Machine Learning Approaches to Predict Hepatotoxicity Risk in Patients Receiving Nilotinib

Background: Although nilotinib hepatotoxicity can cause severe clinical conditions and may alter treatment plans, risk factors affecting nilotinib-induced hepatotoxicity have not been investigated. This study aimed to elucidate the factors affecting nilotinib-induced hepatotoxicity. Methods: This retrospective cohort study was performed on patients using nilotinib from July of 2015 to June of 2020. We estimated the odds ratio and adjusted odds ratio from univariate and multivariate analyses, respectively. Several machine learning models were developed to predict risk factors of hepatotoxicity occurrence. The area under the curve (AUC) was analyzed to assess clinical performance. Results: Among 353 patients, the rate of patients with grade I or higher hepatotoxicity after nilotinib administration was 40.8%. Male patients and patients who received nilotinib at a dose of ≥300 mg had a 2.3-fold and a 3.5-fold increased risk for hepatotoxicity compared to female patients and compared with those who received <300 mg, respectively. H2 blocker use decreased hepatotoxicity by 11.6-fold. The area under the curve (AUC) values of machine learning methods ranged between 0.61–0.65 in this study. Conclusion: This study suggests that the use of H2 blockers was a reduced risk of nilotinib-induced hepatotoxicity, whereas male gender and a high dose were associated with increased hepatotoxicity.

Download Full-text

sLOX-1: A Molecule for Evaluating the Prognosis of Recurrent Ischemic Stroke

Neural Plasticity ◽

10.1155/2021/6718184 ◽

2021 ◽

Vol 2021 ◽

pp. 1-9

Author(s):

Yangmin Zheng ◽

Yuyou Huang ◽

Lingzhi Li ◽

Pingping Wang ◽

Rongliang Wang ◽

...

Keyword(s):

Risk Factors ◽

Recurrent Stroke ◽

Characteristic Curve ◽

Low Density Lipoprotein ◽

Area Under The Curve ◽

Density Lipoprotein ◽

Factors Affecting ◽

Independent Prognostic Marker ◽

Increased Risk ◽

Scale Scores

Several clinical parameters and biomarkers have been proposed as prognostic markers for stroke. However, it has not been clarified whether the risk factors affecting the prognosis of patients with recurrent and first-ever stroke are similar. In this study, we aimed to explore the relationship between soluble lectin-like oxidized low-density lipoprotein receptor 1 (sLOX-1) levels and the prediction of the functional outcome in patients with recurrent and first-ever stroke. A total of 266 patients with recurrent and first-ever stroke, who underwent follow-up for 3 months, were included in this study. Plasma samples were collected within 24 h after onset. The results showed that biomarkers for the prognosis of patients with recurrent stroke were different from that of those with first-ever stroke. sLOX-1 levels were correlated with modified Rankin Scale scores of patients with recurrent stroke alone ( r = 0.3232 , p = 0.001 ). sLOX-1 levels were also associated with an increased risk of unfavorable outcomes in patients with recurrent stroke with an adjusted odds ratio of 1.489 (95% confidence interval, 1.204–1.842, p < 0.0001 ). Combining the risk factors showed greater accuracy for prognosis, yielding a sensitivity of 93.2% and a specificity of 75%, with an area under the curve of 0.916, evaluated by the receiver operating characteristic curve. These findings suggest that the diagnosis and prognosis are different between patients with recurrent stroke and those with first-ever stroke, and sLOX-1 level is an independent prognostic marker in patients with recurrent stroke.

Download Full-text

Importance of GWAS risk loci and clinical data in predicting asthma using machine-learning approaches

10.21203/rs.3.rs-21271/v1 ◽

2020 ◽

Author(s):

Si-Qiao Liang ◽

Jian-Xiong Long ◽

Jingmin Deng ◽

Xuan Wei ◽

Mei-Ling Yang ◽

...

