| Part ? Exosomal lncRNA ENST00000462352 mediates radioresistance of esophageal cancer and combines omics data to build novel model for predicting radiotherapy efficacySection ? The role and mechanism of exosomal lncRNA ENST00000462352 in the transmission of radioresistance for esophageal cancerEsophageal cancer(EC)is a common malignant gastrointestinal tumor with the seventh highest incidence rate and the sixth highest mortality rate in worldwide.Radiation therapy(RT)plays a vital role in multimodality treatment strategies for EC.Due to the tumor biological characteristic of radioresistance,uncontrolled local disease and regional recurrence have become the main reasons to lead to RT failure,which seriously harmed patients' treatment efficacy and prognosis.Therefore,effectively curbing radioresistant phenotype is really a major challenge in current clinical practice of EC.Long non-coding RNA(lncRNA)is a type of RNA with a transcript length of greater than 200 nucleotides and does not encode proteins.LncRNA due to sequence or structure diversity could interact with DNA,RNA or protein,and regulate gene expression at three levels of epigenetics,transcription or post-transcription,thereby affecting the biological characteristics of tumor cell proliferation,invasion,metastasis and so on.The latest evidences have revealed its critical regulatory potential in tumor radioresistance.A variety of lncRNAs that were closely related to EC radioresistance have been found,such as lncRNA DNM3OS,FAM201A,POU5F1B,TUG1,etc..Meanwhile,the underlying mechanisms on their induction of radioresistance have been preliminary discussed.However,more novel lncRNAs have yet to be discovered to further explore the predictive markers of diagnosis and treatment as well as clarify the key therapeutic targets and signaling pathways of EC radioresistance.Exosomes are tiny membrane vesicles with a lipid bilayer membrane structure and a diameter of 30-200nm,which can be secreted by living cells.Exosomes participate in the exchange of information between cells through carrying and transporting proteins,RNA,DNA and other substances.They have significantly relationships with various biological processes of tumor progression.Notably,exosomes have also important functions in inducing tumor radioresistance.Numerous evidences supported that exosomes as signal substance carriers could enrich lncRNAs and take a key regulatory part in the transmission of radioresistance between cells.However,few studies regarding the roles and mechanisms of exosomal lncRNAs in mediating EC radioresistance has been conducted.Objective:This study was firstly to screen and identify exosomal lncRNAs related to EC radioresistance,and secondly to demonstrate the effect of candidate lncRNA on EC radioresistance.Besides,it further confirmed that whether exosomes could transfer candidate lncRNA to mediate the transmission of EC radioresistance.Finally,it preliminarily clarified the mechanism of candidate lncRNA in regulating radioresistance of EC.Overall,we aimed to discover novel markers in predicting RT response of EC,and clarify the potential therapeutic targets and mechanisms to overcome radioresistance.Methods:1.Screening and verification of exosomal lncRNAs related to radioresistance of EC:We prospectively collected plasma from EC patients before RT.According to treatment response outcomes,complete response(CR)and partial response(PR)were defined as radiosensitivity,while stable disease(SD)and progression disease(PD)were regarded as radioresistance.The plasma from 5 patients with radiosensitivity and radiores istance were randomly selected.The exosomes were extracted by ultra-centrifugation,and then identified by transmission electron microscopy(TEM),particle size analysis and western blot.The plasma exosomes samples were used for transcriptome sequencing.Based on bioinformation,lncRNAs which were potentially related to radioresistance were distinguished and verified by quantitative real-time polymerase chain reaction(qRT-PCR).The value of candidate lncRNA(ENST00000462352)in predicting RT efficacy was evaluated by the area under curve(AUC).2.LncRNA ENST00000462352 promotes radioresistance of EC:The radioresistant KYSE-30R/150R cells were constructed.Their radioresistance were successfully demonstrated through clone formation assay,CCK-8 assay,and cell cycle and apoptosis analysis by flow cytometry.The overexpression or interference lncRNA ENST00000462352 and negative control lentiviral vectors were built and stably transfected into KYSE-30/R,KYSE-150/R cells.The infection effects were verified by PCR.The influences of overexpression or interference lncRNA ENST00000462352 on radioresistance were observed in vitro based on clone formation assay,CCK-8 assay,and cell cycle and apoptosis analysis by flow cytometry.Besides,nude mouse subcutaneous