways.ConclusionsImmune cell ratios eosinophil/B.cell.memory and NK.cell.activated/NK.cell.resting in sepsis patients can be utilized for disease subtyping, prognosis, and diagnosis. The proposed cell ratios may have higher prognostic values than the neutrophil-to-lymphocyte ratio (NLR).

Identifying candidate drugs based on transcriptional landscape associated with triple-negative breast cancer

Triple-negative breast cancer (TNBC) is a special type of breast cancer in which ER, PR and HER2 are all negative, characterized by high malignancy, strong invasiveness, and high recurrence rate. It is critical to develop novel drugs for improved therapies for triple-negative breast cancers. Here, we unveiled the transcriptional landscape associated with triple-negative breast cancer and identified candidate drugs using a comprehensive connectivity map. The candidate components could induce reverse expressions of the TNBC gene expression signature and trigger transcriptional reprogramming with a positive impact on modulating chemoresistance and the survival probability of breast cancer patients. Our study also provided an example of drug discovery using in silico drug screening followed by further validations, illustrating an effective computational drug discovery strategy.

Integrative Genomic and Transcriptomic Analysis Reveals Genetic Alterations Associated with the Early Progression of Follicular Lymphoma

Background Follicular lymphoma (FL), the most common indolent lymphoma, is a clinically and genetically heterogeneous disease. However, the prognostic value of driver gene mutations and copy number alterations has not been systematically assessed. Patients and Methods Here, we analyzed clinical-biological features of 415 FL patients to identify variables associated with disease within 24 months of first-line therapy (POD24). In addition, we applied whole exome sequencing (WES) to identify genomic alterations that predict POD24 based on 102 FL patients. Furthermore, we characterized the cellular and molecular heterogeneity within and across patients through bulk RNA sequencing and single-cell RNA sequencing. ResultsPOD24 occurred in 21% of evaluable FL patients. To further examine the effect of POD24 on OS, we used a 24-month landmark approach for Kaplan-Meier curve analysis. As expected, POD24 was related to poor OS, and the 3- and 5-year survival probabilities of patients in whom the disease progressed within the first 24 months were 89.2% (95% CI, 84.5-93.9) and 78.6% (95% CI, 70.4-86.8), respectively, compared with 98.5% (95% CI, 97.8-99.2) and 96.2% (95% CI, 95.0-97.4) in patients who were progression free at 24 months, respectively. Moreover, patients without progression at 24 months were more likely to exhibit a CR to front-line treatment (54% vs. 35%, P = 0.001) compared to the POD24 cohort. Patients with B symptoms, elevated lactate dehydrogenase and β2-microglobulin levels, unfavorable baseline haemoglobin levels, advanced stage, and high-risk FL International Prognostic Index (FLIPI) scores had an increased risk of POD24, with FLIPI being the most important factor in logistic regression. HIST1H1D, known as a driver mutation, was correlated with POD24. Gains of 6q22.2 ( HIST1H1D) and 18q21.33 ( BCL2) and loss of 1p36.13 ( NBPF1) predicted POD24 independent of FLIPI. Integration of the four variants led to the identification of 76% of POD24 patients. Gene expression profiling of FL samples showed that the POD24 cohort was significantly enriched in the inflammatory response (mediated by interferon and tumor necrosis factor), cell cycle regulation (transcription, replication and proliferation) sets and PI3K-AKT-mTOR signaling. This result was further validated with transcriptome-wide information provided by RNA-seq at single-cell resolution. ConclusionOur study, performed on a large cohort of FL patients, highlights the importance of distinctive genetic alterations and gene expression relevant to disease diagnosis and early progression.

A prognostic model based on Scissor+ cancer associated fibroblasts identified from bulk and single cell RNA sequencing data in head and neck squamous cell carcinoma

