Performance evaluation of a new prognostic-efficacy-combination design in the context of telemedical interventions.

AIMS: Telemedical interventions in heart failure patients intend to avoid unfavourable, indication-related events by an early, individualized care, which reacts to the current patients need. However, telemedical support is an expensive intervention, and usually only patients with high risk for unfavourable follow-up events will be able to profit from it. Möckel et al. therefore adapted a new design which we call 'prognostic-efficacy-combination design'. This design allows to define a biomarker cut-off and to perform a randomized controlled trial (RCT) in a biomarker-selected population within a single study. However, so far, it has not been evaluated if this double use of the control group for biomarker cut-off definition and efficacy assessment within the RCT leads to a bias in treatment effect estimation. In this methodological research work, we therefore want to evaluate whether the 'prognostic-efficacy-combination design' leads to biased treatment effect estimates and also compare it to alternative designs. If there is a bias, we further want to analyse its magnitude under different parameter settings. METHODS: We perform a systematic Monte Carlo simulation study to investigate among others potential bias, root mean square error and sensitivity, and specificity as well as the total treatment effect estimate in various realistic trial scenarios that mimic and vary the true data characteristics of the published TIM-HF2 Trial. In particular, we vary the event proportion, the sample size, the biomarker distribution, and the lower bound for the sensitivity. RESULTS: The results show that indeed the proposed design leads to some bias in the effect estimators, indicating an overestimation of the effect. However, this bias is relatively small in most scenarios. CONCLUSIONS: The 'prognostic-efficacy-combination design' can generally be recommended for clinical applications due to its efficiency compared to two separate trials. We recommend a sufficiently large sample size depending on the trial scenario. Our simulation code can be adapted to explore suitable sample sizes for other settings.

Development and validation of a decision support tool for the diagnosis of acute heart failure: systematic review, meta-analysis, and modelling study.

OBJECTIVES: To evaluate the diagnostic performance of N-terminal pro-B-type natriuretic peptide (NT-proBNP) thresholds for acute heart failure and to develop and validate a decision support tool that combines NT-proBNP concentrations with clinical characteristics. DESIGN: Individual patient level data meta-analysis and modelling study. SETTING: Fourteen studies from 13 countries, including randomised controlled trials and prospective observational studies. PARTICIPANTS: Individual patient level data for 10 369 patients with suspected acute heart failure were pooled for the meta-analysis to evaluate NT-proBNP thresholds. A decision support tool (Collaboration for the Diagnosis and Evaluation of Heart Failure (CoDE-HF)) that combines NT-proBNP with clinical variables to report the probability of acute heart failure for an individual patient was developed and validated. MAIN OUTCOME MEASURE: Adjudicated diagnosis of acute heart failure. RESULTS: Overall, 43.9% (4549/10 369) of patients had an adjudicated diagnosis of acute heart failure (73.3% (2286/3119) and 29.0% (1802/6208) in those with and without previous heart failure, respectively). The negative predictive value of the guideline recommended rule-out threshold of 300 pg/mL was 94.6% (95% confidence interval 91.9% to 96.4%); despite use of age specific rule-in thresholds, the positive predictive value varied at 61.0% (55.3% to 66.4%), 73.5% (62.3% to 82.3%), and 80.2% (70.9% to 87.1%), in patients aged <50 years, 50-75 years, and >75 years, respectively. Performance varied in most subgroups, particularly patients with obesity, renal impairment, or previous heart failure. CoDE-HF was well calibrated, with excellent discrimination in patients with and without previous heart failure (area under the receiver operator curve 0.846 (0.830 to 0.862) and 0.925 (0.919 to 0.932) and Brier scores of 0.130 and 0.099, respectively). In patients without previous heart failure, the diagnostic performance was consistent across all subgroups, with 40.3% (2502/6208) identified at low probability (negative predictive value of 98.6%, 97.8% to 99.1%) and 28.0% (1737/6208) at high probability (positive predictive value of 75.0%, 65.7% to 82.5%) of having acute heart failure. CONCLUSIONS: In an international, collaborative evaluation of the diagnostic performance of NT-proBNP, guideline recommended thresholds to diagnose acute heart failure varied substantially in important patient subgroups. The CoDE-HF decision support tool incorporating NT-proBNP as a continuous measure and other clinical variables provides a more consistent, accurate, and individualised approach. STUDY REGISTRATION: PROSPERO CRD42019159407.

