1
|
Keogh RH, Van Geloven N. Prediction Under Interventions: Evaluation of Counterfactual Performance Using Longitudinal Observational Data. Epidemiology 2024; 35:329-339. [PMID: 38630508 DOI: 10.1097/ede.0000000000001713] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 04/19/2024]
Abstract
Predictions under interventions are estimates of what a person's risk of an outcome would be if they were to follow a particular treatment strategy, given their individual characteristics. Such predictions can give important input to medical decision-making. However, evaluating the predictive performance of interventional predictions is challenging. Standard ways of evaluating predictive performance do not apply when using observational data, because prediction under interventions involves obtaining predictions of the outcome under conditions that are different from those that are observed for a subset of individuals in the validation dataset. This work describes methods for evaluating counterfactual performance of predictions under interventions for time-to-event outcomes. This means we aim to assess how well predictions would match the validation data if all individuals had followed the treatment strategy under which predictions are made. We focus on counterfactual performance evaluation using longitudinal observational data, and under treatment strategies that involve sustaining a particular treatment regime over time. We introduce an estimation approach using artificial censoring and inverse probability weighting that involves creating a validation dataset mimicking the treatment strategy under which predictions are made. We extend measures of calibration, discrimination (c-index and cumulative/dynamic AUCt) and overall prediction error (Brier score) to allow assessment of counterfactual performance. The methods are evaluated using a simulation study, including scenarios in which the methods should detect poor performance. Applying our methods in the context of liver transplantation shows that our procedure allows quantification of the performance of predictions supporting crucial decisions on organ allocation.
Collapse
Affiliation(s)
- Ruth H Keogh
- From the Department of Medical Statistics, London School of Hygiene & Tropical Medicine, London, United Kingdom
| | - Nan Van Geloven
- Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands
| |
Collapse
|
2
|
Beyene KM, Chen DG, Kifle YG. A novel nonparametric time-dependent precision-recall curve estimator for right-censored survival data. Biom J 2024; 66:e2300135. [PMID: 38637327 DOI: 10.1002/bimj.202300135] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/20/2023] [Revised: 10/04/2023] [Accepted: 12/27/2023] [Indexed: 04/20/2024]
Abstract
In order to assess prognostic risk for individuals in precision health research, risk prediction models are increasingly used, in which statistical models are used to estimate the risk of future outcomes based on clinical and nonclinical characteristics. The predictive accuracy of a risk score must be assessed before it can be used in routine clinical decision making, where the receiver operator characteristic curves, precision-recall curves, and their corresponding area under the curves are commonly used metrics to evaluate the discriminatory ability of a continuous risk score. Among these the precision-recall curves have been shown to be more informative when dealing with unbalanced biomarker distribution between classes, which is common in rare event, even though except one, all existing methods are proposed for classic uncensored data. This paper is therefore to propose a novel nonparametric estimation approach for the time-dependent precision-recall curve and its associated area under the curve for right-censored data. A simulation is conducted to show the better finite sample property of the proposed estimator over the existing method and a real-world data from primary biliary cirrhosis trial is used to demonstrate the practical applicability of the proposed estimator.
Collapse
Affiliation(s)
- Kassu Mehari Beyene
- College of Health Solutions, Arizona State University, Phoenix, Arizona, USA
| | - Ding-Geng Chen
- College of Health Solutions, Arizona State University, Phoenix, Arizona, USA
- Department of Statistics, University of Pretoria, Pretoria, South Africa
| | - Yehenew Getachew Kifle
- Department of Mathematics and Statistics, University of Maryland Baltimore County, Baltimore, Maryland, USA
| |
Collapse
|
3
|
Beyene KM, Chen DG. Time-dependent receiver operating characteristic curve estimator for correlated right-censored time-to-event data. Stat Methods Med Res 2024; 33:162-181. [PMID: 38130110 DOI: 10.1177/09622802231220496] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/23/2023]
Abstract
In clinical trials, evaluating the accuracy of risk scores (markers) derived from prognostic models for prediction of survival outcomes is of major concern. The time-dependent receiver operating characteristic curve and the corresponding area under the receiver operating characteristic curve are appealing measures to evaluate the predictive accuracy. Several estimation methods have been proposed in the context of classical right-censored data which assumes the event time of individuals are independent. In many applications, however, this may not hold true if, for example, individuals belong to clusters or experience recurrent events. Estimates may be biased if this correlated nature is not taken into account. This paper is then aimed to fill this knowledge gap to introduce a time-dependent receiver operating characteristic curve and the corresponding area under the receiver operating characteristic curve estimation method for right-censored data that take the correlated nature into account. In the proposed method, the unknown status of censored subjects is imputed using conditional survival functions given the marker and frailty of the subjects. An extensive simulation study is conducted to evaluate and demonstrate the finite sample performance of the proposed method. Finally, the proposed method is illustrated using two real-world examples of lung cancer and kidney disease.
Collapse
Affiliation(s)
| | - Ding-Geng Chen
- Arizona State University, College of Health Solutions, AZ, USA
- Department of Statistics, University of Pretoria, Pretoria, South Africa
| |
Collapse
|
4
|
Service SK, De La Hoz J, Diaz-Zuluaga AM, Arias A, Pimplaskar A, Luu C, Mena L, Valencia J, Ramírez MC, Bearden CE, Sabbati C, Reus VI, López-Jaramillo C, Freimer NB, Loohuis LMO. Predicting diagnostic conversion from major depressive disorder to bipolar disorder: an EHR based study from Colombia. MEDRXIV : THE PREPRINT SERVER FOR HEALTH SCIENCES 2023:2023.09.28.23296092. [PMID: 37873340 PMCID: PMC10593019 DOI: 10.1101/2023.09.28.23296092] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 10/25/2023]
Abstract
Bipolar Disorder (BD) is a severe and chronic disorder characterized by recurrent episodes of depression, mania, and/or hypomania. Most BD patients initially present with depressive symptoms, resulting in a delayed diagnosis of BD and poor clinical outcomes. This study leverages electronic health record (EHR) data from the Clínica San Juan de Dios Manizales in Colombia to identify features predictive of the transition from Major Depressive Disorder (MDD) to BD. Analyzing EHR data from 13,607 patients diagnosed with MDD over 15 years, we identified 1,610 cases of conversion to BD. Using a multivariate Cox regression model, we identified severity of the initial MDD episode, the presence of psychosis and hospitalization at first episode, family history of mood or psychotic disorders, female gender to be predictive of the conversion to BD. Additionally, we observed associations with medication classes (prescriptions of mood stabilizers, antipsychotics, and antidepressants) and clinical features (delusions, suicide attempt, suicidal ideation, use of marijuana and alcohol use/abuse) derived from natural language processing (NLP) of clinical notes. Together, these risk factors predicted BD conversion within five years of the initial MDD diagnosis, with a recall of 72% and a precision of 38%. Our study confirms many previously identified risk factors identified through registry-based studies (such as female gender and psychotic depression at the index MDD episode), and identifies novel ones (specifically, suicidal ideation and suicide attempt extracted from clinical notes). These results simultaneously demonstrate the validity of using EHR data for predicting BD conversion as well as underscore its potential for the identification of novel risk factors and improving early diagnosis.
Collapse
Affiliation(s)
- Susan K Service
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Juan De La Hoz
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Ana M Diaz-Zuluaga
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Alejandro Arias
- Research Group in Psychiatry (GIPSI), Institute of Medical Research, Department of Psychiatry, Faculty of Medicine, University of Antioquia, Medellín, Colombia
| | - Aditya Pimplaskar
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Chuc Luu
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Laura Mena
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Johanna Valencia
- Research Group in Psychiatry (GIPSI), Institute of Medical Research, Department of Psychiatry, Faculty of Medicine, University of Antioquia, Medellín, Colombia
| | | | - Carrie E Bearden
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Chiara Sabbati
- Department of Biomedical Data Science, Stanford University, Stanford, USA
| | - Victor I Reus
- Department of Psychiatry, University of California San Francisco, San Francisco, USA
| | - Carlos López-Jaramillo
- Research Group in Psychiatry (GIPSI), Institute of Medical Research, Department of Psychiatry, Faculty of Medicine, University of Antioquia, Medellín, Colombia
| | - Nelson B Freimer
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| | - Loes M Olde Loohuis
- Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, USA
| |
Collapse
|
5
|
Touraine C, Winter A, Castan F, Azria D, Gourgou S. Time-Dependent ROC Curve Analysis for Assessing the Capability of Radiation-Induced CD8 T-Lymphocyte Apoptosis to Predict Late Toxicities after Adjuvant Radiotherapy of Breast Cancer Patients. Cancers (Basel) 2023; 15:4676. [PMID: 37835370 PMCID: PMC10571898 DOI: 10.3390/cancers15194676] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/10/2023] [Revised: 09/15/2023] [Accepted: 09/19/2023] [Indexed: 10/15/2023] Open
Abstract
Late fibrosis can occur in breast cancer patients treated with curative-intent radiotherapy. Predicting this toxicity is of clinical interest in order to adapt the irradiation dose delivered. Radiation-induced CD8 T-lymphocyte apoptosis (RILA) had been proven to be associated with less grade ≥2 late radiation-induced toxicities in patients with miscellaneous cancers. Tobacco smoking status and adjuvant hormonotherapy were also identified as potential factors related to late-breast-fibrosis-free survival. This article evaluates the predictive performance of the RILA using a ROC curve analysis that takes into account the dynamic nature of fibrosis occurrence. This time-dependent ROC curve approach is also applied to evaluate the ability of the RILA combined with the other previously identified factors. Our analysis includes a Monte Carlo cross-validation procedure and the calculation of an expected cost of misclassification, which provides more importance to patients who have no risk of late fibrosis in order to be able to treat them with the maximal irradiation dose. Performance evaluation was assessed at 12, 24, 36 and 50 months. At 36 months, our results were comparable to those obtained in a previous study, thus underlying the predictive power of the RILA. Based on specificity and cost, RILA alone seemed to be the most performant, while its association with the other factors had better negative predictive value results.
Collapse
Affiliation(s)
- Célia Touraine
- Biometrics Unit, Cancer Institute of Montpellier (ICM), University Montpellier, 34090 Montpellier, France; (C.T.); (F.C.); (S.G.)
- French National Platform Quality of Life and Cancer, 34090 Montpellier, France
- Desbrest Institute of Epidemiology and Public Health (IDESP), University Montpellier, INSERM, 34090 Montpellier, France
| | - Audrey Winter
- Biometrics Unit, Cancer Institute of Montpellier (ICM), University Montpellier, 34090 Montpellier, France; (C.T.); (F.C.); (S.G.)
- French National Platform Quality of Life and Cancer, 34090 Montpellier, France
| | - Florence Castan
- Biometrics Unit, Cancer Institute of Montpellier (ICM), University Montpellier, 34090 Montpellier, France; (C.T.); (F.C.); (S.G.)
| | - David Azria
- Radiotherapy Unit, Cancer Institute of Montpellier (ICM), University Montpellier, 34090 Montpellier, France;
| | - Sophie Gourgou
- Biometrics Unit, Cancer Institute of Montpellier (ICM), University Montpellier, 34090 Montpellier, France; (C.T.); (F.C.); (S.G.)
- French National Platform Quality of Life and Cancer, 34090 Montpellier, France
| |
Collapse
|
6
|
Cheng W, Li X. A semi-parametric approach for time-dependent ROC curves with nonignorable missing biomarker. J Biopharm Stat 2023; 33:555-574. [PMID: 36852969 DOI: 10.1080/10543406.2023.2170394] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/15/2021] [Accepted: 12/30/2022] [Indexed: 03/01/2023]
Abstract
The main purpose of this paper is to survey the statistical inference for covariate-specific time-dependent receiver operating characteristic (ROC) curves with nonignorable missing continuous biomarker values. To construct time-dependent ROC curves, we consider a joint model which assumes that the failure time depends on the continuous biomarker and the covariates through a Cox proportional hazards model and that the continuous biomarker depends on the covariates through a semiparametric location model. Assuming a purely parametric model on the propensity score, we utilize instrumental variables to deal with the identifiable issue and estimate the unknown parameters of the propensity score by a simple and efficient method. In addition, when the propensity score is estimated, we develop HT and AIPW approaches to estimate our interested quantities. In the presence of nonignorable missing biomarker, our AIPW estimators of the interested quantities are still doubly robust when the true propensity score is a special parametric logistic model. At last, simulation studies are conducted to assess the performance of our proposed approaches, and a real data analysis is also carried out to illustrate its application.
