Risk Stratification Protocols Research Articles

Predictors of operative ischemic cerebrovascular complications in skull base tumor resections: Experience in low-resource setting

Abstract Background Ischemic cerebrovascular complications following skull base tumor resections remain a significant factor impacting both short-term and long-term patient outcomes. This study aims to improve risk stratification, surgical decision-making, and postoperative care protocols. Methods A retrospective cohort study on predictors of ischemic cerebrovascular complications among patients who underwent skull base tumor resection was conducted at 2 high-volume neurosurgical centers in Ethiopia from 2018 to 2023. Binary logistic analysis was performed to see the association of each predictor variable. Results The study included 266 patients, with 65.5% being female. The median age and tumor size were 37 (± IQR = 17) years and 4.9 cm (± IQR 1.5), respectively. Ischemic cerebrovascular complications occurred in 19.9% of patients. Middle cranial fossa tumors and tumors spanning both anterior and middle cranial fossa (AOR = 6.75, 95% CI: 1.66–27.54, P < .008), grades 3–5 vascular encasement (AOR = 5.04, 95% CI: 1.79–14.12, P < .002), near-total resection and gross total resection (AOR = 2.89, 95% CI: 1.01–8.24, P < .048), and difficult hemostasis (AOR = 9.37, 95% CI: 3.19–27.52, P < .000) were significantly associated with iatrogenic vascular injury. Subarachnoid hemorrhage had a statistically significant association with vasospasm (AOR = 12.27, 95% CI: 1.99–75.37, P = .007). Conclusions Surgery-related ischemic cerebrovascular complications are common. Thorough perioperative risk stratification and proactive treatment planning are crucial to mitigate vascular insults associated with it. In low-resource settings, neurosurgical services are provided without advanced instruments, leading to more complications. Therefore, it is important to focus on improving neurosurgical setup to enhance patient outcomes.

An all-inclusive model for predicting invasive bacterial infection in febrile infants age 7-60 days.

Invasive bacterial infections (IBIs) in febrile infants are rare but potentially devastating. We aimed to derive and validate a predictive model for IBI among febrile infants age 7-60 days. Data were abstracted retrospectively from electronic records of 37 emergency departments (EDs) for infants with a measured temperature >=100.4 F who underwent an ED evaluation with blood and urine cultures. Models to predict IBI were developed and validated respectively using a random 80/20 dataset split, including 10-fold cross-validation. We used precision recall curves as the classification metric. Of 4411 eligible infants with a mean age of 37 days, 29% had characteristics that would likely have excluded them from existing risk stratification protocols. There were 196 patients with IBI (4.4%), including 43 (1.0%) with bacterial meningitis. Analytic approaches varied in performance characteristics (precision recall range 0.04-0.29, area under the curve range 0.5-0.84), with the XGBoost model demonstrating the best performance (0.29, 0.84). The five most important variables were serum white blood count, maximum temperature, absolute neutrophil count, absolute band count, and age in days. A machine learning model (XGBoost) demonstrated the best performance in predicting a rare outcome among febrile infants, including those excluded from existing algorithms. Several models for the risk stratification of febrile infants have been developed. There is a need for a preferred comprehensive model free from limitations and algorithm exclusions that accurately predicts IBIs. This is the first study to derive an all-inclusive predictive model for febrile infants aged 7-60 days in a community ED sample with IBI as a primary outcome. This machine learning model demonstrates potential for clinical utility in predicting IBI.

Co-relation of Portal Vein Tumour Thrombus Response With Survival Function Following Robotic Radiosurgery in Vascular Invasive Hepatocellular Carcinoma

Purpose or ObjectiveProspective evaluation of stereotactic body radiotherapy (SBRT) with Robotic Radiosurgery in hepatocellular carcinoma patients with macrovascular invasion (HCC-PVT). Materials and MethodsPatients with inoperable HCC-PVT, good performance score (PS0-1) and preserved liver function (up to Child Pugh B7) were accrued after ethical and scientific committee approval [CTRI: 2022/01/050234] for treatment on Robotic Radiosurgery (M6) and planned with Multiplan (iDMS V2.0). Triple-phase contrast computed tomography (CT) scan was done for contouring and gross tumor volume (GTV) included contrast enhancing mass within main portal vein and adjacent parenchymal disease. Dose prescription was as per risk stratification protocol (22- 50Gy in 5 fractions) while achieving the constraints of mean liver dose <15Gy, 800 cc liver <8Gy and the duodenum max of <24 Gy). Response assessment done at 2 months follow up for recanalization. Patient and treatment related factors evaluated for influence in survival function. ResultsBetween Jan 2017 till May 2022, 317 consecutive HCC with PVT patients were screened and 219 patients were accrued [male 92%, CP score 5-7 90%, mean age 63 years (38-85yrs), CLIP <3: 84 (40%), 3-6 117 (56%), infective etiology 9.5%, Performance status (PS) 0-37%; 1-56%]. Among 209 consecutive patients after accrual SBRT treatment done (10 patients were excluded after accrual due to ascites and decompensation), 139 were evaluable for response assessment (>2 mo follow up). At mean follow up of 10.2 months (SD 8.43), 88 (63%) patients expired and 51 (36%) were alive. 82 (59%) patients had recanalization of PVT (response), 57 (41%) pts did not recanalize, 28 (17%) had progressive disease had progressive/metastatic disease prior to response evaluation (<2 months). Mean overall survival in responders and non-responders were 18.4 (SE 2.52) and 9.34 month (SE 0.81) respectively (p<0.001). Mean survival in patients with PS0, PS1 and PS2 were 17, 11.7 and 9.7 months (p- 0.019) respectively. Overall survival (OS) in partial recanalization, bland thrombus and complete recanalization was 12.4, 14.1, 30.3 months respectively (p-0.002). Adjuvant sorafenib, Barcelona Clinic Liver Classification (BCLC) stage, gender, age, RT dose did not influence response to treatment. Recanalization rate was higher in good PS patients (p-0.019). OS in patients with response to treatment, no response to treatment, fit but not accrued and not suitable patients were 18.4, 9.34, 5.9, 2.6 months respectively (p-<0.001). 36/139 patients (24%) had RILD [10 (7.2%) had Classic Radiation induced liver disease (RILD) & 26 (19%) had non-classic RILD]. Derangement in Child Pugh Score (CP score change) by more than 2 was seen in 30 (24%) within 2-month period after Robotic Radiosurgery. 18 (13%) had unplanned admissions, two patients required embolization due to fiducial related bleeding, 20 (14%) had ascites of which 9 (6%) patients required abdominocentesis. ConclusionPVT response or recanalization after SBRT is a statistically significant prognostic factor for survival function in HCC-PVT.

