1
|
Zou Y, Yue M, Jia L, Wang Y, Chen H, Zhang A, Xia X, Liu W, Yu R, Yang S, Huang P. Accurate prediction of HCC risk after SVR in patients with hepatitis C cirrhosis based on longitudinal data. BMC Cancer 2023; 23:1147. [PMID: 38007418 PMCID: PMC10676612 DOI: 10.1186/s12885-023-11628-1] [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: 05/28/2023] [Accepted: 11/09/2023] [Indexed: 11/27/2023] Open
Abstract
BACKGROUND Most existing predictive models of hepatocellular carcinoma (HCC) risk after sustained virologic response (SVR) are built on data collected at baseline and therefore have limited accuracy. The current study aimed to construct an accurate predictive model incorporating longitudinal data using a novel modeling strategy. The predictive performance of the longitudinal model was also compared with a baseline model. METHODS A total of 400 patients with HCV-related cirrhosis who achieved SVR with direct-acting antivirals (DAA) were enrolled in the study. Patients were randomly divided into a training set (70%) and a validation set (30%). Informative features were extracted from the longitudinal variables and then put into the random survival forest (RSF) to develop the longitudinal model. A baseline model including the same variables was built for comparison. RESULTS During a median follow-up time of approximately 5 years, 25 patients (8.9%) in the training set and 11 patients (9.2%) in the validation set developed HCC. The areas under the receiver-operating characteristics curves (AUROC) for the longitudinal model were 0.9507 (0.8838-0.9997), 0.8767 (0.6972,0.9918), and 0.8307 (0.6941,0.9993) for 1-, 2- and 3-year risk prediction, respectively. The brier scores of the longitudinal model were also relatively low for the 1-, 2- and 3-year risk prediction (0.0283, 0.0561, and 0.0501, respectively). In contrast, the baseline model only achieved mediocre AUROCs of around 0.6 (0.6113, 0.6213, and 0.6480, respectively). CONCLUSIONS Our longitudinal model yielded accurate predictions of HCC risk in patients with HCV-relate cirrhosis, outperforming the baseline model. Our model can provide patients with valuable prognosis information and guide the intensity of surveillance in clinical practice.
Collapse
Affiliation(s)
- Yanzheng Zou
- Department of Epidemiology, Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, 211166, China
| | - Ming Yue
- Department of Infectious Diseases, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
| | - Linna Jia
- Department of Epidemiology, Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, 211166, China
| | - Yifan Wang
- Department of Infectious Disease, Jurong Hospital Affiliated to Jiangsu University, Jurong, China
| | - Hongbo Chen
- Department of Infectious Disease, Jurong Hospital Affiliated to Jiangsu University, Jurong, China
| | - Amei Zhang
- Faculty of Life Science and Technology, Kunming University of Science and Technology, Yunnan, China
| | - Xueshan Xia
- Faculty of Life Science and Technology, Kunming University of Science and Technology, Yunnan, China
- Kunming Medical University, Kunming, China
| | - Wei Liu
- Department of Epidemiology, Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, 211166, China
- Beijing Institute of Microbiology and Epidemiology, State Key Laboratory of Pathogen and Biosecurity, Beijing, China
| | - Rongbin Yu
- Department of Epidemiology, Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, 211166, China.
| | - Sheng Yang
- Department of Biostatistics, Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, 211166, China.
| | - Peng Huang
- Department of Epidemiology, Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, 211166, China.
