Assessing the incremental value of diagnostic and prognostic markers: a review and illustration

External validation of two clinical prediction models for pediatric pneumonia

OBJECTIVE

To externally validate two prediction models for pediatric radiographic pneumonia.

METHODS

We prospectively evaluated the performance of two prediction models (Pneumonia Risk Score [PRS] and CARPE DIEM models) from a prospective convenience sample of children 90 days - 18 years of age from a pediatric emergency department undergoing chest radiography for suspected pneumonia between January 1, 2022, to December 31st, 2023. We evaluated model performance using the original intercepts and coefficients and evaluated for performance changes when performing recalibration and re-estimation procedures.

RESULTS

We included 202 patients (median age 3 years, IQR 1-6 years), of whom radiographic pneumonia was found in 92 (41.0%). The PRS model had an area under the receiver operator characteristic curve of 0.72 (95% confidence interval [CI] 0.64-0.79), which was higher than the CARPE DIEM (0.59; 95% CI 0.51-0.67) (P<0.01). Using optimal cutpoints, the PRS model showed higher sensitivity (65.2%, 95% CI 54.6-74.9) and specificity (72.7%, 95% CI 63.4-80.8) compared to the CARPE DIEM model (sensitivity 56.5 [95% CI 45.8-66.8]; specificity 60.9 [95% CI 50.2-69.2]). Recalibration and re-estimation of models improved performance, particularly for the CARPE DIEM model, with gains in sensitivity and specificity, and improved calibration.

CONCLUSION

The PRS model demonstrated better performance than the CARPE DIEM model in predicting radiographic pneumonia. Among children with a high rate of pneumonia, these models did not reach a level of performance sufficient to be used independently of clinical judgement. These findings highlight the need for further validation and improvement of models to enhance their utility.

Relative performance evaluation of four bleeding risk scores in atrial fibrillation patients. What does the new DOAC score provide?

BACKGROUND

Recently, the direct oral anticoagulant (DOAC) score was developed and better predicted major bleeding in DOAC-treated patients with atrial fibrillation (AF) than HASBLED did. Little is known on the new score's performance regarding other bleeding risk in AF.

METHODS

We studied 14,672 patients diagnosed with AF between 2014 and 2018. During follow-up, we assessed the performance of DOAC score compared with the HASBLED, ORBIT and SWISS scores at predicting major bleeding in DOACs and non-DOACs users. Discrimination, calibration and decision curve analysis (DCA) were used to assess the risk scorer's performance.

RESULTS

There were 1484 (10.1%) patients on DOACs, 9730 on vitamin K antagonist (VKA), and 3458 on non-oral anticoagulants. Over a median of 3.5 years of follow-up, 79 major bleedings occurred in the DOAC patients, and 486 in the VKA patients (cumulative incidences = 7.4 and 13.9 per 100 patient-years, respectively). Amongst the DOAC patients, the DOAC score discrimination was moderate (C-statistic = 0.711), but significantly higher than HASBLED (C = 0.640; p = 0.03), ORBIT (C = 0.660; p = 0.04), and SWISS scores (C = 0.637; p = 0.002). The DCA showed higher net benefit using DOAC score compared with the remaining scores. In the VKA patients, DOAC score showed the highest discrimination (c-statistic = 0.709), followed by ORBIT (C = 0.692; p = 0.07), HASBLED and SWISS (C = 0.635 and 0.624, respectively; p < 0.01). All risk scores calibrated well, although HASBLED showed relatively poor calibration.

CONCLUSIONS

The new DOAC bleeding risk score is a valid and reasonable predictor of major bleeding over a median of 3.5 years of follow-up. Physicians can be reassured about the applicability of DOAC score for bleeding risk stratification in AF patients.

REGISTRATION

URL: https://www.

CLINICALTRIALS

gov; Unique identifier: NCT04364516.

Tutorial: dos and don'ts in clinical prediction research for venous thromboembolism

Clinical prediction modeling has become an increasingly popular domain of venous thromboembolism research in recent years. Prediction models can help healthcare providers make decisions regarding starting or withholding therapeutic interventions, or referrals for further diagnostic workup, and can form a basis for risk stratification in clinical trials. The aim of the current guide is to assist in the practical application of complicated methodological requirements for well-performed prediction research by presenting key dos and don'ts while expanding the understanding of predictive research in general for (clinical) researchers who are not specifically trained in the topic; throughout we will use prognostic venous thromboembolism scores as an exemplar.

Posttraumatic stress disorder in people with dementia: study protocol

Background: Posttraumatic stress disorder (PTSD) is considered an independent risk factor for dementia. Despite the (clinical) evidence that PTSD is associated with neuropsychiatric symptoms in people with dementia, studies on its prevalence and clinical manifestation are limited, and their quality is affected by the lack of a structured method to diagnose PTSD in this population. The primary aim of the current study is to validate the 'TRAuma and DEmentia' interview as a diagnostic tool for PTSD in people with dementia and to test feasibility of EMDR treatment for people with PTSD and dementia.Methods: This prospective multi-centre study is divided into two parts. In study A, 90 participants with dementia will be included to test the criterion validity, inter-rater reliability and feasibility of the 'TRAuma and DEmentia' interview. In study B, 29 participants with dementia and PTSD will receive eye movement desensitisation and reprocessing therapy by a trained psychologist, and 29 participants with dementia and PTSD will be placed on the waiting list control group.Conclusion: This study aims to improve the diagnostic process of PTSD and to assess the effects of eye movement desensitisation and reprocessing treatment in people with dementia living in Dutch care facilities.Trial registration: NL70479.068.20 / METC 20-063 / OSF registration: https://doi.org/10.17605/OSF.IO/AKW4F.

Predicting Recurrence-Free and Overall Survival for Patients With Stage II Melanoma: The MIA Calculator

PURPOSE

Improvements in recurrence-free survival (RFS) were demonstrated in two recent randomized trials for patients with sentinel node (SN)-negative stage IIB or IIC melanoma receiving adjuvant systemic therapy (pembrolizumab/nivolumab). However, adverse events also occurred. Accurate individualized prognostic estimates of RFS and overall survival (OS) would allow patients to more accurately weigh the risks and benefits of adjuvant therapy. Since the current American Joint Committee on Cancer eighth edition (AJCC-8) melanoma staging system focuses on melanoma-specific survival, we developed a multivariable risk prediction calculator that provides estimates of 5- and 10-year RFS and OS for these patients.

METHODS

Data were extracted from the Melanoma Institute Australia (MIA) database for patients diagnosed with stage II (clinical or pathological) melanoma (n = 3,220). Survival prediction models were developed using multivariable Cox regression analyses (MIA models) and externally validated twice using data sets from the United States and the Netherlands. Each model's performance was assessed using C-statistics and calibration plots and compared with Cox models on the basis of AJCC-8 staging (stage models).

RESULTS

The 5-year and 10-year RFS C-statistics were 0.70 and 0.73 (MIA-model) versus 0.61 and 0.60 (stage-model), respectively. For OS, the 5-year and 10-year C-statistics were 0.71 and 0.75 (MIA-model) compared with 0.62 and 0.61 (stage-model), respectively. The MIA models were well calibrated and externally validated.

CONCLUSION

The MIA models offer accurate and personalized estimates of both RFS and OS in patients with stage II melanoma even in the absence of pathological staging with SN biopsy. These models were robust on external validations and may be used in everyday practice both with (ideally) and without performing SN biopsy to identify high-risk patients for further management strategies. An online tool will be available at the MIA website (Risk Prediction Tools).

Clinical and economic outcomes of assigning percutaneous coronary intervention patients to contrast-sparing strategies based on the predicted risk of contrast-induced acute kidney injury

OBJECTIVE

Contrast-sparing strategies have been developed for percutaneous coronary intervention (PCI) patients at increased risk of contrast-induced acute kidney injury (CI-AKI), and numerous CI-AKI risk prediction models have been created. However, the potential clinical and economic consequences of using predicted CI-AKI risk thresholds for assigning patients to contrast-sparing regimens have not been evaluated. We estimated the clinical and economic consequences of alternative CI-AKI risk thresholds for assigning Medicare PCI patients to contrast-sparing strategies.

METHODS

Medicare data were used to identify inpatient PCI from January 2017 to June 2021. A prediction model was developed to assign each patient a predicted probability of CI-AKI. Multivariable modeling was used to assign each patient two marginal predicted values for each of several clinical and economic outcomes based on (1) their underlying clinical and procedural characteristics plus their true CI-AKI status in the data and (2) their characteristics plus their counterfactual CI-AKI status. Specifically, CI-AKI patients above the predicted risk threshold for contrast-sparing were reassigned their no CI-AKI (counterfactual) outcomes. Expected event rates, resource use, and costs were estimated before and after those CI-AKI patients were reassigned their counterfactual outcomes. This entailed bootstrapped sampling of the full cohort.

RESULTS

Of the 542,813 patients in the study cohort, 5,802 (1.1%) had CI-AKI. The area under the receiver operating characteristic curve for the prediction model was 0.81. At a predicted risk threshold for CI-AKI of >2%, approximately 18.0% of PCI patients were assigned to contrast-sparing strategies, resulting in (/100,000 PCI patients) 121 fewer deaths, 58 fewer myocardial infarction readmissions, 4,303 fewer PCI hospital days, $11.3 million PCI cost savings, and $25.8 million total one-year cost savings, versus no contrast-sparing strategies.

LIMITATIONS

Claims data may not fully capture disease burden and are subject to inherent limitations such as coding inaccuracies. Further, the dataset used reflects only individuals with fee-for-service Medicare, and the results may not be generalizable to Medicare Advantage or other patient populations.

CONCLUSIONS

Assignment to contrast-sparing regimens at a predicted risk threshold close to the underlying incidence of CI-AKI is projected to result in significant clinical and economic benefits.

