Cognitive Outcomes After TAVI

Editor's Note: Commentary based on Khan MM, Herrmann N, Gallagher D, et al. Cognitive outcomes after transcatheter aortic valve implantation: a metaanalysis. J Am Geriatr Soc 2018;66:254-62.

Rationale for Study/Background:
To quantitatively assess and summarize changes in cognitive function after Transcatheter Aortic Valve Implantation (TAVI).

Funding: None.


Design: Meta-analysis

Inclusion Criteria: Studies that included assessment of cognition using standardized neuropsychological measures (including Mini-Mental State Examination [MMSE], Montreal Cognitive Assessment [MoCA]) before and after TAVI in patients with severe aortic stenosis.

Exclusion Criteria: Studies that did not allow extraction of cognitive scores pre and post TAVI.

Exposure: TAVI

Primary Outcome(s): Changes in cognitive scores from pre-TAVI to postoperatively and at 1, 3, 6, 12 and 12-34 months follow-up.

Secondary Outcomes: Cognitive score changes by preoperative cognitive status, perioperative embolic lesions and post-procedural stroke.

Statistical Analysis: Standardized mean differences (SMD) were calculated to compare cognitive scores pre and post TAVI, and data were pooled using a random-effects meta-analysis. Meta-regression was used to assess factors influencing postoperative cognitive score changes.


A total of 18 studies were included in meta-analysis. Total number of patients was 1065 (sample size per study ranging from 10 to 229 patients); 15 studies had a follow-up of 6 months or shorter (follow-up ranging from 3 days to 34 months); MMSE was most commonly used. Compared with preoperative baseline, there was a statistically significant improvement in cognitive function at 1 month (SMD for baseline minus follow-up, -0.33; 95% CI, -0.50 to -0.16; p < 0.001; I2 = 0%). Otherwise no statistically significant changes from baseline were observed at 3, 6 or 12-34 months.

In meta-regression, preoperative cognitive deficits, perioperative embolic lesions and post-procedural stroke were not associated with short-term cognitive score changes; however, there was insufficient data to assess long-term cognitive score changes.

Conclusion: Compared to baseline, cognitive function seems stable peri-operatively and long term after TAVI.

Limitations of Study: The meta-analysis should be interpreted with consideration of limitations of the included studies: pre-to-post comparison design without a non-surgical group, modest sample sizes (only 7 studies had >50 patients), moderate non-response rates to follow-up cognitive assessment and only three of 18 studies reporting long-term (12-34 months) cognitive data. Across the studies, cognitive function was measured using various instruments, which makes the pooled SMD clinically difficult to interpret.

Cognitive trajectories could not be determined due to limited numbers of longitudinal follow-up assessments in most studies. Without individual patient data, meta-regression had limited ability to examine the impact of preoperative cognitive status, perioperative embolic events and post-procedural stroke on postoperative cognitive changes.

Geriatric Perspective for the Cardiovascular Clinician:

As TAVI becomes an increasingly available therapeutic option for frail older adults, postoperative cognitive decline is a main concern for patients undergoing TAVI. Cognitive function is critical in postoperative recovery, functional independence and quality of life. This meta-analysis of 18 studies provides some reassurance that TAVI may not be detrimental to cognitive function.

This meta-analysis found a statistically significant improvement in cognitive function at 1 month, roughly equivalent to a 1-point change in MMSE or MoCA, based on conversion from reported change in SMD. Although this might be a true change due to improved cerebral perfusion after TAVI, this improvement was not observed at subsequent cognitive assessments. Alternative explanations for a measured transient improvement include learning effect, selective drop-out of those with impaired cognitive function, and depressed pre-operative scores due to anxiety about testing or the procedure. Therefore, the clinical significance of finding such a modest improvement at 1 month remains unclear and difficult to interpret.

It is important to understand that the pooled estimates from meta-analysis may not represent the experience for individual patients. The effect of preexisting cognitive impairment, embolic events and stroke on postoperative cognitive function cannot be excluded based on statistically non-significant results from meta-regression of a modest number of studies. Also, the effects of postoperative delirium, which can have both immediate and long-term negative effects on cognitive function, were not examined in this meta-analysis. Thus due to the variation in cognition and health status that exists among older adults, stable mean cognitive scores after TAVI do not indicate stability at an individual patient level.

In summary, the current body of evidence suggests that cognitive function remains stable at least in the year following TAVI. While this finding is reassuring, more research is needed to study long-term cognitive function and to identify characteristics of patients at risk for cognitive decline after TAVI. Ultimately, cautious interpretation guided by careful clinical judgment remains essential to personalized decision making for patients.

Clinical Topics: Cardiac Surgery, Geriatric Cardiology, Invasive Cardiovascular Angiography and Intervention, Valvular Heart Disease, Aortic Surgery, Cardiac Surgery and VHD, Interventions and Structural Heart Disease

Keywords: Meta-Analysis, Geriatrics, Judgment, Frail Elderly, Transcatheter Aortic Valve Replacement, Follow-Up Studies, Quality of Life, Aortic Valve Stenosis, Cognition, Aortic Valve, Heart Valve Prosthesis, Stroke, Anxiety, Decision Making

< Back to Listings