Late Bleeding Not Rare, Ups Mortality After TAVI in All-Asian Cohort: OCEAN-TAVI

Late bleeding up to 3 years after TAVI is not uncommon—in fact is more likely than stroke—and portends worse mortality over time, even with minor bleeding events, according to the OCEAN-TAVI study.

One in 20 patients experienced major bleeding after being discharged from the hospital, researchers found.

“The results highlight the bleeding risk rather than stroke events in specific elderly TAVR cohorts,” lead author Masanori Yamamoto, MD (Toyohashi Heart Center, Japan), told TCTMD in an email. “We should pay attention to the risk-benefit balance of aggressive antithrombotic therapy for preventing stroke and thrombotic events after TAVR.”

Also, beyond what heart failure status and platelet counts can predict, “our data suggest the importance of the relationship between the late bleeding and frailty status,” he continued. “We should emphasize the importance of frailty assessment in the elderly TAVR cohort.”

The all-Asian patient population is notable as most TAVI data so far have come from Europe and the United States. Commenting on the results for TCTMD, Jurriën ten Berg, MD, PhD (St. Antonius Hospital, Nieuwegein, the Netherlands), said “it's very important to also have data from Japan and that they are consistent with the Western population.” Previous data in this cohort looking at those who have undergone PCI, he added, have shown that “they might be a little more susceptible to bleeds because they're smaller, . . . so especially in TAVI patients, you should be even more aware that bleeding might occur.”

More Bleeding Than Stroke

For the study, Yamamoto and colleagues included 2,518 patients (mean age 84.3 years; 30.5% male) enrolled in the Japanese multicenter registry who underwent TAVI between October 2013 and May 2017. Notably, 21.6% of the population was prescribed no or single antiplatelet therapy, while 54.1% were taking dual antiplatelet therapy (DAPT), 22.9% were taking oral anticoagulation (OAC) plus no or single antiplatelet therapy, and 1.4% were taking both OAC and DAPT.

Within 3 years after hospital discharge, the cumulative incidence rates of all late bleeding, major late bleeding, and ischemic stroke were 7.4%, 5.2%, and 3.4%, respectively. The all and major bleeding risks were significantly higher than the risk of stroke (P < 0.001). Platelet count < 14.9 x 10⁴/µL (OR 1.94), score ≥ 4 on the clinical frailty scale (OR 1.55), and NYHA functional class III/IV (OR 1.58) were all independent predictors of postdischarge bleeding.

Among the TAVI patients who experienced late bleeding, 6.4% reported events within 30 days, 22.1% at 30-60 days, 17.9% at 90-180 days, 17.1% at 180-365 days, 26.4% at 365-730 days, and 10% beyond 730 days. GI bleeding (40.7%) was the most common, followed by hemorrhagic stroke (25.75%) and bleeding caused by trauma (15.0%).

Patients taking oral anticoagulants saw higher rates of major bleeding than those not on the medications (P = 0.041), as did patients with atrial fibrillation compared with those without the condition (P = 0.042). Also, at the time of the bleeding event, 33.8% of patients who had been on DAPT at discharge had been de-escalated to none or a single agent, but a full 93.0% of patients on OAC maintained their medication from discharge to the time of bleeding.

Those with late bleeding had higher mortality over 3 years compared with those without bleeding (P < 0.001). When the bleeding types were broken down, GI bleeding, hemorrhagic stroke, and major bleeding all led to significantly higher mortality rates compared with no events (all P < 0.001), but even minor bleeding significantly increased the risk of death (P = 0.002).

On multivariate analysis, late bleeding as a time-varying covariate was associated with a more than fivefold increased risk of mortality following TAVR (HR 5.63; 95% CI 4.28-7.41). Specifically, GI bleeding led to a more than threefold increased risk in mortality (HR 3.38), while hemorrhagic stroke led to a more than tenfold increased risk (HR 10.5). Minor (HR 2.78) and major bleeding (HR 6.81) independently upped the mortality risk, as well.

“The data concerning late bleeding after TAVR is scarce in the Asian cohort,” Yamamoto said. “The East Asian paradox means the Asian cohort has an increased risk of bleeding over ischemic events when compared with Western patients. We needed to clarify the actual incidence, timing, and causes of late bleeding after TAVR in our registry data. As a result, the incidence of late bleeding in elderly Asian-Japanese cohorts is not significantly elevated in comparison with previous Western data. However, the worse prognosis is similarly observed regardless of the race differences.”

He stressed that aggressive antithrombotic therapy could be “harmful” for elderly TAVI patients and urged physicians to use only a short duration of DAPT like they would following PCI. “The [left atrial appendage] closure procedure to stop the OAC may be a good option in patients with AF and high bleeding risk,” Yamamoto added.

‘Substantial’ Contribution

In an accompanying editorial, Hasan Jilaihawi, MD (Cedars-Sinai Medical Center, Los Angeles, CA), writes that because of its size, enrollment of only Asian patients, and focus on late bleeding, the study stands out in a way that “make its contribution to the field substantial.”

The fact that GI bleeding emerged as the most common form of postdischarge bleeding means that future research should focus on “screening of GI and intracranial pathologies and conceivably for primary prevention,” he says. Also, “Heyde’s syndrome and other incidental GI pathologies in elderly patients, exacerbated by anticoagulation and antiplatelet therapy, may be important contributors to GI bleeding that should be actively investigated in the presence of even minor degrees of pre-TAVR anemia of unknown etiology.”

Jilaihawi stresses that intracranial bleeding holds implications even though the data don’t indicate whether these events were spontaneous or related to falls. “We often observe an increase in mobility and confidence post-TAVR with substantial cardiovascular symptom amelioration; it is important to have vigilance and physiotherapy support to ensure patient mobility is safe and patients are not at high risk of falls post-TAVR,” he writes.

For ten Berg, the incidence of bleeding was not unexpected. “But I was a bit surprised that even minor bleeding is predicting mortality,” he said. “That’s new if you compare with other populations, such as PCI patients. Most of the time it’s major bleeding which predicts mortality and especially, of course, intracranial bleeding, which was also affecting this TAVI population.”

He guessed that the high proportion of comorbidities in the study population as well as a great degree of frailty contributed. “The other surprise was the medical treatment of the patients, especially the antithrombotic treatment,” ten Berg said. “Because although there's lots of evidence that single antiplatelet therapy is the way to go, most of these patients were still treated with dual antiplatelet therapy, even at the time of the bleeding. . . . And patients who were on oral anticoagulant because of atrial fibrillation also were additionally treated with single antiplatelet therapy or even DAPT.”

It’s true that the population was enrolled before the release of the POPular TAVI data showing the benefits of single antiplatelet therapy after TAVI, but this suggests that antithrombotic therapy might be different in Japan compared with the US and Europe, he said.

Going forward, ten Berg said he would like to see these data compared with other registries of Asian patients, but he would most like to see a randomized trial of “very low doses of antithrombotic therapy” in Japanese people.