Accurate Prediction Of Prognosis Research Articles

Prognosis prediction of patients with malignant pleural mesothelioma using conditional variational autoencoder on 3D PET images and clinical data.

Deep learning (DL) has been widely used for diagnosis and prognosis prediction of numerous frequently occurring diseases. Generally, DL models require large datasets to perform accurate and reliable prognosis prediction and avoid overlearning. However, prognosis prediction of rare diseases is still limited owing to the small number of cases, resulting in small datasets. This paper proposes a multimodal DL method to predict the prognosis of patients with malignant pleural mesothelioma (MPM) with a small number of 3D positron emission tomography-computed tomography (PET/CT) images and clinical data. A 3D convolutional conditional variational autoencoder (3D-CCVAE), which adds a 3D-convolutional layer and conditional VAE to process 3D images, was used for dimensionality reduction of PET images. We developed a two-step model that performs dimensionality reduction using the 3D-CCVAE, which is resistant to overlearning. In the first step, clinical data were input to condition the model and perform dimensionality reduction of PET images, resulting in more efficient dimension reduction. In the second step, a subset of the dimensionally reduced features and clinical data were combined to predict 1-year survival of patients using the random forest classifier. To demonstrate the usefulness of the 3D-CCVAE, we created a model without the conditional mechanism (3D-CVAE), one without the variational mechanism (3D-CCAE), and one without an autoencoder (without AE), and compared their prediction results. We used PET images and clinical data of 520 patients with histologically proven MPM. The data were randomly split in a 2:1 ratio (train : test) and three-fold cross-validation was performed. The models were trained on the training set and evaluated based on the test set results. The area under the receiver operating characteristic curve (AUC) for all models was calculated using their 1-year survival predictions, and the results were compared. We obtained AUC values of 0.76 (95% confidence interval [CI], 0.72-0.80) for the 3D-CCVAE model, 0.72 (95% CI, 0.68-0.77) for the 3D-CVAE model, 0.70 (95% CI, 0.66-0.75) for the 3D-CCAE model, and 0.69 (95% CI 0.65-0.74) for the without AE model. The 3D-CCVAE model performed better than the other models (3D-CVAE, p=0.039; 3D-CCAE, p=0.0032; and without AE, p=0.0011). This study demonstrates the usefulness of the 3D-CCVAE in multimodal DL models learned using a small number of datasets. Additionally, it shows that dimensionality reduction via AE can be used to learn a DL model without increasing the overlearning risk. Moreover, the VAE mechanism can overcome the uncertainty of the model parameters that commonly occurs for small datasets, thereby eliminating the risk of overlearning. Additionally, more efficient dimensionality reduction of PET images can be performed by providing clinical data as conditions and ignoring clinical data-related features.

Development and validation of a machine learning-based prognostic risk stratification model for acute ischemic stroke

Acute ischemic stroke (AIS) is a most prevalent cause of serious long-term disability worldwide. Accurate prediction of stroke prognosis is highly valuable for effective intervention and treatment. As such, the present retrospective study aims to provide a reliable machine learning-based model for prognosis prediction in AIS patients. Data from AIS patients were collected retrospectively from the Second Affiliated Hospital of Xuzhou Medical University between August 2017 and July 2019. Independent prognostic factors were identified by univariate and multivariate logistic analysis and used to develop machine learning (ML) models. The ML model performance was assessed by area under the receiver operating characteristic curve (AUC) and radar plot. Shapley Additive explanations (SHAP) values were used to interpret the importance of all features included in the predictive model. A total of 677 AIS patients were included in the present study. Poor prognosis was observed in 209 patients (30.9%). Six variables, including neuron specific enolase (NSE), homocysteine (HCY), S-100β, dysphagia, C-reactive protein (CRP), and anticoagulation were included to establish ML models. Six different ML algorithms were tested, and Random Forest model was selected as the final predictive model with the greatest AUC of 0.908. Moreover, according to SHAP results, NSE impacted the predictive model the most, followed by HCY, S-100β, dysphagia, CRP and anticoagulation. Based on the RF model, an online tool was constructed to predict the prognosis of AIS patients and assist clinicians in optimizing patient treatment. The present study revealed that NSE, HCY, CRP, S-100β, anticoagulation, and dysphagia were important factors for poor prognosis in AIS patients. ML algorithms were used to develop predictive models for predicting the prognosis of AIS patients, with the RF model presenting the optimal performance.

Comprehensive analysis of ferroptosis-related genes in immune infiltration and prognosis in multiple myeloma.

