New Microbleeds Not Uncommon Post-TAVI

An observational study shows no short-term harm with these acute bleeds but doesn’t rule out possible cognitive decline.

New Microbleeds Not Uncommon Post-TAVI

Nearly one in four patients who undergo TAVI with a balloon-expandable valve develop new microbleeds visible on MRI in the days thereafter, according to results from the small, observational METHYSTROKE study.

Yet researchers caution that the clinical importance of these acute bleeds, especially over the longer term, is not yet known. In METHYSTROKE, there were no differences in in-hospital outcomes, 6-month neurologic function and quality of life, or 12-month mortality based on new microbleeds.

Even in the general population of elderly individuals, screening with MRI can reveal chronic microbleeds, lead author Eric Van Belle, MD, PhD (CHU Lille, France), told TCTMD. Earlier studies have linked these sorts of chronic cases, which are more likely to occur in patients with cerebral small-vessel disease, to “increased risk for future ischemic strokes, cognitive impairment or dementia, and intracerebral hemorrhage, especially in the setting of antithrombotic use,” he and his colleagues note in their paper, which was published recently in Circulation.

Acute cases are known to occur in relation to endocarditis or after certain cardiac procedures, like surgery or left ventricular device implantation. What was unknown prior to this study, he said, is how such bleeding might manifest in TAVI patients and what its effects might be.

With TAVI, “we give a significant amount of anticoagulation” that can decrease thrombotic risk but increase hemorrhagic risk, Van Belle explained. On top of this, aortic stenosis (AS) in and of itself can lead to acquired von Willebrand factor (VWF) defects, which in turn raise the risk of bleeding and may persist after TAVI’s completion.

“So it’s why we decided to investigate this as much as we could” by imaging patients before their AS was treated and a few days thereafter, he said, acknowledging that their study was small, with fewer than 90 patients. “If we believe what we [see] with the chronic situation, it is very likely [acute cases] will have some impact,” said Van Belle. “If we want to prove this, we will need to do a larger study to investigate this specific question.”

It’s an additional argument for a simplified approach to TAVI. Eric Van Belle

Rishi Puri, MBBS, PhD (Cleveland Clinic, OH), who didn’t take part in the current study, said its results remind him of when, back in 2015, data began to roll out showing reduced leaflet motion with TAVI. At that time, he said, “we all kind of stopped to think: what’s really going on? While this is similar, I don’t think we necessarily need to put the brakes on TAVI or change too much what we’re doing. I think we need to understand a lot more what this means.”

Though the findings are “really interesting,” he said, they may not be specific to TAVI. Rather, “it may relate to vulnerable patients undergoing periprocedural anticoagulation,” which would also apply to other procedures like endovascular aneurysm repair or thoracic endovascular aortic repair, he commented to TCTMD. Or there could be other factors in this elderly population, like amyloid angiopathy, that put patients at risk of bleeding.

This mechanistic study “probably raises more questions than answers, but it’s something we need to think about,” Puri said, adding, “I think there’s a lot more to learn than what we’ve learned thus far. This is just another piece in that puzzle.”


For METHYSTROKE, Van Belle and colleagues obtained 1.5T MRIs on the day before and 3 days after TAVI for 84 patients (mean age 81 years; 47% men). All had concomitant shear-induced acquired VWF defect. Procedures were transfemoral and exclusively employed balloon-expandable Sapien valves (Edwards Lifesciences). No embolic protection devices were used.

Before the TAVI procedure, 22 (26%) of the patients had at least one microbleed and five (6%) had at least one cerebral embolism. After TAVI, rates were 50% and 65%, respectively. This amounted to new microbleeds occurring in 19 patients (23%) and new emboli in 55 patients (65%).

Whether patients had a microbleed at baseline did not impact their likelihood of developing one after the procedure. Nor did cerebral emboli (either before or after TAVI) affect microbleed risk.

On univariable analysis, post-TAVI microbleeds were associated with higher total dose of heparin, longer procedure time, absence of protamine reversion, higher final activated partial thromboplastin time, signs of VWF defect immediately after TAVI, and lower final closure time with adenosine–diphosphate.

Independent predictors of new microbleeds on multivariable analysis included both lengthier procedure times (OR 1.22 per 5 minutes of fluoroscopy; 95% CI 1.03-1.73) and postprocedural acquired VWF defect (OR 1.42 per 0.1-unit reduction in high-molecular-weight:multimer ratio; 95% CI 1.08-1.89).

In-hospital outcomes and 12-month survival rate did not differ between patients who did and didn’t develop post-TAVI microbleeds. Also, occurrence of these bleeds did not, at least within 6 months of follow-up, affect either neurologic functional outcome or quality of life as measured by National Institutes of Health Stroke Scale, modified Rankin Scale, Mini-Mental State Examination, and EQ-5D.

Van Belle said that, based on an observational study alone, it’s too early to change practice. “But it’s an additional argument for a simplified approach to TAVI,” keeping procedure times short, not using general anesthesia, and perhaps not using embolic protection devices, he noted, adding that their work suggests “we should be very cautious (and not too liberal) in our anticoagulation management during TAVR procedures. It further supports a strategy of tightly controlled duration of anticoagulation involving protamine reversion.”

Van Belle said he and other Paris-based researchers are planning a randomized study of protamine that will likely begin later this year.

It’s also possible that the TAVI-related microbleeds could have worse effects in younger and lower-risk patients, Van Belle noted, or that there might be differences among valve types, given that VWF dysfunction after TAVI is mainly related to paravalvular leak.

  • This work was supported by Lille-II University and by the National Research Agency (Programme d’Investissement d’Avenir) with the Hospital-University Research in Health program.
  • Van Belle reports no relevant conflicts of interest.
  • Puri reports consulting for Protembis.