xenograft tumor models were established using K Y S E-30R/150R cells with stable transfection of interference lncRNA ENST00000462352 and negative control.After RT,the tumors growth volumes and weights were recorded.The influence of lncRNA ENST00000462352 on tumor radioresistance could be investigated in vivo.3.Exosome transfers lncRNA ENST00000462352 to mediate the transmission of radiation tolerance in EC:The FISH probe,RNA nucleocytoplasmic separation and PCR assays were used to determine the existence pattern of intracellular and extracellular lncRNA ENST0000462352.Using PKH67 dye to label the exosomes isolated from radioresistant KYSE-30R/150R cells,and co-culturing them with radiosensitive KYSE-30/150 cells to observe whether the labeled exosomes could be uptaken by recipient cells.PCR was used to demonstrate the expression levels changes of lncRNA ENST00000462352 in the cytoplasm of recipient cells.The exosomes from radioresistant or radiosensitive cells were co-cultivated with radiosensitive cells,respectively.The recipient cells' radioresistance under different intervention conditions were observed through clone formation assay,CCK-8 assay,and cell cycle and apoptosis analysis by flow cytometry.Moreover,the exosome inhibitor GW4869 was used to further confirm the intervention effect.The nude mouse subcutaneous xenograft tumor models with radiosensitive KYSE-30/150 cells were established.They were dealt with exosomes of radioresistant KYSE-30R/150R cells and radiosensitive KYSE-30/150 cells.The growth volume and weight of tumors were recorded after RT.4.The preliminary mechanism of lncRNA ENST00000462352 targets the CyclinDl/E1-pRb-E2F1 signaling pathway through miR-424-5p to modulate radioresistance of EC:Based on the bioinformatics algorithm,it predicted that lncRNA ENST00000462352 might bind to miRNA-424-5p,and the dual luciferase reporter would be used for confirmation.According to Targetscan database and previous literatures,miRNA-424-5p targeted the downstream protein CyclinDl/E1 and regulated the signaling pathway of CyclinDl/E1-pRb-E2F1.PCR and western blot were applied to observe the relationships among lncRNA ENST00000462352,miRNA-424-5p,CyclinDl/E1,pRb and E2F1.Results:1.The exosomes were successfully verified by TEM,particle size analysis,and CD63 and CD81 proteins detection.According to sequencing results,the top 10 lncRNAs closely related to radioresistance,including ENST00000462352,ENST00000469143,ENST00000470017,ENST00000480354,ENST00000531702,ENST00000606993,ENST00000602436,ENST00000609090,ENST00000624324 and ENST00000624705 were chosen for verification.The results showed that compared with other lncRNAs,lncRNA ENST00000462352 had the most significant difference between radioresistant and radiosensitive patients.The expression level of exosomal lncRNA ENST00000462352 in radioresistant cases was significantly higher than that of in radiosensitive cases(P=0.001).Additionally,the AUC of exosomal lncRNA ENST00000462352 in predicting RT efficacy was 0.729(95%CI 0.618-0.840),In summary,lncRNA ENST00000462352 was strongly correlated with EC radioresistance and determined as object for follow-up research.2.In KYSE-30/R and KYSE-150/R cells stably transfected with overexpression or interference IncRNA ENST00000462352 and negative controls,according to clone formation assay,CCK-8 assay,and cell cycle and apoptosis analysis by flow cytometry,the results showed that after irradiation,compared with control groups,the overexpression groups of lncRNA ENST00000462352 enhanced clonal formation ability,increased cell viability,decreased apoptotic ratio,increased S phase cells ratio,but decreased G0/G1 phase cells ratio(P<0.05),obviously exhibiting radioresistance phenotype.However,the interference lncRNA ENST00000462352 groups greatly weakened the cells resistance to RT.The opposite results were observed(P<0.05).Moreover,animal experiments revealed that compared with the control group,the interference lncRNA ENST00000462352 group sharply reduced radioresistance but increased radiosensitivity.This resulted in slowing tumor proliferation after RT,which markedly reduced volumes(P=0.009)and weights(P=0.004).3.FISH probe and RNA nucleocytoplasmic separation assays confirmed that lncRNA ENST00000462352 was mainly highly expressed in cytoplasm.PCR was used to determine the expression levels of lncRNA ENST00000462352 in cells supernatant,supernatant exosomes and supernatant after exosomes removal.The results found that exosomes were the main existence pattern of extracellular lncRNA ENST00000462352.Moreover,the expression level of exosomal lncRNA ENST00000462352 in radioresistant cells was significantly higher than that of in radiosensitive cells(P<0.05).4.After the exosomes of radioresistant KYSE-30R/150R cells being co-incubated with radiosensitive KYSE-30/150 cells,we found