Siga este enlace para ver otros tipos de publicaciones sobre el tema: ML prognostic model.
Artículos de revistas sobre el tema "ML prognostic model"
Crea una cita precisa en los estilos APA, MLA, Chicago, Harvard y otros
Consulte los 50 mejores artículos de revistas para su investigación sobre el tema "ML prognostic model".
Junto a cada fuente en la lista de referencias hay un botón "Agregar a la bibliografía". Pulsa este botón, y generaremos automáticamente la referencia bibliográfica para la obra elegida en el estilo de cita que necesites: APA, MLA, Harvard, Vancouver, Chicago, etc.
También puede descargar el texto completo de la publicación académica en formato pdf y leer en línea su resumen siempre que esté disponible en los metadatos.
Explore artículos de revistas sobre una amplia variedad de disciplinas y organice su bibliografía correctamente.
1
Uneno, Yu, Tadayuki Kou, Masashi Kanai, et al. "Prognostic model for survival in patients with advanced pancreatic cancer receiving palliative chemotherapy." Journal of Clinical Oncology 33, no. 3_suppl (2015): 248. http://dx.doi.org/10.1200/jco.2015.33.3_suppl.248.
Texto completoResumen
248 Background: The prognosis of patients with advanced pancreatic cancer (APC) is extremely poor. Several clinical and laboratory factors have been known to be associated with prognosis of APC patients. However, there are few clinically available prognostic models predicting survival in APC patients receiving palliative chemotherapy. Methods: To construct a prognostic model to predict survival in APC patients receiving palliative chemotherapy, we analyzed the clinical data from 306 consecutive patients with pathologically confirmed APC who received palliative chemotherapy. We selected six independent prognostic factors which remained independent prognostic factors after multivariate analysis. Thereafter, we rounded the regression coefficient (β) for each independent prognostic factor derived from the Cox regression equation (HR = eβ) and developed a prognostic index (PI). Results: Developed prognostic index (PI) was as follows: PI = 2 (if performance status score 2–3) + 1 (if metastatic disease) + 1 (if initially unresectable disease) + 1 (if carcinoembryonic antigen level ≥5.0 ng/ml) + 1 (if carbohydrate antigen 19-9 level ≥1000 U/ml) + 2 (if neutrophil–lymphocyte ratio ≥5). The patients were classified into three prognostic groups: favorable (PI 0–1, n = 73), intermediate (PI 2–3, n = 145), and poor prognosis (PI 4–8, n = 88). The median overall survival for each prognostic group was 16.5, 12.3 and 6.2 months, respectively, and the 1-year survival rates were 67.3%, 51.3%, and 19.1%, respectively (P < 0.01). The c index of the model was 0.658. This model was well calibrated to predict 1-year survival, in which overestimation (2.4% and 0.2% in the favorable and poor prognosis groups, respectively) and underestimation (3.6% in the intermediate prognosis group) were observed. Conclusions: This prognostic model based on readily available clinical factors would help clinicians in estimating the overall survival in APC patients receiving palliative chemotherapy.
2
Shen, Ziyuan, Shuo Zhang, Yaxue Jiao, et al. "LASSO Model Better Predicted the Prognosis of DLBCL than Random Forest Model: A Retrospective Multicenter Analysis of HHLWG." Journal of Oncology 2022 (September 16, 2022): 1–10. http://dx.doi.org/10.1155/2022/1618272.
Texto completoResumen
Background. Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous non-Hodgkin’s lymphoma with great clinical challenge. Machine learning (ML) has attracted substantial attention in diagnosis, prognosis, and treatment of diseases. This study is aimed at exploring the prognostic factors of DLBCL by ML. Methods. In total, 1211 DLBCL patients were retrieved from Huaihai Lymphoma Working Group (HHLWG). The least absolute shrinkage and selection operator (LASSO) and random forest algorithm were used to identify prognostic factors for the overall survival (OS) rate of DLBCL among twenty-five variables. Receiver operating characteristic (ROC) curve and decision curve analysis (DCA) were utilized to compare the predictive performance and clinical effectiveness of the two models, respectively. Results. The median follow-up time was 43.4 months, and the 5-year OS was 58.5%. The LASSO model achieved an Area under the curve (AUC) of 75.8% for the prognosis of DLBCL, which was higher than that of the random forest model (AUC: 71.6%). DCA analysis also revealed that the LASSO model could augment net benefits and exhibited a wider range of threshold probabilities by risk stratification than the random forest model. In addition, multivariable analysis demonstrated that age, white blood cell count, hemoglobin, central nervous system involvement, gender, and Ann Arbor stage were independent prognostic factors for DLBCL. The LASSO model showed better discrimination of outcomes compared with the IPI and NCCN-IPI models and identified three groups of patients: low risk, high-intermediate risk, and high risk. Conclusions. The prognostic model of DLBCL based on the LASSO regression was more accurate than the random forest, IPI, and NCCN-IPI models.
3
Qin, Yuchao, Ahmed Alaa, Andres Floto, and Mihaela van der Schaar. "External validity of machine learning-based prognostic scores for cystic fibrosis: A retrospective study using the UK and Canadian registries." PLOS Digital Health 2, no. 1 (2023): e0000179. http://dx.doi.org/10.1371/journal.pdig.0000179.
Texto completoResumen
Precise and timely referral for lung transplantation is critical for the survival of cystic fibrosis patients with terminal illness. While machine learning (ML) models have been shown to achieve significant improvement in prognostic accuracy over current referral guidelines, the external validity of these models and their resulting referral policies has not been fully investigated. Here, we studied the external validity of machine learning-based prognostic models using annual follow-up data from the UK and Canadian Cystic Fibrosis Registries. Using a state-of-the-art automated ML framework, we derived a model for predicting poor clinical outcomes in patients enrolled in the UK registry, and conducted external validation of the derived model using the Canadian Cystic Fibrosis Registry. In particular, we studied the effect of (1) natural variations in patient characteristics across populations and (2) differences in clinical practice on the external validity of ML-based prognostic scores. Overall, decrease in prognostic accuracy on the external validation set (AUCROC: 0.88, 95% CI 0.88-0.88) was observed compared to the internal validation accuracy (AUCROC: 0.91, 95% CI 0.90-0.92). Based on our ML model, analysis on feature contributions and risk strata revealed that, while external validation of ML models exhibited high precision on average, both factors (1) and (2) can undermine the external validity of ML models in patient subgroups with moderate risk for poor outcomes. A significant boost in prognostic power (F1 score) from 0.33 (95% CI 0.31-0.35) to 0.45 (95% CI 0.45-0.45) was observed in external validation when variations in these subgroups were accounted in our model. Our study highlighted the significance of external validation of ML models for cystic fibrosis prognostication. The uncovered insights on key risk factors and patient subgroups can be used to guide the cross-population adaptation of ML-based models and inspire new research on applying transfer learning methods for fine-tuning ML models to cope with regional variations in clinical care.
4
Filipow, Nicole, Eleanor Main, Neil J. Sebire, et al. "Implementation of prognostic machine learning algorithms in paediatric chronic respiratory conditions: a scoping review." BMJ Open Respiratory Research 9, no. 1 (2022): e001165. http://dx.doi.org/10.1136/bmjresp-2021-001165.
Texto completoResumen
Machine learning (ML) holds great potential for predicting clinical outcomes in heterogeneous chronic respiratory diseases (CRD) affecting children, where timely individualised treatments offer opportunities for health optimisation. This paper identifies rate-limiting steps in ML prediction model development that impair clinical translation and discusses regulatory, clinical and ethical considerations for ML implementation. A scoping review of ML prediction models in paediatric CRDs was undertaken using the PRISMA extension scoping review guidelines. From 1209 results, 25 articles published between 2013 and 2021 were evaluated for features of a good clinical prediction model using the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) guidelines.Most of the studies were in asthma (80%), with few in cystic fibrosis (12%), bronchiolitis (4%) and childhood wheeze (4%). There were inconsistencies in model reporting and studies were limited by a lack of validation, and absence of equations or code for replication. Clinician involvement during ML model development is essential and diversity, equity and inclusion should be assessed at each step of the ML pipeline to ensure algorithms do not promote or amplify health disparities among marginalised groups. As ML prediction studies become more frequent, it is important that models are rigorously developed using published guidelines and take account of regulatory frameworks which depend on model complexity, patient safety, accountability and liability.
5
Ferroni, Patrizia, Fabio Zanzotto, Silvia Riondino, Noemi Scarpato, Fiorella Guadagni, and Mario Roselli. "Breast Cancer Prognosis Using a Machine Learning Approach." Cancers 11, no. 3 (2019): 328. http://dx.doi.org/10.3390/cancers11030328.
Texto completoResumen
Machine learning (ML) has been recently introduced to develop prognostic classification models that can be used to predict outcomes in individual cancer patients. Here, we report the significance of an ML-based decision support system (DSS), combined with random optimization (RO), to extract prognostic information from routinely collected demographic, clinical and biochemical data of breast cancer (BC) patients. A DSS model was developed in a training set (n = 318), whose performance analysis in the testing set (n = 136) resulted in a C-index for progression-free survival of 0.84, with an accuracy of 86%. Furthermore, the model was capable of stratifying the testing set into two groups of patients with low- or high-risk of progression with a hazard ratio (HR) of 10.9 (p < 0.0001). Validation in multicenter prospective studies and appropriate management of privacy issues in relation to digital electronic health records (EHR) data are presently needed. Nonetheless, we may conclude that the implementation of ML algorithms and RO models into EHR data might help to achieve prognostic information, and has the potential to revolutionize the practice of personalized medicine.
6
Muscas, Giovanni, Tommaso Matteuzzi, Eleonora Becattini, et al. "Development of machine learning models to prognosticate chronic shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage." Acta Neurochirurgica 162, no. 12 (2020): 3093–105. http://dx.doi.org/10.1007/s00701-020-04484-6.
Texto completoResumen
Abstract Background Shunt-dependent hydrocephalus significantly complicates subarachnoid hemorrhage (SAH), and reliable prognosis methods have been sought in recent years to reduce morbidity and costs associated with delayed treatment or neglected onset. Machine learning (ML) defines modern data analysis techniques allowing accurate subject-based risk stratifications. We aimed at developing and testing different ML models to predict shunt-dependent hydrocephalus after aneurysmal SAH. Methods We consulted electronic records of patients with aneurysmal SAH treated at our institution between January 2013 and March 2019. We selected variables for the models according to the results of the previous works on this topic. We trained and tested four ML algorithms on three datasets: one containing binary variables, one considering variables associated with shunt-dependency after an explorative analysis, and one including all variables. For each model, we calculated AUROC, specificity, sensitivity, accuracy, PPV, and also, on the validation set, the NPV and the Matthews correlation coefficient (ϕ). Results Three hundred eighty-six patients were included. Fifty patients (12.9%) developed shunt-dependency after a mean follow-up of 19.7 (± 12.6) months. Complete information was retrieved for 32 variables, used to train the models. The best models were selected based on the performances on the validation set and were achieved with a distributed random forest model considering 21 variables, with a ϕ = 0.59, AUC = 0.88; sensitivity and specificity of 0.73 (C.I.: 0.39–0.94) and 0.92 (C.I.: 0.84–0.97), respectively; PPV = 0.59 (0.38–0.77); and NPV = 0.96 (0.90–0.98). Accuracy was 0.90 (0.82–0.95). Conclusions Machine learning prognostic models allow accurate predictions with a large number of variables and a more subject-oriented prognosis. We identified a single best distributed random forest model, with an excellent prognostic capacity (ϕ = 0.58), which could be especially helpful in identifying low-risk patients for shunt-dependency.
7
Park, Hyung Soon, Ji Soo Park, Yun Ho Roh, Jieun Moon, Dong Sup Yoon, and Hei-Cheul Jeung. "Prognostic factors and scoring model for survival in advanced biliary tract cancer." Journal of Clinical Oncology 35, no. 4_suppl (2017): 264. http://dx.doi.org/10.1200/jco.2017.35.4_suppl.264.
Texto completoResumen
264 Background: Metastatic biliary tract cancer (BTC) has dismal prognosis. We herein presented multivariate analysis using routinely evaluated clinico-laboratory parameters at the time of initial diagnosis, to implement a scoring model that can effectively identify risk groups, and we finally validated the model using independent dataset. Methods: From September 2006 to February 2015, 482 patients with metastatic BTC were analyzed. Patients were randomly assigned (7:3) into investigational (n = 340) and validation dataset (n = 142). Continuous variables were dichotomized according to the normal range or the best cutoff values statistically determined by Contal and O’Quigley method. Multivariate analysis using Cox’s proportional hazard model was done to find independent prognostic factors, and scoring model were derived by summing the rounded χ2 scores for the factors emerged in the multivariate analysis. Results: Performance status (ECOG 3-4), hypoalbuminemia ( < 3.4 mg/dL), carcinoembryonic antigen (≥9 ng/mL), neutrophil-lymphocyte ratio (≥3.0), and carbohydrate antigen 19-9 (≥120 U/mL) were identified as independent factors for poor survival in investigational dataset. When assigning patients into three risk groups based on these factors, survival was 14.0, 7.3, and 2.3 months for the low, intermediate, and high-risk groups, respectively (P < 0.001). Harrell’s C-index and integrated AUC for scoring model were 0.682 and 0.653, respectively. In validation dataset, prognosis was also well-divided according to the risk groups (median OS, 16.7, 7.5 and 1.9 months, respectively, P < 0.001). Chemotherapy gave a survival benefit in low and intermediate-risk group (11.4 vs. 4.8 months; P< 0.001), but not in high-risk group (median OS, 4.3 vs. 1.1 months; P = 0.105). Conclusions: We propose a set of prognostic criteria for metastatic BTC, which can help accurate patient risk stratification and aid in treatment selection.