Head and neck squamous cell carcinoma (HNSCC) is one of the most lethal diseases in the world, which often recur after multimodality treatment approaches, leading to a poor prognosis. Fibroblasts, a heterogeneous component of the tumor microenvironment, can modulate numerous aspects of tumor biology and have been increasingly acknowledged in dictating the clinical outcome of patients with HNSCC. However, the subpopulation of fibroblasts that are related to the prognosis of HNSCC has not yet been fully explored. To do so, we combined a single-cell RNA sequencing (scRNA-seq) dataset and bulk RNA-sequencing dataset with clinical information, identifying the fibroblast population that are related to poor prognosis of HNSCC. We found these specific population of fibroblasts are less differentiated. In addition, to identify the prognostic signatures of HNSCC, bioinformatics analysis included least absolute shrinkage and selection operator (LASSO) analyses and univariate cox and were performed. We selected 12 prognosis-related genes for constructing a risk model using The Cancer Genome Atlas (TCGA). The AUC values and calibration plots of this model indicated good prognostic prediction efficacy. This model also was validated in two Gene Expression Omnibus (GEO) datasets. In conclusion, we constructed an optimal model that was derived from single cell RNA-seq and bulk RNA-seq to predict the survival probability of HNSCC patients. Among this model, AKR1C3 higher expression in cancer associated fibroblasts (CAFs) of HNSCC has been confirmed by preliminary experiments.

Survival probabilities and prognostic factors of pregnancy-associated breast cancer in Malaysian women.

Pregnancy-associated breast cancer (PABC) is a rare cancer. This study aimed to determine the survival probabilities and prognostic factors in patients with PABC. A retrospective cohort study was conducted in two tertiary care hospitals in Kota Bharu. We included all patients with breast cancer who were diagnosed by histopathology while pregnant or within 2 years post-partum from 2001 through 2020. We matched patients with PABC to non-pregnant patients with breast cancer by age and year of diagnosis. The data were analyzed using Cox proportional hazard regression. A total of 35 cases of PABC and 70 non-PABC controls were recruited. The 3-year, 5-year, and 10-year survival probabilities for patients with PABC were 58.6%, 47.54%, and 38.03%, respectively. The patients with PABC had a non-significant difference in survival probabilities compared with non-PABC patients. The significant prognostic factors of PABC were age (adjusted hazard ratio [aHR], 0.91; 95% confidence interval [CI], 0.86-0.96; P=0.001), advanced stage of cancer (aHR, 9.97; 95% CI, 3.96-25.2; P<0.001), and no surgery (aHR, 3.16; 95% CI, 1.01-9.85; P=0.047). Pregnancy was not found to be an independent factor in the prognosis of PABC (aHR, 0.72; 95% CI, 0.39-1.28; P=0.266). Women diagnosed with PABC had similar survival probabilities compared with non-PABC patients. Pregnancy was not an independent prognostic factor for breast cancer. This information can be useful when women with breast cancer are counseled and supported with the option of beginning treatment with pregnancy continuation.

Assessing the prognostic impact of prostatic urethra involvement and developing a nomogram for T1 stage bladder cancer

PurposeTo investigate prognostic values of prostatic urethra involvement (PUI) and construct a prognostic model that estimates the probability of cancer-specific survival for T1 bladder cancer patients.Method and materialsWe investigated the national Surveillance, Epidemiology, and End Results (SEER) database (2004–2015) to get patients diagnosed with T1 bladder cancer. An external validation cohort was obtained from the First Affiliated Hospital of Nanchang University. The Kaplan–Meier method with the log-rank test was applied to assess cancer-specific survival (CSS) and overall survival (OS). Moreover, the propensity score matching (PSM) and multivariable Cox proportional hazard model were performed. All patients were randomly divided into the development cohort and validation group at the ratio of 7:3. The performance of the model was internally validated by calibration curves and the concordance index (C-index).ResultsThe PUI group had a lower survival rate of both CSS and overall survival OS before and after PSM when compared to non-involved patients (All P < 0.05). Multivariate analysis revealed a poor prognosis in the PUI group for cancer-specific mortality (CSM) and all-cause mortality (ACM) analyses before and after PSM (All P < 0.05). Seven variables, including age, surgery, radiotherapy, tumour size, PUI, and marital status, were incorporated in the final nomogram. The C-index in the development cohort was 0.715 (0.711–0.719), while it was 0.672 (0.667–0.677) in the validation group. Calibration plots for 3- and 5-year cancer-specific survival showed good concordance in the development and validation cohorts.ConclusionsPUI was an independent risk factor of ACM and CSM in T1 bladder cancer patients. In addition, a highly discriminative and precise nomogram that predicted the individualized probability of cancer-specific survival for patients with T1 bladder cancer was constructed.