Keyword(s):

Machine Learning ◽

Risk Factors ◽

Clinical Data ◽

Genome Wide Association Study ◽

Prediction Models ◽

Area Under The Curve ◽

Gradient Boosting ◽

Support Vector ◽

Learning Approaches ◽

Extreme Gradient Boosting

Abstract Asthma is a serious immune-mediated respiratory airway disease. Its pathological processes involve genetics and the environment, but it remains unclear. To understand the risk factors of asthma, we combined genome-wide association study (GWAS) risk loci and clinical data in predicting asthma using machine-learning approaches. A case–control study with 123 asthma patients and 100 healthy controls was conducted in Zhuang population in Guangxi. GWAS risk loci were detected using polymerase chain reaction, and clinical data were collected. Machine-learning approaches (e.g., extreme gradient boosting [XGBoost], decision tree, support vector machine, and random forest algorithms) were used to identify the major factors that contributed to asthma. A total of 14 GWAS risk loci with clinical data were analyzed on the basis of 10 times of 10-fold cross-validation for all machine-learning models. Using GWAS risk loci or clinical data, the best performances were area under the curve (AUC) values of 64.3% and 71.4%, respectively. Combining GWAS risk loci and clinical data, the XGBoost established the best model with an AUC of 79.7%, indicating that the combination of genetics and clinical data can enable improved performance. We then sorted the importance of features and found that the top six risk factors for predicting asthma were rs3117098, rs7775228, family history, rs2305480, rs4833095, and body mass index. Asthma-prediction models based on GWAS risk loci and clinical data can accurately predict asthma and thus provide insights into the disease pathogenesis of asthma. Further research is required to evaluate more genetic markers and clinical data and predict asthma risk.

Download Full-text

A proof-of-concept study applying machine learning methods to putative risk factors for eating disorders: results from the multi-centre European project on healthy eating

Psychological Medicine ◽

10.1017/s003329172100489x ◽

2021 ◽

pp. 1-10

Author(s):

I. Krug ◽

J. Linardon ◽

C. Greenwood ◽

G. Youssef ◽

J. Treasure ◽

...

Keyword(s):

Machine Learning ◽

Risk Factors ◽

Logistic Regression ◽

Predictive Accuracy ◽

Area Under The Curve ◽

Prediction Rule ◽

Predictive Performance ◽

Individual Risk ◽

European Project ◽

Wide Range

Abstract Background Despite a wide range of proposed risk factors and theoretical models, prediction of eating disorder (ED) onset remains poor. This study undertook the first comparison of two machine learning (ML) approaches [penalised logistic regression (LASSO), and prediction rule ensembles (PREs)] to conventional logistic regression (LR) models to enhance prediction of ED onset and differential ED diagnoses from a range of putative risk factors. Method Data were part of a European Project and comprised 1402 participants, 642 ED patients [52% with anorexia nervosa (AN) and 40% with bulimia nervosa (BN)] and 760 controls. The Cross-Cultural Risk Factor Questionnaire, which assesses retrospectively a range of sociocultural and psychological ED risk factors occurring before the age of 12 years (46 predictors in total), was used. Results All three statistical approaches had satisfactory model accuracy, with an average area under the curve (AUC) of 86% for predicting ED onset and 70% for predicting AN v. BN. Predictive performance was greatest for the two regression methods (LR and LASSO), although the PRE technique relied on fewer predictors with comparable accuracy. The individual risk factors differed depending on the outcome classification (EDs v. non-EDs and AN v. BN). Conclusions Even though the conventional LR performed comparably to the ML approaches in terms of predictive accuracy, the ML methods produced more parsimonious predictive models. ML approaches offer a viable way to modify screening practices for ED risk that balance accuracy against participant burden.

Download Full-text

Positive Signs on Physical Examination Are Not Always Indications for Endotracheal Tube Intubation in Patients with Facial Burn

10.21203/rs.3.rs-128508/v1 ◽

2020 ◽

Author(s):

Ruo-Yi Huang ◽

Szu-Jen Chen ◽

Yen-Chang Hsiao ◽

Ling-Wei Kuo ◽

Chien-Hung Liao ◽

...