Improve Management of acute heart failure with ProcAlCiTonin in EUrope: results of the randomized clinical trial IMPACT EU Biomarkers in Cardiology (BIC) 18.

AIM: To determine whether initiation of antibiotic therapy (ABX) by procalcitonin (PCT) within 8 h of admission in patients presenting to the emergency department with symptoms and signs of acute heart failure (AHF) and elevated natriuretic peptides would improve clinical outcomes. METHODS AND RESULTS: The study was a randomized multicentre clinical trial conducted at 16 sites in Europe. Patients were randomized to either a PCT-guided strategy or standard care. Patients with PCT-guided strategy (n = 370) had ABX initiated if PCT was > 0.2 µg/L. Patients with standard care (n = 372) had AHF care in accordance with published guidelines without PCT. The primary endpoint was 90-day all-cause mortality. Pre-specified secondary endpoints included 30-day all-cause mortality and readmission and rate of pneumonia. The Data Safety and Review Committee recommended stopping the study for futility when 762 of the planned 792 patients had been enrolled. A total of 742 patients could be analysed. Patients were elderly (median age: 77 years), 38% were women, and had typical signs and symptoms of AHF. All-cause mortality at 90 days was 10.3% in the PCT-guided group vs. 8.2% in standard care (P = 0.316). Thirty-day readmission was significantly higher in the PCT-guided group vs. standard care but the difference vanished until day 90. The rate of pneumonia was overall low (7.5%) and not different between groups. CONCLUSIONS: In patients with AHF, a strategy of PCT-guided initiation of ABX was not more effective than a standard care strategy in improving clinical outcomes.

Multicentre cross-sectional observational registry to monitor the safety of early discharge after rule-out of acute myocardial infarction by copeptin and troponin: the Pro-Core registry.

OBJECTIVES: There is sparse information on the safety of early primary discharge from the emergency department (ED) after rule-out of myocardial infarction in suspected acute coronary syndrome (ACS). This prospective registry aimed to confirm randomised study results in patients at low-to-intermediate risk, with a broader spectrum of symptoms, across different institutional standards and with a range of local troponin assays including high-sensitivity cTn (hs-cTn), cardiac troponin (cTn) and point-of-care troponin (POC Tn). DESIGN: Prospective, multicentre European registry. SETTING: 18 emergency departments in nine European countries (Germany, Austria, Switzerland, France, Spain, UK, Turkey, Lithuania and Hungary) PARTICIPANTS: The final study cohort consisted of 2294 patients (57.2% males, median age 57 years) with suspected ACS. INTERVENTIONS: Using the new dual markers strategy, 1477 patients were eligible for direct discharge, which was realised in 974 (42.5%) of patients. MAIN OUTCOME MEASURES: The primary endpoint was all-cause mortality at 30 days. RESULTS: Compared with conventional workup after dual marker measurement, the median length of ED stay was 60 min shorter (228 min, 95% CI: 219 to 239 min vs 288 min, 95% CI: 279 to 300 min) in the primary dual marker strategy (DMS) discharge group. All-cause mortality was 0.1% (95% CI: 0% to 0.6%) in the primary DMS discharge group versus 1.1% (95% CI: 0.6% to 1.8%) in the conventional workup group after dual marker measurement. Conventional workup instead of discharge despite negative DMS biomarkers was observed in 503 patients (21.9%) and associated with higher prevalence of ACS (17.1% vs 0.9%, p<0.001), cardiac diagnoses (55.2% vs 23.5%, p<0.001) and risk factors (p<0.01), but with a similar all-cause mortality of 0.2% (95% CI: 0% to 1.1%) versus primary DMS discharge (p=0.64). CONCLUSIONS: Copeptin on top of cardiac troponin supports safe discharge in patients with chest pain or other symptoms suggestive of ACS under routine conditions with the use of a broad spectrum of local standard POC, conventional and high-sensitivity troponin assays. TRIAL REGISTRATION NUMBER: NCT02490969.