Collapse
Affiliation(s)
- Weili Cheng
- School of Mathematics and Statistics, North China University of Water Resources and Electric Power, Zhengzhou, China
| | - Xiaorui Li
- School of Mathematics and Statistics, North China University of Water Resources and Electric Power, Zhengzhou, China
| |
Collapse
|
7
|
Díaz LA, Fuentes-López E, Ayares G, Idalsoaga F, Arnold J, Valverde MA, Perez D, Gómez J, Escarate R, Villalón A, Ramírez CA, Hernandez-Tejero M, Zhang W, Qian S, Simonetto DA, Ahn JC, Buryska S, Dunn W, Mehta H, Agrawal R, Cabezas J, García-Carrera I, Cuyàs B, Poca M, Soriano G, Sarin SK, Maiwall R, Jalal PK, Abdulsada S, Higuera-de-la-Tijera F, Kulkarni AV, Rao PN, Salazar PG, Skladaný L, Bystrianska N, Clemente-Sanchez A, Villaseca-Gómez C, Haider T, Chacko KR, Romero GA, Pollarsky FD, Restrepo JC, Castro-Sanchez S, Toro LG, Yaquich P, Mendizabal M, Garrido ML, Marciano S, Dirchwolf M, Vargas V, Jiménez C, Louvet A, García-Tsao G, Roblero JP, Abraldes JG, Shah VH, Kamath PS, Arrese M, Singal AK, Bataller R, Arab JP. MELD 3.0 adequately predicts mortality and renal replacement therapy requirements in patients with alcohol-associated hepatitis. JHEP Rep 2023; 5:100727. [PMID: 37456675 PMCID: PMC10339256 DOI: 10.1016/j.jhepr.2023.100727] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 08/06/2022] [Revised: 02/22/2023] [Accepted: 02/27/2023] [Indexed: 07/18/2023] Open
Abstract
Background & Aims Model for End-Stage Liver Disease (MELD) score better predicts mortality in alcohol-associated hepatitis (AH) but could underestimate severity in women and malnourished patients. Using a global cohort, we assessed the ability of the MELD 3.0 score to predict short-term mortality in AH. Methods This was a retrospective cohort study of patients admitted to hospital with AH from 2009 to 2019. The main outcome was all-cause 30-day mortality. We compared the AUC using DeLong's method and also performed a time-dependent AUC with competing risks analysis. Results A total of 2,124 patients were included from 28 centres from 10 countries on three continents (median age 47.2 ± 11.2 years, 29.9% women, 71.3% with underlying cirrhosis). The median MELD 3.0 score at admission was 25 (20-33), with an estimated survival of 73.7% at 30 days. The MELD 3.0 score had a better performance in predicting 30-day mortality (AUC:0.761, 95%CI:0.732-0.791) compared with MELD sodium (MELD-Na; AUC: 0.744, 95% CI: 0.713-0.775; p = 0.042) and Maddrey's discriminant function (mDF) (AUC: 0.724, 95% CI: 0.691-0.757; p = 0.013). However, MELD 3.0 did not perform better than traditional MELD (AUC: 0.753, 95% CI: 0.723-0.783; p = 0.300) and Age-Bilirubin-International Normalised Ratio-Creatinine (ABIC) (AUC:0.757, 95% CI: 0.727-0.788; p = 0.765). These results were consistent in competing-risk analysis, where MELD 3.0 (AUC: 0.757, 95% CI: 0.724-0.790) predicted better 30-day mortality compared with MELD-Na (AUC: 0.739, 95% CI: 0.708-0.770; p = 0.028) and mDF (AUC:0.717, 95% CI: 0.687-0.748; p = 0.042). The MELD 3.0 score was significantly better in predicting renal replacement therapy requirements during admission compared with the other scores (AUC: 0.844, 95% CI: 0.805-0.883). Conclusions MELD 3.0 demonstrated better performance compared with MELD-Na and mDF in predicting 30-day and 90-day mortality, and was the best predictor of renal replacement therapy requirements during admission for AH. However, further prospective studies are needed to validate its extensive use in AH. Impact and implications Severe AH has high short-term mortality. The establishment of treatments and liver transplantation depends on mortality prediction. We evaluated the performance of the new MELD 3.0 score to predict short-term mortality in AH in a large global cohort. MELD 3.0 performed better in predicting 30- and 90-day mortality compared with MELD-Na and mDF, but was similar to MELD and ABIC scores. MELD 3.0 was the best predictor of renal replacement therapy requirements. Thus, further prospective studies are needed to support the wide use of MELD 3.0 in AH.
Collapse
Affiliation(s)
- Luis Antonio Díaz
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
| | - Eduardo Fuentes-López
- Departamento de Ciencias de la Salud, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
| | - Gustavo Ayares
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
| | - Francisco Idalsoaga
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
| | - Jorge Arnold
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
| | | | - Diego Perez
- Servicio Medicina Interna, Hospital El Pino, Santiago, Chile
| | - Jaime Gómez
- Servicio Medicina Interna, Hospital El Pino, Santiago, Chile
| | | | - Alejandro Villalón
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
- Departamento de Ciencias Médicas, Facultad de Medicina y Odontología, Universidad de Antofagasta, Antofagasta, Chile
| | - Carolina A. Ramírez
- Department of Anesthesia & Perioperative Medicine, Western University, London, ON, Canada
| | - Maria Hernandez-Tejero
- Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA
- Liver Unit, Hospital Clinic, Barcelona, Spain
| | - Wei Zhang
- Division of Gastroenterology and Hepatology, University of Florida, Gainesville, FL, USA
- Gastroenterology Unit, Massachusetts General Hospital, Boston, MA, USA
| | - Steve Qian
- Division of Gastroenterology and Hepatology, University of Florida, Gainesville, FL, USA
| | | | - Joseph C. Ahn
- Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA
| | - Seth Buryska
- Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA
| | | | - Heer Mehta
- University of Kansas Medical Center, KS, USA
| | - Rohit Agrawal
- Division of Gastroenterology and Hepatology, University of Illinois, Chicago, IL, USA
| | - Joaquín Cabezas
- Gastroenterology and Hepatology Department. University Hospital Marqués de Valdecilla, Santander, Spain
- Research Institute Valdecilla (IDIVAL), Santander, Spain
| | - Inés García-Carrera
- Gastroenterology and Hepatology Department. University Hospital Marqués de Valdecilla, Santander, Spain
- Research Institute Valdecilla (IDIVAL), Santander, Spain
| | - Berta Cuyàs
- Department of Gastroenterology, Hospital de la Santa Creu i Sant Pau, Institut de Recerca Hospital de Sant Pau-IIB Sant Pau, Universitat Autònoma de Barcelona, CIBERehd, Barcelona, Spain
| | - Maria Poca
- Department of Gastroenterology, Hospital de la Santa Creu i Sant Pau, Institut de Recerca Hospital de Sant Pau-IIB Sant Pau, Universitat Autònoma de Barcelona, CIBERehd, Barcelona, Spain
| | - German Soriano
- Department of Gastroenterology, Hospital de la Santa Creu i Sant Pau, Institut de Recerca Hospital de Sant Pau-IIB Sant Pau, Universitat Autònoma de Barcelona, CIBERehd, Barcelona, Spain
| | - Shiv K. Sarin
- Department of Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India
| | - Rakhi Maiwall
- Department of Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India
| | - Prasun K. Jalal
- Department of Gastroenterology and Hepatology, Baylor College of Medicine, Houston, TX, USA
| | - Saba Abdulsada
- Department of Gastroenterology and Hepatology, Baylor College of Medicine, Houston, TX, USA
| | - Fátima Higuera-de-la-Tijera
- Servicio de Gastroenterología, Hospital General de México ‘Dr. Eduardo Liceaga’, Facultad de Medicina, Universidad Nacional Autónoma de México, México City, Mexico
| | - Anand V. Kulkarni
- Department of Hepatology, Asian Institute of Gastroenterology, Hyderabad, India
| | - P. Nagaraja Rao
- Department of Hepatology, Asian Institute of Gastroenterology, Hyderabad, India
| | | | - Lubomir Skladaný
- Division of Hepatology, Gastroenterology and Liver Transplantation, Department of Internal Medicine II, Slovak Medical University, F.D. Roosevelt University Hospital, Banska Bystrica, Slovak Republic
| | - Natália Bystrianska
- Division of Hepatology, Gastroenterology and Liver Transplantation, Department of Internal Medicine II, Slovak Medical University, F.D. Roosevelt University Hospital, Banska Bystrica, Slovak Republic
| | - Ana Clemente-Sanchez
- Liver Unit, Department of Digestive Diseases Hospital General Universitario Gregorio Marañón Madrid, Madrid, Spain
- CIBERehd Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas Madrid, Madrid, Spain
| | - Clara Villaseca-Gómez
- Liver Unit, Department of Digestive Diseases Hospital General Universitario Gregorio Marañón Madrid, Madrid, Spain
- CIBERehd Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas Madrid, Madrid, Spain
| | - Tehseen Haider
- Division of Gastroenterology and Hepatology, Montefiore Medical Center, Bronx, NY, USA
| | - Kristina R. Chacko
- Division of Gastroenterology and Hepatology, Montefiore Medical Center, Bronx, NY, USA
| | - Gustavo A. Romero
- Sección Hepatología, Hospital de Gastroenterología Dr. Carlos Bonorino Udaondo, Buenos Aires, Argentina
| | - Florencia D. Pollarsky
- Sección Hepatología, Hospital de Gastroenterología Dr. Carlos Bonorino Udaondo, Buenos Aires, Argentina
| | - Juan Carlos Restrepo
- Unidad de Hepatología del Hospital Pablo Tobon Uribe, Grupo de Gastrohepatología de la Universidad de Antioquia, Medellín, Colombia
| | - Susana Castro-Sanchez
- Unidad de Hepatología del Hospital Pablo Tobon Uribe, Grupo de Gastrohepatología de la Universidad de Antioquia, Medellín, Colombia
| | - Luis G. Toro
- Hepatology and Liver Transplant Unit, Hospitales de San Vicente Fundación de Medellín y Rionegro, Medellín, Colombia
| | - Pamela Yaquich
- Departamento de Gastroenterología, Hospital San Juan de Dios, Santiago, Chile
| | - Manuel Mendizabal
- Hepatology and Liver Transplant Unit, Hospital Universitario Austral, Buenos Aires, Argentina
| | | | | | - Melisa Dirchwolf
- Unidad de Hígado, Hospital Privado de Rosario, Rosario, Argentina
| | - Victor Vargas
- Liver Unit, Hospital Vall d’Hebron, Vall d'Hebron Research Institute (VHIR), Universitat Autonoma Barcelona, CIBEREHD, Barcelona, Spain
| | - César Jiménez
- Liver Unit, Hospital Vall d’Hebron, Vall d'Hebron Research Institute (VHIR), Universitat Autonoma Barcelona, CIBEREHD, Barcelona, Spain
| | - Alexandre Louvet
- Hôpital Claude Huriez, Services des Maladies de l'Appareil Digestif, CHRU Lille, and Unité INSERM 995, Lille, France
| | - Guadalupe García-Tsao
- Section of Digestive Diseases, Yale University School of Medicine/VA-CT Healthcare System, New Haven/West Haven, CT, USA
| | - Juan Pablo Roblero
- Sección Gastroenterología, Hospital Clínico Universidad de Chile, Escuela de Medicina Universidad de Chile, Santiago, Chile
| | - Juan G. Abraldes
- Division of Gastroenterology, Liver Unit, University of Alberta, Edmonton, AB, Canada
| | - Vijay H. Shah
- Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA
| | - Patrick S. Kamath
- Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA
| | - Marco Arrese
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
| | - Ashwani K. Singal
- Department of Medicine, University of South Dakota Sanford School of Medicine and Transplant Hepatology, Avera Transplant Institute, Sioux Falls, SD, USA
| | | | - Juan Pablo Arab
- Departamento de Gastroenterología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
- Division of Gastroenterology, Department of Medicine, Schulich School of Medicine, Western University & London Health Sciences Centre, London, ONT, Canada
- Department of Epidemiology and Biostatistics, Schulich School of Medicine, Western University, London, ONT, Canada
| |
Collapse
|
8
|
Karamouza E, Glasspool RM, Kelly C, Lewsley LA, Carty K, Kristensen GB, Ethier JL, Kagimura T, Yanaihara N, Cecere SC, You B, Boere IA, Pujade-Lauraine E, Ray-Coquard I, Proust-Lima C, Paoletti X. CA-125 Early Dynamics to Predict Overall Survival in Women with Newly Diagnosed Advanced Ovarian Cancer Based on Meta-Analysis Data. Cancers (Basel) 2023; 15:1823. [PMID: 36980708 PMCID: PMC10047009 DOI: 10.3390/cancers15061823] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/04/2023] [Revised: 03/10/2023] [Accepted: 03/13/2023] [Indexed: 03/19/2023] Open
Abstract
(1) Background: Cancer antigen 125 (CA-125) is a protein produced by ovarian cancer cells that is used for patients' monitoring. However, the best ways to analyze its decline and prognostic role are poorly quantified. (2) Methods: We leveraged individual patient data from the Gynecologic Cancer Intergroup (GCIG) meta-analysis (N = 5573) to compare different approaches summarizing the early trajectory of CA-125 before the prediction time (called the landmark time) at 3 or 6 months after treatment initiation in order to predict overall survival. These summaries included observed and estimated measures obtained by a linear mixed model (LMM). Their performances were evaluated by 10-fold cross-validation with the Brier score and the area under the ROC (AUC). (3) Results: The estimated value and the last observed value at 3 months were the best measures used to predict overall survival, with an AUC of 0.75 CI 95% [0.70; 0.80] at 24 and 36 months and 0.74 [0.69; 0.80] and 0.75 [0.69; 0.80] at 48 months, respectively, considering that CA-125 over 6 months did not improve the AUC, with 0.74 [0.68; 0.78] at 24 months and 0.71 [0.65; 0.76] at 36 and 48 months. (4) Conclusions: A 3-month surveillance provided reliable individual information on overall survival until 48 months for patients receiving first-line chemotherapy.