System analysis based on the pyroptosis-related genes identifes GSDMD as a novel therapy target for skin cutaneous melanoma

BackgroundSkin cutaneous melanoma (SKCM) is the most aggressive skin cancer, accounting for more than 75% mortality rate of skin-related cancers. As a newly identified programmed cell death, pyroptosis has been found to be closely associated with tumor progression. Nevertheless, the prognostic significance of pyroptosis in SKCM remains elusive.MethodsA total of 469 SKCM samples and 812 normal samples were obtained from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases. Firstly, differentially expressed pyroptosis-related genes (PRGs) between normal samples and SKCM samples were identified. Secondly, we established a prognostic model based on univariate Cox and LASSO Cox regression analyses, which was validated in the test cohort from GSE65904. Thirdly, a nomogram was used to predict the survival probability of SKCM patients. The R package “pRRophetic” was utilized to identify the drug sensitivity between the low- and high-risk groups. Tumor immune infiltration was evaluated using “immuneeconv” R package. Finally, the function of GSDMD and SB525334 was explored in A375 and A2058 cells.ResultsBased on univariate Cox and LASSO regression analyses, we established a prognostic model with identified eight PRGs (AIM2, CASP3, GSDMA, GSDMC, GSDMD, IL18, NLRP3, and NOD2), which was validated in the test cohort. SKCM patients were divided into low- and high-risk groups based on the median of risk score. Kaplan–Meier survival analysis showed that high-risk patients had shorter overall survival than low-risk patients. Additionally, time-dependent ROC curves validated the accuracy of the risk model in predicting the prognosis of SKCM. More importantly, 4 small molecular compounds (SB525334, SR8278, Gemcitabine, AT13387) were identified, which might be potential drugs for patients in different risk groups. Finally, overexpression of GSDMD and SB525334 treatment inhibit the proliferation, migration, and invasion of SKCM cells.ConclusionIn this study, we constructed a prognostic model based on PRGs and identified GSDMD as a potential therapeutic target, which provide new insights into SKCM treatment.

Kidney-type glutaminase is a biomarker for the diagnosis and prognosis of hepatocellular carcinoma: a prospective study

PurposeThe pathological diagnosis and prognosis prediction of hepatocellular carcinoma (HCC) is challenging due to the lack of specific biomarkers. This study aimed to validate the diagnostic and prognostic efficiency of Kidney-type glutaminase (GLS1) for HCC in prospective cohorts with a large sample size.MethodsA total of 1140 HCC patients were enrolled in our prospective clinical trials. Control cases included 114 nontumour tissues. The registered clinical trial (ChiCTR-DDT-14,005,102, chictr.org.cn) was referred to for the exact protocol. GLS1 immunohistochemistry was performed on the whole tumour section. The diagnostic and prognostic performances of GLS1 was evaluated by the receiver operating characteristic curve and Cox regression model.ResultsThe sensitivity, specificity, positive predictive value, negative predictive value, Youden index, and area under the curve of GLS1 for the diagnosis of HCC were 0.746, 0.842, 0.979, 0.249, 0.588, and 0.814, respectively, which could be increased to 0.846, 0.886, 0.987,0.366, 0.732, and 0.921 when combined with glypican 3 (GPC3) and alpha-fetoprotein (AFP), indicating better diagnostic performance. Further, we developed a nomogram with GPC3 and GLS1 for identifying HCC which showed good discrimination and calibration. GLS1 expression was also related with age, T stage, TNM stage, Edmondson–Steiner grade, microvascular invasion, Ki67, VEGFR2, GPC3, and AFP expression in HCC. GLS1 expression was negatively correlated with disease-free survival (P < 0.001) probability of patients with HCC.ConclusionsIt was validated that GLS1 was a sensitive and specific biomarker for pathological diagnosis of HCC and had prognostic value, thus having practical value for clinical application.

