Development of a machine learning-based risk model for postoperative complications of lung cancer surgery.

PURPOSE: To develop a comorbidity risk score specifically for lung resection surgeries. METHODS: We reviewed the medical records of patients who underwent lung resections for lung cancer, and developed a risk model using data from 2014 to 2017 (training dataset), validated using data from 2018 to 2019 (validation dataset). Forty variables were analyzed, including 35 factors related to the patient's overall condition and five factors related to surgical techniques and tumor-related factors. The risk model for postoperative complications was developed using an elastic net regularized generalized linear model. The performance of the risk model was evaluated using receiver operating characteristic curves and compared with the Charlson Comorbidity Index (CCI). RESULTS: The rate of postoperative complications was 34.7% in the training dataset and 21.9% in the validation dataset. The final model consisted of 20 variables, including age, surgical-related factors, respiratory function tests, and comorbidities, such as chronic obstructive pulmonary disease, a history of ischemic heart disease, and 12 blood test results. The area under the curve (AUC) for the developed risk model was 0.734, whereas the AUC for the CCI was 0.521 in the validation dataset. CONCLUSIONS: The new machine learning model could predict postoperative complications with acceptable accuracy. CLINICAL REGISTRATION NUMBER: 2020-0375.

CHST4 Gene as a Potential Predictor of Clinical Outcome in Malignant Pleural Mesothelioma.

Malignant pleural mesothelioma (MPM) develops primarily from asbestos exposures and has a poor prognosis. In this study, The Cancer Genome Atlas was used to perform a comprehensive survival analysis, which identified the CHST4 gene as a potential predictor of favorable overall survival for patients with MPM. An enrichment analysis of favorable prognostic genes, including CHST4, showed immune-related ontological terms, whereas an analysis of unfavorable prognostic genes indicated cell-cycle-related terms. CHST4 mRNA expression in MPM was significantly correlated with Bindea immune-gene signatures. To validate the relationship between CHST4 expression and prognosis, we performed an immunohistochemical analysis of CHST4 protein expression in 23 surgical specimens from surgically treated patients with MPM who achieved macroscopic complete resection. The score calculated from the proportion and intensity staining was used to compare the intensity of CHST4 gene expression, which showed that CHST4 expression was stronger in patients with a better postoperative prognosis. The median overall postoperative survival was 107.8 months in the high-expression-score group and 38.0 months in the low-score group (p = 0.044, log-rank test). Survival after recurrence was also significantly improved by CHST4 expression. These results suggest that CHST4 is useful as a prognostic biomarker in MPM.

Prognostic impact of highly solid component in early-stage solid lung adenocarcinoma.

Background: Based on computed tomography (CT) findings of lung cancer, solid nodules have a much worse prognosis than subsolid nodules, even if the nodules are subcentimeter in size. There is, however, no systematic method for determining the prognosis of solid tumors on CT. This study aimed to discover the prognostic factor of early-stage solid lung adenocarcinoma using three-dimensional CT volumetry. Methods: Patients with pathological stage I solid lung adenocarcinoma who underwent complete resection between 2007 and 2012 were selected in this retrospective study. Clinicopathological data and preoperative multidetector CT findings, such as tumor size on the two-dimensional axial image, three-dimensional tumor volume between -600 and 199 HU, and three-dimensional solid volume between 0 and 199 HU, which corresponded to highly solid components, were compared between recurrence and non-recurrence. Furthermore, these radiological values were compared to pathological invasive volume (PIV). Results: During this time, 709 patients had their lung cancer completely removed. From this cohort, 90 patients with pathological stage I solid lung adenocarcinoma were selected. In addition, recurrence was found in 26 patients (28.9%). Although two-dimensional axial image, serum carcinoembryonic antigen (CEA) level, and SUVmax on 18F fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) did not differ statistically between recurrent and non-recurrent patients, three-dimensional tumor and solid tumor volume did. Multivariate analysis indicated that three-dimensional solid tumor volume [hazard ratio: 2.440; 95% confidence interval (CI): 1.110-5.361, P=0.026] and epidermal growth factor receptor (EGFR) mutation (hazard ratio: 4.307; 95% CI: 1.328-13.977, P=0.015) were significantly associated with disease-free survival (DFS). When three-dimensional tumor and solid tumor volume were compared to PIV, three-dimensional solid tumor volume (3,091 mm3 on average) showed a highly similar value with PIV (2,930 mm3 on average), whereas three-dimensional tumor volume (6,175 mm3 on average) was significantly larger than PIV (P<0.001). Conclusions: In patients with early-stage solid lung adenocarcinoma, the measurement of three-dimensional solid tumor volume, which is correlated with PIV, accurately predicted the postoperative outcome.