Biomarker guidance allows a more personalized allocation of patients for remote patient management in heart failure: results from the TIM-HF2 trial.

AIMS: The TIM-HF2 study showed less days lost due to unplanned cardiovascular hospitalization or all-cause death and improved survival in patients randomly assigned to remote patient management (RPM) instead of standard of care. METHODS AND RESULTS: This substudy explored whether the biomarkers mid-regional pro-adrenomedullin (MR-proADM) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) could be used to identify low-risk patients unlikely to benefit from RPM, thereby allowing more efficient allocation of the intervention. For 1538 patients of the trial (median age 73 years, interquartile range 64-78 years, 30% female), baseline biomarkers were used to select subpopulations recommended for RPM with various safety endpoints (100%, 98%, 95% sensitivity), and efficacy of RPM was assessed. Both biomarkers were strongly associated with events. The primary endpoint of lost days increased from 1.0% (1.4%) in the lowest to 17.3% (17.6%) in the highest quintile of NT-proBNP (MR-proADM). After combining biomarkers to identify patients recommended for RPM with 95% sensitivity, in the most efficient scenario (excluding 27% of patients; NT-proBNP < 413.7 pg/mL and MR-proADM < 0.75 nmol/L), the effect of RPM on patients was highly similar to the original trial (ratio of lost days: 0.78, hazard ratio for all-cause death: 0.68). Number needed to treat for all-cause death was lowered from 28 to 21. Rates of emergencies and telemedical efforts were significantly lower among patients not recommended for RPM. Biomarker guidance would have saved about 150 h effort/year per 100 patients of the eligible population. CONCLUSIONS: The combined use of MR-proADM and NT-proBNP may allow safe, more precise, effective and cost-saving allocation of patients with heart failure to RPM and warrants further prospective studies.

Adherence to a procalcitonin-guided antibiotic treatment protocol in patients with severe sepsis and septic shock.

BACKGROUND: In randomised controlled trials, procalcitonin (PCT)-guided antibiotic treatment has been proven to significantly reduce length of antibiotic therapy in intensive care unit (ICU) patients. However, concern was raised on low protocol adherence and high rates of overruling, and thus the value of PCT-guided treatment in real clinical life outside study conditions remains unclear. In this study, adherence to a PCT protocol to guide antibiotic treatment in patients with severe sepsis and septic shock was analysed. METHODS: From 2012 to 2014, surgical ICU patients with severe sepsis or septic shock were retrospectively screened for PCT measurement series appropriate to make treatment decisions on antibiotic therapy. We compared (1) patients with appropriate PCT measurement series to patients without appropriate series; (2) patients who reached the antibiotic stopping advice threshold (PCT < 0.5 ng/mL and/or decrease to 10% of peak level) to patients who did not reach a stopping advice threshold; and (3) patients who were treated adherently to the PCT protocol to non-adherently treated patients. The groups were compared in terms of antibiotic treatment duration, PCT kinetics, and other clinical outcomes. RESULTS: Of 81 patients with severe sepsis or septic shock, 14 were excluded due to treatment restriction or short course in the ICU. The final analysis was performed on 67 patients. Forty-two patients (62.7%) had appropriate PCT measurement series. In patients with appropriate PCT series, median initial PCT (p = 0.001) and peak PCT levels (p < 0.001) were significantly higher compared to those with non-appropriate series. In 26 patients with appropriate series, PCT levels reached an antibiotic stopping advice. In 8 of 26 patients with stopping advice, antibiotics were discontinued adherently to the PCT protocol (30.8%). Patients with adherently discontinued antibiotics had a shorter antibiotic treatment (7d [IQR 6-9] vs. 12d [IQR 9-16]; p = 0.002). No differences were seen in terms of other clinical outcomes. CONCLUSION: In patients with severe sepsis and septic shock, procalcitonin testing was irregular and adherence to a local PCT protocol was low in real clinical life. However, adherently treated patients had a shorter duration of antibiotic treatment without negative clinical outcomes. Procalcitonin peak values and kinetics had a clear impact on the regularity of PCT testing.

