Application of a postnatal prediction model of survival in CDH in the era of fetal therapy.
Clohse K,Rayyan M,Deprest J,Decaluwe H,Gewillig M,Debeer A
The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstetricians
The disease severity in patients with a congenital diaphragmatic hernia (CDH) is highly variable. To compare patient outcomes, set up clinical trials and come to severity-based treatment guidelines, a performant prediction tool early in neonatal life is needed. The primary purpose of this study was to validate the CDH study group (SG) prediction model for survival in neonates with CDH, including patients who had fetal therapy. Secondary, we aimed to assess its predictive value for early morbidity. This is a retrospective single-center study at the University Hospitals Leuven on all infants with a diagnosis of CDH live-born between April 2002 and December 2016. The prediction model of the CDHSG was applied to evaluate its performance in determining mortality risk. Besides, we examined its predictive value for early morbidity parameters, including duration of ventilation, respiratory support on day 30, time to full enteral feeding and length of hospital stay. The CDHSG prediction model predicted survival well, with an area under the curve of 0.796 (CI: 0.720-0.871). It had poor value in predicting infants who needed respiratory support on day 30 (area under the curve (AUC) 0.606; CI: 0.493-0.719), and correlated poorly with duration of ventilation, time to full enteral feeding and length of hospital stay. The CDHSG prediction model was in our hands also a useful tool in predicting mortality in neonates with CDH in the fetal treatment era. Correlation with early morbidity was poor. (1) Validation of the CDHSG prediction model for survival in a cohort of neonates with CDH, in whom fetal endoscopic tracheal occlusion was applied according to the severity of lung hypoplasia. (2) Evaluation of performance of the model in the prediction of early morbidity. (1) Confirmation of the predictive value of the model for survival in neonates with CDH in the era of fetal therapy. (2) No correlation of the model with early morbidity parameters.
Prediction of pre-eclampsia-related complications in women with suspected/confirmed pre-eclampsia: development and internal validation of a clinical prediction model.
Saleh L,Alblas M,Nieboer D,Neuman R,Vergouwe Y,Brussé I,Duvekot J J,Steyerberg E W,Versendaal H J,Danser J A H,van den Meiracker A H,Verdonk K,Visser W
Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology
BACKGROUND:A clinical prediction model that could reliably predict the risk of preeclampsia (PE)-related pregnancy complications does not exist. METHODS:We aimed to develop a model to predict the composite outcome of PE-related pregnancy complications, consisting of maternal and fetal adverse within 7, 14 and 30 days in women with suspected or confirmed PE. Data of 384 women from a prospective, multicenter, observational cohort study (n=620) were used. For the development of the prediction model the possible contribution of clinical and standard laboratory variables as well as the biomarkers soluble Fms like tyrosine kinase-1 (sFlt-1), placental growth factor (PlGF) and their ratio was explored using a multivariable competing risk regression analysis. We assessed the discriminative ability of the model with the concordance (c-) statistic. A bootstrap validation procedure with 500 replications was used to correct the estimate of the prediction model performance for optimism and to compute a shrinkage factor for the regression coefficients to correct for overfitting. RESULTS:Among 384 women with suspected/confirmed PE, 96 had PE-related adverse outcomes at any time after hospital admission. Important predictors of PE-related outcomes included sFlt-1/PlGF ratio (continuous), gestational age at time of biomarker measurement (continuous) and protein-to-creatinine ratio (continuous). The c-statistics (corrected for optimism) for developing a PE-related complication within 7, 14 and 30 days were 0.89, 0.88 and 0.87 respectively. There was limited overfitting as indicated by a shrinkage factor of 0.91. CONCLUSIONS:We propose a simple clinical prediction model with good discriminative performance to predict short-term and longer term PE-related complications. Its usefulness in clinical practice awaits further investigation and external validation. This article is protected by copyright. All rights reserved.
Prospective Validation of a Risk Prediction Model to Identify High-Risk Patients for Medication Errors at Hospital Admission.