Investigation of Risk Factors Predicting Cataract Surgery Complications in Patients with Pseudoexfoliation Syndrome: A Systematic Review.

(1) Background: The present review aims to identify risk factors with predictive value for differentiating between pseudoexfoliation patients at risk of developing intra- or postoperative complications and those without operative risk during cataract surgery. (2) Methods: The review protocol was registered at PROSPERO, registration no. CRD42023417721. The following databases were searched for studies between 2000 and 2023: PubMed/Medline, Scopus, Springer, Science Direct, Web of Science, Cochrane Database of Systematic Reviews, TRIP database, LILACS, Clinical Trials, and reference lists of articles. We included analytical studies of any design examining cataract surgery complications in pseudoexfoliation patients across two population groups, one who underwent uneventful cataract surgery and the other who experienced intra- or postoperative complications. The paper will follow PRISMA 2020 criteria for reporting. Effect measure was assessed using odds ratios (ORs) and corresponding 95% confidence interval (CI) for qualitative variables and means with their respective standard deviation (SD) for quantitative variables. The risk of bias was assessed using the method presented in the Cochrane Handbook for Systematic Reviews. The GRADE scale was used for quality of evidence and certainty. (3) Results: The initial search of published and gray literature databases retrieved 1435 articles, six of which were included in this report. A total of 156 intra- or postoperative incidents were reported in 999 eyes with pseudoexfoliation. The identified predictive factors were a shallow anterior chamber, cataract grade, neutrophil-to-lymphocyte ratio, preoperative intraocular pressure, and symmetry of the exfoliation material. Limitations include heterogeneity of data and limited number of studies identified in our search. (4) Conclusions: These findings suggest the potential to refine risk stratification protocols in clinical settings and assist surgeons in personalized decision-making among individuals with pseudoexfoliation syndrome.

Race and Ethnic Group Enrollment and Outcomes for Wilms Tumor: Analysis of the Current Era Children's Oncology Group Study, AREN03B2.

To review race and ethnic group enrollment and outcomes for Wilms tumor (WT) across all 4 risk-assigned therapeutic trials from the current era Children's Oncology Group Renal Tumor Biology and Risk Stratification Protocol, AREN03B2. For patients with WT enrolled in AREN03B2 (2006 to 2019), disease and biologic features, therapeutic study-specific enrollment, and event-free (EFS) and overall (OS) 4-year survival were compared between institutionally reported race and ethnic groups. Among 5,146 patients with WT, no statistically significant differences were detected between race and ethnic groups regarding subsequent risk-assigned therapeutic study enrollment, disease stage, histology, biologic factors, or overall EFS or OS, except the following variables: Black children were older and had larger tumors at enrollment, whereas Hispanic children had lower rates of diffuse anaplasia WT and loss of heterozygosity at 1p. The only significant difference in EFS or OS between race and ethnic groups was observed among the few children treated for diffuse anaplasia WT with regimen UH-1 and -2 on high-risk protocol, AREN0321. On this therapeutic arm only, Black children showed worse EFS (hazard ratio = 3.18) and OS (hazard ratio = 3.42). However, this finding was not replicated for patients treated with regimen UH-1 and -2 under AREN03B2 but not on AREN0321. Race and ethnic group enrollment appeared constant across AREN03B2 risk-assigned therapeutic trials. EFS and OS on these therapeutic trials when analyzed together were comparable regarding race and ethnicity. Black children may have experienced worse stage-specific survival when treated with regimen UH-1 and -2 on AREN0321, but this survival gap was not confirmed when analyzing additional high-risk AREN03B2 patients.

The Association of Anticoagulation Intensity with Outcomes in Hospitalized COVID-19 Patients.