Determining clinical predictors to identify non-specific abdominal pain and the added value of laboratory examinations: A prospective derivation study in a paediatric emergency department

AIM

To develop a model to discriminate non-specific abdominal pain (NSAP) from organic pain in the paediatric emergency department (PED) and evaluate the added value of laboratory markers.

METHODS

Prospective cohort study in an urban French PED including all patients aged ≥4 years with abdominal pain between November 2020 and May 2021. The outcome was the discrimination between NSAP (patients coded to have only "pain" or "constipation") and organic pain (all other diagnoses) using stepwise backward multivariate logistic regression method with bootstrap resampling.

RESULTS

The study enrolled 246 patients. Overall, 163 patients (66.2%) had NSAP. Four variables associated with organic pain: pain in the epigastric region (OR 0.48 [0.23-0.99]), worsening pain (0.57 [0.32-0.99]), pain migration (0.42 [0.17-0.99]) and vomiting (0.47 [0.26-0.84]) were integrated in a clinical model. To discriminate NSAP with a probability of 65%, model sensitivity was 71.8% (64.9-78.7), specificity was 53.0% (42.3-63.7), and the Net Benefit (NB) was 15.4%. White Blood Count and C-reactive protein results improved discriminative capacity of the model (AUC 0.708 [0.643-0.773] vs. 0.654 [0.585-0.723], p = 0.01) with a supplementary NB of 12%. Patient follow-up showed 95% diagnostic accuracy.

CONCLUSION

This study reveals a four-clinical predictor model with a NB of 15% in predicting NSAP. Validation studies are necessary.

Improved Re-estimation of Perioperative Cardiac Risk Using the Surgical Apgar Score: A Retrospective Cohort Study

OBJECTIVE

To assess whether the Surgical Apgar Score (SAS) improves re-estimation of perioperative cardiac risk.

BACKGROUND

The SAS is a novel risk index that integrates three relevant and easily measurable intraoperative parameters (blood loss, heart rate, mean arterial pressure) to predict outcomes. The incremental prognostic value of the SAS when used in combination with standard preoperative risk indices is unclear.

METHODS

We conducted a retrospective cohort study of adults (18 years and older) who underwent elective noncardiac surgery at a quaternary care hospital in Canada (2009-2014). The primary outcome was postoperative acute myocardial injury. The SAS (range 0-10) was calculated based on intraoperative estimated blood loss, lowest mean arterial pressure, and lowest heart rate documented in electronic medical records. Incremental prognostic value of the SAS when combined with the Revised Cardiac Risk Index was assessed based on discrimination (c-statistic), reclassification (integrated discrimination improvement, net reclassification index), and clinical utility (decision curve analysis).

RESULTS

The cohort included 16,835 patients, of whom 607 (3.6%) patients had acute postoperative myocardial injury. Addition of the SAS to the Revised Cardiac Risk Index improved risk estimation based on the integrated discrimination improvement [2.0%; 95% confidence interval (CI): 1.5%-2.4%], continuous net reclassification index (54%; 95% CI: 46%-62%), and c-index, which increased from 0.68 (95% CI: 0.65-0.70) to 0.75 (95% CI: 0.73-0.77). On decision curve analysis, addition of the SAS to the Revised Cardiac Risk Index resulted in a higher net benefit at all decision thresholds.

CONCLUSIONS

When combined with a validated preoperative risk index, the SAS improved the accuracy of cardiac risk assessment for noncardiac surgery. Further research is needed to delineate how intraoperative data can better guide postoperative decision-making.

Cardiovascular risk indicators among depressed persons: A special case?

BACKGROUND

Traditional cardiovascular risk indicators only partially explain cardiovascular risks in depressed persons. Depressed persons may exhibit a profile of cardiovascular risk indicators that goes beyond traditional cardiovascular risk indicators, such as symptom severity, insomnia, loneliness and neuroticism, yet research on the added value of these depression-related characteristics in predicting cardiovascular risks of depressed persons is scarce.

METHODS

Data from N = 1028 depressed Dutch adults without prevalent CVD were derived from two longitudinal depression cohort studies. The outcome was medication-confirmed self-reported CVD. Fifteen depression-related clinical and psychological characteristics were included and tested against traditional cardiovascular risk indicators. Data were analysed using Cox regression models. Incremental values of these characteristics were calculated using c-statistics.

RESULTS

After a median follow-up of 65.3 months, 12.7% of the participants developed CVD. Only anxiety and depressive symptom severity were associated with incident CVD beyond traditional cardiovascular risk indicators. The c-statistic of the model with traditional cardiovascular risk indicators was 85.47%. This increased with 0.56 or 0.33 percentage points after inclusion of anxiety or depression severity, respectively.

LIMITATIONS

Other relevant depression-related characteristics were not available in the datasets used.

CONCLUSION

Anxiety and depressive symptom severity were indicative of an increased cardiovascular risk. Including these as additional risk indicators barely improved the ability to assess cardiovascular risks in depressed persons. Although traditional cardiovascular risk indicators performed well in depressed persons, existing risk prediction algorithms need to be validated in depressed persons.

A New Clinical Prediction Rule for Infective Endocarditis in Emergency Department Patients With Fever: Definition and First Validation of the CREED Score

Background Infective endocarditis (IE) could be suspected in any febrile patients admitted to the emergency department (ED). This study was aimed at assessing clinical criteria predictive of IE and identifying and prospectively validating a sensible and easy-to-use clinical prediction score for the diagnosis of IE in the ED. Methods and Results We conducted a retrospective observational study, enrolling consecutive patients with fever admitted to the ED between January 2015 and December 2019 and subsequently hospitalized. Several clinical and anamnestic standardized variables were collected and evaluated for the association with IE diagnosis. We derived a multivariate prediction model by logistic regression analysis. The identified predictors were assigned a score point value to obtain the Clinical Rule for Infective Endocarditis in the Emergency Department (CREED) score. To validate the CREED score we conducted a prospective observational study between January 2020 and December 2021, enrolling consecutive febrile patients hospitalized after the ED visit, and evaluating the association between the CREED score values and the IE diagnosis. A total of 15 689 patients (median age, 71 [56-81] years; 54.1% men) were enrolled in the retrospective cohort, and IE was diagnosed in 267 (1.7%). The CREED score included 12 variables: male sex, anemia, dialysis, pacemaker, recent hospitalization, recent stroke, chest pain, specific infective diagnosis, valvular heart disease, valvular prosthesis, previous endocarditis, and clinical signs of suspect endocarditis. The CREED score identified 4 risk groups for IE diagnosis, with an area under the receiver operating characteristic curve of 0.874 (0.849-0.899). The prospective cohort included 13 163 patients, with 130 (1.0%) IE diagnoses. The CREED score had an area under the receiver operating characteristic curve of 0.881 (0.848-0.913) in the validation cohort, not significantly different from the one calculated in the retrospective cohort (P=0.578). Conclusions In this study, we propose and prospectively validate the CREED score, a clinical prediction rule for the diagnosis of IE in patients with fever admitted to the ED. Our data reflect the difficulty of creating a meaningful tool able to identify patients with IE among this general and heterogeneous population because of the complexity of the disease and its low prevalence in the ED setting.

Decision curve analysis confirms higher clinical utility of multi-domain versus single-domain prediction models in patients with open abdomen treatment for peritonitis

BACKGROUND

Prediction modelling increasingly becomes an important risk assessment tool in perioperative systems approaches, e.g. in complex patients with open abdomen treatment for peritonitis. In this population, combining predictors from multiple medical domains (i.e. demographical, physiological and surgical variables) outperforms the prediction capabilities of single-domain prediction models. However, the benefit of these prediction models for clinical decision-making remains to be investigated. We therefore examined the clinical utility of mortality prediction models in patients suffering from peritonitis with a decision curve analysis.

METHODS

In this secondary analysis of a large dataset, a traditional logistic regression approach, three machine learning methods and a stacked ensemble were employed to examine the predictive capability of demographic, physiological and surgical variables in predicting mortality under open abdomen treatment for peritonitis. Calibration was examined with calibration belts and predictive performance was assessed with the area both under the receiver operating characteristic curve (AUROC) and under the precision recall curve (AUPRC) and with the Brier Score. Clinical utility of the prediction models was examined by means of a decision curve analysis (DCA) within a treatment threshold range of interest of 0-30%, where threshold probabilities are traditionally defined as the minimum probability of disease at which further intervention would be warranted.

RESULTS

Machine learning methods supported available evidence of a higher prediction performance of a multi- versus single-domain prediction models. Interestingly, their prediction performance was similar to a logistic regression model. The DCA demonstrated that the overall net benefit is largest for a multi-domain prediction model and that this benefit is larger compared to the default "treat all" strategy only for treatment threshold probabilities above about 10%. Importantly, the net benefit for low threshold probabilities is dominated by physiological predictors: surgical and demographics predictors provide only secondary decision-analytic benefit.

CONCLUSIONS

DCA provides a valuable tool to compare single-domain and multi-domain prediction models and demonstrates overall higher decision-analytic value of the latter. Importantly, DCA provides a means to clinically differentiate the risks associated with each of these domains in more depth than with traditional performance metrics and highlighted the importance of physiological predictors for conservative intervention strategies for low treatment thresholds. Further, machine learning methods did not add significant benefit either in prediction performance or decision-analytic utility compared to logistic regression in these data.

Temporal validation of a multivariable surgical mortality prediction model (NZRisk): a New Zealand national cohort study

OBJECTIVES

Clinical risk calculators (CRCs), such as NZRisk, are used daily by clinicians to guide clinical decisions and explain individual risk to patients. The utility and robustness of these tools depends on the methods used to create the underlying mathematical model, as well as the stability of that model in relation to changing clinical practice and patient populations over time. The later should be checked by temporal validation using external data. Few if any of the clinical prediction models in current clinical use have published temporal validation. Here, we use a large external dataset to temporally validate NZRisk; a perioperative risk prediction model used in the New Zealand population.