Evaluating glucose variability through OGTT in early pregnancy and its association with hypertensive disorders of pregnancy in non-diabetic pregnancies: a large-scale multi-center retrospective study.

BACKGROUND: Recent evidence suggests increased glucose variability (GV) causes endothelial dysfunction, a central pathology of hypertensive disorders of pregnancy (HDP). We aimed to investigate the association between GV in early pregnancy and subsequent HDP development among non-diabetes mellitus (DM) pregnancies. METHODS: This multicenter retrospective study used data from singleton pregnancies between 2009 and 2019. Among individuals who had 75 g-OGTT before 20 weeks of gestation, we evaluated GV by 75 g-OGTT parameters and examined its relationship with HDP development, defining an initial-increase from fasting-plasma glucose (PG) to 1-h-PG and subsequent-decrease from 1-h-PG to 2-h-PG. RESULTS: Approximately 3.0% pregnancies (802/26,995) had 75 g-OGTT before 20 weeks of gestation, and they had a higher prevalence of HDP (14.3% vs. 7.5%). The initial-increase was significantly associated with overall HDP (aOR 1.20, 95% CI 1.02-1.42), and the subsequent-decrease was associated with decreased and increased development of early-onset (EoHDP: aOR 0.56, 95% CI 0.38-0.82) and late-onset HDP (LoHDP: aOR 1.38, 95% CI 1.11-1.73), respectively. CONCLUSIONS: A pattern of marked initial-increase and minor subsequent-decrease (i.e., sustained hyperglycemia) was associated with EoHDP. Contrarily, the pattern of marked initial-increase and subsequent-decrease (i.e., increased GV) was associated with LoHDP. This provides a new perspective for future study strategies.

The prognostic impact of limited-staging surgery in patients with stage IA epithelial ovarian cancer: a multi-center study with a propensity score-adjusted analysis.

OBJECTIVE: Complete-staging surgery is recommended for stage IA ovarian cancer, but may be omitted for various reasons, including the preservation of fertility and an advanced age. We herein investigated the prognostic impact of limited-staging surgery in patients with stage IA epithelial ovarian cancer. METHODS: We retrospectively collected data on 4730 patients with malignant ovarian tumors from the databases of multiple institutions and ultimately included 293 with stage IA epithelial ovarian cancer. Limited-staging surgery was defined as one that did not involve hysterectomy, systematic retroperitoneal lymphadenectomy or the collection of ascites cytology. We used an inverse probability of treatment weighting analysis with propensity scores and estimated the hazard ratios of recurrence and death with limited-staging surgery. RESULTS: In total, 176 out of 293 patients (39.9%) were assigned to the limited-staging surgery group. After propensity score adjustments, no significant differences were observed in recurrence-free survival or overall survival between the limited- and complete-staging surgery groups. Even in the subgroup analysis with age stratification, recurrence-free survival and overall survival were similar in the limited- and complete-staging surgery groups. CONCLUSIONS: The present results indicate the limited prognostic impact of limited-staging surgery for stage IA epithelial ovarian cancer.

Factors associated with adverse drug reactions or death in very elderly hospitalized patients with pulmonary tuberculosis.

The aging of patients with tuberculosis and better therapeutic management for them are recent concerns. This study aimed to identify risk factors for adverse drug reactions (ADRs) or death in very elderly patients with pulmonary tuberculosis and to assess the association between the dosage of antituberculosis drugs and outcomes. We conducted a multicenter retrospective study at two hospitals. Hospitalized patients (≥ 80 years old) with pulmonary tuberculosis who were treated with antituberculosis drugs were enrolled. Multivariate analysis was performed to assess factors associated with ADRs or death within 60 days after treatment initiation. In total, 632 patients were included. The primary endpoint occurred in 268 patients (190 ADRs and 78 deaths). A serum albumin level < 2.5 g/dL, respiratory failure, and dependent activities of daily living were independent risk factors for ADRs or death. However, a low dosage (< 8 mg/kg/day) of rifampicin was associated with a lower risk of the primary outcomes. Delayed time to negative sputum culture conversion was not observed in the lower dosage of rifampicin group. Very elderly hospitalized tuberculosis patients with the aforementioned risk factors should be carefully monitored to receive safer treatment. Rifampicin dosage reduction may be considered for very elderly tuberculosis patients to prevent ADRs/death.