Keyword(s):

Risk Factors ◽

Endotracheal Tube ◽

Odds Ratio ◽

Inhalation Injury ◽

Burn Patients ◽

Shortness Of Breath ◽

Airway Protection ◽

Increased Risk ◽

Independent Risk Factors ◽

Facial Burn

Abstract BackgroundAfter clinical evaluation in the emergency department (ED), facial burn patients are usually intubated to protect their airways. However, the possibility of unnecessary intubation or delayed intubation after admission exists. Objective criteria for the evaluation of inhalation injury and the need for airway protection in facial burn patients are needed.MethodsFacial burn patients between January 2013 and May 2016 were reviewed. Patients who were and were not intubated in the ED were compared. All intubated patients received routine bronchoscopy to evaluate whether they had inhalation injuries. Patients with and without confirmed inhalation injuries were compared. Multivariate logistic regression analysis was used to identify the independent risk factors for inhalation injuries in facial burn patients. The reasons for intubation in patients without inhalation injuries were also investigated.ResultsDuring the study period, 121 patients were intubated in the ED among a total of 335 facial burn patients. Only 73 (60.3%) patients were later confirmed to have inhalation injuries on bronchoscopy. The comparison between patients with and without inhalation injuries showed that shortness of breath (odds ratio=3.376, p=0.027) and high total body surface area (TBSA) (odds ratio=1.038, p=0.001) were independent risk factors for inhalation injury. Other physical signs (e.g., hoarseness, burned nostril hair, etc.), laboratory examinations and chest X-ray findings were not predictive of inhalation injury in facial burn patients. All patients with a TBSA over 60% were intubated in the ED even if they did not have inhalation injuries.ConclusionIn the management of facial burn patients, positive signs on conventional physical examinations may not always be predictive of inhalation injury and the need for endotracheal tube intubation in the ED. More attention should be paid to facial burn patients with shortness of breath and a high TBSA because they have an increased risk of inhalation injuries. Airway protection is needed in facial burn patients without inhalation injuries because of their associated injuries and treatment.

Download Full-text

Stroke in hemodialysis patients randomized to different intravenous iron strategies: a prespecified analysis from the PIVOTAL trial

Kidney360 ◽

10.34067/kid.0004272021 ◽

2021 ◽

pp. 10.34067/KID.0004272021

Author(s):

Patrick B. Mark ◽

Pardeep S. Jhund ◽

Matthew R. Walters ◽

Mark C. Petrie ◽

Albert Power ◽

...

Keyword(s):

Risk Factors ◽

Stroke Risk ◽

Controlled Trial ◽

Intravenous Iron ◽

Hemodialysis Patients ◽

High Dose ◽

Pivotal Trial ◽

Increased Risk ◽

Iv Iron

Background: People with kidney failure treated with hemodialysis (HD) are at increased risk of stroke compared to similarly aged people with normal kidney function. One concern is that treatment of renal anemia might increase stroke risk. We studied risk factors for stroke in a prespecified secondary analysis of a randomized controlled trial of intravenous iron treatment strategies in HD. Methods: We analyzed data from the Proactive IV IrOn Therapy in HaemodiALysis Patients (PIVOTAL) trial focusing on variables associated with risk of stroke. The trial randomized 2,141 adults, who had started hemodialysis <12 months earlier and who were receiving an erythropoiesis-stimulating agent (ESA), to high-dose IV iron administered proactively or low-dose IV iron administered reactively in a 1:1 ratio. Possible stroke events were independently adjudicated. We performed analyses to identify variables associated with stroke during follow-up and assessed survival following stroke. Results: During a median 2.1 years follow-up, 69 (3.2%) patients experienced a first post randomization stroke. 57 (82.6%) were ischemic strokes and 12 (17.4%) hemorrhagic strokes. There were 34 post randomization strokes in the proactive arm and 35 in the reactive arm (hazard ratio (95% confidence interval): 0.90 (0.56, 1.44), p=0.66). In multivariable models, female gender, diabetes, history of prior stroke at baseline, higher baseline systolic blood pressure, lower serum albumin and higher C-reactive protein were independently associated with stroke events during follow up. Hemoglobin, total iron or ESA dose were not associated with risk of stroke. 58% of patients with a stroke event died during follow-up, compared to 23% without a stroke. Conclusions: In hemodialysis patients, stroke risk is broadly associated with risk factors previously described to increase cardiovascular risk in this population. Proactive intravenous iron does not increase stroke risk.