Collapse
Affiliation(s)
- Eleni Karamouza
- Gustave Roussy, Office of Biostatistics and Epidemiology, Université Paris-Saclay, 94805 Villejuif, France
- Oncostat, Labeled Ligue Contre le Cancer, CESP U1018, Inserm, Université Paris-Saclay, 94805 Villejuif, France
| | - Rosalind M. Glasspool
- Beatson West of Scotland Cancer Centre, NHS Greater Glasgow and Clyde, Glasgow G12 0XH, UK
| | - Caroline Kelly
- Cancer Research UK Clinical Trials Unit, Institute of Cancer Sciences, University of Glasgow, Glasgow G12 0YN, UK
| | - Liz-Anne Lewsley
- Cancer Research UK Clinical Trials Unit, Institute of Cancer Sciences, University of Glasgow, Glasgow G12 0YN, UK
| | - Karen Carty
- Cancer Research UK Clinical Trials Unit, Institute of Cancer Sciences, University of Glasgow, Glasgow G12 0YN, UK
| | - Gunnar B. Kristensen
- Department of Gynecologic Oncology, Institute for Cancer Genetics and Informatics, Oslo University Hospital, 0424 Oslo, Norway
| | - Josee-Lyne Ethier
- Department of Medical Oncology, Cancer Centre of Southeastern Ontario, Queen’s University, Kingston, ON K7L 3N6, Canada
| | - Tatsuo Kagimura
- Foundation for Biomedical Research and Innocation, Translational Research Center for Medical Innovation, Kobe 650-0047, Japan
| | | | - Sabrina Chiara Cecere
- Department of Urology and Gynecology, Istituto Nazionale Tumori IRCCS Fondazione G. Pascale, 80131 Napoli, Italy
| | - Benoit You
- EMR UCBL/HCL 3738, Faculté de Médecine Lyon-Sud, Université Lyon, Université Claude Bernard Lyon 1, 69100 Lyon, France
- Medical Oncology, Institut de Cancérologie des Hospices Civils de Lyon (IC-HCL), CITOHL, Centre Hospitalier Lyon-Sud, GINECO, GINEGEPS, 69495 Lyon, France
| | - Ingrid A. Boere
- Department of Medical Oncology, Erasmus MC Cancer Institute, 3015 GD Rotterdam, The Netherlands
| | | | | | - Cécile Proust-Lima
- UMR1219, Bordeaux Population Health Research Center, Inserm, University of Bordeaux, 33000 Bordeaux, France
| | - Xavier Paoletti
- Faculty of Medicine, University of Versailles Saint-Quentin, Université Paris Saclay, 78000 Versailles, France
- INSERM U900, Statistics for Personalized Medicine, Institut Curie, 92210 Saint-Cloud, France
| |
Collapse
|
9
|
McLernon DJ, Giardiello D, Van Calster B, Wynants L, van Geloven N, van Smeden M, Therneau T, Steyerberg EW. Assessing Performance and Clinical Usefulness in Prediction Models With Survival Outcomes: Practical Guidance for Cox Proportional Hazards Models. Ann Intern Med 2023; 176:105-114. [PMID: 36571841 DOI: 10.7326/m22-0844] [Citation(s) in RCA: 25] [Impact Index Per Article: 25.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 12/27/2022] Open
Abstract
Risk prediction models need thorough validation to assess their performance. Validation of models for survival outcomes poses challenges due to the censoring of observations and the varying time horizon at which predictions can be made. This article describes measures to evaluate predictions and the potential improvement in decision making from survival models based on Cox proportional hazards regression. As a motivating case study, the authors consider the prediction of the composite outcome of recurrence or death (the "event") in patients with breast cancer after surgery. They developed a simple Cox regression model with 3 predictors, as in the Nottingham Prognostic Index, in 2982 women (1275 events over 5 years of follow-up) and externally validated this model in 686 women (285 events over 5 years). Improvement in performance was assessed after the addition of progesterone receptor as a prognostic biomarker. The model predictions can be evaluated across the full range of observed follow-up times or for the event occurring by the end of a fixed time horizon of interest. The authors first discuss recommended statistical measures that evaluate model performance in terms of discrimination, calibration, or overall performance. Further, they evaluate the potential clinical utility of the model to support clinical decision making according to a net benefit measure. They provide SAS and R code to illustrate internal and external validation. The authors recommend the proposed set of performance measures for transparent reporting of the validity of predictions from survival models.
Collapse
Affiliation(s)
- David J McLernon
- Institute of Applied Health Sciences, University of Aberdeen, Aberdeen, United Kingdom (D.J.M.)
| | - Daniele Giardiello
- Netherlands Cancer Institute, Amsterdam, the Netherlands, Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands, and Institute of Biomedicine, Eurac Research, Affiliated Institute of the University of Lübeck, Bolzano, Italy (D.G.)
| | - Ben Van Calster
- Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands, and Department of Development and Regeneration, Katholieke Universiteit Leuven, Leuven, Belgium (B.V.)
| | - Laure Wynants
- School for Public Health and Primary Care, Maastricht University, Maastricht, the Netherlands (L.W.)
| | - Nan van Geloven
- Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands (N.V., E.W.S.)
| | - Maarten van Smeden
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands (M.V.)
| | - Terry Therneau
- Department of Quantitative Health Sciences, Mayo Clinic, Rochester, Minnesota (T.T.)
| | - Ewout W Steyerberg
- Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands (N.V., E.W.S.)
| |
Collapse
|
10
|
Barcellini A, Fontana G, Filippini DM, Ronchi S, Bonora M, Vischioni B, Ingargiola R, Camarda AM, Loap P, Facchinetti N, Licitra L, Baroni G, Orlandi E. Exploring the role of neutrophil-to-lymphocyte ratio and blood chemistry in head and neck adenoid cystic carcinomas treated with carbon ion radiotherapy. Radiother Oncol 2022; 177:143-151. [PMID: 36328091 DOI: 10.1016/j.radonc.2022.10.027] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/02/2022] [Revised: 09/26/2022] [Accepted: 10/23/2022] [Indexed: 11/06/2022]
Abstract
BACKGROUND AND PURPOSE In recent years, there is an emerging interest in the prognostic role of chemistry blood biomarkers in oncological patients but their role in adenoid cystic carcinomas (ACCs) is still unknown. This study aims to assess the prognostic significance of baseline neutrophil-to-lymphocyte ratio (NLR) and blood chemistry in a series of head and neck ACC patients treated with carbon ion radiotherapy (CIRT). MATERIAL AND METHODS We retrospectively retrieved the data of 49 consecutive head and neck ACC patients treated with CIRT. Univariable and multivariable Cox proportional hazard regression (Cox-ph) analyses were performed to look for a potential association of NLR, and other blood biomarker values, with disease-free survival (DFS), Local Control (LC), Metastasis Free Survival (MFS) and overall survival (OS). RESULTS No significant association between NLR > 2,5 and DFS, LC, MFS and OS was found with univariable analysis although a trend was reported for DFS (Hazard ratio [HR]: 2,10, 95 % CI: 0,85 - 5,08, p-value = 0,11). Patients with hemoglobin (hb) ≤ 14 g/dL showed significantly better DFS, MFS and OS. Multivariable regression Cox-ph analysis for DFS, adjusted for margin status, clinical target volume and Absolute Number of Monocytes, reported the following statistically significant HRs, for both NLR > 2,5 and hb > 14 g/dL respectively: 4,850 (95 % CI = 1,408 - 16,701, p = 0,012) and 3,032 (95 % CI = 1,095 - 8,393, p = 0,033). Moreover, hb > 14 with HR = 3,69 (95 % CI: 1,23 - 11,07, p-value = 0,02), was a negative independent prognostic predictor for MFS. CONCLUSIONS Pre-treatment NLR and hb values seem to be independent prognostic predictor for clinical outcomes in head and neck ACC patients. If their role will be validated in a larger prospective cohort, they might be worthwhile for a pre-treatment risk stratification in patients treated with CIRT.
Collapse
Affiliation(s)
- Amelia Barcellini
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Giulia Fontana
- Clinical Bioengineering Unit, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Daria Maria Filippini
- Division of Medical Oncology, IRCCS Azienda Ospedaliero-Universitaria Policlinico Sant'Orsola Malpighi, Bologna, Italy
| | - Sara Ronchi
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy.
| | - Maria Bonora
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Barbara Vischioni
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Rossana Ingargiola
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Anna Maria Camarda
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Pierre Loap
- Department of Radiation Oncology, Institut Curie, Paris, France
| | - Nadia Facchinetti
- Scientific Direction, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| | - Lisa Licitra
- Scientific Direction, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy; Head and Neck Medical Oncology 3 Department, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy; Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
| | - Guido Baroni
- Clinical Bioengineering Unit, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy; Department of Electronics, Information and Bioengineering (DEIB), Politecnico di Milano, Milan, Italy
| | - Ester Orlandi
- Radiation Oncology Unit, Clinical Department, National Center for Oncological Hadrontherapy (CNAO), Pavia, Italy
| |
Collapse
|
11
|
Suresh K, Severn C, Ghosh D. Survival prediction models: an introduction to discrete-time modeling. BMC Med Res Methodol 2022; 22:207. [PMID: 35883032 PMCID: PMC9316420 DOI: 10.1186/s12874-022-01679-6] [Citation(s) in RCA: 14] [Impact Index Per Article: 7.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/01/2022] [Accepted: 07/08/2022] [Indexed: 12/05/2022] Open
Abstract
Background Prediction models for time-to-event outcomes are commonly used in biomedical research to obtain subject-specific probabilities that aid in making important clinical care decisions. There are several regression and machine learning methods for building these models that have been designed or modified to account for the censoring that occurs in time-to-event data. Discrete-time survival models, which have often been overlooked in the literature, provide an alternative approach for predictive modeling in the presence of censoring with limited loss in predictive accuracy. These models can take advantage of the range of nonparametric machine learning classification algorithms and their available software to predict survival outcomes. Methods Discrete-time survival models are applied to a person-period data set to predict the hazard of experiencing the failure event in pre-specified time intervals. This framework allows for any binary classification method to be applied to predict these conditional survival probabilities. Using time-dependent performance metrics that account for censoring, we compare the predictions from parametric and machine learning classification approaches applied within the discrete time-to-event framework to those from continuous-time survival prediction models. We outline the process for training and validating discrete-time prediction models, and demonstrate its application using the open-source R statistical programming environment. Results Using publicly available data sets, we show that some discrete-time prediction models achieve better prediction performance than the continuous-time Cox proportional hazards model. Random survival forests, a machine learning algorithm adapted to survival data, also had improved performance compared to the Cox model, but was sometimes outperformed by the discrete-time approaches. In comparing the binary classification methods in the discrete time-to-event framework, the relative performance of the different methods varied depending on the data set. Conclusions We present a guide for developing survival prediction models using discrete-time methods and assessing their predictive performance with the aim of encouraging their use in medical research settings. These methods can be applied to data sets that have continuous time-to-event outcomes and multiple clinical predictors. They can also be extended to accommodate new binary classification algorithms as they become available. We provide R code for fitting discrete-time survival prediction models in a github repository. Supplementary Information The online version contains supplementary material available at (10.1186/s12874-022-01679-6).
Collapse
Affiliation(s)
- Krithika Suresh
- Department of Biostatistics and Informatics, Colorado School of Public Health, Aurora, USA.
| | - Cameron Severn
- Child Health Biostatistics Core Department of Pediatrics, Section of Endocrinology, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, USA
| | - Debashis Ghosh
- Department of Biostatistics and Informatics, Colorado School of Public Health, Aurora, USA
| |
Collapse
|
12
|
Geraili Z, Hajian-Tilaki K, Bayani M, Hosseini SR, Khafri S, Ebrahimpour S, Javanian M, Babazadeh A, Shokri M. Prognostic accuracy of inflammatory markers in predicting risk of ICU admission for COVID-19: application of time-dependent receiver operating characteristic curves. J Int Med Res 2022; 50:3000605221102217. [PMID: 35701893 PMCID: PMC9208048 DOI: 10.1177/03000605221102217] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/08/2023] Open
Abstract
Objective Intensive care unit (ICU) admission occurs at different times during hospitalization among patients with COVID-19. We aimed to evaluate the time-dependent receive operating characteristic (ROC) curve and area under the ROC curve, AUC(t), and accuracy of baseline levels of inflammatory markers C-reactive protein (CRP) and neutrophil-to-lymphocyte ratio (NLR) in predicting time to an ICU admission event in patients with severe COVID-19 infection. Methods In this observational study, we evaluated 724 patients with confirmed severe COVID-19 referred to Ayatollah Rohani Hospital, affiliated with Babol University of Medical Sciences, Iran. Results The AUC(t) of CRP and NLR reached 0.741 (95% confidence interval [CI]: 0.661–0.820) and 0.690 (95% CI: 0.607–0.772), respectively, in the first 3 days after hospital admission. The optimal cutoff values of CRP and NLR for stratification of ICU admission outcomes in patients with severe COVID-19 were 78 mg/L and 5.13, respectively. The risk of ICU admission was significantly greater for patients with these cutoff values (CRP hazard ratio = 2.98; 95% CI: 1.58–5.62; NLR hazard ratio = 2.90; 95% CI: 1.45–5.77). Conclusions Using time-dependent ROC curves, CRP and NLR values at hospital admission were important predictors of ICU admission. This approach is more efficient than using standard ROC curves.
Collapse
Affiliation(s)
- Zahra Geraili
- Social Determinants of Health Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| | - Karimollah Hajian-Tilaki
- Social Determinants of Health Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran.,Department of Biostatistics and Epidemiology, School of Public Health, Babol University of Medical Sciences, Babol, Iran
| | - Masomeh Bayani
- Infectious Diseases and Tropical Medicine Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| | - Seyed Reza Hosseini
- Social Determinants of Health Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| | - Soraya Khafri
- Department of Biostatistics and Epidemiology, School of Public Health, Babol University of Medical Sciences, Babol, Iran
| | - Soheil Ebrahimpour
- Infectious Diseases and Tropical Medicine Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| | - Mostafa Javanian
- Infectious Diseases and Tropical Medicine Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| | - Arefeh Babazadeh
- Infectious Diseases and Tropical Medicine Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| | - Mehran Shokri
- Infectious Diseases and Tropical Medicine Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran
| |
Collapse
|
13
|
Bond MJG, Hamers PAH, Vink GR, van Grevenstein WMU, Laclé MM, van Smeden M, Koopman M, Roodhart JML, Punt CJA, May AM. External validation of the MSKCC nomogram to estimate five-year overall survival after surgery for stage I-III colon cancer in a Dutch population. Acta Oncol 2022; 61:560-565. [PMID: 35253593 DOI: 10.1080/0284186x.2022.2044514] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/01/2022]
Abstract
INTRODUCTION The Memorial Sloan Kettering Cancer Centre (MSKCC) nomogram has been developed to estimate five-year overall survival (OS) after curative-intent surgery of colon cancer based on age, sex, T stage, differentiation grade, number of positive and examined regional lymph nodes. This is the first evaluation of the performance of the MSKCC model in a European population regarding prediction of OS. MATERIAL AND METHODS Population-based data from patients with stage I-III colon cancer diagnosed between 2010 and 2016 were obtained from the Netherlands Cancer Registry (NCR) for external validation of the MSKCC prediction model. Five-year survival probabilities were estimated for all patients in our dataset by using the MSKCC prediction equation. Histogram density plots were created to depict the distribution of the estimated probability and prognostic index. The performance of the model was evaluated in terms of its overall performance, discrimination, and calibration. RESULTS A total of 39,805 patients were included. Five-year OS was 71.9% (95% CI 71.5; 72.3) (11,051 events) with a median follow up of 5.6 years (IQR 4.1; 7.7). The Brier score was 0.10 (95% CI 0.10; 0.10). The C-index was 0.75 (95% CI 0.75; 0.76). The calibration measures and plot indicated that the model slightly overestimated observed mortality (observed/expected ratio = 0.86 [95% CI 0.86; 0.87], calibration intercept = -0.14 [95% CI -0.16; -0.11], and slope 1.07 [95% CI 1.05; 1.09], ICI = 0.04, E50 = 0.04, and E90 = 0.05). CONCLUSIONS The external validation of the MSKCC prediction nomogram in a large Dutch cohort supports the use of this practical tool in the European patient population. These personalised estimated survival probabilities may support clinicians when informing patients about prognosis. Adding potential relevant prognostic factors to the model, such as primary tumour location, might further improve the model.