that the exosomes could be effectively taken up by the recipient cells.Meanwhile,the expression level of lncRNA ENST00000462352 in the cytoplasm of recipient cells was significantly increased(P<0.05).5.The radioresistant or radiosensitive cells exosomes were co-cultured respectively with radiosensitive cells.Based on clone formation assay,CCK-8 assay,and cell cycle and apoptosis analysis by flow cytometry,the results showed that compared with the control group,the recipient cells dealt with exosomes isolated from radioresistant cells have significantly improved resistance to RT after irradiation,resulting in superior clonal formation ability,improved cell viability,decreased apoptotic ratio.Moreover,the proportion of S phase cells was increased,while the proportion of G0/G1 and G2/M phases cells was decreased(P<0.05).After using the GW4869 inhibitor to block the secretion of exosomes from radioresistant cells,the radioresistance of recipient cells could be effectively reversed and the above intervention effect disappeared.The results of animal experiments reported that after irradiation,compared with the radiosensitive cells exosomes intervention group,the radioresistant cells exosomes intervention group effectively developed the recipient cells radioresistance,resulting in accelerating tumor proliferation with significantly increased tumor size(P=0.031)and weight(P=0.001).6.The dual luciferase reporter confirmed the interaction between lncRNAENST00000462352 and miRNA-424-5p.The results of PCR and western blot consistently found that in KYSE-30/R cells,compared to control,overexpression lncRNA ENST00000462352 would reduce miRNA-424-5p expression level,but increase CyclinDl/E1,pRb and E2F1 expression level.Interference IncRNA ENST00000462352 increased the expression level of miRNA-424-5p,but lowered the expression level of CyclinD1/E1,pRb and E2F1.In addition,compared with non-irradiation,the miR-424-5p levels would increase to varying degrees,while the CyclinDl/E1,pRb and E2F1 levels decreased at different extent after irradiation.Conclusions:The potential of lncRNA ENST00000462352 as a new predictive marker of RT efficacy was firstly clarified,and its role in EC radioresistance were confirmed.We initially demonstrated that the exosomes could transfer lncRNA ENST00000462352 to mediate the transmission of radiation tolerance in EC.The potential molecular mechanism of lncRNA ENST00000462352 on affecting EC radioresistance was through miR-424-5p to target and regulate the cell cycle signaling pathway of CyclinD1/El-pRb-E2F1.Overall,this study indicated that targeting exosomes,IncRNA ENST0000046232 and signaling pathway of CyclinDl/E1-pRb-E2F1 or combined with RT might be able to effectively overcome EC radioresistance.Section ? Exosomal lncRNA ENST00000462352 combines omics data to build novel model for predicting radiotherapy efficacy of esophageal cancerIn section 1.we confirmed the value of exosomal lncRNA ENST00000462352 in predicting RT efficacy for EC patients.Objectively speaking,its evaluation performance was still limited.More new markers with high sensitivity and specificity needed to be discovered.The internal biological characteristics of tumors such as proliferation,necrosis,hemorrhage,hypoxia and other microscopic phenotypes were strongly correlated with poor treatment response.The latest evidences showed that radiomics could process traditional medical images into a set of high-dimensional and quantitative features.Furthermore,these features objectively and accurately quantified the above phenotypes,and were expected to become new promising markers for predicting treatment efficacy.However,the researches on computed tomography(CT)-based radiomics applied to evaluate RT response for EC were extremely limited,and its predictive activity remained to be investigated.Additionally,the occurrence and development of EC were complex and heterogeneous.Considering the multidimensional information from diverse tumors biological characteristics should be intimate and complementary,and their integrated application might greatly improve the predictive power of RT efficacy.However,it still needed to carry out in-depth exploration.Objective:On one hand,this study intended to prove the value of radiomics feature extracted from CT in monitoring RT response of EC.On the other hand,the radiomics features,exosomal lncRNA ENST00000462352 and other clinical data significantly related to RT efficacy were effectively combined to innovatively develop models,aiming to improve performance for predicting RT response and guide the formulation of individualized treatment strategies for EC.Methods:86 EC patients receiving RT were prospectively enrolled from our hospital.The esophageal lesions were delineated as regions of interest(ROIs)on the diagnostic chest enhanced CT before RT,and then