Time-dependent Receiver Operating Characteristic Curve Analysis Research Articles

Predictive value of insulin resistance surrogates for the development of diabetes in individuals with baseline normoglycemia: findings from two independent cohort studies in China and Japan

BackgroundInsulin resistance (IR) plays a crucial role in the occurrence and progression of diabetes. This study aimed to evaluate and compare the predictive value of four IR surrogates, including the triglycerides glucose (TyG) index, TyG and body mass index (TyG-BMI), triglycerides/high-density lipoprotein cholesterol (TG/HDL-C) ratio, and the metabolic score for IR (MetS-IR) for diabetes in two large cohorts.MethodsA total of 116,661 adult participants from the China Rich Healthcare Group and 15,464 adult participants from the Japanese NAGALA cohort were included in the study. Multivariable Cox proportional hazards models were used to assess the standardized hazard ratio (HR) of the TyG index, TyG-BMI, TG/HDL-C ratio, and MetS-IR directly associated with diabetes. Receiver operating characteristic (ROC) curve and time-dependent ROC curve analysis were performed to evaluate and compare the predictive value of the four IR surrogates for diabetes.ResultsIn the two independent cohorts, the average follow-up time was 3.1 years in the China cohort, with 2681(2.30%) incident cases of diabetes recorded, and 6.13 years in the Japan cohort, with 373 incident cases (2.41%) of diabetes recorded. After adjusting for potential confounding factors, we found that among the four IR surrogates, TyG-BMI and MetS-IR showed stronger associations with diabetes. The stronger associations persisted even after further stratification by age, sex, hypertension, and obese subgroups. In terms of diabetes prediction, based on ROC analysis, TyG-BMI demonstrated the highest predictive accuracy for diabetes in the Chinese population, while both TyG-BMI and MetS-IR showed the highest predictive accuracy in the Japanese population. The results of further subgroup ROC analysis confirmed the robustness of these findings. Furthermore, the time-dependent ROC results indicated that among the four IR surrogates, MetS-IR exhibited the highest accuracy in predicting future diabetes at various time intervals in the Japanese population.ConclusionOur findings suggest that evaluating TyG-BMI and MetS-IR as IR surrogates may be the most useful for predicting diabetes events and assessing the risk of developing diabetes in East Asian populations.

Risk of non-colorectal cancer-related death in elderly patients with the disease: A comparison of five preoperative risk assessment indices.

A considerable number of elderly patients with colorectal cancer (CRC) die of non-CRC-related causes. The Controlling Nutritional Status (CONUT) score, American Society of Anesthesiologists Physical Status classification, Charlson Comorbidity Index, National Institute on Aging, and National Cancer Institute Comorbidity Index, and Adult Comorbidity Evaluation-27 score are all known predictors of survival in patients with CRC. However, the utility of these indices for predicting non-CRC-related death in elderly CRC patients is not known. The study population comprised 364 patients aged 80 years or more who received curative resection for stage I-III CRC between 2000 and 2016. The association of each index with non-CRC-related death was compared by competing-risks analysis such as the cumulative incidence function and proportional subdistribution hazards regression analysis as well as time-dependent receiver-operating characteristic (ROC) analysis. There were 85 deaths (40 CRC-related and 45 non-CRC-related) during a median observation period of 53.2months. Cumulative incidence function analysis identified CONUT score as the most suitable for risk stratification for non-CRC-related death. In proportional subdistribution hazards regression, risk of non-CRC-related death increased significantly as CONUT score worsened (2/3/4 vs. 0/1, hazard ratio 1.73, 95% confidence interval [CI] 0.91-3.15; ≥5 vs. 2/3/4, hazard ratio 2.71, 95% CI 1.08-6.81). Time-dependent ROC curve analysis showed that CONUT score were consistently superior to other indices during the 5-year observation period. The majority of deaths in elderly patients with CRC were not CRC-related. CONUT score was the most useful predictor of non-CRC-related death in these patients.

Prognostic Impact of Radiological Splenic Artery Involvement in Pancreatic Ductal Adenocarcinoma of the Body and Tail.

Splenic artery (SpA) involvement heralds poor prognosis in pancreatic ductal adenocarcinoma (PDAC) of the body and tail but is not included in the resectability criteria. This study evaluated the prognostic impact of radiological SpA involvement in PDAC of the body and tail. Preoperative computed tomography images of patients who underwent distal pancreatectomy for resectable PDAC of the body and tail (n = 242) at our hospital between 2004 and 2018 were graded according to splenic vessel involvement status as clear, abutment, or encasement. Clinicopathological prognostic factors and overall survival (OS) and recurrence-free survival (RFS) rates were compared between the three groups. The prognostic value of radiological involvement status was assessed using Harrell's concordance statistic (C-index) and time-dependent receiver-operating characteristic curve analysis and compared with pathological findings. The diagnostic concordance rate was 0.87 (weighted κ statistic). Prognosis worsened with progression from clear, abutment, to encasement status. SpA encasement (hazard ratio [HR] 1.97, p = 0.04) predicted poor OS in multivariate Cox hazard regression analysis. SpA abutment (HR 1.77, p = 0.017) and encasement (HR 1.86, p = 0.034) independently predicted poor RFS. Splenic vein abutment and encasement were not significant predictors of poor OS or RFS. SpA encasement without adjuvant chemotherapy had the poorest prognosis because of early distant metastasis. The prognostic value was higher for radiological SpA involvement than for pathological SpA invasion. Radiological SpA involvement status is a meaningful and reproducible prognostic indicator that can be used preoperatively for determining the treatment strategy in PDAC of the body and tail.

