A modified Wald interval for the area under the ROC curve (AUC) in diagnostic case-control studies

Systematic review and meta-analysis of the prognostic value of 18F-Fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) and/or computed tomography (CT)-based radiomics in head and neck cancer

AIM

Radiomics involves the extraction of quantitative data from medical images to facilitate the diagnosis, prognosis, and staging of tumors. This study provides a comprehensive overview of the efficacy of radiomics in prognostic applications for head and neck cancer (HNC) in recent years. It undertakes a systematic review of prognostic models specific to HNC and conducts a meta-analysis to evaluate their predictive performance.

MATERIALS AND METHODS

This study adhered rigorously to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for literature searches. The literature databases, including PubMed, Embase, Cochrane, and Scopus were systematically searched individually. The methodological quality of the incorporated studies underwent assessment utilizing the radiomics quality score (RQS) tool. A random-effects meta-analysis employing the Harrell concordance index (C-index) was conducted to evaluate the performance of all radiomics models.

RESULTS

Among the 388 studies retrieved, 24 studies encompassing a total of 6,978 cases were incorporated into the systematic review. Furthermore, eight studies, focusing on overall survival as an endpoint, were included in the meta-analysis. The meta-analysis revealed that the estimated random effect of the C-index for all studies utilizing radiomics alone was 0.77 (0.71-0.82), with a substantial degree of heterogeneity indicated by an I2 of 80.17%.

CONCLUSIONS

Based on this review, prognostic modeling utilizing radiomics has demonstrated enhanced efficacy for head and neck cancers; however, there remains room for improvement in this approach. In the future, advancements are warranted in the integration of clinical parameters and multimodal features, balancing multicenter data, as well as in feature screening and model construction within this field.

A diagnostic phase III/IV seamless design to investigate the diagnostic accuracy and clinical effectiveness using the example of HEDOS and HEDOS II

The development process of medical devices can be streamlined by combining different study phases. Here, for a diagnostic medical device, we present the combination of confirmation of diagnostic accuracy (phase III) and evaluation of clinical effectiveness regarding patient-relevant endpoints (phase IV) using a seamless design. This approach is used in the Thyroid HEmorrhage DetectOr Study (HEDOS & HEDOS II) investigating a post-operative hemorrhage detector named ISAR-M THYRO® in patients after thyroid surgery. Data from the phase III trial are reused as external controls in the control group of the phase IV trial. An unblinded interim analysis is planned between the two study stages which includes a recalculation of the sample size for the phase IV part after completion of the first stage of the seamless design. The study concept presented here is the first seamless design proposed in the field of diagnostic studies. Hence, the aim of this work is to emphasize the statistical methodology as well as feasibility of the proposed design in relation to the planning and implementation of the seamless design. Seamless designs can accelerate the overall trial duration and increase its efficiency in terms of sample size and recruitment. However, careful planning addressing numerous methodological and procedural challenges is necessary for successful implementation as well as agreement with regulatory bodies.

Development and validation of a multimodal neuroimaging biomarker for electroconvulsive therapy outcome in depression: a multicenter machine learning analysis

BACKGROUND

Electroconvulsive therapy (ECT) is the most effective intervention for patients with treatment resistant depression. A clinical decision support tool could guide patient selection to improve the overall response rate and avoid ineffective treatments with adverse effects. Initial small-scale, monocenter studies indicate that both structural magnetic resonance imaging (sMRI) and functional MRI (fMRI) biomarkers may predict ECT outcome, but it is not known whether those results can generalize to data from other centers. The objective of this study was to develop and validate neuroimaging biomarkers for ECT outcome in a multicenter setting.

METHODS

Multimodal data (i.e. clinical, sMRI and resting-state fMRI) were collected from seven centers of the Global ECT-MRI Research Collaboration (GEMRIC). We used data from 189 depressed patients to evaluate which data modalities or combinations thereof could provide the best predictions for treatment remission (HAM-D score ⩽7) using a support vector machine classifier.

RESULTS

Remission classification using a combination of gray matter volume and functional connectivity led to good performing models with average 0.82-0.83 area under the curve (AUC) when trained and tested on samples coming from the three largest centers (N = 109), and remained acceptable when validated using leave-one-site-out cross-validation (0.70-0.73 AUC).

CONCLUSIONS

These results show that multimodal neuroimaging data can be used to predict remission with ECT for individual patients across different treatment centers, despite significant variability in clinical characteristics across centers. Future development of a clinical decision support tool applying these biomarkers may be feasible.

New use of preoperative fibrinogen in ovarian cancer management

Background

Ovarian cancer (OC) is often diagnosed at an advanced stage due to the absence of specific symptoms in its early stages. And the prognosis greatly depends on when the disease is diagnosed. Thus, we conducted to evaluate the value of preoperative fibrinogen (Fib) levels for the diagnosis of OC in the hope of improving its diagnostic efficiency.

Methods

A total of 126 ovarian tumor patients were retrospectively included in this study. Four candidate OC markers, including cancer antigen 125 (CA125), Fib, platelet (PLT) and homocysteine (Hcy) were employed to establish a diagnosis model for OC. The diagnostic performance of the model was evaluated using the area under the receiver operating characteristic curve (AUC) and Youden index.

Results

All included markers could be used for the diagnosis of OC. The AUCs of CA125, Fib, PLT and Hcy were 0.881, 0.825, 0.676 and 0.647, respectively. The new diagnosis model combining CA125 and Fib (CA125-Fib) had a higher AUC (0.924), Youden index (0.730), and best sensitivity (SN) (74.6%) and specificity (SP) (98.41%). CA125-Fib also had a high value in the diagnosis of stage I-II OC (AUC, Youden index, SN and SP: 0.853, 0.624, 81.48% and 80.95%).

Conclusions

Fib could be used for OC diagnosis. In particular, the combination of Fib and CA125 could further improve the diagnostic efficiency. And the diagnostic value of PLT and Hcy was found to be poor.

Hospital length of stay prediction tools for all hospital admissions and general medicine populations: systematic review and meta-analysis

Background

Unwarranted extended length of stay (LOS) increases the risk of hospital-acquired complications, morbidity, and all-cause mortality and needs to be recognized and addressed proactively.

Objective

This systematic review aimed to identify validated prediction variables and methods used in tools that predict the risk of prolonged LOS in all hospital admissions and specifically General Medicine (GenMed) admissions.

Method

LOS prediction tools published since 2010 were identified in five major research databases. The main outcomes were model performance metrics, prediction variables, and level of validation. Meta-analysis was completed for validated models. The risk of bias was assessed using the PROBAST checklist.