Collapse
Affiliation(s)
- Marinde J. G. Bond
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Patricia A. H. Hamers
- Department of Medical Oncology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Geraldine R. Vink
- Department of Medical Oncology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
- Department of Research and Development, Netherlands Comprehensive Cancer Organisation, Utrecht, The Netherlands
| | | | - Miangela M. Laclé
- Department of Pathology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Maarten van Smeden
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Miriam Koopman
- Department of Medical Oncology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Jeanine M. L. Roodhart
- Department of Medical Oncology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Cornelis J. A. Punt
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| | - Anne M. May
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
| |
Collapse
|
14
|
Neumann JT, Thao LTP, Murray AM, Callander E, Carr PR, Nelson MR, Wolfe R, Woods RL, Reid CM, Shah RC, Newman AB, Williamson JD, Tonkin AM, McNeil JJ. Prediction of disability-free survival in healthy older people. GeroScience 2022; 44:1641-1655. [PMID: 35420334 PMCID: PMC9213595 DOI: 10.1007/s11357-022-00547-x] [Citation(s) in RCA: 9] [Impact Index Per Article: 4.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/08/2021] [Accepted: 03/16/2022] [Indexed: 11/04/2022] Open
Abstract
Prolonging survival in good health is a fundamental societal goal. However, the leading determinants of disability-free survival in healthy older people have not been well established. Data from ASPREE, a bi-national placebo-controlled trial of aspirin with 4.7 years median follow-up, was analysed. At enrolment, participants were healthy and without prior cardiovascular events, dementia or persistent physical disability. Disability-free survival outcome was defined as absence of dementia, persistent disability or death. Selection of potential predictors from amongst 25 biomedical, psychosocial and lifestyle variables including recognized geriatric risk factors, utilizing a machine-learning approach. Separate models were developed for men and women. The selected predictors were evaluated in a multivariable Cox proportional hazards model and validated internally by bootstrapping. We included 19,114 Australian and US participants aged ≥65 years (median 74 years, IQR 71.6-77.7). Common predictors of a worse prognosis in both sexes included higher age, lower Modified Mini-Mental State Examination score, lower gait speed, lower grip strength and abnormal (low or elevated) body mass index. Additional risk factors for men included current smoking, and abnormal eGFR. In women, diabetes and depression were additional predictors. The biased-corrected areas under the receiver operating characteristic curves for the final prognostic models at 5 years were 0.72 for men and 0.75 for women. Final models showed good calibration between the observed and predicted risks. We developed a prediction model in which age, cognitive function and gait speed were the strongest predictors of disability-free survival in healthy older people.Trial registration Clinicaltrials.gov (NCT01038583).
Collapse
Affiliation(s)
- Johannes Tobias Neumann
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia. .,Department of Cardiology, University Heart & Vascular Centre Hamburg, Hamburg, Germany. .,German Centre for Cardiovascular Research (DZHK), Partner Site Hamburg/Kiel/Lübeck, Hamburg, Germany.
| | - Le T P Thao
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | - Anne M Murray
- Division of Geriatrics, Department of Medicine, Hennepin Healthcare, and Berman Centre for Outcomes and Clinical Research, Hennepin Healthcare Research Institute, Minneapolis, USA
| | - Emily Callander
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | - Prudence R Carr
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | - Mark R Nelson
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia.,Menzies Institute for Medical Research, University of Tasmania, Hobart, Australia
| | - Rory Wolfe
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | - Robyn L Woods
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | - Christopher M Reid
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia.,Curtin School of Population Health, Curtin University, Perth, WA, Australia
| | - Raj C Shah
- Department of Family Medicine and Rush Alzheimer's Disease Centre, Rush University Medical Centre, Chicago, IL, USA
| | - Anne B Newman
- Centre for Aging and Population Health, Department of Epidemiology, University of Pittsburgh, Pittsburgh, USA
| | - Jeff D Williamson
- Sticht Centre on Health Aging and Alzheimer's Prevention, Section on Gerontology and Geriatric Medicine, Department of Internal Medicine, Wake Forest School of Medicine, Winston-Salem, NC, USA
| | - Andrew M Tonkin
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | - John J McNeil
- Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, 99 Commercial Road, Melbourne, Victoria, 3004, Australia
| | | |
Collapse
|
15
|
Dynamic Prediction of Near-Term Overall Survival in Patients with Advanced NSCLC Based on Real-World Data. Cancers (Basel) 2022; 14:cancers14030690. [PMID: 35158958 PMCID: PMC8833771 DOI: 10.3390/cancers14030690] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/11/2021] [Revised: 01/18/2022] [Accepted: 01/25/2022] [Indexed: 02/01/2023] Open
Abstract
Simple Summary Patients near the end of life often receive aggressive care, which may be of low value. For patients with advanced cancers, it is standard clinical practice to estimate the prognosis to inform treatment decisions and improve end-of-life care. However, clinical estimates of prognosis may be imprecise and rapidly become out-of-date if clinical factors that evolve over time are not incorporated. Patient prognosis is commonly estimated based on a clinician’s subjective assessment of patient reserve, such as performance status. We propose a spline-smoothed landmarking approach to dynamically estimate survival probabilities based on objective, evolving patient features. The proposed method allows predictions at any time during the patient disease course and demonstrates dramatically improved prediction accuracy compared to methods using clinical features at a fixed time. The proposed approaches can assist clinicians and patients in appropriately regulating treatments to improve outcomes and quality of life. Abstract Patients with terminal cancers commonly receive aggressive and sub-optimal treatment near the end of life, which may not be beneficial in terms of duration or quality of life. To improve end-of-life care, it is essential to develop methods that can accurately predict the short-term risk of death. However, most prediction models for patients with cancer are static in the sense that they only use patient features at a fixed time. We proposed a dynamic prediction model (DPM) that can incorporate time-dependent predictors. We apply this method to patients with advanced non-small-cell lung cancer from a real-world database. Inverse probability of censoring weighted AUC with bootstrap inference was used to compare predictions among models. We found that increasing ECOG performance status and decreasing albumin had negative prognostic associations with overall survival (OS). Moreover, the negative prognostic implications strengthened over the patient disease course. DPMs using both time-independent and time-dependent predictors substantially improved short-term prediction accuracy compared to Cox models using only predictors at a fixed time. The proposed model can be broadly applied for prediction based on longitudinal data, including an estimation of the dynamic effects of time-dependent features on OS and updating predictions at any follow-up time.
Collapse
|
16
|
Nuño MM, Gillen DL. Censoring-robust time-dependent receiver operating characteristic curve estimators. Stat Med 2021; 40:6885-6899. [PMID: 34658036 PMCID: PMC8671363 DOI: 10.1002/sim.9216] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/05/2021] [Revised: 09/03/2021] [Accepted: 09/18/2021] [Indexed: 01/01/2023]
Abstract
Time-dependent receiver operating characteristic curves are often used to evaluate the classification performance of continuous measures when considering time-to-event data. When one is interested in evaluating the predictive performance of multiple covariates, it is common to use the Cox proportional hazards model to obtain risk scores; however, previous work has shown that when the model is mis-specified, the estimand corresponding to the partial likelihood estimator depends on the censoring distribution. In this manuscript, we show that when the risk score model is mis-specified, the AUC will also depend on the censoring distribution, leading to either over- or under-estimation of the risk score's predictive performance. We propose the use of censoring-robust estimators to remove the dependence on the censoring distribution and provide empirical results supporting the use of censoring-robust risk scores.
Collapse
Affiliation(s)
- Michelle M. Nuño
- Department of Preventive Medicine, University of Southern California, CA, United States
- Children’s Oncology Group, CA, United States
| | - Daniel L. Gillen
- Department of Statistics, University of California, Irvine, CA, United States
| |
Collapse
|
17
|
Receiver operating characteristic (ROC) movies, universal ROC (UROC) curves, and coefficient of predictive ability (CPA). Mach Learn 2021. [DOI: 10.1007/s10994-021-06114-3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/25/2022]
Abstract
AbstractThroughout science and technology, receiver operating characteristic (ROC) curves and associated area under the curve ($$\mathrm{AUC}$$
AUC
) measures constitute powerful tools for assessing the predictive abilities of features, markers and tests in binary classification problems. Despite its immense popularity, ROC analysis has been subject to a fundamental restriction, in that it applies to dichotomous (yes or no) outcomes only. Here we introduce ROC movies and universal ROC (UROC) curves that apply to just any linearly ordered outcome, along with an associated coefficient of predictive ability ($${\mathrm{CPA}}$$
CPA
) measure. $${\mathrm{CPA}}$$
CPA
equals the area under the UROC curve, and admits appealing interpretations in terms of probabilities and rank based covariances. For binary outcomes $${\mathrm{CPA}}$$
CPA
equals $$\mathrm{AUC}$$
AUC
, and for pairwise distinct outcomes $${\mathrm{CPA}}$$
CPA
relates linearly to Spearman’s coefficient, in the same way that the C index relates linearly to Kendall’s coefficient. ROC movies, UROC curves, and $${\mathrm{CPA}}$$
CPA
nest and generalize the tools of classical ROC analysis, and are bound to supersede them in a wealth of applications. Their usage is illustrated in data examples from biomedicine and meteorology, where rank based measures yield new insights in the WeatherBench comparison of the predictive performance of convolutional neural networks and physical-numerical models for weather prediction.
Collapse
|
18
|
Liu C, Sun Y, Yang Y, Feng Y, Xie X, Qi L, Liu K, Wang X, Zhu Q, Zhao X. Gadobenate dimeglumine-enhanced biliary imaging from the hepatobiliary phase can predict progression in patients with liver cirrhosis. Eur Radiol 2021; 31:5840-5850. [PMID: 33533990 DOI: 10.1007/s00330-021-07702-6] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/15/2020] [Revised: 12/02/2020] [Accepted: 01/19/2021] [Indexed: 11/27/2022]
Abstract
OBJECTIVES To determine the value of gadobenate dimeglumine (Gd-BOPTA)-enhanced biliary imaging from the hepatobiliary phase in predicting hepatic decompensation and insufficiency for patients with cirrhosis. METHODS This single-center retrospective study included 270 patients who underwent Gd-BOPTA-enhanced magnetic resonance imaging. The relative enhancement ratios of the biliary system (REB) and liver parenchyma (REL) in patients with normal liver function without underlying chronic liver disease and three groups of patients with Child-Pugh A, Child-Pugh B, and Child-Pugh C disease were measured. After a mean follow-up of 38.5 ± 22.5 months, prognostic factors were evaluated using the Cox proportional hazards regression model. Receiver operating characteristic (ROC) curve analyses were performed to assess the capacity of the REB and REL to predict the development of hepatic decompensation and insufficiency. RESULTS During the follow-up period, nine of 79 patients with Child-Pugh A disease developed hepatic decompensation. The REB was a significant predictive factor (hazard ratio (HR) = 0.40 (0.19-0.84); p = 0.016), but the REL showed no association with hepatic decompensation. Moreover, the areas under the ROC curves (AUCs) were 0.83 and 0.52 for the REB and REL, respectively. Thirty-eight of 207 patients with cirrhosis developed hepatic insufficiency. The REB was a significant predictive factor (HR = 0.24 (0.13-0.46); p < 0.0001), but the REL did not show statistically significant association with hepatic insufficiency. The AUCs were 0.82 and 0.57 for the REB and REL, respectively. CONCLUSIONS Gd-BOPTA-enhanced biliary imaging from the hepatobiliary phase was valuable in predicting hepatic decompensation and insufficiency for cirrhotic patients. KEY POINTS • Gd-BOPTA-enhanced biliary imaging was a significant predictive factor for hepatic decompensation in patients with cirrhosis. • Gd-BOPTA-enhanced biliary imaging was a significant predictive factor for hepatic insufficiency in patients with cirrhosis. • Gd-BOPTA-enhanced biliary imaging showed superior predictive values for adverse clinical outcomes compared to liver parenchymal imaging at the hepatobiliary phase.
Collapse
Affiliation(s)
- Chenxi Liu
- Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
- Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Yan Sun
- Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Shandong Province, 324#, Jing 5 Rd, Ji'nan, 250021, Shandong Province, China
- Department of Radiology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
| | - Yao Yang
- Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
- Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Yuemin Feng
- Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
- Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Xiaoyu Xie
- Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
- Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Lingyu Qi
- Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
- Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Keke Liu
- Shandong Academy of Clinical Medicine, Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Ximing Wang
- Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Shandong Province, 324#, Jing 5 Rd, Ji'nan, 250021, Shandong Province, China
- Department of Radiology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
| | - Qiang Zhu
- Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China
- Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong province, China
| | - Xinya Zhao
- Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Shandong Province, 324#, Jing 5 Rd, Ji'nan, 250021, Shandong Province, China.
- Department of Radiology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, 250021, Shandong province, China.
| |
Collapse
|
19
|
van Geloven N, He Y, Zwinderman A, Putter H. Estimation of incident dynamic AUC in practice. Comput Stat Data Anal 2021. [DOI: 10.1016/j.csda.2020.107095] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/23/2022]
|
20
|
Beyene KM, El Ghouch A. Smoothed time-dependent receiver operating characteristic curve for right censored survival data. Stat Med 2020; 39:3373-3396. [PMID: 32687225 DOI: 10.1002/sim.8671] [Citation(s) in RCA: 7] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/26/2019] [Revised: 03/30/2020] [Accepted: 06/05/2020] [Indexed: 11/08/2022]
Abstract
The prediction reliability is of primary concern in many clinical studies when the objective is to develop new predictive models or improve existing risk scores. In fact, before using a model in any clinical decision making, it is very important to check its ability to discriminate between subjects who are at risk of, for example, developing certain disease in a near future from those who will not. To that end, the time-dependent receiver operating characteristic (ROC) curve is the most commonly used method in practice. Several approaches have been proposed in the literature to estimate the ROC nonparametrically in the context of survival data. But, except one recent approach, all the existing methods provide a nonsmooth ROC estimator whereas, by definition, the ROC curve is smooth. In this article we propose and study a new nonparametric smooth ROC estimator based on a weighted kernel smoother. More precisely, our approach relies on a well-known kernel method used to estimate cumulative distribution functions of random variables with bounded supports. We derived some asymptotic properties for the proposed estimator. As bandwidth is the main parameter to be set, we present and study different methods to appropriately select one. A simulation study is conducted, under different scenarios, to prove the consistency of the proposed method and to compare its finite sample performance with a competitor. The results show that the proposed method performs better and appear to be quite robust to bandwidth choice. As for inference purposes, our results also reveal the good performances of a proposed nonparametric bootstrap procedure. Furthermore, we illustrate the method using a real data example.
