PHILADELPHIA – The rate of neurologic events in aortic stenosis patients at high risk for surgery was higher in those treated with transcatheter aortic valve replacement than in those who underwent open replacement in the PARTNER trial, and patients who had the procedure transapically had the greatest incidence.
However, the rates of neurologic events were low in all groups in the PARTNER (Placement of Aortic Transcatheter Valves) trial, Dr. D. Craig Miller of Stanford (Calif.) University reported at the meeting.
The neurologic substudy compared the risk of neurologic complications in 657 high-risk symptomatic patients with critical aortic stenosis receiving transcatheter aortic valve replacement (TAVR) or open replacement (AVR), as well as the difference between using a transapical or a transfemoral approach.
The PARTNER investigators ran two parallel trials in 1,057 patients with severe, symptomatic aortic stenosis: Cohort A comprised 699 patients at high risk for AVR, while cohort B comprised 358 inoperable patients.
Cohort B patients were randomized to either TAVR, using a transfemoral approach, or medical therapy. Results demonstrated superior survival and quality of life following TAVR compared with standard medical therapy; however, stroke occurred significantly more frequently with TAVR (N. Engl. J. Med. 2010;363:1597-607).
Cohort A patients were first assessed for suitability for either a transfemoral or transapical approach in TAVR, then randomized within those groups to either TAVR or AVR. Dr. Miller presented the neurologic outcomes of cohort A.
The main results of the cohort A trial, reported in April by Dr. Craig Smith, chief of cardiothoracic surgery at New York–Presbyterian Hospital/Columbia University Medical Center in New York, at the American College of Cardiology annual meeting, showed that all-cause mortality at 30 days favored (TAVR) at 3.4%, compared with 6.5% for open surgery, and was similar at 1 year (24% and 27%, respectively). The difference at 1 year reached the trial’s prespecified noninferiority margin with a P value of .001. But again, neurologic complications occurred more frequently after TAVR.
Stroke was a major secondary end point and was defined as a focal deficit lasting greater than 24 hours or a focal deficit lasting greater than 24 hours with positive neuron-imaging studies. Each event was adjudicated in a blinded review by an independent clinical events committee and subclassified (ischemic, hemorrhagic, or undetermined etiology).
There were 47 patients who had 49 neurologic events, and most (72%) were ischemic.
There was, however, a statistically significant difference in all neurologic events between AVR and TAVR at 30 days (2.4% vs. 5.5%, respectively) and 1 year (4.3% vs. 8.3%).
In the current study presented by Dr. Miller, a hazard analysis showed that there was an early high-peaking hazard phase and a later constant hazard phase for the likelihood of neurologic events.
In the early hazard phase, TAVR was a significant risk factor for neurologic events, compared with AVR, as was a smaller aortic valve area (AVA) index in the TAVR group.
In the late constant hazard phase, significant risk factors were patient- and disease-related. They included higher New York Heart Association functional class, stroke or TIA within 6-12 months, and nontransfemoral TAVR assessment.
Dr. Miller emphasized that there were "remarkably low 30-day mortality rates in these very elderly, very-high-risk aortic stenosis patients in both arms of the study." All neurologic events occurred more frequently in the TAVR arm compared with the AVR arm at 30 days and 1 year. In the more favorable transfemoral stratum, the rate of neurologic events was about threefold higher in the transfemoral TAVR group, at 4.6%, compared with the open AVR control group, at 1.4%, at 30 days, which persisted out to 1 year (6.1% and 1.9%, respectively). He also stated that "major stroke rates at 30 days were also very low (AVR, 2.3%; TAVR, 3.8%)." But he cautioned that the number of major strokes (29) was too small to support robust statistical hazard analysis or parametric modeling.
He emphasized that the higher observed incidence of neurologic events in the nontransfemoral candidate stratum reflected the patient substrate, and was not related to the transapical TAVR or AVR procedures per se.
Overall, taking the competing hazard of death into consideration, which is a strong force in these elderly patients with many comorbidities, the likelihood of being alive and having sustained a neurologic event was lowest in AVR patients in the transfemoral stratum, and highest in the transapical TAVR group, he concluded.
The PARTNER Trial was sponsored by Edwards Lifesciences. Dr. Miller is a consultant for Medtronic Cardiovascular and St. Jude Medical. He is a faculty member on the PARTNER U.S. Pivotal Trial Executive Committee and the Stanford Principal Investigator for the PARTNER Trial.