METHODS

A sample of 1 976 362 adult non-cardiac surgical procedures collected over 15 years from the New Zealand Ministry of Health National Minimum Dataset, was used to temporally validate NZRisk. We divided the dataset into 15 single year cohorts and compared 13 of these to our NZRisk model (2 years used for the model building were excluded). We compared the area under the curve (AUC) value, calibration slope and intercept for each single year cohort, to the same values produced by the data used to create NZRisk, by fitting a random effects meta-regression with each year cohort acting as a separate study point. In addition, we used two-sided t-tests to compare each measure across the cohorts.

RESULTS

The AUC values for the 30-day NZRisk model applied to our single year cohorts ranged from 0.918 to 0.940 (NZRisk AUC was 0.921). There were eight statistically different AUC values for the following years 2007-2009, 2016 and 2018-2021. The intercept values ranged from -0.004 to 0.007 and 7 years had statistically significant different intercepts during leave-one-out t-tests; 2007-2010, 2012, 2018 and 2021. The slope values ranged from 0.72 to 1.12 and 7 years had statistically significant different slopes during leave-one-out t-tests; 2010, 2011, 2017, 2018 and 2019-2021. The random effects meta-regression upheld our results related to AUC (0.54 (95% CI 0.40 to 0.99), I² 67.57 (95% CI 40.67 to 88.50), Cochran's Q<0.001) and slope (τ 0.14 (95% CI 0.01 to 0.23), I² 98.61 (95% CI 97.31 to 99.50), Cochran's Q<0.001) between year difference.

CONCLUSION

The NZRisk model shows differences in AUC and slope but not intercept values over time. The biggest differences were in the calibration slope. The models maintained excellent discrimination over time as shown by the AUC values. These findings suggest we update our model in the next 5 years. To our knowledge, this is the first temporal validation of a CRC in current use.

One Heartbeat Away from a Prediction Model for Cardiovascular Diseases in Patients with Chronic Kidney Disease: A Systematic Review

INTRODUCTION

Patients with chronic kidney disease (CKD) have a high risk of cardiovascular disease (CVD). Prediction models, combining clinical and laboratory characteristics, are commonly used to estimate an individual's CVD risk. However, these models are not specifically developed for patients with CKD and may therefore be less accurate. In this review, we aim to give an overview of CVD prognostic studies available, and their methodological quality, specifically for patients with CKD.

METHODS

MEDLINE was searched for papers reporting CVD prognostic studies in patients with CKD published between 2012 and 2021. Characteristics regarding patients, study design, outcome measurement, and prediction models were compared between included studies. The risk of bias of studies reporting on prognostic factors or the development/validation of a prediction model was assessed with, respectively, the QUIPS and PROBAST tool.

RESULTS

In total, 134 studies were included, of which 123 studies tested the incremental value of one or more predictors to existing models or common risk factors, while only 11 studies reported on the development or validation of a prediction model. Substantial heterogeneity in cohort and study characteristics, such as sample size, event rate, and definition of outcome measurements, was observed across studies. The most common predictors were age (87%), sex (75%), diabetes (70%), and estimated glomerular filtration rate (69%). Most of the studies on prognostic factors have methodological shortcomings, mostly due to a lack of reporting on clinical and methodological information. Of the 11 studies on prediction models, six developed and internally validated a model and four externally validated existing or developed models. Only one study on prognostic models showed a low risk of bias and high applicability.

CONCLUSION

A large quantity of prognostic studies has been published, yet their usefulness remains unclear due to incomplete presentation, and lack of external validation of prognostic models. Our review can be used to select the most appropriate prognostic model depending on the patient population, outcome, and risk of bias. Future collaborative efforts should aim at improving existing models by externally validating them, evaluating the addition of new predictors, and assessment of the clinical impact.

REGISTRATION

We have registered the protocol of our systematic review on PROSPERO (CRD42021228043).

Use of platelet inhibitors for digital ulcers related to systemic sclerosis: EUSTAR study on derivation and validation of the DU-VASC model

OBJECTIVE

To develop and validate the prognostic prediction model DU-VASC to assist the clinicians in decision-making regarding the use of platelet inhibitors (PIs) for the management of digital ulcers in patients with systemic sclerosis. Secondly, to assess the incremental value of PIs as predictor.

METHODS

We analysed patient data from the European Scleroderma Trials and Research group registry (one time point assessed). Three sets of derivation/validation cohorts were obtained from the original cohort. Using logistic regression, we developed a model for prediction of digital ulcers (DUs). C-Statistics and calibration plots were calculated to evaluate the prediction performance. Variable importance plots and the decrease in C-statistics were used to address the importance of the predictors.

RESULTS

Of 3710 patients in the original cohort, 487 had DUs and 90 were exposed to PIs. For the DU-VASC model, which includes 27 predictors, we observed good calibration and discrimination in all cohorts (C-statistic = 81.1% [95% CI: 78.9%, 83.4%] for the derivation and 82.3% [95% CI: 779.3%, 85.3%] for the independent temporal validation cohort). Exposure to PIs was associated with absence of DUs and was the most important therapeutic predictor. Further important factors associated with absence of DUs were lower modified Rodnan skin score, anti-Scl-70 negativity and normal CRP. Conversely, the exposure to phosphodiesterase-5 inhibitor, prostacyclin analogues or endothelin receptor antagonists seemed to be associated with the occurrence of DUs. Nonetheless, previous DUs remains the most impactful predictor of DUs.

CONCLUSION

The DU-VASC model, with good calibration and discrimination ability, revealed that PI treatment was the most important therapy-related predictor associated with reduced DU occurrence.

Development and External Validation of a Radiomics Model Derived from Preoperative Gadoxetic Acid-Enhanced MRI for Predicting Histopathologic Grade of Hepatocellular Carcinoma

Histopathologic grade of hepatocellular carcinoma (HCC) is an important predictor of early recurrence and poor prognosis after curative treatments. This study aims to develop a radiomics model based on preoperative gadoxetic acid-enhanced MRI for predicting HCC histopathologic grade and to validate its predictive performance in an independent external cohort. Clinical and imaging data of 403 consecutive HCC patients were retrospectively collected from two hospitals (265 and 138, respectively). Patients were categorized into poorly differentiated HCC and non-poorly differentiated HCC groups. A total of 851 radiomics features were extracted from the segmented tumor at the hepatobiliary phase images. Three classifiers, logistic regression (LR), support vector machine, and Adaboost were adopted for modeling. The areas under the curve of the three models were 0.70, 0.67, and 0.61, respectively, in the external test cohort. Alpha-fetoprotein (AFP) was the only significant clinicopathological variable associated with HCC grading (odds ratio: 2.75). When combining AFP, the LR+AFP model showed the best performance, with an AUC of 0.71 (95%CI: 0.59-0.82) in the external test cohort. A radiomics model based on gadoxetic acid-enhanced MRI was constructed in this study to discriminate HCC with different histopathologic grades. Its good performance indicates a promise in the preoperative prediction of HCC differentiation levels.

Assessing Performance and Clinical Usefulness in Prediction Models With Survival Outcomes: Practical Guidance for Cox Proportional Hazards Models

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.

Diagnostic accuracy and added value of blood-based protein biomarkers for pancreatic cancer: A meta-analysis of aggregate and individual participant data

BACKGROUND

Novel blood-based protein biomarkers may be of value for efficient, accurate, and non-invasive diagnosis of pancreatic cancer. This study assesses the diagnostic accuracy of newly recognized blood-based protein biomarkers for detecting pancreatic cancer, and investigates their added value to CA19-9, the common blood-based biomarker in clinical use for pancreatic cancer.

METHODS

PubMed, Embase, Web of Science, and the Wiley/Cochrane Library were systematically searched from inception until June 2022. A meta-analysis of aggregate and individual participant data was conducted using frequentist and Bayesian hierarchical random-effects models. The added clinical utility of protein biomarkers was investigated using bootstrap bias-corrected decision curve analyses.

FINDINGS

Aggregate data from 28 primary studies (6127 participants) were included, of which 8 studies (1790 participants) provided individual participant data. CA19-9 was significantly more accurate than MIC-1 for distinguishing pancreatic cancer from benign disease (AUC, 0.83 vs 0.74; relative diagnostic odds ratio [rDOR], 2.10 [95% CI, 0.98-4.48]; p = 0.002), THBS2 (AUC, 0.87 vs 0.69; rDOR, 4.53 [2.16-9.39]; p < 0.0001), TIMP-1 (AUC, 0.91 vs 0.70; rDOR, 8.00 [3.81-16.9]; p < 0.0001), OPN (AUC, 0.89 vs 0.74; rDOR, 4.22 [1.13-15.6]; p < 0.0001), ICAM-1 (AUC, 0.91 vs 0.68; rDOR 9.30 [0.87-99.5]; p < 0.0001), and IGFBP2 (AUC, 0.91 vs 0.68; rDOR, 4.48 [0.78-24.3]; p < 0.0001). The addition of these novel protein biomarkers to CA19-9 did not significantly improve the AUC, and resulted in minor increases or limited decreases in clinical utility.

INTERPRETATION

Novel protein biomarkers have moderate diagnostic accuracy, do not outperform CA19-9 in differentiating pancreatic cancer from benign disease, and show limited added clinical value to CA19-9. We propose recommendations to aid the development of minimally invasive diagnostic tests with sufficient clinical utility to improve the management of patients with suspected pancreatic cancer.

FUNDING

Bennink Foundation, Dutch Cancer Foundation (KWF Kankerbestrijding), and AIRC.