Collapse
Affiliation(s)
- Kassu Mehari Beyene
- Institute of Statistics, Biostatistics and Actuarial Sciences, Catholic University of Louvain, Louvain la Neuve, Belgium
| | - Anouar El Ghouch
- Institute of Statistics, Biostatistics and Actuarial Sciences, Catholic University of Louvain, Louvain la Neuve, Belgium
| |
Collapse
|
21
|
Amico M, Van Keilegom I, Han B. Assessing cure status prediction from survival data using receiver operating characteristic curves. Biometrika 2020. [DOI: 10.1093/biomet/asaa080] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/13/2022] Open
Abstract
Summary
Survival analysis relies on the hypothesis that, if the follow-up is long enough, the event of interest will eventually be observed for all observations. This assumption, however, is often not realistic. The survival data then contain a cure fraction. A common approach to modelling and analysing this type of data consists in using cure models. Two types of information can therefore be obtained: the survival at a given time and the cure status, both possibly modelled as a function of the covariates. The cure status is often of interest to medical practitioners, and one is usually interested in predicting it based on markers. Receiver operating characteristic, Roc, curves are one way to evaluate the predicted performance; however, the classical Roc curve method is not appropriate since the cure status is partially unobserved due to the presence of censoring in survival data. We propose a Roc curve estimator that aims to evaluate the cured/noncured status classification performance from cure survival data. This estimator, which handles the presence of censoring, decomposes sensitivity and specificity by means of the definition of conditional probability, and estimates these two quantities by means of weighted empirical distribution functions. The mixture cure model is used to calculate the weights. Based on simulations, we demonstrate good performance of the proposed method, and compare it with the classical Roc curve nonparametric estimator that would be obtained if the cure status was fully observed. We also compare our proposed method with the Roc curves of Heagerty et al. (2000) for classical survival analysis. Finally, we illustrate the methodology on a breast cancer dataset.
Collapse
Affiliation(s)
- M Amico
- Research Centre for Operations Research and Statistics, KU Leuven, Naamsestaat 69, 3000 Leuven, Belgium
| | - I Van Keilegom
- Research Centre for Operations Research and Statistics, KU Leuven, Naamsestaat 69, 3000 Leuven, Belgium
| | - B Han
- Research Centre for Operations Research and Statistics, KU Leuven, Naamsestaat 69, 3000 Leuven, Belgium
| |
Collapse
|
22
|
Díaz-Coto S, Martínez-Camblor P, Pérez-Fernández S. smoothROCtime: an R package for time-dependent ROC curve estimation. Comput Stat 2020. [DOI: 10.1007/s00180-020-00955-7] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/01/2022]
|
23
|
Díaz-Coto S, Corral-Blanco NO, Martínez-Camblor P. Two-stage receiver operating-characteristic curve estimator for cohort studies. Int J Biostat 2020; 17:117-137. [PMID: 32862149 DOI: 10.1515/ijb-2019-0097] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/27/2019] [Accepted: 05/25/2020] [Indexed: 12/22/2022]
Abstract
The receiver operating-characteristic (ROC) curve is a graphical statistical tool routinely used for studying the classification accuracy in both, diagnostic and prognosis problems. Given the different nature of these situations, ROC curve estimation has been separately considered for binary (diagnostic) and time-to-event (prognosis) outcomes, even for data coming from the same study design. In this work, the authors propose a two-stage ROC curve estimator which allows to link both contexts through a general prediction model (first-stage) and the empirical cumulative estimator of the distribution function (second-stage) of the considered test (marker) on the total population. The so-called two-stage Mixed-Subject (sMS) approach proves its behavior on both, large-samples (theoretically) and finite-samples (via Monte Carlo simulations). Besides, a useful asymptotic distribution for the concomitant area under the curve is also computed. Results show the ability of the proposed estimator to fit non-standard situations by considering flexible predictive models. Two real-world examples, one with binary and one with time-dependent outcomes, help us to a better understanding of the proposed methodology on usual practical circumstances. The R code used for the practical implementation of the proposed methodology and its documentation is provided as supplementary material.
Collapse
Affiliation(s)
| | | | - Pablo Martínez-Camblor
- Biomedical Data Science Department, Geisel school of Medicine at Dartmouth, Hanover, NH, USA
| |
Collapse
|
24
|
Díaz-Coto S, Martínez-Camblor P, Corral-Blanco NO. Cumulative/dynamic ROC curve estimation under interval censorship. J STAT COMPUT SIM 2020. [DOI: 10.1080/00949655.2020.1736071] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/20/2022]
Affiliation(s)
- Susana Díaz-Coto
- Department of Statistics, University of Oviedo, Oviedo, Asturias, Spain
| | | | | |
Collapse
|
25
|
Blangero Y, Rabilloud M, Laurent-Puig P, Le Malicot K, Lepage C, Ecochard R, Taieb J, Subtil F. The area between ROC curves, a non-parametric method to evaluate a biomarker for patient treatment selection. Biom J 2020; 62:1476-1493. [PMID: 32346912 DOI: 10.1002/bimj.201900171] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/05/2019] [Revised: 09/26/2019] [Accepted: 01/10/2020] [Indexed: 12/19/2022]
Abstract
Treatment selection markers are generally sought for when the benefit of an innovative treatment in comparison with a reference treatment is considered, and this benefit is suspected to vary according to the characteristics of the patients. Classically, such quantitative markers are detected through testing a marker-by-treatment interaction in a parametric regression model. Most alternative methods rely on modeling the risk of event occurrence in each treatment arm or the benefit of the innovative treatment over the marker values, but with assumptions that may be difficult to verify. Herein, a simple non-parametric approach is proposed to detect and assess the general capacity of a quantitative marker for treatment selection when no overall difference in efficacy could be demonstrated between two treatments in a clinical trial. This graphical method relies on the area between treatment-arm-specific receiver operating characteristic curves (ABC), which reflects the treatment selection capacity of the marker. A simulation study assessed the inference properties of the ABC estimator and compared them with other parametric and non-parametric indicators. The simulations showed that the estimate of the ABC had low bias, power comparable to parametric indicators, and that its confidence interval had a good coverage probability (better than the other non-parametric indicator in some cases). Thus, the ABC is a good alternative to parametric indicators. The ABC method was applied to data of the PETACC-8 trial that investigated FOLFOX4 versus FOLFOX4 + cetuximab in stage III colon adenocarcinoma. It enabled the detection of a treatment selection marker: the DDR2 gene.
Collapse
Affiliation(s)
- Yoann Blangero
- Service de Biostatistique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| | - Muriel Rabilloud
- Service de Biostatistique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| | - Pierre Laurent-Puig
- Université Paris Descartes, Sorbonne Paris Cité, Paris, France.,Service de génétique, Hôpital Européen Georges Pompidou, Paris, France.,INSERM UMR-S 1147, Paris, France
| | | | - Côme Lepage
- Fédération Francophone de Cancérologie Digestive, Dijon, France.,Hépato-gastroentérologie et cancérologie digestive, Centre hospitalier universitaire Dijon Bourgogne, Dijon, France.,INSERM U 866, Dijon, France
| | - René Ecochard
- Service de Biostatistique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| | - Julien Taieb
- Université Paris Descartes, Sorbonne Paris Cité, Paris, France.,Chirurgie digestive générale et cancérologique, Hôpital Européen Georges Pompidou, Paris, France
| | - Fabien Subtil
- Service de Biostatistique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| |
Collapse
|
26
|
Jacquemont L, Tilly G, Yap M, Doan-Ngoc TM, Danger R, Guérif P, Delbos F, Martinet B, Giral M, Foucher Y, Brouard S, Degauque N. Terminally Differentiated Effector Memory CD8 + T Cells Identify Kidney Transplant Recipients at High Risk of Graft Failure. J Am Soc Nephrol 2020; 31:876-891. [PMID: 32165419 DOI: 10.1681/asn.2019080847] [Citation(s) in RCA: 38] [Impact Index Per Article: 9.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/27/2019] [Accepted: 01/16/2020] [Indexed: 12/17/2022] Open
Abstract
BACKGROUND Identifying biomarkers to predict kidney transplant failure and to define new therapeutic targets requires more comprehensive understanding of the immune response to chronic allogeneic stimulation. METHODS We investigated the frequency and function of CD8+ T cell subsets-including effector memory (EM) and terminally differentiated EM (TEMRA) CD8+ T cells-in blood samples from 284 kidney transplant recipients recruited 1 year post-transplant and followed for a median of 8.3 years. We also analyzed CD8+ T cell reactivity to donor-specific PBMCs in 24 patients who had received living-donor kidney transplants. RESULTS Increased frequency of circulating TEMRA CD8+ T cells at 1 year post-transplant associated with increased risk of graft failure during follow-up. This association remained after adjustment for a previously reported composite of eight clinical variables, the Kidney Transplant Failure Score. In contrast, increased frequency of EM CD8+ T cells associated with reduced risk of graft failure. A distinct TEMRA CD8+ T cell subpopulation was identified that was characterized by expression of FcγRIIIA (CD16) and by high levels of proinflammatory cytokine secretion and cytotoxic activity. Although donor-specific stimulation induced a similar rapid, early response in EM and TEMRA CD8+ T cells, CD16 engagement resulted in selective activation of TEMRA CD8+ T cells, which mediated antibody-dependent cytotoxicity. CONCLUSIONS At 1 year post-transplant, the composition of memory CD8+ T cell subsets in blood improved prediction of 8-year kidney transplant failure compared with a clinical-variables score alone. A subpopulation of TEMRA CD8+ T cells displays a novel dual mechanism of activation mediated by engagement of the T-cell receptor or of CD16. These findings suggest that TEMRA CD8+ T cells play a pivotal role in humoral and cellular rejection and reveal the potential value of memory CD8+ T cell monitoring for predicting risk of kidney transplant failure.
Collapse
Affiliation(s)
- Lola Jacquemont
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Gaëlle Tilly
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Michelle Yap
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Tra-My Doan-Ngoc
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Richard Danger
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | | | | | - Bernard Martinet
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Magali Giral
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Yohann Foucher
- INSERM, Université de Nantes, methodS in Patient-centered outcomes and HEalth ResEarch (SPHERE), UMR1246, Nantes, France
| | - Sophie Brouard
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France.,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| | - Nicolas Degauque
- Université de Nantes, Inserm, Centre de Recherche en Transplantation et Immunologie (CRTI), UMR 1064, Nantes, France; .,CHU Nantes, Université de Nantes, ITUN, Nantes, France
| |
Collapse
|
27
|
Prognostic Value of Lymph Node-To-Primary Tumor Standardized Uptake Value Ratio in Esophageal Squamous Cell Carcinoma Treated with Definitive Chemoradiotherapy. Cancers (Basel) 2020; 12:cancers12030607. [PMID: 32155748 PMCID: PMC7139766 DOI: 10.3390/cancers12030607] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/28/2020] [Revised: 03/01/2020] [Accepted: 03/05/2020] [Indexed: 02/08/2023] Open
Abstract
We aimed to investigate the prognostic value of the relative maximum standardized uptake value (SUV) of metastatic lymph node (LN) compared with that of primary tumor (SUVLN / SUVTumor) based on a pretreatment [18F]-FDG PET/CT scan in patients with clinically node-positive esophageal squamous cell carcinoma (cN+ ESCC) treated with definitive chemoradiotherapy (dCRT). We retrospectively evaluated cN+ ESCC patients who underwent a PET/CT scan before dCRT. Time-dependent receiver operating characteristics analysis was performed to identify the optimal cutoff value for SUVLN / SUVTumor. Prognostic influences of SUVLN / SUVTumor on distant metastasis-free survival (DMFS) and overall survival (OS) were evaluated using the Kaplan-Meier method and log-rank test for univariate analysis and Cox's proportional hazards regression model for multivariate analysis. We identified 112 patients with newly diagnosed cN+ ESCC. After a median follow-up of 32.0 months, 50 (44.6%) patients had distant failure and 84 (75.0%) patients died. Patients with high SUVLN / SUVTumor (≥ 0.39) experienced worse outcomes than low SUVLN / SUVTumor (< 0.39) (two-year DMFS: 26% vs. 70%, p < 0.001; two-year OS: 21% vs. 48%, p = 0.001). Multivariate analysis showed that SUVLN / SUVTumor was an independent prognostic factor for both DMFS (adjusted HR 2.24, 95% CI 1.34-3.75, p = 0.002) and OS (adjusted HR 1.61, 95% CI 1.03-2.53, p = 0.037). Pretreatment of SUVLN / SUVTumor is a simple and useful marker for prognosticating DMFS and OS in cN+ ESCC patients treated with dCRT, which may help in tailoring treatment and designing future clinical trials.
