Author + information
- Tullio Palmerini, MD,
- Giuseppe Biondi-Zoccai, MD and
- Gregg W. Stone, MD∗ ()
- ↵∗Columbia University Medical Center, Cardiovascular Research Foundation, 1700 Broadway, 8th Floor, New York, New York 10019
We thank Dr. Guo and colleagues for their interest in our pairwise and individual patient data meta-analysis showing increased rates of bleeding-related deaths with longer dual antiplatelet therapy (DAPT) duration compared with shorter DAPT in patients undergoing drug-eluting stent implantation (1). The network meta-analysis performed by the authors is interesting and, in some ways, complementary to our findings. It appears that the greater the difference in the duration of DAPT, the higher the likelihood of a significant difference in bleeding-related death and noncardiac mortality between treatment arms, a finding that is not surprising and is consistent with our results. We would caution the authors, however, in not overinterpreting the modest between-group differences noted in their meta-analysis. By dividing the studies into 3 discrete groups and performing multiple comparisons, sufficient statistical power may not be present to exclude clinically relevant differences in mortality among groups. Nonetheless, across each of the ≤6-month DAPT versus 1-year DAPT, ≤6-month DAPT versus ≥18-month DAPT, and 1-year DAPT versus ≥18-month DAPT comparisons, there was a consistent trend suggesting lower rates of bleeding-related deaths with shorter DAPT compared with longer DAPT, independent of the type of comparison. Given the same direction of the results and the overlapping 95% confidence intervals, it is likely that interaction testing would be negative. The same would be true for noncardiac deaths. These data are thus consistent with our meta-analysis, in which there was no significant heterogeneity in treatment effect for the relative rates of bleeding-related deaths and all-cause mortality across studies comparing different DAPT durations.
In addition, the authors did not report direct versus indirect estimates of risk for each individual endpoint; therefore, consistency between direct versus indirect evidence remains undetermined. Finally, reliance on a Bayesian network, lack of individual patient data, and missing adjustments for stent type and other baseline variables may have reduced precision and validity (2,3). In contrast, an individual patient data analysis provides substantially greater power to uncover relationships and adjust for differences between patients and studies.
Thus, in the large population of patients enrolled in these studies, prolonging DAPT increased the risk of bleeding and noncardiac mortality in large part because of bleeding-related deaths, an association that became more evident when DAPT was prolonged for more than 18 months. Nonetheless, patients at high risk for ischemic events may still benefit from prolonged DAPT. The optimal DAPT duration for individual patients should be determined by considering the specific demographic and anatomic factors that determine their absolute and relative risks of ischemic versus bleeding events.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose. P.K. Shah, MD, served as Guest Editor-in-Chief for this paper. Kalyanam Shivkumar, MD, PhD, served as Guest Editor for this paper.
- 2017 American College of Cardiology Foundation
- Palmerini T.,
- Bacchi Reggiani L.,
- Della Riva D.,
- et al.
- Biondi-Zoccai G.
- Piccolo R.,
- Gargiulo G.,
- Franzone A.,
- et al.