Ebbens Marieke M,Laar Sylvia A van,Wesselink Elsbeth J,Gombert-Handoko Kim B,van den Bemt Patricia M L A
The Annals of pharmacotherapy
BACKGROUND:Pharmacy-led medication reconciliation in elective surgery patients is often performed at the preoperative screening (POS). Because of the time lag between POS and admission, changes in medication may lead to medication errors at admission (MEAs). In a previous study, a risk prediction model for MEA was developed. OBJECTIVE:To validate this risk prediction model to identify patients at risk for MEAs in a university hospital setting. METHODS:The risk prediction model was derived from a cohort of a Dutch general hospital and validated within a comparable cohort from a Dutch University Medical Centre. MEAs were assessed by comparing the POS medication list with the reconciled medication list at hospital admission. This was considered the gold standard. For every patient, a risk score using the risk prediction model was calculated and compared with the gold standard. The risk prediction model was assessed with receiver operating characteristic (ROC) analysis. RESULTS:Of 368 included patients, 167 (45.4%) had at least 1 MEA. ROC analysis revealed significant differences in the area under the curve of 0.535 ( P = 0.26; validation cohort) versus 0.752 ( P < 0.0001; derivation cohort). The sensitivity in this validating cohort was 66%, with a specificity of 40%. Conclusion and Relevance: The risk prediction model developed in a general hospital population is not suitable to identify patients at risk for MEA in a university hospital population. However, number of medications is a common risk factor in both patient populations and should, thus, form the basis of an adapted risk prediction model.
Development and validation of a prediction model for axial length elongation in myopic children treated with overnight orthokeratology.
Xu Shengsong,Li Zhouyue,Hu Yin,Zhao Wenchen,Jiang Jinyun,Feng Zhibin,Chen Weiyin,Li Cong,Chen Linxing,Fang Binglan,Wang Huarong,Zhai Zhou,Li Bin,Zeng Junwen,Yang Xiao
PURPOSE:To develop and validate a standardized prediction model aiming at 1-year axial length elongation and to guide the orthokeratology lens practice. METHODS:This retrospective study was based on medical records of myopic children treated with orthokeratology. Individuals aged 8-15 years (n = 1261) were included and divided into the primary cohort (n = 757) and validation cohort (n = 504). Feature selection was primarily performed to sort out influential predictors by high-throughput extraction. Then, the prediction model was developed using multivariable linear regression analysis completed by backward stepwise selection. Finally, the validation of the prediction model was performed by evaluation metrics (mean-square error, root-mean-square error, mean absolute error and ). RESULTS:No significant difference was found between primary and validation cohort (all p > 0.05). After the feature selection, the crude model was adjusted by demographic information in multivariable linear regression analysis, and five final predictors were identified (all p < 0.01). The interaction effect of age with 1-month change of zone-3 mm flat K was detected (p < 0.01); hence, two final prediction models were developed based on two age subgroups. The validation proved an acceptable performance. CONCLUSION:An effective multivariable prediction model aiming at 1-year axial length elongation was developed and validated. It can potentially help clinicians to predict orthokeratology efficacy and make valid adjustments. The influential variables revealed in this model can also provide designers directions to optimize the design of lens to improve the efficacy of myopia control.
Early prediction of live birth for assisted reproductive technology patients: a convenient and practical prediction model.
Gao Hong,Liu Dong-E,Li Yumei,Wu Xinrui,Tan Hongzhuan
Live birth is the most important concern for assisted reproductive technology (ART) patients. Therefore, in the medical reproductive centre, obstetricians often need to answer the following question: "What are the chances that I will have a healthy baby after ART treatment?" To date, our obstetricians have no reference on which to base the answer to this question. Our research aimed to solve this problem by establishing prediction models of live birth for ART patients. Between January 1, 2010, and May 1, 2017, we conducted a retrospective cohort study of women undergoing ART treatment at the Reproductive Medicine Centre, Xiangya Hospital of Central South University, Hunan, China. The birth of at least one live-born baby per initiated cycle or embryo transfer procedure was defined as a live birth, and all other pregnancy outcomes were classified as no live birth. A live birth prediction model was established by stepwise multivariate logistic regression. All eligible subjects were randomly allocated to two groups: group 1 (80% of subjects) for the establishment of the prediction models and group 2 (20% of subjects) for the validation of the established prediction models. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of each prediction model at different cut-off values were calculated. The prediction model of live birth included nine variables. The area under the ROC curve was 0.743 in the validation group. The sensitivity, specificity, PPV, and NPV of the established model ranged from 97.9-24.8%, 7.2-96.3%, 44.8-83.8% and 81.7-62.5%, respectively, at different cut-off values. A stable, reliable, convenient, and satisfactory prediction model for live birth by ART patients was established and validated, and this model could be a useful tool for obstetricians to predict the live rate of ART patients. Meanwhile, it is also a reference for obstetricians to create good conditions for infertility patients in preparation for pregnancy.