Background: One particular type of cellular death that is known as ferroptosis is caused by the excessive lipid peroxidation. It is a regulated form of cell death that can affect the response of the tumor cells. Currently, it is not known if the presence of this condition can affect the prognosis of patients with multiple myeloma (MM). Methods: In this study, we studied the expression differences and prognostic value of ferroptosis-related genes (FRGs) in MM, and established a ferroptosis risk scoring model. In order to improve the prediction accuracy and clinical applicability, a nomogram was also established. Through gene enrichment analysis, pathways closely related to high-risk groups were identified. We then explored the differences in risk stratification in drug sensitivity and immune patterns, and evaluated their value in prognostic prediction and treatment response. Lastly, we gathered MM cell lines and samples from patients to confirm the expression of marker FRGs using quantitative real-time PCR (qRT-PCR). Results: The ability to predict the survival of MM patients is a challenging issue. Through the use of a risk model derived from ferroptosis, we were able to develop a more accurate prediction of the disease's prognosis. They were then validated by a statistical analysis, which showed that the model is an independent factor in the prognosis of MM. Patients of high ferroptosis risk scores had a much worse chance of survival than those in the low-risk groups. The calibration and power of the nomogram were also strong. We noted that the link between the ferroptosis risk score and the clinical treatment was suggested by the FRG's significant correlation with the immune checkpoint genes and the medication sensitivity. We validated the predictive model using qRT-PCR. Conclusion: We demonstrated the association between FRGs and MM, and developed a new risk model for prognosis in MM patients. Our study sheds light on the potential clinical relevance of ferroptosis in MM and highlights its potential as a therapeutic target for patients with this disease.

Adaptive boosting model for breast cancer prediction

The aggressive form of cancer commonly in breast cells is breast cancer. The highly aggressive form of cancer is frequently created in breast cells. The need for the predictive model to accurately measure the prognosis prediction of breast cancer in the earlier stage is highly recommended. This development of methods for protecting people from fatal diseases by the researchers from the different disciplines who are all working altogether. An accurate breast cancer prognosis prediction is made by using a good predictive model to assist Medical Internet of Things (mIoT). Various advantages such as cancer detection in an earlier stage, medical expenses related to treatment, and having unwanted treatment gives the accurate prediction attains spare patients. Existing models lie on the uni-modal data such as chosen gene expression to predict the model’s design. Few learning-based predictive models are used in the proposed method to improve breast cancer prognosis prediction from the current data sets. Most of the peculiar benefits of the suggested method rely on the model’s architecture. Here, a novel adaptive boosting model (a-BM) is used to measure the loss function of every individual and intends to reduce the error rate. Various performances metrics are used to evaluate the predictive performance, which provides the model gives a good outcome rather than the previous techniques.

Development and validation of a vesicle-mediated transport-associated gene signature for predicting prognosis and immune therapy response in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is a malignant tumor with a poor prognosis. The progression of numerous malignancies has been linked to abnormal vesicle-mediated transport-related gene (VMTRG) expression. The prognostic importance of VMTRGs in HCC is uncertain nonetheless. Utilizing HCC data from TCGA and ICGC, we employed univariate cox analysis, unsupervised clustering, and lasso analysis to construct molecular subtypes and prognostic signature of HCC based on the prognostic-associated VMTRGs expression levels. Subsequently, we validated the expression levels of the signature genes. We investigated the probable pathways using gene set variation analysis (GSVA) and gene set enrichment analysis (GSEA). Six methods were utilized to compare immune cell infiltration between two risk groups. Moreover, the "pRRophetic" algorithm was utilized to test the drug sensitivity of both groups. We identified two distinct subtypes with divergent biological behaviors and immune functionality through unsupervised clustering. Subtype C1 demonstrated a poorer prognosis. A prognostic signature incorporating two VMTRGs (KIF2C and RAC1) was formulated. Immunohistochemistry and qRT-PCR analyses unveiled a significant upregulation of these pivotal genes within HCC tissues. The prognosis was worse for the high-risk group, which also had a higher clinicopathological grade, higher levels of tumor mutation burden (TMB), a higher immunological infiltration of CD8 + T cells, a higher expression of immune checkpoints, and enhanced immunotherapy efficacy. These two risk groups also have varied chemotherapy drug sensitivities. Based on VMTRGs, we have developed a signature that assists in accurate prognosis prediction and formulating personalized treatment strategies for HCC patients.

