New-Onset A-fib Tied to Particularly Poor Outcomes After TAVR

Patients with new-onset A-fib after TAVR have worse outcomes than those with preexisting A-fib or those in sinus rhythm at the time of the procedure, a study of Medicare beneficiaries shows.

A-fib has been associated with poorer TAVR outcomes before, but “the thing that was surprising for us was how bad the prognosis of new-onset AF was,” lead author Amgad Mentias, MD (University of Iowa Carver College of Medicine, Iowa City), told TCTMD in an email.

In fact, patients who developed A-fib after the procedure carried higher risks of mortality and hospitalizations for bleeding, stroke, and heart failure compared with those who already had the arrhythmia at the time of TAVR.

“Physicians should consider new-onset AF patients a high-risk population that warrants close monitoring and observation in the first few months after the TAVR procedure,” Mentias said.

“Decisions about prescribing anticoagulation or dual antiplatelet therapy after TAVR or placing [a left atrial appendage] closure device in these patients should probably be tailored to individual patients after determining thromboembolic and bleeding risk and after an informed discussion with the patient,” he continued. “It is not currently clear whether providers should even go a further step and utilize an event monitor to detect subclinical new-onset AF in the first 30 days in high-risk patients after the procedure, as our study did not include patients with subclinical AF.”

Commenting for TCTMD, Matthew Sherwood, MD (Inova Heart and Vascular Institute, Falls Church, VA), said new-onset A-fib is not necessarily more harmful than preexisting A-fib, pointing to the limitations inherent to any observational study.

But, he said, “certainly it seems that AF in general, new-onset or preexisting, is a marker for patients that indeed have worse outcomes after TAVR. And that may be related to other comorbidities that go along with AF or a signature that the AF is correlated with sicker patients in general who have worse outcomes.”

As for what should be done about that, Sherwood said “significant and special care should be taken for these patients to watch for those bad outcomes, [with] perhaps closer follow-up, especially in the new-onset atrial fibrillation patients, to determine what additional medical therapies they might need.”

New-Onset A-fib Declines Over Time

Nearly half of patients undergoing TAVR either have A-fib at the time of the procedure or will develop it later, with multiple studies showing that the arrhythmia signals a worse outcome. A study conducted in Israel, for example, showed that rates of stroke or death at 1 year were higher in patients with preexisting—but not new-onset—atrial fibrillation. Another study found that both new-onset and preexisting A-fib were associated with higher 1-year mortality. However, there is little data comparing the impact of new-onset versus preexisting A-fib.

For this study, published online October 16, 2019, ahead of print in JACC: Cardiovascular Interventions, Mentias et al examined Medicare inpatient claims data on 72,660 patients 65 and older (mean age 81.9 years; 53% men) who underwent non-apical TAVR between 2014 and 2016. Four out of every 10 patients had A-fib at the time of the procedure and another 6.8% developed it afterwards. The rate of new-onset A-fib declined from 10.8% in 2014 to 4.9% in 2016.

Patients who developed the arrhythmia after TAVR tended to be older and to have longer stays in the hospital and the intensive care unit, were most likely to need a new pacemaker, and were least likely to get discharged on the next day or within 72 hours. Comorbidity burden and average CHA₂DS₂-VASc score were highest in patients with preexisting A-fib.

Despite the difference in comorbidities, the all-cause mortality rate (per 100 person-years) through a median follow-up of 305 days was highest in the new-onset group (29.7), followed by the patients with preexisting A-fib (22.6) and those who remained in sinus rhythm (12.8).

After adjustment for patient characteristics and hospital TAVR volume, new-onset A-fib was associated with a greater risk of mortality compared with both preexisting A-fib (HR 1.35; 95% CI 1.26-1.45) and no A-fib (HR 2.06; 95% CI 1.92-2.20). The findings were similar when it came to hospitalizations for bleeding, stroke, and heart failure. Preexisting A-fib also was associated with poorer outcomes compared with no A-fib, although the difference in stroke failed to reach statistical significance.

Asked why new-onset A-fib would carry worse outcomes compared with preexisting disease, Mentias noted that “it is hard to determine causality or intermediary pathways with such an observational study design using administrative database.”

But, he said, there are some possible explanations. For one, new-onset A-fib could be caused by direct injuries from the TAVR procedure. Or, “the higher risk could be a consequence of the AF itself as a disease, with the resulting irregular fast heart rate, the loss of atrial kick, and the potential drop in cardiac output,” Mentias speculated. “New-onset AF would differ from preexisting AF in that aspect in that preexisting AF patients are probably on established rate control and anticoagulation therapy, while new-onset AF patients may not have been on optimal medical therapy.”

Best Antithrombotic Therapy an Open Issue

The impact of treatment could not be addressed by this study because the database lacked information on use of anticoagulants and other medications.

“It would be interesting to have a study look into whether antithrombotic therapy would mitigate the risk of stroke in these patients, and what effect would that have on bleeding and mortality risk,” Mentias said. He added that there is also ongoing research—the WATCH-TAVR trial, for instance—to explore the utility of adding left atrial appendage closure at the time of TAVR in patients with A-fib who are not good candidates for long-term anticoagulation. Those results are “eagerly awaited,” he said.

According to Sherwood, there’s uncertainty about the best way to manage TAVR patients who either have or will develop A-fib. “The problem here is that there’s not enough data to tell us what the right thing is to do for these patients,” he said. “And so it’s a really important need for large-scale studies for TAVR patients.”

The dearth of data, which is in stark contrast to the bounty of randomized data informing the treatment of patients with A-fib who are undergoing PCI that has come out in recent years, has led to substantial variation in treatment out in practice, Sherwood said. Based on a prior study from his group that was published as an abstract, he said “there is significant variation, including what I believe is underuse of oral anticoagulation, in patients with AF who undergo TAVR, and then a lot of variation in the combinations of oral anticoagulation and antiplatelet therapy in these patients. And those may or may not contribute to differences in clinical outcomes.”

Ongoing studies like POPULAR-TAVI might help provide some insight into the best treatment strategies, he said.