8
Hulsbergen, Alexander, Yu Tung Lo, Vasileios Kavouridis, et al. "SURG-02. SURVIVAL PREDICTION AFTER NEUROSURGICAL RESECTION OF BRAIN METASTASES: A MACHINE LEARNING APPROACH." Neuro-Oncology 22, Supplement_2 (2020): ii203. http://dx.doi.org/10.1093/neuonc/noaa215.849.
Texto completoResumen
Abstract INTRODUCTION Survival prediction in brain metastases (BMs) remains challenging. Current prognostic models have been created and validated almost completely with data from patients receiving radiotherapy only, leaving uncertainty about surgical patients. Therefore, the aim of this study was to build and validate a model predicting 6-month survival after BM resection using different machine learning (ML) algorithms. METHODS An institutional database of 1062 patients who underwent resection for BM was split into a 80:20 training and testing set. Seven different ML algorithms were trained and assessed for performance. Moreover, an ensemble model was created incorporating random forest, adaptive boosting, gradient boosting, and logistic regression algorithms. Five-fold cross validation was used for hyperparameter tuning. Model performance was assessed using area under the receiver-operating curve (AUC) and calibration and was compared against the diagnosis-specific graded prognostic assessment (ds-GPA); the most established prognostic model in BMs. RESULTS The ensemble model showed superior performance with an AUC of 0.81 in the hold-out test set, a calibration slope of 1.14, and a calibration intercept of -0.08, outperforming the ds-GPA (AUC 0.68). Patients were stratified into high-, medium- and low-risk groups for death at 6 months; these strata strongly predicted both 6-months and longitudinal overall survival (p &lt; 0.001). CONCLUSIONS We developed and internally validated an ensemble ML model that accurately predicts 6-month survival after neurosurgical resection for BM, outperforms the most established model in the literature, and allows for meaningful risk stratification. Future efforts should focus on external validation of our model.
9
Xiao, Changhu, Yuan Guo, Kaixuan Zhao, et al. "Prognostic Value of Machine Learning in Patients with Acute Myocardial Infarction." Journal of Cardiovascular Development and Disease 9, no. 2 (2022): 56. http://dx.doi.org/10.3390/jcdd9020056.
Texto completoResumen
(1) Background: Patients with acute myocardial infarction (AMI) still experience many major adverse cardiovascular events (MACEs), including myocardial infarction, heart failure, kidney failure, coronary events, cerebrovascular events, and death. This retrospective study aims to assess the prognostic value of machine learning (ML) for the prediction of MACEs. (2) Methods: Five-hundred patients diagnosed with AMI and who had undergone successful percutaneous coronary intervention were included in the study. Logistic regression (LR) analysis was used to assess the relevance of MACEs and 24 selected clinical variables. Six ML models were developed with five-fold cross-validation in the training dataset and their ability to predict MACEs was compared to LR with the testing dataset. (3) Results: The MACE rate was calculated as 30.6% after a mean follow-up of 1.42 years. Killip classification (Killip IV vs. I class, odds ratio 4.386, 95% confidence interval 1.943–9.904), drug compliance (irregular vs. regular compliance, 3.06, 1.721–5.438), age (per year, 1.025, 1.006–1.044), and creatinine (1 µmol/L, 1.007, 1.002–1.012) and cholesterol levels (1 mmol/L, 0.708, 0.556–0.903) were independent predictors of MACEs. In the training dataset, the best performing model was the random forest (RDF) model with an area under the curve of (0.749, 0.644–0.853) and accuracy of (0.734, 0.647–0.820). In the testing dataset, the RDF showed the most significant survival difference (log-rank p = 0.017) in distinguishing patients with and without MACEs. (4) Conclusions: The RDF model has been identified as superior to other models for MACE prediction in this study. ML methods can be promising for improving optimal predictor selection and clinical outcomes in patients with AMI.
10
Dou, Guanhua, Dongkai Shan, Kai Wang, et al. "Integrating Coronary Plaque Information from CCTA by ML Predicts MACE in Patients with Suspected CAD." Journal of Personalized Medicine 12, no. 4 (2022): 596. http://dx.doi.org/10.3390/jpm12040596.
Texto completoResumen
Conventional prognostic risk analysis in patients undergoing noninvasive imaging is based upon a limited selection of clinical and imaging findings, whereas machine learning (ML) algorithms include a greater number and complexity of variables. Therefore, this paper aimed to explore the predictive value of integrating coronary plaque information from coronary computed tomographic angiography (CCTA) with ML to predict major adverse cardiovascular events (MACEs) in patients with suspected coronary artery disease (CAD). Patients who underwent CCTA due to suspected coronary artery disease with a 30-month follow-up for MACEs were included. We collected demographic characteristics, cardiovascular risk factors, and information on coronary plaques by analyzing CCTA information (plaque length, plaque composition and coronary artery stenosis of 18 coronary artery segments, coronary dominance, myocardial bridge (MB), and patients with vulnerable plaque) and follow-up information (cardiac death, nonfatal myocardial infarction and unstable angina requiring hospitalization). An ML algorithm was used for survival analysis (CoxBoost). This analysis showed that chest symptoms, the stenosis severity of the proximal anterior descending branch, and the stenosis severity of the middle right coronary artery were among the top three variables in the ML model. After the 22nd month of follow-up, in the testing dataset, ML showed the largest C-index and AUC compared with Cox regression, SIS, SIS score + clinical factors, and clinical factors. The DCA of all the models showed that the net benefit of the ML model was the highest when the treatment threshold probability was between 1% and 9%. Integrating coronary plaque information from CCTA based on ML technology provides a feasible and superior method to assess prognosis in patients with suspected coronary artery disease over an approximately three-year period.
11
Bel’skaya, L. V., and V. K. Kosenok. "A new field of application of saliva tests for prognostic purpose: focus on lung cancer." Biomedical Chemistry: Research and Methods 3, no. 3 (2020): e00133. http://dx.doi.org/10.18097/bmcrm00133.
Texto completoResumen
The aim of this work was to determine the potential prognostic role of the biochemical parameters of saliva in lung cancer. The study included 425 patients with lung cancer of various histological types. Saliva samples were collected before treatment and 34 biochemical parameters were determined. Prognostic factors were analyzed by multivariate analysis using Cox’s proportional hazard model. Results have shown that for squamous cell carcinoma, LDH activity below 1748 U/L was an independent prognostically unfavorable factor (HR = 2.89; 95% CI 1.28 – 6.46; р = 0.00330). For adenocarcinoma, there was a combination of factors: the content of imidazole compounds above 0.311 mmol/L, seromucoids below 0.098 c.u. and uric acid less than 83 nmol/mL (HR = 7.91; 95% CI 2.52 – 24.11; р = 0.00004). For neuroendocrine lung cancer, an unfavorable prognosis was associated with the urea content below 8 mmol/L, NO below 24 nmol/mL and ALP activity below 74 E/L (HR = 12.50; 95% CI 1.09 – 138.7; р = 0.01184). For radical surgery, an unfavorable prognosis was associated with LDH activity below 545 U/L (HR = 7.20; 95% CI 2.13 – 23.70; p = 0.00968), for combined treatment, with the NO level below 15 nmol/mL (HR = 5.64; 95% CI 1.89 – 16.46; p = 0.02797). The worst prognosis for palliative treatment was observed at the NO level below 24 nmol/mL and the imidazole compound content above 0.311 mmol/L (HR = 2.73; 95% CI 1.07 – 12.92; р = 0.01827). The identified parameters can be used to predict lung cancer and personalized patient’s treatment.
12
Nadali, Gianpaolo, Luisa Tavecchia, Elisabetta Zanolin, et al. "Serum Level of the Soluble Form of the CD30 Molecule Identifies Patients With Hodgkin's Disease at High Risk of Unfavorable Outcome." Blood 91, no. 8 (1998): 3011–16. http://dx.doi.org/10.1182/blood.v91.8.3011.3011_3011_3016.
Texto completoResumen
Preliminary reports suggested a prognostic significance for serum levels of soluble CD30 (sCD30) in patients with Hodgkin's disease (HD). In this study, we investigated the prognostic impact of sCD30 concentration at diagnosis in relation to the other recognized prognostic parameters in 303 patients with HD observed in three different institutions between 1984 and 1996. sCD30 levels were correlated with stage, presence of B symptoms, and tumor burden. High sCD30 levels entailed a higher risk of poor outcome, and the event-free survival (EFS) probability at 5 years for patients with sCD30 levels ≥100 and less than 100 U/mL was 59.9% (95% confidence interval [CI], 40.6% to 65.9%) and 87.5% (95% CI, 81.5% to 91.6%), respectively (P &lt; .001). On the basis of the results of univariate analysis of 14 pretreatment characteristics, we included five prognostic factors (high sCD30 serum level, stage III-IV, B symptoms, low hemoglobin level, and age ≥50 years) into a multivariate model. High sCD30 and advanced stage were independently associated with an unfavorable prognosis. Their combined evaluation identified patients at high risk (stages III and IV and sCD30 ≥100 U/mL: EFS, 46.9%) and low risk (stages I and II with sCD30 &lt;100 U/mL: EFS, 88.7%) of treatment failure (P &lt; .001). We conclude that the combined evaluation of sCD30 serum level and stage at presentation identifies patients with HD at high risk of an unfavorable outcome.
13
Andaur Navarro, Constanza L., Johanna A. A. G. Damen, Toshihiko Takada, et al. "Protocol for a systematic review on the methodological and reporting quality of prediction model studies using machine learning techniques." BMJ Open 10, no. 11 (2020): e038832. http://dx.doi.org/10.1136/bmjopen-2020-038832.
Texto completoResumen
IntroductionStudies addressing the development and/or validation of diagnostic and prognostic prediction models are abundant in most clinical domains. Systematic reviews have shown that the methodological and reporting quality of prediction model studies is suboptimal. Due to the increasing availability of larger, routinely collected and complex medical data, and the rising application of Artificial Intelligence (AI) or machine learning (ML) techniques, the number of prediction model studies is expected to increase even further. Prediction models developed using AI or ML techniques are often labelled as a ‘black box’ and little is known about their methodological and reporting quality. Therefore, this comprehensive systematic review aims to evaluate the reporting quality, the methodological conduct, and the risk of bias of prediction model studies that applied ML techniques for model development and/or validation.Methods and analysisA search will be performed in PubMed to identify studies developing and/or validating prediction models using any ML methodology and across all medical fields. Studies will be included if they were published between January 2018 and December 2019, predict patient-related outcomes, use any study design or data source, and available in English. Screening of search results and data extraction from included articles will be performed by two independent reviewers. The primary outcomes of this systematic review are: (1) the adherence of ML-based prediction model studies to the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD), and (2) the risk of bias in such studies as assessed using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). A narrative synthesis will be conducted for all included studies. Findings will be stratified by study type, medical field and prevalent ML methods, and will inform necessary extensions or updates of TRIPOD and PROBAST to better address prediction model studies that used AI or ML techniques.Ethics and disseminationEthical approval is not required for this study because only available published data will be analysed. Findings will be disseminated through peer-reviewed publications and scientific conferences.Systematic review registrationPROSPERO, CRD42019161764.
14
Nuñez-Garcia, Jean C., Antonio Sánchez-Puente, Jesús Sampedro-Gómez, et al. "Outcome Analysis in Elective Electrical Cardioversion of Atrial Fibrillation Patients: Development and Validation of a Machine Learning Prognostic Model." Journal of Clinical Medicine 11, no. 9 (2022): 2636. http://dx.doi.org/10.3390/jcm11092636.
Texto completoResumen
Background: The integrated approach to electrical cardioversion (EC) in atrial fibrillation (AF) is complex; candidates can resolve spontaneously while waiting for EC, and post-cardioversion recurrence is high. Thus, it is especially interesting to avoid the programming of EC in patients who would restore sinus rhythm (SR) spontaneously or present early recurrence. We have analyzed the whole elective EC of the AF process using machine-learning (ML) in order to enable a more realistic and detailed simulation of the patient flow for decision making purposes. Methods: The dataset consisted of electronic health records (EHRs) from 429 consecutive AF patients referred for EC. For analysis of the patient outcome, we considered five pathways according to restoring and maintaining SR: (i) spontaneous SR restoration, (ii) pharmacologic-cardioversion, (iii) direct-current cardioversion, (iv) 6-month AF recurrence, and (v) 6-month rhythm control. We applied ML classifiers for predicting outcomes at each pathway and compared them with the CHA2DS2-VASc and HATCH scores. Results: With the exception of pathway (iii), all ML models achieved improvements in comparison with CHA2DS2-VASc or HATCH scores (p < 0.01). Compared to the most competitive score, the area under the ROC curve (AUC-ROC) was: 0.80 vs. 0.66 for predicting (i); 0.71 vs. 0.55 for (ii); 0.64 vs. 0.52 for (iv); and 0.66 vs. 0.51 for (v). For a threshold considered optimal, the empirical net reclassification index was: +7.8%, +47.2%, +28.2%, and +34.3% in favor of our ML models for predicting outcomes for pathways (i), (ii), (iv), and (v), respectively. As an example tool of generalizability of ML models, we deployed our algorithms in an open-source calculator, where the model would personalize predictions. Conclusions: An ML model improves the accuracy of restoring and maintaining SR predictions over current discriminators. The proposed approach enables a detailed simulation of the patient flow through personalized predictions.