Venous thromboembolism (VTE) risk is increased in patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A key question was whether increased intensity of anticoagulation would help prevent VTE and improve patient outcomes, including transfer to the intensive care unit (ICU) and mortality. At the start of the coronavirus disease-19 (COVID-19) pandemic, our institution, Boston Medical Center, instituted a VTE risk stratification protocol based on patients' initial D-dimer levels, medical history, and presence of thrombosis to determine whether they should receive standard-dose prophylaxis, high-dose prophylaxis, or therapeutic anticoagulation. We performed a retrospective observational cohort study examining the association of degree of anticoagulation with outcomes in 915 hospitalized COVID-19 patients hospitalized initially on the general inpatient wards between March 1,, 2020, and June 1, 2020. Patients directly hospitalized in the ICU were excluded. Most, 813 patients (89%), in our cohort were on standard-dose prophylaxis; 32 patients (3.5%) received high-dose prophylaxis; 70 patients (7.7%), were treated with therapeutic anticoagulation. VTE occurred in 45 patients (4.9%), and the overall in-hospital mortality rate was 5.4% (49 deaths). On multivariable analysis of clinical outcomes in relation to type of anticoagulation, in the high-dose prophylaxis group, there was a trend towards increased in-hospital mortality (odds ratio 2.4 (0.8-7.5, 95% CI)) and increased ICU transfer (odds ratio 2.2 (0.9-5.7, 95% CI)). Our results suggest that patients receiving high-dose prophylaxis had more severe disease that was not mitigated by intermediate-dose anticoagulation.

Diagnosis and staging of small intestinal neuroendocrine tumors with CT enterography and PET with Gallium-68: preoperative risk stratification protocol.

Pre-operative diagnosis and staging of small intestine neuroendocrine tumors (SI-NETs) remain sub-optimal, with open palpation during surgery still considered the gold standard. This limits a standardized implementation of minimally invasive surgery (MIS). The aim of this single-center retrospective study was to assess a tailored diagnostic work-up to identify candidates at low risk of undetected disease who may benefit from MIS. Patients diagnosed with SI-NETs between 2013 and 2022 who underwent contrast-enhanced computed tomography enterography (CTE) and Ga68-DOTATOC-positron emission tomography-CT (68Ga DOTATATE PET/CT) preoperatively and subsequently underwent open surgical resection were included. Imaging studies were reassessed by two radiologists. Combined use of CTE and 68Ga DOTATATE PET/CT in determining primary lesion disease burden (number of lesions) and LN disease stage (distal and proximal relative to superior mesenteric vessels) was assessed, using surgical reports and pathology as gold standard. Overall, 56 patients were included. Sensitivity of CTE and 68Ga DOTATATE PET/CT for at least one primary SI-NET was 100% and 94%, respectively. In the presence of concordance between studies, combined use of CTE and 68Ga DOTATATE PET/CT for detection of single primary tumors improved specificity to 89% (n = 25/28) with a positive predictive value of 87.5% (n = 21/24). Distal LN disease was identified in 89.2% of cases (n = 33/37). The association of single lesion and distal LN disease was found pre-operatively in 32% of patients (n = 18). Combined use of CTE and 68Ga DOTATATE PET/CT enables identifying low-risk surgical candidates (single SI-NET lesions with distal LN disease).

European Society of Cardiology Core Curriculum for cardio-oncology.

Cardio-oncology is a rapidly growing field of cardiovascular (CV) medicine that has resulted from the continuously increasing clinical demand for specialized CV evaluation, prevention and management of patients suffering or surviving from malignant diseases. Dealing with CV disease in patients with cancer requires special knowledge beyond that included in the general core curriculum for cardiology. Therefore, the European Society of Cardiology (ESC) has developed a special core curriculum for cardio-oncology, a consensus document that defines the level of experience and knowledge required for cardiologists in this particular field. It is structured into 8 chapters, including (i) principles of cancer biology and therapy; (ii) forms and definitions of cancer therapy-related cardiovascular toxicity (CTR-CVT); (iii) risk stratification, prevention and monitoring protocols for CTR-CVT; (iv) diagnosis and management of CV disease in patients with cancer; (v) long-term survivorship programmes and cardio-oncology rehabilitation; (vi) multidisciplinary team management of special populations; (vii) organization of cardio-oncology services; (viii) research in cardio-oncology. The core curriculum aims at promoting standardization and harmonization of training and evaluation in cardio-oncology, while it further provides the ground for an ESC certification programme designed to recognize the competencies of certified specialists.

Prognostic Outcomes of Cutaneous Squamous Cell Carcinoma in Solid Organ Transplant Recipients: A Retrospective Comparative Cohort Study.