Extended-spectrum antibiotics for community-acquired pneumonia with a low risk for drug-resistant pathogens.

OBJECTIVES: The potential hazards of extended-spectrum antibiotic therapy for patients with community-acquired pneumonia (CAP) with low risk for drug-resistant pathogens (DRPs) remain unclear; however, risk assessment for DRPs is essential to determine the initial antibiotics to be administered. The study objective was to assess the effect of unnecessary extended-spectrum therapy on the mortality of such patients. METHODS: A post hoc analysis was conducted after a prospective multicenter observational study for CAP. Multivariable logistic regression analysis was performed to assess the effect of extended-spectrum therapy on 30-day mortality. Three sensitivity analyses, including propensity score analysis to confirm the robustness of findings, were also performed. RESULTS: Among 750 patients with CAP, 416 with CAP with a low risk for DRPs were analyzed; of these, 257 underwent standard therapy and 159 underwent extended-spectrum therapy. The 30-day mortality was 3.9% and 13.8% in the standard and extended-spectrum therapy groups, respectively. Primary analysis revealed that extended-spectrum therapy was associated with increased 30-day mortality compared with standard therapy (adjusted odds ratio 2.82; 95% confidence interval 1.20-6.66). The results of the sensitivity analyses were consistent with those of the primary analysis. CONCLUSION: Physicians should assess the risk for DRPs when determining the empirical antibiotic therapy and should refrain from administering unnecessary extended-spectrum antibiotics for patients with CAP with a low risk for DRPs.

Who are the long-term survivors of recurrent ovarian carcinoma?: a retrospective analysis of a multicenter study.

BACKGROUND: The aim of the present study was to investigate the incidence and hallmarks of long-term survivors of recurrent ovarian carcinoma (LTSROC) in a large-scale retrospective cohort of patients from a multicenter study group. METHODS: We performed a regional multicenter retrospective study between January 1986 and September 2021 using clinical data collected under the central pathological review system. Patients who underwent surgery for primary OC at diagnosis and developed recurrent tumors after the initial treatment were included. We defined LTSROC as patients who survived for 5 years or longer after initial tumor recurrence and examined factors affecting the long-term survival of ROC and outcomes of LTSROC. RESULTS: We collected information on patients with malignant ovarian tumors and finally 657 of them that developed ROC were included in the study population. Sixty-eight (10.4%) patients were LTSROC while 399 (60.7%) were short-term survivors of recurrent ovarian carcinoma. In a multivariate logistic regression analysis, negative ascites cytology [odds ratio (OR) 1.865; 95% CI 1.026-3.393; p = 0.041] and a recurrence-free interval (RFI) of 1 year or longer (OR 2.896; 95% CI 1.546-5.425; p < 0.001) were identified as independent factors associated with LTSROC. Approximately 80% of LTSROC presented with solitary recurrent tumors. Furthermore, more than 50% of LTSROC underwent tumor debulking surgery for the first recurrent tumor with or without chemotherapy. CONCLUSION: RFI of 1 year or longer and negative ascites cytology in the initial surgery were identified as independent predictive factors for LTSROC.

Obesity contributes to the stealth peritoneal dissemination of ovarian cancer: a multi-institutional retrospective cohort study.