PTEN-related risk classification models for predicting prognosis and immunotherapy response of hepatocellular carcinoma

IntroductionPTEN often mutates in tumors, and its manipulation is suggested to be used in the development of preclinical tools in cancer research. This study aims to explore the biological impact of gene expression related to PTEN mutations and to develop a prognostic classification model based on the heterogeneity of PTEN expression, and to explore its sensitivity as an indicator of prognosis and molecular and biologic features in hepatocellular carcinoma (HCC).Material and methodsRNA-seq data and mutation data of the LIHC cohort sample downloaded from The Cancer Genome Atlas (TCGA). The HCC samples were grouped according to the mean expression of PTEN, and the tumor microenvironment (TME) was evaluated by ESTIMATE and ssGSEA. The prognostic classification model related to PTEN were constructed by COX and LASSO regression analysis of differentially expressed genes (DEGs) between PTEN-high and -low expressed group.ResultsThe expression of PTEN was affected by copy number variation (CNV) and negatively correlated with immune score, IFNγ score and immune cell infiltration. 1281 DEGs were detected between PTEN-high and PTEN-low expressed group, 8 of the DEGs were finally filtered for developing a prognosis classification model. This model showed better prognostic value than other clinicopathological parameters, and the prediction accuracy of prognosis and ICB treatment for immunotherapy cohorts was better than that of TIDE model.ConclusionsThis study demonstrated the effect of CNV on PTEN expression and the negative immune correlation of PTEN, and constructed a classification model related to the expression of PTEN, which was of guiding significance for evaluating prognostic results of HCC patients and ICB treatment response of cancer immunotherapy cohorts.

An Optimal Prognostic Model Based on Multiparameter Ultrasound for Acute-on-Chronic Liver Failure

Acute-on-chronic liver failure (ACLF) is associated with a considerably high mortality, and accurate prognosis prediction is critical to navigate intervention decisions and improve clinical outcomes. The objective of this study was to establish a better prognostic model for ACLF based on multiparameter ultrasound in combination with clinical features. A total of 149 patients with ACLF were prospectively enrolled and underwent conventional ultrasound, 2-D shear wave elastography (SWE), attenuation imaging, color Doppler sonography, superb microvascular imaging and contrast-enhanced ultrasound (CEUS). Univariate and multivariate analyses were performed to identify independent ultrasound signatures for the prognosis of ACLF, which, when integrated with clinical characteristics, were used to establish a prognostic model. Hepatic perfusion features of CEUS differed significantly between the poor and good prognosis groups, among which the time interval (TI) between peak portal vein (PV) velocity and liver parenchyma (LP) enhancement, TI(PV, LP), was independently associated with the prognosis of ACLF. A prediction model comprising TI(PV, LP) and the international normalized ratio was established, and the area under the curve (AUC) was 0.851, which is greater than those of the Model for End-stage Liver Disease (0.785), fall time of LP model (0.754), 2-D SWE nomogram (0.708) and TI(PV, LP) (0.352). Furthermore, the performance of the model was verified in an independent validation cohort (AUC=0.920). The newly developed model performs better than existing tested models; thus, it has potential as a better non-invasive model for predicting the prognosis of patients with ACLF. A future multicenter, large-sample study is required to validate the performance of this model.

Molecular subtypes of lung adenocarcinoma patients for prognosis and therapeutic response prediction with machine learning on 13 programmed cell death patterns.