Early identification of acute heart failure at the time of presentation: do natriuretic peptides make the difference?

BACKGROUND: The early identification of patients with acute heart failure (AHF) is challenging as many other diseases lead to a clinical presentation with dyspnea. AIM: The aim of the study was to evaluate the impact of natriuretic peptides at common HF study cut-offs on the diagnosis of patients with dyspnea at admission. METHODS AND RESULTS: For this post hoc analysis, we analysed n = 726 European Union (EU) patients from the prospective BACH (Biomarkers in Acute Heart Failure) study. Cut-offs were 350 ng/L (BNP), 300 pmol/L [pro-atrial natriuretic peptide (proANP)], and 1800 ng/L (NT-proBNP). These cut-offs had equivalent 90 days' mortality in the EU cohort of BACH. We analysed the effect of selection using these cut-offs on the prevalence of the gold standard diagnoses made in the BACH study and the respective mortality. The prevalence of AHF is increased from 47.5 to 75.6% (NT-proBNP criteria) up to 79.7% (BNP criteria). With the use of the proANP criteria, 90 days' mortality of patients with AHF rose from 14 to 17% (P = 0.029). In the group with no-AHF diagnoses, mortality rose from 10 to 25% (P < 0.001). CONCLUSIONS: The prevalence of patients with the gold standard diagnoses of AHF among those presenting with dyspnea to the emergency department is significantly increased by the use of natriuretic peptides with common cut-offs used in prospective HF studies. Nevertheless, in the selected groups, patients with no AHF diagnosis have the highest mortality, and therefore, the addition of a natriuretic peptide alone is insufficient to start specific therapies.

Rationale and design of the IMPACT EU-trial: improve management of heart failure with procalcitonin biomarkers in cardiology (BIC)-18.

OBJECTIVES: To evaluate the effectiveness of procalcitonin (PCT)-guided antibiotic treatment compared to current treatment practice to reduce 90-day all-cause mortality in emergency patients with shortness of breath (SOB) and suspected acute heart failure (AHF). BACKGROUND: Concomitant AHF and lower respiratory tract (or other bacterial) infection in emergency patients with dyspnea are common and can be difficult to diagnose. Early and adequate initiation of antibiotic therapy (ABX) significantly improves patient outcome, but superfluous prescription of ABX maybe harmful. METHODS: In a multicentre, prospective, randomized, controlled process trial with an open intervention, adult emergency patients with SOB and increased levels of natriuretic peptides will be randomized to either a standard care group or a PCT-guided group with respect to the initiation of antibiotic treatment. In the PCT-guided group, the initiation of antibiotic therapy is based on the results of acute PCT measurements at admission, using a cut-off of 0.2 ng/ml. A two-stage sample-size adaptive design is used; an interim analysis was done after completion of 50% of patients and the final sample size remained unchanged. Primary endpoint is 90-day all-cause mortality. CONCLUSIONS: The current study will provide evidence, whether the routine use of PCT in patients with suspected AHF improves outcome.

The Randomized Controlled STRAWINSKI Trial: Procalcitonin-Guided Antibiotic Therapy after Stroke.

