A Machine Learning Model for the Accurate Prediction of 1-Year Survival in TAVI Patients: A Retrospective Observational Cohort Study

BACKGROUND

predicting the 1-year survival of patients undergoing transcatheter aortic valve implantation (TAVI) is indispensable for managing safe early discharge strategies and resource optimization.

METHODS

Routinely acquired data (134 variables) were used from 629 patients, who underwent transfemoral TAVI from 2012 up to 2018. Support vector machines, neuronal networks, random forests, nearest neighbour and Bayes models were used with new, previously unseen patients to predict 1-year mortality in TAVI patients. A genetic variable selection algorithm identified a set of predictor variables with high predictive power.

RESULTS

Univariate analyses revealed 19 variables (clinical, laboratory, echocardiographic, computed tomographic and ECG) that significantly influence 1-year survival. Before applying the reject option, the model performances in terms of negative predictive value (NPV) and positive predictive value (PPV) were similar between all models. After applying the reject option, the random forest model identified a subcohort showing a negative predictive value of 96% (positive predictive value = 92%, accuracy = 96%).

CONCLUSIONS

Our model can predict the 1-year survival with very high negative and sufficiently high positive predictive value, with very high accuracy. The "reject option" allows a high performance and harmonic integration of machine learning in the clinical decision process.

A New Definition of Thrombocytopenia Following Transcatheter Aortic Valve Implantation: Incidence, Outcome, and Predictors

Background: The aim of this study was to assess the incidence, outcomes, and risk factors associated with thrombocytopenia following TAVI according to a corrected platelet count (CPC), to avoid the bias of hemodilution/concentration. Methods: We analyzed patients who underwent TAVI in our center between 2009 and 2018. The study population were divided into three groups: none (NT), mild (MT), and severe (ST) postoperative thrombocytopenia. Primary outcomes were bleedings, length of hospital stay, and mortality. A multivariate logistic regression was performed to assess risk factors for ST. Results: A total of 907 patients were included in the analysis. MT was observed in 28.1% and ST in 2.6% of all patients. The following clinical outcomes were recorded: incidence of life-threatening and major bleeding (NT = 14.2%, MT = 20.8%, ST = 58.3%), the median length of hospital stay (NT = 8, MT = 10, ST = 14 days), in-hospital mortality (NT = 3.9%, MT = 6.3%, ST = 16.7%), and the overall significance in comparison with NT (p < 0.05). The logistic regression showed ST was associated with preoperative CPC, transapical access, diabetes mellitus, and the critical preoperative state. Conclusions: Worse clinical outcomes are associated with both MT and ST after TAVI. In particular, ST is associated with higher in-hospital and 30-day mortality. Management of modifiable baseline and procedural variables may improve this outcome.

Two approaches-one phenomenon-thrombocytopenia after surgical and transcatheter aortic valve replacement

BACKGROPUND AND AIM

Postoperative thrombocytopenia after surgical aortic valve replacement (SAVR) or transcatheter aortic valve replacement (TAVR) and aggravating causes were the aim of this retrospective study.

METHODS

Data of all patients treated with SAVR (n = 1068) and TAVR (n = 816) due to severe aortic valve stenosis was collected at our center from 2010 to 2017. Preprocedural and postprocedural values were collected from electronic patient records.

RESULTS

There was a significant drop in platelets in both groups, the TAVR group showed overall superior platelet preservation compared to the AVR group (P < .001). In the SAVR subgroup analysis, a significant difference in platelet preservation was observed between the valve types (P < .001), particularly with the Freedom SOLO valve. In the TAVR subgroup analysis, the valve type did not influence platelet count (PLT) reduction (P = .13). In the SAVR subgroup analyses, PLT was found to be worsened with cardiopulmonary bypass (CPB) duration.

CONCLUSION

Thrombocytopenia frequently occurs after implantation of a biological heart valve prosthesis, with a higher frequency observed in patients after cardiac surgery rather than TAVR. Although some surgical bioprosthetic models are more susceptible to this phenomenon, CPB duration seems to be a major determinant for the development of postoperative thrombocytopenia.