HPV E6/E7 mRNA in situ hybridization in endocervical adenocarcinoma: implications for prognosis and diagnosis

BackgroundHuman papillomavirus (HPV) E6/E7 mRNA in situ hybridization (HPV E6/E7 RNAscope) appears to be a sensitive and specific technique in detecting transcriptionally active HPV. We aimed to examine the diagnostic utility of this technique in endocervical adenocarcinoma (ECA), to explore the prognostic factors for ECA patients and develop a clinically useful nomogram based on clinicopathological parameters to predict it.MethodsWe retrospectively analyzed 200 patients with ECA who had undergone surgery at Sun Yat-sen University Cancer Center from 2010 and 2014. The diagnostic performance of HPV E6/E7 RNAscope were evaluated by receiver operating characteristic (ROC) curve. A prognostic nomogram model including HPV E6/E7 RNAscope was generated based on multivariate Cox regression analysis, then we compared the predictive accuracy of the prognostic model with FIGO staging and treatment using concordance index (C-index), time-dependent ROC (tdROC), and decision curve analysis (DCA).ResultsThe sensitivity and specificity of HPV E6/E7 RNAscope for distinguishing HPV-associated adenocarcinoma (HPVA) from non-HPV-associated adenocarcinoma (NHPVA) in the whole cohort were 75.8% and 80%, respectively. According the univariate analysis and multivariate logistic regression analysis, age, lymphovascular invasion (LVI), lymph node involvement (LNI), and HPV E6/E7 RNAscope were valuable predictive factors for OS. These parameters were incorporated into the nomogram model (nomogram A) compared with FIGO stage and treatment. The C-index of nomogram A for predicting OS was 0.825, which was significantly higher than FIGO stage (C-index = 0.653, p = 0.002) and treatment (C-index = 0.578, p < 0.001).Conclusions HPV E6/E7 RNAscope is highly specific for ECA, and the 4-variable nomogram showed more accurate prognostic outcomes in patients with ECA.

Identification of an EMT-Related Gene Signature for Predicting Overall Survival in Gastric Cancer.

BackgroundIt has been widely reported that epithelial-mesenchymal transition (EMT) is associated with malignant progression in gastric cancer (GC). Integration of the molecules related to EMT for predicting overall survival (OS) is meaningful for understanding the role of EMT in GC. Here, we aimed to establish an EMT-related gene signature in GC.MethodsTranscriptional profiles and clinical data of GC were downloaded from The Cancer Genome Atlas (TCGA). We constructed EMT-related gene signature for predicting OS by using univariate Cox regression and least absolute shrinkage and selection operator (LASSO) regression analyses. Time-dependent receiver operating characteristic (ROC), Kaplan-Meier analysis were performed to assess its predictive value. A nomogram combining the prognostic signature with clinical characteristics for OS prediction was established. And its predictive power was estimated by concordance index (C-index), time-dependent ROC curve, calibration curve and decision curve analysis (DCA). GSE62254 dataset from Gene Expression Omnibus (GEO) was used for external validation. Quantitative real-time PCR (qRT-PCR) was used to detected the mRNA expression of the five EMT-related genes in human normal gastric mucosal and GC cell lines. To further understand the potential mechanisms of the signature, Gene Set Enrichment Analysis (GSEA), pathway enrichment analysis, predictions of transcription factors (TFs)/miRNAs were performed.ResultsA novel EMT-related gene signature (including ITGAV, DAB2, SERPINE1, MATN3, PLOD2) was constructed for OS prediction of GC. With external validation, ROC curves indicated the signature’s good performance. Patients stratified into high- and low-risk groups based on the signature yielded significantly different prognosis. Univariate and multivariate Cox regression suggested that the signature was an independent prognostic variable. Nomogram for prognostication including the signature presented better predictive accuracy and clinical usefulness than the similar model without risk score to some extent with external validation. The qRT-PCR assays suggested that high expression of the five EMT-related genes could be found in human GC cell lines compared with normal gastric mucosal cell line. GSEA and pathway enrichment analysis revealed that focal adhesion and ECM-receptor interaction might be the two important pathways to the signature.ConclusionOur EMT-related gene signature may have practical application as an independent prognostic factor in GC.

Overexpression of GINS4 Is Associated With Tumor Progression and Poor Survival in Hepatocellular Carcinoma.

PurposeOur research was aimed to identify the expression, clinical value and biological significance of GINS complex subunit 4 (GINS4) in hepatocellular carcinoma (HCC).Materials and MethodsGINS4 was initially screened through weighted gene co-expression network analysis (WGCNA). The TCGA, GEO, and TIMER databases were applied for analyzing the GINS4 mRNA expression in HCC. GINS4 protein levels were detected via immunohistochemistry (IHC). Receiver operating characteristic (ROC) curve was applied for estimating the diagnostic significance of GINS4 in HCC. Kaplan-Meier plots, Cox model, and nomogram were used to assess the prognostic performance of GINS4 in HCC. Nomogram validation was conducted through time-dependent ROC and decision curve analysis (DCA). The Wanderer, UALCAN, and DiseaseMeth databases were utilized to identify GINS4 methylation levels in HCC. Genes co-expressed with GINS4 in HCC were estimated through the TCGA, cBioPortal, and GEPIA. GO, KEGG, and GSEA unraveled the possible biological mechanisms of GINS4 in HCC.ResultsWGCNA confirmed that GINS4 was one of hub genes significantly associated with histological grade of HCC. Multiple databases confirmed the significant upregulation of GINS4 in HCC tissues compared with non-tumor controls. IHC analysis of 35 HCC patients demonstrated that overexpressed GINS4 positively correlated with advanced TNM stage and poor pathological differentiation. GINS4 could effectively differentiate HCC cases from healthy individuals, with an AUC of 0.865. Increased GINS4 expression predicted unsatisfactory prognosis in HCC patients, especially in age >60 years, histological grade 1, HBV infection-negative, and occurring relapse subgroup. Nomogram incorporating GINS4 level and TNM stage displayed satisfactory predictive accuracy and clinical utility in predicting HCC prognosis. Upregulated GINS4 exhibited hypomethylated levels in HCC. Functional analysis indicated that GINS4 potentially positively modulated cell cycle and PI3K/AKT/mTOR pathway.ConclusionGINS4 is overexpressed in HCC and is correlated with undesirable survival of HCC patients.