Results

Overall, 25 all admission studies and 14 GenMed studies were identified. Statistical and machine learning methods were used almost equally in both groups. Calibration metrics were reported infrequently, with only 2 of 39 studies performing external validation. Meta-analysis of all admissions validation studies revealed a 95% prediction interval for theta of 0.596 to 0.798 for the area under the curve. Important predictor categories were co-morbidity diagnoses and illness severity risk scores, demographics, and admission characteristics. Overall study quality was deemed low due to poor data processing and analysis reporting.

Conclusion

To the best of our knowledge, this is the first systematic review assessing the quality of risk prediction models for hospital LOS in GenMed and all admissions groups. Notably, both machine learning and statistical modeling demonstrated good predictive performance, but models were infrequently externally validated and had poor overall study quality. Moving forward, a focus on quality methods by the adoption of existing guidelines and external validation is needed before clinical application.

Systematic review registration

https://www.crd.york.ac.uk/PROSPERO/, identifier: CRD42021272198.

Classification of breast microcalcifications with GaAs photon-counting spectral mammography using an inverse problem approach

The purpose of this study was to investigate the use of a Gallium Arsenide (GaAs) photon-counting spectral mammography system to differentiate between Type I and Type II calcifications. Type I calcifications, consisting of calcium oxalate dihydrate (CO) or weddellite compounds are more often associated with benign lesions in the breast, and Type II calcifications containing hydroxyapatite (HA) are associated with both benign and malignant lesions in the breast. To be able to differentiate between these two calcification types, it is necessary to be able to estimate the full spectrum of the x-ray beam transmitted through the breast. We propose a novel method for estimating the energy-dependent x-ray transmission fraction of a beam using a photon counting detector with a limited number of energy bins. Using the estimated x-ray transmission through microcalcifications, it was observed that calcification type can be accurately estimated with machine learning. The study was carried out on a custom-built laboratory benchtop system using the SANTIS 0804 GaAs detector prototype system from DECTRIS Ltd with two energy thresholds enabled. Four energy thresholds detector was simulated by taking two separate acquisitions in which two energy thresholds were enabled for each acquisition and set at (12 keV, 21 keV) and then (29 keV, 36 keV). Measurements were performed using BR3D (CIRS, Norfolk, VA) breast imaging phantoms mimicking 100% adipose and 100% glandular tissues swirled together in an approximate 50/50 ratio by weight with the addition of in-house-developed synthetic microcalcifications. First, an inverse problem-based approach was used to estimate the full energy x-ray transmission fraction factor using known basis transmission factors from varying thicknesses of aluminum and polymethyl methacrylate (PMMA). Second, the classification of Type I and Type II calcifications was performed using the estimated energy-dependent transmission fraction factors for the pixels containing calcifications. The results were analyzed using receiver operating characteristic (ROC) analysis and demonstrated good discrimination performance with the area under the ROC curve greater than 84%. They indicated that GaAs photon-counting spectral mammography has potential use as a non-invasive method for discrimination between Type I and Type II calcifications. Results from this study suggested that GaAs-based spectral mammography could serve as a non-invasive measure for ruling out malignancy of calcifications found in the breast. Additional studies in more clinically realistic conditions involving breast tissues samples with smaller microcalcification specks should be performed to further explore the feasibility of this approach.

Hospital length of stay prediction for general surgery and total knee arthroplasty admissions: Systematic review and meta-analysis of published prediction models

Objective

Systematic review of length of stay (LOS) prediction models to assess the study methods (including prediction variables), study quality, and performance of predictive models (using area under receiver operating curve (AUROC)) for general surgery populations and total knee arthroplasty (TKA).

Method

LOS prediction models published since 2010 were identified in five major research databases. The main outcomes were model performance metrics including AUROC, prediction variables, and level of validation. Risk of bias was assessed using the PROBAST checklist.

Results

Five general surgery studies (15 models) and 10 TKA studies (24 models) were identified. All general surgery and 20 TKA models used statistical approaches; 4 TKA models used machine learning approaches. Risk scores, diagnosis, and procedure types were predominant predictors used. Risk of bias was ranked as moderate in 3/15 and high in 12/15 studies. Discrimination measures were reported in 14/15 and calibration measures in 3/15 studies, with only 4/39 externally validated models (3 general surgery and 1 TKA). Meta-analysis of externally validated models (3 general surgery) suggested the AUROC 95% prediction interval is excellent and ranges between 0.803 and 0.970.

Conclusion

This is the first systematic review assessing quality of risk prediction models for prolonged LOS in general surgery and TKA groups. We showed that these risk prediction models were infrequently externally validated with poor study quality, typically related to poor reporting. Both machine learning and statistical modelling methods, plus the meta-analysis, showed acceptable to good predictive performance, which are encouraging. Moving forward, a focus on quality methods and external validation is needed before clinical application.

DCBLD1 Overexpression Is Associated With a Poor Prognosis in Head and Neck Squamous Cell Carcinoma

Background

DCBLD1 is highly expressed in several kinds of cancer and plays a potential prognostic factor. However, the prognostic value and immune infiltration in head and neck squamous cell carcinoma remain unclear and need further research.

Materials and Methods

DCBLD1 expression and clinical information were obtained from the Cancer Genome Atlas (TCGA) database. The mRNA level in cell lines (SCC25 and CAL27) and gingival fibroblasts were detected using quantitative PCR. Cox regression analysis was used to evaluate the prognostic values of DCBLD1 and clinical data in HNSCC. A nomogram was also established to predict the impact of DCBLD1 on prognosis based on Cox multivariate results. The methylation level of DCBLD1 in HNSC and its prognosis were analyzed in UALACN and MethSurv. Finally, the potential biological functions of DCBLD1 were investigated using gene set enrichment analysis (GSEA) and single-sample GSEA (ssGSEA).

Results

The mRNA and protein expression levels of DCBLD1 were highly expressed in HNSCC tissue and cell lines. The Cox analyses demonstrate that highly expressed DCBLD1 is an independent prognosis marker (p < 0.05). ROC curve analysis showed the performance of DCBLD1 (area under the ROC curve: 0.948, sensitivity: 93.2%, specificity: 84.7%). The methylation was increased in HNSCC patients compared with normal subjects (p < 0.05) and was associated with poor prognosis at sites cg27642470 and cg21104965. Additionally, DCBLD1 expression is poorly associated with immune cell infiltration and immunological checkpoints PD-L1 and TIM-3.

Conclusion

In head and neck squamous cell carcinoma, DCBLD1 is overexpressed, associated with poor patient prognosis. The detailed underlying mechanism merits further research.