Development and Validation of a Dynamic Risk Prediction Model to Forecast Psychosis Onset in Patients at Clinical High Risk.
Studerus Erich,Beck Katharina,Fusar-Poli Paolo,Riecher-Rössler Anita
The prediction of outcomes in patients at Clinical High Risk for Psychosis (CHR-P) almost exclusively relies on static data obtained at a single snapshot in time (ie, baseline data). Although the CHR-P symptoms are intrinsically evolving over time, available prediction models cannot be dynamically updated to reflect these changes. Hence, the aim of this study was to develop and internally validate a dynamic risk prediction model (joint model) and to implement this model in a user-friendly online risk calculator. Furthermore, we aimed to explore the prognostic performance of extended dynamic risk prediction models and to compare static with dynamic prediction. One hundred ninety-six CHR-P patients were recruited as part of the "Basel Früherkennung von Psychosen" (FePsy) study. Psychopathology and transition to psychosis was assessed at regular intervals for up to 5 years using the Brief Psychiatric Rating Scale-Expanded (BPRS-E). Various specifications of joint models were compared with regard to their cross-validated prognostic performance. We developed and internally validated a joint model that predicts psychosis onset from BPRS-E disorganization and years of education at baseline and BPRS-E positive symptoms during the follow-up with good prognostic performance. The model was implemented as online risk calculator (http://www.fepsy.ch/DPRP/). The use of extended joint models slightly increased the prognostic accuracy compared to basic joint models, and dynamic models showed a higher prognostic accuracy than static models. Our results confirm that extended joint modeling could improve the prediction of psychosis in CHR-P patients. We implemented the first online risk calculator that can dynamically update psychosis risk prediction.
Prediction model of artificial neural network for the risk of hyperuricemia incorporating dietary risk factors in a Chinese adult study.
Zeng Jie,Zhang Junguo,Li Ziyi,Li Tianwang,Li Guowei
Food & nutrition research
Background:Risk of hyperuricemia (HU) has been shown to be strongly associated with dietary factors. However, there is scarce evidence on prediction models incorporating dietary factors to estimate the risk of HU. Objective:The aim of this study was to develop a prediction model to predict the risk of HU in Chinese adults based on dietary information. Design:Our study was based on a cross-sectional survey, which recruited 1,488 community residents aged 18 to 60 years in Beijing from October 2010 to January 2011. The eligible participants were randomly divided into a training set ( = 992) and a validation set ( = 496) in the ratio of 2:1. We developed the prediction model in three stages. We first used a logistic regression model (LRM) based on the training set to select a set of dietary risk factors which were related to the risk of HU. Artificial neural network (ANN) was then used to construct the prediction model using the training set. Finally, we used receiver operating characteristic (ROC) curve analysis to assess the accuracy of the prediction model using training and validation sets. Results:In the training set, the mean age of participants with and without HU was 39.3 (standard deviation [SD]: 9.65) and 38.2 (SD: 9.38) years, respectively. Patients with HU consisted of 101 males (77.7%) and 29 females (22.3%). The LRM found that food frequency (vegetables [odds ratio (OR) = 0.73], meat [0.72], eggs [0.80], plant oil [0.78], tea [0.51], eating habits (breakfast [OR = 1.28]), and the salty cooking style (OR = 1.33) were associated with risk of HU. In the ANN analysis, we selected a three-layer back propagation neural network (BPNN) model with 14, 3, and 1 neuron in the input, hidden, and output layers, respectively, as the best prediction model. The areas under the ROC of the training and validation sets were 0.827 and 0.814, respectively. HU would occur when the incidence probability is greater than 0.128. The indicators of accuracy, sensitivity, specificity, and Yuden Index suggested that the ANN model in our study is successful and valuable. Conclusions:This study suggests that the ANN model could be used to predict the risk of HU in Chinese adults. Further prospective studies are needed to improve the accuracy and to generalize the use of model.
Derivation and validation of a mortality risk prediction model using global longitudinal strain in patients with acute heart failure.