Construction and Validation of a Prognostic Gene-Based Model for Overall Survival Prediction in Hepatocellular Carcinoma Using an Integrated Statistical and Bioinformatic Approach.

Hepatocellular carcinoma (HCC) is one of the most common lethal cancers worldwide and is often related to late diagnosis and poor survival outcome. More evidence is demonstrating that gene-based prognostic models can be used to predict high-risk HCC patients. Therefore, our study aimed to construct a novel prognostic model for predicting the prognosis of HCC patients. We used multivariate Cox regression model with three hybrid penalties approach including least absolute shrinkage and selection operator (Lasso), adaptive lasso and elastic net algorithms for informative prognostic-related genes selection. Then, the best subset regression was used to identify the best prognostic gene signature. The prognostic gene-based risk score was constructed using the Cox coefficient of the prognostic gene signature. The model was evaluated by Kaplan–Meier (KM) and receiver operating characteristic curve (ROC) analyses. A novel four-gene signature associated with prognosis was identified and the risk score was constructed based on the four-gene signature. The risk score efficiently distinguished the patients into a high-risk group with poor prognosis. The time-dependent ROC analysis revealed that the risk model had a good performance with an area under the curve (AUC) of 0.780, 0.732, 0.733 in 1-, 2- and 3-year prognosis prediction in The Cancer Genome Atlas (TCGA) dataset. Moreover, the risk score revealed a high diagnostic performance to classify HCC from normal samples. The prognosis and diagnosis prediction performances of risk scores were verified in external validation datasets. Functional enrichment analysis of the four-gene signature and its co-expressed genes involved in the metabolic and cell cycle pathways was constructed. Overall, we developed a novel-gene-based prognostic model to predict high-risk HCC patients and we hope that our findings can provide promising insight to explore the role of the four-gene signature in HCC patients and aid risk classification.

The association between platelet indices and cardiovascular events in chronic kidney disease patients without dialysis.

Previous studies have indicated that platelet indices are related to the pathogenesis of cardiovascular diseases (CVD). However, it is unclear which platelet-related indicators are associated with CVD events in patients with chronic kidney disease (CKD) without dialysis. We performed a single-center prospective cohort study involved 1391 CKD patients to explore the relationship between platelet indices and CVD events in CKD patients. A nomogram was generated to predict CVD-free survival after 3 and 5years of follow-up in terms of the fitted Cox regression model. And the time-dependent receiver-operating characteristic (ROC) curves were applied to evaluate the prediction accuracy of platelet indices on CVD events. During a median follow-up of 3.41years, 211 (15.2%) patients experienced CVD events. Results showed that platelet counts (PLT), plateletcrit (PCT), platelet-large cell ratio (P-LCR), and platelet distribution width (PDW) among 5 platelet indices were significantly lower in advanced CKD stages. Cox regression model showed that PLT, PDW, and PCT were associated with CVD events. However, after multivariable-adjusted, low level of PLT, hazard ratio (HR) 0.994 and 95% confidence interval (95% CI 0.989-1.000, p = 0.04), and PDW, HR 0.936 (95% CI 0.878-0.998, p = 0.044) predicted CVD events. The area under the ROC curve (AUC) of platelet indices assessed by time-dependent ROC curve analysis showed that only PLT and PDW were significant for predicting CVD events for 5years. We demonstrated that PLT and PDW among 5 platelet indices were independently associated with CVD events in patients with CKD.

Identification of a Transcriptional Prognostic Signature From Five Metabolic Pathways in Oral Squamous Cell Carcinoma.

Dysregulated metabolic pathways have been appreciated to be intimately associated with tumorigenesis and patient prognosis. Here, we sought to develop a novel prognostic signature based on metabolic pathways in patients with primary oral squamous cell carcinoma (OSCC). The original RNA-seq data of OSCC from The Cancer Genome Atlas (TCGA) project and Gene Expression Omnibus (GEO) database were transformed into a metabolic pathway enrichment score matrix by single-sample gene set enrichment analysis (ssGSEA). A novel prognostic signature based on metabolic pathways was constructed by LASSO and stepwise Cox regression analysis in the training cohort and validated in both testing and validation cohorts. The optimal cut-off value was obtained using the Youden index by receiver operating characteristic (ROC) curve. The overall survival curves were plotted by the Kaplan-Meier method. A time-dependent ROC curve analysis with 1, 3, 5 years as the defining point was performed to evaluate the predictive value of this prognostic signature. A 5-metabolic pathways prognostic signature (5MPS) for OSCC was constructed which stratified patients into subgroups with favorable or inferior survival. It served as an independent prognostic factor for patient survival and had a satisfactory predictive performance for OSCC. Our results developed a novel prognostic signature based on dysregulated metabolic pathways in OSCC and provided support for aberrant metabolism underlying OSCC tumorigenesis.