Development and validation of the Chinese Geriatric Depression Risk calculator (CGD-risk): A screening tool to identify elderly Chinese with depression

BACKGROUND

The prevalence of depression among China's elderly is high, but stigma surrounding mental illness and a shortage of psychiatrists limit widespread screening and diagnosis of geriatric depression. We sought to develop a screening tool using easy-to-obtain and minimally sensitive predictors to identify elderly Chinese with depressive symptoms (depression hereafter) for referral to mental health services and determine the most important factors for effective screening.

METHODS

Using nationally representative survey data, we developed and externally validated the Chinese Geriatric Depression Risk calculator (CGD-Risk). CGD-Risk, a gradient boosting machine learning model, was evaluated based on discrimination (Concordance (C) statistic), calibration, and through a decision curve analysis. We conducted a sensitivity analysis on a cohort of middle-aged Chinese, a sub-group analysis using three data sets, and created predictor importance and partial dependence plots to enhance interpretability.

RESULTS

A total of 5681 elderly Chinese were included in the development data and 12,373 in the external validation data. CGD-Risk showed good discrimination during internal validation (C: 0.81, 95 % CI 0.79 to 0.84) and external validation (C: 0.77, 95 % CI: 0.76, 0.78). Compared to an alternative screening strategy CGD-Risk would correctly identify 17.8 more elderly with depression per 100 people screened.

LIMITATIONS

We were only able to externally validate a partial version of CGD-Risk due to differences between the internal and external validation data.

CONCLUSIONS

CGD-Risk is a clinically viable, minimally sensitive screening tool that could identify elderly Chinese at high risk of depression while circumventing issues of response bias from stigma surrounding emotional openness.

Methodological guidance for the evaluation and updating of clinical prediction models: a systematic review

BACKGROUND

Clinical prediction models are often not evaluated properly in specific settings or updated, for instance, with information from new markers. These key steps are needed such that models are fit for purpose and remain relevant in the long-term. We aimed to present an overview of methodological guidance for the evaluation (i.e., validation and impact assessment) and updating of clinical prediction models.

METHODS

We systematically searched nine databases from January 2000 to January 2022 for articles in English with methodological recommendations for the post-derivation stages of interest. Qualitative analysis was used to summarize the 70 selected guidance papers.

RESULTS

Key aspects for validation are the assessment of statistical performance using measures for discrimination (e.g., C-statistic) and calibration (e.g., calibration-in-the-large and calibration slope). For assessing impact or usefulness in clinical decision-making, recent papers advise using decision-analytic measures (e.g., the Net Benefit) over simplistic classification measures that ignore clinical consequences (e.g., accuracy, overall Net Reclassification Index). Commonly recommended methods for model updating are recalibration (i.e., adjustment of intercept or baseline hazard and/or slope), revision (i.e., re-estimation of individual predictor effects), and extension (i.e., addition of new markers). Additional methodological guidance is needed for newer types of updating (e.g., meta-model and dynamic updating) and machine learning-based models.

CONCLUSION

Substantial guidance was found for model evaluation and more conventional updating of regression-based models. An important development in model evaluation is the introduction of a decision-analytic framework for assessing clinical usefulness. Consensus is emerging on methods for model updating.

Hypochloremia as a novel adverse prognostic factor in acute liver failure

BACKGROUND AND AIMS

Emerging evidence has identified hypochloremia as an independent predictor for mortality in multiple conditions including cirrhosis. Acute liver failure (ALF) is frequently complicated by electrolyte abnormalities. We investigated the prognostic value of hypochloremia in a large cohort of ALF patients from North America.

METHODS

The Acute Liver Failure Study Group (ALFSG) registry is a longitudinal cohort study involving 2588 ALF patients enrolled prospectively from 32 North American academic centres. The primary outcome was a composite of 21-day all-cause mortality or requirement for liver transplantation (death/LT).

RESULTS

Patients with hypochloremia (<98 mEq/L) had a significantly higher 21-day mortality rate (42.1%) compared with those with normal (27.5%) or high (>107 mEq/L) chloride (28.0%) (p < .001). There was lower transplant-free cumulative survival in the hypochloremic group than in the normo- or hyper-chloremic groups (log-rank, χ2 24.2, p < .001). Serum chloride was inversely associated with the hazard of 21-day death/LT with multivariable adjustment for known prognostic factors (adjusted hazard ratio [aHR]: 0.977; 95% CI: 0.969-0.985; p < .001). Adding chloride to the ALFSG Prognostic Index more accurately predicted risk of death/LT in 19% of patients (net reclassification improvement [NRI] = 0.19, 95% CI: 0.13-0.25) but underestimated the probability of transplant-free survival in 34% of patients (NRI = -0.34, 95% CI: -0.39 to -0.28).

CONCLUSIONS

Hypochloremia is a novel independent adverse prognostic factor in ALF. A new ALFSG-Cl Prognostic Index may improve the sensitivity to identify patients at risk for death without LT.

Epidemiology of diabetes in Iran: A scoping review of studies published during 2015–2019

Background

Proper synthesis of existing epidemiologic studies on diabetes in Iran can guide future research efforts. We aimed to conduct a comprehensive scoping review on all research articles that investigated any aspect of diabetes epidemiology in Iran during 2015-2019.

Methods

This work was conducted as a part of the Iran Diabetes Research Roadmap and completed under Arksey and O'Malley's framework for scoping reviews. The Scopus and PubMed databases were searched on Feb 15th, 2020. Eligible document types on diabetes epidemiology in the Iranian population, in Persian or English, that published during the 2015-2019 period underwent eligibility assessment. A total of 315 relevant articles were included and further analysis was performed on the original studies (n = 268). Through classifying them into six domains: Diabetes incidence; the prevalence of diabetes and associated factors; the incidence/prevalence of complications/comorbid conditions; mortality/survival; burden; and prediction modeling.

Results

In total, 64 (20.3%) papers were published in Q1 journals, and 40 (12.6%) were international collaborations. No clear annual trend was present in the number of published primary or secondary articles, the portion of papers published in Q1 journals, international collaborations or relative domain proportions. Few review articles were found on prediction modeling, mortality or burden (excluding global studies).

Conclusions

Our findings show a minor portion of works on diabetic epidemiology in Iran meets the quality standards of Q1 journals. Researchers have neglected some critical subjects and have occasionally fallen for common pitfalls of epidemiologic research. In particular, adhering to established guidelines can help authors implement rigorous methods to develop, validate, and deploy practical clinical prediction models. Researchers should prioritize investigating longitudinally collected data that aid in measuring disease incidence and enable casual inference.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40200-022-01094-0.

Frailty assessment tools for use by surgeons when evaluating older adults prior to surgery: a scoping review protocol

INTRODUCTION

Despite growing evidence, uncertainty persists about which frailty assessment tools are best suited for routine perioperative care. We aim to understand which frailty assessment tools perform well and are feasible to implement.

METHODS AND ANALYSIS

Using a registered protocol following Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA), we will conduct a scoping review informed by the Joanna Briggs Institute Guide for Scoping Reviews and reported using PRISMA extension for Scoping Reviews recommendations. We will develop a comprehensive search strategy with information specialists using the Peer Review of Electronic Search Strategies checklist, and implement this across relevant databases from 2005 to 13 October 2021 and updated prior to final review publication. We will include all studies evaluating a frailty assessment tool preoperatively in patients 65 years or older undergoing intracavitary, non-cardiac surgery. We will exclude tools not assessed in clinical practice, or using laboratory or radiologic values alone. After pilot testing, two reviewers will independently assess information sources for eligibility first by titles and abstracts, then by full-text review. Two reviewers will independently chart data from included full texts using a piloted standardised electronic data charting. In this scoping review process, we will (1) index frailty assessment tools evaluated in the preoperative clinical setting; (2) describe the level of investigation supporting each tool; (3) describe useability of each tool and (4) describe direct comparisons between tools. The results will inform ready application of frailty assessment tools in routine clinical practice by surgeons and other perioperative clinicians.

ETHICS AND DISSEMINATION

Ethic approval is not required for this secondary data analysis. This scoping review will be published in a peer-review journal. Results will be used to inform an ongoing implementation study focused on geriatric surgery to overcome the current lack of uptake of older adult-oriented care recommendations and ensure broad impact of research findings.

Elevated Plasma Levels of Matrix Metalloproteinase-3 and Tissue-Inhibitor of Matrix Metalloproteinases-1 Associate With Organ Dysfunction and Mortality in Sepsis

BACKGROUND

Matrix Metalloproteinases (MMP) respond to tissue damage during sepsis. Higher plasma concentrations of MMPs and the tissue-inhibitor of matrix metalloproteinases (TIMP) have been reported in sepsis compared with healthy controls. The objective of this study was to examine if plasma levels of MMP-3, MMP-9, and TIMP-1 associate with mortality and organ dysfunction during sepsis.

METHODS

We conducted a prospective cohort study of critically ill patients with sepsis adjudicated per Sepsis-3 criteria at a tertiary academic medical center. We measured plasma concentrations of MMP-3, MMP-9, and TIMP-1 on intensive care unit admission. We phenotyped the subjects for shock, acute respiratory distress syndrome (ARDS), acute kidney injury (AKI), and mortality at 30 days. We used logistic regression to test the associations between the MMPs and TIMP-1 with shock, ARDS, AKI, and mortality.

RESULTS

Higher plasma TIMP-1 levels were associated with shock (odds ratio [OR] 1.51 per log increase [95% CI 1.25, 1.83]), ARDS (OR 1.24 [95% CI 1.05, 1.46]), AKI (OR 1.18 [95% CI 1.01, 1.38]), and mortality (OR 1.20 [95% CI 1.05, 1.46]. Higher plasma MMP-3 concentrations were associated with shock (OR 1.40 [95% CI 1.12, 1.75]) and mortality (OR 1.24 [95% CI 1.03, 1.48]) whereas MMP-9 levels were not associated with outcomes. Higher plasma TIMP-1 to MMP-3 ratios were associated with shock (OR 1.41 [95% CI 1.15, 1.72], P = 0.02).