OBJECTIVE: The clinical significance of a higher BMI on the prognosis of ovarian cancer remains controversial; therefore, a more detailed analysis is demanded. This study investigated the impact of BMI on peritoneum-specific recurrence to clarify the involvement of adipose tissue in the proliferation of cancer cells at sites of peritoneal dissemination. METHODS: Among 4,730 patients with malignant ovarian tumors, 280 diagnosed with International Federation of Gynecology and Obstetrics (FIGO) stage IIB to IIIC epithelial ovarian cancer and who underwent complete resection in the primary surgery were included in the present study. RESULTS: There were 42, 201, and 37 women in the low, normal, and high BMI groups, respectively. Peritoneum-specific recurrence-free survival and overall survival were both significantly shorter in patients with a high BMI than in those with a normal BMI (p = 0.028 and 0.018, respectively). No significant differences were observed in the distribution of sites of recurrence between these two groups. A multivariate analysis identified obesity as an independent prognostic factor in addition to pT3 tumor staging and positive ascites cytology. CONCLUSIONS: Patients with a high BMI had a significantly worse prognosis than those with a normal BMI, and peritoneal adipose tissue may have contributed to this difference.

Prediction of Prehospital Change of the Cardiac Rhythm From Nonshockable to Shockable in Out-of-Hospital Patients With Cardiac Arrest: A Post Hoc Analysis of a Nationwide, Multicenter, Prospective Registry.

Background Predicting a spontaneous rhythm change from nonshockable to shockable before hospital arrival in patients with out-of-hospital cardiac arrest can help emergency medical services develop better strategies for prehospital treatment. The aim of this study was to identify predictors of spontaneous rhythm change before hospital arrival in patients with out-of-hospital cardiac arrest and develop a predictive scoring system. Methods and Results We retrospectively reviewed data of eligible patients with out-of-hospital cardiac arrest with an initial nonshockable rhythm registered in a nationwide registry between June 2014 and December 2017. We performed a multivariable analysis using a Cox proportional hazards model to identify predictors of a spontaneous rhythm change, and a ridge regression model for predicting it. The data of 25 804 patients were analyzed (derivation cohort, n=17 743; validation cohort, n=8061). The rhythm change event rate was 4.1% (724/17 743) in the derivation cohort, and 4.0% (326/8061) in the validation cohorts. Age, sex, presence of a witness, initial rhythm, chest compression by a bystander, shock with an automated external defibrillator by a bystander, and cause of the cardiac arrest were all found to be independently associated with spontaneous rhythm change before hospital arrival. Based on this finding, we developed and validated the Rhythm Change Before Hospital Arrival for Nonshockable score. The Harrell's concordance index values of the score were 0.71 and 0.67 in the internal and external validations, respectively. Conclusions Seven factors were identified as predictors of a spontaneous rhythm change from nonshockable to shockable before hospital arrival. We developed and validated a score to predict rhythm change before hospital arrival.

Machine learning prediction models for postpartum depression: A multicenter study in Japan.

AIM: Postpartum depression (PPD) and perinatal mental health care are of growing importance worldwide. Here we aimed to develop and validate machine learning models for the prediction of PPD, and to evaluate the usefulness of the recently adopted 2-week postpartum checkup in some parts of Japan for the identification of women at high risk of PPD. METHODS: A multicenter retrospective study was conducted using the clinical data of 10 013 women who delivered at ≥35 weeks of gestation at 12 maternity care hospitals in Japan. PPD was defined as an Edinburgh Postnatal Depression Scale score of ≥9 points at 4 weeks postpartum. We developed prediction models using conventional logistic regression and four machine learning algorithms based on the information that can be routinely collected in daily clinical practice. The model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). RESULTS: In the machine learning models developed using clinical data before discharge, the AUROCs were similar to those in the conventional logistic regression models (AUROC, 0.569-0.630 vs. 0.626). The incorporation of additional 2-week postpartum checkup data into the model significantly improved the predictive performance for PPD compared to that without in the Ridge regression and Elastic net (AUROC, 0.702 vs. 0.630 [p < 0.01] and 0.701 vs. 0.628 [p < 0.01], respectively). CONCLUSIONS: Our machine learning models did not achieve better predictive performance for PPD than conventional logistic regression models. However, we demonstrated the usefulness of the 2-week postpartum checkup for the identification of women at high risk of PPD.

The rCAST score is useful for estimating the neurological prognosis in pediatric patients with post-cardiac arrest syndrome before ICU admission: External validation study using a nationwide prospective registry.

