Low-risk Stratum Research Articles

Application of the REVEAL risk score calculator 2.0 in the CHEST study

BackgroundCurrently there are no risk assessment recommendations for chronic thromboembolic pulmonary hypertension (CTEPH). The Registry to Evaluate Early and Long-Term PAH Disease Management (REVEAL) risk score (RRS), developed for risk assessment in patients with pulmonary arterial hypertension, has previously predicted outcomes in CTEPH. RRS 2.0 was developed to refine the RRS. MethodsThis post hoc analysis of the CHEST study (n = 237), which assessed riociguat in patients with inoperable and persistent/recurrent CTEPH, evaluated RRS 2.0 and its relationship with survival and clinical worsening-free survival (CWFS). ResultsAt CHEST-1 Week 16, RRS 2.0 significantly improved and more patients moved into the low-risk stratum with riociguat versus placebo; these improvements were maintained at CHEST-2 Week 12. RRS 2.0 at CHEST-1 baseline and Week 16, and change in RRS 2.0 from CHEST-1 baseline to Week 16 were significant predictors of survival and CWFS in CHEST-2. ConclusionsOur data suggest that RRS 2.0 may have utility in predicting outcomes and monitoring treatment response in patients with inoperable or persistent/recurrent CTEPH.

Validation of Different Prognostic Scores in Allogeneic Hematopoietic Cell Transplantation in the Post-Transplant Cyclophosphamide Era

INTRODUCTIONThe use of post-transplantation cyclophosphamide (PTCy) for graft-versus-host disease (GvHD) prophylaxis has decreased the rates of this complication, resulting on an improvement of transplant-related toxicity and survival. Secondary to its efficacy, the use of PTCy has been almost universally integrated for allogeneic hematopoietic cell transplantation (alloHCT), independently of the selected donor source.Clinical decisions in alloHCT are supported by the use of prognostic scores for outcome prediction. However, capability of prediction by diverse scores can vary depending on their features and on the composition of the study cohort. Additionally, the continuous innovation on alloHCT techniques and practices leads to an ongoing need to update risk indices aimed at improving risk stratification of patients undergoing alloHCT. This study explores the predictive capacity of different prognostic scores routinely used in alloHCT, in a contemporaneous cohort of adults undergoing peripheral blood (PB) alloHCT using PTCy-based GvHD prophylaxis.METHODSBetween 2014 and 2020, 230 consecutive adults with hematological malignancies underwent PB-alloHCT with PTCy-based GvHD prophylaxis at our Institution. Data related to Hematopoietic Cell Transplantation-Comorbidity Index (HCT-CI), Karnosfky Performance Status (KPS), Disease Risk Index (DRI), European Bone Marrow Transplantation (EBMT) score, and Endothelial Activation and Stress Index (EASIX) were collected retrospectively. Complete information was available for 216 patients. Overall survival (OS) was considered the main outcome variable. Patients were grouped into two risk groups based on the optimal cut-off value for each score. In the case of EASIX, 1.578 was the most discriminating cut-off for OS. The score discrimination for OS was measured independently for each index using the receiver operating characteristic curve (AUC) calculated using receiver operating characteristic (ROC) curves, and determined at different time-points after alloHCT.RESULTSOf the 216 patients included, the median age was 52 years (range: 18-70), acute myeloid leukemia (36.1%) was the most prevalent baseline diagnosis, 42.1% of adults underwent reduced-intensity conditioning alloHCT, 69.4% received grafts from unrelated donors, and 23.0% from haploidentical donors. With a median follow-up of 22.6 months, 24.1% patients relapsed, and 2-y OS and non-relapse mortality were 67.3% and 19.9%.DRI, HCT-CI, KPS, and EASIX successfully grouped patients into higher and lower risk strata, supporting their use for risk classification. HCT-CI [(score>3 (vs 0-3): HR 2.02, p<0.01], DRI [High - Very High risk (vs Low - Int): HR 2.08, p<0.01], and EASIX [>1.578 (vs ≤ 1.578): HR 1.73, p<0.02], maintained an optimal discrimination capacity during the entire post-transplant follow-up (median AUC ranges > 55%).DRI was the most accurate prognostic index during the entire post-transplant period (median AUC ranges > 60%). KPS score was found to be a useful predictor of mortality up to the first year after alloHCT and with the highest prognostic accuracy at 3 months (AUC 67.09%). HCT-CI score was found to present a better discrimination capacity once elapsed 6 months after alloHCT and with a peak of prediction capacity at 2 years (AUC 60.3%). EASIX, when measured at the pre-transplant evaluation, demonstrated to have acceptable predictive ability during the entire post-transplant period (median AUC > 55%), and with a peak of prediction at 3 months (AUC 62.6%). The EBMT score had the lowest predictive capacity in our analysis (Figure 1). CONCLUSIONThis study validates, for the first time, the risk stratification capacity for OS of DRI, HCT-CI, KPS, and EASIX in PB-alloHCT with PCTy-based prophylaxis. Interestingly, the prediction accuracy of the prognostic scores differed depending on the time-period. This result can be taken into consideration to enhance the applicability of these scores and refine the clinical decisions taken based on the information provided from their use in routine clinical practice. [Display omitted] DisclosuresLozano: Terumo BCT: Honoraria, Research Funding; Macopharma: Research Funding; Grifols: Honoraria. Rosinol: Janssen, Celgene, Amgen and Takeda: Honoraria. Esteve: Novartis: Consultancy, Research Funding; Astellas: Consultancy; Jazz: Consultancy; Pfizer: Consultancy; Novartis: Research Funding; Abbvie: Consultancy; Bristol Myers Squibb/Celgene: Consultancy.

Risk Stratification in Cost-Effectiveness Analyses of Cancer Screening: Intervention Eligibility, Strategy Choice, and Optimality.

IntroductionThere is increasing interest in risk-stratified approaches to cancer screening in cost-effectiveness analysis (CEA). Current CEA practice regarding risk stratification is heterogeneous and guidance on the best approach is lacking. This article suggests how stratification in CEA can be improved.MethodsI use a simple example of a hypothetical screening intervention with 3 potential recipient risk strata. The screening intervention has 6 alternative intensities, each with different costs and effects, all of which vary between strata. I consider a series of alternative stratification approaches, demonstrating the consequences for estimated costs, effects, and the choice of optimal strategy. I supplement this analysis with applied examples from the literature.ResultsAdopting the same screening policy for all strata yields the least efficient strategies, where efficiency is understood as the volume of net health benefit generated across a range of cost-effectiveness threshold values. Basic stratification that withholds screening from lower-risk strata while adopting a common strategy for those screened increases efficiency. Greatest efficiency is achieved when different strata receive separate strategies. While complete optimization can be achieved within a single analysis by considering all possible policy combinations, the resulting number of strategy combinations may be inconveniently large. Optimization with separate strata-specific analyses is simpler and more transparent. Despite this, there can be good reasons to simulate all strata together in a single analysis.ConclusionsIf the benefits of risk stratification are to be fully realized, policy makers need to consider the extent to which stratification is feasible, and modelers need to simulate those choices adequately. It is hoped this analysis will clarify those policy and modeling choices and therefore lead to improved population health outcomes.

Integrated prognostication of intrahepatic cholangiocarcinoma by contrast-enhanced computed tomography: the adjunct yield of radiomics

PurposeTo test radiomics for prognostication of intrahepatic mass-forming cholangiocarcinoma (IMCC) and to develop a comprehensive risk model.MethodsHistologically proven IMCC (representing the full range of stages) were retrospectively analyzed by volume segmentation on baseline hepatic venous phase computed tomography (CT), by two readers with different experience (R1 and R2). Morphological CT features included: tumor size, hepatic satellite lesions, lymph node and distant metastases. Radiomic features (RF) were compared across CT protocols and readers. Univariate analysis against overall survival (OS) warranted ranking and selection of RF into radiomic signature (RSign), which was dichotomized into high and low-risk strata (RSign*). Models without and with RSign* (Model 1 and 2, respectively) were compared.ResultsAmong 78 patients (median follow-up 262 days, IQR 73–957), 62/78 (79%) died during the study period, 46/78 (59%) died within 1 year. Up to 10% RF showed variability across CT protocols; 37/108 (34%) RF showed variability due to manual segmentation. RSign stratified OS (univariate: HR 1.37 for R1, HR 1.28 for R2), RSign* was different between readers (R1 0.39; R2 0.57). Model 1 showed AUC 0.71, which increased in Model 2: AUC 0.81 (p < 0.001) and AIC 89 for R1, AUC 0.81 (p = 0.001) and AIC 90.2 for R2.ConclusionThe use of RF into a unified RSign score stratified OS in patients with IMCC. Dichotomized RSign* classified survival strata, its inclusion in risk models showed adjunct yield. The cut-off value of RSign* was different between readers, suggesting that the use of reference values is hampered by interobserver variability.

POS-223 Clinical Utility of KidneyIntelX in Patients with Early Stages of Diabetic Kidney Disease in CANVAS Participants