CONCLUSION

Elevated plasma concentrations of TIMP-1 associate with organ dysfunction and mortality in sepsis. Higher plasma levels of MMP-3 associate with shock and mortality. Plasma MMP and TIMP-1 may warrant further investigation as emerging sepsis theragnostic biomarkers.

A Validated Risk Prediction Model for Bone Fragility in Children With Acute Lymphoblastic Leukemia

Although bone fragility may already be present at diagnosis of pediatric acute lymphoblastic leukemia (ALL), routine performance of dual-energy X-ray absorptiometry (DXA) in every child is not universally feasible. The aim of this study was to develop and validate a risk prediction model for low lumbar spine bone mineral density (LS BMD Z-score ≤ -2.0) at diagnosis, as an important indicator for fracture risk and further treatment-related BMD aggravation. Children with ALL (4-18 years), treated according to the Dutch Childhood Oncology Group protocol (DCOG-ALL9; model development; n = 249) and children from the Canadian Steroid-Associated Osteoporosis in the Pediatric Population cohort (STOPP; validation; n = 99) were included in this study. Multivariable logistic regression analyses were used to develop the prediction model and to confirm the association of low LS BMD at diagnosis with symptomatic fractures during and shortly after cessation of ALL treatment. The area under the receiver operating characteristic curve (AUC) was used to assess model performance. The prediction model for low LS BMD at diagnosis using weight (β = -0.70) and age (β = -0.10) at diagnosis revealed an AUC of 0.71 (95% CI, 0.63-0.78) in DCOG-ALL9 and 0.74 (95% CI, 0.63-0.84) in STOPP, and resulted in correct identification of 71% of the patients with low LS BMD. We confirmed that low LS BMD at diagnosis is associated with LS BMD at treatment cessation (OR 5.9; 95% CI, 3.2-10.9) and with symptomatic fractures (OR 1.7; 95% CI, 1.3-2.4) that occurred between diagnosis and 12 months following treatment cessation. In meta-analysis, LS BMD at diagnosis (OR 1.6; 95% CI, 1.1-2.4) and the 6-month cumulative glucocorticoid dose (OR 1.9; 95% CI, 1.1-3.2) were associated with fractures that occurred in the first year of treatment. In summary, a prediction model for identifying pediatric ALL patients with low LS BMD at diagnosis, as an important indicator for bone fragility, was successfully developed and validated. This can facilitate identification of future bone fragility in individual pediatric ALL patients. © 2021 American Society for Bone and Mineral Research (ASBMR).

Lymph Node Ratio after Neoadjuvant Chemotherapy for Stage II/III Breast Cancer: Prognostic Value Measured with Gini's Mean Difference of Restricted Mean Survival Times

Restricted mean survival time (RMST), recommended for reporting survival, lacks a tool to evaluate multilevel factors. The potential of the Gini’s mean difference of RMSTs (Δ) is explored in a comparison of a lymph node ratio-based classification (LNRc) versus a number-based classification (ypN) applied to stage II/III breast cancer patients who received neoadjuvant chemotherapy and underwent axillary dissection. Number of positive nodes (npos) classified patients into ypN0, npos = 0, ypN1, npos = [1,3], ypN2, npos = [4,9], and ypN3, npos ⩾ 10. Ratio npos/(number of nodes examined) of 0, (0,0.20], (0.20,0.65], and >0.65, classified patients into Lnr0 to Lnr3, respectively. Unadjusted and Cox-adjusted RMSTs were computed for the ypN and LNRc’s. At a follow-up time horizon of 72 months for 114 node-negative and 254 node-positive patients, unadjusted ypN0-ypN3 RMSTs were 62.4-41.4 months, Δ = 11.9 months (95%CI: 7.4-16.9), and Lnr0-Lnr3 62.4 to 36.3 months, Δ = 14.0 months (95%CI: 10.1-18.1). Cox models’ ypN1-ypN3 hazard ratios were 1.81-3.30, and Lnr1-Lnr3 1.52-4.39. Δ from Cox-fitted survival were ypN 8.1 months (95%CI: 5.9-10.5), LNRc 10.5 months (95%CI: 8.4-12.8). In conclusion, Gini’s mean difference is applicable to well established data in keeping with the literature on LNRc. It provides an alternative view on the improvement gained with a lymph node ratio-classification over using a number-classification.

Prognostic Research in Traumatic Brain Injury: Markers, Modeling, and Methodological Principles

Prognostic assessment in traumatic brain injury (TBI) is embedded deeply in clinical care. Considering the limitations of current prognostic indicators, there is increasing interest in understanding the role of new biomarkers, and in finding other prognostic indicators of long-term outcomes following TBI. New prognostic indicators may result in the development of more accurate prediction models that could be useful for both risk stratification and clinical decision making. We aimed to review methodological issues and provide tentative guidelines for prognostic research in TBI. Prognostic factor research focuses on the role of a specific patient or disease-related characteristic in relation to outcome. Typically, univariable relations of the prognostic factor are studied, followed by analyses adjusting for other variables related to the outcome. Following existing guidelines, we emphasize the importance of transparent reporting of patient and specimen characteristics, study design, clinical end-points, and statistical analysis. Prognostic model research considers combinations of predictors, with challenges for model specification, estimation, evaluation, validation, and presentation. We highlight modern approaches and opportunities related to missing values, exploration of non-linear effects, and assessing between-study heterogeneity. Prognostic research in TBI can be improved if key methodological principles are adhered to and when research is performed in collaboration among multiple centers to ensure generalizability.

Predicting recurrence in patients with sentinel node-negative melanoma: validation of the EORTC nomogram using population-based data

BACKGROUND

Identifying patients with sentinel node (SN)-negative melanoma who are at greatest risk of recurrence is important. The European Organization for Research and Treatment of Cancer (EORTC) Melanoma Group proposed a prognostic model that has not been validated in population-based data. The EORTC nomogram includes Breslow thickness, ulceration status and anatomical location as parameters. The aim of this study was to validate the EORTC model externally using a large national data set.

METHODS

Adults with histologically proven, invasive cutaneous melanoma with a negative SN biopsy in the Netherlands between 2000 and 2014 were identified from the Dutch Pathology Registry, and relevant data were extracted. The EORTC nomogram was used to predict recurrence-free survival. The predictive performance of the nomogram was assessed by discrimination (C-statistic) and calibration.

RESULTS

A total of 8795 patients met the eligibility criteria, of whom 14·7 per cent subsequently developed metastatic disease. Of these recurrences, 20·9 per cent occurred after the first 5 years of follow-up. Validation of the EORTC nomogram showed a C-statistic of 0·70 (95 per cent c.i. 0·68 to 0·71) for recurrence-free survival, with excellent calibration (R2 = 0·99; P = 0·999, Hosmer-Lemeshow test).

CONCLUSION

This population-based validation confirmed the value of the EORTC nomogram in predicting recurrence-free survival in patients with SN-negative melanoma. The EORTC nomogram could be used in clinical practice for personalizing follow-up and selecting high-risk patients for trials of adjuvant systemic therapy.

Development and Validation of Nomograms to Predict Local, Regional, and Distant Recurrence in Patients With Thin (T1) Melanomas

PURPOSE

Although the prognosis of patients with thin primary cutaneous melanomas (T1, ≤ 1.0 mm) is generally excellent, some develop recurrence. We sought to develop and validate a model predicting recurrences in patients with thin melanomas.

METHODS

A Dutch population-based cohort (n = 25,930, development set) and a cohort from an Australian melanoma treatment center (n = 2,968, validation set) were analyzed (median follow-up 6.7 and 12.0 years, respectively). Multivariable Cox models were generated for local, regional, and distant recurrence-free survival (RFS). Discrimination was assessed using Harrell's C-statistic for each outcome. Each nomogram performance was evaluated using calibration plots defining low-risk and high-risk groups as the lowest and top 5% of the nomogram risk score, respectively. The nomograms' C-statistics were compared with those of a model including the current American Joint Committee on Cancer staging parameters (T-stage and sentinel node status).

RESULTS

Local, regional, and distant recurrences were found in 209 (0.8%), 503 (1.9%), and 203 (0.8%) Dutch patients, respectively, and 23 (0.8%), 61 (2.1%), and 75 (2.5%) Australian patients, respectively. C-statistics of 0.79 (95% CI, 0.75 to 0.82) for local RFS, 0.77 (95% CI, 0.75 to 0.78) for regional RFS, and 0.80 (95% CI, 0.77 to 0.83) for distant RFS were obtained for the development model. External validation showed C-statistics of 0.80 (95% CI, 0.69 to 0.90), 0.76 (95% CI, 0.70 to 0.82), and 0.74 (95% CI, 0.69 to 0.80), respectively. Calibration plots showed a good match between predicted and observed rates. Using the nomogram, the C-statistic was increased by 9%-12% for the development cohort and by 11%-15% for the validation cohort, compared with a model including only T-stage and sentinel node status.

CONCLUSION

Most patients with thin melanomas have an excellent prognosis, but some develop recurrence. The presented nomograms can accurately identify a subgroup at high risk. An online calculator is available at www.melanomarisk.org.au.

The discriminative ability of Prostate Health Index to detect prostate cancer is enhanced in combination with miR-222-3p

BACKGROUND

There is an urgent need for better prostate cancer (PCa) biomarkers due to the low specificity of prostate specific antigen (PSA).

OBJECTIVE

Prostate Health Index (PHI) is an advanced PSA-based test for early detection of PCa. The present study aim was to investigate the potential improvement of diagnostic accuracy of PHI by its combination with suitable discriminative microRNAs (miRNAs).

METHODS

A two-phase study was performed. In a discovery phase, a panel of 177 miRNAs was measured in ten men with biopsy proven PCa and ten men with histologically no evidence of malignancy (NEM). These results were validated in a second phase including 25 patients in each group. The patients of all groups were matched regarding their PSA values and PHI were measured.