Download Full-text

Retrospective analysis of risk factors of hypotensive bradycardic events during shoulder arthroscopic surgery under interscalene blockade in the sitting position

Korean Journal of Anesthesiology ◽

10.4097/kja.20035 ◽

2020 ◽

Vol 73 (6) ◽

pp. 542-549

Author(s):

Taeha Ryu ◽

Baek Jin Kim ◽

Seong Jun Woo ◽

So Young Lee ◽

Jung A Lim ◽

...

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Odds Ratio ◽

Arthroscopic Surgery ◽

Sitting Position ◽

Past Medical History ◽

Frequent Adverse Event ◽

Increased Risk ◽

Potential Risk Factors ◽

Independent Risk Factors

Background: Hypotensive bradycardic events (HBEs) are a frequent adverse event in patients who underwent shoulder arthroscopic surgery under interscalene block (ISB) in the sitting position. This retrospective study was conducted to investigate the independent risk factors of HBEs in shoulder arthroscopic surgery under ISB in the sitting position. Methods: A total of 2549 patients who underwent shoulder arthroscopic surgery under ISB and had complete clinical data were included in the study. The 357 patients who developed HBEs were included in the HBEs group, and the remaining 2192 in the non-HBEs group. The potential risk factors for HBEs, such as age, sex, past medical history, anesthetic characteristics, and intraoperative medications were collected and compared between the groups. Statistically significant variables were included in a logistic regression model to further evaluate the independent risk factors for HBEs in shoulder arthroscopic surgery under ISB. Results: The incidence of HBEs was 14.0% (357/2549). Logistic regression analysis revealed that the intraoperative use of hydralazine (odds ratio [OR] 4.2; 95% confidence interval [CI] 2.9–6.3), propofol (OR 2.1; 95% CI 1.3–3.6), and dexmedetomidine (OR 3.9; 95% CI 1.9–7.8) before HBEs were independent risk factors for HBEs in patients who received shoulder arthroscopic surgery under ISB. Conclusions: The intraoperative use of antihypertensives such as hydralazine and sedatives such as propofol or dexmedetomidine leads to increased risk of HBEs during shoulder arthroscopic surgery under ISB in the sitting position.

Download Full-text

Bortezomib administration is a risk factor associated with the development of tumor lysis syndrome in male patients with multiple myeloma: A retrospective study

10.21203/rs.3.rs-22557/v4 ◽

2020 ◽

Author(s):

Masahiro Kondo ◽

Yuji Hotta ◽

Karen Yamauchi ◽

Akimasa Sanagawa ◽

Hirokazu Komatsu ◽

...