Isolated Sphenoid Sinusitis: Anatomical Features for Choosing a Method of Treatment, a Case-Control Study

Isolated sphenoid sinusitis (ISS) is a group of pathologies characterized by inflammation in one or both sphenoid sinuses. The gold standard for analyzing and diagnosing ISS is computer tomography. Many researchers have discussed the treatment of patients with ISS variants such as fully opacified sinus, mostly with surgery. A retrospective analysis of clinical data of 59 patients (21 male (35%), 38 female (65%)) with ISS, who were treated in the Otorhinolaryngological Department of Pavlov First Saint Petersburg State Medical University between January 2017 and January 2020, was conducted. All patients were in the first stage of the same medical therapy. In cases where there was no recovery, these patients were referred to surgery. For the control group, we analyzed patients without any disorders according to CT-scan examination. After analyzing the obtained clinical and radiological data, we found indicators that were common in patients who did not recover after medical therapy. According to the reverse regression method statistical model, in male patients with a diffuse headache and nasal discharge it was shown that medical therapy was highly effective (more than 78%). The presence of nasal septum deviation and adenoids in male and female patients leads to the highest risk of surgical treatment (83% probability of the logistic model). The detailed analysis of CT-scans and the complaints of patients with ISS can be the key to determining the preferred therapy choice. Not all cases need to have an endoscopic opening of the sphenoid sinus, according to our research.

A Clinical Diagnostic Study: Fibulin-2 is a Novel Promising Biomarker for Predicting Infection

Introduction

Infection remains a major cause of morbidity and mortality in hospital. As uncontrolled early infection may develop into systemic infection and eventually progress to sepsis, it is important to address infection at an early stage. Furthermore, early detection and prompt diagnosis of infection are the basis of clinical intervention. However, as a result of the interference of complex aetiologies, including fever and trauma, problems regarding the sensitivity and specificity of current diagnostic indices remain, such as for C-reactive protein (CRP), procalcitonin (PCT), white blood cells (WBC), neutrophil ratio (NEU%), interleukin-6 (IL-6) and D-dimer. As a result, there is an urgent need to develop new biomarkers to diagnose infection.

Methods

From January to October 2021, consecutive patients in the emergency department (ED) were recruited to investigate the feasibility of fibulin-2 as a diagnostic indicator of early infection. Fibulin-2 concentrations in plasma were determined with enzyme-linked immunosorbent assay (ELISA). The performance of fibulin-2 for predicting infection was analysed by receiver operating characteristic (ROC) curves.

Results

We found that the plasma fibulin-2 level was elevated in patients with infection compared with those without infection. ROC curve analysis showed that the area under the curve (AUC) for fibulin-2 was 0.712. For all patients included, the diagnostic ability of fibulin-2 (AUC 0.712) performed as well as CRP (AUC 0.667) and PCT (AUC 0.632), and better than WBC (AUC 0.620), NEU% (AUC 0.619), IL-6 (AUC 0.561) and D-dimer (AUC 0.630). In patients with fever, fibulin-2 performed as well as PCT and better than the other biomarkers in infection diagnosis. In particular, fibulin-2 performed better than all these biomarkers in patients with trauma.

Conclusion

Fibulin-2 is a novel promising diagnostic biomarker for predicting infection.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40121-022-00622-y.

Risk prediction of diabetes and pre-diabetes based on physical examination data

Diabetes is a metabolic disorder caused by insufficient insulin secretion and insulin secretion disorders. From health to diabetes, there are generally three stages: health, pre-diabetes and type 2 diabetes. Early diagnosis of diabetes is the most effective way to prevent and control diabetes and its complications. In this work, we collected the physical examination data from Beijing Physical Examination Center from January 2006 to December 2017, and divided the population into three groups according to the WHO (1999) Diabetes Diagnostic Standards: normal fasting plasma glucose (NFG) (FPG < 6.1 mmol/L), mildly impaired fasting plasma glucose (IFG) (6.1 mmol/L ≤ FPG < 7.0 mmol/L) and type 2 diabetes (T2DM) (FPG > 7.0 mmol/L). Finally, we obtained1,221,598 NFG samples, 285,965 IFG samples and 387,076 T2DM samples, with a total of 15 physical examination indexes. Furthermore, taking eXtreme Gradient Boosting (XGBoost), random forest (RF), Logistic Regression (LR), and Fully connected neural network (FCN) as classifiers, four models were constructed to distinguish NFG, IFG and T2DM. The comparison results show that XGBoost has the best performance, with AUC (macro) of 0.7874 and AUC (micro) of 0.8633. In addition, based on the XGBoost classifier, three binary classification models were also established to discriminate NFG from IFG, NFG from T2DM, IFG from T2DM. On the independent dataset, the AUCs were 0.7808, 0.8687, 0.7067, respectively. Finally, we analyzed the importance of the features and identified the risk factors associated with diabetes.

A Modified AUC for Training Convolutional Neural Networks: Taking Confidence Into Account

Receiver operating characteristic (ROC) curve is an informative tool in binary classification and Area Under ROC Curve (AUC) is a popular metric for reporting performance of binary classifiers. In this paper, first we present a comprehensive review of ROC curve and AUC metric. Next, we propose a modified version of AUC that takes confidence of the model into account and at the same time, incorporates AUC into Binary Cross Entropy (BCE) loss used for training a Convolutional neural Network for classification tasks. We demonstrate this on three datasets: MNIST, prostate MRI, and brain MRI. Furthermore, we have published GenuineAI, a new python library, which provides the functions for conventional AUC and the proposed modified AUC along with metrics including sensitivity, specificity, recall, precision, and F1 for each point of the ROC curve.

Use and performance of machine learning models for type 2 diabetes prediction in clinical and community care settings: Protocol for a systematic review and meta-analysis of predictive modeling studies

Objective

Machine learning involves the use of algorithms without explicit instructions. Of late, machine learning models have been widely applied for the prediction of type 2 diabetes. However, no evidence synthesis of the performance of these prediction models of type 2 diabetes is available. We aim to identify machine learning prediction models for type 2 diabetes in clinical and community care settings and determine their predictive performance.

Methods

The systematic review of English language machine learning predictive modeling studies in 12 databases will be conducted. Studies predicting type 2 diabetes in predefined clinical or community settings are eligible. Standard CHARMS and TRIPOD guidelines will guide data extraction. Methodological quality will be assessed using a predefined risk of bias assessment tool. The extent of validation will be categorized by Reilly–Evans levels. Primary outcomes include model performance metrics of discrimination ability, calibration, and classification accuracy. Secondary outcomes include candidate predictors, algorithms used, level of validation, and intended use of models. The random-effects meta-analysis of c-indices will be performed to evaluate discrimination abilities. The c-indices will be pooled per prediction model, per model type, and per algorithm. Publication bias will be assessed through funnel plots and regression tests. Sensitivity analysis will be conducted to estimate the effects of study quality and missing data on primary outcome. The sources of heterogeneity will be assessed through meta-regression. Subgroup analyses will be performed for primary outcomes.

Ethics and dissemination

No ethics approval is required, as no primary or personal data are collected. Findings will be disseminated through scientific sessions and peer-reviewed journals.

