Author + information
- Salem Badr, MD,
- Danny Dvir, MD and
- Ron Waksman, MD⁎ ()
- ↵⁎Cardiovascular Research Institute, Washington Hospital Center, 110 Irving Street Northwest, Suite 4B-1, Washington, DC 20010
Mehran et al. (1) recently reported the results of a multicenter observational study examining long-term outcomes of 1,791 patients after percutaneous coronary intervention (PCI) for chronic total occlusion (CTO) lesions, comparing the patients who succeeded in the procedure with those who failed. The authors report an overall procedural success rate of 68% and detected in their model that a successful CTO procedure was an independent predictor of reduced cardiac mortality with a strong trend toward lower all-cause mortality. Although the authors should be congratulated for reporting on such a large cohort of patients undergoing PCI to CTO lesions, we found the analysis biased against the patients who failed PCI. Furthermore, there are several methodological deficiencies in the study that significantly impair the power of this study and put into question the accuracy of their conclusion.
To address the question of whether treating CTO by PCI impacts on late clinical events, the control group should have appropriately included patients assigned to medical therapy and not those who failed PCI. Comparing the treatment effect of a device between a group that succeeded in a procedure and another that failed might directly lead to a major bias and does not offer any meaningful conclusion other than the intuitive fact that when the procedure fails it is bad for the patient.
Second, the authors also reported that the rate of coronary artery bypass graft procedures for the failed PCI group was higher in patients whose occlusions could not be opened (13.3% vs. 3.2%, p < 0.01), leading to an impression that such an event is more frequent when the attempt to open a difficult CTO has failed; however, it might be related to vascular injuries that were more frequent in patients with failed PCI, such as coronary perforation (7.4% vs. 1.7%, p < 0.01) and residual dissection (9.4% vs. 4.3%, p < 0.01), thereby exaggerating the relative benefits of a successful opening of the occluded artery. Consistent with that, our group previously reported analysis of a cohort of patients with failed but uncomplicated CTO PCI procedures, showing similar rates of death and myocardial infarction at a mean follow-up of 2 years (2). It would be appropriate to repeat the analysis of the authors and compare the successful PCI group with the noncomplicated failed group and examine whether their conclusion still holds.
Finally, with regard to the use of drug-eluting stents (DES) versus bare-metal stents (BMS), the authors reported that treatment with DES in comparison with BMS resulted in similar definite/probable stent thrombosis rates (1.7% vs. 2.3%, p = 0.58); however, only 4.2% of patients in the DES group versus 42% of the patients in the BMS group reached 5-year follow-up. This major difference in follow-up time could lead to a bias as well.
We agree that performing a randomized clinical trial comparing PCI for CTO and conservative therapy with medications only, such as in the upcoming DECISION-CTO (Drug-Eluting Stent Implantation vs. Optimal Medical Treatment in Patients with Chronic Total Occlusion) trial, might reveal whether treating these complex lesions has an effect on clinical result.
- American College of Cardiology Foundation