Hwang In-Chang,Cho Goo-Yeong,Choi Hong-Mi,Yoon Yeonyee E,Park Jin Joo,Park Jun-Bean,Park Jae-Hyeong,Lee Seung-Pyo,Kim Hyung-Kwan,Kim Yong-Jin,Sohn Dae-Won
European heart journal cardiovascular Imaging
AIMS:To develop a mortality risk prediction model in patients with acute heart failure (AHF), using left ventricular (LV) function parameters with clinical factors. METHODS AND RESULTS:In total, 4312 patients admitted for AHF were retrospectively identified from three tertiary centres, and echocardiographic parameters including LV ejection fraction (LV-EF) and LV global longitudinal strain (LV-GLS) were measured in a core laboratory. The full set of risk factors was available in 3248 patients. Using Cox proportional hazards model, we developed a mortality risk prediction model in 1859 patients from two centres (derivation cohort) and validated the model in 1389 patients from one centre (validation cohort). During 32 (interquartile range 13-54) months of follow-up, 1285 patients (39.6%) died. Significant predictors for mortality were age, diabetes, diastolic blood pressure, body mass index, natriuretic peptide, glomerular filtration rate, failure to prescribe beta-blockers, failure to prescribe renin-angiotensin system blockers, and LV-GLS; however, LV-EF was not a significant predictor. Final model including these predictors to estimate individual probabilities of mortality had C-statistics of 0.75 [95% confidence interval (CI) 0.73-0.78; P < 0.001] in the derivation cohort and 0.78 (95% CI 0.75-0.80; P < 0.001) in the validation cohort. The prediction model had good performance in both heart failure (HF) with reduced EF, HF with mid-range EF, and HF with preserved EF. CONCLUSION:We developed a mortality risk prediction model for patients with AHF incorporating LV-GLS as the LV function parameter, and other clinical factors. Our model provides an accurate prediction of mortality and may provide reliable risk stratification in AHF patients.
Development and Internal Validation of a Multivariable Prediction Model for Adrenocortical-Carcinoma-Specific Mortality.
Ettaieb Madeleine H T,van Kuijk Sander M J,de Wit-Pastoors Annelies,Feelders Richard A,Corssmit Eleonora P M,Eekhoff Elisabeth M W,van der Valk Paul,Timmers Henri J L M,Kerstens Michiel N,Klümpen Heinz-Josef,Leeuwaarde van Rachel S,Havekes Bas,Haak Harm R
Adrenocortical carcinoma (ACC) has an incidence of about 1.0 per million per year. In general, survival of patients with ACC is limited. Predicting survival outcome at time of diagnosis is a clinical challenge. The aim of this study was to develop and internally validate a clinical prediction model for ACC-specific mortality. Data for this retrospective cohort study were obtained from the nine centers of the Dutch Adrenal Network (DAN). Patients who presented with ACC between 1 January 2004 and 31 October 2013 were included. We used multivariable Cox proportional hazards regression to compute the coefficients for the prediction model. Backward stepwise elimination was performed to derive a more parsimonious model. The performance of the initial prediction model was quantified by measures of model fit, discriminative ability, and calibration. We undertook an internal validation step to counteract the possible overfitting of our model. A total of 160 patients were included in the cohort. The median survival time was 35 months, and interquartile range (IQR) 50.7 months. The multivariable modeling yielded a prediction model that included age, modified European Network for the Study of Adrenal Tumors (mENSAT) stage, and radical resection. The c-statistic was 0.77 (95% Confidence Interval: 0.72, 0.81), indicating good predictive performance. We developed a clinical prediction model for ACC-specific mortality. ACC mortality can be estimated using a relatively simple clinical prediction model with good discriminative ability and calibration.
Protocol for a systematic review on the methodological and reporting quality of prediction model studies using machine learning techniques.
Andaur Navarro Constanza L,Damen Johanna A A G,Takada Toshihiko,Nijman Steven W J,Dhiman Paula,Ma Jie,Collins Gary S,Bajpai Ram,Riley Richard D,Moons Karel Gm,Hooft Lotty
INTRODUCTION:Studies 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 ANALYSIS:A 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 DISSEMINATION:Ethical 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 REGISTRATION:PROSPERO, CRD42019161764.
[Value of joint prediction model based on the modified systemic inflammatory response syndrome score for predicting mortality risk of patients with large area burns at early stage after admission].