15
Wang, Xin, Yilun Han, Wei Xue, Guangwen Yang, and Guang J. Zhang. "Stable climate simulations using a realistic general circulation model with neural network parameterizations for atmospheric moist physics and radiation processes." Geoscientific Model Development 15, no. 9 (2022): 3923–40. http://dx.doi.org/10.5194/gmd-15-3923-2022.
Texto completoResumen
Abstract. In climate models, subgrid parameterizations of convection and clouds are one of the main causes of the biases in precipitation and atmospheric circulation simulations. In recent years, due to the rapid development of data science, machine learning (ML) parameterizations for convection and clouds have been demonstrated to have the potential to perform better than conventional parameterizations. Most previous studies were conducted on aqua-planet and idealized models, and the problems of simulation instability and climate drift still exist. Developing an ML parameterization scheme remains a challenging task in realistically configured models. In this paper, a set of residual deep neural networks (ResDNNs) with a strong nonlinear fitting ability is designed to emulate a super-parameterization (SP) with different outputs in a hybrid ML–physical general circulation model (GCM). It can sustain stable simulations for over 10 years under real-world geographical boundary conditions. We explore the relationship between the accuracy and stability by validating multiple deep neural network (DNN) and ResDNN sets in prognostic runs. In addition, there are significant differences in the prognostic results of the stable ResDNN sets. Therefore, trial and error is used to acquire the optimal ResDNN set for both high skill and long-term stability, which we name the neural network (NN) parameterization. In offline validation, the neural network parameterization can emulate the SP in mid- to high-latitude regions with a high accuracy. However, its prediction skill over tropical ocean areas still needs improvement. In the multi-year prognostic test, the hybrid ML–physical GCM simulates the tropical precipitation well over land and significantly improves the frequency of the precipitation extremes, which are vastly underestimated in the Community Atmospheric Model version 5 (CAM5), with a horizontal resolution of 1.9∘ × 2.5∘. Furthermore, the hybrid ML–physical GCM simulates the robust signal of the Madden–Julian oscillation with a more reasonable propagation speed than CAM5. However, there are still substantial biases with the hybrid ML–physical GCM in the mean states, including the temperature field in the tropopause and at high latitudes and the precipitation over tropical oceanic regions, which are larger than those in CAM5. This study is a pioneer in achieving multi-year stable climate simulations using a hybrid ML–physical GCM under actual land–ocean boundary conditions that become sustained over 30 times faster than the target SP. It demonstrates the emerging potential of using ML parameterizations in climate simulations.
16
Dzis, Ivan, Oleksandra Tomashevska, Yevhen Dzis, and Zoryana Korytko. "Prediction of survival in non-Hodgkin lymphoma based on markers of systemic inflammation, anemia, hypercoagulability, dyslipidemia, and Eastern Cooperative Oncology Group performance status." Acta Haematologica Polonica 51, no. 1 (2020): 34–41. http://dx.doi.org/10.2478/ahp-2020-0008.
Texto completoResumen
AbstractBackgroundThe International Prognostic Index and its modifications are used to estimate prognosis in non-Hodgkin lymphoma. However, the outcome is often different in patients with similar index scores.AimThe aim of this study was to elaborate a prognostic model for patients with mature B-cell non-Hodgkin lymphoma using a combination of predictive markers.Material and methodsThe study included 45 patients with mature B-cell non-Hodgkin lymphoma. Before the administration of treatment, clinical and laboratory parameters were measured. After the 35-month follow-up period, overall survival was studied in relation to the data obtained at initial examination.ResultsWe revealed nine adverse predictive markers for overall survival of enrolled patients: Eastern Cooperative Oncology Group (ECOG) performance status >1; erythrocyte sedimentation rate >30 mm/h; levels of hemoglobin <120 g/L, fibrinogen ≥6 g/L, interleukin-6 ≥2 pg/mL, tumor necrosis factor ≥1.45 pg/mL, soluble fibrin monomer complexes >4 mg/dL, high-density lipoprotein cholesterol <1.03 mmol/L in men, and <1.29 mmol/L in women; and short activated partial thromboplastin time. A prognostic model for the estimation of the risk of death within the ensuing 1.5–2 years in patients with non-Hodgkin lymphoma was constructed.ConclusionMarkers of inflammation, anemia, hypercoagulability, dyslipidemia, and poor ECOG status are associated with worse survival in patients with mature B-cell non-Hodgkin lymphoma.
17
Bruschetta, Roberta, Gennaro Tartarisco, Lucia Francesca Lucca, et al. "Predicting Outcome of Traumatic Brain Injury: Is Machine Learning the Best Way?" Biomedicines 10, no. 3 (2022): 686. http://dx.doi.org/10.3390/biomedicines10030686.
Texto completoResumen
One of the main challenges in traumatic brain injury (TBI) patients is to achieve an early and definite prognosis. Despite the recent development of algorithms based on artificial intelligence for the identification of these prognostic factors relevant for clinical practice, the literature lacks a rigorous comparison among classical regression and machine learning (ML) models. This study aims at providing this comparison on a sample of TBI patients evaluated at baseline (T0), after 3 months from the event (T1), and at discharge (T2). A Classical Linear Regression Model (LM) was compared with independent performances of Support Vector Machine (SVM), k-Nearest Neighbors (k-NN), Naïve Bayes (NB) and Decision Tree (DT) algorithms, together with an ensemble ML approach. The accuracy was similar among LM and ML algorithms on the analyzed sample when two classes of outcome (Positive vs. Negative) approach was used, whereas the NB algorithm showed the worst performance. This study highlights the utility of comparing traditional regression modeling to ML, particularly when using a small number of reliable predictor variables after TBI. The dataset of clinical data used to train ML algorithms will be publicly available to other researchers for future comparisons.
18
Setiawan, Rinaldy T., Eko Prasetyo, Maximillian Ch Oley, and Fredrik G. Langi. "Relationship between Serum Fibronectin and Level of Consciousness according to FOUR Score in Traumatic Brain Injury Patients." e-CliniC 10, no. 2 (2022): 160. http://dx.doi.org/10.35790/ecl.v10i2.39165.
Texto completoResumen
Abstract: Traumatic brain injuries (TBI) are determined by the severity of the primary and secondary brain damage. Fibronectin and FOUR score are suggested to be diagnostic and prognostic predictors in patients with traumatic brain injuries (TBI). This study aimed to evaluate the relationship between serum fibronectin level and FOUR score in TBI patients. This was an observational study with a prospective cohort method design, conducted on TBI patients admitted to the emergency room at Prof. Dr. R. D. Kandou Hospital. Serum fibronectin examination and assessment of the level of consciousness determined by the FOUR score were performed when the patient entered the emergency room <24 hours. A proportional regression model was used to assess the relationship between serum fibronectin levels and the FOUR score. The results obtained 65 patients, median FOUR score of 13, and 8 patients (12%) with high-risk TBI, median serum fibronectin level of 4 ng/ml, and seven patients (11%) died. The ability of fibronectin as a prognostic factor, especially mortality, did not differ from FOUR score. Logistic regression estimated that serum fibronectin levels >7 ng/ml would increase mortality 33 times and the incidence of mortality increased 23 times. A FOUR score of 8 or less had mortality 34 times and a relative risk of 28 times. In conclusion, there is a significant relationship between serum fibronectin level and FOUR score in terms of stratification of TBI patients. Elevated serum fibronectin level can be used as a diagnostic biomarker and prognostic evaluation of mortality in TBI patients.Keywords: fibronectin; FOUR score; traumatic brain injury Abstrak: Fibronektin dan Skor FOUR disarankan sebagai prediktor diagnostik dan prognositk pada pasien COT. Penelitian ini bertujuan untuk mengevaluasi hubungan antara kadar fibronektin serum dan skor FOUR pada pasien COT. Jenis penelitian ialah observasional dengan desain metode kohort prospektif, dilakukan pada pasien COT yang masuk ke IGD RSUP Prof. Dr. R. D. Kandou Manado. Pemeriksaan fibronektin serum dan penilaian tingkat kesadaran ditentukan dengan skor FOUR dilakukan saat pasien masuk ke IGD <24 jam. Model regresi proporsional digunakan untuk menilai hubungan antara kadar fibronektin serum dan skor FOUR. Hasil penelitian mendapatkan 65 pasien COT. Median skor FOUR 13, 8 pasien (12%) COT risiko tinggi (FOUR 0-7), median kadar serum fibronektin 4 ng/ml, 7 pasien (11%) meninggal. Fibronektin sebagai faktor prognostik, khususnya mortalitas, tidak berbeda dengan skor FOUR, regresi logistik mengestimasi bahwa kadar serum fibronektin >7 ng/ml mening-katkan OR mortalitas 33 kali dan insidens mortalitas 23 kali. skor FOUR 8 memiliki odds mortalitas 34 kali dan resiko relatif 28 kali. Simpulan penelitian ini ialah terdapat hubungan bermakna antara kadar serum fibronektin dan Skor FOUR dalam hal stratifikasi pasien COT. Peningkatan kadar serum fibronektin dapat dijadikan sebagai biomarker diagnostik dan evaluasi prognostik mortalitas pasien COT.Kata kunci: fibronektin; skor FOUR; cedera otak traumatik
19
Lin, Weiyuan, Lifeng Que, Guisen Lin, et al. "Using Machine Learning to Predict Five-Year Reintervention Risk in Type B Aortic Dissection Patients After Thoracic Endovascular Aortic Repair." Journal of Medical Imaging and Health Informatics 11, no. 6 (2021): 1560–67. http://dx.doi.org/10.1166/jmihi.2021.3813.
Texto completoResumen
Purpose: Type B aortic dissection (TBAD) is a high-risk disease, commonly treated with thoracic endovascular aortic repair (TEVAR). However, for the long-term follow-up, it is associated with a high 5-year reintervention rate for patients after TEVAR. There is no accurate definition of prognostic risk factors for TBAD in medical guidelines, and there is no scientific judgment standard for patients’ quality of life or survival outcome in the next five years in clinical practice. A large amount of medical data features makes prognostic analysis difficult. However, machine learning (ML) permits lots of objective data features to be considered for clinical risk stratification and patient management. We aimed to predict the 5-year prognosis in TBAD after TEVAR by Ml, based on baseline, stent characteristics and computed tomography angiography (CTA) imaging data, and provided a certain degree of scientific basis for prognostic risk score and stratification in medical guidelines. Materials and Methods: Dataset we recorded was obtained from 172 TBAD patients undergoing TEVAR. Totally 40 features were recorded, including 14 baseline, 5 stent characteristics and 21 CTA imaging data. Information gain (IG) was used to select features highly associated with adverse outcome. Then, the Gradient Boost classifier was trained using grid search and stratified 5-fold cross-validation, and Its predictive performance was evaluated by the area under the curve (AUC) in the receiver operating characteristic (ROC). Results: Totally 60 patients underwent reintervention during follow-up. Combing 24 features selected by IG, ML model predicted prognosis well in TBAD after TEVAR, with an AUC of 0.816 and a 95% confidence interval of 0.797 to 0.837. Reintervention rate of prediction was slightly higher than the actual (48.2% vs. 34.8%). Conclusion: Machine learning, which combined with baseline, stent characteristics and CTA imaging data for personalized risk computations, effectively predicted reintervention risk in TBAD patients after TEVAR in 5-year follow-up. The model could be used to efficiently assist the clinical management of TBAD patients and prompt high-risk factors.
20
Kumar, Shaji, Angela Dispenzieri, Martha Q. Lacy, et al. "Revised Prognostic Staging System for Light Chain Amyloidosis Incorporating Cardiac Biomarkers and Serum Free Light Chain Measurements." Journal of Clinical Oncology 30, no. 9 (2012): 989–95. http://dx.doi.org/10.1200/jco.2011.38.5724.
Texto completoResumen
Purpose Cardiac involvement predicts poor prognosis in light chain (AL) amyloidosis, and the current prognostic classification is based on cardiac biomarkers troponin-T (cTnT) and N-terminal pro–B-type natriuretic peptide (NT-ProBNP). However, long-term outcome is dependent on the underlying plasma cell clone, and incorporation of clonal characteristics may allow for better risk stratification. Patients and Methods We developed a prognostic model based on 810 patients with newly diagnosed AL amyloidosis, which was further examined in two other datasets: 303 patients undergoing stem-cell transplantation, and 103 patients enrolled onto different clinical trials. Results We examined the prognostic value of plasma cell–related characteristics (ie, difference between involved and uninvolved light chain [FLC-diff], marrow plasma cell percentage, circulating plasma cells, plasma cell labeling index, and β2 microglobulin). In a multivariate model that included these characteristics as well as cTnT and NT-ProBNP, only FLC-diff, cTnT, and NT-ProBNP were independently prognostic for overall survival (OS). Patients were assigned a score of 1 for each of FLC-diff ≥ 18 mg/dL, cTnT ≥ 0.025 ng/mL, and NT-ProBNP ≥ 1,800 pg/mL, creating stages I to IV with scores of 0 to 3 points, respectively. The proportions of patients with stages I, II, III and IV disease were 189 (25%), 206 (27%), 186 (25%) and 177 (23%), and their median OS from diagnosis was 94.1, 40.3, 14, and 5.8 months, respectively (P < .001). This classification system was validated in the other datasets. Conclusion Incorporation of serum FLC-diff into the current staging system improves risk stratification for patients with AL amyloidosis and will help develop risk-adapted therapies for AL amyloidosis.