Keyword(s):

Risk Factors ◽

Multiple Myeloma ◽

Odds Ratio ◽

Tumor Lysis Syndrome ◽

Low Risk ◽

Novel Therapies ◽

Tumor Lysis ◽

Factors Associated ◽

Increased Risk ◽

Male Patients

Abstract Background: Novel agents such as proteasome inhibitors have been developed for several years to treat multiple myeloma. Although multiple myeloma is a low-risk disease for developing tumor lysis syndrome (TLS), treatment with these novel therapies might increase TLS risk. Previous studies, mostly case reports or case series, have reported bortezomib-induced TLS in patients with multiple myeloma. This study aimed to investigate risk factors associated with TLS development in multiple myeloma patients.Methods: We retrospectively investigated incidences of laboratory and clinical TLS (LTLS and CTLS, respectively) in patients who received primary therapy for treatment-naive, symptomatic multiple myeloma between May 2007 and January 2018. We used multivariate logistic regression analyses to evaluate the associations between TLS and several parameters previously reported to be associated with increased risk.Results: This study included 210 patients with multiple myeloma, of which ten (4.8%) had LTLS and seven (3.3%) had CTLS. The characteristics of the administered anticancer or prophylactic antihyperuricemic agents were similar between patients with and without TLS. Multivariate analyses revealed that TLS was most strongly associated with bortezomib-containing therapy (odds ratio = 3.40, P = 0.069), followed by male sex (odds ratio = 2.29, P = 0.153). In a subgroup analysis focused on men, treatment with bortezomib-containing therapy was significantly associated with increased risk of TLS (odds ratio = 8.51, P = 0.046).Conclusion: In the present study, we investigated the risk factors associated with TLS development in 210 multiple myeloma patients, which, to the best of our knowledge, is the largest number of patients reported to date. Furthermore, this study is the first to evaluate TLS risk factors in MM by adjusting for the effects of potential confounding factors in patients’ backgrounds. Consequently, we found that bortezomib-containing therapy increases the risk of TLS in male patients with multiple myeloma. TLS risk should be evaluated further in low-risk diseases such as multiple myeloma, since a significant number of novel therapies can achieve high antitumor responses.

Download Full-text

Abstract WP164: Role of Risk Factors in the Paradoxical Effect of Smoking on Peri-procedural Stroke in SAMMPRIS

Stroke ◽

10.1161/str.44.suppl_1.awp164 ◽

2013 ◽

Vol 44 (suppl_1) ◽

Author(s):

Christine A Holmstedt ◽

Tanya N Turan ◽

Michael J Lynn ◽

Bethany F Lane ◽

Jean Montgomery ◽

...

Keyword(s):

Risk Factors ◽

Odds Ratio ◽

Multivariate Analyses ◽

Smoking Status ◽

Active Form ◽

Intracranial Stenosis ◽

Contributory Factor ◽

Increased Risk ◽

Never Smokers

Background: A previous SAMMPRIS analysis of patients randomized to stenting showed that peri-procedural ischemic infarcts were significantly associated with diabetes, basilar stenosis, age, and smoking status with never smokers having a higher risk (odds ratio = 8.8, p< 0.001). We sought to determine if this finding could be due to a higher burden of other risk factors in never smokers. Method: Baseline features in 213 patients undergoing stenting in SAMMPRIS were compared between never smokers vs. former and current smokers in univariate and multivariate analyses. Logistic regression was used to determine the effect of smoking on peri-procedural ischemic infarcts after adjusting for factors related to smoking. Data: Univariate results are shown in Table 1. Never smokers were significantly (P<0.05) more likely to be female, diabetic, hypertensive, and have another intracranial stenosis, but in multivariate analyses only hypertension and another intracranial stenosis remained significantly (P<0.05) associated with smoking status. In a multivariate model that incorporated hypertension and another intracranial stenosis along with smoking status, diabetes, basilar stenosis, and age, smoking status remained significant with an increased risk among patients who never smoked (odds ratio = 5.3, p = 0.005). Conclusion: While never smokers had significantly higher rates of some risk factors compared to active or previous smokers, these risk factors do not explain all the increased risk of early stroke in never smokers after stenting in SAMMPRIS. Another contributory factor may be that smoking accelerates the conversion of clopidogrel to its active form.