INTRODUCTION: The objective of this cohort study was to investigate whether the revised post-Cardiac Arrest Syndrome for Therapeutic hypothermia score (rCAST), which we previously developed as a prognostic score for adult patients with post-cardiac arrest syndrome (PCAS), is also applicable to pediatric patients. METHODS: Pediatric PCAS patients were included from an out-of-hospital cardiac arrest (OHCA) registry of the Japanese Association for Acute Medicine (JAAM). We validated the predictive accuracy of the rCAST for the neurological outcomes at 30 and 90 days. We also evaluated the probability of a good neurological outcome in each of the three specified severity categories based on the rCAST (low severity: ≤5.5; moderate severity: 6.0-14.0; high severity: ≥14.5). RESULTS: Among the 737 pediatric patients with OHCA, the data of 179 pediatric PCAS patients in whom return of spontaneous circulation was achieved were analyzed. The areas under the curve (AUC) of the rCAST for predicting the neurological outcomes at 30 days and 90 days were 0.95 (95% CI: 0.90-0.99) and 0.96 (0.91-1.00), respectively. The proportions of patients with a good neurological outcome at 30 days were 100% (12/12) in the low severity group, 36.1% (13/36) in the moderate severity group, and 2.3% (3/131) in the high severity group. CONCLUSIONS: The AUC of the rCAST for pediatric PCAS patients was found to be greater than 0.9 in the external validation, which corresponds to excellent predictive accuracy. There was no patient with good neurological outcome among the patients with more than 17.0 points (extremely high severity group).

A large-scale multi-institutional study evaluating prognostic aspects of positive ascites cytology and effects of therapeutic interventions in epithelial ovarian cancer.

Positive ascites cytology is a strong prognostic factor in patients with early-stage ovarian cancer (OvCa). However, limited information is currently available on the impact of positive ascites cytology on patient prognoses under each clinical background. We herein investigated the comprehensive impact of positive ascites cytology on patients with epithelial OvCa and the effectiveness of additional therapeutic interventions, including complete staging surgery and chemotherapy. Among 4730 patients with malignant ovarian neoplasms, retrospectively identified in multiple institutions, 1906 with epithelial OvCa were included. In the investigation of its effects on clinical factors using a multivariate analysis, positive ascites cytology correlated with a poor prognosis. Positive ascites cytology had a significantly worse prognosis than those with negative cytology in all subgroups except for patients with stage IV tumors and a mucinous histology. Chemotherapy may be effective in reducing the negative impact of positive ascites cytology on the prognosis of patients in terms of progression-free and overall survivals, while complete staging surgery did not improve the prognosis of patients with positive ascites cytology. Collectively, our findings suggested that positive ascites cytology had a negative impact on the prognosis of patients with epithelial OvCa, but not those with stage IV tumors or a mucinous histology.

The interplay between bystander cardiopulmonary resuscitation and ambient temperature on neurological outcome after cardiac arrest: A nationwide observational cohort study.

BACKGROUND: At lower ambient temperature, patients with out-of-hospital cardiac arrest (OHCA) easily experience hypothermia. Hypothermia has shown to improve the rate of successful return of spontaneous circulation (ROSC) in animal models. We hypothesized that lower temperature affects the impact of bystander cardiopulmonary resuscitation (CPR) on the increased odds of a favorable neurological outcome post-OHCA. METHODS: This study used information collected by the prospective, nationwide, Utstein registry to examine data from 352,689 adult patients who experienced OHCA from 2012 to 2016 in Japan. The primary outcome was a 1-month favorable neurological outcomes. Multivariable logistic regression analyses were conducted to test the impact of bystander CPR according to the temperature on the favorable outcome. RESULTS: A total of 201,111 patients with OHCA were included in the complete case analysis. The lower temperature group had lower proportions of receiving bystander CPR (46.5 vs. 47.9%) and having favorable outcome (2.1 vs 2.8%) than those in the higher group. Multivariable analysis revealed that bystander CPR at lower temperatures was significantly associated with favorable outcomes (adjusted odds ratio, 1.22; 95% CI, 1.09-1.37), whereas bystander CPR at higher temperatures was not associated with favorable outcomes (1.02; 0.92-1.13). The nonlinear relationship using a spline curve in the multivariable model revealed that odds ratio of favorable neurological outcomes associated with bystander CPR increased as the temperature decreased. CONCLUSION: Bystander CPR was associated with favorable neurological outcomes at lower temperatures. The odds of a favorable outcome associated with bystander CPR increased as the temperature decreased.

