Author + information
- J. Matthew Brennan, MD, MPH∗ ()
- ↵∗Division of Cardiology, Duke University School of Medicine, 2400 Pratt Street, Room 0311 Terrace Level, Durham, North Carolina 27705
While I appreciate the concerns Dr. Hu and colleagues expressed regarding the inclusion of patients undergoing concomitant surgical revascularization in our comparison of transcatheter aortic valve replacement (TAVR) versus surgical aortic valve replacement (SAVR) outcomes (1), it is unlikely that this has meaningfully biased the analysis.
Dr. Hu and colleagues note 2 important considerations. First, surgical revascularization is a marker of more advanced coronary artery disease. As such, it is an important potential source of confounding and should be accounted for in any TAVR-SAVR comparison. Both the STS ACSD (Society of Thoracic Surgeons Adult Cardiac Surgery Database) and STS/ACC TVT (Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy) registry collect data on the number of diseased coronary arteries, and this variable (as well as the presence of “prior percutaneous coronary intervention” and “prior coronary artery bypass graft [CABG]”) was included in the propensity score used to balance risk in our 2 treatment cohorts. Our analysis achieved excellent balance across each of these variables. To the extent that these are appropriate surrogates for “severity of coronary artery disease (CAD),” it is unlikely that this is an important source of confounding in our analysis.
Second, concomitant CABG carries additional procedural risk. This is a more difficult issue to address. However, existing data from our analysis (particularly, among patients with “no CAD”) suggests that this is also unlikely to be an important source of bias, here. Revascularization is common practice prior to TAVR (for proximal lesions affecting large regions of myocardium) and at the time of SAVR among patients with CAD. Just as percutaneous revascularization is part of the TAVR strategy in these patients, concomitant CABG is a part of the SAVR strategy—and, the additional risk should be included when considering the effectiveness of the strategy. Whereas “severity of CAD” was balanced across the 2 cohorts (TAVR, SAVR), there is no good way to balance for revascularization using our existing data sources. In the United States, most percutaneous revascularization procedures were performed in a staged fashion prior to the intended TAVR procedure during the study era. However, the TVT registry does not collect information on patients who may have had a complication following percutaneous revascularization before they reached the intended TAVR procedure. To the extent that major complications are common in this setting, our analysis will have underestimated the risk of an intended percutaneous coronary intervention/TAVR approach. However, the incidence of major complications with percutaneous coronary intervention is rare in contemporary practice, and there is no evidence to suggest that this would be a sufficiently common issue among pre-TAVR patients to bias our results. In fact, comparison of TAVR versus SAVR outcomes among the 4,618 patients (49%) with “no CAD” (who would not have required either percutaneous or surgical revascularization) demonstrated similar results as among patients with 1-, 2-, or 3-vessel CAD—with no differences observed for 1-year outcomes across the 2 treatments.
For these reasons, I would argue that “severity of CAD” and concomitant surgical revascularization are not likely to be important sources of bias in our analysis, as suggested by Dr. Hu and colleagues. Furthermore, I would assert that inclusion of patients undergoing concomitant CABG (with adjustment for “severity of CAD”) is the most appropriate choice in this and similar analyses.
Please note: Dr. Brennan has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation