A transcatheter aortic-valve implantation (TAVI)-first strategy is noninferior to the standard percutaneous coronary intervention (PCI)-first approach in terms of clinical outcomes at 1 year in patients with severe aortic stenosis and concomitant coronary artery disease (CAD), according to the TAVI PCI trial presented at ESC 2026.
“About half of patients with severe aortic stenosis undergoing TAVI also have CAD … Performing PCI before TAVI represents the current standard approach, but the preferred treatment strategy has not previously been studied,” said lead author Prof Barbara Elisabeth Stähli from the University Hospital Zurich, Switzerland.
Therefore, the TAVI PCI trial was designed to address this evidence gap in this patient population.
The trial enrolled 986 patients (median age 82 years, 66 percent male) with severe aortic stenosis and CAD who were deemed eligible for both TAVI and PCI by a multidisciplinary heart team. Participants were randomized in a 1:1 ratio to undergo either TAVI before PCI (TAVI-first group; n=498) or PCI before TAVI (PCI-first group; n=488), with both procedures performed within 1–45 days.
The primary endpoint of the study was the composite of death, nonfatal myocardial infarction, ischemia-driven revascularization, valve-, procedure-, or heart failure-related rehospitalization, or life-threatening, disabling, or major bleeding at 1 year.
At 1 year, the primary endpoint occurred in 22.2 percent of patients in the TAVI-first group compared with 24.2 percent in the PCI-first group. The risk difference between the groups was –2 percentage points (95 percent confidence interval, –7.4 to 3.4), which fell within the noninferiority margin of 6.6 percentage points (p<0.001 for noninferiority). [ESC 2026, Hot Line Session 7]
The results showed no heterogeneity in the predefined subgroups, including age, sex, diabetes status, Syntax score, timing of procedures, access site, renal function, or left ventricular ejection fraction, Stähli noted.
Notably, the TAVI-first group experienced a numerically lower incidence of life-threatening, disabling, or major bleeding compared with the PCI-first group (6.6 percent vs 9.7 percent), and BARC* type 3a bleeding was also less frequent (3.5 percent vs 6.8 percent), “although this secondary finding should be interpreted cautiously. Antiplatelet therapy and the interval between the two procedures may play a role, and these questions need further study,” noted Stähli.
“Interestingly, fewer patients in the TAVI-first group ultimately underwent PCI. This was not because coronary access was impaired after TAVI. Rather, once the valve had been treated, the clinical decision about whether PCI was still necessary sometimes changed,” said Stähli.
Serious adverse events (AEs) occurred in 53 percent of patients in the TAVI-first group and in 55.9 percent of those in the PCI-first group. The overall incidence of serious AES was comparable between the two groups. [NEJM 2026;doi:10.1056/NEJMoa2606924]
Overall, “the results of the TAVI PCI trial demonstrate that we as clinicians now have a choice. The trial provides randomized evidence to individualize the treatment strategy according to what is best for each patient. In patients for whom both TAVI and PCI were planned, a TAVI-first strategy was noninferior to a PCI-first strategy at 1 year,” said Stähli.
“PCI therefore does not necessarily have to come first. For some patients, there may be good reasons to treat the coronary arteries first. For others, treating the valve first may make more sense. Symptoms, anatomy, bleeding risk, coronary disease, and the overall clinical situation all matter,” she added.