RESULTS

Based on data in the discovery phase, four elevated miRNAs were selected as potential miRNA candidates for further validation. A combination of miR-222-3p as the best discriminative miRNA with PHI extended the diagnostic accuracy of PHI from an AUC value of 0.690 to 0.787 and resulted in a sensitivity of 72.0% and a specificity of 84.0%.

CONCLUSION

Circulating microRNAs show useful diagnostic potential in combination with common used biomarkers to enhance their diagnostic power.

A systematic review of the status and methodological considerations for estimating risk of first ever stroke in the general population

AIMS

The methodological quality of development, validation, and modification of those models have not been evaluated via a thoroughly literature review. This study aims to describe the overall status and evaluate the methodological quality of risk prediction models for stroke incidence in the general population.

METHODS

We searched the database of EMBASE and MEDLINE by the combination of subject words and key words to collect the research on stroke risk prediction model in the general population. The retrieval time was from the establishment of the database to September 2019. It should be mentioned that risk of bias for each model was assessed, and data on population characteristics and model performance was also extracted.

RESULTS

The search screened 11,386 peer-reviewed publications and 57 citation searching, of which 48 were included in the review, describing the development of 51 prediction models, 47 external validation models, and 12 modification models. Among 51 development models, the predicted outcome concentrated on fatal or non-fatal stroke (n = 37, 73%). Thirty-nine development models (76%) were without internal validation. C-statistic or AUC was adopted for discrimination in 80% models, and Hosmer-Lemeshow test (n = 25, 49%) was also performed for calibration. Twenty-six development models (53%) were externally validated, among which only 2 (8%) were validated by independent researchers. Risk prediction performance was improved when models were modified by adding novel risk factors, such as the internal carotid artery plaque and intima-media thickness.

CONCLUSION

Models for predicting stroke occurrence need further external validation, recalibration, or modification in different populations, to help interpret those models in the practice of stroke prevention.

Toward Personalized Radiation Therapy of Liver Metastasis: Importance of Serial Blood Biomarkers

PURPOSE

To assess the added value of serial blood biomarkers in liver metastasis stereotactic body radiation therapy (SBRT).

MATERIALS AND METHODS

Eighty-nine patients were retrospectively included. Pre- and midtreatment blood samples were analyzed for potential biomarkers of the treatment response. Three biomarker classes were studied: gene mutation status, complete blood count, and inflammatory cytokine concentration in plasma. One-year local failure (LF) and 2-year overall survival (OS) were chosen as study end points. Multivariate logistic regression was used for response prediction. Added predictive benefit was assessed by quantifying the difference between the predictive performance of a baseline model (clinicopathologic and dosimetric predictors) and that of the biomarker-enhanced model, using three metrics: (1) likelihood ratio, (2) predictive variance, and (3) area under the receiver operating characteristic curve (AUC).

RESULTS

The most important predictors of LF were mutation in KRAS gene (hazard ratio [HR] = 2.92, 95% CI, [1.17 to 7.28], P = .02) and baseline and midtreatment concentration of plasma interleukin-6 (HR = 1.15 [1.04 to 1.26] and 1.06 [1.01 to 1.13], P = .01). Absolute lymphocyte count and platelet-to-lymphocyte ratio at baseline as well as neutrophil-to-lymphocyte ratio at baseline and before fraction 3 (HR = 1.33 [1.16 to 1.51] and 1.19 [1.09 to 1.30]) had the most significant association with OS (P = .0003). Addition of baseline GEN and inflammatory plasma cytokine biomarkers in predicting LF, respectively, increased AUC by 0.06 (from 0.73 to 0.79) and 0.07 (from 0.77 to 0.84). In predicting OS, inclusion of midtreatment complete blood count biomarkers increased AUC from 0.72 to 0.80, along with significant boosts in likelihood ratio and predictive variance.

CONCLUSION

Inclusion of serial blood biomarkers leads to significant gain in predicting response to liver metastasis stereotactic body radiation therapy and can guide treatment personalization.

Validation and verification of predictive salivary biomarkers for oral health

Oral health is important not only due to the diseases emerging in the oral cavity but also due to the direct relation to systemic health. Thus, early and accurate characterization of the oral health status is of utmost importance. There are several salivary biomarkers as candidates for gingivitis and periodontitis, which are major oral health threats, affecting the gums. These need to be verified and validated for their potential use as differentiators of health, gingivitis and periodontitis status, before they are translated to chair-side for diagnostics and personalized monitoring. We aimed to measure 10 candidates using high sensitivity ELISAs in a well-controlled cohort of 127 individuals from three groups: periodontitis (60), gingivitis (31) and healthy (36). The statistical approaches included univariate statistical tests, receiver operating characteristic curves (ROC) with the corresponding Area Under the Curve (AUC) and Classification and Regression Tree (CART) analysis. The main outcomes were that the combination of multiple biomarker assays, rather than the use of single ones, can offer a predictive accuracy of > 90% for gingivitis versus health groups; and 100% for periodontitis versus health and periodontitis versus gingivitis groups. Furthermore, ratios of biomarkers MMP-8, MMP-9 and TIMP-1 were also proven to be powerful differentiating values compared to the single biomarkers.

PHI density prospectively improves prostate cancer detection

Purpose

To evaluate the Prostate Health Index (PHI) density (PHID) in direct comparison with PHI in a prospective large cohort.

Methods

PHID values were calculated from prostate-specific antigen (PSA), free PSA and [− 2]proPSA and prostate volume. The 1057 patients included 552 men with prostate cancer (PCa) and 505 with no evidence of malignancy (NEM). In detail, 562 patients were biopsied at the Charité Hospital Berlin and 495 patients at the Sana Hospital Offenbach. All patients received systematic or magnetic resonance imaging (MRI)/ultrasound fusion-guided biopsies. The diagnostic accuracy was evaluated by receiver operating characteristic (ROC) curves comparing areas under the ROC-curves (AUC). The decision curve analysis (DCA) was performed with the MATLAB Neural Network Toolbox.

Results

PHID provided a significant larger AUC than PHI (0.835 vs. 0.801; p = 0.0013) in our prospective cohort of 1057 men from 2 centers. The DCA had a maximum net benefit of ~ 5% for PHID vs. PHI between 35 and 65% threshold probability. In those 698 men within the WHO-calibrated PSA grey-zone up to 8 ng/ml, PHID was also significantly better than PHI (AUC 0.819 vs. 0.789; p = 0.0219). But PHID was not different from PHI in the detection of significant PCa.

Conclusions

Based on ROC analysis and DCA, PHID had an advantage in comparison with PHI alone to detect any PCa but PHI and PHID performed equal in detecting significant PCa.

Validation of Risk Prediction Models to Inform Clinical Decisions After Acute Kidney Injury

Rationale & Objective

There is limited evidence to guide follow-up after acute kidney injury (AKI). Knowledge gaps include which patients to prioritize, at what time point, and for mitigation of which outcomes. In this study, we sought to compare the net benefit of risk model–based clinical decisions following AKI.

Study Design

External validation of 2 risk models of AKI outcomes: the Grampian -Aberdeen (United Kingdom) AKI readmissions model and the Alberta (Canada) kidney disease risk model of chronic kidney disease (CKD) glomerular (G) filtration rate categories 4 and 5 (CKD G4 and G5). Process mining to delineate existing care pathways.

Setting & Participants

Validation was based on data from adult hospital survivors of AKI from Grampian, 2011-2013.

Predictors

KDIGO-based measures of AKI severity and comorbidities specified in the original models.

Outcomes

Death or readmission within 90 days for all hospital survivors. Progression to new CKD G4-G5 for patients surviving at least 90 days after AKI.

Analytical Approach

Decision curve analysis to assess the “net benefit” of use of risk models to guide clinical care compared to alternative approaches (eg, prioritizing all AKI, severe AKI, or only those without kidney recovery).

Results

26,575 of 105,461 hospital survivors in Grampian (mean age, 60.9 ± 19.8 [SD] years) were included for validation of the death or readmission model, and 9,382 patients (mean age, 60.9 ± 19.8 years) for the CKD G4-G5 model. Both models discriminated well (area under the curve [AUC], 0.77 and 0.86, respectively). Decision curve analysis showed greater net benefit for follow up of all AKI than only severe AKI in most cases. Both original and refitted models provided net benefit superior to any other decision strategy. In process mining of all hospital discharges, 41% of readmissions and deaths occurred among people recovering after AKI. 1,464 of 3,776 people (39%) readmitted after AKI had received no intervening monitoring.

Limitations

Both original models overstated risks, indicating a need for regular updating.

Conclusions

Follow up after AKI has potential net benefit for preempting readmissions, death, and subsequent CKD progression. Decisions could be improved by using risk models and by focusing on AKI across a full spectrum of severity. The current lack of monitoring among many with poor outcomes indicates possible opportunities for implementation of decision support.

Is "pain before vomiting" useful?: Diagnostic performance of the classic patient history item in acute appendicitis

BACKGROUND

In most surgical textbooks, it has been stated that pain almost always precedes vomiting in patients with appendicitis. However, the usefulness of this classic history item, "pain before vomiting", has been investigated in only one study nearly 50 years ago, in which the cause of abdominal pain could not be identified in more than 40% of patients. Accordingly, our objective was to evaluate the performance of pain before vomiting for the diagnosis of acute appendicitis in patients who presented with both acute abdominal pain and vomiting.

METHODS

A retrospective chart review of adult outpatients with abdominal pain and vomiting at three acute care hospitals was performed. The reference standard for appendicitis was a CT scan evaluated by two radiologists. Diagnostic performance of pain before vomiting and the value it added to the Alvarado score were evaluated.

RESULTS

Among 310 patients, 24 patients were diagnosed with appendicitis. Diagnostic performance of pain before vomiting was a sensitivity of 95.8% (95% confidence interval [CI] 79.8-99.3) and a specificity of 16.6% (95% CI 12.6-21.4). When combined with the Alvarado score, it ruled out appendicitis in an additional 12% (increased from 32% to 44%) of patients without any false negatives.

CONCLUSIONS

"Pain before vomiting" is useful for ruling out appendicitis in patients with abdominal pain and vomiting.

Placental growth factor for the prognosis of women with preeclampsia (fullPIERS model extension): context matters

Background

The fullPIERS risk prediction model was developed to identify which women admitted with confirmed diagnosis of preeclampsia are at highest risk of developing serious maternal complications. The model discriminates well between women who develop (vs. those who do not) adverse maternal outcomes. It has been externally validated in several populations. We assessed whether placental growth factor (PlGF), a biomarker associated with preeclampsia risk, adds incremental value to the fullPIERS model.

Methods

Using a cohort of women admitted into tertiary hospitals in well-resourced settings (the USA and Canada), between May 2010 to February 2012, we evaluated the incremental value of PlGF added to fullPIERS for prediction of adverse maternal outcomes within 48 h after admission with confirmed preeclampsia. The discriminatory performance of PlGF and the fullPIERS model were assessed in this cohort using the area under the receiver’s operating characteristic curve (AUROC) while the extended model (fullPIERS +PlGF) was assessed based on net reclassification index (NRI) and integrated discrimination improvement (IDI) performances.

Results

In a cohort of 541 women delivered shortly (< 1 week) after presentation, 8.1% experienced an adverse maternal outcome within 48 h of admission. Prediction of adverse maternal outcomes was not improved by addition of PlGF to fullPIERS (NRI: -8.7, IDI − 0.06). Discriminatory performance (AUROC) was 0.67 [95%CI: 0.59–0.75] for fullPIERS only and 0.67 [95%CI: 0.58–0.76]) for fullPIERS extended with PlGF, a performance worse than previously documented in fullPIERS external validation studies (AUROC > 0.75).

Conclusions

While fullPIERS model performance may have been affected by differences in healthcare context between this study cohort and the model development and validation cohorts, future studies are required to confirm whether PlGF adds incremental benefit to the fullPIERS model for prediction of adverse maternal outcomes in preeclampsia in settings where expectant management is practiced.

Circular RNAs and Their Linear Transcripts as Diagnostic and Prognostic Tissue Biomarkers in Prostate Cancer after Prostatectomy in Combination with Clinicopathological Factors

As new biomarkers, circular RNAs (circRNAs) have been largely unexplored in prostate cancer (PCa). Using an integrative approach, we aimed to evaluate the potential of circRNAs and their linear transcripts (linRNAs) to act as (i) diagnostic biomarkers for differentiation between normal and tumor tissue and (ii) prognostic biomarkers for the prediction of biochemical recurrence (BCR) after radical prostatectomy. In a first step, eight circRNAs (circATXN10, circCRIM1, circCSNK1G3, circGUCY1A2, circLPP, circNEAT1, circRHOBTB3, and circSTIL) were identified as differentially expressed via a genome-wide circRNA-based microarray analysis of six PCa samples. Additional bioinformatics and literature data were applied for this selection process. In total, 115 malignant PCa and 79 adjacent normal tissue samples were examined using robust RT-qPCR assays specifically established for the circRNAs and their linear counterparts. Their diagnostic and prognostic potential was evaluated using receiver operating characteristic curves, Cox regressions, decision curve analyses, and C-statistic calculations of prognostic indices. The combination of circATXN10 and linSTIL showed a high discriminative ability between malignant and adjacent normal tissue PCa. The combination of linGUCY1A2, linNEAT1, and linSTIL proved to be the best predictive RNA-signature for BCR. The combination of this RNA signature with five established reference models based on only clinicopathological factors resulted in an improved predictive accuracy for BCR in these models. This is an encouraging study for PCa to evaluate circRNAs and their linRNAs in an integrative approach, and the results showed their clinical potential in combination with standard clinicopathological variables.

Development and validation of a prediction model for incident hand osteoarthritis in the HUNT study

OBJECTIVE

To develop and externally validate prediction models for incident hand osteoarthritis (OA) in a large population-based cohort of middle aged and older men and women.

DESIGN

We included 17,153 men and 18,682 women from a population-based cohort, aged 35-70 years at baseline (1995-1997). Incident hand OA were obtained from diagnostic codes in the Norwegian National Patient Register (1995-2018). We studied whether a range of self-reported and clinically measured predictors could predict hand OA, using the Area Under the receiver-operating Curve (AUC) from logistic regression. External validation of an existing prediction model for male hand OA was tested on discrimination in a sample of men. Bootstrapping was used to avoid overfitting.

RESULTS

The model for men showed modest discriminatory ability (AUC = 0.67, 95% CI 0.62-0.71). Adding a genetic risk score did not improve prediction. Similar discrimination was observed in the model for women (AUC = 0.62, 95% CI 0.59-0.64). Prediction was not improved by adding a genetic risk score or hormonal and reproductive factors. Applying external validation, similar results were observed among men in HUNT (The Nord-Trøndelag Health Study) as in the developmental sample (AUC = 0.62, 95% CI 0.57-0.65).

CONCLUSION

We developed prediction models for incident hand OA in men and women. For women, the model included body mass index (BMI), heavy physical work, high physical activity and perceived poor health. The model showed moderate discrimination. For men, we have shown that a prediction model including BMI, education and information on sleep can predict incident hand OA in several populations with moderate discriminative ability.

The Added Value of Lactate and Lactate Clearance in Prediction of In-Hospital Mortality in Critically Ill Patients With Sepsis

Supplemental Digital Content is available in the text.

We investigated the added predictive value of lactate and lactate clearance to the Acute Physiology and Chronic Health Evaluation IV model for predicting in-hospital mortality in critically ill patients with sepsis.

Validation of Random Forest Machine Learning Models to Predict Dementia-Related Neuropsychiatric Symptoms in Real-World Data

BACKGROUND

Neuropsychiatric symptoms (NPS) are the leading cause of the social burden of dementia but their role is underestimated.

OBJECTIVE

The objective of the study was to validate predictive models to separately identify psychotic and depressive symptoms in patients diagnosed with dementia using clinical databases representing the whole population to inform decision-makers.

METHODS

First, we searched the electronic health records of 4,003 patients with dementia to identify NPS. Second, machine learning (random forest) algorithms were applied to build separate predictive models for psychotic and depressive symptom clusters in the training set (N = 3,003). Third, calibration and discrimination were assessed in the test set (N = 1,000) to assess the performance of the models.

RESULTS

Neuropsychiatric symptoms were noted in the electronic health record of 58% of patients. The area under the receiver operating curve reached 0.80 for the psychotic cluster model and 0.74 for the depressive cluster model. The Kappa index and accuracy also showed better discrimination in the psychotic model. Calibration plots indicated that both types of model had less predictive accuracy when the probability of neuropsychiatric symptoms was <25%. The most important variables in the psychotic cluster model were use of risperidone, level of sedation, use of quetiapine and haloperidol and the number of antipsychotics prescribed. In the depressive cluster model, the most important variables were number of antidepressants prescribed, escitalopram use, level of sedation, and age.

CONCLUSION

Given their relatively good performance, the predictive models can be used to estimate prevalence of NPS in population databases.

Detection of Malignant Melanoma Using Artificial Intelligence: An Observational Study of Diagnostic Accuracy

BACKGROUND

Malignant melanoma can most successfully be cured when diagnosed at an early stage in the natural history. However, there is controversy over screening programs and many advocate screening only for high-risk individuals.

OBJECTIVES

This study aimed to evaluate the accuracy of an artificial intelligence neural network (Deep Ensemble for Recognition of Melanoma [DERM]) to identify malignant melanoma from dermoscopic images of pigmented skin lesions and to show how this compared to doctors' performance assessed by meta-analysis.

METHODS

DERM was trained and tested using 7,102 dermoscopic images of both histologically confirmed melanoma (24%) and benign pigmented lesions (76%). A meta-analysis was conducted of studies examining the accuracy of naked-eye examination, with or without dermoscopy, by specialist and general physicians whose clinical diagnosis was compared to histopathology. The meta-analysis was based on evaluation of 32,226 pigmented lesions including 3,277 histopathology-confirmed malignant melanoma cases. The receiver operating characteristic (ROC) curve was used to examine and compare the diagnostic accuracy.

RESULTS

DERM achieved a ROC area under the curve (AUC) of 0.93 (95% confidence interval: 0.92-0.94), and sensitivity and specificity of 85.0% and 85.3%, respectively. Avoidance of false-negative results is essential, so different decision thresholds were examined. At 95% sensitivity DERM achieved a specificity of 64.1% and at 95% specificity the sensitivity was 67%. The meta-analysis showed primary care physicians (10 studies) achieve an AUC of 0.83 (95% confidence interval: 0.79-0.86), with sensitivity and specificity of 79.9% and 70.9%; and dermatologists (92 studies) 0.91 (0.88-0.93), 87.5%, and 81.4%, respectively.

CONCLUSIONS

DERM has the potential to be used as a decision support tool in primary care, by providing dermatologist-grade recommendation on the likelihood of malignant melanoma.

Developing Biomarker Panels to Predict Progression of Acute Kidney Injury After Cardiac Surgery

Introduction

Acute kidney injury (AKI) is a frequent complication of cardiac surgery, but only a fraction of cardiac surgery patients that experience postoperative AKI have progression to more severe stages. Biomarkers that can distinguish patients that will experience progression of AKI are potentially useful for clinical care and/or the development of therapies.