Combining total and differential somatic cell count to screen for mastitis

Submitted 2020-06-30 | Accepted 2020-07-23 | Available 2020-12-01 https://doi.org/10.15414/afz.2020.23.mi-fpap.88-96 Somatic cell count (SCC) has been extensively used as indicator of udder health and milk quality. Recent developments in milk-testing technology have led to cell differentiation in milk in a high throughput manner. Information on the proportion of the different cell types in milk would represent a valuable asset for a more precise definition of udder health status. The aim of the present study was to apply receiver-operating characteristic curve analysis to define the most accurate thresholds of milk differential somatic cell count (DSCC), which represents the percentage of neutrophils plus lymphocytes in the total SCC. The dataset accounted for 117,482 test-day records of 60,009 Holstein Friesian, Brown Swiss and Simmental cows. Different thresholds were defined so that DSCC trends were analysed throughout the lactation, considering also the classification factors of breed and parity. Finally, cows were classified as healthy, susceptible, mastitic or chronic on the basis of their health status, which was defined combining the information of SCC (below or above 200,000 cells/mL) and DSCC (below or above the specific cut-off). Our findings offered new insights for a practical use of DSCC to screen for mastitis, in order to help farmers make decisions to reduce the use of antimicrobials in the herd. Keywords: differential somatic cell count; receiver-operating characteristic (ROC) curve; cut-off; mastitis; cattle References Adkins, P.R.F. and Middleton, J.R. (2018) Methods for diagnosing mastitis. Veterinary Clinics of North America: Food Animal Practice, 34, 479–491. https://doi.org/10.1016/j.cvfa.2018.07.003 Bobbo, T. et al. (2016) The nonlinear effect of somatic cell count on milk composition, coagulation properties, curd firmness modeling, cheese yield, and curd nutrient recovery. Journal of Dairy Science, 99, 5104-5119. https://doi.org/10.3168/jds.2015-10512 Bobbo, T. et al. (2019) Short communication: Genetic aspects of milk differential somatic cell count in Holstein cows: A preliminary analysis. Journal of Dairy Science, 102, 4275–4279. https://doi.org/10.3168/jds.2018-16092 Bobbo, T. et al. (2018) Alternative somatic cell count traits exploitable in genetic selection for mastitis resistance in Italian Holsteins. Journal of Dairy Science, 101, 10001–10010. https://doi.org/10.3168/jds.2018-14827 Cecchinato, A. et al. (2018) Genetic variation in serum protein pattern and blood β-hydroxybutyrate and their relationships with udder health traits, protein profile, and cheese-making properties in Holstein cows. Journal of Dairy Science, 101, 11108-11119. https://doi.org/10.3168/jds.2018-14907 Cassandro, M. et al. (2008) Genetic parameters of milk coagulation properties and their relationships with milk yield and quality traits in Italian Holstein cows. Journal of Dairy Science, 91, 371-376. https://10.3168/jds.2007-0308. Damm, M. et al. (2017) Differential somatic cell count - A novel method for routine mastitis screening in the frame of Dairy Herd Improvement testing programs. Journal of Dairy Science, 100, 4926–4940. https://doi.org/10.3168/jds.2016-12409 Dohoo, I.R. and Leslie, K.E. (1991) Evaluation of changes in somatic cell counts as indicators of new intra-mammary infections. Journal of Preventive Veterinary Medicine, 10, 225-237. https://doi.org/10.1016/0167-5877(91)90006-N. Kamarudin, A.N. et al. (2017) Time-dependent ROC curve analysis in medical research: current methods and applications. BMC Medical Research and Methodology, 17, 53. https://doi.org/10.1186/s12874-017-0332-6 Kehrli, M.E. and Shuster, D.E. (1994) Factors affecting milk somatic cells and their role in health of the bovine mammary gland. Journal of Dairy Science, 77, 619–627. https://doi.org/10.3168/jds.S0022-0302(94)76992-7 Kirkeby, C. et al. (2019) Differential somatic cell count as an additional indicator for intramammary infections in dairy cows. Journal of Dairy Science, 103, 1759-1775. https://doi.org/10.3168/jds.2019-16523 Lee, C.S. et al. (1980) Identification properties and differential counts of cell populations using electron microscopy of dry cow secretions, colostrum and milk from normal cows. Journal of Dairy Research, 47, 39–50. https://doi.org/10.1017/S0022029900020860 Leitner, G. et al. (2008) Milk leucocyte population patterns in bovine udder infection of different aetiology. Journal of Veterinary Medicine B, 47, 581–589. https://doi.org/10.1046/j.1439-0450.2000.00388.x Liu, H. and Wu, T. (2003) Estimating the area under a receiver operating characteristic (ROC) curve for repeated measures design. Journal of Statistical Software, 8, 1–18. https://doi.org/10.18637/jss.v008.i12. Lopez-Raton, M. et al. (2014) OptimalCutpoints: An R package for selecting optimal cutpoints in diagnostic tests. Journal of Statistical Software, 61, 1-36. https://doi.org/10.18637/jss.v061.i08 Michael, H. et al. (2019) The ROC curve for regularly measured longitudinal biomarkers. Biostatistics, 20, 433-451. https://doi.org/10.1093/biostatistics/kxy010 Pilla, R. et al. (2012) Microscopic differential cell counting to identify inflammatory reactions in dairy cow quarter milk samples. Journal of Dairy Science, 95, 4410–4420. https://doi.org/10.3168/jds.2012-5331 Pilla, R. et al. (2013) Differential cell count as an alternative method to diagnose dairy cow mastitis. Journal of Dairy Science, 96, 1653–1660. https://doi.org/10.3168/jds.2012-6298 Pillai, S.R. et al. (2001) Application of differential inflammatory cell count as a tool to monitor udder health. Journal of Dairy Science, 84, 1413–1420. https://doi.org/10.3168/jds.S0022-0302(01)70173-7 R Core Team (2018) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/. Ruegg, P.L. and Pantoja, J.C.F. (2013) Understanding and using somatic cell counts to improve milk quality. Irish Journal of Agricultural and Food Research, 52, 101-117. Schukken, Y.H. et al. (2003) Monitoring udder health and milk quality using somatic cell counts. Veterinary Research, 34, 579-596. https://doi.org/10.1051/vetres:2003028 Schwarz, D. et al. (2011) Microscopic differential cell counts in milk for the evaluation of inflammatory reactions in clinically healthy and subclinically infected bovine mammary glands. Journal of Dairy Research, 78, 448–455. https://doi.org/10.1017/S0022029911000574 Schwarz, D. et al. (2019) Investigation of differential somatic cell count as a potential new supplementary indicator to somatic cell count for identification of intramammary infection in dairy cows at the end of the lactation period. Preventive Veterinary Medicine, 172, 104803. https://doi.org/10.1016/j.prevetmed.2019.104803 Viale, E. et al. (2017) Association of candidate gene polymorphisms with milk technological traits, yield, composition, and somatic cell score in Italian Holstein-Friesian sires. Journal of Dairy Science, 100, 7271–7281. https://doi.org/10.3168/jds.2017-12666 Wall, S. K. et al. (2018) Differential somatic cell count in milk before, during, and after artificially induced immune reactions of the mammary gland. Journal of Dairy Science, 101, 5362–5373. https://doi.org/10.3168/jds.2017-14152 Youden, W.J. (1950) An index for rating diagnostic tests. Cancer, 3, 32–35. https://doi.org/10.1002/1097-0142(1950)3:1&lt;32::AID-CNCR2820030106&gt;3.0.CO;2-3 Zecconi, A. et al. (2019) Assessment of subclinical mastitis diagnostic accuracy by differential cell count in individual cow milk. Italian Journal of Animal Science, 18, 460-465. https://doi.org/10.1080/1828051X.2018.1533391 Â