PROSPERO registration number

CRD42019130886

Teaching Cognitive Behavior Therapy to Postgraduate Health Care Professionals in Times of COVID 19 - An Asynchronous Blended Learning Environment Proved to Be Non-inferior to In-Person Training

Training of postgraduate health professionals on their way to becoming licensed therapists for Cognitive Behavior Therapy (CBT) came to a halt in Germany in March 2020 when social distancing regulations came into effect. Since the German healthcare system almost exclusively relies on this profession when it comes to the implementation of CBT and 80% of those therapists active in 2010 will have retired at the end of 2030, it is critical to assess whether online CBT training is as satisfactory as classroom on-site CBT training. An asynchronous, blended, inverted-classroom online learning environment for CBT training (CBT for psychosis) was developed as an emergency solution. It consisted of pre-recorded CBT video lectures, exercises to train interventions in online role-plays, and regular web conferences. Training was provided at five different training institutes in Germany (duration 8–16 h). Postgraduate health care professionals (psychiatrists and psychologists) (n = 43) who received the online CBT training filled out standard self-report evaluations that assessed satisfaction and didactic quality. These evaluations were compared to those evaluations of students (n = 142) who had received in-person CBT training with identical content offered by the same CBT trainer at the same training institutes before the COVID-19 crisis. Both groups were comparable with respect to interest in the subject and prior knowledge. We tested non-inferiority hypotheses using Wilcoxon-Mann-Whitney ROC-curve analyses with an equivalence margin corresponding to a small-to-medium effect size (d = 0.35). The online training evaluations were non-inferior concerning information content, conception of content, didactic presentation, assessment of the trainer as a suitable role-model, working atmosphere, own commitment, and practical relevance. In contrast, we could not exclude a small effect in favor of in-person training in professional benefit and room for active participation. Our results suggest that delivering substantial CBT knowledge online to postgraduate health-professionals is sufficient, and at most incurs minimal loss to the learning experience. These encouraging findings indicate that integrating online elements in CBT teaching is an acceptable option even beyond social distancing requirements.

Effects of ergotamine on the central nervous system using untargeted metabolomics analysis in a mouse model

The ergot alkaloid ergotamine is produced by Claviceps purpurea, a parasitic fungus that commonly infects crops and pastures of high agricultural and economic importance. In humans and livestock, symptoms of ergotism include necrosis and gangrene, high blood pressure, heart rate, thermoregulatory dysfunction and hallucinations. However, ergotamine is also used in pharmaceutical applications to treat migraines and stop post-partum hemorrhage. To define its effects, metabolomic profiling of the brain was undertaken to determine pathways perturbed by ergotamine treatment. Metabolomic profiling identified the brainstem and cerebral cortex as regions with greatest variation. In the brainstem, dysregulation of the neurotransmitter epinephrine, and the psychoactive compound 2-arachidonylglycerol was identified. In the cerebral cortex, energy related metabolites isobutyryl-L-carnitine and S-3-oxodecanoyl cysteamine were affected and concentrations of adenylosuccinate, a metabolite associated with mental retardation, were higher. This study demonstrates, for the first time, key metabolomic pathways involved in the behavioural and physiological dysfunction of ergot alkaloid intoxicated animals.

Deep learning for predicting COVID-19 malignant progression

As COVID-19 is highly infectious, many patients can simultaneously flood into hospitals for diagnosis and treatment, which has greatly challenged public medical systems. Treatment priority is often determined by the symptom severity based on first assessment. However, clinical observation suggests that some patients with mild symptoms may quickly deteriorate. Hence, it is crucial to identify patient early deterioration to optimize treatment strategy. To this end, we develop an early-warning system with deep learning techniques to predict COVID-19 malignant progression. Our method leverages CT scans and the clinical data of outpatients and achieves an AUC of 0.920 in the single-center study. We also propose a domain adaptation approach to improve the generalization of our model and achieve an average AUC of 0.874 in the multicenter study. Moreover, our model automatically identifies crucial indicators that contribute to the malignant progression, including Troponin, Brain natriuretic peptide, White cell count, Aspartate aminotransferase, Creatinine, and Hypersensitive C-reactive protein.

Identification of robust diagnostic and prognostic gene signatures in different grades of gliomas: a retrospective study

Background

Gliomas are the most common primary tumors of the central nervous system. The complexity and heterogeneity of the tumor makes it difficult to obtain good biomarkers for drug development. In this study, through The Cancer Genome Atlas (TCGA) and Chinese Glioma Genome Atlas (CGGA), we analyze the common diagnostic and prognostic moleculer markers in Caucasian and Asian populations, which can be used as drug targets in the future.

Methods

The RNA-seq data from Genotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) were analyzed to identify signatures. Based on the signatures, the prognosis index (PI) of every patient was constructed to predict the prognostic risk. Also, gene ontology (GO) functional enrichment analysis and KEGG analysis were conducted to investigate the biological functions of these mRNAs. Glioma patients’ data in the CGGA database were introduced to validate the effectiveness of the signatures among Chinese populations. Excluding the previously reported prognostic markers of gliomas from this study, the expression of HSPA5 and MTPN were examined by qRT-PCR and immunohistochemical assay.

Results

In total, 20 mRNAs were finally selected to build PI for patients from TCGA, including 16 high-risk genes and four low-risk genes. For Chinese patients, the log-rank test p values of PI were both less than 0.0001 in two independent datasets. And the AUCs were 0.831 and 0.907 for 3 years of two datasets, respectively. Moreover, among these 20 mRNAs, 10 and 15 mRNAs also had a significant predictive effect via univariate COX analysis in CGGA_693 and CGGA_325, respectively. qRT-PCR and Immunohistochemistry assay indicated that HSPA5 and MTPN over-expressed in Glioma samples compared to normal samples.

Conclusion

The 20-gene signature can forecast the risk of Glioma in TCGA effectively, moreover it can also predict the risks of Chinese patients through validation in the CGGA database. HSPA5 and MTPN are possible biomarkers of gliomas suitable for all populations to improve the prognosis of these patients.

Procalcitonin measurement using antibody-conjugated fluorescent microspheres distinguishes atypical bacterial meningitis from viral encephalitis in children

Examination of cerebrospinal fluid in atypical bacterial meningitis (ABM) is similar to that of viral encephalitis (VE), so ABM can easily be misdiagnosed as VE, which can delay diagnosis and treatment. We developed a simple, rapid hand-held lateral flow immunoassay detection system based on fluorescent microspheres (FMS) for procalcitonin (PCT) detection, which provides an indicator to differentiate between ABM and VE. With this novel method, the antigen-antibody reaction systems involve different species, making the test strips more stable than those utilizing one species. The strips exhibited a wide dynamic range (0.04-50 ng/mL) and good sensitivity (0.03 ng/mL). The function of PCT in the identification of ABM and VE in children was further studied. A significant difference in PCT levels was observed between the ABM and VE groups (P = 0.00) and between the ABM and the normal control groups (P = 0.00). PCT levels were not significantly different between the VE and normal control groups (P = 0.30). The area under the receiver operating characteristic curve of PCT for the diagnosis of ABM was 0.95. These findings collectively indicate the usefulness of the PCT detection method based on FMS for clinically differentiating between ABM and VE.