Cutaneous squamous cell carcinoma (cSCC) is the second most common cutaneous neoplasm, and its incidence is on the rise. While most cSCCs have an excellent prognosis, certain risk factors, especially immunosuppression, have been associated with higher rates of local recurrence (LR), metastasis, and poor prognosis. This study aims to assess the risk factors for LR and metastasis development in cSCC among solid organ transplant recipients (SOTRs) and compare these rates with those in immunocompetent patients. A retrospective observational study included cSCC cases from the University Hospital Reina Sofía in Córdoba, Spain, between 2002 and 2019. Demographic, clinical, and histopathological data were collected. Local recurrence and metastasis rates were analyzed, along with progression-free survival. Univariate analyses were performed to identify prognostic factors in SOTRs. Among 849 cSCC cases, we found higher rates of local recurrence and metastasis in tumors developed by SOTRs compared to those in immunocompetent individuals. However, no significant differences in local recurrence, metastasis, or progression-free survival were observed between the two groups. Risk factors for adverse outcomes in SOTRs included tumor size > 2 cm, depth > 4 mm, and a higher Clark level. A total of 34.4% of SOTRs developed a second primary cSCC during the follow-up. In our study, cSCCs in SOTRs did not exhibit statistically significant differences in the rates of adverse outcomes compared to immunocompetent patients. The prognosis of cSCCs in SOTRs may be more related to other tumor-dependent risk factors than to the immunosuppression status itself. Future studies are needed to refine risk stratification and follow-up protocols to ensure the optimal management of high-risk cSCC cases, particularly among immunosuppressed patients.

Abstract 18568: Machine Learning Model for the Prediction of Early Discharge of Patient Underwent Transcatheter Aortic Valve Replacement Using Electronic Medical Record

Introduction: With the increasing experience in TAVR, early discharge (within 24 to 72 h) has become a goal to accelerate patient recovery. Several studies suggest the possibility of safe, early discharges. But there is little consensus and considerable variation in early discharge protocols. Optimizing patient selection for early discharge can avoid prolonged hospital stays and save substantial costs without sacrificing patient safety. Hypothesis: The decision for early discharge after TAVR can be modeled and predicted by a machine learning model based on EMR data. Methods: Detailed demography, laboratory, medications, and comorbidities data from the EMR of all 1075 TAVR patients at Brigham and Women's Hospital from 2017-22 were included in a Random Forest machine learning model to make a binary prediction of whether the patient will be discharged within 36 hours after the procedure. We performed a 10-fold cross-validation experiment using the proposed model on the 1075-patient data. Receiver-operating characteristics (ROC) evaluated the prediction accuracy and uncertainty of the model, and clinical outcomes of patients with early discharge were evaluated by 6-month readmissions. Results: The mean age was 81± 9 years; 57% were male. The median length of stay was 51.3 hours. A total of 481 (44.7%) were discharged within 36 hours. Only 12 patients in this cohort were admitted with 24 hours. The average area under the ROC curve (AUC) of the 10-fold cross-validation was 0.83 (SD 0.04). (Figure) A total of 76 (7.1%) patients were readmitted due to cardiovascular reasons within 6 months, including 16/481 (3.3%) early-discharge patients vs. 60/594 (10.1%) non-early discharge patients (p&lt;0.001). Conclusions: Our model accurately predicts early (36 hours) discharge after TAVR, indicating the feasibility of utilizing machine learning to develop an EMR-based patient risk stratification and early discharge protocol.

High Success Rates Utilizing Local Standard of Care Cytogenetic Testing for Risk Stratified Frontline Therapy for Newly Diagnosed Multiple Myeloma: Experience from the Ukmra RADAR Study

Background. Recent studies in genetically high risk (HR) multiple myeloma (MM) such as the UKMRA OPTIMUM study have reported high response rates with intensive induction and post-ASCT consolidation and maintenance strategies, providing the rationale for stratified therapy based on genetic risk. Whilst centralised testing in clinical trials offers ready uniformity and quality control, the use of local laboratories for testing within a clinical trial paves the way for genetic testing to be available for all patients, facilitating risk stratified therapy in routine care. We designed the UKMRA RADAR study to use local laboratories for cytogenetic (CGN) testing to assign treatment according to genetic risk. Methods. RADAR is an ongoing national, multi-center, risk-adapted, response-guided multi-arm, multi-stage (MAMS) phase II/III trial for patients with newly diagnosed MM eligible for ASCT. Patients receive induction with RCyBorD (lenalidomide, cyclophosphamide, bortezomib, dexamethasone), followed by high-dose melphalan and stem cell rescue. Post ASCT, HR patients are randomised to receive isatuximab or not, alongside RBorD consolidation and R maintenance whilst standard risk (SR) patients receive MRD response guided regimens. A protocol amendment (2022) included isatuximab throughout the HR pathway, removing the post-ASCT randomisation. Risk assessment, failure rates and turnaround time are overseen by a genomics working group (GWG) including clinical and specialist cytogenetic representation. Genetic risk testing is undertaken in local genetic labs on CD138 selected bone marrow (BM) plasma cells. HR is assigned based on the presence of at least two of: t(4;14), t(14;16), t(14:20), del(17p), del(1p) and gain(1q). Clone size cut-offs are 10% for IgH translocations and 20% for copy number changes. CGN results from local laboratories are centrally reviewed to assign risk in real-time. Indeterminate cases are discussed by the GWG. Repeat testing is mandated if the initial sample fails to yield a definitive risk result for risk. Subjects with no definitive result despite 2 samples are treated as SR. Results. As of 19 th June 2023, the trial has recruited 472 patients from 72 sites across the UK. Median age is 61 years, 59.3% male, 79.9% white, 11.7% other ethnicity and 8.5% missing ethnicity data. 25 local laboratories were used for CGN testing. Of those with CGN results 354/453 patients (78.1%) were assigned standard risk, 84 (18.5%) high risk and 15 (3.3%) risk category undefined. Risk stratification is ongoing for 19 patients. Of the high risk lesions gain(1q) was most prevalent (35%), followed by t(4;14)(10.6%), del(17p)(8.7%), del(1p32)(8.7%), t(14;16)(3.8%) and t(14;20)(0.8%). In 376/453 patients (83.0%), risk was assigned from a first BM sample. 77 (17.0%) patients required a second sample; the success rate of repeat samples is 71.4%, 22 (4.9%) patients were unable to be risk-stratified despite 2 samples. Figure 1 shows the success rate of first and repeat tests. The overall median turnaround time (from D1 cycle 1 until site notified of result) was 11 days (IQR 0 to 21). Sites were sent risk status prior to cycle 1 in 99 (21.9%) patients , 285 (62.9%) prior to cycle 2 and 341 (75.3%) cycle 3. Note this is calculated where treatment data is available. Conclusion. The RADAR study demonstrates that risk-stratified treatment approaches are feasible in the context of a national phase II/III clinical trial with the use of local laboratories for risk assignment. The vast majority of patients have been successfully allocated to a risk-adapted treatment pathway following a first FISH test (83.0%). Protocol v4 requires time-critical risk-stratification for addition of isatuximab to induction treatment from cycle 2 in high-risk patients, and this has proven to be achievable in almost 2/3 of patients with over ¾ by the end of cycle 2. The RADAR study also enables hospitals across the UK to access CGN testing via a network model of laboratories with standardization of testing and clinically relevant turnaround times. This model is applicable internationally in both public and privately funded healthcare systems and supports the infrastructure for accessible genetic risk stratification for all newly diagnosed MM patients. Ongoing studies within the RADAR trial are evaluating novel and clinically deliverable risk stratification protocols to inform risk stratified approaches in newly diagnosed MM.