extracted radiomics features.In order to reduce operator bias,two delineations were carried out successively with an interval of 2 months.Intra-class correlation coefficient(ICC)was used to compare the stability of these two sets of features.The ICC>0.8 features were chosen for further study.The least absolute shrinkage and selection operator(LASSO)algorithm was applied to distinguish radiomics features potentially related to RT efficacy,and generate radiomics signature.AUC was computed to evaluate the ability of radiomics signature for predicting RT response.Univariate and multivariate logistic regression analyses were used to demonstrate the potential relationships between candidate indicators and RT efficacy.The logistic regression classifier models were built utilizing the parameters closely associated with the effect of RT.The models'prediction performance for RT efficacy were observed by AUC.Results:Totally,850 CT-based radiomics features were extracted.7 features which were strongly associated with RT response were chosen,including original_shape_Maximum2DDiameterColumn?original_glrlm_ShortRunLowGrayLevelEmphasis?wavelet_LHH_glcm_Imc1?wavelet_HLH_gldm_LargeDependenceHighGrayLevel Emphasis?wavelet_HLH_gldm_SmallDependenceLowGrayLevelEmphasis?wavelet_HLH_firstorder_TotalEnergy?wavelet_LLL_firstorder_Skewness.The radiomics signature was built by using the above-mentioned features,which its AUC in predicting RT response was 0.807(95%CI 0.710-0.904).Univariate logistic regression analysis showed that radiomics signature,exosomal lncRNA ENST00000462352,tumor grade,length,TNM stage,platelet-lymphocyte ratio(PLR),neutrophils-lymphocyte ratio(NLR)and lymphocyte-monocyte ratio(LMR)were potential predictors of RT efficacy.Multivariate logistic analysis found that tumor length(P=0.008,OR 0.129,95%CI 0.028-0.589),exosomal lncRNA ENST00000462352(P=0.047,OR 0.246,95%CI 0.061-0.981)and radiomics signature(P<0.001,OR 36.397,95%CI 4.935-268.428)were independent predictors of RT efficacy.According to the results of univariate and multivariate logistic analy ses,the optimal indicators were selected for fitting.4 logistic regression classifiers were built to predict the efficacy of RT,which were defined as the traditional clinical model(integration of tumor length and TNM stage),the lncRNA ENST00000462352 model(integration of exosomal lncRNA ENST00000462352,tumor length and NLR),the radiomics model(integration of radiomics signature and tumor length)and the multidimensional omics data model(integration of radiomics signature,tumor length and exosomal lncRNA ENST00000462352).The corresponding AUC of these models was 0.726(95%CI 0.617-0.836)?0.847(95%CI 0.758-0.937)?0.880(95%CI 0.804-0.956)?0.894(95%CI 0.827-0.960),respectively.Compared with the other 3 models,the novel multidimensional omics data-based logistic classifier model,which comprehensively integrated tumor information at different levels,exhibited the strongest efficacy assessment ability and had excellent clinical practical value.Conclusions:The CT-based radiomics signature was developed and confirmed to be a powerful novel imaging marker for predicting RT response.We creatively merged radiomics signature,exosomal lncRNA ENST00000462352,tumor length,TNM staging and NLR to construct intelligent and individualized logistic classifiers for forecasting RT efficacy.These models were precious and convenient tools for early identification of RT response,and provided important technical support for effectively improving clinical treatment effects of EC.Part ? Developing multidimensional information-based Nomogram model and risk classification system using machine learning to predict survival for esophageal cancer patients receiving chemoradiotherapyFor unresectable locally advanced EC,radical concurrent chemoradiotherapy(CRT)was the standard treatment,but the clinical outcomes were not satisfactory.About half of patients have uncontrolled local diseases and recurrence,and the 5-year survival rate was only 20%.It might be beneficial to distinguish people at high risk of recurrence or death at an early stage and then take active interventions.At present,the TNM staging system was the most used method for evaluating tumors prognosis.However,only relying on the traditional and simple TNM staging strategy to judge prognosis was one-sided,and ignored individual heterogeneity as well.Eagerly exploring novel multidimensional prognostic markers has broad research prospects.In recent years,the role of radiomics in the diagnosis and treatment of EC has attracted much attention.Radiomics provided strong technical support for questing prognostic markers,but it still needed to be fully demonstrated its clinical application value.Additionally,considering the complex and heterogeneity of EC biological characteristics,the effective incorporation of multidimensional