Stroke Rates During First 5 Years of the TAVR Era: Low but Stubbornly Stable

Postprocedure stroke occurred in just 2.3% of patients undergoing transcatheter aortic valve replacement (TAVR) in the first five years after the procedure’s FDA approval, but this rate remained notably constant throughout the period despite increasing operator experience and improved device technology. So finds an analysis of the Transcatheter Valve Therapy (TVT) Registry recently reported in JAMA (2019;321:2306-2315) by a team led by Cleveland Clinic researchers.

“One-month stroke risk after TAVR was low during the first five years of the procedure’s clinical use, but because of stroke’s potentially devastating consequences, it is imperative to identify how to improve,” says the study’s senior and corresponding author, Samir Kapadia, MD, Chair of Cardiovascular Medicine at Cleveland Clinic. “The lack of improvement in stroke rates over time indicates that we have more work to do to refine stroke prevention strategies.”

A nearly complete national sample

The study drew data from the joint Society of Thoracic Surgeons/American College of Cardiology TVT Registry from November 9, 2011, through May 31, 2017, with 30-day follow-up ending June 30, 2017. Since participation in the registry is required for reimbursement by the Centers for Medicare & Medicaid Services, nearly all patients undergoing TAVR procedures in the U.S. were included.

A total of 101,430 patients from 521 hospitals were assessed. Their median age was 83 years, and nearly half (47.1%) were women. Treatment was via femoral access in 83.9% of patients.

Outcomes of interest were rates of stroke and transient ischemic attack (TIA) 30 days after TAVR, as well as stroke’s association with 30-day mortality and medical therapy at discharge.

Findings yield insights into stroke risk

The analysis generated several key findings:

• The 30-day stroke rate was lower than expected. Within 30 days post-TAVR, 2.3% of patients had a stroke (95% CI, 2.2%-2.4%) and 0.4% had a TIA (95% CI, 0.3%-0.4%). This stroke rate is considerably lower than the rates in clinical trials leading to TAVR approval, which ranged from 3.4% to 6.7%. A possible explanation for the difference, the authors note, is that neurological assessment following TAVR was likely not as thorough in clinical practice as during the trials. Moreover, they add, registry reporting of neurological events is voluntary, which likely leads to underreporting.
• Stroke risk was remarkably stable over the five-year period. The annual stroke rate varied only slightly across the study period — from 2.2% to 2.4%. The lack of improvement over time puzzled the researchers, as progressively lower-risk patients were being treated, operator experience was mounting and device technology was improving as the study progressed. It is possible that the ongoing addition of inexperienced sites to the TVT Registry might have offset these factors, the authors speculate.
• Stroke risk was highest immediately postprocedure. Among patients who had a stroke, 48.9% of the events occurred within the first day after TAVR and 68.4% occurred within three days.
• Stroke was linked to a sixfold increase in mortality risk. The 30-day mortality rate was significantly higher among patients who had a stroke (16.7%) versus those who did not (3.7%), for a risk-adjusted hazard ratio of 6.1.
• Type of medical therapy was not associated with stroke risk. After propensity score matching, stroke risk was not found to be linked with use of either dual antiplatelet therapy or oral anticoagulant therapy. This held true whether patients underwent TAVR via femoral or nonfemoral access. The authors note, however, that because the registry recorded only discharge medications, they were unable to assess the potential effect of therapies on in-hospital strokes, which represented the majority of 30-day strokes.

The analysis also identified factors associated with higher risk of stroke, which included older age, female sex, prior stroke, presence of other vascular disease, nonfemoral access, and the use of a self-expanding valve rather than a balloon expandable valve.

“This snapshot of the first five years of TAVR experience in the U.S. brings up many intriguing questions,” says Dr. Kapadia. “We know from prior studies that the risk of stroke is not higher with TAVR compared with surgical AVR, but we still want to reduce the risk further. This study provides a framework for future research on how to do so.”

Limitations and future directions

Registry-based studies have inherent limitations, the study’s authors note, including reliance on incomplete information, voluntary reporting of some factors and often nonstandardized protocols among institutions.

Randomized investigation is warranted, Dr. Kapadia says, particularly regarding use of cerebral embolic protection during TAVR since most strokes in this setting are procedural strokes. He adds that further study is also needed of intensive antithrombotic therapy after TAVR, which was limited in scope in this analysis, and of the association of post-TAVR stroke risk with atrial fibrillation.

In an editorial accompanying the study in JAMA, vascular neurologist Steven Messé, MD, and cardiac surgeon Gorav Ailawadi, MD, write that the study “provides important insight into stroke complicating TAVR in clinical practice.” They point out that even with the higher stroke risk reported in clinical trials of TAVR, “there was a clear mortality benefit and quality of life was improved.” They conclude: “There is little doubt that TAVR is a major step forward for patient care.”