Prediction of progression during early stages of diabetic kidney disease (DKD) and determination of those most likely to derive benefits from therapies a priori is challenging. KidneyIntelX is a new test that incorporates clinical data and three plasma biomarkers using machine learning into a composite risk score for DKD progression that has yet to be applied to a clinical trial cohort. We performed a post-hoc analysis of CANVAS Study participants with diabetes and prevalent DKD (defined by estimated glomerular filtration rate (eGFR) <60 ml/min/1.73m2 or UACR > 30 mg/g at enrollment). Three circulating biomarkers in plasma, soluble Tumor Necrosis Factor Receptor 1 (sTNFR-1), soluble Tumor Necrosis Factor Receptor 2 (sTNFR-2), and Kidney Injury Molecule (KIM)-1, were measured with proprietary multiplex assays. We calculated chronic eGFR slope from 6 weeks in the canagliflozin arm and from baseline in the placebo arm and compared the proportion of individuals experiencing the composite kidney outcome of estimated glomerular filtration rate (eGFR) decline of ≥5 ml/min/year, ≥40% sustained decline in eGFR, or kidney failure within 5 years in CANVAS participants stratified by KidneyIntelX risk categories based on the existing model vs. the KDIGO risk strata. We also assessed clinical response to canagliflozin vs. placebo within KidneyIntelX and KDIGO risk strata. We included 1325 (31% of original cohort of 4330) CANVAS participants with prevalent DKD and available baseline plasma samples (mean age 64 years, 32% female, mean eGFR 65 mL/min/1.73 m2, median uACR 56 mg/g). During a mean follow-up of 5.6 years, 131 (9.9%) of this subgroup with baseline DKD experienced the composite kidney outcome. KDIGO categorization based on eGFR and uACR stratified 69%, 23% and 8% of the DKD population into “moderately increased risk,” “high risk,” and “very high risk,” with event rates of 7%, 17%, and 18% (relative risk [RR] of 2.3, 95% CI 1.3-4.2 for the “very high risk” vs. “moderately high risk” in the treatment arm and RR of 3.4, 95% CI 1.5-7.5 in the placebo arm). Using risk cutoffs from prior validation studies, KidneyIntelX stratified patients to low (40%), intermediate (45%), and high risk (15%) strata with event rates of 3%, 10%, and 26% (RR of 6.0 , 95% CI 3.2-11.3 for the high vs. low risk groups in the treatment arm and RR 15.2, 95% CI 4.8-37.8 in the placebo arm) (Figure). The treatment effects of canagliflozin vs. placebo varied by baseline strata of KidneyIntelX. In the low risk stratum, the RR was 1.65 (95% CI 0.54-5.08), in the intermediate stratum the RR was 0.59 (95% CI 0.36-0.98), and in the high-risk stratum the RR was 0.67 (95% CI 0.38-1.18). In the high-risk stratum, the absolute risk reduction (ARR) for canagliflozin vs. placebo for the composite kidney outcome was 8% vs. an ARR of 2.6% in the overall DKD population. KidneyIntelX successfully risk-stratified a large multi-national external cohort for risk of progression of DKD, with larger differences in observed events across KidneyIntelX risk strata compared to KDIGO risk strata. Moreover, individuals scored as high risk by KidneyIntelX have the potential to achieve greater absolute risk reductions for the outcome with canagliflozin treatment compared to non-stratified DKD participants.

Analysis of Risk Assessment Tools in the REPLACE Study

Purpose The REPLACE study showed that switching to riociguat (rio) can benefit pulmonary arterial hypertension (PAH) patients treated with phosphodiesterase type 5 inhibitors (PDE5i) but still at intermediate risk. Improvements to lower risk status with rio versus PDE5i were seen with all three risk assessment tools analyzed, although the difference was not statistically significant with the modified REVEAL risk score (mRRS). Here, we report changes in the components of the mRRS, COMPERA, and FPHN non-invasive risk assessment tools, and a post hoc analysis of RRS Lite 2 in REPLACE. Methods Individual components of mRRS, COMPERA, and FPHN non-invasive assessment tools, and RRS Lite 2 risk scores and risk strata were calculated for the rio group (n=111) and PDE5i group (n=113) at REPLACE baseline and Week 24. WHO/NYHA functional class (FC), 6MWD, and NT-proBNP were components of all risk scores. Vital signs (systolic blood pressure [SBP] and heart rate [HR]) and renal insufficiency were additional components for mRRS and RRS Lite 2. PAH etiology and demographics were also components of mRRS. Results For mRRS, COMPERA, and FPHN, 44% of patients in the rio group vs 23% in the PDE5i group improved their WHO/NYHA FC component score at Week 24. Similar results were found for the 6MWD component (1 point removed from mRRS: rio 41% vs PDE5i 27%; COMPERA grade 1/FPHN low risk at Week 24: rio 37% vs PDE5i 27%). The incidences of lowest-risk NT-proBNP component scores at Week 24 were rio 41% vs PDE5i 40% for mRRS, rio 21% vs PDE5i 12% for COMPERA, and rio 41% vs PDE5i 40% for FPHN. At baseline, mRRS vital sign scores were similar between the two groups. At Week 24, more patients in the rio group (50%) versus the PDE5i group (36%) had 1−2 points added to their mRRS score due to SBP 96 bpm. The overall result was similar for the mRRS and the RRS Lite 2. With RRS Lite 2, the proportion of patients achieving a low-risk stratum from baseline to Week 24 was 41% to 61% for rio and 37% to 50% for PDE5i. Conclusion Risk assessment tools showed similar trends in shared components between treatment groups at baseline and Week 24. Results with RRS Lite 2 were similar to mRRS. Differences in discrimination of the risk sores, weighting of components, and differences in the vital signs component of mRRS could explain between score differences at baseline and Week 24 in REPLACE. SBP and HR will be individually analyzed in future.