information could improve the potential for prognosis evaluation.Objective:This study firstly verified the significance of dynamically changing CT radiomics features during RT in monitoring EC prognosis,and found innovative imaging markers to efficiently predict survival.Secondly,combining the radiomics features,clinicopathological parameters and hematological indicators that were closely associated with EC prognosis to build multidimensional information-based nomogram models and risk classification systems.We aimed to individually and accurately predict survival and powerfully discern patients with different prognostic risks.Furthermore,targeted treatment strategies could be formulated for improving survival.Methods:298 patients with EC from two independent medical institutions were enrolled,and divided into the training cohort(n=168)and the validation cohort(n=130).The primary tumors of esophagus were regarded as ROIs.A total of 850 radiomics features were extracted from CT images of RT planning.The longitudinal dynamic changes of these features between RT positioning and resetting were further calculated.In order to reduce bias,30 patients were randomly selected for re-delineation at 2 months after the initial delineation.ICC was computed based on the two delineation results.The robust features of ICC>0.8 were used for further analysis.LASSO algorithm was applied to select the delta radiomics features(?RF)closely related to overall survival(OS)and progression-free survival(PFS),and then construct radiomics signature for forecasting survival.According to univariate and multivariate Cox regression analyses,the potential high-risk factors for OS and PFS were clarified.Reference to the results of multivariate analysis,multidimensional information-based OS and PFS related nomogram models were developed through integrating valuable prognostic data from radiomics signature,clinicopathological and hematological parameters.The clinical application value of these models was evaluated by concordance index(C index),calibration curve and decision curve analysis(DCA)in both the training and validation cohorts.The net reclassification index(NRI)was used to compare the power of nomogram models and traditional clinical model in predicting survival.A risk classification system was developed by recursive partitioning analysis(RPA).Kaplan-Meier curves were utilized to demonstrate the performance of this system to distinguish patients with different prognostic risks.Results:8 ?RF were closely related to OS.7 ?RF were strongly associated with PFS.OS and PFS related radiomics signatures were generated using these features.In the training cohort,we found that the radiomics signature of patients with longer-term OS was significantly higher than that of patients with shorter-term OS(0.29±0.34 vs-0.17±0.36,P<0.001).Similarly,the radiomics signature of longer PFS cases was also obviously higher than that of shorter PFS cases(0.28±0.35 vs-0.07±0.26,P<0.001).The results were further confirmed in the validation cohort.In short,the conclusions revealed that radiomics signatures were potentially correlated with OS and PFS.The multivariate Cox analysis showed that tumor grade,length,TNM stage,the change ratio of systemic immune inflammation index pre-and post-treatment(?S?)and radiomic signature were independent risk factors for OS and PFS(P<0.05).Subsequently,the nomogram prediction models for OS and PFS were constructed by integrating the above factors using machine learning.In the training cohort,the C index of OS and PFS nomogram models were 0.951(95%CI 0.922-0.981)and 0.902(0.857-0.947),respectively.In the validation cohort,the C index of nomogram models for predicting OS and PFS were 0.917(95%CI 0.866-0.968)and 0.922(0.873-0.972),respectively.Calibration curve and DCA confirmed the clinical feasibility and practicality of models.The NRI proved that nomogram models were more superior than traditional clinical model in terms of OS and PFS monitor.Moreover,the Kaplan-Meier analysis showed that the risk classification system had excellent power,and could divide patients into different prognosis risk subgroups(P<0.001).Conclusions:The radiomics signatures developed by the dynamic changes of ?RF during RT were novel quantitative imaging markers for OS and PFS.For the first time,the radiomics signatures,clinicopathological and hematological indicators were comprehensively combined using machine learning to creatively construct OS and PFS nomogram models that individually evaluated survival,and generate risk classification system that efficiently distinguished different prognostic subgroups.These models might be important auxiliary tools to guide individualized management for EC in clinical practice.Part ? Building an intelligent Nomogram model and risk classification system based on multimodal omics