A New Risk Index Combining d-Dimer, Fibrinogen, HE4, and CA199 Differentiates Suspecting Endometrial Cancer From Patients With Abnormal Vaginal Bleeding or Discharge

PURPOSE

To establish an efficient new risk index for screening patients with endometrial cancer from patients with abnormal vaginal bleeding or discharge.

METHOD

A total of 254 patients with abnormal vaginal bleeding or discharge were included in this study. Several candidate markers, including HE4, CA125, CA199, CA153, AFP, CEA, d-dimer, and fibrinogen, were employed. A new risk index for endometrial cancer screening was established by binary logistic regression. The diagnostic value of the candidate markers and the new risk index were assessed by a receiver operating characteristic curve, sensitivity, and specificity.

RESULTS

The most valuable diagnostic indicator for endometrial cancer was HE4, followed by d-dimer and then fibrinogen (area under the receiver operating characteristic curve: HE4 = 0.794, d-dimer = 0.717, fibrinogen = 0.690). The new risk index was superior to a single application of markers and a widely used combination (HE4 and CA125). At the ideal cutoff level, the sensitivity and specificity were 91.34% and 70.08%, respectively. In addition, only patients without organic disease served as controls, which further increase its performance (area under the receiver operating characteristic curve = 0.932, sensitivity = 94.49%, and specificity = 77.42%).

CONCLUSIONS

The new risk index combining HE4, d-dimer, fibrinogen, and CA199 was the ideal combination for the screening of endometrial cancer. As a simple, rapid, nondestructive detection method, the new risk index is worth promotion in clinical practice, especially in primary medical institutions.

Use and performance of machine learning models for type 2 diabetes prediction in community settings: A systematic review and meta-analysis

OBJECTIVE

We aimed to identify machine learning (ML) models for type 2 diabetes (T2DM) prediction in community settings and determine their predictive performance.

METHOD

Systematic review of ML predictive modelling studies in 13 databases since 2009 was conducted. Primary outcomes included metrics of discrimination, calibration, and classification. Secondary outcomes included important variables, level of validation, and intended use of models. Meta-analysis of c-indices, subgroup analyses, meta-regression, publication bias assessments and sensitivity analyses were conducted.

RESULTS

Twenty-three studies (40 prediction models) were included. Studies with high-, moderate-, and low- risk of bias were 3, 14, and 6 respectively. All studies conducted internal validation whereas none conducted external validation of their models. Twenty studies provided classification metrics to varying extents whereas only 7 studies performed model calibration. Eighteen studies reported information on both the variables used for model development and the feature importance. Twelve studies highlighted potential applicability of their models for T2DM screening. Meta-analysis produced a good pooled c-index (0.812). Sources of heterogeneity were identified through subgroup analyses and meta-regression. Issues pertaining to methodological quality and reporting were observed.

CONCLUSIONS

We found evidence of good performance of ML models for T2DM prediction in the community. Improvements to methodology, reporting and validation are needed before they can be used at scale.

Sample size calculation and re-estimation based on the prevalence in a single-arm confirmatory diagnostic accuracy study

INTRODUCTION

In a confirmatory diagnostic accuracy study, sensitivity and specificity are considered as co-primary endpoints. For the sample size calculation, the prevalence of the target population must be taken into account to obtain a representative sample. In this context, a general problem arises. With a low or high prevalence, the study may be overpowered in one subpopulation. One further issue is the correct pre-specification of the true prevalence. With an incorrect assumption about the prevalence, an over- or underestimated sample size will result.

METHODS

To obtain the desired power independent of the prevalence, a method for an optimal sample size calculation for the comparison of a diagnostic experimental test with a prespecified minimum sensitivity and specificity is proposed. To face the problem of an incorrectly pre-specified prevalence, a blinded one-time re-estimation design of the sample size based on the prevalence and a blinded repeated re-estimation design of the sample size based on the prevalence are evaluated by a simulation study. Both designs are compared to a fixed design and additionally among each other.

RESULTS

The type I error rates of both blinded re-estimation designs are not inflated. Their empirical overall power equals the desired theoretical power and both designs offer unbiased estimates of the prevalence. The repeated re-estimation design reveals no advantages concerning the mean squared error of the re-estimated prevalence or sample size compared to the one-time re-estimation design. The appropriate size of the internal pilot study in the one-time re-estimation design is 50% of the initially calculated sample size.

CONCLUSIONS

A one-time re-estimation design of the prevalence based on the optimal sample size calculation is recommended in single-arm diagnostic accuracy studies.

Tremorgenic effects and functional metabolomics analysis of lolitrem B and its biosynthetic intermediates

The neuroactive mycotoxin lolitrem B causes a neurological syndrome in grazing livestock resulting in hyperexcitability, muscle tremors, ataxia and, in severe cases, clonic seizures and death. To define the effects of the major toxin lolitrem B in the brain, a functional metabolomic study was undertaken in which motor coordination and tremor were quantified and metabolomic profiling undertaken to determine relative abundance of both toxin and key neurotransmitters in various brain regions in male mice. Marked differences were observed in the duration of tremor and coordination between lolitrem B pathway members, with some showing protracted effects and others none at all. Lolitrem B was identified in liver, kidney, cerebral cortex and thalamus but not in brainstem or cerebellum which were hypothesised previously to be the primary site of action. Metabolomic profiling showed significant variation in specific neurotransmitter and amino acid profiles over time. This study demonstrates accumulation of lolitrem B in the brain, with non-detectable levels of toxin in the brainstem and cerebellum, inducing alterations in metabolites such as tyrosine, suggesting a dynamic catecholaminergic response over time. Temporal characterisation of key pathways in the pathophysiological response of lolitrem B in the brain were also identified.

Quantitative Multiparametric MRI May Augment the Response to Radiotherapy in Mid-Treatment Assessment of Patients with Esophageal Carcinoma

OBJECTIVE

The purpose of this study was to assess the mid-treatment response to radiotherapy (RT) using dynamic contrast-enhanced MRI (DCE-MRI) and diffusion-weighted imaging (DWI) in patients with esophageal cancer (EC).

METHODS

42 patients with squamous EC were prepared for DCE-MRI and DWI scans both before treatment (NRT) and after the fifth radiotherapy (5th RT). The patients were classified into two groups (complete response [CR] and partial response [PR]) according to tumor regression after treatment. The quantitative parameters of DCE-MRI (Ktrans, Kep, Ve, and ADC) were measured. A receiver operating characteristic curve (ROC) was used to detect the efficiency of the above parameters.