Comparing Practice and Outcome of Laparoscopic Liver Resection Between High-volume Expert Centres and Nationwide Low-medium Volume Centres

Purpose: It is unknown how use and outcome of laparoscopic liver resection (LLR) compare between high-volume centres (HV) and low-medium volume centres (LMV). This study aimed to compare use and outcome of LLR according to surgical difficulty in three leading European HV and twenty Dutch LMV. Method: An international, retrospective multicentre cohort study including LLR data from three European HV and twenty Dutch LMV from January 2011 to December 2016. HV are defined by performing >50 LLR annually. Patients were retrospectively stratified into low-, moderate- and high-risk Southampton difficulty strata. Results: 2425 patients were included (1540 HV; 885 LMV). The annual volume of LLR was higher in HV than LMV (42.9% versus 7.2%). Patients in HV experienced fewer conversions (7.4 versus 13.1%; P<0.001) and less intraoperative incidents (9.3% versus 14.6%; P=0.002) compared to LMV. Whereas morbidity and mortality rates were similar, intervention rate (5.1% versus 7.2%; P=0.034) and postoperative hospital stay (3 versus 5 days; P<0.001) were lower in HV compared to LMV. Per difficulty score stratum, the conversion rate and postoperative hospital stay were lower in HV. The rate of intraoperative incidents was similar in the low-risk stratum, however lower in HV in the moderate-risk and high-risk strata (absolute difference 6.7% and 14.2%; P<0.004). Conclusion: HV had a sixfold higher use of LLR, less conversions, and shorter hospital stay, as compared to a nationwide LMV. Stratification into difficulty scores identified some differences but largely outcomes appeared better for HV in each risk group.

Drivers of Cost Differences Between Nurse Practitioner and Physician Attributed Medicare Beneficiaries.

Although recent research suggests that primary care provided by nurse practitioners costs less than primary care provided by physicians, little is known about underlying drivers of these cost differences. Identify the drivers of cost differences between Medicare beneficiaries attributed to primary care nurse practitioners (PCNPs) and primary care physicians (PCMDs). Cross-sectional cost decomposition analysis using 2009-2010 Medicare administrative claims for beneficiaries attributed to PCNPs and PCMDs with risk stratification to control for beneficiary severity. Cost differences between PCNPs and PCMDs were decomposed into payment, service volume, and service mix within low-risk, moderate-risk and high-risk strata. Overall, the average PCMD cost of care is 34% higher than PCNP care in the low-risk stratum, and 28% and 21% higher in the medium-risk and high-risk stratum. In the low-risk stratum, the difference is comprised of 24% service volume, 6% payment, and 4% service mix. In the high-risk stratum, the difference is composed of 7% service volume, 9% payment, and 4% service mix. The cost difference between PCNP and PCMD attributed beneficiaries is persistent and significant, but narrows as risk increases. Across the strata, PCNPs use fewer and less expensive services than PCMDs. In the low-risk stratum, PCNPs use markedly fewer services than PCMDs. There are differences in the costs of primary care of Medicare beneficiaries provided by nurse practitioners and MDs. Especially in low-risk populations, the lower cost of PCNP provided care is primarily driven by lower service volume.

Developing a toolbox for identifying when to engage senior surgeons in emergency general surgery: A multicenter cohort study

BackgroundHaving a senior surgeon present for high-risk patients is an important safety measure in emergency surgery, but 24-h consultant cover is not efficient. We aimed to develop a user-friendly toolbox (risk identification, outcome prediction and patient stratification) to support when to involve a senior surgeon. Materials and methodsWe included 11,901 general surgery patients (10.0% emergencies) in a multicenter prospective cohort in China (2015–2016). Patient information and surgeons' seniority were compared between emergency and elective surgery with the same procedure codes. Risk indicators common in these two surgical timings and specific to emergency surgery were identified, and their clinical importance was evaluated by a working group of 48 experienced surgeons. Predictive models for mortality and morbidity were built using logistic regression models. Stratification rules were created to balance patients' risk and surgeons’ caseload with an Acute Call Team (ACT) model. ResultsEmergency patients had significantly higher risks of mortality (3.6% vs 0.6%) and morbidity (7.8% vs 4.3%) than elective patients, but disproportionally fewer senior surgeons (59.9% vs 91.4%) were present. Using three risk indicators (American Society of Anesthesiologists score, age, blood urea nitrogen), C-statistic (95% CI) for prediction of emergency mortality was high [0.90 (0.84–0.96)]. It was less complex but equally accurate as two existing and validated models (0.86 [0.79–0.93] and 0.86 [0.77–0.95]). Using five indicators, C-statistic (95% CI) was moderate for prediction of overall morbidity [0.77 (0.72–0.83)], but high for severe morbidity [0.92 (0.88–0.97)]. Based on stratification rules of the ACT model, patient mortality and morbidity were 0.5% and 5.3% in the low-risk stratum (composing 64.6% of emergency caseload), and 15.9% and 29.0% in the very high-risk stratum (6.9% of caseload). ConclusionThese findings show the practical feasibility of using a risk assessment tool to direct senior surgeons’ involvement in emergency general surgery.