Collapse
|
28
|
Debray TPA, Damen JAAG, Riley RD, Snell K, Reitsma JB, Hooft L, Collins GS, Moons KGM. A framework for meta-analysis of prediction model studies with binary and time-to-event outcomes. Stat Methods Med Res 2019; 28:2768-2786. [PMID: 30032705 PMCID: PMC6728752 DOI: 10.1177/0962280218785504] [Citation(s) in RCA: 101] [Impact Index Per Article: 20.2] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/17/2022]
Abstract
It is widely recommended that any developed-diagnostic or prognostic-prediction model is externally validated in terms of its predictive performance measured by calibration and discrimination. When multiple validations have been performed, a systematic review followed by a formal meta-analysis helps to summarize overall performance across multiple settings, and reveals under which circumstances the model performs suboptimal (alternative poorer) and may need adjustment. We discuss how to undertake meta-analysis of the performance of prediction models with either a binary or a time-to-event outcome. We address how to deal with incomplete availability of study-specific results (performance estimates and their precision), and how to produce summary estimates of the c-statistic, the observed:expected ratio and the calibration slope. Furthermore, we discuss the implementation of frequentist and Bayesian meta-analysis methods, and propose novel empirically-based prior distributions to improve estimation of between-study heterogeneity in small samples. Finally, we illustrate all methods using two examples: meta-analysis of the predictive performance of EuroSCORE II and of the Framingham Risk Score. All examples and meta-analysis models have been implemented in our newly developed R package "metamisc".
Collapse
Affiliation(s)
- Thomas PA Debray
- Julius Center for Health Sciences and
Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
- Cochrane Netherlands, University Medical
Center Utrecht, Utrecht, The Netherlands
| | - Johanna AAG Damen
- Julius Center for Health Sciences and
Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
- Cochrane Netherlands, University Medical
Center Utrecht, Utrecht, The Netherlands
| | - Richard D Riley
- Research Institute for Primary Care and
Health Sciences, Keele University, Staffordshire, UK
| | - Kym Snell
- Research Institute for Primary Care and
Health Sciences, Keele University, Staffordshire, UK
| | - Johannes B Reitsma
- Julius Center for Health Sciences and
Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
- Cochrane Netherlands, University Medical
Center Utrecht, Utrecht, The Netherlands
| | - Lotty Hooft
- Julius Center for Health Sciences and
Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
- Cochrane Netherlands, University Medical
Center Utrecht, Utrecht, The Netherlands
| | - Gary S Collins
- Centre for Statistics in Medicine,
University of Oxford, Oxford, UK
| | - Karel GM Moons
- Julius Center for Health Sciences and
Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
- Cochrane Netherlands, University Medical
Center Utrecht, Utrecht, The Netherlands
| |
Collapse
|
29
|
Blanche P, Kattan MW, Gerds TA. The c-index is not proper for the evaluation of $t$-year predicted risks. Biostatistics 2019; 20:347-357. [PMID: 29462286 DOI: 10.1093/biostatistics/kxy006] [Citation(s) in RCA: 75] [Impact Index Per Article: 15.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/18/2017] [Accepted: 01/17/2018] [Indexed: 12/16/2022] Open
Abstract
We show that the widely used concordance index for time to event outcome is not proper when interest is in predicting a $t$-year risk of an event, for example 10-year mortality. In the situation with a fixed prediction horizon, the concordance index can be higher for a misspecified model than for a correctly specified model. Impropriety happens because the concordance index assesses the order of the event times and not the order of the event status at the prediction horizon. The time-dependent area under the receiver operating characteristic curve does not have this problem and is proper in this context.
Collapse
Affiliation(s)
- Paul Blanche
- Section of Biostatistics, Department of Public Health, University of Copenhagen, Oester Farimagsgade 5, 1014 Copenhagen, Denmark
| | - Michael W Kattan
- Department of Quantitative Health Sciences, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH 44195, USA
| | - Thomas A Gerds
- Section of Biostatistics, Department of Public Health, University of Copenhagen, Oester Farimagsgade 5, 1014 Copenhagen, Denmark
| |
Collapse
|
30
|
Kazmirczak F, Chen KHA, Adabag S, von Wald L, Roukoz H, Benditt DG, Okasha O, Farzaneh-Far A, Markowitz J, Nijjar PS, Velangi PS, Bhargava M, Perlman D, Duval S, Akçakaya M, Shenoy C. Assessment of the 2017 AHA/ACC/HRS Guideline Recommendations for Implantable Cardioverter-Defibrillator Implantation in Cardiac Sarcoidosis. Circ Arrhythm Electrophysiol 2019; 12:e007488. [PMID: 31431050 DOI: 10.1161/circep.119.007488] [Citation(s) in RCA: 57] [Impact Index Per Article: 11.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 01/08/2023]
Abstract
BACKGROUND Implantable cardioverter-defibrillators are used to prevent sudden cardiac death in patients with cardiac sarcoidosis. The most recent recommendations for implantable cardioverter-defibrillator implantation in these patients are in the 2017 American Heart Association/American College of Cardiology/Heart Rhythm Society Guideline for Management of Patients With Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death. These recommendations, based on observational studies or expert opinion, have not been assessed. We aimed to assess them. METHODS We performed a large retrospective cohort study of patients with biopsy-proven sarcoidosis and known or suspected cardiac sarcoidosis that underwent cardiovascular magnetic resonance imaging. Patients were followed for a composite end point of significant ventricular arrhythmia or sudden cardiac death. The discriminatory performance of the Guideline recommendations was tested using time-dependent receiver operating characteristic analyses. The optimal cutoff for the extent of late gadolinium enhancement predictive of the composite end point was determined using the Youden index. RESULTS In 290 patients, the class I and IIa recommendations identified all patients who experienced the composite end point during a median follow-up of 3.0 years. Patients meeting class I recommendations had a significantly higher incidence of the composite end point than those meeting class IIa recommendations. Left ventricular ejection fraction (LVEF) >35% with >5.7% late gadolinium enhancement on cardiovascular magnetic resonance imaging was as sensitive as and significantly more specific than LVEF >35% with any late gadolinium enhancement. Patients meeting 2 class IIa recommendations, LVEF >35% with the need for a permanent pacemaker and LVEF >35% with late gadolinium enhancement >5.7%, had high annualized event rates. Excluding 2 class IIa recommendations, LVEF >35% with syncope and LVEF >35% with inducible ventricular arrhythmia, resulted in improved discrimination for the composite end point. CONCLUSIONS We assessed the Guideline recommendations for implantable cardioverter-defibrillator implantation in patients with known or suspected cardiac sarcoidosis and identified topics for future research.
Collapse
Affiliation(s)
- Felipe Kazmirczak
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Ko-Hsuan Amy Chen
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Selcuk Adabag
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis.,Division of Cardiology, Department of Medicine, Veterans Affairs Health Care System, Minneapolis, MN (S.A.)
| | - Lisa von Wald
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Henri Roukoz
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - David G Benditt
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Osama Okasha
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Afshin Farzaneh-Far
- Section of Cardiology, Department of Medicine, University of Illinois at Chicago (A.F.-F.)
| | - Jeremy Markowitz
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Prabhjot S Nijjar
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Pratik S Velangi
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Maneesh Bhargava
- Division of Pulmonary, Allergy, Critical Care, and Sleep Medicine (M.B., D.P.), University of Minnesota Medical School, Minneapolis
| | - David Perlman
- Division of Pulmonary, Allergy, Critical Care, and Sleep Medicine (M.B., D.P.), University of Minnesota Medical School, Minneapolis
| | - Sue Duval
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| | - Mehmet Akçakaya
- Department of Electrical and Computer Engineering, Center for Magnetic Resonance Research, University of Minnesota, Minneapolis (M.A.)
| | - Chetan Shenoy
- Cardiovascular Division, Department of Medicine (F.K., K.-H.A.C., S.A., L.v.W., H.R., D.G.B., O.O., J.M., P.S.N., P.S.V., S.D., C.S.), University of Minnesota Medical School, Minneapolis
| |
Collapse
|
31
|
Beyene KM, El Ghouch A, Oulhaj A. On the validity of time-dependent AUC estimation in the presence of cure fraction. Biom J 2019; 61:1430-1447. [PMID: 31310019 DOI: 10.1002/bimj.201800376] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/09/2018] [Revised: 04/16/2019] [Accepted: 06/04/2019] [Indexed: 11/09/2022]
Abstract
During the last decades, several approaches have been proposed to estimate the time-dependent area under the receiver operating characteristic curve (AUC) of risk tools derived from survival data. The validity of these estimators relies on some regularity assumptions among which a survival function being proper. In practice, this assumption is not always satisfied because a fraction of the population may not be susceptible to experience the event of interest even for long follow-up. Studying the sensitivity of the proposed estimators to the violation of this assumption is of substantial interest. In this paper, we investigate the performance of a nonparametric simple estimator, developed for classical survival data, in the case when the population exhibits a cure fraction. Motivated from the current practice of deriving risk tools in oncology and cardiovascular disease prevention, we also assess the loss, in terms of predictive performance, when deriving risk tools from survival models that do not acknowledge the presence of cure. The simulation results show that the investigated method is valid even under the presence of cure. They also show that risk tools derived from survival models that ignore the presence of cure have smaller AUC compared to those derived from survival models that acknowledge the presence of cure. This was also attested with a real data analysis from a breast cancer study.
Collapse
Affiliation(s)
- Kassu M Beyene
- Institute of Statistics, Biostatistics and Actuarial Sciences, Catholic University of Louvain, Louvain la Neuve, Belgium
| | - Anouar El Ghouch
- Institute of Statistics, Biostatistics and Actuarial Sciences, Catholic University of Louvain, Louvain la Neuve, Belgium
| | - Abderrahim Oulhaj
- Institute of Public Health, College of Medicine and Health Sciences, UAE University, Al-Ain, United Arab Emirates
| |
Collapse
|
32
|
Tardivon C, Desmée S, Kerioui M, Bruno R, Wu B, Mentré F, Mercier F, Guedj J. Association Between Tumor Size Kinetics and Survival in Patients With Urothelial Carcinoma Treated With Atezolizumab: Implication for Patient Follow-Up. Clin Pharmacol Ther 2019; 106:810-820. [PMID: 30985002 DOI: 10.1002/cpt.1450] [Citation(s) in RCA: 27] [Impact Index Per Article: 5.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/02/2019] [Accepted: 03/21/2019] [Indexed: 12/14/2022]
Abstract
We characterized the association between tumor size kinetics and survival in patients with advanced urothelial carcinoma treated with atezolizumab (anti-programmed death-ligand 1, Tecentriq) using a joint model. The model, developed on data from 309 patients of a phase II clinical trial, identified the time-to-tumor growth and the instantaneous changes in tumor size as the best on-treatment predictors of survival. On the validation dataset containing data from 457 patients from a phase III study, the model predicted individual survival probability using 3-month or 6-month tumor size follow-up data with an area under the receptor-occupancy curve between 0.75 and 0.84, as compared with values comprised between 0.62 and 0.75 when the model included only information available at treatment initiation. Including tumor size kinetics in a relevant statistical framework improves the prediction of survival probability during immunotherapy treatment and may be useful to identify most-at-risk patients in "real-time."
Collapse
Affiliation(s)
| | - Solène Desmée
- UMR 1246, Université de Tours, Université de Nantes, Inserm SPHERE, Tours, France
| | - Marion Kerioui
- Université de Paris, IAME, INSERM, F-75018 Paris, France.,UMR 1246, Université de Tours, Université de Nantes, Inserm SPHERE, Tours, France
| | - René Bruno
- Clinical Pharmacology, Roche/Genentech, Marseille, France
| | - Benjamin Wu
- Clinical Pharmacology, Genentech Inc., South San Francisco, California, USA
| | - France Mentré
- Université de Paris, IAME, INSERM, F-75018 Paris, France
| | - François Mercier
- Clinical Pharmacology, Roche Innovation Center, Basel, Switzerland
| | - Jérémie Guedj
- Université de Paris, IAME, INSERM, F-75018 Paris, France
| |
Collapse
|
33
|
He Y, Ong Y, Li X, Din FV, Brown E, Timofeeva M, Wang Z, Farrington SM, Campbell H, Dunlop MG, Theodoratou E. Performance of prediction models on survival outcomes of colorectal cancer with surgical resection: A systematic review and meta-analysis. Surg Oncol 2019; 29:196-202. [PMID: 31196488 DOI: 10.1016/j.suronc.2019.05.014] [Citation(s) in RCA: 14] [Impact Index Per Article: 2.8] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/12/2019] [Revised: 05/07/2019] [Accepted: 05/18/2019] [Indexed: 01/19/2023]
Abstract
Prediction models allow accurate estimate of individualized prognosis. Increasing numbers of models on survival of CRC patients with surgical resection are being published. However, their performance and potential clinical utility have been unclear. A systematic search in MEDLINE and Embase databases (until 9th April 2018) was performed. Original model development studies and external validation studies predicting any survival outcomes from CRC (follow-up ≥1 year after surgery) were included. We conducted random-effects meta-analyses in external validation studies to estimate the performance of each model. A total of 83 original prediction models and 52 separate external validation studies were identified. We identified five models (Basingstoke score, Fong score, Nordinger score, Peritoneal Surface Disease Severity Score and Valentini nomogram) that were validated in at least two external datasets with a median summarized C-statistic of 0.67 (range: 0.57-0.74). These models can potentially assist clinical decision-making. Besides developing new models, future research should also focus on validating existing prediction models and investigating their real-word impact and cost-effectiveness for CRC prognosis in clinical practice.