21
Ho, Shu-Yein, Po-Hong Liu, Chia-Yang Hsu, et al. "A New Tumor Burden Score and Albumin–Bilirubin Grade-Based Prognostic Model for Hepatocellular Carcinoma." Cancers 14, no. 3 (2022): 649. http://dx.doi.org/10.3390/cancers14030649.
Texto completoResumen
The prognosis of hepatocellular carcinoma (HCC) varies widely due to variable tumor extent and liver reserve. We aimed to develop and validate a new prognostic model based on tumor burden score (TBS) and albumin–bilirubin (ALBI) grade for HCC. We prospectively identified 3794 HCC patients who were randomized into derivation and validation groups. Survival predictors were evaluated by a multivariate Cox model. The TBS–ALBI system allocated two points for high TBS and ALBI grade 3, and one point each for the presence of ascites, serum α-fetoprotein ≥ 400 ng/mL, vascular invasion or distant metastasis, performance status 2–4, medium TBS, and ALBI grade 2, with a maximal score of 8 points. Significant survival differences were found across different TBS–ALBI score groups in the validation cohort (all p < 0.001). The TBS–ALBI system had the lowest corrected Akaike information criterion (AICc) and the highest homogeneity compared with other proposed staging models. The discriminative ability of the TBS–ALBI system was consistently stable across different viral etiologies, cancer stages, and treatment strategies. Conclusions: This new TBS–ALBI system is a feasible and robust prognostic system in comparison with other systems; it is a user-friendly tool for long-term outcome assessment independent of treatment modality and cancer stage in HCC.
22
Kantauskaitė, Marta, Agnė Laučytė-Cibulskienė, and Marius Miglinas. "Histopathological Classification—A Prognostic Tool for Rapidly Progressive Glomerulonephritis." Medicina 54, no. 2 (2018): 17. http://dx.doi.org/10.3390/medicina54020017.
Texto completoResumen
Background: Recently proposed histopathological classification may predict patient outcome in pauci-immune glomerulonephritis. This study sought to prove that the prognostic effect could be extended to all types of rapidly progressive glomerulonephritis. Methods: Retrospective analysis of patients diagnosed with rapidly progressive glomerulonephritis between April 1999 and August 2015 was performed. Epidemiological and clinical data were collected from medical records. The descriptions of renal biopsies were reviewed and classified into focal, sclerotic, crescentic and mixed class according to classification proposed by Berden et al. The study end points were end stage renal disease (ESRD) or death. Survival analyses were modelled using Cox regression. Results: 73 renal biopsies with diagnosis of rapidly progressive glomerulonephritis were included in the study. 25 (34.2%), 16 (21.9%), 24 (32.9%) and 8 (11%) patients were assigned to focal, crescentic, mixed and sclerotic class, respectively. Thirty-two (42.5%) patients were anti-neutrophil cytoplasmic antibody (ANCA) negative, of which eight (10.9%) were anti–glomerular basement membrane antibody (anti–GBM) positive and 24 (32.8%) were negative for autoimmune antibodies. Six (8.2%) patients died within one year. Among patients who survived, median change in estimated glomerular filtration rate (eGFR) values were: −10.5 mL/min in focal, 4.2 mL/min in crescentic, −4.3 mL/min in mixed and 4.1 mL/min in sclerotic group, p > 0.05. In the Cox regression model, there was no significant predictor of patient survival whereas the sclerotic group (HR 3.679, 95% CI, 1.164–11.628, p < 0.05) and baseline eGFR of <15 mL/min (HR 4.832, 95% CI, 1.55–15.08, p < 0.01) had an unfavorable effect for renal survival. Conclusions: Predominant glomerular sclerosis and low eGFR at baseline are associated with higher risk of ESRD in cases with crescentic glomerulonephritis. Therefore, despite the origin of injury, histological classification might aid in prediction of patient outcomes in rapidly progressive glomerulonephritis.
23
Lorenzi, M., B. Lorenzi, and R. Vernillo. "Serum Ferritin in Colorectal Cancer Patients and its Prognostic Evaluation." International Journal of Biological Markers 21, no. 4 (2006): 235–41. http://dx.doi.org/10.1177/172460080602100407.
Texto completoResumen
The aim of this study was to investigate the relationship between preoperative serum ferritin levels, clinicopathological parameters and survival analysis of patients with colorectal cancer. Ninety-four patients (57 males) with a mean age of 65 years (39–87 years) underwent 63 curative and 31 palliative operations. Follow-up was at least 5 years. Patients were categorized with normal (30-215 ng/mL in men and 11-148 ng/mL in women), low, or high serum ferritin levels. Prognostic evaluation was undertaken with stratified survival analysis and Cox's regression model. Twenty-nine of the patients (30.9%) had raised ferritin levels and 14 (14.9%) had low values. Comparisons of the survival curves showed significant differences in stage C disease; specifically, patients with either low or high ferritin levels had a shorter survival than patients with normal levels. Patients who underwent palliative surgery and had high ferritin serum values also had a shorter survival. In multivariate analysis, the variables with a negative effect on survival were stage, serum ferritin levels and age. Our data suggest that patients with advanced colorectal cancer having normal preoperative serum ferritin levels may have a better prognosis, although the prognostic value related to this association requires further investigation.
24
Koller, Charles Asa, B. Nebiyou Bekele, Xian Zhou, et al. "Thrombopoietin as an Independent Prognostic Marker in Chronic Lymphocytic Leukemia." Blood 104, no. 11 (2004): 1905. http://dx.doi.org/10.1182/blood.v104.11.1905.1905.
Texto completoResumen
Abstract Several cytokines and growth factors are involved in the regulation of megakaryocytopoieses and platelet formation. Interleukin-11 (IL-11), IL-6, IL-3, IL-1b, and thrombopoietin (TPO) act synergistically to promote proliferation and maturation of megakaryocytes. Recombinant IL-11 and TPO are under clinical investigation as supplements to stimulate thrombopoiesis in patients with cancer. We investigated the plasma levels of TPO in 127 patients with chronic lymphocytic leukemia (CLL) and correlated these levels with platelet counts and various laboratory and clinical characteristics. TPO levels were significantly higher in patients with CLL (median, 232 pg/mL; range, 26.4–2714.5 pg/mL) than in healthy control subjects (P <0.01). More importantly, higher levels of TPO were associated with advanced Rai stage (P <0.0001), higher b-2-microglublin (b2-M) levels (P<0.001), and the presence of mutation in the IgVH gene (P <0.001). TPO levels were inversely correlated with platelet count (P = 0.002). Univariate Cox proportional hazard models demonstrated that high TPO levels in CLL patients were associated with shorter survival (P <0.0001); multivariate analysis demonstrated that this association was independent of the IgVH mutation status, b2-M levels, and Rai stage. Recursive partitioning showed that a cut-point of 639.4 pg/mL separated CLL patients into two major groups with a significant difference in survival (P <0.0001). When b2-M level was added to this model, its effect was masked by the effects of TPO in patients with TPO >639.4 pg/mL. However, among patients with TPO levels =639.4 pg/mL, those with b2-M levels >4.95 mg/L had significantly shorter survival than those with lower b2-M levels. These data not only emphasize the importance of TPO in reflecting the aggressiveness of CLL, but also suggest that TPO and b2-M can be used to predict clinical behavior in this disease and may replace the need for determining IgVH mutation status. Further studies are needed to determine if the prognostic role of TPO is due to its direct effects on the growth of the leukemic clone or simply reflects a general failure in the homeostatic regulation of various cytokines. Figure Figure
25
Kneev, A. Y., M. I. Shkol’nik, O. A. Bogomolov, and G. M. Zharinov. "Prostate specif c antigen density as a prognostic factor in patients with prostate cancer treated with combined hormonal radiation therapy." Siberian journal of oncology 21, no. 3 (2022): 12–23. http://dx.doi.org/10.21294/1814-4861-2022-21-3-12-23.
Texto completoResumen
Purpose. To evaluate prostate specifc antigen density (PSAD) as a predictor of overall (OS) and cancerspecifc survival (CSS) in patients with prostate cancer (PC) who have undergone combined hormonal-radiation therapy.Material and Methods. In order to assess the prognostic signifcance of PSAD we retrospectively analyzed outcomes of 714 PCa patients who received combined hormonal-radiation therapy at the A.M. Granov Russian Scientifc Center of Radiology and Surgical Technologies, Ministry of Healthcare of Russia, between January 1996 and December 2016. Since the prognosis and management differ according to the extent of tumor involvement, patients were categorized into localized (n=272), locally advanced (n=231) and metastatic (n=211) PC groups. We equentially applied ROC-analysis, Kaplan-Meier product limit estimator and Cox proportional hazards model to assess the prognostic relevance and establish threshold values of PSAD that had a signifcant impact on survival rates.Results. In the localized PC group, PSAD threshold values of 0.34 ng/mL/cc and 0.36 ng/mL/cc were associated with a decrease in OS and CSS, respectively. Patients with “low” PSAD had signifcantly better OS and CSS survival rates in both uni- and multivariate analyses. In locally advanced PC group, PSAD threshold values were 0.28 ng/mL/cc and 0.63 ng/mL/cc for OS and CSS, respectively. However, exceeding the specifed values, in the locally advanced PC group, was not accompanied by a statistically signifcant decrease in survival rates. Finally, in the metastatic PC group, established PSAD threshold values were 2.25 ng/mL/cc and 2.30 ng/mL/ccfor OS and CSS, respectively. According to the results of univariate analysis, patients with “low” PSA tend to demonstrate statistically signifcant better OS and CSS rates. The results of multivariate analysis, however, failed to prove PSAD as an independent prognostic factor within the metastatic PC cohort. Conclusion. PSA density is a reliable tool for assessing survival rates in patients with localized PC who have undergone combined hormonal-radiation therapy.
26
Shin, Kabsoo, Joori Kim, Juyeon Park, Ok Ran Kim, Nahyeon Kang, and In-Ho Kim. "Prognostic significance of exosomal programmed death-ligand 1 in advanced gastric cancer patients treated with first-line chemotherapy." Journal of Clinical Oncology 40, no. 4_suppl (2022): 665. http://dx.doi.org/10.1200/jco.2022.40.4_suppl.665.
Texto completoResumen
665 Background: Recent studies have found that exosomal PD-L1 could be associated with prognosis in several malignancies by its potential immunosuppressive role. Prognostic role of exosomal PD-L1 in advanced gastric cancer patients treated with systemic chemotherapy has not been well explored. The aim of the present study was to explore prognostic and predictive significance of exosomal PD-L1 in advanced gastric cancer patients. Methods: We prospectively collected plasma samples of patients with advanced gastric cancer receiving first-line chemotherapy at pre and post treatment. Combined ultracentrifugation with polymer-based precipitation was applied to isolate exosome and enzyme-linked immunosorbent assay(ELISA) was applied to measure exosomal PD-L1. The correlation between exosomal PD-L1 and circulating immune cells in the plasma was evaluated. Patients were divided into two groups according to pretreatment exosomal PD-L1 level. Survival analysis was performed using Kaplan-Meier method and multivariate analyses based on Cox-proportional hazards regression model. Results: A total of 99 patients were enrolled in the study, and serum samples collected at pre-treatment and post-treatment were pooled. Median value of the pre-treatment exosomal PD-L1 was 0.42 pg/ml (range 0.01 ̃3.75). Patients with pre-treatment exoPD-L1 < 0.42 pg/mL (exoPD-L1low) showed significantly better overall survival than those with exoPD-L1 ≥ 0.42 pg/mL(exoPD-L1high) (p = 0.033). Progression-free survival was not different between the groups (p = 0.754). exoPD-L1lowgroup showed significantly higher percentages of CD8+T cells than exoPD-L1high group. (p = 0.024). In multivariate analysis, pretreatment exosomal PD-L1, poorly differentiated tumors and No. of metastatic sites(≥2) were independent prognostic factors for overall survival. Conclusions: In summary, the level of pre-treatment exosomal PD-L1 could be used as a prognostic marker for patients with advanced gastric cancer receiving systemic chemotherapy.
27
Jia, Jing, MinZhe Li, Wenhao Teng, et al. "Prognostic Significance of Preoperative Serum Carcinoembryonic Antigen Varies with Lymph Node Metastasis Status in Colorectal Cancer." Journal of Oncology 2021 (December 27, 2021): 1–8. http://dx.doi.org/10.1155/2021/4487988.