Study of demographic, hematological profile and risk stratification in chronic myeloid leukemia patients, in and around North-West Punjab, India

Background: Chronic myeloid leukemia (CML) is a hematopoietic stem cell disorder with cytogenic profile and tyrosine kinase inhibitors used as its therapy. Objective of the present study was to determine the demographic, haematological profile and to characterize them in low and high-risk group on presentation with European treatment and outcome study score (EUTOS) in CML patients in and around North-West Punjab, India.Methods: Diagnosed cases of CML were taken. Investigations were done. Molecular and cytogenetic studies were also done whenever required and with EUTOS patients were stratified and then treatment was individualized.Results: Total 100 patients were enrolled. The mean age of presentation of CML was 44.7 years with M:F ratio was 1:1. 20% cases were of Hindu religion and 80% cases were Sikh by religion. 68% cases from rural area and 32% cases were from urban area. 92 patients were in chronic phase, 5 patients in accelerated phase, and 3 patients were found in blast crisis phase. Out of total 100 cases, 32% cases were of high-risk group and 68% cases were of low-risk group on presentation according to EUTOS.Conclusions: Most CML patients in north west Punjab are young (31-40 years) with male: female ratio is 1:1. Majority of them were Sikh by religion and from rural area. Most of them presented in Chronic phase of the disease and with low-risk strata according to EUTOS. Conclusion is that most patients presents in early phase of disease, with anaemia, leucocytosis or splenomegaly.

Integration of EHR and Cancer Registry Data to Construct a Childhood Cancer Survivorship Cohort to Improve Long-Term Follow-up Care for Leukemia and Lymphoma Survivors

Background and aims: Marked improvements in pediatric oncology care for children with acute lymphoblastic leukemia (ALL) and leukemia emphasize the importance of appropriate follow-up care for childhood cancer survivors (CCS) to monitor for late effects. The electronic health record (EHR) sparks unique approaches to construct, maintain, and leverage CCS cohorts to improve care. The primary purpose of this study is to utilize a novel approach to integrate EHR and cancer registry data to construct a CCS cohort for ALL and lymphoma survivors. Data mining in the EHR facilitates stratification of patients into risk cohorts and identification of high risk patients with inadequate subspecialty survivorship care Methods Cancer registry data from January 1, 1994 to November 30, 2012 provided a base cohort from which pediatric oncology clinic visits were extracted from the EHR. Explanatory variables included gender, race/ethnicity, risk stratification, distance to the medical center, rural-urban commuting area code, and area deprivation index as a measure of socioeconomic status. Primary outcomes included date of last clinic visit to determine appropriate follow-up, defined as being seen in a subspecialty clinic between five and seven years after initial diagnosis. Patients who died or relapsed during the seven year follow-up period were excluded from the analysis. Results Between January 1, 1994 and November 30, 2012, there were 262 pediatric oncology patients evaluated at our institution and reported in the cancer registry who were alive and without evidence of recurrence seven years after initial diagnosis. Of these patients, 22% of patients (n=57) were considered to have inadequate follow-up care. In univariate analysis, younger age (p&amp;lt;0.001), primary diagnosis of ALL compared to lymphoma (p=0.007), low risk strata (p=0.014) and closer proximity to primary treatment center or Children's Oncology Group-affiliate site (p=0.013 and 0.026, respectively) were associated with a higher likelihood of follow-up care. Multivariable logistic regression modeling is currently ongoing. Conclusions Integration of EHR and cancer registry data represents a feasible and novel approach to construct a cohort of childhood ALL and lymphoma survivors, risk stratify patients based on treatment exposures, and assess adequate subspecialty survivorship care. Future applications include refinement of treatment exposures through the EHR, adherence to guideline recommendations, and other late effects outcomes. Disclosures No relevant conflicts of interest to declare.

High-Risk Multiple Myeloma Patients Are Missed without Gene Expression Profiling

High-Risk Multiple Myeloma Patients Are Missed without Gene Expression Profiling

Association Between Early Severe Cardiovascular Events and the Initiation of Treatment With the Anti–Interleukin 12/23p40 Antibody Ustekinumab

Ustekinumab, a monoclonal antibody targeting interleukin 12/23p40 (IL-12/23p40), is effective in the treatment of moderate to severe psoriasis, psoriatic arthritis, and Crohn disease. In 2011, a meta-analysis of randomized clinical trials reported a potential risk of severe cardiovascular events (SCEs) within the first few months after the initiation of anti-IL-12/23p40 antibodies. To assess whether the initiation of ustekinumab treatment is associated with increased risk of SCEs. This case-time-control study used data from the French national health insurance database, covering 66 million individuals, on all patients exposed to ustekinumab between April 1, 2010, and December 31, 2016, classified according to their cardiovascular risk level (high- and low-risk strata). The risk period was the 6 months before the SCE, defined as acute coronary syndrome or stroke, and the reference period was the 6 months before the risk period. Statistical analysis was performed from September 20, 2017, to July 6, 2018. The initiation of ustekinumab treatment was screened during the risk and reference periods. Odds ratios for the risk of SCE after the initiation of ustekinumab treatment were calculated. Of the 9290 patients exposed to ustekinumab (4847 men [52%]; mean [SD] age, 43 [14] years), 179 experienced SCEs (65 cases of acute coronary syndrome, 68 cases of unstable angina, and 46 cases of stroke). Among patients with a high cardiovascular risk, a statisically significant association between initiaton of ustekinumab treatment and SCE occurrence was identified (odds ratio, 4.17; 95% CI, 1.19-14.59). Conversely, no statistically significant association was found among patients with a low cardiovascular risk (odds ratio, 0.30; 95% CI, 0.03-3.13). This study suggests that the initiation of ustekinumab treatment may trigger SCEs among patients at high cardiovascular risk. In line with the current mechanistic models for atherosclerotic disease, the period after the initiation of anti-IL-12/23p40 may be associated with atherosclerotic plaque destabilization via the inhibition of helper T cell subtype 17. Although the study interpretation is limited by its observational design, these results suggest that caution may be needed in the prescription of ustekinumab to patients at high cardiovascular risk.