Collapse
Affiliation(s)
- Yazhou He
- Centre for Global Health Research, Usher Institute of Population Health Sciences and Informatics, The University of Edinburgh, Edinburgh, UK; Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh, UK
| | - Yuhan Ong
- Western General Hospital, Edinburgh, UK
| | - Xue Li
- Centre for Global Health Research, Usher Institute of Population Health Sciences and Informatics, The University of Edinburgh, Edinburgh, UK
| | - Farhat Vn Din
- Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh, UK; Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK
| | - Ewan Brown
- Edinburgh Cancer Centre NHS Lothian, Edinburgh, UK
| | - Maria Timofeeva
- Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh, UK; Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK
| | - Ziqiang Wang
- Department of Gastrointestinal Surgery, West China Hospital, Sichuan University, Chengdu, 610041, PR China
| | - Susan M Farrington
- Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh, UK; Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK
| | - Harry Campbell
- Centre for Global Health Research, Usher Institute of Population Health Sciences and Informatics, The University of Edinburgh, Edinburgh, UK
| | - Malcolm G Dunlop
- Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh, UK; Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK
| | - Evropi Theodoratou
- Centre for Global Health Research, Usher Institute of Population Health Sciences and Informatics, The University of Edinburgh, Edinburgh, UK; Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK.
| |
Collapse
|
34
|
Blangero Y, Rabilloud M, Ecochard R, Subtil F. A Bayesian method to estimate the optimal threshold of a marker used to select patients' treatment. Stat Methods Med Res 2019; 29:29-43. [PMID: 30599802 DOI: 10.1177/0962280218821394] [Citation(s) in RCA: 6] [Impact Index Per Article: 1.2] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/17/2022]
Abstract
The use of a quantitative treatment selection marker to choose between two treatment options requires the estimate of an optimal threshold above which one of these two treatments is preferred. Herein, the optimal threshold expression is based on the definition of a utility function which aims to quantify the expected utility of the population (e.g. life expectancy, quality of life) by taking into account both efficacy (success or failure) and toxicity of each treatment option. Therefore, the optimal threshold is the marker value that maximizes the expected utility of the population. A method modelling the marker distribution in patient subgroups defined by the received treatment and the outcome is proposed to calculate the parameters of the utility function so as to estimate the optimal threshold and its 95% credible interval using the Bayesian inference. The simulation study found that the method had low bias and coverage probability close to 95% in multiple settings, but also the need of large sample size to estimate the optimal threshold in some settings. The method is then applied to the PETACC-8 trial that compares the efficacy of chemotherapy with a combined chemotherapy + anti-epidermal growth factor receptor in stage III colorectal cancer.
Collapse
Affiliation(s)
- Yoann Blangero
- Service de Biostatistique-Bioinformatique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| | - Muriel Rabilloud
- Service de Biostatistique-Bioinformatique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| | - René Ecochard
- Service de Biostatistique-Bioinformatique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| | - Fabien Subtil
- Service de Biostatistique-Bioinformatique, Pôle Santé Publique, Hospices Civils de Lyon, Lyon, France.,Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Evolutive UMR 5558, Villeurbanne, France
| |
Collapse
|
35
|
Wu C, Li L. Quantifying and estimating the predictive accuracy for censored time-to-event data with competing risks. Stat Med 2018; 37:3106-3124. [PMID: 29766537 DOI: 10.1002/sim.7806] [Citation(s) in RCA: 9] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/23/2017] [Revised: 03/29/2018] [Accepted: 04/11/2018] [Indexed: 01/13/2023]
Abstract
This paper focuses on quantifying and estimating the predictive accuracy of prognostic models for time-to-event outcomes with competing events. We consider the time-dependent discrimination and calibration metrics, including the receiver operating characteristics curve and the Brier score, in the context of competing risks. To address censoring, we propose a unified nonparametric estimation framework for both discrimination and calibration measures, by weighting the censored subjects with the conditional probability of the event of interest given the observed data. The proposed method can be extended to time-dependent predictive accuracy metrics constructed from a general class of loss functions. We apply the methodology to a data set from the African American Study of Kidney Disease and Hypertension to evaluate the predictive accuracy of a prognostic risk score in predicting end-stage renal disease, accounting for the competing risk of pre-end-stage renal disease death, and evaluate its numerical performance in extensive simulation studies.
Collapse
Affiliation(s)
- Cai Wu
- Department of Biostatistics, The University of Texas Health Science Center at Houston, Houston, TX, USA.,Department of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
| | - Liang Li
- Department of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
| |
Collapse
|
36
|
Can functional parameters from hepatobiliary phase of gadoxetate MRI predict clinical outcomes in patients with cirrhosis? Eur Radiol 2018; 28:4215-4224. [DOI: 10.1007/s00330-018-5366-6] [Citation(s) in RCA: 10] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/31/2017] [Revised: 01/12/2018] [Accepted: 02/01/2018] [Indexed: 12/26/2022]
|
37
|
Li L, Luo S, Hu B, Greene T. Dynamic Prediction of Renal Failure Using Longitudinal Biomarkers in a Cohort Study of Chronic Kidney Disease. STATISTICS IN BIOSCIENCES 2017; 9:357-378. [PMID: 29250207 PMCID: PMC5726783 DOI: 10.1007/s12561-016-9183-7] [Citation(s) in RCA: 16] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/16/2015] [Revised: 04/27/2016] [Accepted: 10/27/2016] [Indexed: 12/17/2022]
Abstract
In longitudinal studies, prognostic biomarkers are often measured longitudinally. It is of both scientific and clinical interest to predict the risk of clinical events, such as disease progression or death, using these longitudinal biomarkers as well as other time-dependent and time-independent information about the patient. The prediction is dynamic in the sense that it can be made at any time during the follow-up, adapting to the changing at-risk population and incorporating the most recent longitudinal data. One approach is to build a joint model of longitudinal predictor variables and time to the clinical event, and draw predictions from the posterior distribution of the time to event conditional on longitudinal history. Another approach is to use the landmark model, which is a system of prediction models that evolve with the follow-up time. We review the pros and cons of the two approaches, and present a general analytical framework using the landmark approach. The proposed framework allows the measurement times of longitudinal data to be irregularly spaced and differ between subjects. We propose a unified kernel weighting approach for estimating the model parameters, calculating predicted probabilities, and evaluating prediction accuracy through double time-dependent Receiver Operating Characteristics (ROC) curves. We illustrate the proposed analytical framework using the African American Study of Kidney Disease and Hypertension (AASK) to develop a landmark model for dynamic prediction of end stage renal diseases or death among patients with chronic kidney disease.
Collapse
Affiliation(s)
- Liang Li
- Department of Biostatistics, University of Texas MD Anderson Cancer Center, 1400 Pressler Street, Houston, TX 77030, USA.
| | - Sheng Luo
- Department of Biostatistics, University of Texas School of Public Health, Houston, TX, USA
| | - Bo Hu
- Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA
| | - Tom Greene
- Department of Population Health Sciences, University of Utah, Salt Lake City, UT, USA
| |
Collapse
|
38
|
Martínez-Camblor P, Pardo-Fernández JC. Smooth time-dependent receiver operating characteristic curve estimators. Stat Methods Med Res 2017; 27:651-674. [DOI: 10.1177/0962280217740786] [Citation(s) in RCA: 13] [Impact Index Per Article: 1.9] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/17/2022]
Abstract
The receiver operating characteristic curve is a popular graphical method often used to study the diagnostic capacity of continuous (bio)markers. When the considered outcome is a time-dependent variable, two main extensions have been proposed: the cumulative/dynamic receiver operating characteristic curve and the incident/dynamic receiver operating characteristic curve. In both cases, the main problem for developing appropriate estimators is the estimation of the joint distribution of the variables time-to-event and marker. As usual, different approximations lead to different estimators. In this article, the authors explore the use of a bivariate kernel density estimator which accounts for censored observations in the sample and produces smooth estimators of the time-dependent receiver operating characteristic curves. The performance of the resulting cumulative/dynamic and incident/dynamic receiver operating characteristic curves is studied by means of Monte Carlo simulations. Additionally, the influence of the choice of the required smoothing parameters is explored. Finally, two real-applications are considered. An R package is also provided as a complement to this article.
Collapse
Affiliation(s)
- Pablo Martínez-Camblor
- The Dartmouth Institute of Health Police and Clinical Practice, Geisel School of Medicine, Dartmouth College, Hanover, NH, USA
- Universidad Autónoma de Chile, Santiago, Chile
| | - Juan Carlos Pardo-Fernández
- Department of Statistics and Operational Research and Biomedical Research Centre CINBIO, Universidade de Vigo, Vigo, Spain
| |
Collapse
|
39
|
A sudden death risk score specifically for hypertension: based on 25 648 individual patient data from six randomized controlled trials. J Hypertens 2017. [PMID: 28650919 DOI: 10.1097/hjh.0000000000001451] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/26/2022]
Abstract
OBJECTIVE To construct a sudden death risk score specifically for hypertension (HYSUD) patients with or without cardiovascular history. METHODS Data were collected from six randomized controlled trials of antihypertensive treatments with 8044 women and 17 604 men differing in age ranges and blood pressure eligibility criteria. In total, 345 sudden deaths (1.35%) occurred during a mean follow-up of 5.16 years. Risk factors of sudden death were examined using a multivariable Cox proportional hazards model adjusted on trials. The model was transformed to an integer system, with points added for each factor according to its association with sudden death risk. RESULTS Antihypertensive treatment was not associated with a reduction of the sudden death risk and had no interaction with other factors, allowing model development on both treatment and placebo groups. A risk score of sudden death in 5 years was built with seven significant risk factors: age, sex, SBP, serum total cholesterol, cigarette smoking, diabetes, and history of myocardial infarction. In terms of discrimination performance, HYSUD model was adequate with areas under the receiver operating characteristic curve of 77.74% (confidence interval 95%, 74.13-81.35) for the derivation set, of 77.46% (74.09-80.83) for the validation set, and of 79.17% (75.94-82.40) for the whole population. CONCLUSION Our work provides a simple risk-scoring system for sudden death prediction in hypertension, using individual data from six randomized controlled trials of antihypertensive treatments. HYSUD score could help assessing a hypertensive individual's risk of sudden death and optimizing preventive therapeutic strategies for these patients.
Collapse
|
40
|
Le Borgne F, Combescure C, Gillaizeau F, Giral M, Chapal M, Giraudeau B, Foucher Y. Standardized and weighted time-dependent receiver operating characteristic curves to evaluate the intrinsic prognostic capacities of a marker by taking into account confounding factors. Stat Methods Med Res 2017. [PMID: 28633603 DOI: 10.1177/0962280217702416] [Citation(s) in RCA: 6] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
Time-dependent receiver operating characteristic curves allow to evaluate the capacity of a marker to discriminate between subjects who experience the event up to a given prognostic time from those who are free of this event. In this article, we propose an inverse probability weighting estimator of a standardized and weighted time-dependent receiver operating characteristic curve. This estimator provides a measure of the prognostic capacities by taking into account potential confounding factors. We illustrate the robustness of the estimator by a simulation-based study and its usefulness by two applications in kidney transplantation.
Collapse
Affiliation(s)
- Florent Le Borgne
- 1 Université de Nantes, Université de Tours, INSERM, SPHERE U1246, Nantes, France.,2 IDBC/A2com, Pace, France.,3 ITUN, INSERM U1064, Nantes, France
| | - Christophe Combescure
- 4 CRC and Division of Clinical Epidemiology, Department of Health and Community Medicine, University of Geneva, University Hospitals of Geneva, Geneva, Switzerland
| | - Florence Gillaizeau
- 1 Université de Nantes, Université de Tours, INSERM, SPHERE U1246, Nantes, France.,5 Department of Statistical Science, University College London, London, UK
| | - Magali Giral
- 1 Université de Nantes, Université de Tours, INSERM, SPHERE U1246, Nantes, France.,3 ITUN, INSERM U1064, Nantes, France
| | - Marion Chapal
- 1 Université de Nantes, Université de Tours, INSERM, SPHERE U1246, Nantes, France.,6 Médecine Néphrologie - Hémodialyse, Centre Hospitalier Départemental Vendée Site de La Roche sur Yon, La Roche-sur-Yon, France
| | - Bruno Giraudeau
- 7 Centre d'Investigation clinique (CIC), INSERM 1415, Tours, France.,8 Université de Tours, Université de Nantes, INSERM, SPHERE U1246, Tours, France.,9 CHRU de Tours, Tours, France
| | - Yohann Foucher
- 1 Université de Nantes, Université de Tours, INSERM, SPHERE U1246, Nantes, France.,10 Nantes University Hospital, Nantes, France
| |
Collapse
|
41
|
Kamarudin AN, Cox T, Kolamunnage-Dona R. Time-dependent ROC curve analysis in medical research: current methods and applications. BMC Med Res Methodol 2017; 17:53. [PMID: 28388943 PMCID: PMC5384160 DOI: 10.1186/s12874-017-0332-6] [Citation(s) in RCA: 435] [Impact Index Per Article: 62.1] [Reference Citation Analysis] [Abstract] [Key Words] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/27/2016] [Accepted: 03/28/2017] [Indexed: 01/30/2023] Open
Abstract
BACKGROUND ROC (receiver operating characteristic) curve analysis is well established for assessing how well a marker is capable of discriminating between individuals who experience disease onset and individuals who do not. The classical (standard) approach of ROC curve analysis considers event (disease) status and marker value for an individual as fixed over time, however in practice, both the disease status and marker value change over time. Individuals who are disease-free earlier may develop the disease later due to longer study follow-up, and also their marker value may change from baseline during follow-up. Thus, an ROC curve as a function of time is more appropriate. However, many researchers still use the standard ROC curve approach to determine the marker capability ignoring the time dependency of the disease status or the marker. METHODS We comprehensively review currently proposed methodologies of time-dependent ROC curves which use single or longitudinal marker measurements, aiming to provide clarity in each methodology, identify software tools to carry out such analysis in practice and illustrate several applications of the methodology. We have also extended some methods to incorporate a longitudinal marker and illustrated the methodologies using a sequential dataset from the Mayo Clinic trial in primary biliary cirrhosis (PBC) of the liver. RESULTS From our methodological review, we have identified 18 estimation methods of time-dependent ROC curve analyses for censored event times and three other methods can only deal with non-censored event times. Despite the considerable numbers of estimation methods, applications of the methodology in clinical studies are still lacking. CONCLUSIONS The value of time-dependent ROC curve methods has been re-established. We have illustrated the methods in practice using currently available software and made some recommendations for future research.
Collapse
Affiliation(s)
| | - Trevor Cox
- Department of Biostatistics, University of Liverpool, Liverpool, L69 3GL, UK
| | | |
Collapse
|
42
|
Li L, Greene T, Hu B. A simple method to estimate the time-dependent receiver operating characteristic curve and the area under the curve with right censored data. Stat Methods Med Res 2016; 27:2264-2278. [DOI: 10.1177/0962280216680239] [Citation(s) in RCA: 37] [Impact Index Per Article: 4.6] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/03/2023]
Abstract
The time-dependent receiver operating characteristic curve is often used to study the diagnostic accuracy of a single continuous biomarker, measured at baseline, on the onset of a disease condition when the disease onset may occur at different times during the follow-up and hence may be right censored. Due to right censoring, the true disease onset status prior to the pre-specified time horizon may be unknown for some patients, which causes difficulty in calculating the time-dependent sensitivity and specificity. We propose to estimate the time-dependent sensitivity and specificity by weighting the censored data by the conditional probability of disease onset prior to the time horizon given the biomarker, the observed time to event, and the censoring indicator, with the weights calculated nonparametrically through a kernel regression on time to event. With this nonparametric weighting adjustment, we derive a novel, closed-form formula to calculate the area under the time-dependent receiver operating characteristic curve. We demonstrate through numerical study and theoretical arguments that the proposed method is insensitive to misspecification of the kernel bandwidth, produces unbiased and efficient estimators of time-dependent sensitivity and specificity, the area under the curve, and other estimands from the receiver operating characteristic curve, and outperforms several other published methods currently implemented in R packages.