Lung adenocarcinoma (LUAD) seriously threatens people's health worldwide. Programmed cell death (PCD) plays a critical role in regulating LUAD growth and metastasis as well as in therapeutic response. However, currently, there is a lack of integrative analysis of PCD-related signatures of LUAD for accurate prediction of prognosis and therapeutic response. The bulk transcriptome and clinical information of LUAD were obtained from TCGA and GEO databases. A total of 1382 genes involved in regulating 13 various PCD patterns (apoptosis, necroptosis, pyroptosis, ferroptosis, cuproptosis, netotic cell death, entotic cell death, lysosome-dependent cell death, parthanatos, autophagy-dependent cell death, oxeiptosis, alkaliptosis and disulfidptosis) were included in the study. Weighted gene co-expression network analysis (WGCNA) and differential expression analysis were performed to identify PCD-associated differential expression genes (DEGs). An unsupervised consensus clustering algorithm was used to explore the potential subtypes of LUAD based on the expression profiles of PCD-associated DEGs. Univariate Cox regression analysis, Least Absolute Shrinkage and Selection Operator (LASSO) regression, Random Forest (RF) analysis and stepwise multivariate Cox analysis were performed to construct a prognostic gene signature. The "oncoPredict" algorithm was utilized for drug-sensitive analysis. GSVA and GSEA were utilized to perform function enrichment analysis. MCPcounter, quanTIseq, Xcell and ssGSEA algorithms were used for tumor immune microenvironment analysis. A nomogram incorporating PCDI and clinicopathological characteristics was established to predict the prognosis of LUAD patients. Forty PCD-associated DEGs related to LUAD were obtained by WGCNA analysis and differential expression analysis, followed by unsupervised clustering to identify two LUAD molecular subtypes. A programmed cell death index (PCDI) with a five-gene signature was established by machine learning algorithms. LUAD patients were then divided into a high PCDI group and a low PCDI group using the median PCDI as a cutoff. Survival and therapeutic analysis revealed that the high PCDI group had a poor prognosis and was more sensitive to targeted drugs but less sensitive to immunotherapy compared to the low PCDI group. Further enrichment analysis showed that B cell-related pathways were significantly downregulated in the high PCDI group. Accordingly, the decreased tumor immune cell infiltration and the lower tumor tertiary lymphoid structure (TLS) scores were also found in the high PCDI group. Finally, a nomogram with reliable predictive performance PCDI was constructed by incorporating PCDI and clinicopathological characteristics, and a user-friendly online website was established for clinical reference ( https://nomogramiv.shinyapps.io/NomogramPCDI/ ). We performed the first comprehensive analysis of the clinical relevance of genes regulating 13 PCD patterns in LUAD and identified two LUAD molecular subtypes with distinct PCD-related gene signature which indicated differential prognosis and treatment sensitivity. Our study provided a new index to predict the efficacy of therapeutic interventions and the prognosis of LUAD patients for guiding personalized treatments.

History of Various Classifications of Gastritis

Gastritis is common worldwide. The combination of Helicobacter pylori (H. pylori) infection with background gastritis, including atrophic gastritis or intestinal metaplasia is implicated as an important etiopathogenetic contributor to gastric cancer. Since the gastritis classification proposed by Schindler, research has focused on classification of gastritis for accurate diagnosis and prediction of prognosis. Advances in endoscopic technology have enabled more accurate visualization of the gastric mucosa and ‘targeted’ biopsies with the emergence of newer classifications based on visual findings (Kimura-Takemoto classification) and more specific histopathological findings using targeted biopsies (Whitehead classification). Following the discovery of H. pylori, it is mandatory to consider its role as an important contributor to gastritis. Therefore, it was necessary to redefine the classification of gastritis and arrive at a consensus, which led to the establishment of an international consensus classification, referred to as the Sydney system. However, the Sydney system alone cannot predict the gastric cancer risk, and scoring systems such as the Operative Link for Gastritis Assessment and the Operative Link on Gastritis Assessment based on Intestinal Metaplasia were proposed. These systems are based on histopathological findings observed in endoscopic biopsy specimens. However, availability of high-definition images following technological advances has facilitated the emergence of a visual classification, the Kyoto classification. In contrast to the Sydney system, the Kyoto classification is based exclusively on interpretation of visual findings and focuses on detection of H. pylori infection and gastric cancer prediction. In this review, we summarize the history and background of the various classifications of gastritis.

Machine learning-based models to predict one-year mortality among Chinese older patients with coronary artery disease combined with impaired glucose tolerance or diabetes mellitus

PurposeAn accurate prediction of survival prognosis is beneficial to guide clinical decision-making. This prospective study aimed to develop a model to predict one-year mortality among older patients with coronary artery disease (CAD) combined with impaired glucose tolerance (IGT) or diabetes mellitus (DM) using machine learning techniques.MethodsA total of 451 patients with CAD combined with IGT and DM were finally enrolled, and those patients randomly split 70:30 into training cohort (n = 308) and validation cohort (n = 143).ResultsThe one-year mortality was 26.83%. The least absolute shrinkage and selection operator (LASSO) method and ten-fold cross-validation identified that seven characteristics were significantly associated with one-year mortality with creatine, N-terminal pro-B-type natriuretic peptide (NT-proBNP), and chronic heart failure being risk factors and hemoglobin, high density lipoprotein cholesterol, albumin, and statins being protective factors. The gradient boosting machine model outperformed other models in terms of Brier score (0.114) and area under the curve (0.836). The gradient boosting machine model also showed favorable calibration and clinical usefulness based on calibration curve and clinical decision curve. The Shapley Additive exPlanations (SHAP) found that the top three features associated with one-year mortality were NT-proBNP, albumin, and statins. The web-based application could be available at https://starxueshu-online-application1-year-mortality-main-49cye8.streamlitapp.com/.ConclusionsThis study proposes an accurate model to stratify patients with a high risk of one-year mortality. The gradient boosting machine model demonstrates promising prediction performance. Some interventions to affect NT-proBNP and albumin levels, and statins, are beneficial to improve survival outcome among patients with CAD combined with IGT or DM.