Initial Validation of a Machine Learning-Derived Prognostic Test (KidneyIntelX) Integrating Biomarkers and Electronic Health Record Data To Predict Longitudinal Kidney Outcomes.

Individuals with type 2 diabetes (T2D) or the apolipoprotein L1 high-risk (APOL1-HR) genotypes are at increased risk of rapid kidney function decline (RKFD) and kidney failure. We hypothesized that a prognostic test using machine learning integrating blood biomarkers and longitudinal electronic health record (EHR) data would improve risk stratification. We selected two cohorts from the Mount Sinai BioMe Biobank: T2D (n=871) and African ancestry with APOL1-HR (n=498). We measured plasma tumor necrosis factor receptors (TNFR) 1 and 2 and kidney injury molecule-1 (KIM-1) and used random forest algorithms to integrate biomarker and EHR data to generate a risk score for a composite outcome: RKFD (eGFR decline of ≥5 ml/min per year), or 40% sustained eGFR decline, or kidney failure. We compared performance to a validated clinical model and applied thresholds to assess the utility of the prognostic test (KidneyIntelX) to accurately stratify patients into risk categories. Overall, 23% of those with T2D and 18% of those with APOL1-HR experienced the composite kidney end point over a median follow-up of 4.6 and 5.9 years, respectively. The area under the receiver operator characteristic curve (AUC) of KidneyIntelX was 0.77 (95% CI, 0.75 to 0.79) in T2D, and 0.80 (95% CI, 0.77 to 0.83) in APOL1-HR, outperforming the clinical models (AUC, 0.66 [95% CI, 0.65 to 0.67] and 0.72 [95% CI, 0.71 to 0.73], respectively; P<0.001). The positive predictive values for KidneyIntelX were 62% and 62% versus 46% and 39% for the clinical models (P<0.01) in high-risk (top 15%) stratum for T2D and APOL1-HR, respectively. The negative predictive values for KidneyIntelX were 92% in T2D and 96% for APOL1-HR versus 85% and 93% for the clinical model, respectively (P=0.76 and 0.93, respectively), in low-risk stratum (bottom 50%). In patients with T2D or APOL1-HR, a prognostic test (KidneyIntelX) integrating biomarker levels with longitudinal EHR data significantly improved prediction of a composite kidney end point of RKFD, 40% decline in eGFR, or kidney failure over validated clinical models.

Derivation and validation of a clinical prediction rule for thrombolysis-associated major bleeding in patients with acute pulmonary embolism: the BACS score.

Improved prediction of the risk of major bleeding in patients with acute pulmonary embolism (PE) receiving systemic thrombolysis is crucial to guide the choice of therapy. The study included consecutive patients with acute PE who received systemic thrombolysis in the RIETE registry. We used multivariable logistic regression analysis to create a risk score to predict 30-day major bleeding episodes. We externally validated the risk score in patients from the COMMAND VTE registry. We also compared the newly created risk score against the Kuijer and RIETE scores. Multivariable logistic regression identified four predictors for major bleeding: recent major Bleeding (3 points), Age >75 years (1 point), active Cancer (1 point), and Syncope (1 point) (BACS). Among 1172 patients receiving thrombolytic therapy in RIETE, 446 (38%) were classified as having low-risk (none of the variables present, 0 points) of major bleeding according to the BACS score, and the overall 30-day major bleeding rate of this group was 2.9% (95% CI, 1.6-4.9%), compared with 44% (95% CI, 14-79%) in the high-risk group (>3 points). In the validation cohort, 51% (149/290) of patients were classified as having low-risk, and the overall 30-day major bleeding rate of this group was 1.3%. In RIETE, the 30-day major bleeding event rates in the Kuijer and RIETE low-risk stratum were 5.3% and 4.4%, respectively. The BACS score is an easily applicable aid for prediction of the risk of major bleeding in the population of PE patients who receive systemic thrombolysis.

Diagnostic Performance of Emergency Physician Gestalt for Predicting Acute Appendicitis in Patients Age 5 to 20Years.