Germline genetic variants and pediatric rhabdomyosarcoma outcomes: a report from the Children's Oncology Group.

Relative to other pediatric cancers, survival for rhabdomyosarcoma (RMS) has not improved in recent decades, suggesting the need to enhance risk stratification. Therefore, we conducted a genome-wide association study for event-free survival (EFS) and overall survival (OS) to identify genetic variants associated with outcomes in individuals with RMS. The study included 920 individuals with newly diagnosed RMS who were enrolled in Children's Oncology Group protocols. To assess the association of each single nucleotide polymorphism (SNP) with EFS and OS, we estimated hazard ratios (HRs) and 95% confidence intervals (CIs) using multivariable Cox proportional hazards models, adjusted for clinical covariates. All statistical tests were two sided. We also performed stratified analyses by histological subtype (alveolar and embryonal RMS) and carried out sensitivity analyses of statistically significant SNPs by PAX3/7-FOXO1 fusion status and genetic ancestry group. We identified that rs17321084 was associated with worse EFS (HR = 2.01, 95% CI = 1.59 to 2.53, P = 5.39 × 10-9) and rs10094840 was associated with worse OS (HR = 1.84, 95% CI = 1.48 to 2.27, P = 2.13 × 10-8). Using publicly available data, we found that rs17321084 lies in a binding region for transcription factors GATA2 and GATA3, and rs10094840 is associated with SPAG1 and RNF19A expression. We also identified that CTNNA3 rs2135732 (HR = 3.75, 95% CI = 2.34 to 5.99, P = 3.54 × 10-8) and MED31 rs74504320 (HR = 3.21, 95% CI = 2.12 to 4.86, P = 3.60 × 10-8) were associated with worse OS among individuals with alveolar RMS. We demonstrated that common germline variants are associated with EFS and OS among individuals with RMS. Additional replication and investigation of these SNP effects may further support their consideration in risk stratification protocols.

Yurf-Bangash-Yousafzai Augur: Proposal of Carotid Plaque Formation Predictor Among General Population by Exploration of State-Of-The-Art for Automated Machine Learning

BackgroundCarotid plaque, recognized as a predictor of adverse cardiovascular and cerebrovascular complications, has been explored via machine learning. With Artificial intelligence in general and machine learning in particular exhibiting superior predictive and prognosticative capabilities, when compared with conventional statistical techniques, we explored the current state-of-the-art for automated Machine Learning to develop prediction algorithmic models for Carotid Plaque (CP) formation with the aim to optimise risk stratification and subsequent complication triaging among general population. MethodologyThe study population comprised 122 entrants— variables including subjective measures such as weight, height and smoking history as well as serum levels of HbA1c and HDL-C. The current state of the art (SOTA) for automated Machine Learning (aML) was adopted to develop predictive models using algorithms including Neural Network, eXtreme Gradient Boosting and CatBoost with the employment of hyperparameter tuning. Ensemble approach, which is the amalgamation of two or more than two algorithmic models to develop such a model which is better than either of its computive components, was superimposed. The training was achieved on 40% of the loaded dataset with the theretofore unexposed 60% of the said dataset used for testing that respective developed algorithmic model which came out to exhibit the highest macro weighted average Area Under the Receiver Operating Curve (mWA-AUROC) among the trained models. mWA-AUROC and accuracy, along with other parameters, were adopted to assess the predictive discriminative ability of the tested algorithmic model. ResultsA Decision Tree algorithmic model predicted CP formation with a perfect mWA-AUROC of 1.00 and an accuracy of 100% on the training partition (40% of the original dataset) along with an mWA-AUROC of 0.99 and an accuracy of 98.6% on the testing partition (60% of the original dataset). (Figure 1) A precision of 100% and a recall of 98.5%— such that an F1 score of 99.2%— was achieved on the testing partition by the respective algorithmic model— to be heretofore termed as the “Yurf-Bangash-Yousafzai augur”. The Yurf-Bangash-Yousafzai augur thus outsmarts the CP predictive model (having AUROC, accuracy and an F1 score each of 0.86) developed by Wu D et al. ConclusionsOur novel approach to developing a predictive risk model for CP formation by exploring SOTA for automated machine learning provides optimised predictions which, when incorporated into the respective risk stratification protocols, shall translate into a decrease in the morbidity and mortality associated with cardiovascular atherosclerotic events by assisting in risk stratification and subsequent complication triaging among the general population.These authors contributed equally to data curation, development of methodology, statistical analyses and writing with subsequent reviewing and editing of the abstract draft.These authors, indicated alphabetically, contributed equally to statistical analyses and validation.