BACKGROUND: Pneumonia is among the most common acute complications after stroke and is associated with poor long-term outcome. Biomarkers may help identifying stroke patients at high risk for developing stroke-associated pneumonia (SAP) and to guide early treatment. AIMS: This trial investigated whether procalcitonin (PCT) ultrasensitive (PCTus)-guided antibiotic treatment of SAP can improve functional outcome after stroke. METHODS: In this international, multicenter, randomized, controlled clinical trial with blinded assessment of outcomes, patients with severe ischemic stroke in the middle cerebral artery territory were randomly assigned within 40 h after symptom onset to PCTus-based antibiotic therapy guidance in addition to stroke unit care or standard stroke unit care alone. The primary endpoint was functional outcome at 3 months, defined according to the modified Rankin Scale (mRS) and dichotomized as acceptable (≤4) or unacceptable (≥5). Secondary endpoints included usage of antibiotics, infection rates, days of fever, and mortality. The trial was registered with http://ClinicalTrials.gov (Identifier NCT01264549). RESULTS: In the intention-to-treat-analysis based on 227 patients (112 in PCT and 115 in control group), 197 patients completed the 3-month follow-up. Adherence to PCT guidance was 65%. PCT-guided therapy did not improve functional outcome as measured by mRS (odds ratio 0.79; 95% confidence interval 0.45-1.35, p = 0.47). Pneumonia rate and mortality were similar in both groups. Days with fever tended to be lower (p = 0.055), whereas total number of days treated with antibiotics were higher (p = 0.004) in PCT compared to control group. A post hoc analysis including all PCT values in the intention-to-treat population demonstrated a significant increase on the first day of infection in patients with pneumonia and sepsis compared to patients with urinary tract infections or without infections (p < 0.0001). CONCLUSION: PCTus-guided antibiotic therapy did not improve functional outcome at 3 months after severe ischemic stroke. PCT is a promising biomarker for early detection of pneumonia and sepsis in acute stroke patients.

Serial Procalcitonin Predicts Mortality in Severe Sepsis Patients: Results From the Multicenter Procalcitonin MOnitoring SEpsis (MOSES) Study.

OBJECTIVES: To prospectively validate that the inability to decrease procalcitonin levels by more than 80% between baseline and day 4 is associated with increased 28-day all-cause mortality in a large sepsis patient population recruited across the United States. DESIGN: Blinded, prospective multicenter observational clinical trial following an Food and Drug Administration-approved protocol. SETTING: Thirteen U.S.-based emergency departments and ICUs. PATIENTS: Consecutive patients meeting criteria for severe sepsis or septic shock who were admitted to the ICU from the emergency department, other wards, or directly from out of hospital were included. INTERVENTIONS: Procalcitonin was measured daily over the first 5 days. MEASUREMENTS AND MAIN RESULTS: The primary analysis of interest was the relationship between a procalcitonin decrease of more than 80% from baseline to day 4 and 28-day mortality using Cox proportional hazards regression. Among 858 enrolled patients, 646 patients were alive and in the hospital on day 4 and included in the main intention-to-diagnose analysis. The 28-day all-cause mortality was two-fold higher when procalcitonin did not show a decrease of more than 80% from baseline to day 4 (20% vs 10%; p = 0.001). This was confirmed as an independent predictor in Cox regression analysis (hazard ratio, 1.97 [95% CI, 1.18-3.30; p < 0.009]) after adjusting for demographics, Acute Physiology and Chronic Health Evaluation II, ICU residence on day 4, sepsis syndrome severity, antibiotic administration time, and other relevant confounders. CONCLUSIONS: Results of this large, prospective multicenter U.S. study indicate that inability to decrease procalcitonin by more than 80% is a significant independent predictor of mortality and may aid in sepsis care.

Metabolite profiling identifies candidate markers reflecting the clinical adaptations associated with Roux-en-Y gastric bypass surgery.