| |
Collapse
|
2
|
Kim HN, Nance RM, Lo Re V, Silverberg MJ, Franco R, Sterling TR, Cachay ER, Horberg MA, Althoff KN, Justice AC, Moore RD, Klein M, Crane HM, Delaney JA, Kitahata MM. Development and Validation of a Model for Prediction of End-Stage Liver Disease in People With HIV. J Acquir Immune Defic Syndr 2022; 89:396-404. [PMID: 35202048 PMCID: PMC8887786 DOI: 10.1097/qai.0000000000002886] [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: 09/30/2021] [Accepted: 12/06/2021] [Indexed: 11/26/2022]
Abstract
BACKGROUND End-stage liver disease (ESLD) is a leading cause of non-AIDS-related death among people with HIV (PWH). Factors that increase the progression of liver disease include comorbidities and HIV-specific factors, but we currently lack a tool to apply this evidence into clinical practice. METHODS We developed and validated a risk prediction model for ESLD among PWH who received care in 12 cohorts of the North American AIDS Cohort Collaboration on Research and Design between 2000 and 2016 and had fibrosis-4 index > 1.45. The first occurrence of ascites, variceal bleed, spontaneous bacterial peritonitis, or hepatic encephalopathy was verified by standardized medical record review. The Bayesian model averaging was used to select predictors among biomarkers and diagnoses and the Harrell C statistic to assess model discrimination. RESULTS Among 13,787 PWH in the training set, 82% were men and 54% were Black with a mean age of 48 years. Three hundred ninety ESLD events occurred over a mean 5.4 years. Among the ESLD cases, 52% had hepatitis C virus, 15% hepatitis B virus, and 31% alcohol use disorder. Twelve factors together predicted ESLD risk moderately well (C statistic 0.79, 95% confidence interval: 0.76 to 0.81): age, sex, race/ethnicity, chronic hepatitis B or C, and routinely collected laboratory values reflecting hepatic impairment (serum albumin, aspartate aminotransferase, total bilirubin, and platelets) and lipid metabolism (triglycerides, high-density lipoprotein, and total cholesterol). Our model performed well in the test set (C statistic 0.81, 95% confidence interval: 0.76 to 0.86). CONCLUSION This model of readily accessible clinical parameters predicted ESLD in a large diverse population of PWH.
Collapse
Affiliation(s)
| | | | | | | | | | | | | | | | | | - Amy C. Justice
- Yale University Schools of Medicine and Public Health, New Haven, CT, USA and Veterans Administration Connecticut Healthcare System, USA
| | | | - Marina Klein
- McGill University Health Centre, Montreal, Quebec, Canada
| | | | - Joseph A. Delaney
- University of Washington, Seattle, WA, USA
- University of Manitoba, Winnipeg, Manitoba, Canada
| | | |
Collapse
|
3
|
Bull LM, Lunt M, Martin GP, Hyrich K, Sergeant JC. Harnessing repeated measurements of predictor variables for clinical risk prediction: a review of existing methods. Diagn Progn Res 2020; 4:9. [PMID: 32671229 PMCID: PMC7346415 DOI: 10.1186/s41512-020-00078-z] [Citation(s) in RCA: 6] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/06/2020] [Accepted: 04/28/2020] [Indexed: 12/19/2022] Open
Abstract
BACKGROUND Clinical prediction models (CPMs) predict the risk of health outcomes for individual patients. The majority of existing CPMs only harness cross-sectional patient information. Incorporating repeated measurements, such as those stored in electronic health records, into CPMs may provide an opportunity to enhance their performance. However, the number and complexity of methodological approaches available could make it difficult for researchers to explore this opportunity. Our objective was to review the literature and summarise existing approaches for harnessing repeated measurements of predictor variables in CPMs, primarily to make this field more accessible for applied researchers. METHODS MEDLINE, Embase and Web of Science were searched for articles reporting the development of a multivariable CPM for individual-level prediction of future binary or time-to-event outcomes and modelling repeated measurements of at least one predictor. Information was extracted on the following: the methodology used, its specific aim, reported advantages and limitations, and software available to apply the method. RESULTS The search revealed 217 relevant articles. Seven methodological frameworks were identified: time-dependent covariate modelling, generalised estimating equations, landmark analysis, two-stage modelling, joint-modelling, trajectory classification and machine learning. Each of these frameworks satisfies at least one of three aims: to better represent the predictor-outcome relationship over time, to infer a covariate value at a pre-specified time and to account for the effect of covariate change. CONCLUSIONS The applicability of identified methods depends on the motivation for including longitudinal information and the method's compatibility with the clinical context and available patient data, for both model development and risk estimation in practice.