data to predict radiation pneumonitis in patients with esophageal cancerRadiation pneumonitis(RP)was a common dose limiting adverse event of thoracic RT,which could result in diminished treatment efficacy,impaired quality of life and even life threatening.Various groups have previously reported that lung dosimetric parameters and patient-specific characteristics were high-risk contributors to RP onset,but the findings were still controversial.New markers needed to be explored for effectively monitoring RP.The assessment criteria of RP in current clinical practice were qualitative and subjective.It would be of great value and necessity to identify RP based on quantitative measurements.Recently,the development of radiomics brought dawn to quantify clinically significant RP.What's more,the integration analysis of multimodal data from diverse dimensions might be the most prospective approach to precisely predict RP.Objective:In this study,we aimed to evaluate the capability of CT-based dynamic radiomics features in characterizing lungs damage resulted from RT,and demonstrated that the radiomics signature built by integrating multidimensional features could be novel quantitative imaging marker to forecast severe acute radiation pneumonitis(SARP).Besides,incorporating multimodal data potentially related to SARP,such as radiomics signature,clinicopathological,hematological and dosimetric parameters to construct a nomogram model and risk classification system.Based on these simple and easy-to-operate intelligent tools,the purposes were to individually predict SARP and accurately divide patients into different risk groups.Furthermore,the most appropriate intervention measures and follow-up strategies could be formulated for heterogeneous entity.Methods:400 EC patients were enrolled from two independent medical institutions and divided into the training(n=200)and validation cohorts(n=200).850 radiomics features of lungs were extracted from positioning and resetting CT images of RT planning.The longitudinal dynamic changes in this set of radiomics features during RT were further calculated.40 cases were randomly selected to re-delineate and extract radiomics features at 2 months after the initial delineation.These two series of features data from the 40 patients were used to calculate ICC.Only the robust features with ICC>0.8 were applied in subsequent analysis.LASSO algorithm was performed to select valuable ?RF related to SARP and then build radiomics signature.The multimodal data which were radiomics signature,clinicopathological,dosimetric and hematological predictors were combined to establish a nomogram prediction model for SARP based on multivariate analysis.The predictive performance and clinical application value of nomogram were both evaluated in the training and validation cohorts by C index,calibration curves and DCA.RPA algorithm was utilized to generate a risk classification system regarding SARP.Results:24 ?RF were significantly associated with SARP status(P<0.001),and were applied to construct a radiomics signature.In the training and validation cohorts,the radiomics signatures of SARP patients were both evidently higher than those of patients without SARP(0.02±1.06 vs-2.10±1.06,P<0.001;-0.34±1.02 vs-2.17±1.23,P<0.001),perfectly demonstrating their power to distinguish SARP.The multivariate analysis showed that radiomics feature(OR 196.366;95%CI 29.934-1288.165;P<0.001),subjective global assessment score(SGA;OR 11.107;95%CI 1.928-63.998;P=0.007;OR 30.869;95%CI 4.087-233.139,P=0.001),pulmonary fibrosis score(PFS;OR 6.921;95%CI 1.358-35.273;P=0.020),mean lung dose(MLD;OR 9.383;95%CI 2.176-40.463;P=0.003)and the change ratio of systemic immune inflammation index at 4 weeks during RT ?4w S?;OR 16.437;95%CI 2.551-105.901;P=0.003)are independent predictors of SARP.The above-mentioned multimodal data were merged to develop an intelligent nomogram for predicting SARP.The C index of model was 0.975(95%CI 0.953-0.996)and 0.921(95%CI 0.876-0.966)in the training and validation cohorts,respectively,which effi ciently discerning SARP occurrence.Calibration curves and DCA confirmed the satisfactory clinical feasibility and utility of nomogram.The risk classification system displayed remarkable performance in the stratification of SARP subgroups(P<0.001).Conclusions:It was initially confirmed that CT-based radiomics signature was a promising non-invasive imaging marker for predicting SARP.Integrating radiomics signature,clinicopathological,hematological and dosimetric parameters to innovatively construct multimodal intelligent nomogram model and risk stratification system,aiming to individually and accurately predict SARP and effectively classify people into different risk levels.These tools could guide clinical decision making and provide consulting services. |
PDF Full Text Request