BACKGROUND: Roux-en-Y gastric bypass (RYGB) surgery is associated with weight loss, improved insulin sensitivity and glucose homeostasis, and a reduction in co-morbidities such as diabetes and coronary heart disease. To generate further insight into the numerous metabolic adaptations associated with RYGB surgery, we profiled serum metabolites before and after gastric bypass surgery and integrated metabolite changes with clinical data. METHODOLOGY AND PRINCIPAL FINDINGS: Serum metabolites were detected by gas and liquid chromatography-coupled mass spectrometry before, and 3 and 6 months after RYGB in morbidly obese female subjects (n = 14; BMI = 46.2+/-1.7). Subjects showed decreases in weight-related parameters and improvements in insulin sensitivity post surgery. The abundance of 48% (83 of 172) of the measured metabolites changed significantly within the first 3 months post RYGB (p<0.05), including sphingosines, unsaturated fatty acids, and branched chain amino acids. Dividing subjects into obese (n = 9) and obese/diabetic (n = 5) groups identified 8 metabolites that differed consistently at all time points and whose serum levels changed following RYGB: asparagine, lysophosphatidylcholine (C18:2), nervonic (C24:1) acid, p-Cresol sulfate, lactate, lycopene, glucose, and mannose. Changes in the aforementioned metabolites were integrated with clinical data for body mass index (BMI) and estimates for insulin resistance (HOMA-IR). Of these, nervonic acid was significantly and negatively correlated with HOMA-IR (p = 0.001, R = -0.55). CONCLUSIONS: Global metabolite profiling in morbidly obese subjects after RYGB has provided new information regarding the considerable metabolic alterations associated with this surgical procedure. Integrating clinical measurements with metabolomics data is capable of identifying markers that reflect the metabolic adaptations following RYGB.

Identification of gastric cancer patients by serum protein profiling.

Using surface-enhanced laser desorption ionization mass spectrometry (SELDI/TOF-MS) and ProteinChip technology, coupled with a pattern-matching algorithm and serum samples, we screened for protein patterns to differentiate gastric cancer patients from noncancer patients. A classifier ensemble, consisting of 50 decision trees, correctly classified all gastric cancers and all controls of a training set (100% sensitivity and 100% specificity). Eight of 9 stage I gastric cancers (88.9% sensitivity for stage I) were correctly classified. In addition, 28 sera from gastric cancer patients taken in different hospitals were correctly classified (100% sensitivity). Furthermore, all 11 control sera obtained from patients without gastric cancer (100% specificity) were classified correctly and 29 of 30 healthy blood-donors were classified as noncancerous. ProteinChip technology in conjunction with bioinformatics allows the highly sensitive and specific recognition of gastric cancer patients.

Bioinformatics in proteomics: application, terminology, and pitfalls.

Bioinformatics applies data mining, i.e., modern computer-based statistics, to biomedical data. It leverages on machine learning approaches, such as artificial neural networks, decision trees and clustering algorithms, and is ideally suited for handling huge data amounts. In this article, we review the analysis of mass spectrometry data in proteomics, starting with common pre-processing steps and using single decision trees and decision tree ensembles for classification. Special emphasis is put on the pitfall of overfitting, i.e., of generating too complex single decision trees. Finally, we discuss the pros and cons of the two different decision tree usages.

The time-organized map algorithm: extending the self-organizing map to spatiotemporal signals.

The new time-organized map (TOM) is presented for a better understanding of the self-organization and geometric structure of cortical signal representations. The algorithm extends the common self-organizing map (SOM) from the processing of purely spatial signals to the processing of spatiotemporal signals. The main additional idea of the TOM compared with the SOM is the functionally reasonable transfer of temporal signal distances into spatial signal distances in topographic neural representations. This is achieved by neural dynamics of propagating waves, allowing current and former signals to interact spatiotemporally in the neural network. Within a biologically plausible framework, the TOM algorithm (1) reveals how dynamic neural networks can self-organize to embed spatial signals in temporal context in order to realize functional meaningful invariances, (2) predicts time-organized representational structures in cortical areas representing signals with systematic temporal relation, and (3) suggests that the strength with which signals interact in the cortex determines the type of signal topology realized in topographic maps (e.g., spatially or temporally defined signal topology). Moreover, the TOM algorithm supports the explanation of topographic reorganizations based on time-to-space transformations (Wiemer, Spengler, Joublin, Stagge, & Wacquant, 2000).