Aortic valve calcification as a risk factor for major complications and reduced survival after transcatheter replacement

BACKGROUND

Aortic valve calcification is supposed to be a possible cause of embolic stroke or subclinical valve thrombosis after transcatheter aortic valve replacement (TAVR). We aimed to assess the role of aortic valve calcification in the occurrence of in-hospital clinical complications and survival after TAVR.

METHODS

We retrospectively analyzed preoperative contrast-enhanced multidetector computed tomography scans of patients who underwent TAVR on the native aortic valve in our center. Calcium volume was calculated for each aortic cusp, above and below the aortic annulus. Outcomes were recorded according to VARC-2 criteria.

RESULTS

Overall, 581 patients were included in the study (SapienXT = 192; Sapien3 = 228; CoreValve/EvolutR = 45; Engager = 5; Acurate = 111). Median survival was 4.98 years (interquartile range 4.41-5.54). Logistic regression identified calcium load beneath the right coronary cusp in left ventricular outflow tract (LVOT) as significantly associated with stroke (odds ratio [OR] 1.2; 95% confidence interval [CI] 1.03-1.3; p = 0.0019) and in-hospital mortality (OR 1.1; 95% CI 1.004-1.2; p = 0.04), whereas total calcium volume of the LVOT was associated with both in-hospital and 30 day-mortality (OR 1.2; 95% CI 1.01-1.4; p = 0.03, and OR 1.2; 95% CI 1.02-1.43; p = 0.029, respectively). Cox regression identified total calcium of LVOT (hazard ratio [HR] 1.18; 95% CI 1.02-1.38; p = 0.026), male sex (HR 1.88; 95% CI 1.06-3.32; p = 0.031), baseline creatinine clearance (HR 0.96; 95% CI 0.93-0.98; p < 0.001), and baseline severe aortic regurgitation (HR 7.48; 95% CI 2.76-20.26; p < 0.001) as risk factors associated with lower survival.

CONCLUSION

LVOT calcification is associated with increased risk of peri-procedural stroke and mortality as well as shorter long-term survival.

Cell/surface interactions and adhesion on Ti-6Al-4V: Effects of surface texture

This paper presents the results of an experimental study of the effects of surface texture on the interactions between human osteo-sarcoma (HOS) cells and Ti-6Al-4V. These include the Ti-6Al-4V with polished (smooth); Al(2)O(3) blasted (rough); and laser micro-grooved geometries with controlled spacings and depths. Immuno-fluorescence staining of adhesion proteins (actin and vinculin) was used to study the spreading and adhesion of HOS cells in 2 day culture experiments. Quantitative measures of adhesion were also obtained using an enzymatic detachment assay. The results are discussed within the context of existing theories of cell adhesion. The implications of the results are also examined for the design of textured surfaces in biomedical systems.

Superconductivity in the filled cage compounds Ba6Ge25 and Ba4Na2Ge25

The clathrate compound Ba 6Ge25 and its relatives consist of a rigid germanium skeleton, into which barium or other metal atoms are coordinated. These guest atoms can "rattle" freely at high temperatures, but in Ba 6Ge25 some of them lock randomly into split positions below T(S) approximately 200 K. The resulting bad metal undergoes a BCS-like superconducting transition at T(c) approximately 0.24 K. T(c) increases more than 16-fold, as T(S) is suppressed by hydrostatic pressure p, but changes only slightly with p from T(c) approximately 0.84 K in the undistorted sister compound Ba 4Na2Ge25. The large enhancement of T(c) in Ba 6Ge25 may be attributed mainly to the pressure tuning of strong disorder caused by the random displacement of Ba atoms at T(S).

YbRh(2)Si(2): pronounced non-fermi-liquid effects above a low-lyingmagnetic phase transition

We report the first observation of non-Fermi-liquid (NFL) effects in a clean Yb compound at ambient pressure and zero magnetic field. The electrical resistivity and the specific-heat coefficient of high-quality single crystals of YbRh(2)Si(2) present a linear and a logarithmic temperature dependence, respectively, in more than a decade in temperature. We ascribe this NFL behavior to the presence of (presumably) quasi-2D antiferromagnetic spin fluctuations related to a very weak magnetic phase transition at T(N) approximately 65 mK. Application of hydrostatic pressure induces anomalies in the electrical resistivity, indicating the stabilization of magnetic order.