Systemic Immune-Inflammation Index Predicts Restenosis after Interventions for Lower Extremity Arteriosclerosis Obliterans.

To investigate the association between the pretreatment systemic immune-inflammation index (SII) and restenosis after interventions for lower extremity arteriosclerosis obliterans (ASO). We retrospectively evaluated 309 patients with ASO who underwent endovascular interventions between January 2018 and December 2021. Pretreatment inflammatory markers, including the SII, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), systemic inflammation response index (SIRI), aggregate index of systemic inflammation (AISI), and C-reactive protein (CRP) were collected. The logistic regression model was used to determine the associations between these inflammatory markers and restenosis. Clinical manifestations, ankle-brachial index (ABI), and quality of life after intervention also were compared. The pretreatment SII (p < 0.001), NLR (p < 0.001), PLR (p < 0.001), SIRI (p = 0.002), AISI (p < 0.001), and CRP (p = 0.036) were significantly higher in patients with restenosis than in those without restenosis. Among the four markers, SII had the highest area under the curve (AUC) in predicting restenosis (SII vs. NLR vs. PLR vs. SIRI vs. AISI vs. CRP: 0.715 vs. 0.689 vs. 0.695 vs. 0.643 vs. 0.691 vs. 0.596). Multivariate analysis revealed that the pretreatment SII was the only independent factor for restenosis (hazard ratio [HR]: 4.102; 95% confidence interval [CI]: 1.155-14.567; p = 0.029). Moreover, a lower SII was associated with significantly better improvements in clinical manifestations (Rutherford classification 1-2: 67.5% vs. 52.9%, p = 0.038) and ABI (median: 0.29 vs. 0.22; p = 0.029), together with better quality of life (p < 0.05 for physical functioning, social functioning, pain, and mental health). The pretreatment SII is an independent predictor of restenosis after interventions in patients with lower extremity ASO, providing more accurate prognosis prediction than other inflammatory markers.

CSRP2 transcript levels after consolidation therapy increase prognostic prediction ability in B-cell acute lymphoblastic leukaemia.

Quantification of measurable residual disease (MRD) correlates with the risk of leukemia recurrence in adults with B-cell acute lymphoblastic leukemia (ALL). However, it remains unknown whether collecting data on cysteine and glycine-rich protein 2 (CSRP2) transcript levels, after completing the second course of consolidation, improves prognosis prediction accuracy. A total of 204 subjects with B-cell ALL were tested for CSPR2 transcripts after completing the second course of consolidation using quantitative real-time polymerase chain reaction (qRT-PCR) and divided into high (N = 32) and low (N = 172) CSRP2 expression cohorts. In multivariable analyses, subjects with high expression of CSRP2 had a higher 5-year cumulative incidence of relapse (CIR) (hazard ratio [HR] = 2.57, 95% confidence interval [CI] 1.38-4.76; P = 0.003), lower 5-year relapse-free survival (RFS) (HR = 3.22, 95% CI 1.75-5.93; P < 0.001), and overall survival (OS) (HR = 4.59, 95% CI 2.64-7.99; P < 0.001) in the whole cohort, as well as in the multi-parameter flow cytometry (MPFC) MRD-negative cohort (for CIR, HR = 2.70, 95% CI 1.19-6.12; for RFS, HR = 4.37, 95% CI 1.94-9.85; for OS, HR = 4.90, 95% CI 2.43-9.90; all P < 0.05). Prognostic analysis showed that allogeneic hematopoietic stem cell transplantation (allo-HSCT) could significantly improve the prognosis of patients with high CSRP2 expression (allo-HSCT vs chemotherapy: 5-year CIR, 52% vs 91%; RFS, 41% vs 9%; OS, 38% vs 20%; all P< 0.05). Our data indicate that incorporating data from CSPR2 transcript levels to the MRD-testing at the end of the second course of consolidation therapy enhances prognosis prediction accuracy in adults with B-cell ALL.

Computed tomography-based radiomic analysis for predicting pathological response and prognosis after neoadjuvant chemotherapy in patients with locally advanced esophageal cancer.