Download Full-text

Computational prediction of implantation outcome after embryo transfer

Health Informatics Journal ◽

10.1177/1460458219892138 ◽

2019 ◽

Vol 26 (3) ◽

pp. 1810-1826 ◽

Cited By ~ 3

Author(s):

Behnaz Raef ◽

Masoud Maleki ◽

Reza Ferdousi

Keyword(s):

Machine Learning ◽

Feature Selection ◽

Prediction Model ◽

Embryo Transfer ◽

Area Under The Curve ◽

Computational Prediction ◽

Support Vector ◽

Human Menopausal Gonadotropin ◽

Optimum Number ◽

Learning Approaches

The aim of this study is to develop a computational prediction model for implantation outcome after an embryo transfer cycle. In this study, information of 500 patients and 1360 transferred embryos, including cleavage and blastocyst stages and fresh or frozen embryos, from April 2016 to February 2018, were collected. The dataset containing 82 attributes and a target label (indicating positive and negative implantation outcomes) was constructed. Six dominant machine learning approaches were examined based on their performance to predict embryo transfer outcomes. Also, feature selection procedures were used to identify effective predictive factors and recruited to determine the optimum number of features based on classifiers performance. The results revealed that random forest was the best classifier (accuracy = 90.40% and area under the curve = 93.74%) with optimum features based on a 10-fold cross-validation test. According to the Support Vector Machine-Feature Selection algorithm, the ideal numbers of features are 78. Follicle stimulating hormone/human menopausal gonadotropin dosage for ovarian stimulation was the most important predictive factor across all examined embryo transfer features. The proposed machine learning-based prediction model could predict embryo transfer outcome and implantation of embryos with high accuracy, before the start of an embryo transfer cycle.

Download Full-text

Clinical indicators of treatment-resistant psychosis

The British Journal of Psychiatry ◽

10.1192/bjp.2019.120 ◽

2019 ◽

Vol 216 (5) ◽

pp. 259-266 ◽

Cited By ~ 5

Author(s):

Sophie E. Legge ◽

Charlotte A. Dennison ◽

Antonio F. Pardiñas ◽

Elliott Rees ◽

Amy J. Lynham ◽

...

Keyword(s):

Machine Learning ◽

Social Adjustment ◽

Odds Ratio ◽

Age At Onset ◽

Learning Approaches ◽

Regression Analyses ◽

Treatment Resistant ◽

Genetic Liability ◽

Illness Onset ◽

Risk Of Treatment

BackgroundAround 30% of individuals with schizophrenia remain symptomatic and significantly impaired despite antipsychotic treatment and are considered to be treatment resistant. Clinicians are currently unable to predict which patients are at higher risk of treatment resistance.AimsTo determine whether genetic liability for schizophrenia and/or clinical characteristics measurable at illness onset can prospectively indicate a higher risk of treatment-resistant psychosis (TRP).MethodIn 1070 individuals with schizophrenia or related psychotic disorders, schizophrenia polygenic risk scores (PRS) and large copy number variations (CNVs) were assessed for enrichment in TRP. Regression and machine-learning approaches were used to investigate the association of phenotypes related to demographics, family history, premorbid factors and illness onset with TRP.ResultsYounger age at onset (odds ratio 0.94,P= 7.79 × 10−13) and poor premorbid social adjustment (odds ratio 1.64,P= 2.41 × 10−4) increased risk of TRP in univariate regression analyses. These factors remained associated in multivariate regression analyses, which also found lower premorbid IQ (odds ratio 0.98,P= 7.76 × 10−3), younger father's age at birth (odds ratio 0.97,P= 0.015) and cannabis use (odds ratio 1.60,P= 0.025) increased the risk of TRP. Machine-learning approaches found age at onset to be the most important predictor and also identified premorbid IQ and poor social adjustment as predictors of TRP, mirroring findings from regression analyses. Genetic liability for schizophrenia was not associated with TRP.ConclusionsPeople with an earlier age at onset of psychosis and poor premorbid functioning are more likely to be treatment resistant. The genetic architecture of susceptibility to schizophrenia may be distinct from that of treatment outcomes.

Download Full-text