Texto completoResumen
Background. Preoperative serum level of carcinoembryonic antigen (pCEA) is generally recognized as a prognostic factor for colorectal cancer (CRC), but the stage-specific role of pCEA in colorectal cancer remains unclear. Objective. We investigated the prognostic significance of pCEA levels in different tumor stages of nonmetastatic CRC patients. Methods. Six hundred and fifteen CRC patients at stage I–III were retrospectively analyzed. All of them received curative tumor resection. The X-tile program was used to generate stage-specific cutoff values of pCEA for all patients and two subpopulations (lymph node-positive or -negative). The prognostic significance of pCEA was assessed using Kaplan–Meier analysis and Cox proportional hazards regression analysis. A nomogram model that combined pCEA score and clinical feature indexes was established and evaluated. Results. Two cutoff values were identified in the study population. At a cutoff value of 4.9 ng/mL, a significantly higher 5-year overall survival (OS) rate (82.16%) was observed in the pCEA-low group (<4.9 ng/mL) compared with 65.52% in the pCEA-high group (≥4.9 ng/mL). Furthermore, at the second cutoff value of 27.2 ng/mL, 5-year OS was found to be only 40.9%. Stratification analysis revealed that preoperative serum level of pCEA was an independent prognostic factor (OR = 1.991, P < 0.01 ) in the subpopulation of lymph node metastasis (stage III) patients, and the relative survival rates in the pCEA-low (≤4.9 ng/mL), pCEA-medium (4.9–27.2 ng/mL), and pCEA-high (≥27.2 ng/mL) groups were 73.4%, 60.5%, and 24.8%, respectively ( P < 0.05 ). However, no such effect was observed in the lymph node nonmetastasis (stage I and II) subgroup. The established nomogram showed acceptable predictive power of the 5-year OS rate (C-index: 0.612) in lymph node-positive CRC patients, with an area under the curve value of 0.772, as assessed by ROC curve analysis. Conclusions. Pretreatment serum CEA levels had different prognostic significance based on the lymph node metastasis status. Among stage III CRC patients, pCEA was an independent prognostic factor. Five-year OS rates could be predicted according to the individual pCEA level at the different cutoff values.
28
Shi, Na, Lan Lan, Jiawei Luo, et al. "Predicting the Need for Therapeutic Intervention and Mortality in Acute Pancreatitis: A Two-Center International Study Using Machine Learning." Journal of Personalized Medicine 12, no. 4 (2022): 616. http://dx.doi.org/10.3390/jpm12040616.
Texto completoResumen
Background: Current approaches to predicting intervention needs and mortality have reached 65–85% accuracy, which falls below clinical decision-making requirements in patients with acute pancreatitis (AP). We aimed to accurately predict therapeutic intervention needs and mortality on admission, in AP patients, using machine learning (ML). Methods: Data were obtained from three databases of patients admitted with AP: one retrospective (Chengdu) and two prospective (Liverpool and Chengdu) databases. Intervention and mortality differences, as well as potential predictors, were investigated. Univariate analysis was conducted, followed by a random forest ML algorithm used in multivariate analysis, to identify predictors. The ML performance matrix was applied to evaluate the model’s performance. Results: Three datasets of 2846 patients included 25 potential clinical predictors in the univariate analysis. The top ten identified predictors were obtained by ML models, for predicting interventions and mortality, from the training dataset. The prediction of interventions includes death in non-intervention patients, validated with high accuracy (96%/98%), the area under the receiver-operating-characteristic curve (0.90/0.98), and positive likelihood ratios (22.3/69.8), respectively. The post-test probabilities in the test set were 55.4% and 71.6%, respectively, which were considerably superior to existing prognostic scores. The ML model, for predicting mortality in intervention patients, performed better or equally with prognostic scores. Conclusions: ML, using admission clinical predictors, can accurately predict therapeutic interventions and mortality in patients with AP.
29
Tschepe, Merle, Valerie Seeber, Isabella Zwiener, et al. "A novel H-FABP assay and a fast prognostic score for risk assessment of normotensive pulmonary embolism." Thrombosis and Haemostasis 111, no. 05 (2014): 996–1003. http://dx.doi.org/10.1160/th13-08-0663.
Texto completoResumen
SummaryWe tested whether heart-type fatty acid binding protein (H-FABP) measured by a fully-automated immunoturbidimetric assay in comparison to ELISA provides additive prognostic value in patients with pulmonary embolism (PE), and validated a fast prognostic score in comparison to the ESC risk prediction model and the simplified Pulmonary Embolism Severity Index (sPESI). We prospectively examined 271 normotensive patients with PE; of those, 20 (7%) had an adverse 30-day outcome. H-FABP levels determined by immunoturbidimetry were higher (median, 5.2 [IQR; 2.7–9.8] ng/ml) than those by ELISA (2.9 [1.1–5.4] ng/ml), but Bland-Altman plot demonstrated a good agreement of both assays. The area under the curve for H-FABP was greater for immunoturbidimetry than for ELISA (0.82 [0.74–0.91] vs 0.78 [0.68–0.89]; P=0.039). H-FABP measured by immunoturbidimetry (but not by ELISA) provided additive prognostic information to other predictors of 30-day outcome (OR, 12.4 [95% CI, 1.6–97.6]; P=0.017). When H-FABP determined by immunoturbidimetry was integrated into a novel prognostic score (H-FABP, Syncope, and Tachycardia; FAST score), the score provided additive prognostic information by multivariable analysis (OR, 14.2 [3.9–51.4]; p<0.001; c-index, 0.86) which were superior to information obtained by the ESC model (c-index, 0.62; net reclassification improvement (NRI), 0.39 [0.21–0.56]; P<0.001) or the sPESI (c-index, 0.68; NRI, 0.24 [0.05–0.43]; P=0.012). In conclusion, determination of H-FABP by immunoturbidimetry provides prognostic information superior to that of ELISA and, if integrated in the FAST score, appears more suitable to identify patients with an adverse 30-day outcome compared to the ESC model and sPESI.
30
Nishimura, Noriko, Masahiro Yokoyama, Kengo Takeuchi, et al. "Soluble Interleukin-2 Receptors (sIL-2R) Is An Independent Prognostic Factor for Patients with Diffuse Large B Cell Lymphoma Treated with R-CHOP." Blood 118, no. 21 (2011): 2678. http://dx.doi.org/10.1182/blood.v118.21.2678.2678.
Texto completoResumen
Abstract Abstract 2678 Background: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease characterized by a wide range of clinical outcomes. Rituximab added to CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) chemotherapy, R-CHOP has made a marked improvement in outcome in patients with DLBCL. The International Prognostic Index (IPI), which consists of age > 60 years, stage III/IV, elevated lactate dehydrogenase (LDH) level, Eastern Cooperative Oncology Group (ECOG) performance status (PS) † 2, and more than one extranodal (EN) site of disease, remains the most commonly used system for risk classification in DLBCL. However, recent studies suggested that new agent has altered the significance of previously recognized risk factors. Here we investigate the prognostic impact of reported risk factors in a large DLBCL patient cohort in a single institute to determine a better prognostic model in rituximab era. Patients and Methods: In total, 250 newly diagnosed DLBCL patients treated with R-CHOP regimen at the Cancer Institute Hospital of JFCR between October 2003 and December 2008 were included and analyzed. Progression free survival (PFS) and overall survival (OS) were estimated using the Kaplan-Meier method and compared among risk groups using the log rank test. The Cox proportional hazards model was used to test the significance of prognostic factors. ROC curve was used to determine optimal serum level of sIL-2R and LDH as a cut off value for 4-year mortality risk. Results: The median age of patients was 65 years (range 23–88 years), 56% were male. The median follow-up time was 49 months (range 1–90 months) and 39 deaths had been recorded by the time of the last follow-up. The IPI still remains predictive with an OS ranging from 52.4% to 91.6% at 4 years; however it cannot discriminate between low and low-intermediate group. Revised IPI was valid as well with an OS ranging from 63.3% to 97%. In univariate analysis, elevated sIL-2R level, B symptom, elevated LDH level, PS>2, age>65, stage III/IV, CD5 positive, and EN>1 were significant as poor prognostic factors whereas sex, bulky mass, MIB1 index >90%, Non-GCB were not. Furthermore, multivariate analysis showed that only sIL-2R>924U/ml, CD5 expression, and EN>1 were significant with relative hazard 1.4∼17.5, 1.4∼8.9, and 1.3∼4.7, respectively. As elevated sIL-2R was the most powerful prognostic factor, we performed further analysis on this parameter. Average serum sIL-2R level was 2,775U/ml (range from 220U/ml to 43,100U/ml) with a normal limit of upper is 230U/ml. ROC curve demonstrated that serum sIL-2R was more optimal value than serum LDH to identify high risk patients for 4-year mortality after initiation of R-CHOP therapy and cutoff value of sIL-2R was 924U/ml (1.73 upper limit of normal). sIL-2R level can be divided into three distinctprognostic groups. Patients with sIL-2R<925U/ml fall into a very good group with a 4-year OS:98% and 4-year PFS:90.7%, patients with 925U/ml<=sIL-2R<4,625U/mlfall into a good group with a 4-year OS:82% and 4-year PFS:77.7%, and patients with sIL-2R>=4,625U/ml fall into a poor group with a 4-year OS:59.6% and 4-year PFS:54.7% (P < 0.001). Conclusions: sIL-2R level is an independent and powerful prognostic factor in serum level dependent manner in DLBCL patients treated with R-CHOP. This prognostic model should be reassessed on a larger scale and prospective study. Disclosures: No relevant conflicts of interest to declare.
31
Chen, Fangyue, Piyawat Kantagowit, Tanawin Nopsopon, Arisa Chuklin, and Krit Pongpirul. "Prediction and diagnosis of chronic kidney disease development and progression using machine-learning: Protocol for a systematic review and meta-analysis of reporting standards and model performance." PLOS ONE 18, no. 2 (2023): e0278729. http://dx.doi.org/10.1371/journal.pone.0278729.
Texto completoResumen
Chronic Kidney disease (CKD) is an important yet under-recognized contributor to morbidity and mortality globally. Machine-learning (ML) based decision support tools have been developed across many aspects of CKD care. Notably, algorithms developed in the prediction and diagnosis of CKD development and progression may help to facilitate early disease prevention, assist with early planning of renal replacement therapy, and offer potential clinical and economic benefits to patients and health systems. Clinical implementation can be affected by the uncertainty surrounding the methodological rigor and performance of ML-based models. This systematic review aims to evaluate the application of prognostic and diagnostic ML tools in CKD development and progression. The protocol has been prepared using the Preferred Items for Systematic Review and Meta-analysis Protocols (PRISMA-P) guidelines. The systematic review protocol for CKD prediction and diagnosis have been registered with the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42022356704, CRD42022372378). A systematic search will be undertaken of PubMed, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), the Web of Science, and the IEEE Xplore digital library. Studies in which ML has been applied to predict and diagnose CKD development and progression will be included. The primary outcome will be the comparison of the performance of ML-based models with non-ML-based models. Secondary analysis will consist of model use cases, model construct, and model reporting quality. This systematic review will offer valuable insight into the performance and reporting quality of ML-based models in CKD diagnosis and prediction. This will inform clinicians and technical specialists of the current development of ML in CKD care, as well as direct future model development and standardization.
32
Kastritis, Efstathios, Ioannis Papassotiriou, Evangelos Terpos, et al. "Growth Differentiation Factor-15 in Patients with Light Chain (AL) Amyloidosis Has Independent Prognostic Significance and Adds Prognostic Information Related to Risk of Early Death and Renal Outcomes." Blood 124, no. 21 (2014): 306. http://dx.doi.org/10.1182/blood.v124.21.306.306.
Texto completoResumen
Abstract Growth differentiation factor-15 (GDF-15) is a member of the TGF-beta family, which is involved in several pathological conditions, including inflammation, cancer, cardiovascular, pulmonary and renal diseases. GDF-15 has prognostic value in patients with cardiovascular disorders and adds prognostic information to conventional prognostic factors, such as NT-proBNP and high-sensitivity troponin (hs-TnT). Cardiac involvement is the most important determinant of prognosis in patients with AL amyloidosis and cardiac biomarkers have major prognostic importance in AL. The aim of the study was to explore the value of GDF-15 in patients with AL amyloidosis. We measured the circulating levels of GDF-15, NT-proBNP and hs-TnT in 77 patients with newly diagnosed AL amyloidosis, before and 3 months post frontline treatment. GDF-15 was measured by a novel pre-commercial immunoassay (Roche Diagnostics). Patients' median age was 68 years; most patients had cardiac (61%) or renal involvement (74%); 61% had NT-proBNP >1284 pg/ml and 46% had hsTnT>54 ng/ml. Median eGFR was 57 ml/min/1.73m2, 52% had eGFR <60 ml/min/1.73m2, while 12% required dialysis at the time of treatment initiation. All patients received primary therapy with bortezomib- (49%) or lenalidomide-based regimens (51%). Median levels of GDF-15 were 3594 pg/ml (range 626-71,475pg/ml); 95% of patients with AL had GDF-15 levels >1200 pg/ml (the upper limit of normal for individuals without cardiovascular disease). GDF-15 correlated with NT-proBNP (r=0.538, p<0.001), hs-TnT (r=0.447, p=0.02) and eGFR (r=-0.570, p<0.001). Patients with GDF-15 levels within the upper quartile (>7575 pg/ml) had a very poor outcome (median overall survival (OS) 3 months) compared to patients with GDF-15 levels below the upper quartile (p=0.01; see the Figure). Among other cardiac markers, hs-TnT >54 ng/ml (12 vs >48 months, p=0.001) and NT-proBNP >1284 pg/ml (11 vs >48 months, p<0.001) were also associated with shorter OS. Higher cut-off levels for NT-proBNP and hs-TnT did not discriminate patients at high risk for early death more accurately. In a multiple logistic regression model which included GDF-15, NT-proBNP and hs-TnT, only GDF-15 in the upper quartile (HR: 8.427, 95% CI 1.73-41.1, p=0.008) was independently predictive of early death at 3 months. Similar results were obtained when these biomarkers were treated as continuous variables. Regarding OS, GDF-15 had independent prognostic significance in a multivariate model that included both NT-proBNP and hs-TnT. We also evaluated changes in the levels of GDF-15, NT-proBNP and hs-TnT in patients who received lenalidomide after 3 months of treatment. In these patients NT-proBNP often increases without obvious deterioration of cardiac function, thus complicating the assessment of cardiac response early, during the course of therapy. GDF-15 levels did not change significantly either in patients with hematological response (p=0.998) or those without hematological response (p=0.774). However, NT-proBNP levels increased substantially both in those with hematological response (p=0.05) and in those without hematological responses (p=0.013). Similarly, hs-TnT levels increased in non-responders (p=0.006) and did not change in patients with hematological response (p=0.251). As GDF-15 reflects heart and renal defects, we further evaluated whether GDF-15 could be associated with the risk of progression to ESRD and need for dialysis. Using ROC analysis, GDF-15 >median was identified to better discriminate patients which had a shorter time to dialysis (29 months vs not reached, p=0.001, see the Figure; with 38% vs. 8% progressing to ESRD, respectively). eGFR< 60 ml/min/m2 was also a strong predictor of ESRD (p=0.004). However, in multivariate analysis which included GDF-15 >median, eGFR <60 ml/min/m2 and proteinuria >5 g/day, only GDF-15 was independently associated with a higher risk of ESRD requiring dialysis (HR: 4.25, 95% CI 1.01-18, p=0.045). In conclusion, GDF-15 is a novel biomarker with prognostic implications for different outcomes in patients with AL; it is associated with a high risk of early death, with OS and also with renal outcome. More importantly GDF-15 adds prognostic information independent of the traditional cardiac biomarkers and thus, its measurement in larger series of patients is recommended. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.