Pediatric appendicitis remains a challenging diagnosis in the emergency department (ED). Available risk prediction algorithms may contribute to excessive ED imaging studies. Incorporation of physician gestalt assessment could help refine predictive tools and improve diagnostic imaging decisions. This study was a subanalysis of a parent study that prospectively enrolled patients ages 5 to 20.9years with a chief complaint of abdominal pain presenting to 11 community EDs within an integrated delivery system between October 1, 2016, and September 30, 2018. Prior to diagnostic imaging, attending emergency physicians enrolled patients with ≤5days of right-sided or diffuse abdominal pain using a Web-based application embedded in the electronic health record. Predicted risk (gestalt) of acute appendicitis was prospectively entered using a sliding scale from 1% to 100%. As a planned secondary analysis, we assessed the performance of gestalt via c-statistics of receiver operating characteristic (ROC) curves; tested associations between gestalt performance and patient, physician, and facility characteristics; and examined clinical characteristics affecting gestalt estimates. Of 3,426 patients, 334 (9.8%) had confirmed appendicitis. Physician gestalt had excellent ROC curve characteristics (c-statistic=0.83, 95% confidence interval=0.81 to 0.85), performing particularly well in the low-risk strata (appendicitis rate=1.1% in gestalt 1%-10% range, negative predictive value of 98.9% for appendicitis diagnosis). Physicians with ≥5years since medical school graduation demonstrated improved gestalt performance over those with less experience (p=0.007). All clinical characteristics tested, except pain <24hours, were significantly associated with physician gestalt value (p<0.05). Physician gestalt for acute appendicitis diagnosis performed well, especially in low-risk patients and when employed by experienced physicians.

P327 Concordance between regadenoson stress echocardiography and spect-MIBI (99mTc) in patients with suspected myocardial ischemia

Abstract Background Regadenoson is a selective A2A adenosine receptor agonist approved for the detection of myocardial with SPECT-MIBI. To date, few studies have appraised the utility of this drug using transthoracic echocardiography with the same purpose. We designed this prospective study to evaluate the diagnostic agreement between these two techniques used simultaneously to detect myocardial ischemia. We report our first results evaluating the concordance between both techniques. Methods Patients enrolled were referred to our clinic for the evaluation of chest pain. Myocardial perfusion imaging was performed with regadenoson (400 µg ) which was injected before 99mTc-MIBI. Stress and rest sets of images were evaluated for relative uptake of the radiotracer in order to detect and characterize perfusion defects. The echocardiogram was acquired 60-90 seconds after the administration of the drug using a standardized technique. Two independent observers (one cardiologist and one specialist in nuclear medicine) interpreted the images, both unaware of the result of the complementary technique. Results One hundred twenty five patients were included, 69 (55%) of them males. Median age was 73 years (IQR 65-79). One hundred seventeen patients had both studies interpretable. Thirty-nine (32.5%) patients had had a diagnosis of ischemic heart disease before the study (either a myocardial infarction or a coronary revascularization). The median pre-test probability of coronary artery disease in those without a true previous diagnosis was 30.5% (16.0-50.8), being 50% in the intermediate risk (15-85%) and 49% in the low risk stratum (less than 15%). Significant reductions is systolic and diastolic blood pressure were detected with regadenoson [systolic 134 (21) mmHg Vs 131 (23) mmHg, p &amp;lt; 0.001; diastolic: 78 (12) mmHg Vs 73 (13) mmHg, p &amp;lt; 0.001], with a striking increase in heart rate: 67 (13) bpm Vs 91 (17) bpm, p &amp;lt; 0.001. More patients had a SPECT test showing myocardial ischemia (28% Vs 16%; p &amp;lt; 0.001). Agreement between both tests were 84%. The kappa index obtained from this sample was 0.34 (CI95% 0.15-0.53). Segregating those patients without a history of coronary artery disease (n = 81), the positive test rate were 14% for echo and 20% for SPECT-MIBI (p = 0.019). Kappa index was even lower: 0.29 (CI95% 0.02-0.56). Conclusions. SPECT-MIBI provides more positive tests than stress echocardiography when regadenoson is used as the stressor agent. The concordance between tests is low.

A Predictive Model for Early Death after Frontline Hypomethylating Agents in Elderly Unfit Acute Myeloid Leukemia Patients: Results from the Pethema Group