Accurate prediction of prognosis and pathological response to neoadjuvant chemotherapy (NAC) is crucial for optimizing treatment strategies for patients with locally advanced esophageal cancer (LA-EC). This study aimed to investigate the use of radiomics for pretreatment CT in predicting the pathological response of patients with LA-EC to NAC. Overall, 144 patients (145 lesions) with LA-EC who underwent pretreatment contrast-enhanced CT and then received NAC followed by surgery with pathological tumor regression grade (TRG) analysis were enrolled. The obtained dataset was randomly divided into training and validation cohorts using fivefold cross-validation. CT-based radiomic features were extracted followed by the feature selection process using the variance threshold, SelectKBest, and least absolute shrinkage and selection operator methods. The radiomic model was constructed using six machine learning classifiers, and predictive performance was evaluated using ROC curve analysis in the training and validation cohorts. All patients were divided into responders (n = 40, 28%) and non-responders (n = 104, 72%) based on the TRG results and a statistically significant split by overall survival analysis (0.899 [0.754-0.961] vs. 0.630 [0.510-0.729], respectively). There were no significant differences between responders and non-responders in terms of age, sex, tumor size, tumor location, or histopathology. The mean AUC of fivefold in the validation cohort was 0.720 (confidence interval [CI]: 0.594-0.982), and the best AUC of the radiomic model using logistic regression to predict the non-responders was 0.815 (CI: 0.626-1.000, sensitivity 0.620, specificity 0.860). A radiomic model derived from contrast-enhanced CT may help stratify chemotherapy effect prediction and improve clinical decision-making.

Can peritumoral edema evaluated by Magnetic Resonance Imaging before neoadjuvant chemotherapy predict complete pathological response in breast cancer?

The complete pathological response (pCR) following neoadjuvant chemotherapy (NAC) in breast cancer is essential for the accurate prediction of prognosis. We aimed to evaluate the efficacy of the presence and type of peritumoral edema detected by magnetic resonance imaging (MRI) in predicting pCR to NAC in breast cancer patients. One hundred five patients with the diagnosis of invasive carcinoma were evaluated by MRI before NAC. Edema was evaluated in fat-suppressed T2-weighted images. The patients were categorized into three groups: patients with no peritumoral edema, patients with peritumoral edema, and patients demonstrating subcutaneous edema. The cases were categorized as being pCR and non-pCR. Molecular subtypes, lymphovascular invasion (LVI), tumor size, and apparent diffusion coefficient (ADC) were evaluated. A positive relationship was found between the presence of edema and tumor size. Subcutaneous edema was found to be statistically higher in non-pCR patients. While the number of pCR patients with subcutaneous edema was 17 (30.4%), the number of non-pCR patients with subcutaneous edema was 26 (53.1%) (p = 0.018). LVI was found to be statistically higher in patients with edema. The number of edema-negative and LVI (+) patients was 4 (15.4%), while the number of edema-positive and LVI (+) patients was 28 (35.4%) (p = 0.042). Intratumoral and peritumoral ADC values were significantly higher in tumors with edema. The presence of subcutaneous edema and LVI may be utilized for the prediction of pCR outcomes in breast cancer patients scheduled for NAC treatment.

A multiparameter radiomic model for accurate prognostic prediction of glioma

AbstractAn accurate prediction of prognosis is important for clinical treatments of glioma. In this study, a multiparameter radiomic model is proposed for accurate prognostic prediction of glioma. Three kinds of region of interest were extracted from preoperative postcontrast T1‐weighted images and T2 fluid‐attenuated inversion recovery images acquired from 140 glioma patients. Radiomics score (Radscore) was calculated and the conventional image features and clinical molecular characteristics that may be related to progression‐free survival (PFS) were evaluated. Five uniparameter and various combinations of biparameter and multiparameter models based on above characteristics were built. The performance of these models was evaluated by concordance index (C index), and the nomogram of the multiparameter radiomic model was constructed. The results show that the proposed multiparameter radiomic model has a better prediction performance than other models. In the training and validation sets, the calibration curves of the multiparameter radiomic model for the 1‐, 2‐, and 3‐year PFS probability demonstrate a high consistence between predictions and observations. In conclusion, this study demonstrates that the multiparameter radiomic model based on Radscore, conventional image features and clinical molecular characteristics can improve the prediction accuracy of glioma prognosis, which could be informative for individualized treatments.

Prevalence and Prognostic Significance of Malnutrition in Patients with Brain Metastasis.

