Abstract 437 Manual support during robotic-assisted PCI in Germany

Background

Robotic-assisted percutaneous coronary intervention (rPCI) has proven feasible and safe while reducing radiation exposure for the operator. In some rPCI cases, manual support is required for successful completion of PCI but data regarding this important aspect are scarce. We therefore aimed to (1) evaluate the need for manual support during rPCI and (2) identify associated procedure-related factors.

Methods

We used individual-level data from four sites across Germany (University Hospital Hamburg-Eppendorf, University Heart Center Freiburg - Bad Krozingen, University Hospital of Giessen and Marburg, St. Johannes Hospital Dortmund). At all institutions, rPCI was performed using the 2nd generation CorPath GRX Vascular Robotic System from 2020 to 2022. Manual support was defined as the composite of manual assistance (partial manipulation with ultimate completion of the procedure robotically) and manual conversion (required bedside manipulation of either the guide catheter, guidewire, or delivery system to complete the procedure). Data harmonisation and statistical analyses were performed at the University Heart & Vascular Center Hamburg. To determine a best performing multivariable model predicting manual support we used a two-step selection process. Due to small data, we added Firth’s logit to the final model. The AUC was corrected for overoptimism by bootstrap correction with 250 bootstrap samples.

Results

A total of 231 coronary lesions were treated with rPCI in 210 patients (median age 69.0 years; 25.7% female). Baseline characteristics including age, sex, cardiovascular risk factors and comorbidities were similar among patient with and without manually assisted rPCI. Manual support was required in 19.9 % of lesions and was more often required in rPCI for complex coronary lesions (Type C: 78.3% vs. 41.6%, p<0.001), chronic total occlusions (CTO) PCI (17.4% vs. 0.5%, p<0.001), aorto-ostial lesions (15.2% vs. 3.8%, p=0.009) and true bifurcations (26.1% vs. 7.7%, p=0.001). Multivariate analysis identified aorto-ostial lesions (OR 1.22 [1.00-1.50], CTO (OR 1.96 [1.46-2.65], true bifurcations (OR 1.44 [1.23-1.68), severe calcification (OR 1.15 [0.99-1.33], and the number of implanted DES (OR 1.11 [1.03-1.19], as predictors for manual assistance (AUC=0.80).

Conclusion

In this multi-center analysis, we found manual support to be necessary in approximately one-fifth of rPCI procedures. This was particularly frequent in complex coronary lesions, providing additional insights into the utility of rPCI in different clinical scenarios.

Clin Res Cardiol (2023). https://doi.org/10.1007/s00392-023-02180-w
Manual support during robotic-assisted PCI in Germany – a multi-center analysis
F. J. Brunner¹, J. Rilinger², F. J. Hofmann³, V. Tiyerili⁴, A. Goßling¹, B. Bay¹, H. Möllmann⁴, H. Nef³, C. von zur Mühlen², M. Seiffert¹
¹Klinik für Kardiologie, Universitäres Herz- und Gefäßzentrum Hamburg, Hamburg; ²Klinik für Kardiologie und Angiologie, Universitäts-Herzzentrum Freiburg - Bad Krozingen, Freiburg im Breisgau; ³Medizinische Klinik I - Kardiologie und Angiologie, Universitätsklinikum Gießen und Marburg GmbH, Gießen; ⁴Klinik für Innere Medizin I, Kath. St. Paulus Gesellschaft, Dortmund;
Background Robotic-assisted percutaneous coronary intervention (rPCI) has proven feasible and safe while reducing radiation exposure for the operator. In some rPCI cases, manual support is required for successful completion of PCI but data regarding this important aspect are scarce. We therefore aimed to (1) evaluate the need for manual support during rPCI and (2) identify associated procedure-related factors. Methods We used individual-level data from four sites across Germany (University Hospital Hamburg-Eppendorf, University Heart Center Freiburg - Bad Krozingen, University Hospital of Giessen and Marburg, St. Johannes Hospital Dortmund). At all institutions, rPCI was performed using the 2nd generation CorPath GRX Vascular Robotic System from 2020 to 2022. Manual support was defined as the composite of manual assistance (partial manipulation with ultimate completion of the procedure robotically) and manual conversion (required bedside manipulation of either the guide catheter, guidewire, or delivery system to complete the procedure). Data harmonisation and statistical analyses were performed at the University Heart & Vascular Center Hamburg. To determine a best performing multivariable model predicting manual support we used a two-step selection process. Due to small data, we added Firth’s logit to the final model. The AUC was corrected for overoptimism by bootstrap correction with 250 bootstrap samples. Results A total of 231 coronary lesions were treated with rPCI in 210 patients (median age 69.0 years; 25.7% female). Baseline characteristics including age, sex, cardiovascular risk factors and comorbidities were similar among patient with and without manually assisted rPCI. Manual support was required in 19.9 % of lesions and was more often required in rPCI for complex coronary lesions (Type C: 78.3% vs. 41.6%, p<0.001), chronic total occlusions (CTO) PCI (17.4% vs. 0.5%, p<0.001), aorto-ostial lesions (15.2% vs. 3.8%, p=0.009) and true bifurcations (26.1% vs. 7.7%, p=0.001). Multivariate analysis identified aorto-ostial lesions (OR 1.22 [1.00-1.50], CTO (OR 1.96 [1.46-2.65], true bifurcations (OR 1.44 [1.23-1.68), severe calcification (OR 1.15 [0.99-1.33], and the number of implanted DES (OR 1.11 [1.03-1.19], as predictors for manual assistance (AUC=0.80). Conclusion In this multi-center analysis, we found manual support to be necessary in approximately one-fifth of rPCI procedures. This was particularly frequent in complex coronary lesions, providing additional insights into the utility of rPCI in different clinical scenarios.
https://dgk.org/kongress_programme/jt2023/aV1089.html