Abnormal workload-indexed blood pressure response to exercise in patients at risk of or with overt heart failure

Abstract Background Conflicting evidence exists regarding the functional and prognostic significance of hypertensive response to exercise (HRE), commonly defined as peak systolic blood pressure (SBP) &amp;gt;210 mmHg in males and &amp;gt;190 mmHg in females. No study yet has explored the impact of sex-specific, workload-indexed HRE in patients at different stages of the heart failure (HF) spectrum, as defined by the American College of Cardiology/American Heart Association HF staging system. Purpose We evaluated the SBP/Workload slope during exercise across the HF spectrum and investigated the hemodynamic, functional and prognostic correlates of HRE. Methods Fifty-eight healthy controls, 201 subjects at risk of developing HF (HF Stages A-B), and 369 patients with a definite diagnosis of HF (Stage C) were enrolled; among the latter, 143 had preserved (HFpEF) and 226 reduced (HFrEF) left ventricular ejection fraction. All patients underwent a combined cardiopulmonary-exercise stress echocardiography test. For each patient, we recorded at least two BP measurements during exercise: the first within the first minute of exercise and the last at peak effort. The SBP/Workload slope was calculated as the difference in SBP over the corresponding increment in watts (W): (SBPpeak – SBPfirst)/(Wpeak – Wfirst). In each ACC/AHA HF Stage, we divided patients according to SBP/Workload slope tertiles; HRE was defined as the highest sex-specific SBP/Workload slope tertile. A composite of all-cause mortality and hospitalization for cardiovascular reasons was chosen as the primary outcome. Results The median SBP/Workload slope was 0.53 (interquartile range 0.36-0.72) mmHg/W and was steeper in women than in men in all subgroups (p&amp;lt;0.0001). After adjusting for age and sex, SBP/Workload slope was significantly higher in Stages A-B (0.61, 0.47-0.75 mmHg/W) and HFpEF (0.63, 0.42-0.86 mmHg/W) compared to controls (0.43, 0.35-0.57 mmHg/W) and HFrEF (0.47, 0.30-0.63 mmHg/W, Figure 1). In all subgroups, patients with HRE showed significantly lower peak oxygen consumption and peripheral oxygen extraction (Table). After a median follow-up of 16 months, HRE was an independent predictor of adverse outcomes (Hazard Ratio 2.05, 95% confidence interval 1.81-5.18), while rest and peak absolute SBP values did not. Kaplan-Meier analysis confirmed a worse survival probability for patients with HRE in Stages A-B and HFpEF but not in HFrEF. After adjusting for confounding factors, the patients with the highest SBP/Workload slope had a 47% higher risk of the primary outcome than subjects in the lowest tertile (Figure 2). Conclusions Across the HF spectrum, patients with HRE, as evaluated by the SBP/Workload slope, present with peculiar pathophysiological features and impaired functional capacity. This is especially true for those in Stages A-B and patients with HFpEF. In these subgroups, HRE may portend an increased risk of adverse cardiovascular outcomes.SBP/workload slopesKaplan-Meier curves and Hazard Ratios

The mitotic cell cycle-associated nomogram predicts overall survival in lung adenocarcinoma.

This study aimed to develop a prognostic model for lung adenocarcinoma (LUAD) associated with mitotic cell cycle. The model will predict the probability of survival at different time points and serve as a reference tool to evaluate the effectiveness of LUAD treatment. A cohort of 442 patients with LUAD from the gene expression omnibus (GEO) database was randomly divided into a training group (n = 299) and a validation group (n = 99). The least absolute shrinkage and selection operator (LASSO)-COX algorithm was used to reduce the number of predictors based on the clinicopathological and RNA sequencing data to establish mutant characteristics that could predict patient survival. Additionally, gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), gene set variation analysis (GSVA), and gene set enrichment analysis (GSEA) analyses were conducted on the mutant characteristics. The performance of the developed nomogram was evaluated using calibration curves and the C-index. The mutant characteristics had prognostic value for LUAD and acted as an independent prognostic factor. The mutant characteristics profile derived from the LASSO-COX algorithm demonstrated a significant association with overall survival in patients with LUAD. Functional annotation based on the mutant score, its involvement in the phase transition of the mitotic cell cycle, and its regulatory processes. The nomogram, which combined the mutant score with clinical factors associated with prognosis, showed robust accuracy in both the training and validation groups. This study presents the first individualized model that establishes a mutant score for predicting survival in LUAD. This model can be used as a predictive tool for determining 1-, 2-, 3-, and 5-year survival probabilities in patients with LUAD.

The development and experimental validation of hypoxia-related long noncoding RNAs prognostic signature in predicting prognosis and immunotherapy of cutaneous melanoma.