Incorporating Ensembled Stacked Approach into Automated Machine Learning in an Attempt to Predict Acute Ischemic Heart Disease in Patients with Atypical Chest Pain: Secondary Analysis of a Single-Center Retrospective Cohort Study

Background, Rationale & ObjectiveConsiderable research has been carried out to explore the adoption of novel methodologies in a bid to optimize the management protocols for acute ischemic heart disease (AIHD) that is prevalent worldwide and has significant morbidity and mortality associated with it. With Artificial intelligence in general and machine learning in particular exhibiting superior predictive and prognositicative capabilities, when compared with conventional statistical techniques, we explored automated Machine Learning to predict AIHD in such patients presenting with atypical chest pain. MethodologyThe study population comprised 3,833 patients presenting with chest pain who visited the ED of a tertiary-care hospital from January 2014 to December 2018. Such patients who exhibited typical chest pain were excluded. (Mean age = 60.3; Female subjects = 45.6%; Diagnosed with AIHD = 11.8%). Variables included a plethora of variables obtained on a routine ED evaluation including vital signs and laboratory data. The current state of the art for automated Machine Learning was adopted to develop classification models using algorithms including Extreme Gradiant Boosting, Random Forest and Neural Network. The models were stacked after implementation of hyperparameter tuning. Ensemble approach, which is the amalgamation of two or more than two algorithmic models to develop such a model which is better than either of its computive components, was superimposed on stacked models. The macro-weighted average Area under the Response-operating Curve (mwA-AUROC) gauged the discriminating ability of models with the highest value of 1 indicating perfect discriminative classification ability. ResultsAn ensemble of stacked Extra Trees, Neural Network, CatBoost, Light Gradient Boosted Machine, Random Forest algorithmic models achieved an mwA-AUROC of 0.78 and an accuracy of 87.9%. (Figure 1) Our model outperformed those developed by Kim KH et al. The ensemble model exhibited the least log loss among all of the developed algorithmic models. ConclusionSuch predictive models, when deployed on cloud, have the capability of significantly influencing the prognosticating protocols for AIHD by providing instantaneous predictions. Such an optimization of the risk stratification protocols achieved via incorporation of aML would serve to improve the management of not only the respective patient but also that of the resource allocation. This would ultimately translate into significant reduction of morbidity and mortality associated with AIHD.These authors contributed equally to data curation, development of methodology, statistical analyses and writing with subsequent reviewing and editing of the abstract draft.These authors, indicated alphabetically, contributed equally to statistical analyses and validation.

Ultrasound classification of non-mass breast lesions following BI-RADS presents high positive predictive value.

To investigate the positive predictive value of ultrasound classification of non-mass breast lesions (NMLs) following breast imaging reporting and data system (BI-RADS), and enhance understanding of NMLs. Fifty-nine women with 59 ultrasound-detected breast NMLs were finally enrolled. The ultrasound (US) features of breast NMLs were analyzed; the incidence of malignant NMLs was calculated; the malignancy risk stratification of US for breast NMLs was established using BI-RADS. The incidence of malignant NMLs was 4.59% of all breast carcinoma. Non-ductal hypoechoic area, microcalcifications and posterior shadowing are the main US features of malignant NMLs, and there were significant differences between malignant and benign NMLs for microcalcifications and posterior shadowing. Taking BI-RADS 4B as a cutoff value, the sensitivity, specificity, area under the receiver operating characteristic curve (AUC), positive and negative predictive values, and odds ratio of the BI-RADS category were 82.98%,41.67%,0.62,84.78%,38.46% and 3.48, respectively. Stratifying the malignancy risk of breast NMLs using the BI-RADS the sensitivity and positive and predictive value are promising, but the likelihood of malignancy of malignant NMLs is underestimated, and that of benign NMLs is overestimated. The solution may be that to separate NMLs from breast masses and use different malignancy risk stratification protocols.