Fan J H,Sun Y F,Wu G S,Wang K A,Wei J,Sun Y
Zhonghua shao shang za zhi = Zhonghua shaoshang zazhi = Chinese journal of burns
To investigate the predictive value of the joint prediction model based on the modified systemic inflammatory response syndrome (SIRS) score (hereinafter referred to as the joint prediction model) for the mortality risk of patients with large area burns within 24 hours after admission. The clinical data of 158 patients [111 males, 47 females, aged 40 (28, 50) years] admitted to the Department of Burn Surgery of the First Affiliated Hospital of Naval Medical University from January 2005 to January 2018, conforming to the study criteria, were analyzed retrospectively by the method of case-control study. The age, gender, total burn area, full-thickness burn area, injury cause, with or without inhalation injury, severity of inhalation injury, and tracheotomy condition of patients were recorded, and the modified SIRS score and the modified Baux score of patients were calculated. According to the final outcome, all patients were divided into survival group (=123) and death group (=35). The clinical data of patients between two groups, except for modified Baux score, were compared by chi-square test or Mann-Whitney test to screen the death-related factors of patients. The indexes with statistically significant difference between the two groups were included in the multivariate logistic regression analysis to screen the independent risk factors related to the death of patients, and the prediction model was constructed by combining the modified SIRS score. The receiver's operating characteristic curves of the modified SIRS score, the modified Baux score, and the joint prediction model of 158 patients were drawn to analyze their ability to predict death of patients. The area under curve (AUC) of the receiver's operating characteristic and the sensitivity and specificity of optimal threshold were calculated, and the quality of AUC of the three prediction indexes was compared with Jonckheere-Terpstra test. (1) There were statistically significant differences between the two groups in the modified SIRS score, age, total burn area, full-thickness burn area, severity of inhalation injury, with or without inhalation injury, and tracheotomy condition of patients (=-4.356, -3.568, -5.291, -6.052, -4.720, (2)=12.967, 19.692, <0.01). (2) The modified SIRS score, age, full-thickness burn area were the independent risk factors for the death of patients with large area burn (odds ratio=2.699, 1.069, 1.029, 95% confidence interval=1.447-5.033, 1.029-1.109, 1.005-1.054, <0.05). (3) The AUC of modified SIRS score, the joint prediction model, and the modified Baux score for predicting death of 158 patients within 24 hours after admission were 0.730, 0.879, and 0.895 respectively (95% confidence interval=0.653-0.797, 0.818-0.926, 0.836-0.938, <0.01). The sensitivities of the three optimal threshold values to death prediction were 54.3%, 91.4%, and 82.9% respectively, while the specificities were 81.3%, 76.4%, and 84.6% respectively. The AUC quality of the joint prediction model was similar to that of the modified Baux score (95% confidence interval=-0.057-0.088, >0.05), and both of them were significantly better than that of the modified SIRS score (95% confidence interval=0.072-0.259, 0.023-0.276, <0.05 or <0.01). Both the joint prediction model and the modified Baux score are considered to be good to predict the death rate of patients with large area burns at early stage after admission. However, the joint prediction model has better clinical practice value due to its advantage of simple scoring and easier access to data acquisition.
An integrated time adaptive geographic atrophy prediction model for SD-OCT images.
Zhang Yuhan,Zhang Xiwei,Ji Zexuan,Niu Sijie,Leng Theodore,Rubin Daniel L,Yuan Songtao,Chen Qiang
Medical image analysis
The automated prediction of geographic atrophy (GA) lesion growth can help ophthalmologists understand how the GA progresses, and assess the efficiency of current treatment and the prognosis of the disease. We developed an integrated time adaptive prediction model for identifying the location of future GA growth. The proposed model was comprised of bi-directional long short-term memory (BiLSTM) network-based prediction module and convolutional neural network (CNN)-based refinement module. Considering the discontinuity of time intervals among sequential follow-up visits, we integrated time factors into BiLSTM-based prediction module to control the time attribute expediently. Then, the results from prediction module were refined by a CNN-based strategy to obtain the final locations of future GA growth. The 10 scenarios were designed to evaluate the prediction accuracy of our proposed model. The 1-6th scenarios demonstrated the importance of the prior information similarity, the 7-8th scenarios verified the effect of time factors and refinement methods respectively and the 9th scenario compared the prediction results between those using a single follow-up visit for training and using 2 sequential follow-up visits for training. The 10th scenario showed the model generalization performance across regions. The average dice indexes (DI) of the predicted GA regions in the 1-6th scenarios are 0.86, 0.89, 0.89, 0.92 and 0.88, 0.90, respectively. By integrating time factors to the BiLSTM models, the prediction accuracy was improved by almost 10%. The CNN-based refinement strategy can remove the wrong GA regions effectively to preserve the actual GA regions and improve the prediction accuracy further. The prediction results based on 2 sequential follow-up visits showed higher correlations than that based on single follow-up visit. The proposed model presented a good generalization performance while training patients and testing patients were from different regions. Experimental results demonstrated the importance of prior information to the prediction accuracy. We demonstrate the feasibility of creating a model for disease prediction.
