P2273Predictors of short- and mid-term outcomes after TAVR in low-flow, low-gradient aortic stenosis

Background

There is a lack of data on outcomes in classical (C-LFLG) and paradoxical low-flow, low-gradient aortic stenosis (P-LFLG) patients undergoing TAVR.

Purpose

We aim to compare baseline characteristic and procedural outcomes between C-LFLG, P-LFLG and high-gradient aortic stenosis (HG-AS) patients undergoing TAVR.

Methods

Patients included in the Transcatheter RegistrY of aorTic valve biOprosthesis in Latin-AMerica (TRYTOM Registry) were divided in 3 groups: 1) HG-AS: mean transaortic gradient (MG) ≥40 mmHg; 2) P-LFLG: MG <40 mmHg and left ventricular ejection fraction (LVEF) ≥50%; 3) C-LFLG: MG <40 mmHg and LVEF <50%. The outcomes were evaluated at 30-days and 1-year and were classified according to definitions of the VARC-2.

Results

1040 patients were included, 677 (65%) classified as HG-AS, 223 (21%) as P-LFLG and 140 (14%) as C-LFLG. Median follow-up was 16 months (range 0–109). There were baseline differences between HG-AS, P-LFLG and C-LFLG regarding age (80±7 vs 80±5 vs 78±8 years, respectively; p=0.017), NYHA FC III and IV (61.0 vs 72.6 vs 83.6%, respectively; p<0.001), coronary artery disease (44.1 vs 47.1 vs 57.9%, respectively; p=0.012), EuroSCORE II (7.2±6.3 vs 7.5±5.0 vs 12.9±10.4%, respectively; p<0.001), LVEF (56±11 vs 61±7 vs 32±9%, respectively; p<0.001), MG (53±13 vs 30±6 vs 27±7 mmHg, respectively; p<0.001), aortic valve area (0.65±0.16 vs 0.74±0.15 vs 0.70±0.16 cm2, respectively; p<0.001) and creatinine (1.2±0.7 vs 1.1±0.5 vs 1.5±1.3 mg/dl, respectively; p<0.001). Despite these significant baseline differences, we found similar outcomes after TAVR between HG-AS, P-LFLG and C-LFLG regarding device success (89.8 vs 95.1 vs 90.7%, respectively; p=0.057), in-hospital mortality (6.1 vs 5.9 vs 11.8%, respectively; p=0.144) and all other VARC-2 major outcomes, including major bleeding, major vascular complication and disabling stroke. In addition, female sex (OR 2.13, 95% CI 1.16–3.92, p=0.014), LVEF (OR 1.02, 95% CI 1.00–1.04, p=0.039) and MG (OR 0.97, 95% CI 0.95–0.99, p=0.004) were the only predictor of device success by multivariate analysis. Furthermore, 1-year mortality was similar among the groups (9.5 vs 8.3 vs 14.3%, respectively; p=0.358; Figure 1), and by multivariate analysis, diabetes (HR 2.44, 95% CI 1.10–5.41, p=0.028), creatinine (HR 1.65, 95% CI 1.17–2.33, p=0.004), conversion to general anesthesia (HR 7.93, 95% CI 2.08–30.20, p=0.002) and post-procedure disabling stroke (HR 12.84, 95% CI 3.09–53.40, p<0.001) predicted increased 1-year mortality, irrespective on the LVEF and MG.

Conclusions

Apart from baseline differences, TAVR in P-LFLG and C-LFLG was feasible and with similar clinical outcomes when compared to HG-AS. Mid-term mortality rates was associated with diabetes, creatinine and procedure complications.

P837Neural networks algorithms improve GRACE Score performance

Background

Global Registry of Acute Coronary Events (GRACE) score is a well-known model used to predict the probability of events in acute coronary syndrome (ACS). GRACE model was developed using a logistic regression approach that can only model linear functions, a limitation that could be prevented using artificial neural networks (NN) a recognized tool for nonlinear statistical modeling. The aim of this study was to develop, train and test different NN algorithm-based models to improve the GRACE score performance.

Methods

We analyzed a prospective database including 1,255 patients admitted with diagnosis of ACS in a community hospital, between June 2008 and June 2017. The database included 40 demographic and laboratory admission variables. In the guided approach, only the individual predictors included in the GRACE score were used to train and test three NN algorithm-based models, one- and two-hidden layer multilayer perceptron (MLP), and a radial basis function network. In addition, three extra unguided models were built using the 40 admission variables. Finally, expected mortality according to the GRACE score was calculated using the logistic regression equation.

The database was split into 2 datasets: 70% for model training and 30% for validation. In order to choose the best model, the training process was repeated 50 times. Every time the models were tested on the validation cohort, accuracy, receiver operating characteristic (ROC) area, negative predictive value (NPV), and positive predictive value (PPV) were recorded. Only models showing the best discrimination power were selected for comparison with logistic regression outcomes. The end point was in-hospital all-cause mortality.

Results

In terms of accuracy, ROC area and NPV, almost all NN algorithms outperformed the logistic regression approach (accuracy 97.1, 96.7, 96.2, 97.3 and 94.1%, p<0.001; ROC area 0.89, 0.86, 0.84, 0.84 and 0.75, Hanley-McNeil p≤0.05; for guided and unguided one- and two-hidden layers MLP and GRACE score, respectively). Only radial basis function models obtained a better accuracy level based on NPV improvement (100 vs. 98.8%, p=0.0001), at the expense of PPV reduction (0.0% vs. 13.2%, p<0.0001) (ROC are 0.84 vs. 0.75, p=0.043). Compared with the logistic regression approach, one- and two-hidden layers in guided and unguided MLP models improved PPV from 13.2 to 18.2% (38% increase), 15.4% (17% increase), 27.3% (107% increase), and 25.0% (89% increase), respectively, although these differences were not statistically significant.

Conclusions

NN algorithms improve GRACE score performance in terms of discriminatory power for the prediction of in-hospital mortality. Its application should become a useful tool for the decision making in ACS patients

[Significance of matrix metalloproteinase-9, tissue inhibitor of metalloproteinase and protein Ki-67 in prostate tumors]

Immunohistochemical evaluation of localization of matrix metalloproteinase-9 (MMP-9) and an inhibitor of matrix metalloproteinase-1 (TIMP-1) and cell proliferative activity in the production Ki-67 protein in benign prostatic hyperplasia (BPH), and adenocarcinoma with different Gleason scores was performed. Moderate positive correlation between the Gleason scores and cell proliferation index usind Ki-67 antigen (rs = 0.674), moderate negative correlation between Gleason scores and levels of MMP-9 production (rs = -0.660), and weak significant negative correlation between the levels of cell proliferative activity and MMP-9 production by tumor cells (rs = -0.369) were established. Invasive properties of tumor cells, expressed in the destruction of type IV collagen in basement membrane and connective tissue of the stroma of the prostate, are associated with imbalance in MMP-9 protein, and blocking enzyme, TIMP-1; and TIMP-1 production is reduced significantly in adenocarcinomas with different Gleason scores compared with BPH.