Collapse
Affiliation(s)
- Lucy M. Bull
- grid.5379.80000000121662407Centre for Epidemiology Versus Arthritis, Centre for Musculoskeletal Research, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK
- grid.5379.80000000121662407Centre for Biostatistics, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK
| | - Mark Lunt
- grid.5379.80000000121662407Centre for Epidemiology Versus Arthritis, Centre for Musculoskeletal Research, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK
| | - Glen P. Martin
- grid.5379.80000000121662407Division of Informatics, Imaging and Data Science, Faculty of Biology, Medicine and Health, University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
| | - Kimme Hyrich
- grid.5379.80000000121662407Centre for Epidemiology Versus Arthritis, Centre for Musculoskeletal Research, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK
- grid.498924.aNational Institute for Health Research Manchester Biomedical Research Centre, Manchester University NHS Foundation Trust, Manchester Academic Health Science Centre, Manchester, UK
| | - Jamie C. Sergeant
- grid.5379.80000000121662407Centre for Epidemiology Versus Arthritis, Centre for Musculoskeletal Research, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK
- grid.5379.80000000121662407Centre for Biostatistics, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK
| |
Collapse
|
4
|
Jonas MC, Loftus B, Horberg MA. The Road to Hepatitis C Virus Cure: Practical Considerations from a Health System's Perspective. Infect Dis Clin North Am 2019; 32:481-493. [PMID: 29778267 DOI: 10.1016/j.idc.2018.02.007] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/07/2023]
Abstract
Hepatitis C virus infection remains a significant global health problem. Many individuals are unaware of their infection or disease stage. Innovations in care that promote rapid and easy identification of at-risk populations for screening, comprehensive diagnostic screening, and triage to curative direct-acting antiviral medications will accelerate efforts to eradicate hepatitis C.
Collapse
Affiliation(s)
- M Cabell Jonas
- Mid-Atlantic Permanente Medical Group, PC, 2101 East Jefferson Street, Rockville, MD 20852, USA.
| | - Bernadette Loftus
- Mid-Atlantic Permanente Medical Group, PC, 2101 East Jefferson Street, Rockville, MD 20852, USA
| | - Michael A Horberg
- Mid-Atlantic Permanente Medical Group, PC, Mid-Atlantic Permanente Research Institute, 2101 East Jefferson Street, Rockville, MD 20852, USA
| |
Collapse
|
5
|
Konerman MA, Beste LA, Van T, Liu B, Zhang X, Zhu J, Saini SD, Su GL, Nallamothu BK, Ioannou GN, Waljee AK. Machine learning models to predict disease progression among veterans with hepatitis C virus. PLoS One 2019; 14:e0208141. [PMID: 30608929 PMCID: PMC6319806 DOI: 10.1371/journal.pone.0208141] [Citation(s) in RCA: 39] [Impact Index Per Article: 7.8] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/19/2018] [Accepted: 11/07/2018] [Indexed: 12/15/2022] Open
Abstract
Background Machine learning (ML) algorithms provide effective ways to build prediction models using longitudinal information given their capacity to incorporate numerous predictor variables without compromising the accuracy of the risk prediction. Clinical risk prediction models in chronic hepatitis C virus (CHC) can be challenging due to non-linear nature of disease progression. We developed and compared two ML algorithms to predict cirrhosis development in a large CHC-infected cohort using longitudinal data. Methods and findings We used national Veterans Health Administration (VHA) data to identify CHC patients in care between 2000–2016. The primary outcome was cirrhosis development ascertained by two consecutive aspartate aminotransferase (AST)-to-platelet ratio indexes (APRIs) > 2 after time zero given the infrequency of liver biopsy in clinical practice and that APRI is a validated non-invasive biomarker of fibrosis in CHC. We excluded those with initial APRI > 2 or pre-existing diagnosis of cirrhosis, hepatocellular carcinoma or hepatic decompensation. Enrollment was defined as the date of the first APRI. Time zero was defined as 2 years after enrollment. Cross-sectional (CS) models used predictors at or closest before time zero as a comparison. Longitudinal models used CS predictors plus longitudinal summary variables (maximum, minimum, maximum of slope, minimum of slope and total variation) between enrollment and time zero. Covariates included demographics, labs, and body mass index. Model performance was evaluated using concordance and area under the receiver operating curve (AuROC). A total of 72,683 individuals with CHC were analyzed with the cohort having a mean age of 52.8, 96.8% male and 53% white. There are 11,616 individuals (16%) who met the primary outcome over a mean follow-up of 7 years. We found superior predictive performance for the longitudinal Cox model compared to the CS Cox model (concordance 0.764 vs 0.746), and for the longitudinal boosted-survival-tree model compared to the linear Cox model (concordance 0.774 vs 0.764). The accuracy of the longitudinal models at 1,3,5 years after time zero also showed superior performance compared to the CS model, based on AuROC. Conclusions Boosted-survival-tree based models using longitudinal information are statistically superior to cross-sectional or linear models for predicting development of cirrhosis in CHC, though all four models were highly accurate. Similar statistical methods could be applied to predict outcomes in other non-linear chronic disease states.