Methods

Data come from a prospective cohort study of cardiac surgery patients; the analytic dataset contained data from 354 cardiac surgery patients meeting criteria for AKI following surgery. Candidate predictors were 38 biomarkers of kidney function, insult, or injury measured at the time of AKI diagnosis. The outcome was AKI progression, defined as worsening of AKI Network stage. We investigated combining biomarkers with Bayesian model averaging (BMA) and random forests of classification trees, with and without center transformation. For both approaches, we used resampling-based methods to avoid optimistic bias in our assessment of model performance.

Results

BMA yielded a combination of 3 biomarkers and an optimism-corrected estimated area under the receiver operating characteristic curve (AUC) of 0.75 (95% confidence interval [CI]: 0.68, 0.82). The random forests approach, which nominally uses all biomarkers, had an estimated AUC of 0.74 (95% CI: 0.66, 0.82). A second application of random forests applied to biomarker values after a center-specific transformation had an estimated AUC of 0.80 (95% CI: 0.72, 0.88).

Conclusion

These findings suggest that the application of advanced statistical techniques to combine biomarkers offers only modest improvements over use of single biomarkers alone. This exemplifies a common experience in biomarker research: combinations of modestly performing biomarkers often also have modest performance.

Determinants of diagnostic discordance for non-diabetic hyperglycaemia and Type 2 diabetes using paired glycated haemoglobin measurements in a large English primary care population: cross-sectional study

AIM

To investigate factors influencing diagnostic discordance for non-diabetic hyperglycaemia and Type 2 diabetes.

METHODS

Some 10 000 adults at increased risk of diabetes were screened with HbA_1c and fasting plasma glucose (FPG). The 2208 participants with initial HbA_1c ≥ 42 mmol/mol (≥ 6.0%) or FPG ≥ 6.1 mmol/l were retested after a median 40 days. We compared the first and second HbA_1c results, and consequent diagnoses of non-diabetic hyperglycaemia and Type 2 diabetes, and investigated predictors of discordant diagnoses.

RESULTS

Of 1463 participants with non-diabetic hyperglycaemia and 394 with Type 2 diabetes on first testing, 28.4% and 21.1% respectively had discordant diagnoses on repeated testing. Initial diagnosis of non-diabetic hyperglycaemia and/or impaired fasting glucose according to both HbA_1c and FPG criteria, or to FPG only, made reclassification as Type 2 diabetes more likely than initial classification according to HbA_1c alone. Initial diagnosis of Type 2 diabetes according to both HbA_1c and FPG criteria made reclassification much less likely than initial classification according to HbA_1c alone. Age, and anthropometric and biological measurements independently but inconsistently predicted discordant diagnoses and changes in HbA_1c .

CONCLUSIONS

Diagnosis of non-diabetic hyperglycaemia or Type 2 diabetes with a single measurement of HbA_1c in a screening programme for entry to diabetes prevention trials is unreliable. Diagnosis of non-diabetic hyperglycaemia and Type 2 diabetes should be confirmed by repeat testing. FPG results could help prioritise retesting. These findings do not apply to people classified as normal on a single test, who were not retested.

New horizons in arginine metabolism, ageing and chronic disease states

The elucidation of the metabolic pathways of the amino acid arginine and their role in health and disease have been an intensive focus of basic and clinical research for over a century. The recent advent of robust analytical techniques for biomarker assessment in large population cohorts has allowed the investigation of the pathophysiological role of specific arginine metabolites in key chronic disease states in old age, particularly those characterised by a reduced synthesis of endothelial nitric oxide, with consequent vascular disease and atherosclerosis. Two arginine metabolites have been increasingly studied in regard to their potential role in risk stratification and in the identification of novel therapeutic targets: the methylated arginine asymmetric dimethylarginine (ADMA) and the arginine analogue homoarginine. Higher circulating concentrations of ADMA, a potent inhibitor of nitric oxide synthesis, have been shown to predict adverse cardiovascular outcomes. By contrast, there is emerging evidence that homoarginine might exert cardioprotective effects. This review highlights recent advances in the biological and clinical role of ADMA and homoarginine in cardiovascular disease and other emerging fields, particularly chronic obstructive pulmonary disease, dementia, and depression. It also discusses opportunities for future research directions with the ultimate goal of translating knowledge of arginine metabolism, and its role in health and disease, into the clinical care of older adults.

A Novel Predictor Tool of Biochemical Recurrence after Radical Prostatectomy Based on a Five-MicroRNA Tissue Signature

Within five to ten years after radical prostatectomy (RP), approximately 15–34% of prostate cancer (PCa) patients experience biochemical recurrence (BCR), which is defined as recurrence of serum levels of prostate-specific antigen >0.2 µg/L, indicating probable cancer recurrence. Models using clinicopathological variables for predicting this risk for patients lack accuracy. There is hope that new molecular biomarkers, like microRNAs (miRNAs), could be potential candidates to improve risk prediction. Therefore, we evaluated the BCR prognostic capability of 20 miRNAs, which were selected by a systematic literature review. MiRNA expressions were measured in formalin-fixed, paraffin-embedded (FFPE) tissue RP samples of 206 PCa patients by RT-qPCR. Univariate and multivariate Cox regression analyses were performed, to assess the independent prognostic potential of miRNAs. Internal validation was performed, using bootstrapping and the split-sample method. Five miRNAs (miR-30c-5p/31-5p/141-3p/148a-3p/miR-221-3p) were finally validated as independent prognostic biomarkers. Their prognostic ability and accuracy were evaluated using C-statistics of the obtained prognostic indices in the Cox regression, time-dependent receiver-operating characteristics, and decision curve analyses. Models of miRNAs, combined with relevant clinicopathological factors, were built. The five-miRNA-panel outperformed clinically established BCR scoring systems, while their combination significantly improved predictive power, based on clinicopathological factors alone. We conclude that this miRNA-based-predictor panel will be worth to be including in future studies.

Circular RNAs in Clear Cell Renal Cell Carcinoma: Their Microarray-Based Identification, Analytical Validation, and Potential Use in a Clinico-Genomic Model to Improve Prognostic Accuracy

Circular RNAs (circRNAs) may act as novel cancer biomarkers. However, a genome-wide evaluation of circRNAs in clear cell renal cell carcinoma (ccRCC) has yet to be conducted. Therefore, the objective of this study was to identify and validate circRNAs in ccRCC tissue with a focus to evaluate their potential as prognostic biomarkers. A genome-wide identification of circRNAs in total RNA extracted from ccRCC tissue samples was performed using microarray analysis. Three relevant differentially expressed circRNAs were selected (circEGLN3, circNOX4, and circRHOBTB3), their circular nature was experimentally confirmed, and their expression-along with that of their linear counterparts-was measured in 99 malignant and 85 adjacent normal tissue samples using specifically established RT-qPCR assays. The capacity of circRNAs to discriminate between malignant and adjacent normal tissue samples and their prognostic potential (with the endpoints cancer-specific, recurrence-free, and overall survival) after surgery were estimated by C-statistics, Kaplan-Meier method, univariate and multivariate Cox regression analysis, decision curve analysis, and Akaike and Bayesian information criteria. CircEGLN3 discriminated malignant from normal tissue with 97% accuracy. We generated a prognostic for the three endpoints by multivariate Cox regression analysis that included circEGLN3, circRHOBT3 and linRHOBTB3. The predictive outcome accuracy of the clinical models based on clinicopathological factors was improved in combination with this circRNA-based signature. Bootstrapping as well as Akaike and Bayesian information criteria confirmed the statistical significance and robustness of the combined models. Limitations of this study include its retrospective nature and the lack of external validation. The study demonstrated the promising potential of circRNAs as diagnostic and particularly prognostic biomarkers in ccRCC patients.

Prediction of Low and Very Low Bone Mineral Density Among Adult Survivors of Childhood Cancer

PURPOSE

To develop and validate prediction models for low and very low bone mineral density (BMD) on the basis of clinical and treatment characteristics that identify adult survivors of childhood cancer who require screening by dual-energy x-ray absorptiometry.

PATIENTS AND METHODS

White survivors of childhood cancer (n = 2,032; median attained age, 29.3 years [range, 18.1 to 40.9 years]) enrolled in the St Jude Lifetime Cohort (SJLIFE; development) and survivors treated at the Erasmus Medical Center (validation) in the Netherlands (n = 403; median age, 24.2 years [range, 18.0 to 40.9 years]) were evaluated with dual-energy x-ray absorptiometry to determine lumbar spine BMD and total-body BMD. Low and very low BMD were defined as lumbar spine BMD and/or total-body BMD z scores of -1 or lower or -2 or lower, respectively. Multivariable logistic regression was used to build prediction models; performance was assessed using receiver operating characteristic curves. Diagnostic values were calculated at different probabilities.

RESULTS

Low BMD was present in 51% and 45% of SJLIFE and Dutch participants, respectively, and very low BMD was present in 20% and 10%, respectively. The model for low BMD included male sex (odds ratio [OR], 3.07), height (OR, 0.95), weight (OR, 0.98), attained age (OR, 0.97), current smoking status (OR, 1.48), and cranial irradiation (OR, 2.11). Areas under the curve were 0.72 (95% CI, 0.70 to 0.75) in the SJLIFE cohort and 0.69 (95% CI, 0.64 to 0.75) in the Dutch cohort. The sum of the sensitivity (69.0%) and specificity (64.0%) was maximal at the predicted probability of 50%. The model for very low BMD included male sex (OR, 3.28), height (OR, 0.95), weight (OR, 0.97), attained age (OR, 0.98), cranial irradiation (OR, 2.07), and abdominal irradiation (OR, 1.61), yielding areas under the curve of 0.76 (95% CI, 0.73 to 0.78; SJLIFE cohort) and 0.75 (95% CI, 0.67 to 0.83; Dutch cohort).

CONCLUSION

Validated prediction models for low and very low BMD, using easily measured patient and treatment characteristics, correctly identified BMD status in most white adult survivors through age 40 years.