Adjustment for the measurement error in evaluating biomarker performances at baseline for future survival outcomes: Time-dependent receiver operating characteristic curve within a joint modelling framework

The performance of a biomarker is defined by how well the biomarker is capable to distinguish between healthy and diseased individuals. This assessment is usually based on the baseline value of the biomarker; the value at the earliest time point of the patient follow-up, and quantified by ROC (receiver operating characteristic) curve analysis. However, the observed baseline value is often subjected to measurement error due to imperfect laboratory conditions and limited machine precision. Failing to adjust for measurement error may underestimate the true performance of the biomarker, and in a direct comparison, useful biomarkers could be overlooked. We develop a novel approach to account for measurement error when calculating the performance of the baseline biomarker value for future survival outcomes. We adopt a joint longitudinal and survival data modelling formulation and use the available longitudinally repeated values of the biomarker to make adjustment of the measurement error in time-dependent ROC curve analysis. Our simulation study shows that the proposed measurement error-adjusted estimator is more efficient for evaluating the performance of the biomarker than estimators ignoring the measurement error. The proposed method is illustrated using Mayo Clinic primary biliary cirrhosis (PBC) study.

Prognostic significance of preoperative plasma fibrinogen levels in primary gastrointestinal stromal tumours: a retrospective cohort study

Improved prediction of prognosis for gastrointestinal stromal tumours (GISTs) has become increasingly important since the introduction of targeted therapy. Here, we aimed to evaluate the prognostic significance of preoperative plasma fibrinogen (Fib) levels in patients with primary GISTs and to analyse their correlations with clinicopathological characteristics. A total of 201 previously untreated patients with primary GISTs who had undergone radical surgery at our institution between October 2004 and July 2018 were enrolled. The optimal cut-off value for Fib levels was calculated using time-dependent receiver-operating characteristic curve analysis. RFS, the primary endpoint, was calculated by the Kaplan-Meier method and compared by the log-rank test. Univariate and multivariate Cox regression models were calculated. High preoperative plasma Fib levels were detected as an independent adverse prognostic factor (p = 0.008, hazard ratio 3.136, 95% CI 1.356‒7.256). Furthermore, high preoperative plasma Fib levels also indicated a poor prognosis within the modified National Institutes of Health (mNIH) high-risk subgroup (p = 0.041). In addition, preoperative plasma Fib levels showed a positive correlation with several prognostic factors and even a linear relationship with tumour size (Spearman correlation coefficient [r] = 0.411, p < 0.001). Our results suggest that high preoperative plasma Fib levels may indicate a poor prognosis in patients with primary GISTs. As a cost-effective biomarker, preoperative assessment of plasma Fib levels may help to further risk stratify patients with mNIH high-risk GISTs and instruct the application of targeted therapy.

Superiority of Epstein-Barr Virus (EBV) DNA in the Plasma over Whole Blood in Prognostication of Extranodal NK/T Cell Lymphoma, Nasal Type (ENKTL)