Maximum Threshold Genomic Prediction Model for Ordinal Traits.
Montesinos-López Abelardo,Gutierrez-Pulido Humberto,Montesinos-López Osval Antonio,Crossa José
G3 (Bethesda, Md.)
Due to the ever-increasing data collected in genomic breeding programs, there is a need for genomic prediction models that can deal better with big data. For this reason, here we propose a Maximum Threshold Genomic Prediction (MAPT) model for ordinal traits that is more efficient than the conventional Bayesian Threshold Genomic Prediction model for ordinal traits. The MAPT performs the predictions of the Threshold Genomic Prediction model by using the maximum estimation of the parameters, that is, the values of the parameters that maximize the joint posterior density. We compared the prediction performance of the proposed MAPT to the conventional Bayesian Threshold Genomic Prediction model, the multinomial Ridge regression and support vector machine on 8 real data sets. We found that the proposed MAPT was competitive with regard to the multinomial and support vector machine models in terms of prediction performance, and slightly better than the conventional Bayesian Threshold Genomic Prediction model. With regard to the implementation time, we found that in general the MAPT and the support vector machine were the best, while the slowest was the multinomial Ridge regression model. However, it is important to point out that the successful implementation of the proposed MAPT model depends on the informative priors used to avoid underestimation of variance components.
Prognosis prediction model based on competing endogenous RNAs for recurrence of colon adenocarcinoma.
Jin Li Peng,Liu Tao,Meng Fan Qi,Tai Jian Dong
BACKGROUND:Colon adenocarcinoma (COAD) patients who develop recurrence have poor prognosis. Our study aimed to establish effective prognosis prediction model based on competing endogenous RNAs (ceRNAs) for recurrence of COAD. METHODS:COAD expression profilings downloaded from The Cancer Genome Atlas (TCGA) were used as training dataset, and expression profilings of GSE29623 retrieved from Gene Expression Omnibus (GEO) were set as validation dataset. Differentially expressed RNAs (DERs) between non-recurrent and recurrent specimens in training dataset were screened, and optimum prognostic signature DERs were revealed to establish prognostic score (PS) model. Kaplan-Meier survival analysis was conducted for PS model, and GEO dataset was used for validation. Prognosis prediction efficiencies were evaluated by area under curve (AUC) and C-index. Meanwhile, ceRNA regulatory network was constructed by using signature mRNAs, lncRNAs and miRNAs. RESULTS:We identified 562 DERs including 42 lncRNAs, 36 miRNAs, and 484 mRNAs. PS prediction model, consisting of 17 optimum prognostic signature DERs, showed that high risk group had significantly poorer prognosis (5-year AUC = 0.951, C-index = 0.788), which also validated in GSE29623. Prognosis prediction model incorporating multi-RNAs with pathologic distant metastasis (M) and pathologic primary tumor (T) (5-year AUC = 0.969, C-index = 0.812) had better efficiency than clinical prognosis prediction model (5-year AUC = 0.712, C-index = 0.680). In the constructed ceRNA regulatory network, lncRNA NCBP2-AS1 could interact with hsa-miR-34c and hsa-miR-363, and lncRNA LINC00115 could interact with hsa-miR-363 and hsa-miR-4709. SIX4, GRAP, NKAIN4, MMAA, and ERVMER34-1 are regulated by hsa-miR-4709. CONCLUSION:Prognosis prediction model incorporating multi-RNAs with pathologic M and pathologic T may have great value in COAD prognosis prediction.