RESULTS

After 1 month of RT, 29 patients were classified as CR and 11 patients were classified as PR. In the NRT group, the p values of Ktrans, Kep, Ve, and ADC were 0.004, 0.078, 0.0008, and <0.0001, respectively. After the 5th RT, the p values of the above parameters were <0.001, 0.005, 0.108, and 0.365, respectively. In the NRT group, the areas under the ROC curves of Ktrans, Ve, and ADC were 0.790, 0.617, and 0.737; the sensitivity values were 89.3, 92.5, and 90.0%; the specificity values were 69.4, 27.5, and 50.0%. In the 5th RT group, the areas under the ROC curves of Ktrans and Kep were 0.816 and 0.804; the sensitivity values were 71.2 and 95.0%; the specificity values were 81.6 and 50.0%.

CONCLUSION

DCE-MRI combined with DWI is effective in the early prediction of radiotherapeutic response of EC after the 5th RT other than after the traditional final treatment.

iTRAQ-based proteomics reveals SOD2 as a potential salivary biomarker in liver cancer

BACKGROUND

Salivary proteomic analysis has been extensively used in a wide range of cancer, but not in hepatocellular carcinoma. The aim of this study was to identify potential salivary biomarkers for hepatocellular carcinoma clinical screening.

METHODS

In this study, we performed isobaric tags for relative and absolute quantitation (iTRAQ)-based quantitative proteomics analysis to detect differentially expressed proteins between saliva samples from 15 hepatocellular carcinoma patients and 15 healthy controls. Enzyme-linked immunosorbent assay (ELISA) verification was undertaken in saliva samples from 14 hepatocellular carcinoma patients and 14 healthy controls.

RESULTS

Overall, 133 proteins with significant differential expression level (ratio > 1.5 or < 0.67) were detected. Using bioinformatic analysis, two candidate proteins were selected and subsequently verified by ELISA. The increased expression of superoxide dismutase 2, mitochondrial (SOD2) in hepatocellular carcinoma patients was confirmed by ELISA, with an area under the curve value of 0.9082.

CONCLUSIONS

iTRAQ-based quantitative proteomics revealed that SOD2 might serve as a potential salivary biomarker for hepatocellular carcinoma detection. Our results indicated that a noninvasive and inexpensive salivary test might be established for hepatocellular carcinoma detection.

Screening for active pulmonary tuberculosis: Development and applicability of artificial neural network models

Tuberculosis (TB) remains a significant public health challenge, motivated by the diversity of healthcare epidemiological settings, as other factors. Cost-effective screening has substantial importance for TB control, demanding new diagnostic tools. This paper proposes a decision support tool (DST) for screening pulmonary TB (PTB) patients at a secondary clinic. The DST is composed of an adaptive resonance model (iART) for risk group identification (low, medium and high) and a multilayer perceptron (MLP) neural network for classifying patients as active or inactive PTB. Our tool attains an overall sensitivity (SE) and specificity (SP) of 92% (95% CI; 79-97) and 58% (95% CI; 47-68), respectively. SE values for smear-positive and smear-negative patients are 96% (95% CI; 80-99) and 82% (95% CI; 52-95), as well as higher than 83% (95% CI; 43-97) in low and high-risk cases. Even in scenarios with prevalence up to 20%, negative predictive values superior to 95% are obtained. The proposed DST provides a quick and low-cost pretest for presumptive PTB patients, which is useful to guide confirmatory testing and patient management, especially in settings with limited resources in low and middle-incoming countries.

Prediction models for patients with esophageal or gastric cancer: A systematic review and meta-analysis

Background

Clinical prediction models are increasingly used to predict outcomes such as survival in cancer patients. The aim of this study was threefold. First, to perform a systematic review to identify available clinical prediction models for patients with esophageal and/or gastric cancer. Second, to evaluate sources of bias in the included studies. Third, to investigate the predictive performance of the prediction models using meta-analysis.

Methods

MEDLINE, EMBASE, PsycINFO, CINAHL, and The Cochrane Library were searched for publications from the year 2000 onwards. Studies describing models predicting survival, adverse events and/or health-related quality of life (HRQoL) for esophageal or gastric cancer patients were included. Potential sources of bias were assessed and a meta-analysis, pooled per prediction model, was performed on the discriminative abilities (c-indices).

Results

A total of 61 studies were included (45 development and 16 validation studies), describing 47 prediction models. Most models predicted survival after a curative resection. Nearly 75% of the studies exhibited bias in at least 3 areas and model calibration was rarely reported. The meta-analysis showed that the averaged c-index of the models is fair (0.75) and ranges from 0.65 to 0.85.

Conclusion

Most available prediction models only focus on survival after a curative resection, which is only relevant to a limited patient population. Few models predicted adverse events after resection, and none focused on patient’s HRQoL, despite its relevance. Generally, the quality of reporting is poor and external model validation is limited. We conclude that there is a need for prediction models that better meet patients’ information needs, and provide information on both the benefits and harms of the various treatment options in terms of survival, adverse events and HRQoL.

A Novel and Practical Chromatographic "Fingerprint-ROC-SVM" Strategy Applied to Quality Analysis of Traditional Chinese Medicine Injections: Using KuDieZi Injection as a Case Study

Fingerprinting is widely and commonly used in the quality control of traditional Chinese medicine (TCM) injections. However, current studies informed that the fingerprint similarity evaluation was less sensitive and easily generated false positive results. For this reason, a novel and practical chromatographic “Fingerprint-ROC-SVM” strategy was established by using KuDieZi (KDZ) injection as a case study in the present article. Firstly, the chromatographic fingerprints of KDZ injection were obtained by UPLC and the common characteristic peaks were identified with UPLC/Q-TOF-MS under the same chromatographic conditions. Then, the receiver operating characteristic (ROC) curve was used to optimize common characteristic peaks by the AUCs value greater than 0.7. Finally, a support vector machine (SVM) model, with the accuracy of 97.06%, was established by the optimized characteristic peaks and applied to monitor the quality of KDZ injection. As a result, the established model could sensitively and accurately distinguish the qualified products (QPs) with the unqualified products (UPs), high-temperature processed samples (HTPs) and high-illumination processed samples (HIPs) of KDZ injection, and the prediction accuracy was 100.00%, 93.75% and 100.00%, respectively. Furthermore, through the comparison with other chemometrics methods, the superiority of the novel analytical strategy was more prominent. It indicated that the novel and practical chromatographic “Fingerprint-ROC-SVM” strategy could be further applied to facilitate the development of the quality analysis of TCM injections.