Comparison of Uveitis Incidence by Medication in Juvenile Idiopathic Arthritis and Implications for Screening

To determine uveitis incidence in juvenile idiopathic arthritis (JIA) patients treated with disease-modifying antirheumatic drugs (DMARD) medications, and to evaluate uveitis risk-stratification protocols. Retrospective clinical cohort study. Medical records of patients with JIA seen by ophthalmology at a single institution from April 2014 to April 2022 and ≥18 months' follow-up were reviewed. Exclusion criteria included uveitis history prior to study period, Still disease, or <18 months' follow-up. Patient characteristics, medications, and uveitis status were recorded. Factors associated with uveitis development were analyzed and statistically significant metrics used to determine empiric risk-stratification criteria. These criteria and American College of Rheumatology (ACR) risk-stratification guidelines were applied retroactively to determine predictive power. One hundred eighty-four patients met inclusion criteria and were included. Twenty-one new cases of uveitis developed during the study period. There were no statistically significant differences between no DMARD treatment, methotrexate (MTX), and etanercept (ETA) groups in uveitis incidence, whereas the adalimumab (ADA) and other biologics groups had no uveitis cases. Under the empirically determined criteria, the ratio of uveitis incidence between high- and low-risk groups was 8.21 (2.68-33.55; P < .0001), whereas it was 1.90 (0.72-4.93; P=.15) under the ACR criteria. Patients on MTX, ETA, and no DMARD treatment were comparable in JIA-associated uveitis incidence, whereas there were no new cases with ADA or other biologics. Further, we found increased predictive power in the empiric criteria in comparison to current ACR risk stratification.

Mitral Valve Prolapse and Mitral Annular Disjunction Arrhythmic Syndromes: Diagnosis, Risk Stratification and Management.

Although mitral valve prolapse (MVP) is usually considered a benign clinical condition, it has been linked with ventricular arrhythmias and sudden cardiac death in patients with a certain "arrhythmic" phenotype, raising awareness and mandating a specific risk stratification protocol. Mitral annular disjunction (MAD) is considered a "red flag" in malignant MVP syndrome along with bileaflet myxomatous prolapse, female gender, negative or biphasic T waves in the inferior leads, fibrosis in the papillary muscles or inferobasal wall detected by cardiac magnetic resonance imaging and complex arrhythmias of right bundle branch morphology. MAD seems to play a critical role in the chain of morphofunctional abnormalities which lead to increased mechanical stretch and subsequent fibrosis mainly in the papillary muscles, forming the vulnerable anatomic substrate prone to arrhythmogenesis, and associated with long-term severe ventricular arrhythmias. Arrhythmogenesis in MVP/MAD patients is not fully understood but a combination between a substrate and a trigger has been established with premature ventricular contraction triggered ventricular fibrillation being the main mechanism of sudden cardiac death (SCD). Certain characteristics mostly recognized by non-invasive imaging modalities serve as risk factors and can be used to diagnose and identify high risk patients with MAD, while treatment options include catheter ablation, device therapy and surgical intervention. This review focuses on the clinical presentation, the arrhythmogenic substrate, and the incidence of ventricular arrhythmias and SCD in MAD population. The current risk stratification tools in MAD arrhythmogenic entity are discussed.

Cost-minimisation analysis of polysomnography and pulse oximetry in a risk stratification protocol for paediatric adenotonsillectomy.

Severe obstructive sleep apnoea (OSA) is associated with increased risk of respiratory compromise in the post-operative period following adenotonsillectomy (AT). This study analyses the economic cost of polysomnography or overnight oximetry as part of pre-operative risk stratification in paediatric AT, supplementing previously published research demonstrating the efficacy of this protocol in predicting respiratory complications. This cost-minimisation analysis examines costs associated with pre-operative overnight oximetry and polysomnography in triaging paediatric patients older than 2 years old, with no major comorbidities except for OSA, undergoing AT for OSA (n=1801) to either a secondary or quaternary Australian hospital. Decision analysis modelling via probability trees were utilized to estimate pre- and peri-operative costs. A third hypothetical 'no investigation' model based upon conducting all AT at a secondary hospital was performed. Costs are derived from the financial year 2020-2021, censored at discharge. The total cost per patient of AT including pre-operative investigations of oximetry and polysomnography, and associated inpatient costs, were AUD4181.34 and 5013.99 respectively. This is more expensive compared to a hypothetical no-investigation model (AUD3958.98). Within the scope of this partial economic evaluation, this study finds a small additional cost for a model of care involving overnight oximetry as a pre-operative triage tool, balanced by the reduced cost of care in a lower acuity centre for low-risk patients and potential high cost of complications if all children are treated in a low acuity centre. This supports oximetry in peri-operative risk stratification for paediatric AT from a financial perspective.

Abstract 683: Identification of common germline variants associated with pediatric rhabdomyosarcoma survival: A report from the Children's Oncology Group (COG)