33
Tanabe, H., N. Katsumata, K. Matsumoto, et al. "CA125 nadir as a prognostic factor in advanced ovarian carcinoma: A retrospective study of 84 patients achieving clinical CR." Journal of Clinical Oncology 24, no. 18_suppl (2006): 5060. http://dx.doi.org/10.1200/jco.2006.24.18_suppl.5060.
Texto completoResumen
5060 Background: There have been few reports of the relationship between CA125 nadir and the prognosis in patients with advanced ovarian carcinoma (AOC) who achieve a complete clinical response by multivariate analysis. Methods: CA125 nadir and the prognosis after the initial therapy (debulking surgery + first line chemotherapy) were retrospectively investigated in patients (pts) who had AOC meeting the criteria below and received therapy at National Cancer Center Hospital, between 1998 and 2004. The eligibility criteria were: 1) histological diagnosis of mullerian carcinoma, 2) FIGO stages III and IV, 3) a combination of platinum and taxane was administered as the first line chemotherapy, and 4) achieved clinically defined complete response (CR) after the initial therapy (ie, no cancer-related symptoms; normal physical examination, computed tomography scan of the abdomen/pelvis and chest x-ray). The possibility of CA125 nadir being a prognostic factor was investigated in pts who met these criteria by multivariate analysis (age (<50 years or 50 years≤), stage (III or IV), histological type (serous adenocarcinoma (serous) or non-serous), residual tumor diameter (<2 cm or 2 cm≤) and CA125 nadir (<10 U/ml or 10 U/ml≤)) using the Cox regression model. Results: There were 84 pts with a median age of 55.5 years (26–74). The stage was III in 59 pts and IV in 25 pts, the histological type was serous in 71 pts and non-serous in 13pts, and the residual tumor diameter was <2 cm in 72 pts and >2 cm in 12 pts and the median of CA125 before the initial therapy was 535 U/ml (13–28190), the CA125 nadir was <10 U/ml in 54 pts and 10 U/ml≤ in 30 pts, respectively. Regarding the prognosis, the median progression free survival (PFS) was 19 months (6–82), and the median overall survival (OS) was 36.5 months (10–82). By multivariate analysis, the CA125 nadir was significantly associated with the prognosis (hazard ratio of PFS was 0.39 (95% CI, 0.21–0.71), hazard ratio of OS was 0.28 (95% CI, 0.11–0.72)). Conclusion: CA125 nadir is a prognostic factor in patients with AOC who achieved clinically defined CR after the initial therapy, when the cut-off value was set to 10 U/ml. CA125 nadir may be an important factor for identifying pts for whom maintenance chemotherapy is effective. [Table: see text]
34
Faderl, Stefan, Kim-Anh Do, Marcella M. Johnson, et al. "Angiogenic factors may have a different prognostic role in adult acute lymphoblastic leukemia." Blood 106, no. 13 (2005): 4303–7. http://dx.doi.org/10.1182/blood-2005-03-1010.
Texto completoResumen
Angiogenesis plays an important role in solid tumors and hematologic malignancies. The prognostic significance of angiogenic factors in adult acute lymphoblastic leukemia (ALL) remains ambiguous. We therefore analyzed the impact of angiogenic factor levels on overall survival of newly diagnosed adult ALL patients. Plasma levels of vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), interleukin-1 receptor alpha (IL-1Rα), IL-6, IL-8, VEGF receptors VEGFR1 and VEGFR2, and thrombopoietin (TPO) were measured in plasma samples of 95 patients by enzyme-linked immunosorbent assay (ELISA). In a univariate Cox proportional hazards model, higher levels of IL-1Rα, IL-8, VEGFR1, and VEGFR2 were predictive of poor survival. In contrast, higher levels of VEGF were predictive of longer survival, and higher levels of bFGF suggested a similar trend (P = .09). The multivariate model simultaneously included VEGF (relative risk [RR] for death, 8.01; P = .001 for levels less than or equal to 19.5 pg/mL), IL-1Rα (RR, 5.12; P = .007 for levels greater than 373 pg/mL), and VEGFR2 (RR, 4.01; P = .04 for levels greater than 8222 pg/mL) as independent factors for survival. Of interest is the association of high levels of VEGF with good prognosis and higher levels of VEGF receptors with poor outcome. These data reflect the complexity by which angiogenic factors may affect the clinical behavior of patients with ALL, and this complexity should be considered in any therapeutic strategy incorporating antiangiogenic agents.
35
Chufal, Kundan S., Irfan Ahmad, Anjali K. Pahuja, Alexis A. Miller, Rajpal Singh, and Rahul L. Chowdhary. "Application of Artificial Neural Networks for Prognostic Modeling in Lung Cancer after Combining Radiomic and Clinical Features." Asian Journal of Oncology 05, no. 02 (2019): 050–55. http://dx.doi.org/10.1055/s-0039-3401438.
Texto completoResumen
Abstract Objective This study was aimed to investigate machine learning (ML) and artificial neural networks (ANNs) in the prognostic modeling of lung cancer, utilizing high-dimensional data. Materials and Methods A computed tomography (CT) dataset of inoperable nonsmall cell lung carcinoma (NSCLC) patients with embedded tumor segmentation and survival status, comprising 422 patients, was selected. Radiomic data extraction was performed on Computation Environment for Radiation Research (CERR). The survival probability was first determined based on clinical features only and then unsupervised ML methods. Supervised ANN modeling was performed by direct and hybrid modeling which were subsequently compared. Statistical significance was set at <0.05. Results Survival analyses based on clinical features alone were not significant, except for gender. ML clustering performed on unselected radiomic and clinical data demonstrated a significant difference in survival (two-step cluster, median overall survival [ mOS]: 30.3 vs. 17.2 m; p = 0.03; K-means cluster, mOS: 21.1 vs. 7.3 m; p < 0.001). Direct ANN modeling yielded a better overall model accuracy utilizing multilayer perceptron (MLP) than radial basis function (RBF; 79.2 vs. 61.4%, respectively). Hybrid modeling with MLP (after feature selection with ML) resulted in an overall model accuracy of 80%. There was no difference in model accuracy after direct and hybrid modeling (p = 0.164). Conclusion Our preliminary study supports the application of ANN in predicting outcomes based on radiomic and clinical data.
36
Li, Qing, Fanfei Kong, Jian Ma, et al. "Nomograms Based on Fibrinogen, Albumin, Neutrophil-Lymphocyte Ratio, and Carbohydrate Antigen 125 for Predicting Endometrial Cancer Prognosis." Cancers 14, no. 22 (2022): 5632. http://dx.doi.org/10.3390/cancers14225632.
Texto completoResumen
Background: This study aimed to determine the prognostic value of the preoperative levels of fibrinogen, albumin (ALB), neutrophil–lymphocyte ratio (NLR), and carbohydrate antigen 125 (CA125) in endometrial cancer and to establish nomograms for predicting patient survival. Methods: Patients with endometrial cancer (n = 1483) who underwent surgery were included in this study, and their preoperative fibrinogen, ALB, NLR, and CA125 levels and clinicopathological characteristics were collected. Patients were randomized into a training cohort (70%, n = 1038) and an external validation cohort (30%, n = 445). The Cox regression analysis was performed using the data for the patients in the training cohort to identify independent prognostic factors; nomograms for predicting prognosis were established and validated. Results: High fibrinogen (≥3.185 g/L), NLR (≥2.521 g/L), and CA125 (≥35 U/mL) levels and low ALB (<4.185 g/L) levels were independently associated with poor progression-free survival (PFS) and poor overall survival (OS) in patients with endometrial cancer. Prognostic prediction model nomograms were developed and validated based on these results. Calibration curves and C-indexes underscored the good predictive power of the nomograms, and both the net reclassification index (NRI) and integrated discrimination improvement (IDI) values of the prognostic prediction model nomograms were improved. Conclusion: Nomograms that are developed based on preoperative fibrinogen, ALB, NLR, and CA125 levels accurately predict PFS and OS in patients with endometrial cancer.
37
Agibetov, Asan, Benjamin Seirer, Theresa-Marie Dachs, et al. "Machine Learning Enables Prediction of Cardiac Amyloidosis by Routine Laboratory Parameters: A Proof-of-Concept Study." Journal of Clinical Medicine 9, no. 5 (2020): 1334. http://dx.doi.org/10.3390/jcm9051334.
Texto completoResumen
(1) Background: Cardiac amyloidosis (CA) is a rare and complex condition with poor prognosis. While novel therapies improve outcomes, many affected individuals remain undiagnosed due to a lack of awareness among clinicians. This study was undertaken to develop an expert-independent machine learning (ML) prediction model for CA relying on routinely determined laboratory parameters. (2) Methods: In a first step, we developed baseline linear models based on logistic regression. In a second step, we used an ML algorithm based on gradient tree boosting to improve our linear prediction model, and to perform non-linear prediction. Then, we compared the performance of all diagnostic algorithms. All prediction models were developed on a training cohort, consisting of patients with proven CA (positive cases, n = 121) and amyloidosis-unrelated heart failure (HF) patients (negative cases, n = 415). Performances of all prediction models were evaluated on a separate prognostic validation cohort with 37 CA-positive and 124 CA-negative patients. (3) Results: Our best model, based on gradient-boosted ensembles of decision trees, achieved an area under the receiver operating characteristic curve (ROC AUC) score of 0.86, with sensitivity and specificity of 89.2% and 78.2%, respectively. The best linear model had an ROC AUC score of 0.75, with sensitivity and specificity of 84.6 and 71.7, respectively. (4) Conclusions: Our work demonstrates that ML makes it possible to utilize basic laboratory parameters to generate a distinct CA-related HF profile compared with CA-unrelated HF patients. This proof-of-concept study opens a potential new avenue in the diagnostic workup of CA and may assist physicians in clinical reasoning.
38
Llovet, Josep M., Amit G. Singal, Augusto Villanueva, et al. "Prognostic and predictive factors in patients treated with ramucirumab (RAM) with advanced hepatocellular carcinoma (aHCC) and elevated alpha-fetoprotein (AFP): Results from two phase III trials." Journal of Clinical Oncology 39, no. 15_suppl (2021): 4146. http://dx.doi.org/10.1200/jco.2021.39.15_suppl.4146.