Collapse
Affiliation(s)
- Monica A. Konerman
- Michigan Medicine, Department of Internal Medicine, Division of Gastroenterology and Hepatology, Ann Arbor, Michigan, United States of America
| | - Lauren A. Beste
- Department of Medicine, Veterans Affairs Puget Sound Healthcare System and University of Washington, Seattle, WA, United States of America
| | - Tony Van
- VA Ann Arbor Health Services Research and Development Center of Clinical Management Research, Ann Arbor, Michigan, United States of America
| | - Boang Liu
- Department of Statistics, University of Michigan, Ann Arbor, MI, United States of America
| | - Xuefei Zhang
- Department of Statistics, University of Michigan, Ann Arbor, MI, United States of America
| | - Ji Zhu
- Department of Statistics, University of Michigan, Ann Arbor, MI, United States of America
| | - Sameer D. Saini
- Michigan Medicine, Department of Internal Medicine, Division of Gastroenterology and Hepatology, Ann Arbor, Michigan, United States of America
- VA Ann Arbor Health Services Research and Development Center of Clinical Management Research, Ann Arbor, Michigan, United States of America
| | - Grace L. Su
- Michigan Medicine, Department of Internal Medicine, Division of Gastroenterology and Hepatology, Ann Arbor, Michigan, United States of America
- VA Ann Arbor Health Services Research and Development Center of Clinical Management Research, Ann Arbor, Michigan, United States of America
| | - Brahmajee K. Nallamothu
- Michigan Medicine, Department of Internal Medicine, Division of Cardiology, Ann Arbor, Michigan, United States of America
- Michigan Integrated Center for Health Analytics and Medical Prediction (MiCHAMP), Ann Arbor, MI, United States of America
| | - George N. Ioannou
- Division of Gastroenterology, Department of Medicine, University of Washington, Seattle, WA, United States of America
- Division of Gastroenterology, Department of Medicine, Veterans Affairs Puget Sound Healthcare System, Seattle, WA, United States of America
| | - Akbar K. Waljee
- Michigan Medicine, Department of Internal Medicine, Division of Gastroenterology and Hepatology, Ann Arbor, Michigan, United States of America
- VA Ann Arbor Health Services Research and Development Center of Clinical Management Research, Ann Arbor, Michigan, United States of America
- Michigan Integrated Center for Health Analytics and Medical Prediction (MiCHAMP), Ann Arbor, MI, United States of America
- * E-mail:
| |
Collapse
|
6
|
Assessing risk of fibrosis progression and liver-related clinical outcomes among patients with both early stage and advanced chronic hepatitis C. PLoS One 2017; 12:e0187344. [PMID: 29108017 PMCID: PMC5673203 DOI: 10.1371/journal.pone.0187344] [Citation(s) in RCA: 20] [Impact Index Per Article: 2.9] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/25/2017] [Accepted: 10/18/2017] [Indexed: 02/07/2023] Open
Abstract
OBJECTIVE Assessing risk of adverse outcomes among patients with chronic liver disease has been challenging due to non-linear disease progression. We previously developed accurate prediction models for fibrosis progression and clinical outcomes among patients with advanced chronic hepatitis C (CHC). The primary aim of this study was to validate fibrosis progression and clinical outcomes models among a heterogeneous patient cohort. DESIGN Adults with CHC with ≥3 years follow-up and without hepatic decompensation, hepatocellular carcinoma (HCC), liver transplant (LT), HBV or HIV co-infection at presentation were analyzed (N = 1007). Outcomes included: 1) fibrosis progression 2) hepatic decompensation 3) HCC and 4) LT-free survival. Predictors included longitudinal clinical and laboratory data. Machine learning methods were used to predict outcomes in 1 and 3 years. RESULTS The external cohort had a median age of 49.4 years (IQR 44.3-54.3); 61% were male, 80% white, and 79% had genotype 1. At presentation, 73% were treatment naïve and 31% had cirrhosis. Fibrosis progression occurred in 34% over a median of 4.9 years (IQR 3.2-7.6). Clinical outcomes occurred in 22% over a median of 4.4 years (IQR 3.2-7.6). Model performance for fibrosis progression was limited due to small sample size. The area under the receiver operating characteristic curve (AUROC) for 1 and 3-year risk of clinical outcomes was 0.78 (95% CI 0.73-0.83) and 0.76 (95% CI 0.69-0.81). CONCLUSION Accurate assessments for risk of clinical outcomes can be obtained using routinely collected data across a heterogeneous cohort of patients with CHC. These methods can be applied to predict risk of progression in other chronic liver diseases.
Collapse
|