Outcome Related to Level of Targeted Temperature Management in Postcardiac Arrest Syndrome of Low, Moderate, and High Severities: A Nationwide Multicenter Prospective Registry.

OBJECTIVES: The optimal target temperature during targeted temperature management for patients after cardiac arrest remains under debate. The aim of this study was to evaluate the association between targeted temperature management at lower target temperatures and the neurologic outcomes among patients classified by the severity of postcardiac arrest syndrome. DESIGN: A multicenter observational study from the out-of-hospital cardiac arrest registry of the Japanese Association for Acute Medicine, which is a nationwide prospective registry of out-of-hospital cardiac arrest patients. SETTING: A total of 125 critical care medical centers or hospitals with an emergency care department across Japan. PATIENTS: A total of 1,111 out-of-hospital cardiac arrest patients who had received targeted temperature management. MEASUREMENTS AND MAIN RESULTS: We divided all 1,111 postcardiac arrest syndrome patients treated with targeted temperature management into two groups: those who received targeted temperature management at a lower target temperature (33-34°C) and those who received targeted temperature management at a higher target temperature (35-36°C). In regard to classification of the patients, we divided the patients into three categories of severity (low, moderate, and high severities) using the risk classification tool, post-Cardiac Arrest Syndrome for Therapeutic hypothermia, which was previously validated. The primary outcome was the percentage of patients with a good neurologic outcome at 30 days, and the secondary outcome was the survival rate at 30 days. Multivariate analysis showed that targeted temperature management at 33-34°C was significantly associated with a good neurologic outcome and survival at 30 days in the moderate severity (odds ratio, 1.70 [95% CI, 1.03-2.83] and 1.90 [95% CI, 1.15-3.16], respectively), but not in the patients of low or high severity (pinteraction = 0.033). Propensity score analysis also showed that targeted temperature management at 33-34°C was associated with a good neurologic outcome in the moderate-severity group (p = 0.022). CONCLUSIONS: Targeted temperature management at 33-34°C was associated with a significantly higher rate of a good neurologic outcome in the moderate-severity postcardiac arrest syndrome group, but not in the low- or high-severity group.

Postoperative stroke and neurological outcomes in the early phase after revascularization surgeries for moyamoya disease: an age-stratified comparative analysis.

Stroke and neurological outcomes in the early phase following revascularization for moyamoya disease (MMD) may depend on the patient's age. In this study, an age-stratified comparative analysis was performed to clarify this issue. We reviewed 105 MMD patients who underwent 179 revascularization surgeries. The demographic characteristics were collected in four age groups (≤ 5 and 6-17 years for pediatric patients and 18-49 and ≥ 50 years for adults). Additionally, we assessed the incidence of subsequent stroke and deterioration of modified Rankin Scale (mRS) score. Then, we evaluated predictors of postoperative stroke and mRS deterioration using logistic regression. The mean patient age was 26.2 ± 18.5 years. No significant difference in the incidence of postoperative stroke was observed between age groups; however, the incidence tended to be increased among patients aged ≤ 5 years (17.9%) and patients aged ≥ 50 years (16.7%). Deterioration of mRS scores was significantly associated with ages ≤ 5 years (17.9%) and ≥ 50 years (11.1%). Logistic regression showed that posterior cerebral artery involvement (odds ratio [OR], 4.6) and postoperative transient neurological events (TNEs) (OR, 5.93) were risk factors for postoperative stroke. Age ≤ 5 years (OR, 9.73), postoperative TNEs (OR, 7.38), and postoperative stroke (OR, 49) were identified as predictors of unfavorable neurological outcomes. The novel feature of this comparative analysis by age group is that membership in the early-childhood MMD patient group (under 5 years old) was an independent risk factor for unfavorable short-term neurological outcomes and was mainly associated with the incidence of postoperative severe cerebral infarction.

Semi-parametric empirical Bayes factor for genome-wide association studies.