Malnutrition is a severe but modifiable risk factor for cancers. However, the relationship between malnutrition and the survival of patients with brain metastases has not been fully revealed. We aimed to evaluate the prevalence of malnutrition and assess its prognostic value on patients with brain metastases. We retrospectively recruited 2,633 patients with brain metastases between January 2014 and September 2020. Three malnutrition scores were used to evaluate patients' malnutrition status at their first admission, including controlling nutritional status, the nutritional risk index, and the prognostic nutritional index. The association between malnutrition and overall survival (OS) was estimated. The three malnutrition scores were associated with each other and with body mass index (BMI). Malnutrition assessed by any of the three scores was significantly associated with poor OS. All three malnutrition scores were better indicators than BMI, and adding malnutrition scores to the Graded Prognostic Assessment (GPA) scoring system could significantly improve the accuracy of prognosis prediction. Malnutrition monitoring using any of the three malnutrition scores on patients' first admission could be a better survival indicator for patients with brain metastases compared with BMI alone. Malnutrition is a more significant indicator of survival stratification compared with BMI. Adding malnutrition to the GPA score system achieves better survival prediction.

Abstract 5562: Prevalence and prognostic significance of malnutrition in patients with brain metastasis

Abstract Background: Malnutrition is a severe but modifiable risk factor for cancers. However, the relationship between malnutrition and the survival of patients with brain metastases has not been fully revealed. We aimed to evaluate the prevalence of malnutrition and assess its prognostic value on patients with brain metastases. Method: We retrospectively recruited 2633 patients with brain metastases between January 2014 and September 2020. Three malnutrition scores were used to evaluate patients' malnutrition status at their first admission, including controlling nutritional status (CONUT), the nutritional risk index (NRI), and the prognostic nutritional index (PNI). The association between malnutrition and overall survival was estimated. Results: A total of 1984 (75.4%) of patients were assessed as any degree malnourished and 963 (36.6%) of patients were assessed as moderate to severe malnutrition by at least one malnutrition score. The three malnutrition scores were associated with each other and with BMI. Malnutrition assessed by any of the three scores was significantly associated with poor overall survival (adjust hazard ratio [HR] and 95% confidence interval [CI] for CONUT: 1.12 [1.10-1.16]; for NRI: 0.97 [0.97-0.98]; for PNI: 0.96 [0.95-0.97], all p values &amp;lt;0.001). Furthermore, malnutrition was a better indicator than BMI, and adding malnutrition to the Graded Prognostic Assessment scoring system could significantly improve the accuracy of prognosis prediction. Conclusion: Malnutrition was prevalent in patients with brain metastases and was significantly associated with overall survival. Malnutrition monitoring on patients' first admission could improve the survival prediction ability for brain metastasis. Citation Format: Zheran Liu, Yu Zhang, Yiyan Pei, Yan He, Jiayi Yu, Renjie Zhang, Jingjing Wang, Weelic Chong, Yang Hai, Xingchen Peng, Fang Fang. Prevalence and prognostic significance of malnutrition in patients with brain metastasis. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5562.

Integrative bioinformatics approaches to establish potential prognostic immune-related genes signature and drugs in the non-small cell lung cancer microenvironment.

Introduction: Research has revealed that the tumor microenvironment (TME) is associated with the progression of malignancy. The combination of meaningful prognostic biomarkers related to the TME is expected to be a reliable direction for improving the diagnosis and treatment of non-small cell lung cancer (NSCLC). Method and Result: Therefore, to better understand the connection between the TME and survival outcomes of NSCLC, we used the "DESeq2" R package to mine the differentially expressed genes (DEGs) of two groups of NSCLC samples according to the optimal cutoff value of the immune score through the ESTIMATE algorithm. A total of 978 up-DEGs and 828 down-DEGs were eventually identified. A fifteen-gene prognostic signature was established via LASSO and Cox regression analysis and further divided the patients into two risk sets. The survival outcome of high-risk patients was significantly worse than that of low-risk patients in both the TCGA and two external validation sets (p-value < 0.05). The gene signature showed high predictive accuracy in TCGA (1-year area under the time-dependent ROC curve (AUC) = 0.722, 2-year AUC = 0.708, 3-year AUC = 0.686). The nomogram comprised of the risk score and related clinicopathological information was constructed, and calibration plots and ROC curves were applied, KEGG and GSEA analyses showed that the epithelial-mesenchymal transition (EMT) pathway, E2F target pathway and immune-associated pathway were mainly involved in the high-risk group. Further somatic mutation and immune analyses were conducted to compare the differences between the two groups. Drug sensitivity provides a potential treatment basis for clinical treatment. Finally, EREG and ADH1C were selected as the key prognostic genes of the two overlapping results from PPI and multiple Cox analyses. They were verified by comparing the mRNA expression in cell lines and protein expression in the HPA database, and clinical validation further confirmed the effectiveness of key genes. Conclusion: In conclusion, we obtained an immune-related fifteen-gene prognostic signature and potential mechanism and sensitive drugs underling the prognosis model, which may provide accurate prognosis prediction and available strategies for NSCLC.