Background Extranodal natural killer T cell lymphoma (ENKTL) is a rare subtype of non-Hodgkin lymphoma, of which pathogenesis is related to Epstein-Barr virus (EBV) infection. The presence of EBV-DNA in blood is a well-known prognosticator, which is also included in a prognostic model, PINK-E (Kim SJ et al, Lancet Oncol 2016). Furthermore, its presence after treatment is known to be a predictor of the treatment failure (Kim SJ et al, Lancet Haematol 2015). However, no consensus have been made on which blood sample (i.e, whole blood vs. plasma) is best to be used for the test. While EBV in plasma is known to have a higher specificity and sensitivity for EBV-related diseases as compared with EBV in peripheral blood mononuclear cells (Kanakry JA et al., Blood 2016), PINK-E model incorporated EBV status regardless of types of samples. We aimed to compare the prognostic performance of EBV DNA titers from whole blood (WB) and plasma. Method The EBV DNA viral load was measured in both WB and plasma samples by real-time quantitative PCR at the time of diagnosis, during and after completion of planned treatment. DNA was extracted using Artus® EBV RG PCR kit (Qiagen Inc.). The limit of detection level was 2.66 log copies/mL according to the manufacturer. Prognostic accuracy was compared by using time-dependent receiver operating characteristic (ROC) curves and corresponding area under the curve (AUC) values. Results A total of 60 patients with newly diagnosed ENKTL between September 2014 and September 2018 were included in this analysis. Twenty nine patients (48.3%) were in stage I, 6 (10.0%) in stage II, and 25 (41.7%) in stage IV. Forty six (76.7 %) patients had nasal involvement. Among them, 33 patients underwent concurrent chemoradiotherapy with weekly cisplatin and 53 received L-asparaginase containing chemotherapy (VIDL, n=46, 86.8%; SMILE, n=7, 13.2%) as definitive treatment or after chemoradiotherapy. Fifty two patients were evaluable for response: 33 (63.5%) complete response, 9 (17.3%) partial response, and 10 (19.2%) progressive disease. With a median follow-up duration of 34.1 months (range, 1.2-57.5 months), 2-year progression-free survival (PFS) and overall survival (OS) rates were 55.0% and 63.0%, respectively. EBV-DNA was detected in 37 (61.7%) in pretreatment WB and 23 patients (38.3%) in plasma samples respectively (Table 1). There was moderate agreement (ƙ=0.49) between the tests. Discordant results were noted in 16 (26.7%) patients, 15 of whom were WB (+) and plasma (-). Presence of EBV DNA in plasma was significantly correlated with unfavorable clinical characteristics than that in WB (Table 1). WB EBV positivity (2-year PFS rates, 42% vs 74%; HR 2.61, 95% CI, 1.04-6.52, p=0.041) and plasma EBV positivity (2-year PFS rates, 23% vs 75%; HR 4.87, 95% CI, 2.15-11.03, p&lt;0.0001) were both associated with poor PFS. Plasma EBV positivity at diagnosis was a highly significant prognosticator of OS (2-year OS rates of 27% vs. 85%; HR 8.38, 95% CI, 3.03-23.19, p&lt;0.0001). EBV status in the WB also tended to be associated with survival but without statistical significance (2-year OS rates, 47% vs 79%, HR 3.34, 95% CI, 1.12-9.96, P=0.071). Furthermore, WB EBV positivity in those with plasma EBV negativity was not prognostic for OS (2-year OS rates, 86% vs 84%, p=0.944) (Figure 1). On the other hand, plasma EBV status in those with WB EBV positivity was a significant predictor of OS (2yr OS rates, 84% vs. 29%, p&lt;0.001). Multivariable analysis showed plasma EBV positivity was significantly association with poor PFS (HR 4.22, 95% CI, 1.79-9.97, p=0.001) and OS (HR 8.39, 95% CI, 3.03-23.19, p=0.002) (Table 2). The prognostic accuracy assessed by the time dependent ROC curve analysis for 2-year OS rates also showed higher AUCs with plasma than WB (0.79, 95% CI, 0.63-0.92 for plasma; 0.66, 95% CI, 0.37-0.81 for WB). After completion of treatment, 22 patients were tested for EBV titer. Among those with plasma EBV positivity, WB EBV positivity was not associated with OS (2-year OS rates, 83% vs 62%, p=0.488). By contrast, plasma EBV positivity among those with WB EBV positivity was a significant predictor of OS (2yr OS rates, 62% vs. 17%, p=0.008) (Figure 2). In addition, post treatment plasma EBV DNA titer was a significant prognostic factors for PFS and OS by multivariable analysis (Table 3). Conclusion EBV DNA test in plasma is more accurate prognostic marker for ENKTL than that in WB as it can prognosticate regardless of EBV status in the WB. Disclosures No relevant conflicts of interest to declare.

A Novel Prognostic Signature of Transcription Factors for the Prediction in Patients With GBM.

Background: Although the diagnosis and treatment of glioblastoma (GBM) is significantly improved with recent progresses, there is still a large heterogeneity in therapeutic effects and overall survival. The aim of this study is to analyze gene expressions of transcription factors (TFs) in GBM so as to discover new tumor markers. Methods: Differentially expressed TFs are identified by data mining using public databases. The GBM transcriptome profile is downloaded from The Cancer Genome Atlas (TCGA). The nonnegative matrix factorization (NMF) method is used to cluster the differentially expressed genes to discover hub genes and signal pathways. The TFs affecting the prognosis of GBM are screened by univariate and multivariate COX regression analysis, and the receiver operating characteristic (ROC) curve is determined. The GBM hazard model and nomogram map are constructed by integrating the clinical data. Finally, the TFs involving potential signaling pathways in GBM are screened by Gene Set Enrichment Analysis (GSEA), Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. Results: There are 68 differentially expressed TFs in GBM, of which 43 genes are upregulated and 25 genes are downregulated. NMF clustering analysis suggested that GBM patients are divided into three groups: Clusters A, B, and C. LHX2, MEOX2, SNAI2, and ZNF22 are identified from the above differential genes by univariate/multivariate regression analysis. The risk score of those four genes are calculated based on the beta coefficient of each gene, and we found that the predictive ability of the risk score gradually increased with the prolonged predicted termination time by time-dependent ROC curve analysis. The nomogram results have showed that the integration of risk score, age, gender, chemotherapy, radiotherapy, and 1p/19q can further improve predictive ability towards the survival of GBM. The pathways in cancer, phosphoinositide 3-kinases (PI3K)–Akt signaling, Hippo signaling, and proteoglycans, are highly enriched in high-risk groups by GSEA. These genes are mainly involved in cell migration, cell adhesion, epithelial–mesenchymal transition (EMT), cell cycle, and other signaling pathways by GO and KEGG analysis. Conclusion: The four-factor combined scoring model of LHX2, MEOX2, SNAI2, and ZNF22 can precisely predict the prognosis of patients with GBM.