Collapse
Affiliation(s)
- Liang Li
- Department of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
| | - Tom Greene
- Department of Population Health Sciences, University of Utah, Salt Lake City, UT, USA
| | - Bo Hu
- Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA
| |
Collapse
|
43
|
Valero C, Pardo L, López M, García J, Camacho M, Quer M, León X. Pretreatment count of peripheral neutrophils, monocytes, and lymphocytes as independent prognostic factor in patients with head and neck cancer. Head Neck 2016; 39:219-226. [PMID: 27534525 DOI: 10.1002/hed.24561] [Citation(s) in RCA: 64] [Impact Index Per Article: 8.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/01/2016] [Revised: 05/12/2016] [Accepted: 07/05/2016] [Indexed: 12/19/2022] Open
Abstract
BACKGROUND The purpose of this study was to analyze the prognostic value of pretreatment count of peripheral neutrophils, lymphocytes, monocytes, and neutrophil-to-lymphocyte ratio (NLR) in patients with head and neck squamous cell carcinoma (HNSCC). METHODS Local, regional, and distant recurrence-free survival and disease-specific survival were analyzed according to the count of neutrophils, lymphocytes, monocytes, and NLR. RESULTS We observed a decrease in disease-specific survival as the quartile category of neutrophils, monocytes, and NLR increased. In the case of lymphocytes, patients in the lower quartile had lower disease-specific survival. Considering the disease-specific survival as the dependent variable, a recursive partitioning analysis classified the patients according to the neutrophil and monocyte counts. CONCLUSION High pretreatment count of peripheral neutrophils and/or monocytes was independently related with worse prognosis in patients with HNSCC. Classification based on pretreatment neutrophil and monocyte counts enabled the identification of different prognostic profiles. © 2016 Wiley Periodicals, Inc. Head Neck 39: 219-226, 2017.
Collapse
Affiliation(s)
- Cristina Valero
- Otorhinolaryngology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain
| | - Laura Pardo
- Otorhinolaryngology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain
| | - Montserrat López
- Otorhinolaryngology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain
| | - Jacinto García
- Otorhinolaryngology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain
| | - Mercedes Camacho
- Laboratory of Angiology, Vascular Biology and Inflammation, Institute of Biomedical Research (IIB Sant Pau), Barcelona, Spain
| | - Miquel Quer
- Otorhinolaryngology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain
| | - Xavier León
- Otorhinolaryngology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain.,Centro de Investigación Biomédica en Red de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Madrid, Spain
| |
Collapse
|
44
|
Henry KE, Hager DN, Pronovost PJ, Saria S. A targeted real-time early warning score (TREWScore) for septic shock. Sci Transl Med 2016; 7:299ra122. [PMID: 26246167 DOI: 10.1126/scitranslmed.aab3719] [Citation(s) in RCA: 282] [Impact Index Per Article: 35.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/25/2023]
Abstract
Sepsis is a leading cause of death in the United States, with mortality highest among patients who develop septic shock. Early aggressive treatment decreases morbidity and mortality. Although automated screening tools can detect patients currently experiencing severe sepsis and septic shock, none predict those at greatest risk of developing shock. We analyzed routinely available physiological and laboratory data from intensive care unit patients and developed "TREWScore," a targeted real-time early warning score that predicts which patients will develop septic shock. TREWScore identified patients before the onset of septic shock with an area under the ROC (receiver operating characteristic) curve (AUC) of 0.83 [95% confidence interval (CI), 0.81 to 0.85]. At a specificity of 0.67, TREWScore achieved a sensitivity of 0.85 and identified patients a median of 28.2 [interquartile range (IQR), 10.6 to 94.2] hours before onset. Of those identified, two-thirds were identified before any sepsis-related organ dysfunction. In comparison, the Modified Early Warning Score, which has been used clinically for septic shock prediction, achieved a lower AUC of 0.73 (95% CI, 0.71 to 0.76). A routine screening protocol based on the presence of two of the systemic inflammatory response syndrome criteria, suspicion of infection, and either hypotension or hyperlactatemia achieved a lower sensitivity of 0.74 at a comparable specificity of 0.64. Continuous sampling of data from the electronic health records and calculation of TREWScore may allow clinicians to identify patients at risk for septic shock and provide earlier interventions that would prevent or mitigate the associated morbidity and mortality.
Collapse
Affiliation(s)
- Katharine E Henry
- Department of Computer Science, Johns Hopkins University, Baltimore, MD 21218, USA
| | - David N Hager
- Division of Pulmonary and Critical Care Medicine, Department of Medicine, School of Medicine, Johns Hopkins University, Baltimore, MD 21205, USA
| | - Peter J Pronovost
- Armstrong Institute for Patient Safety and Quality, Johns Hopkins University, Baltimore, MD 21202, USA. Department of Anesthesiology and Critical Care Medicine, School of Medicine, Johns Hopkins University, Baltimore, MD 21202, USA. Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21205, USA
| | - Suchi Saria
- Department of Computer Science, Johns Hopkins University, Baltimore, MD 21218, USA. Armstrong Institute for Patient Safety and Quality, Johns Hopkins University, Baltimore, MD 21202, USA. Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21205, USA. Department of Applied Math and Statistics, Johns Hopkins University, Baltimore, MD 21218, USA.
| |
Collapse
|
45
|
Affiliation(s)
- Pablo Martínez-Camblor
- Hospital Universitario Central de Asturias (HUCA), Asturies, Spain
- Universidad Autonoma de Chile, Santiago, Chile
| | - Gustavo F. Bayón
- Instituto Universitario de Oncología del Principado de Asturias (IUOPA), Asturies, Spain
| | | |
Collapse
|
46
|
A New Prognostic Score Supporting Treatment Allocation for Multimodality Therapy for Malignant Pleural Mesothelioma. J Thorac Oncol 2015; 10:1634-41. [DOI: 10.1097/jto.0000000000000661] [Citation(s) in RCA: 47] [Impact Index Per Article: 5.2] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/02/2023]
|
47
|
Rodríguez-Álvarez MX, Meira-Machado L, Abu-Assi E, Raposeiras-Roubín S. Nonparametric estimation of time-dependent ROC curves conditional on a continuous covariate. Stat Med 2015; 35:1090-102. [PMID: 26487068 DOI: 10.1002/sim.6769] [Citation(s) in RCA: 22] [Impact Index Per Article: 2.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/15/2014] [Accepted: 09/29/2015] [Indexed: 12/31/2022]
Abstract
The receiver-operating characteristic (ROC) curve is the most widely used measure for evaluating the performance of a diagnostic biomarker when predicting a binary disease outcome. The ROC curve displays the true positive rate (or sensitivity) and the false positive rate (or 1-specificity) for different cut-off values used to classify an individual as healthy or diseased. In time-to-event studies, however, the disease status (e.g. death or alive) of an individual is not a fixed characteristic, and it varies along the study. In such cases, when evaluating the performance of the biomarker, several issues should be taken into account: first, the time-dependent nature of the disease status; and second, the presence of incomplete data (e.g. censored data typically present in survival studies). Accordingly, to assess the discrimination power of continuous biomarkers for time-dependent disease outcomes, time-dependent extensions of true positive rate, false positive rate, and ROC curve have been recently proposed. In this work, we present new nonparametric estimators of the cumulative/dynamic time-dependent ROC curve that allow accounting for the possible modifying effect of current or past covariate measures on the discriminatory power of the biomarker. The proposed estimators can accommodate right-censored data, as well as covariate-dependent censoring. The behavior of the estimators proposed in this study will be explored through simulations and illustrated using data from a cohort of patients who suffered from acute coronary syndrome.
Collapse
Affiliation(s)
- María Xosé Rodríguez-Álvarez
- Department of Statistics and Operations Research, and Biomedical Research Centre (CINBIO), University of Vigo, Campus Lagoas-Marcosende s/n, Vigo, 36310, Spain
| | - Luís Meira-Machado
- Centre of Mathematics and Department of Mathematics and Applications, University of Minho, Campus de Azurém, Guimarães, 4800-058, Portugal
| | - Emad Abu-Assi
- Department of Cardiology, University Clinical Hospital of Santiago de Compostela, Spain
| | | |
Collapse
|
48
|
Comparison of Pathologic Response Evaluation Systems after Anthracycline with/without Taxane-Based Neoadjuvant Chemotherapy among Different Subtypes of Breast Cancers. PLoS One 2015; 10:e0137885. [PMID: 26394326 PMCID: PMC4578929 DOI: 10.1371/journal.pone.0137885] [Citation(s) in RCA: 19] [Impact Index Per Article: 2.1] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/30/2015] [Accepted: 08/24/2015] [Indexed: 11/19/2022] Open
Abstract
Purpose Several methods are used to assess the pathologic response of breast cancer after neoadjuvant chemotherapy (NAC) to predict clinical outcome. However, the clinical utility of these systems for each molecular subtype of breast cancer is unclear. Therefore, we applied six pathologic response assessment systems to specific subtypes of breast cancer and compared the results. Patients and Methods Five hundred and eighty eight breast cancer patients treated with anthracycline with/without taxane-based NAC were retrospectively analyzed, and the ypTNM stage, residual cancer burden (RCB), residual disease in breast and nodes (RDBN), tumor response ratio, Sataloff’s classification, and Miller—Payne grading system were evaluated. The results obtained for each assessment system were analyzed in terms of patient survival. Results In triple-negative tumors, all systems were significantly associated with disease-free survival and Kaplan-Meier survival curves for disease-free survival were clearly separated by all assessment methods. For HR+/HER2- tumors, systems assessing the residual tumor (ypTNM stage, RCB, and RDBN) had prognostic significance. However, for HER2+ tumors, the association between patient survival and the pathologic response assessment results varied according to the system used, and none resulted in distinct Kaplan—Meier curves. Conclusion Most of the currently available pathologic assessment systems used after anthracycline with/without taxane-based NAC effectively classified triple-negative breast cancers into groups showing different prognoses. The pathologic assessment systems evaluating residual tumors only also had prognostic significance in HR+/HER2- tumors. However, new assessment methods are required to effectively evaluate the pathologic response of HR+/HER2+ and HR-/HER2+ tumors to anthracycline with/without taxane-based NAC.
Collapse
|
49
|
Jacqmin-Gadda H, Blanche P, Chary E, Touraine C, Dartigues JF. Receiver operating characteristic curve estimation for time to event with semicompeting risks and interval censoring. Stat Methods Med Res 2014; 25:2750-2766. [PMID: 24803510 DOI: 10.1177/0962280214531691] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/15/2022]
Abstract
Semicompeting risks and interval censoring are frequent in medical studies, for instance when a disease may be diagnosed only at times of visit and disease onset is in competition with death. To evaluate the ability of markers to predict disease onset in this context, estimators of discrimination measures must account for these two issues. In recent years, methods for estimating the time-dependent receiver operating characteristic curve and the associated area under the ROC curve have been extended to account for right censored data and competing risks. In this paper, we show how an approximation allows to use the inverse probability of censoring weighting estimator for semicompeting events with interval censored data. Then, using an illness-death model, we propose two model-based estimators allowing to rigorously handle these issues. The first estimator is fully model based whereas the second one only uses the model to impute missing observations due to censoring. A simulation study shows that the bias for inverse probability of censoring weighting remains modest and may be less than the one of the two parametric estimators when the model is misspecified. We finally recommend the nonparametric inverse probability of censoring weighting estimator as main analysis and the imputation estimator based on the illness-death model as sensitivity analysis.
Collapse
Affiliation(s)
- Hélène Jacqmin-Gadda
- Université Bordeaux Segalen, ISPED, Centre INSERM U897, Bordeaux, France
- INSERM, Centre INSERM U-897, F-33000 Bordeaux, France
| | - Paul Blanche
- Université Bordeaux Segalen, ISPED, Centre INSERM U897, Bordeaux, France
- INSERM, Centre INSERM U-897, F-33000 Bordeaux, France
| | - Emilie Chary
- Université Bordeaux Segalen, ISPED, Centre INSERM U897, Bordeaux, France
- INSERM, Centre INSERM U-897, F-33000 Bordeaux, France
| | - Célia Touraine
- Université Bordeaux Segalen, ISPED, Centre INSERM U897, Bordeaux, France
- INSERM, Centre INSERM U-897, F-33000 Bordeaux, France
| | - Jean-François Dartigues
- Université Bordeaux Segalen, ISPED, Centre INSERM U897, Bordeaux, France
- INSERM, Centre INSERM U-897, F-33000 Bordeaux, France
| |
Collapse
|
50
|
Lorent M, Giral M, Foucher Y. Net time-dependent ROC curves: a solution for evaluating the accuracy of a marker to predict disease-related mortality. Stat Med 2014; 33:2379-89. [DOI: 10.1002/sim.6079] [Citation(s) in RCA: 39] [Impact Index Per Article: 3.9] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/02/2013] [Revised: 10/15/2013] [Accepted: 12/05/2013] [Indexed: 01/27/2023]
Affiliation(s)
- Marine Lorent
- SPHERE EA 4275 Biostatistics, Clinical Research and Subjective Measurements in Health Sciences; University of Nantes; 1 rue Gaston Veil 44035 Nantes France
| | - Magali Giral
- Transplantation, Urology and Nephrology Institute (ITUN); Nantes Hospital and University; Inserm U1064, 30 Bd. Jean Monnet 44093 Nantes France
| | - Yohann Foucher
- SPHERE EA 4275 Biostatistics, Clinical Research and Subjective Measurements in Health Sciences; University of Nantes; 1 rue Gaston Veil 44035 Nantes France
| |
Collapse
|