Abstract Background: Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma in children and has one of the poorest survival rates among pediatric cancers, underscoring the need to identify factors which may be leveraged to improve therapeutic options for these individuals. Methods: We carried out a genome-wide association study of overall survival (OS) and event-free survival (EFS) in 920 RMS patients from COG protocols and randomly divided them into discovery (n=642) and replication (n=278) cohorts. Genotyping was conducted using the Illumina OmniExpress or Global Screening Array and imputed using the Haplotype Reference Consortium. We used Cox proportional hazards regression to calculate an adjusted hazard ratio (aHR) and P value for each common variant (minor allele frequency [MAF]&amp;gt;5%) for OS and EFS while adjusting for age at diagnosis, tumor stage, histological subtype, and the top five principal components. Analyses were also conducted by histological subtype: embryonal RMS (ERMS, n=544) and alveolar RMS (ARMS, n=268). Finally, we performed a meta-analysis of the results from the discovery and replication cohorts to generate a summary aHR and P value for each single nucleotide polymorphism (SNP). Results: We identified an intergenic SNP at chr8q21.13 associated with worse RMS EFS across subtypes (aHR=2.08, P=2.80x10-9), which had consistent effects across the discovery (aHR=1.91, P=5.05x10-6) and replication (aHR=2.62, P=7.16x10-5) cohorts. This SNP lies in a region which spans the genomic binding site for GATA2 and GATA3, transcription factors that are recognized to contribute to cancer development. We also identified a significant association between a SNP at chr12q21.1 and worse EFS (aHR=2.04, P=3.35x10-8) with consistent effects across the discovery and replication cohorts. Based on data from the Genotype-Tissue Expression project (GTEx), this SNP is associated with expression of SLCO1B1, a gene which encodes a liver anion transporter linked to RMS treatment-related toxicities. In subtype-specific analyses, we identified a SNP at chr17q21.32 that was significantly associated with worse ARMS OS (129 events; aHR=3.18, P=3.12x10-8; discovery: aHR=3.19, P=6.23x10-4; replication: aHR=3.16, P=1.43x10-3). In GTEx, this SNP is associated with expression and splicing of genes including PITPNM3, KIAA0753, and MED31 across various tissues. No SNPs were significantly associated with ERMS OS or EFS. Conclusion: In the first GWAS of RMS survival outcomes, we identified two SNPs that were significantly associated with worse EFS across RMS subtypes. Further, we identified a SNP that was associated with OS in ARMS patients, a subtype that is associated with worse outcomes. Additional investigation of the impact of these SNPs may further support their consideration for novel risk stratification protocols. Citation Format: Bailey A. Martin-Giacalone, Michael E. Scheurer, Javed Khan, Stephen J. Chanock, Shengchao Alfred Li, Meredith Yeager, Deborah A. Marquez-Do, Donald A. Barkauskas, David Hall, Matthew T. McEvoy, Melissa A. Richard, Pagna Sok, Austin L. Brown, Aniko Sabo, Stephen X. Skapek, Douglas S. Hawkins, Rajkumar Venkatramani, Lisa Mirabello, Philip J. Lupo. Identification of common germline variants associated with pediatric rhabdomyosarcoma survival: A report from the Children's Oncology Group (COG) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 683.

Abstract 2506: De novo variants associated with neuroblastoma and congenital heart defects: evidence of pleiotropic effect from 1311 WGS trios

Abstract Introduction: Neuroblastoma (NBL) is the most common type of extracranial pediatric solid tumor. NBL exhibits a wide spectrum of phenotypes, from low-risk cases that spontaneously regress to high-risk cases with 50% mortality, and it is responsible for 12% to 15% of deaths due to childhood cancer. While genome-wide association studies have identified a handful of risk loci, the germline genetic origins of sporadic NBL remain largely unknown. Notably, NBL is one of several childhood cancers that tend to co-occur with specific birth defects, which could point to novel genetic origins. For example, the risk of neuroblastoma is increased 3- to 7-fold among children with certain congenital heart defects (CHD). It is hypothesized that this co-occurrence arises from pleiotropic defects in the neural crest cells that are progenitors of both NBL and some heart structures. Based on these lines of evidence, we hypothesize that novel germline risk loci for NBL can be identified through pleiotropic analyses of CHD and NBL cohorts. Therefore, in this study we seek to identify germline risk loci for neuroblastoma by jointly analyzing CHD and NBL data sets with a primary focus on de novo variants. Methods: Using whole genome sequencing data from the Gabriella Miller Kids First Data Resource Center, we will identify NBL and CHD associated loci by testing for high rates of de novo variants (DNV) in each cohort separately and in the two cohorts combined. The testing strategy compares the rate of DNV to the expected locus-specific mutation rate. In our primary analyses, we will focus on genes known to be associated with CHD in order to reduce the multiple testing penalty for that family of loci. A broader analysis will explore coding and regulatory loci genome wide to identify loci associated jointly with NBL and CHD. Results: We acquired whole genome sequencing data of parent-offspring trios for NBL (n=609) and CHD (n=711) from the Gabriella Miller Kids First Data Resource Center. The probands in the NBL cohort was 54% male and 81% Non-Hispanic White, with a median age at diagnosis of 16 months. The probands in the CHD cohort was 59% male and 60% Non-Hispanic White. The most common CHD phenotypes were ventricular septal defect (46%), tetralogy of Fallot (32%), hypoplastic left heart syndrome (28%), atrial septal defect (25%), and right aortic arch (25%). Analyses to identify DNV are ongoing. Conclusion: The identification of germline risk loci associated with NBL will help to inform the developmental origins of this important childhood cancer and may inform novel risk stratification protocols and therapies. This study also aims to help explain the NBL-CHD co-occurrence, and that could shed light on the development of other neural crest cell derived cancers. Citation Format: Alexander Renwick, Yao Yu, Chad Huff, Philip J. Lupo. De novo variants associated with neuroblastoma and congenital heart defects: evidence of pleiotropic effect from 1311 WGS trios [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2506.