Optimization of behavioural tests for the prediction of outcomes in mouse models of focal middle cerebral artery occlusion

Intraluminal middle cerebral artery occlusion (MCAO) is the most widely used model of stroke. We aimed to predict the outcome of MCAO using a combination of fine behavioural tests for the prediction of unsuccessful surgery in mice leading to no infarction, haemorrhage and unexpected death. MCAO was performed on adult mice under the guidance of laser-Doppler flowmetry (LDF) to warrant a decrease in regional cerebral blood flow (rCBF) in the MCA territory. Four outcomes of MCAO were defined according to histological analysis: infarction, no infarction, haemorrhage and unexpected death (death within 24h post-surgery). Fine behavioural tests including the rotarod, modified neurological severity score (mNSS), Clark general and Clark focal tests were performed separately at 6h, 12h and 24h post-stroke. A total of 94 mice were included in the analysis. The infarction rate associated with MCAO was 58.5% (55/94). After optimization of the timing and behavioural tests, we found that higher Clark focal (>17.5) or higher mNSS scores (>10) were markedly related to early death, whereas a lower mNSS score (<3.5) was indicative of a tendency to show no infarction at 6h post-stroke. After 24h post-stroke, there was a positive correlation between the infarct volume and Clark focal results. Behavioural tests could help to predict the outcomes in the MCAO mouse model.

Prediction of near-term increases in suicidal ideation in recently depressed patients with bipolar II disorder using intensive longitudinal data

BACKGROUND

There are substantial gaps in understanding near-term precursors of suicidal ideation in bipolar II disorder. We evaluated whether repeated patient-reported mood and energy ratings predicted subsequent near-term increases in suicide ideation.

METHODS

Secondary data were used from 86 depressed adults with bipolar II disorder enrolled in one of 3 clinical trials evaluating Interpersonal and Social Rhythm Therapy and/or pharmacotherapy as treatments for depression. Twenty weeks of daily mood and energy ratings and weekly Hamilton Depression Rating Scale (HDRS) were obtained. Penalized regression was used to model trajectories of daily mood and energy ratings in the 3 week window prior to HDRS Suicide Item ratings.

RESULTS

Participants completed an average of 68.6 (sd=52) days of mood and energy ratings. Aggregated across the sample, 22% of the 1675 HDRS Suicide Item ratings were non-zero, indicating presence of at least some suicidal thoughts. A cross-validated model with longitudinal ratings of energy and depressed mood within the three weeks prior to HDRS ratings resulted in an AUC of 0.91 for HDRS Suicide item >2, accounting for twice the variation when compared to baseline HDRS ratings. Energy, both at low and high levels, was an earlier predictor than mood.

LIMITATIONS

Data derived from a heterogeneous treated sample may not generalize to naturalistic samples. Identified suicidal behavior was absent from the sample so it could not be predicted.

CONCLUSIONS

Prediction models coupled with intensively gathered longitudinal data may shed light on the dynamic course of near-term risk factors for suicidal ideation in bipolar II disorder.

The Sensitivity and Specificity of Pulmonary Carbon Dioxide Elimination for Noninvasive Assessment of Fluid Responsiveness

BACKGROUND

We sought to determine whether the response of pulmonary elimination of CO2 (VCO2) to a sudden increase in positive end-expiratory pressure (PEEP) could predict fluid responsiveness and serve as a noninvasive surrogate for cardiac index (CI).

METHODS

Fifty-two patients undergoing cardiovascular surgery were included in this study. By using a constant-flow ventilation mode, we performed a PEEP challenge of 1-minute increase in PEEP from 5 to 10 cm H2O. At PEEP of 5 cm H2O, patients were preloaded with 500 mL IV saline solution after which a second PEEP challenge was performed. Patients in whom fluid administration increased CI by ≥15% from the individual baseline value were defined as volume responders. Beat-by-beat CI was derived from arterial pulse contour analysis, and breath-by-breath VCO2 data were collected during the protocol. The sensitivity and specificity of VCO2 for detecting the fluid responders according to CI was performed by the receiver operating characteristic curves.

RESULTS

Twenty-one of 52 patients were identified as fluid responders (40%). The PEEP maneuver before fluid administration decreased CI from 2.65 ± 0.34 to 2.21 ± 0.32 L/min/m (P = 0.0011) and VCO2 from 150 ± 23 to 123 ± 23 mL/min (P = 0.0036) in responders, whereas the changes in CI and VCO2 were not significant in nonresponders. The PEEP challenge after fluid administration induced no significant changes in CI and VCO2, in neither responders nor nonresponders. PEEP-induced decreases in CI and VCO2 before fluid administration were well correlated (r = 0.75, P < 0.0001) but not thereafter. The area under the receiver operating characteristic curves for a PEEP-induced decrease in ΔCI and ΔVCO2 was 0.99, with a 95% confidence interval from 0.96 to 0.99 for ΔCI and from 0.97 to 0.99 for ΔVCO2. During the PEEP challenge, a decrease in VCO2 by 11% predicted fluid responsiveness with a sensitivity of 0.90 (95% confidence interval, 0.87-0.93) and a specificity of 0.95 (95% confidence interval, 0.92-0.98).

CONCLUSIONS

PEEP-induced changes in VCO2 predicted fluid responsiveness with accuracy in patients undergoing cardiac surgery.

Quantifying patterns of brain activity: Distinguishing unaffected siblings from participants with ADHD and healthy individuals

Attention-deficit/hyperactivity disorder (ADHD) is one of the most prevalent and heritable psychiatric disorders. While previous studies have focussed on mapping focal or connectivity differences at the group level, the present study employed pattern recognition to quantify group separation between unaffected siblings, participants with ADHD, and healthy controls on the basis of spatially distributed brain activations. This was achieved using an fMRI-adapted version of the Stop-Signal Task in a sample of 103 unaffected siblings, 184 participants with ADHD, and 128 healthy controls. We used activation maps derived from three task regressors as features in our analyses employing a Gaussian process classifier. We showed that unaffected siblings could be distinguished from participants with ADHD (area under the receiver operating characteristic curve (AUC) = 0.65, p = 0.002, 95% Modified Wald CI: 0.59-0.71 AUC) and healthy controls (AUC = 0.59, p = 0.030, 95% Modified Wald CI: 0.52-0.66 AUC), although the latter did not survive correction for multiple comparisons. Further, participants with ADHD could be distinguished from healthy controls (AUC = 0.64, p = 0.001, 95% Modified Wald CI: 0.58-0.70 AUC). Altogether the present results characterise a pattern of frontolateral, superior temporal and inferior parietal expansion that is associated with risk for ADHD. Unaffected siblings show differences primarily in frontolateral regions. This provides evidence for a neural profile shared between participants with ADHD and their healthy siblings.