Bayes factor analysis has the attractive property of accommodating the risks of both false negatives and false positives when identifying susceptibility gene variants in genome-wide association studies (GWASs). For a particular SNP, the critical aspect of this analysis is that it incorporates the probability of obtaining the observed value of a statistic on disease association under the alternative hypotheses of non-null association. An approximate Bayes factor (ABF) was proposed by Wakefield (Genetic Epidemiology 2009;33:79-86) based on a normal prior for the underlying effect-size distribution. However, misspecification of the prior can lead to failure in incorporating the probability under the alternative hypothesis. In this paper, we propose a semi-parametric, empirical Bayes factor (SP-EBF) based on a nonparametric effect-size distribution estimated from the data. Analysis of several GWAS datasets revealed the presence of substantial numbers of SNPs with small effect sizes, and the SP-EBF attributed much greater significance to such SNPs than the ABF. Overall, the SP-EBF incorporates an effect-size distribution that is estimated from the data, and it has the potential to improve the accuracy of Bayes factor analysis in GWASs.

Desmoid with biweekly methotrexate and vinblastine shows similar effects to weekly administration: A phase II clinical trial.

Low-dose methotrexate (MTX) plus vinblastine (VBL) chemotherapy is an effective treatment for desmoid-type fibromatosis (DF). However, previous reports have described a weekly regimen, with no reports available on a biweekly one. The aim of this study was to determine the clinical outcomes of a biweekly regimen in a cohort prospectively treated in our single institution. Since 2010, we have prospectively treated refractory DF patients with biweekly MTX (30 mg/m2 ) + VBL (6 mg/m2 ). Efficacy, progression-free survival (PFS), and correlating factors were analyzed. Adverse events (AEs) were recorded. In total, 38 patients received low-dose MTX + VBL therapy, and its efficacy was assessed in 37 of them. Nineteen (51%) patients showed partial response (PR). Clinical benefit rate was 95%. PFS at 5 y was 80.8%. In PR cases, median time to response was 10 mo. Longer duration of therapy was significantly associated with the response of PR (P = .007) by univariate analysis. There was no clear association between various clinicopathological factors, including tumor size, location, catenin beta-1 (CTNNB1) mutation status with effect. Only 3 AEs of grade 3/4 were observed. Tumor regrowth after MTX + VBL discontinuation was observed in 5 (20%) of 25 patients. Biweekly administration of MTX + VBL chemotherapy was well tolerated compared with weekly administration, and its efficacy was anticipated in DF patents, although the time needed to achieve a response may be relatively long. The treatment interval should be determined taking into account both the condition of the tumor and the patient's preference.

Effect-Size Estimation Using Semiparametric Hierarchical Mixture Models in Disease-Association Studies with Neuroimaging Data.

In disease-association studies using neuroimaging data, evaluating the biological or clinical significance of individual associations requires not only detection of disease-associated areas of the brain but also estimation of the magnitudes of the associations or effect sizes for individual brain areas. In this paper, we propose a model-based framework for voxel-based inferences under spatial dependency in neuroimaging data. Specifically, we employ hierarchical mixture models with a hidden Markov random field structure to incorporate the spatial dependency between voxels. A nonparametric specification is proposed for the effect size distribution to flexibly estimate the underlying effect size distribution. Simulation experiments demonstrate that compared with a naive estimation method, the proposed methods can substantially reduce the selection bias in the effect size estimates of the selected voxels with the greatest observed associations. An application to neuroimaging data from an Alzheimer's disease study is provided.

Two-stage analysis for selecting fixed numbers of features in omics association studies.

One of main roles of omics-based association studies with high-throughput technologies is to screen out relevant molecular features, such as genetic variants, genes, and proteins, from a large pool of such candidate features based on their associations with the phenotype of interest. Typically, screened features are subject to validation studies using more established or conventional assays, where the number of evaluable features is relatively limited, so that there may exist a fixed number of features measurable by these assays. Such a limitation necessitates narrowing a feature set down to a fixed size, following an initial screening analysis via multiple testing where adjustment for multiplicity is made. We propose a two-stage screening approach to control the false discovery rate (FDR) for a feature set with fixed size that is subject to validation studies, rather than for a feature set from the initial screening analysis. Out of the feature set selected in the first stage with a relaxed FDR level, a fraction of features with most statistical significance is firstly selected. For the remaining feature set, features are selected based on biological consideration only, without regard to any statistical information, which allows evaluating the FDR level for the finally selected feature set with fixed size. Improvement of the power is discussed in the proposed two-stage screening approach. Simulation experiments based on parametric models and real microarray datasets demonstrated substantial increment in the number of screened features for biological consideration compared with the standard screening approach, allowing for more extensive and in-depth biological investigations in omics association studies.