An immune checkpoint-based signature predicts prognosis and chemotherapy response for patients with small cell lung cancer.

Therapeutic options for small cell lung cancer (SCLC), a particularly lethal malignancy, remain limited. Members of the B7-CD28 family are compelling targets for immune checkpoint blockade strategies, which involve activating, inhibiting, and fine-tuning the T cell immune response. However, their clinical features and significance have not been explored comprehensively. We enrolled 228 patients with an initial diagnosis of SCLC, including 77 cases from Cbioportal and a validation cohort of 151 cases with qPCR data. Kaplan-Meier analysis and LASSO Cox model were used to identify a signature based on the B7-CD28 family, which was applied for accurate prediction of chemotherapy benefit and prognosis for SCLC patients. In addition, we applied bioinformatics analysis to explore potential signature-related molecular mechanisms and the immune landscape. The mutation profiles of healthy tissues and SCLC tissues were distinct. A signature consisting of seven genes (CD86, ICOSLG, CD276, CD28, CTLA-4, PDCD1, and TMIGD2) was identified and applied to group patients based on risk level (high-risk and low-risk), producing two groups for which survival outcomes differed significantly (HR=3.81, 95% CI: 2.16-6.74, P<0.001). The immune checkpoint-based signature accurately predicted patient outcomes for the selected training and validation sets. Notably, low-risk patients were more likely to benefit from chemotherapy and showed greater immune activation. Additionally, time-dependent ROC curves and C-index analysis confirmed that the immune checkpoint-based signature has excellent predictive power for prognosis and chemotherapy benefit compared to clinically recognized parameters. Finally, multivariate analysis confirmed the identified signature as an independent risk factor for prognosis and chemotherapeutic response. We systematically obtained a comprehensive molecular profile for B7-CD28 family members in SCLC patients, from which we produced a reliable and robust prognostic immune checkpoint-based signature with the potential to improve prognostic stratification and therapy strategies for SCLC patients.

A novel lysosome-related gene signature coupled with gleason score for prognosis prediction in prostate cancer.

Background: Prostate cancer (PCa) is highly heterogeneous, which makes it difficult to precisely distinguish the clinical stages and histological grades of tumor lesions, thereby leading to large amounts of under- and over-treatment. Thus, we expect the development of novel prediction approaches for the prevention of inadequate therapies. The emerging evidence demonstrates the pivotal role of lysosome-related mechanisms in the prognosis of PCa. In this study, we aimed to identify a lysosome-related prognostic predictor in PCa for future therapies. Methods: The PCa samples involved in this study were gathered from The Cancer Genome Atlas database (TCGA) (n = 552) and cBioPortal database (n = 82). During screening, we categorized PCa patients into two immune groups based on median ssGSEA scores. Then, the Gleason score and lysosome-related genes were included and screened out by using a univariate Cox regression analysis and the least absolute shrinkage and selection operation (LASSO) analysis. Following further analysis, the probability of progression free interval (PFI) was modeled by using unadjusted Kaplan-Meier estimation curves and a multivariable Cox regression analysis. A receiver operating characteristic (ROC) curve, nomogram and calibration curve were used to examine the predictive value of this model in discriminating progression events from non-events. The model was trained and repeatedly validated by creating a training set (n = 400), an internal validation set (n = 100) and an external validation (n = 82) from the cohort. Results: Following grouping by ssGSEA score, the Gleason score and two LRGs-neutrophil cytosolic factor 1 (NCF1) and gamma-interferon-inducible lysosomal thiol reductase (IFI30)-were screened out to differentiate patients with or without progression (1-year AUC = 0.787; 3-year AUC = 0.798; 5-year AUC = 0.772; 10-year AUC = 0.832). Patients with a higher risk showed poorer outcomes (p < 0.0001) and a higher cumulative hazard (p < 0.0001). Besides this, our risk model combined LRGs with the Gleason score and presented a more accurate prediction of PCa prognosis than the Gleason score alone. In three validation sets, our model still achieved high prediction rates. Conclusion: In conclusion, this novel lysosome-related gene signature, coupled with the Gleason score, works well in PCa for prognosis prediction.