Metabonomic analysis of the protective effect of quercetin on the toxicity induced by mixture of organophosphate pesticides in rat urine

The present study aims to investigate the protective effect of quercetin against the joint toxic action induced by the mixture of four organophosphate pesticides (mixture-OPs) (dimethoate, acephate, dichlorvos, and phorate) at their corresponding no observed adverse effect level (NOAEL) using metabonomics. Rats were randomly divided into control, quercetin-treated, mixture-OPs-treated, and quercetin plus mixture-OPs-treated groups. Mixture-OPs and quercetin were given to the rats daily through drinking water and intragastric administration, respectively, for 90 days. The metabonomic profiles of rat urine were analyzed using ultra-performance liquid chromatography–mass spectrometry (UPLC/MS). The 14 metabolites significantly changed in the treatment groups compared with the control group, including the biomarkers of OPs exposure (dimethylphosphate, dimethyldithiophosphate, diethylphosphate) and the metabolites of quercetin (quercetin and isorhamnetina). The intensities of gentisic acid, creatinine, suberic acid, hippuric acid, uric acid, and citric acid significantly decreased, whereas the intensities of 7-methylguanine, estrone sulfate, and cholic acid significantly increased, in the mixture-OPs-treated group compared with the control group ( p < 0.01). The variation tendency of the aforementioned metabolites was significantly ameliorated in the high-dose quercetin (50 mg/(kg bw day)) plus mixture-OPs-treated group compared with the mixture-OPs-treated group ( p < 0.05). However, the intensities of these metabolites in the high-dose quercetin plus mixture-OPs-treated group were still significantly different from those of the control group ( p < 0.05). Results indicated that high dose of quercetin elicits a partial protective effect on the toxicity induced by mixture-OPs, including fatty acid and energy metabolism, antioxidant defense system, DNA damage, and liver and kidney function.

Digital holographic microscopy as screening tool for cervical cancer preliminary study

BACKGROUND

The liquid-based cytology Papanicolaou (Pap) test has gradually replaced the conventional Pap test, as it seems to be more efficient, in part because it uses computer assistance. However, Pap test diagnoses are still subjective and are cost-consuming or time-consuming. The authors describe a new digital holographic microscopy instrument and accompanying software. This technology provides an instantaneous, 3-dimensional image reconstruction of cells directly from the uterine cervical sample vial.

METHODS

Analyses were performed using the proprietary digital holographic microscopy (DHM) instrument, computer, and software. Residual materials from 3 randomly chosen laboratories were analyzed and subjected to DHM, and the sensitivity and specificity of DHM were calculated for the detection neoplasia.

RESULTS

In 41 specimens that yielded normal Pap test results, 1333 cells were analyzed using DHM; and, in 30 specimens that yielded abnormal Pap test results, 494 cells were analyzed using DHM. Available histologic correlation was as follows: 4 histologic samples were diagnosed as grade 1 cervical intraepithelial neoplasia (corresponding to 2 cytologic diagnoses of low-grade squamous intraepithelial lesion and 2 cytologic diagnoses of atypical squamous cells of undetermined significance), 4 histologic samples were diagnosed as grade 3 cervical intraepithelial neoplasia (corresponding to 2 cytologic diagnoses of low-grade squamous intraepithelial lesion and 2 cytologic diagnosis of atypical squamous cells, cannot rule out high-grade squamous intraepithelial lesion), and 11 lesions were diagnosed as benign. Receiver operating characteristic curve analysis indicated that the best criteria were the nuclear-to-cytoplasmic ratio (area under the curve, 0.88; 95% confidence interval, 0.68-1.00) and the greatest nuclear dimension (area under the curve, 0.85; 95% confidence interval, 0.66-1.00).

CONCLUSIONS

This preliminary study demonstrated for the first time that the DHM technique is suitable for the processing of gynecologic cervical samples. Nevertheless, DHM criteria and parameters could be better defined. Hopefully, holographic analysis will be performed automatically and will provide an instantaneous, cost-effective diagnosis from a closed vial with the preservation of all cellular material. Cancer Cytopathol 2016;124:573-80. © 2016 American Cancer Society.

Ratio and log odds of positive lymph nodes in breast cancer patients with mastectomy

PURPOSE

This study aimed to investigate the predictive role of lymph nodes (LNs) and assess the prognostic significance of the ratio of positive LNs (LNR) and log odds of positive LNs (LODDS) in breast cancer patients who have undergone a mastectomy.

PATIENTS AND METHODS

All of the breast cancer patients in the Taiwan Cancer Database during 2002-2006 were considered. We excluded patients who had inflammatory breast cancer, stage 0 and IV disease, breast conservative surgery or survival <1 month. The primary end point was overall survival (OS). A Cox hazards model was constructed and compared via Nagelkerke R(2) (R(2)N) and receiver operating characteristics (ROC).

RESULTS

A total of 11,349 (6042 node-negative, 5307 node-positive) patients were enrolled, and 10.5% patients had a limited number of LNs harvested. In a multivariate Cox model, LNR and LODDS demonstrated prognostic significance (<0.001). For node-positive patients, a model with LNR showed the best fit (P < 0.001; R(2)N = 18.2%) when sufficient LNs were examined. However, a model with LODDS showed the best fit in patients with a limited number of LNs harvested (P < 0.001; R(2)N = 21.1%), even in node-negative patients (P = 0.004; R(2)N = 13.5%). The area under the ROC curve (AUC) was highest for LODDS (AUC: 0.761), followed by LNR (AUC: 0.757). A limited LN harvest induced an AUC value for an approximate 3.6% loss (LNR) or 3.1% loss (LODDS).

CONCLUSION

The prognostic superiority of LNR is confounded by a limited LN harvest, thus making LODDS the most powerful and unified prognostic classifier in breast cancer patients who have had a mastectomy.

A wild bootstrap approach for the selection of biomarkers in early diagnostic trials

BACKGROUND

In early diagnostic trials, particularly in biomarker studies, the aim is often to select diagnostic tests among several methods. In case of metric, discrete, or even ordered categorical data, the area under the receiver operating characteristic (ROC) curve (denoted by AUC) is an appropriate overall accuracy measure for the selection, because the AUC is independent of cut-off points.

METHODS

For selection of biomarkers the individual AUC's are compared with a pre-defined threshold. To keep the overall coverage probability or the multiple type-I error rate, simultaneous confidence intervals and multiple contrast tests are considered. We propose a purely nonparametric approach for the estimation of the AUC's with the corresponding confidence intervals and statistical tests. This approach uses the correlation among the statistics to account for multiplicity. For small sample sizes, a Wild-Bootstrap approach is presented. It is shown that the corresponding intervals and tests are asymptotically exact.

RESULTS

Extensive simulation studies indicate that the derived Wild-Bootstrap approach keeps and exploits the nominal type-I error at best, even for high accuracies and in case of small samples sizes. The strength of the correlation, the type of covariance structure, a skewed distribution, and also a moderate imbalanced case-control ratio do not have any impact on the behavior of the approach. A real data set illustrates the application of the proposed methods.

CONCLUSION

We recommend the new Wild Bootstrap approach for the selection of biomarkers in early diagnostic trials, especially for high accuracies and small samples sizes.