Cutaneous melanoma (CM) is widely acknowledged as a highly aggressive form of malignancy that is associated with a considerable degree of morbidity and poor prognosis. Despite this recognition, the precise role of hypoxia-related long noncoding RNAs (HRLs) in the pathogenesis of CM remains an area of active research. This study sought to elucidate the contribution of HRLs in CM by conducting a thorough screening and extraction of hypoxia-related genes (HRGs). In particular, we conducted univariate and multivariate Cox regression analyses to assess the independence of the prognostic signature of HRLs. Our results demonstrated that a novel risk model could be established based on five prognostic HRLs. Remarkably, patients with low-risk scores exhibited significantly higher overall survival rates compared to their high-risk counterparts, as confirmed by Kaplan-Meier survival analysis. Furthermore, we utilized consensus clustering analysis to categorize CM patients into two distinct subtypes, which revealed marked differences in their prognosis and immune infiltration landscapes. Our nomogram results confirmed that the HRLs prognostic signature served as an independent prognostic indicator, offering an accurate evaluation of the survival probability of CM patients. Notably, our findings from ESTIMATE and ssGSEA analyses highlighted significant disparities in the immune infiltration landscape between low- and high-risk groups of CM patients. Additionally, IPS and TIDE results suggested that CM patients in different risk subtypes may exhibit favorable responses to immunotherapy. Enrichment analysis and GSVA results indicated that immune-related signaling pathways may mediate the role of HRLs in CM. Finally, our tumor mutation burden (TMB) results indicated that patients with low-risk scores had a higher TMB status. In summary, the establishment of a risk model based on HRLs in this study provided an accurate prognostic prediction and correlated with the immune infiltration landscape of CM, thereby providing novel insights for the future clinical management of this disease.

A novel immune-related gene signature for predicting immunotherapy outcomes and survival in clear cell renal cell carcinoma

Clear cell renal carcinoma (ccRCC) is one of the most common cancers worldwide. In this study, a new model of immune-related genes was developed to predict the overall survival and immunotherapy efficacy in patients with ccRCC. Immune-related genes were obtained from the ImmPort database. Clinical data and transcriptomics of ccRCC samples were downloaded from GSE29609 and The Cancer Genome Atlas. An immune-related gene-based prognostic model (IRGPM) was developed using the least absolute shrinkage and selection operator regression algorithm and multivariate Cox regression. The reliability of the developed models was evaluated by Kaplan–Meier survival curves and time-dependent receiver operating characteristic curves. Furthermore, we constructed a nomogram based on the IRGPM and multiple clinicopathological factors, along with a calibration curve to examine the predictive power of the nomogram. Overall, this study investigated the association of IRGPM with immunotherapeutic efficacy, immune checkpoints, and immune cell infiltration. Eleven IRGs based on 528 ccRCC samples significantly associated with survival were used to construct the IRGPM. Remarkably, the IRGPM, which consists of 11 hub genes (SAA1, IL4, PLAUR, PLXNB3, ANGPTL3, AMH, KLRC2, NR3C2, KL, CSF2, and SEMA3G), was found to predict the survival of ccRCC patients accurately. The calibration curve revealed that the nomogram developed with the IRGPM showed high predictive performance for the survival probability of ccRCC patients. Moreover, the IRGPM subgroups showed different levels of immune checkpoints and immune cell infiltration in patients with ccRCC. IRGPM might be a promising biomarker of immunotherapeutic responses in patients with ccRCC. Overall, the established IRGPM was valuable for predicting survival, reflecting the immunotherapy response and immune microenvironment in patients with ccRCC.

Clinical Text Reports to Stratify Patients Affected with Myeloid Neoplasms Using Natural Language Processing