INTRODUCTION: A sizable proportion of elderly acute myeloid leukemia (AML) patients receive frontline hypomethylating agents (HMAs), namely azacitidine (AZA) and decitabine (DAC), as they are deemed unfit for intensive chemotherapy (ICT) by their treating physicians. A foreseeable high early death (ED) rate and lack of overall survival (OS) benefit under ICT are the main drivers for this decision. Several groups have published different predictive tools for ED or OS in elderly patients receiving ICT but, since ED in patients treated by HMAs is lower, the research activity has been restricted to OS in this population. METHODS: 415 elderly AML patients (264 M, 152 F) aged 61-90 receiving frontline HMAs (AZA 297, DAC 118), either in daily practice or within clinical trials (AZA 27, DAC 17), with complete relevant clinical information (see Table I) were available from the PETHEMA epidemiologic AML registry (NCT02006004). We analyzed the predictive value for ED (8wk) of the prognostic factors for OS/ED in AML included in the Walter, MRC/LRF, ALFA and ALMA scoring systems, namely age, WBC count, performance status (PS), MRC 2010 cytogenetics, platelet count and secondary disease, as well as the type of HMA. The potential predictors were categorized following previous published models (Walter, MRC/LRF, ALFA, ALMA). Cumulative early death rate at 8 weeks was calculated by the life-time method and the relevant strata were tested for univariate significance by the Wilcoxon test. All significant covariates were included in a Cox multivariate regression model and those significant for death at 8wk were included in a new predictive tool (HMA-EDS). Patients were assigned randomly in a 1:1 ratio to a training cohort (TC) and a validation cohort (VC). The different scoring systems (Walter, MRC/LRF, ALFA, ALMA, HMA-EDS) were checked for their prognostic impact on ED. Finally the 95% CI for the expected death rate at 8wk for the different strata of the new model was calculated for the full patient series. RESULTS: 51 patients out of 415 died and 13 were lost to follow-up before day 56 (cumulative ED rate at 8wk 13%, 95%CI 9-17%). Age, cytogenetics, secondary AML, platelet count and type of HMA were not significantly associated to ED. PS and WBC count strata confirmed their prognostic utility both in univariate and multivariate analysis (Table II). We developed the HMA-EDS by adding WBC (cutoffs 10 and 50, scores 1/2/3) and PS (0-1/2-4, scores 0/1) that classified patients in low-risk (score 1-2/ 84.6% of patients) and high risk (3-4/ 15.4% of patients) strata. When the prognostic utility for ED in the TC and the VC for the different scoring AML systems were checked, only HMA-EDS predicted ED in both cohorts (see Table III). The new EDS discriminates 2 different strata for ED at 8wk in unfit AML patients treated by HMA (see Figure 1 &amp; Table III), namely a lower-risk group (ED rate 10%, 95% CI 6-14) and a high-risk group (ED rate 26%, 95% CI 14-38). CONCLUSIONS: WBC count and PS are the main predictors for ED in unfit AML patients treated by HMAs. A new tool (HMA-EDS) discriminates two different risk groups and supersedes other previously published prognostic systems (Walter's, Wheatley's MRC/LRF, ALFA and ALMA) for this purpose. This score could be useful to select patients for front-line HMA or even HMAs-based combination therapies, given that several cycles are usually needed to achieve a clinical response. We suggest that other patient-related covariates such as geriatric assessment be checked in future studies. Disclosures Ramos: Daiichi Sankyo: Honoraria; Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen: Consultancy, Honoraria; Abbvie: Honoraria. Fernandez:Incyte: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Teva: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Daiichi Sankyo: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Karyopharm: Membership on an entity's Board of Directors or advisory committees, Research Funding; Abbvie: Membership on an entity's Board of Directors or advisory committees; Pfizer: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Novartis: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau.

External validation and comparison of multiple prognostic scores in allogeneic hematopoietic stem cell transplantation

AbstractClinical decisions in allogeneic hematopoietic stem cell transplantation (allo-HSCT) are supported by the use of prognostic scores for outcome prediction. Scores vary in their features and in the composition of development cohorts. We sought to externally validate and compare the performance of 8 commonly applied scoring systems on a cohort of allo-HSCT recipients. Among 528 patients studied, acute myeloid leukemia was the leading transplant indication (44%) and 46% of patients had a matched sibling donor. Most models successfully grouped patients into higher and lower risk strata, supporting their use for risk classification. However, discrimination varied (2-year overall survival area under the receiver operating characteristic curve [AUC]: revised Pretransplantation Assessment of Mortality [rPAM], 0.64; PAM, 0.63; revised Disease Risk Index [rDRI], 0.62; Endothelial Activation and Stress Index [EASIx], 0.60; combined European Society for Blood and Marrow Transplantation [EBMT]/Hematopoietic Cell Transplantation-specific Comorbidity Index [HCT-CI], 0.58; EBMT, 0.58; Comorbidity-Age, 0.58; HCT-CI, 0.55); AUC ranges from 0.5 (random) to 1.0 (perfect prediction). rPAM and PAM, which had the greatest predictive capacity across all outcomes, are comprehensive models including patient, disease, and transplantation information. Interestingly, EASIx, a biomarker-driven model, had comparable performance for nonrelapse mortality (NRM; 2-year AUC, 0.65) but no predictive value for relapse (2-year AUC, 0.53). Overall, allo-HSCT prognostic systems may be useful for risk stratification, but individual prediction remains a challenge, as reflected by the scores' limited discriminative capacity.

A dynamical systems perspective on driver behavior

Dynamical systems (DS) theoretical and methodological approaches have furthered our understanding of human development and behavior across many different domains, but have not yet been applied to driver behavior. Using a DS lens, a new theory of driver behavior is proposed, the phase transition framework (PTF), and then applied to understanding how novice drivers develop. The PTF can facilitate new lines of research by providing an individual-level explanatory account that yields the widely observed and poorly understood aggregate declines in novice drivers’ crash rates in the early months of licensure. For the purpose of the PTF, learning to drive is defined as changes in the psychological system integral to the acquisition, refinement, automation, and maintenance of competencies necessary for safe driving. In the PTF, the greatest reductions in population-level collision rates observed in the early months of the post-licensure period are hypothesized to be due to qualitative cognitive reorganizations (e.g., adoption of strategies that reduce risk) that cause abrupt shifts in crash risk trajectories into lower risk strata followed by refinement of strategy use and deployment through real-world experience. Critically, the PTF differs from other theories of driver behavior in that it holds that drivers’ behavioral development is self-organizing and emergent.