Serial monitoring of isavuconazole blood levels during prolonged antifungal therapy.

Isavuconazole is the newest triazole antifungal approved for the treatment of invasive aspergillosis (IA) and invasive mucormycosis in adult patients. To characterize the assessment of the blood levels of isavuconazole and their association with efficacy and toxicity. From January 2017 to May 2018, blood samples obtained from patients receiving isavuconazole were analysed for therapeutic drug monitoring. Factors influencing the blood concentrations of isavuconazole, such as weight, length of treatment, route of administration and results of selected liver function tests, were analysed in univariate and multivariate models. The receiver operating characteristic (ROC) curve was analysed to detect the best cut-off for isavuconazole toxicity. A total of 264 isavuconazole blood concentrations in 19 patients were analysed. The median value of isavuconazole concentration in all patients during the first 30 days of therapy was 3.69 mg/L (range 0.64-8.13 mg/L). A linear increase of 0.032 mg/L (range 0.023-0.041 mg/L) for each day of treatment (P = 0.002) was observed. In multivariate analysis the association between the length of treatment and higher levels of isavuconazole (P < 0.001) and higher serum GGT and lower isavuconazole levels (P = 0.001) was confirmed. Adverse events, mainly gastrointestinal, were reported in six patients (31.6%). Based on time-dependent and fixed-time ROC curve analysis, 4.87 mg/L and 5.13 mg/L, respectively, were the identified thresholds for toxicity. Isavuconazole was efficacious and well tolerated. Side effects, mainly gastrointestinal, were associated with prolonged administration and high serum levels.

Seven-senescence-associated gene signature predicts overall survival for Asian patients with hepatocellular carcinoma.

BACKGROUNDCellular senescence is a recognized barrier for progression of chronic liver diseases to hepatocellular carcinoma (HCC). The expression of a cluster of genes is altered in response to environmental factors during senescence. However, it is questionable whether these genes could serve as biomarkers for HCC patients.AIMTo develop a signature of senescence-associated genes (SAGs) that predicts patients’ overall survival (OS) to improve prognosis prediction of HCC.METHODSSAGs were identified using two senescent cell models. Univariate COX regression analysis was performed to screen the candidate genes significantly associated with OS of HCC in a discovery cohort (GSE14520) for the least absolute shrinkage and selection operator modelling. Prognostic value of this seven-gene signature was evaluated using two independent cohorts retrieved from the GEO (GSE14520) and the Cancer Genome Atlas datasets, respectively. Time-dependent receiver operating characteristic (ROC) curve analysis was conducted to compare the predictive accuracy of the seven-SAG signature and serum α-fetoprotein (AFP).RESULTSA total of 42 SAGs were screened and seven of them, including KIF18B, CEP55, CIT, MCM7, CDC45, EZH2, and MCM5, were used to construct a prognostic formula. All seven genes were significantly downregulated in senescent cells and upregulated in HCC tissues. Survival analysis indicated that our seven-SAG signature was strongly associated with OS, especially in Asian populations, both in discovery and validation cohorts. Moreover, time-dependent ROC curve analysis suggested the seven-gene signature had a better predictive accuracy than serum AFP in predicting HCC patients’ 1-, 3-, and 5-year OS.CONCLUSIONWe developed a seven-SAG signature, which could predict OS of Asian HCC patients. This risk model provides new clinical evidence for the accurate diagnosis and targeted treatment of HCC.

Determining the optimal cutoff point for lymph node density and its impact on overall survival in children with Wilms' tumor.

ObjectivePrevious studies showed that the lymph node density (LND) was a predictor of survival in Wilms’ tumor (WT). However, the optimal LND cutoff point is controversial due to methodological shortcomings of previous studies, and no studies have shown the effect of LND on survival in children with WT. The purpose of this study was to remedy this situation.MethodsWe identified 376 children with WT. LND cutoff point was determined using the median value, the X-tile program, the survival-tree algorithm, and the time-dependent ROC curve analysis. Survival functions were estimated by the Kaplan–Meier method. We used Cox regression analysis to determine the impact of LND on survival. Smooth curve fitting between relative mortality risk and LND was performed.ResultsThe LND cutoff point was 0.44, 0.65, 0.65, and 0.64 according to the median value, the X-tile program, the survival-tree algorithm, and the time-dependent ROC curve analysis, respectively. The 5-, 10-, and 20-year overall survival rates were 86.9%, 86.9%, and 84.7%, respectively, in the <0.44 group and 81.3%, 80.3%, and 80.3%, respectively, in the ≥0.44 group. Survival did not differ significantly between the two groups (P=0.185). The 5-, 10-, and 20-year overall survival rates were 87.8%, 87.8%, and 86.0%, respectively, in the < 0.65 or < 0.64 group and 76.5%, 75.1%, and 75.1%, respectively, in the ≥ 0.65 or ≥ 0.64 group. Children with the high LND had a significantly worse survival (P=0.011) if 0.64 or 0.65 was used for the stratification. LND was a significant predictor for overall survival in the multivariate Cox regression analysis (HR =1.797; 95% CI, 1.043–3.097; P=0.035). Smooth curve fitting suggested that the risk of mortality tended to be ascending with the increase in LND in general.ConclusionThe three methods including the X-tile program, the survival-tree algorithm, and the time-dependent receiver operating characteristic (ROC) curve analysis are equivalent in their ability to stratify patients and clearly better than the median method. The results showed that the optimal LND cutoff point was around 0.65 and the LND was a reliable predictor of overall survival in children with WT.