Background: The availability of multimodal patient data, such as demographics, clinical, imaging, treatment, quality of life, outcomes and wearables data, as well as genome sequencing, have paved the way for the development of multimodal clinical solutions that introduce personalized or precision medicine. The clinical report is an information layer that contains relevant information about the disease in addition to the patient's point of view. Natural language processing (NLP) is a branch of artificial intelligence (AI) and its pre-trained language models are the key technology for extracting value from this data layer. Aims: This project was conducted by GenoMed4all and Synthema EU consortia, with the aim to: 1) Build an AI language model specific for the hematology domain. 2) Use NLP technology to extract relevant information from clinical reports and perform unsupervised stratification of patients, in order to 3) demonstrate that the clinical report is earlier access to data relative to disease clinical phenotype and biology and provide important information for patient stratification and prediction of clinical outcomes. Methods: To translate text sentences into numerical embeddings, we implemented bidirectional encoder representations from transformers (BERT) framework. To learn text representations and correlations within data, we performed domain-adaptation by fine-tuned pre-trained model on hematological clinical reports of patients with myeloproliferative neoplasms (MPN), myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). Patient stratification was performed by HDBSCAN clustering on text embedding encoded by BERT (HematoBERT). Clusters validation was performed by assessing patients' diagnosis and survival probability. Finally, we compared domain-tuned HematoBERT vs pre-trained non-contextualized models. Results: We implemented HematoBERT based on the bert-base-multilingual-uncased version of BERT. Training data were hematological text reports of 1,328 patients. During fine-tuning, texts were tokenized, then we randomly replaced 15% of the tokens with masked tokens, training the model to predict them. We performed stratification using clinical reports from a validation cohort of 360 patients. We identified 7 clusters, defined according to similar words in meaning that were placed in a specific topic. We extracted the most important words and concepts for each cluster (topic) and we summarized them into effective descriptions for each group of patients. Two clusters included MDS patients with excess blasts, and without excess blasts with ring sideroblasts and del5q (n=69, n=115). One cluster included patients with excess blasts and MDS/MPN (n= 33). Two clusters included MPN patients with primary and secondary myelofibrosis, and MPN patients most including subjects affected with polycythemia vera and essential thrombocythemia (n=35, n=46). Two clusters included patients with AML from MDS and therapy-related AML, and patients with de novo AML (n=22, n=42). Clinical validation was performed based on the diagnosis and survival probability of patients assigned to clusters. Patients' diagnoses were compatible with the cluster assignment (Figure 1). Frequency of gene mutations (as assessed by targeted Next-Generation Sequencing) among different clusters reflected the well-known genotypic-phenotypic associations in MDS, MPN and AML. Kaplan-Maier curves indicated significative risk stratification in clusters in terms of survival probability (Figure 2), similar to stratifications performed on clinical and genomic data. Finally, we evaluate the domain adaptation by comparing the model to other pre-trained non-contextualized ones. Pseudo perplexity score (PPS), accuracy and F1 score were calculated to quantify how good the models are when they see new data, predicting the next word given the context of the sentence. HematoBERT obtained high PPS, accuracy and F1 scores, outperforming the other models also trained on generic clinical domains. Conclusion: Domain-adapted language models are able to understand contexts and correlations in documents. HematoBERT can be used to extract relevant features from clinical reports. This data layer is relevant to perform disease stratification of patients based on clinical and genomic information and could be integrated into next-generation multimodal models of personalized medicine.

Metastasis, characteristic, and treatment of breast cancer in young women and older women: A study from the Surveillance, Epidemiology, and End Results registration database.

Younger age is an independent risk factor for breast cancer (BC) prognosis, and BC in young women is often considered more aggressive. BC patients with different age and molecular subtypes have different metastasis patterns and survival. Herein, we aim to explore the metastasis patterns, characteristics and treatment methods of young patients with BC, and to compare them with older patients. Data of young patients (aged ≤40 years old) and older patients (aged >40 years old) with BC were extracted from the Surveillance, Epidemiology, and End Results (SEER) registration database in 2010-2019 in this retrospective cohort study. Univariate and multivariate competing risk models and proportional hazard models were used to explore the association between different metastasis patterns and treatments and BC prognoses in young and older patients. Kaplan-Meier (KM) curves were drawn to reflect the survival probability of patients with BC who have different metastasis patterns. Also, we performed subgroup analysis of different metastasis patterns to explore the association between different treatments and overall survival (OS)/cancer specific survival (CSS) in patients with BC. The evaluation index was hazard ratios (HRs) with 95% confidence intervals (CIs). Totally, 5,984 patients died, and 92.56% of them died from BC. There were respectively 1,089 young patients and 9,105 older patients, and we found some differences of characteristics and metastasis patterns between them. After adjusting for covariates, young patients who had brain metastasis and multiple sites metastasis seemed to have high risk of both lower OS and CSS. Among older patients with BC, brain metastasis, liver metastasis, and multiple sites metastasis were all positively associated with both lower OS and CSS. In young and older patients, those who not receive radiotherapy or surgery, or received non-surgery combined with radiotherapy seemed to have high risk of both lower OS and CSS. Breast-conserving surgery (BCS) and surgery combined with radiotherapy were associated with higher OS and CSS in young patients, while only older patients received surgery combined with radiotherapy had higher OS and CSS. Results of subgroup analysis indicated that for patients with different metastasis patterns, developing a personalized treatment plan is necessary. Characteristics of BC between young patients and older patients were different. Clinicians should focus on different metastasis sites and choose appropriate treatments in patients with different ages, which may improve the prognoses.

