The use of robotics in interventional cardiology offers operators the ability to safely and successfully complete percutaneous coronary intervention (PCI) while minimizing occupational hazards due to x-ray exposure and wearing lead protection. Proof-of-concept of the first-generation CorPath 200 robotic system (Corindus Vascular Robotics, Waltham, MA) was demonstrated in the multicenter PRECISE registry of single-vessel robotic PCI (R-PCI) for relatively simple lesions. However, utility of R-PCI in clinical practice, especially with complex coronary anatomy, was unknown.

 

Comparing Robotic & Manual PCI

In a study published in the Journal of the American College of Cardiology: Cardiovascular Interventions, we and our colleagues sought to identify the safety and feasibility of R-PCI for complex coronary artery disease (CAD), as well as compare outcomes with similar manually completed PCI (M-PCI). This was an all-comers study excluding those presenting with ST-segment elevation myocardial infarction (MI) or requiring procedures ineligible for R-PCI.

A total of 334 procedures (108 R-PCI, 226 M-PCI) were performed during the study period (Table). The majority of R-PCI were completed entirely robotically (81.5%). Partial manual assistance occurred in 12 procedures (11.1%) and manual conversion in eight (7.4%) due to either adverse events (3 procedures, without MACE), technical limitation of the robotic platform (8 procedures), or limited guidewire or catheter support (9 procedures), resulting in 91.7% robotic technical success. Periprocedural MI occurred twice following M-PCI and once following R-PCI, resulting in 99.1% clinical success in both groups.

Baseline clinical characteristics for individuals in the R-PCI and M-PCI groups were similar overall. A total of 157 lesions were addressed robotically, including left main (4.6%), ostial (7.4%), bifurcation (7.4%), chronic total occlusion (5.6%), and saphenous vein graft (1.9%) lesions. R-PCI procedures also addressed multiple lesions (38.0%), multiple vessels (13.0%), or utilized hemodynamic support (3.7%).  Compared with M-PCI, the R-PCI group had more complex lesion characteristics with longer primary lesions (22.2 mm vs. 19.4 mm), more type B2/C lesions (81% vs. 69%), and higher overall SYNTAX score (19.6 vs. 15.7). Overall procedure time was longer (44.5 minutes vs. 36.6 minutes), but contrast volume and patient radiation exposure were lower in R-PCI. Procedure time remained longer following a propensity-matched analysis and was predominantly observed during R-PCI of relatively simple lesions.

 

What the Findings Mean

These findings highlight both the high safety (99.1% clinical success) and feasibility (91.7% robotic technical success) of R-PCI in complex CAD. Robotically addressed lesions were predominantly type B2/C complexity, encompassing an array of complex lesion types. R-PCI addressing multiple lesions, multiple vessels, or utilizing hemodynamic support were also included. Though comparisons were non-randomized, these procedural and lesion characteristics were no less complex than those of a “real-world” M-PCI comparator group. Longer procedure time was observed during R-PCI of low complexity lesions, possibly attributable to time spent loading the robotic drive, but this did not translate into increased patient radiation exposure, contrast volume used, or stents deployed.

From these findings, interventional cardiologists can gain confidence in the safe and efficacious application of R-PCI in complex real-world CAD management. Further understanding of current robotic platform limitations and anatomical challenges to R-PCI could lead to additional optimization. Additional analyses regarding specific reasons for manual assistance/conversion and cost effectiveness of this first-generation robotic platform would help inform decision making toward integration into daily practice. A randomized clinical trial of the second-generation CorPath GRX platform with added guide catheter control compared with M-PCI would provide validity to the comparable outcomes as observed in the CORA-PCI study with R-PCI.

Author