Preoperative Proteinuria and Reduced Glomerular Filtration Rate Predicts Renal Replacement Therapy in Patients Supported With Continuous-Flow Left Ventricular Assist Devices.

Renal failure requiring renal replacement therapy (RRT) has detrimental effects on quality of life and survival of patients with continuous-flow left ventricular assist devices (CF-LVADs). Current guidelines do not offer a decision-making algorithm for CF-LVAD candidates with poor baseline renal function. Objective of this study was to identify risk factors associated with RRT after CF-LVAD implantation. Three hundred and eighty-nine consecutive patients underwent contemporary CF-LVAD implantation at the Columbia University Medical Center between January 2004 and August 2015. Baseline demographics, comorbid conditions, clinical risk scores, and renal function were analyzed in patients with or without RRT after CF-LVAD implantation. Time-dependent receiver-operating characteristic curve analysis was performed to define optimal cutoffs for continuous risk factors. Forty-four patients (11.6%) required RRT during a median follow-up of 9.9 months. Patients requiring RRT had significantly worse renal function, lower hemoglobin, and increased proteinuria at baseline. Low estimated glomerular filtration rate (<40 mL/min/1.73 m2) and proteinuria (urine protein to creatinine ratio ≥0.55 mg/mg) were significant predictors of RRT after CF-LVAD support. Dipstick proteinuria was also a significant predictor of RRT after CF-LVAD implantation. Patients with both low estimated glomerular filtration rate and proteinuria had highest risk of RRT (63.6%) compared with those with either low estimated glomerular filtration rate or proteinuria (18.7%) and those with neither of these risk factors (2.7%) at 1-year follow-up (log-rank P<0.001). Estimated glomerular filtration rate and proteinuria are predictors RRT after CF-LVAD implantation and should be routinely assessed in CF-LVAD candidates to guide decision making.

Volume-based assessment by 18F-FDG PET/CT predicts survival in patients with stage III non-small-cell lung cancer

We evaluated the prognostic impact of volume-based assessment by (18)F-FDG PET/CT in patients with stage III non-small-cell lung cancer (NSCLC). We reviewed 194 consecutive patients with stage IIIA NSCLC treated with surgical resection (surgical group) and 115 patients treated with nonsurgical therapy (nonsurgical group: 50 stage IIIA, 65 stage IIIB). Metabolic tumour volume (MTV), total lesion glycolysis (TLG), and maximum standardized uptake value (SUVmax) of primary tumours were measured using pretreatment (18)F-FDG PET/CT. Overall survival was assessed using the Kaplan-Meier method. The prognostic significance of PET parameters and other clinical variables was assessed using Cox proportional hazards regression analyses. To evaluate and compare the predictive performance of PET parameters, time-dependent receiver operating characteristic (ROC) curve analysis was used. In the Cox proportional hazards models, MTV (HR=1.27 for a doubling of MTV, P=0.008) and TLG (HR=1.22 for a doubling of TLG, P=0.035) were significantly associated with an increased risk of death after adjusting for age, gender, histological cell type, T stage, N stage, and treatment variables in the surgical group. SUVmax was not a significant prognostic factor in either the surgical or nonsurgical group. In the time-dependent ROC curve analysis, volume-based PET parameters predicted survival better than SUVmax. The volume-based PET parameters (MTV and TLG) are significant prognostic factors for survival independent of tumour stage and better prognostic imaging biomarkers than SUVmax in patients with stage IIIA NSCLC after surgical resection.

Volume-Based Parameters of 18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography Improve Outcome Prediction in Early-Stage Non–Small Cell Lung Cancer After Surgical Resection

To evaluate the prognostic significance and predictive performance of volume-based parameters of F-fluorodeoxyglucose positron emission tomography/computed tomography (F-FDG PET/CT) in early-stage non-small cell lung cancer (NSCLC). Although surgical resection remains the optimal treatment for early-stage NSCLC, approximately 40% of patients with stage I and 60% of patients with stage II NSCLC relapse and die within 5 years after curative resection. Therefore, identification of additional prognostic biomarkers is needed to develop risk-adapted treatment strategies. We retrospectively reviewed 529 consecutive patients with pathologically proven early-stage NSCLC who underwent preoperative F-FDG PET/CT. Maximum standardized uptake value (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) for the primary tumors were measured. Overall survival (OS) and disease-free survival (DFS) were assessed by the Kaplan-Meier method. The prognostic significance of PET parameters and other clinicopathological variables was assessed by Cox proportional hazards regression analysis. To evaluate and compare the predictive performance of PET parameters, time-dependent receiver operating characteristic (ROC) curve analysis was used. In the multivariate analyses, volume-based parameters of PET (MTV and TLG) that were analyzed as continuous variables were significantly associated with an increased risk of recurrence (P = 0.001 for MTV, P < 0.001 for TLG) and death (P = 0.009 for MTV, P = 0.007 for TLG), after adjusting for age, sex, histology, tumor stage, and type of surgery. SUVmax analyzed as a continuous variable was not a significant prognostic factor for both DFS (P = 0.056) and OS (P = 0.525). In the time-dependent ROC curve analysis, the volume-based parameter of PET showed better predictive performance than SUVmax (P < 0.001). The volume-based parameter of PET is an independent prognostic factor for survival in addition to pathological tumor-node-metastasis stage and a promising tool for better prediction of outcome in patients with early-stage NSCLC.