Texto completoResumen
4146 Background: Elevated AFP in patients with aHCC is a poor prognostic factor with distinct molecular features, including high vascular endothelial growth factor (VEGF) signalling and increased angiogenesis. RAM, a human IgG1 monoclonal antibody, VEGF receptor 2 (VEGFR2) inhibitor, demonstrated improved survival vs placebo among patients with elevated AFP in the REACH-2 trial and is accepted as a standard of care for management of aHCC. We analyzed prognostic factors in patients with AFP ≥400 ng/mL and predictors of clinical benefit to RAM in an individual participant data (IPD) meta-analysis of the REACH and REACH-2 Phase III trials. Methods: Patients with aHCC, Child-Pugh A, ECOG performance status (PS) ≤1, and prior sorafenib were randomized (REACH 1:1; REACH-2 2:1) to RAM 8 mg/kg or Placebo Q2W. Meta-analysis was conducted in patients with AFP ≥400 ng/mL (n = 542). Univariate (UV) and multivariate (MV) analyses were performed using a Cox proportional hazard regression model. MV used the cut-off p-value < 0.1 from UV, irrespective of treatment arm. Overall survival (OS) was evaluated by Kaplan-Meier estimator and Cox models. To define predictors of RAM benefit, treatment-by-covariate interactions terms were evaluated. Results: In terms of prognosis assessed by MV analysis in patients with AFP ≥400 ng/mL, 6 variables among demographic and baseline disease characteristics were associated with poor OS in the RAM cohort (ECOG PS 1, AFP > 1000 ng/mL, Child-Pugh > A5, Extrahepatic site > 1, neutrophil-to-lymphocyte ratio > 3.2 and aspartate aminotransferase > 57 U/L) with an additional 3 factors identified within the whole cohort (macrovascular invasion presence, etiology HCV vs. Other and alkaline phosphatase ≥146). RAM benefit was present across all subgroups, including patients with very aggressive HCCs (AFP > 4000 ng/mL; HR: 0.64; 95% CI: 0.49-0.84) and those with nonalcoholic steatohepatitis /alcohol related aHCC (HR: 0.56; 95% CI: 0.40-0.79). Of note, two treatment-emergent (TE) events were the only factors that were significantly associated with improved RAM-related survival: TE-hypertension (p interaction = 0.0392) and TE-ascites (p interaction = 0.0001). However, these results should be interpreted with caution given that TE events are factors only observed after randomization. Conclusions: Several poor prognostic factors for OS were identified in patients with aHCC and elevated AFP. RAM provided an OS benefit irrespective of baseline prognostic covariates, with greater benefit observed in patients with aggressive HCC and those who experienced TE-hypertension or TE-ascites. Clinical trial information: NCT01140347; NCT02435433.
39
Wiegel, Thomas, Detlef Bartkowiak, Dirk Bottke, Alessandra Siegmann, Volker Budach, and Wolfgang Hinkelbein. "Prognostic significance of the PSA nadir after salvage radiotherapy following radical prostatectomy in prostate cancer." Journal of Clinical Oncology 33, no. 7_suppl (2015): 207. http://dx.doi.org/10.1200/jco.2015.33.7_suppl.207.
Texto completoResumen
207 Background: Salvage radiotherapy (SRT) is a curative approach in recurrent prostate cancer after radical prostatectomy. The outcome depends on various parameters. We report the long term results of SRT with special respect to the course of PSA after SRT. Methods: Between 1997 and 2007, 307 patients received SRT with 66.6 (N=240) or 70.2 Gy (N=67) using CT-based 3D planning. The median pre-SRT PSA was 0.297 ng/ml. Post-SRT progression was defined as either PSA rising >0.2 ng/ml above nadir, or hormone treatment, or clinical recurrence. Data were analyzed with the Kaplan-Meier method (Logrank-test) and with multivariate Cox regression. Results: Patients were followed up for median 7.2 (max. 14.4) years. Recurrence occurred in median 9.4 months post-RP. In 112 patients, SRT was administered before their PSA reached 0.2 ng/ml, 195 men were above that threshold. After SRT, 222 patients achieved a PSA nadir <0.1 ng/ml, 85 retained higher values. SRT given at a PSA <0.2 ng/ml correlated with achieving a post-SRT nadir <0.1 ng/ml (p<0.0001) and with improved freedom from progression (p=0.0133). Men with a post-SRT nadir <0.1 ng/ml (undetectable range) had significantly less recurrences (p<0.0001) and a better overall survival (p=0.0248). In multivariate analysis of pre-SRT parameters, pT≥3, Gleason Score ≥7, a post-RP PSA nadir ≥0.1 ng/ml and pre-SRT PSA ≥0.2 ng/ml increased the risk of progression. If failing the post-SRT nadir <0.1 ng/ml was included in the model, then this was the strongest risk factor (hazard ratio 7.93). Conclusions: Our data suggest early salvage RT at a PSA level below 0.2 ng/ml to be a favorable treatment option for post-RP PSA recurrence. It increases the chances of achieving a post-SRT PSA-nadir <0.1 ng/ml, which is associated with an improved outcome in terms of PSA progression and overall survival.
40
Seidel, Carina, Anders Sundan, Martin Hjorth, et al. "Serum syndecan-1: a new independent prognostic marker in multiple myeloma." Blood 95, no. 2 (2000): 388–92. http://dx.doi.org/10.1182/blood.v95.2.388.
Texto completoResumen
Serum samples drawn at diagnosis from 174 myeloma patients were analyzed for the presence of the heparin sulfate proteoglycan, syndecan-1. Syndecan-1 was elevated in 79% of patients (median, 643 units/mL) compared with 40 healthy controls (median, 128 units/mL),P &lt; .0001. Serum syndecan-1 correlated with the following: serum creatinine, secretion of urine M-component over the course of 24 hours, soluble interleukin-6 (IL-6) receptor, C-terminal telopeptide of type I collagen, β2-microglobulin, percentage of plasma cells in the bone marrow, disease stage, and serum M-component concentration. In order to evaluate syndecan-1 as a prognostic marker in multiple myeloma, it was entered into a multivariate Cox regression model. Data from 138 patients were available for this analysis. As a continuous variable, syndecan-1 was an independent prognostic parameter in addition to serum β2-microglobulin and World Health Organization performance status. When syndecan-1 was dichotomized by the best cutoff (66th percentile, 1170 units/mL), the survival difference between the groups was highly significant: “high” syndecan-1 group had a median survival of 20 months, and the “low” syndecan-1 group had a median of 44 months (P &lt; .0001). We conclude that syndecan-1 is a new independent prognostic parameter in multiple myeloma, and its role in prognostic classification systems should be further investigated.
41
Gershtein, Elena Sergeyevna, E. A. Korotkova, A. P. Petrosyan, E. A. Suleymanov, I. S. Stilidi, and N. E. Kushlinskii. "Prognostic significance of VEGF signaling system components and matrix metalloproteinases in blood serum of gastric cancer patients." Russian Clinical Laboratory Diagnostics 66, no. 11 (2021): 650–54. http://dx.doi.org/10.51620/0869-2084-2021-66-11-650-654.
Texto completoResumen
Analysis of long-term treatment results of 77 primary gastric cancer patients at stage I-IV of the tumor process followed during 1 - 41 months (median - 6.4 months) from the onset of specific treatment are presented depending on the basal levels of VEGF, soluble forms of its receptors (sVEGFR1, sVEGFR2) and matrix metalloproteinases (MMP-2, 7, 9) in blood serum. Overall survival assessed by Kaplan-Meyer analysis and with the help of Cox multiparametric regression model was applied as the criterion of prognostic value. It was found that at high (≥ 420 pg/ml) serum VEGF, the overall survival of patients with gastric cancer was statistically significantly lower than at the marker’s levels below 420 pg/ml (p<0.011): 3-year’s survival comprised 46,3±12,5% and 88,2±7,8% respectively. Median survival of patients with high VEGF level comprised 21.7 months, of those with low VEGF was not achieved during the whole follow-up period. Serum sVEGFR1, sVEGFR2, MMP-2, 7 and 9 levels were not significantly associated with the overall survival of patients included in this study. Only index M of TNM system and serum VEGF level demonstrated an independent prognostic value in multiparametric model (p=0.036). Thus, it was confirmed that VEGF signaling pathway plays an important role in gastric cancer, and its components - in the first place, VEGF A - are substantial factors of disease prognosis, and can also be useful for monitoring of treatment efficiency.
42
Alfraihat, Ausilah, Amer F. Samdani, and Sriram Balasubramanian. "Predicting curve progression for adolescent idiopathic scoliosis using random forest model." PLOS ONE 17, no. 8 (2022): e0273002. http://dx.doi.org/10.1371/journal.pone.0273002.
Texto completoResumen
Background Adolescent Idiopathic Scoliosis (AIS) is a three-dimensional (3D) spinal deformity characterized by coronal curvature and rotational deformity. Predicting curve progression is important for the selection and timing of treatment. Although there is a consensus in the literature regarding prognostic factors associated with curve progression, the order of importance, as well as the combination of factors that are most predictive of curve progression is unknown. Objectives (1) create an ordered list of prognostic factors that most contribute to curve progression, and (2) develop and validate a Machine Learning (ML) model to predict the final major Cobb angle in AIS patients. Methods 193 AIS patients were selected for the current study. Preoperative PA, lateral and lateral bending radiographs were retrospectively obtained from the Shriners Hospitals for Children. Demographic and radiographic features, previously reported to be associated with curve progression, were collected. Sequential Backward Floating Selection (SBFS) was used to select a subset of the most predictive features. Based on the performance of several machine learning methods, a Random Forest (RF) regressor model was used to provide the importance rank of prognostic features and to predict the final major Cobb angle. Results The seven most predictive prognostic features in the order of importance were initial major Cobb angle, flexibility, initial lumbar lordosis angle, initial thoracic kyphosis angle, age at last visit, number of levels involved, and Risser "+" stage at the first visit. The RF model predicted the final major Cobb angle with a Mean Absolute Error (MAE) of 4.64 degrees. Conclusion A RF model was developed and validated to identify the most important prognostic features for curve progression and predict the final major Cobb angle. It is possible to predict the final major Cobb angle value within 5 degrees error from 2D radiographic features. Such methods could be directly applied to guide intervention timing and optimization for AIS treatment.
43
Soof, Camilia M., Sameer Ashok Parikh, Susan L. Slager, et al. "Serum B-cell maturation antigen as a prognostic marker for untreated chronic lymphocytic leukemia." Journal of Clinical Oncology 37, no. 15_suppl (2019): 7525. http://dx.doi.org/10.1200/jco.2019.37.15_suppl.7525.
Texto completoResumen
7525 Background: New prognostic markers in chronic lymphocytic leukemia (CLL) are in demand. Different groups have developed models which combine multiple prognostic markers into a single index to classify CLL patients (pts). The CLL-International Prognostic Index (CLL-IPI) combines five parameters: age, clinical stage, TP53 status, IGHV mutational status, and serum β2 microglobulin levels. B-cell maturation antigen (BCMA) is a cell membrane receptor expressed exclusively on late stage B-cells and plasma cells with elevated serum (s) levels found in B-cell malignancies, such as multiple myeloma (MM). In MM, sBCMA levels can be used to monitor disease status and predict overall survival (OS). To further evaluate this biomarker in other hematologic malignancies, we studied it in CLL. Methods: Untreated (UNTX) CLL pts seen and consented at Mayo Clinic were identified. sBCMA levels were measured on stored sera of 331 UNTX CLL pts using an ELISA-based assay with a polyclonal anti-BCMA antibody from R&D Systems (Minneapolis, MN). The Mann-Whitney analysis was used to assess differences between CLL pts and healthy controls. The relationships between sBCMA and both time to first treatment (TFT) and OS were also assessed using Cox Regression models with an optimal sBCMA cutoff of 40.9 ng/mL. Results: The median age of pts was 61 years, and 71% were male. The distribution of CLL-IPI risk groups was as follows: 135 (41%) Low; 114 (34%) Intermediate; 67 (20%) High; 15 (5%) Very High. The median level of sBCMA in CLL pts (48.6 ng/mL) was higher (P <0.0001) than those of healthy controls (n = 104; 36.03 ng/mL). In CLL pts, sBCMA is significant in univariable analyses of TFT (HR 2.9 (95%CI, 2.0-4.2); P < 0.0001) and OS (HR 2.5 (95%CI, 1.5-4.0); P < 0.0003), and remains significant when adjusting for sex and CLL-IPI factors (HR 2.3 (95%CI, 1.6-3.3), P < 0.0001; HR 1.9 (95%CI 1.1-3.1), P = 0.01, respectively). Conclusions: sBCMA is elevated in CLL pts compared to healthy controls. After adjusting for CLL-IPI and sex, sBCMA levels provided independent prognostic value in predicting TFT and OS in this cohort. Measuring sBCMA with a readily accessible ELISA-based test, provides incremental value over the current CLL-IPI model in predicting prognosis of CLL.
44
Tefferi, Ayalew, Ruben A. Mesa, Jocelin Huang, et al. "Red Blood Cell Transfusion Requirement at Diagnosis Adversely Affects Both Overall and Leukemia-Free Survival in Primary Myelofibrosis - Increased Serum Ferritin or Total Transfusion Burden Does Not." Blood 112, no. 11 (2008): 5232. http://dx.doi.org/10.1182/blood.v112.11.5232.5232.