Knockdown of PROM2 Enhances Paclitaxel Sensitivity in Endometrial Cancer Cells by Regulating the AKT/FOXO1 Pathway.

Endometrial cancer is a very common and highly lethal reproductive malignant tumour in women. Paclitaxel (PTX) is a usual drug utilized in chemotherapy for endometrial cancer. It has been uncovered that PROM2 participates in the progression of various cancers through playing a promoter. However, the regulatory function of PROM2 in PTX treatment for endometrial cancer remains unclear. The cell viability (IC50) was examined through CCK8 assay. The mRNA and protein expressions of genes were measured through RT-qPCR and western blot. The proliferation was evaluated through colony formation and EdU assays. The cell apoptosis was assessed through flow cytometry. In this work, through bioinformatic analysis on online websites, it is found that the up-regulated expression of PROM2 existed in endometrial cancer. In addition, the survival probability of UCEC patients with high PROM2 expression was worse. This study adopted PTX treatment for obtaining the PTX-resistant cells (HEC-1A/PTX and KLE/PTX). Furthermore, suppression of PROM2 enhanced PTX sensitivity through decreasing IC50 and proliferation in endometrial cancer. Additionally, knockdown of PROM2 facilitated cell apoptosis in HEC-1A/PTX and KLE/PTX cells. Next, we found that silencing of PROM2 retards the AKT/FOXO1 pathway. At last, rescue assays reversed the strengthened PTX sensitivity mediated by PROM2 inhibition after SC79 treatment (AKT activator). Knockdown of PROM2 enhanced PTX sensitivity in endometrial cancer through modulating the AKT/FOXO1 pathway. This study hinted that PROM2 may be a useful therapeutic target for PTX treatment in endometrial cancer.

Development of machine learning prognostic models for overall survival of prostate cancer patients with lymph node-positive

Prostate cancer (PCa) patients with lymph node involvement (LNI) constitute a single-risk group with varied prognoses. Existing studies on this group have focused solely on those who underwent prostatectomy (RP), using statistical models to predict prognosis. This study aimed to develop an easily accessible individual survival prediction tool based on multiple machine learning (ML) algorithms to predict survival probability for PCa patients with LNI. A total of 3280 PCa patients with LNI were identified from the Surveillance, Epidemiology, and End Results (SEER) database, covering the years 2000–2019. The primary endpoint was overall survival (OS). Gradient Boosting Survival Analysis (GBSA), Random Survival Forest (RSF), and Extra Survival Trees (EST) were used to develop prognosis models, which were compared to Cox regression. Discrimination was evaluated using the time-dependent areas under the receiver operating characteristic curve (time-dependent AUC) and the concordance index (c-index). Calibration was assessed using the time-dependent Brier score (time-dependent BS) and the integrated Brier score (IBS). Moreover, the beeswarm summary plot in SHAP (SHapley Additive exPlanations) was used to display the contribution of variables to the results. The 3280 patients were randomly split into a training cohort (n = 2624) and a validation cohort (n = 656). Nine variables including age at diagnosis, race, marital status, clinical T stage, prostate-specific antigen (PSA) level at diagnosis, Gleason Score (GS), number of positive lymph nodes, radical prostatectomy (RP), and radiotherapy (RT) were used to develop models. The mean time-dependent AUC for GBSA, RSF, and EST was 0.782 (95% confidence interval [CI] 0.779–0.783), 0.779 (95% CI 0.776–0.780), and 0.781 (95% CI 0.778–0.782), respectively, which were higher than the Cox regression model of 0.770 (95% CI 0.769–0.773). Additionally, all models demonstrated almost similar calibration, with low IBS. A web-based prediction tool was developed using the best-performing GBSA, which is accessible at https://pengzihexjtu-pca-n1.streamlit.app/. ML algorithms showed better performance compared with Cox regression and we developed a web-based tool, which may help to guide patient treatment and follow-up.