Texto completoResumen
Abstract BACKGROUND: In myelodysplastic syndromes (MDS) without excess blasts, red blood cell (RBC) transfusion requirement has been associated with poorer overall (OS) and leukemia-free (LFS) survival, suggesting that transfusion dependency is a marker of more severe disease (JCO2005;23:7594). We recently reported similar results in refractory anemia with ring sideroblasts (RARS): RBC transfusion requirement was an IPSS-independent adverse prognostic factor, but neither serum ferritin nor transfusion burden had prognostic value (AJH2008;83:611). Here, we examine the prognostic relevance of serum ferritin, need for RBC transfusions at time of diagnosis, and total transfusion burden in primary myelofibrosis (PMF). METHODS: We reviewed medical records to ascertain the clinical and laboratory features of 185 consecutive patients diagnosed with PMF according to the 2001 World Health Organization (WHO) criteria. Patients were excluded if ferritin measurements at time of diagnosis were unavailable. OS and LFS curves were constructed by Kaplan-Meier method, taking the interval from the date of diagnosis to death, leukemic transformation, or last contact. Log-rank test was used to test the homogeneity of survival curves over different groups. Cox proportional hazards model was utilized to determine the impact of various clinical and laboratory variables on OS and LFS. RESULTS: Clinical characteristics at diagnosis: Median age was 58 years (range 15–81; 110 males). Median serum ferritin level at diagnosis was 164 ng/mL (range 1–3903) and was ≥1000 ng/mL in 22 patients (12%). 101 (55%), 65 (35%) and 19 (10%) patients were assigned low, intermediate- and high-risk disease category (Blood1996;88:1013). In addition, 32 (17%) patients required RBC transfusions, 33% had constitutional symptoms and 39% displayed ≥ 1% circulating blasts. Where evaluated, 40% of the patients had cytogenetic abnormalities and 56% JAK2V617F. Events during the disease course: At a median follow-up of 28 months (range 0.5–231), 79 (43%) deaths and 15 (8%) leukemic transformations were documented. During this period, 126 (68%) patients required some form of therapy other than transfusion, including splenectomy in 33 (18%); only 4 underwent allogeneic stem cell transplantation. Causes of death were documented in 33 instances and none were attributed to iron overload. Median peak serum ferritin level during the disease course was 231ng/mL (range 9–13,080); 144 (78%) patients had peak levels &lt;1000 ng/mL, 28 (15%) between 1000 and 3000 ng/mL, and 13 (7%) above 3000 ng/mL. Number of total RBC transfusions ranged from none to 121. Iron chelation therapy was reported in 62 patients (34%). Prognostic factors for OS and LFS: Kaplan-Meier projected median survival was 72 months. By univariate analysis, increased serum ferritin level at diagnosis considered as either a continuous (p&lt;0.0001) or categorical (≥ 1000 ng/mL) variable (p&lt;0.0001), RBC transfusion requirement at diagnosis (p&lt;0.0001) and higher number of total RBC units transfused during the course of the disease (p=0.004) were all associated with inferior survival. However, only RBC transfusion requirement at diagnosis sustained its prognostic significance when age and conventional prognostic risk scores were added to the multivariable model as covariates. Similarly, a peak serum ferritin level of &gt; 3000 ng/mL documented during the disease course was not detrimental to survival. History of iron chelation therapy was associated with shortened survival (p=0.003). Multivariable analysis that included previously described risk factors for LFS also identified RBC transfusion requirement at diagnosis as an additional and independent risk factor. CONCLUSIONS: Serum ferritin level at time of diagnosis or during the disease course of PMF lacks independent prognostic value for either OS or LFS. The same is true for total transfusion burden. However, although hemoglobin &lt;10 g/dL is a component of all currently utilized prognostic scoring systems for PMF, the presence of a more severe erythropoietic defect as indicated by RBC transfusion need at time of diagnosis has an additional adverse prognostic value for both OS and LFS.
45
Raoof, Mustafa, Zeljka Jutric, Laleh Golkar Melstrom, et al. "Prognostic significance of chromogranin A in small pancreatic neuroendocrine tumors." Journal of Clinical Oncology 35, no. 4_suppl (2017): 375. http://dx.doi.org/10.1200/jco.2017.35.4_suppl.375.
Texto completoResumen
375 Background: Theincidence of non-functional pancreatic neuroendocrine tumors (PNETs) < 2 cm is rising. The biologic behavior of these tumors is variable and therefore their management remains controversial. We hypothesized that chromogranin A (CgA) levels are prognostic in these patients and may help guide management. Methods: Patients with PNETs measuring < 2 cm, without distant metastases were identified from the National Cancer Database (NCDB) over a ten-year period (2004-2014). Patients were categorized as CgA high (> 36.4 ng/ml) or CgA low (<36.4 ng/ml), and those lacking data on CgA levels were excluded from the study. Univariate and multivariate analyses were performed using Cox proportional hazards model. Results: Of the 445 eligible patients, 149 (33.5%) were CgA Low and 296 (66.5%) were CgA High. Median CgA level was 71 (Inter-quartile range, IQR 24-294) ng/ml. On multivariate analysis, CgA levels independently predicted overall survival after controlling for tumor size, grade, clinical nodal status and academic status of the facility (p = 0.001). At a median follow up of 26.5 months there were no deaths in the CgA Low group whereas 8% of the patients in CgA High group had died (p = 0.0068). Only CgA High patients benefited from surgical resection (HR 0.31, 95% CI 0.11-0.86, p = 0.025). Conclusions: Serum CgA levels can be incorporated in surgical decision making for patients with small PNETs. Patients with a high CgA should be strongly considered for resection.
46
Harrison, Rebecca A., Rongjie Liu, Vikram Rao, et al. "Evaluating the capacity of connectome analysis to predict survival in high-grade astrocytoma." Journal of Clinical Oncology 37, no. 15_suppl (2019): 2049. http://dx.doi.org/10.1200/jco.2019.37.15_suppl.2049.
Texto completoResumen
2049 Background: While factors such as age, histology and tumor molecular variants (e.g. IDH status) contribute to prognosis in patients with high grade astrocytoma (HGA), there remains a wide variability in patient survival outcomes. The connectome, or brain network organization, incorporates biologic, molecular and environmental processes providing a uniquely parsimonious summary of key prognostic factors. This study compared the capacity of machine learning (ML) models based on baseline connectomics and clinical variables to predict patient survival in HGA. Methods: Patients with a new diagnosis of HGA and a presurgical 3D, T1-weighted MRI available were retrospectively identified. Individual patient connectomes were derived from MRI with 90 cortical/subcortical features. Presurgical clinical features included age, gender, histology, tumor grade and IDH status. Three ML algorithms were implemented: extreme learning machine with Buckley–James estimator (ELMBJ), random survival forest (RSF) with logrank splitting and RSF with concordance index (CI) splitting. For each algorithm, we used a 60/40 training/testing split with 50 iterations and CI as the performance metric. We tested three models: 1) connectome only, 2) clinical only, and 3) connectome plus clinical variables. Results: Of patients identified (n = 105), 66 had glioblastoma and 39 had anaplastic astrocytoma. Thirty-eight harbored IDH mutation. Median overall survival was 27.43 months (SD 39.57). Connectome-only models showed better prediction performance compared to clinical-only models across all algorithms. ELMBJ showed the best performance (connectome median CI = 0.522, clinical CI = 0.201). Connectome models also performed as well as combined models (e.g. median CI = 0.523 for ELMBJ). Conclusions: This study demonstrates the potential of a connectome model to predict survival of patients with HGA. Replication in a larger sample is required to validate these results and refine ML models including examination of additional clinical features. If successful, use of a simple T1 MRI could provide additional variables to augment existing prognostic prediction, especially in scenarios where tumor genotyping is not available.
47
Han, Liz Y., Charles N. Landen, Aparna A. Kamat, et al. "Preoperative Serum Tissue Factor Levels Are an Independent Prognostic Factor in Patients With Ovarian Carcinoma." Journal of Clinical Oncology 24, no. 5 (2006): 755–61. http://dx.doi.org/10.1200/jco.2005.02.9181.
Texto completoResumen
Purpose Tissue factor (TF) is a procoagulant that plays an important part in tumor angiogenesis. We sought to determine the role of preoperative serum TF levels in predicting clinical outcome in patients with ovarian cancer. Materials and Methods TF expression was determined by reverse transcriptase polymerase chain reaction in ovarian cell lines. Using enzyme-linked immunosorbent assay, we assessed preoperative serum TF levels in 98 women with invasive epithelial ovarian carcinoma, 30 with low malignant potential (LMP) tumors, 16 with benign tumors, and a separate validation group of 39 women with adnexal masses. Clinical information was gathered from chart review. Results TF was expressed in four of the five ovarian cancer cell lines, but absent in the nontransformed cells. Ovarian cancer patients had a median preoperative serum TF level of 85.2 pg/mL, which was significantly higher than in those with LMP tumors (12.8 pg/mL; P < .01) and benign adnexal disease (30.7 pg/mL; P < .01). TF ≥ 190 pg/mL was significantly associated with decreased patient survival (P < .01). After adjusting for other clinical variables in a multivariate Cox regression model, TF ≥ 190 pg/mL was an independent prognostic factor (P < .01). Analysis of serum TF levels from the validation set confirmed that high TF (≥190 pg/mL) was associated with a 3.4-fold increase in risk of death from disease (P = .02) and shorter survival (P = .01). Conclusion Preoperative serum TF levels are significantly elevated in patients with ovarian carcinoma. Elevated preoperative TF level is an independent prognostic factor for death from disease.
48
Céruse, Philippe, Muriel Rabilloud, Anne Charrié, Christian Dubreuil, and François Disant. "Study of Cyfra 21–1, a Tumor Marker, in Head and Neck Squamous Cell Carcinoma." Annals of Otology, Rhinology & Laryngology 114, no. 10 (2005): 768–76. http://dx.doi.org/10.1177/000348940511401006.
Texto completoResumen
Objectives: We performed a prospective study to determine the cutoff value and the prognostic value of Cyfra 21–1, a serum tumor marker, in head and neck squamous cell carcinoma (HNSCC). Methods: The serum concentration of Cyfra 21–1 was measured in a group of 300 patients (group 1) with HNSCC, in a control group of 71 healthy subjects (group 2), and in a group of 73 patients with a nonmalignant tumor or inflammatory disease (group 3). The concentrations were compared between the various groups and subgroups; the cutoff value was calculated with a receiver operating characteristic curve. Furthermore, the serum concentrations of Cyfra 21–1 before treatment in the group of 300 patients were compared with the stage of the disease and with the evolution of the overall survival rate and the disease-free survival rate. Finally, to determine whether Cyfra 21–1 is an independent prognostic factor, we compared the concentrations, by a Cox model, with the classic prognostic factors of HNSCC. Results: At the cutoff value of 1 ng/mL, the specificity was 94% and the sensitivity was 72%. The serum concentrations of Cyfra 21–1 were statistically correlated with the stage of the disease. The overall survival rate and the disease-free survival rate were lower in patients with high serum concentrations, and these differences were statistically significant (p <.001). The Cox model allows us to conclude that Cyfra 21–1 is a prognostic marker that is independent of other classic prognostic factors. Conclusions: Cyfra 21–1 is an interesting tumor marker that could be proposed for the early detection of HNSCC with a cutoff value of 1 ng/mL. Furthermore, Cyfra 21–1 can be considered an independent prognostic marker.
49
Rujirojindakul, Pairaya, and Arnuparp Lekhakula. "Prognostic Significance of Serum Proangiogenic Molecules in Patients withDe NovoNon-Hodgkin Lymphomas." Scientific World Journal 2012 (2012): 1–5. http://dx.doi.org/10.1100/2012/215231.
Texto completoResumen
This study was aimed to assess the clinical significances of the serum VEGF and bFGF in Thai patients withde novoNHL. Serum VEGF and bFGF concentrations were measured from 79 adult patients with newly diagnosed stage 2–4 non-Hodgkin lymphomas by quantitative sandwich enzyme immunoassay. At the time of diagnosis, the serum VEGF concentrations from 79 patients ranged from 72.0 to 2919.4 pg/mL, with a mean of 668.0 pg/dL. The serum bFGF concentrations ranged from undetectable to 2919.4 pg/mL, with a mean of 12.15 pg/dL. Multivariate analysis identified higher than the mean of serum VEGF, B symptoms, bulky diseases, anemia, and treatment with CHOP or R-CHOP as independent variables influencing the complete remission rate. From a Cox proportional hazards model, variables independently associated with overall survival were bone marrow involvement, more extranodal involvement, poor performance status, anemia, and higher than the mean of serum bFGF.
50
Sarfati, M., S. Chevret, C. Chastang, et al. "Prognostic importance of serum soluble CD23 level in chronic lymphocytic leukemia." Blood 88, no. 11 (1996): 4259–64. http://dx.doi.org/10.1182/blood.v88.11.4259.4259.
Texto completoResumen
Abstract Prognosis of B-cell chronic lymphocytic leukemia (CLL) is based on clinical staging whose limitation is the failure to assess whether the disease will progress or remain stable in early stage (Binet A, or Rai 0, I, II) patients. We previously reported that soluble CD23 (sCD23), a protein derived from the B-cell membrane CD23 Ag, is selectively elevated in the serum of CLL patients. This prospective study assessed the predictive value of serum sCD23 level measured at study entry on the overall survival of all CLL patients and on disease progression of stage Binet A patients. Prognostic value of repeated measurements of sCD23 over time in stage A patients was also analyzed. One hundred fifty-three CLL patients were prospectively followed with a median follow-up of 78 months. Eight clinical or biological parameters were collected from the date of the first sCD23 measurement. At study entry, by Cox model, Binet staging (P = .0001) and serum sCD23 level (P = .03) appeared as prognostic factors for survival. Patients with sCD23 level above median value (> 574 U/mL) had a significantly worse prognosis than those with lower values (median survival of 53 v 100+ months, P = .0001). During follow-up, sCD23 doubling time increased by 3.2 the risk of death (P = .001). Among stage A patients (n = 100), sCD23 determination at study entry was the sole variable predictive of disease progression, patients with sCD23 level above 574 U/mL had a median time progression of 42 months versus 88 months for those with lower levels (P = .0001). Stage A patients who doubled their sCD23 level exhibited a 15-fold increased risk of progression (P = .0001) and, in addition, the sCD23 increase preceded by 48 months disease progression. We conclude that in CLL patients, serum sCD23 level provides significant additional prognostic information in terms of overall survival. Most interestingly, among early stage patients, sCD23 determination at diagnosis and during the course of the disease may help to the early identification of patients who will rapidly progress to upper stages.