In this installment of Med Op-Ed: A late-emerging safety signal highlights opportunities to improve postmarketing surveillance of heart devices; researchers face challenges validating findings based on UK Biobank data in other cohorts; and how to approach nonculprit lesions in ST-segment elevation myocardial infarction (STEMI).

TAVR Safety Signal Took 5 Years to Emerge

An emerging safety signal in the Evolut Low Risk Trial — an increased rate of reintervention among patients who received transcatheter aortic valve replacement (TAVR) compared with those who underwent surgical aortic valve replacement — became apparent only after sustained follow-up beyond 5 years. Although postmarket device surveillance worked to identify the signal in this instance, such follow-up often relies on extraordinary effort and transparency from researchers rather than routine, reliable infrastructure, Harlan M. Krumholz, MD, and Sanket S. Dhruva, MD, wrote in the Journal of the American College of Cardiology.

The Context

• Trial investigators in February reported that the reintervention rate at 6 years was 5.5% for TAVR vs 3.3% for surgery. The primary composite endpoint of all-cause mortality and disabling stroke was not significantly different.
• Permanently implanting any new medical device creates a new phase of care in which the durability of a device and its biological interactions might not be fully understood for years or decades, the authors of the editorial wrote.
• Nevertheless, post-approval studies for high-risk devices can be inconsistent in their timeliness and completeness. Active surveillance based on real-world experience has not yet been implemented at scale due to funding constraints, challenges integrating data, and incomplete information about long-term outcomes.

In Their Own Words

“Late safety or durability signals are not surprising. What is striking is how easily they can be missed in a system that learns slowly, episodically, and often too late. What makes the Evolut experience notable is not that a late signal emerged, but that it was detected, interrogated, and reported with transparency.”

Read it: When the System Works—and Why It so Often Does Not: Building a Learning Infrastructure for High-Risk Medical Devices

Do UK Biobank Findings Apply to Your Patients?

The UK Biobank has transformed cardiovascular research with deep phenotyping and genomic data from half a million participants. Still, external validation in diverse population cohorts is essential to enhance the reliability of research findings and support their translation to the clinic, Ahmed M. Salih, PhD; Adam J. Lewandowski, DPhil; and Zahra Raisi-Estabragh, MBChB, PhD, wrote in JAMA Cardiology.

The Context

• The UK Biobank has generated approximately 16,000 peer-reviewed papers, with nearly one third focused on cardiometabolic research.
• Researchers increasingly use machine learning techniques that leverage genomics, proteomics, metabolomics, multiorgan imaging, and electrocardiogram signals but face challenges replicating these approaches in cohorts where these data are missing or captured differently.
• Efforts to establish validation cohorts should prioritize data accessibility and transparent documentation of data collection and analytic methods, the authors wrote.

In Their Own Words

“The importance of validation analyses is unequivocally acknowledged but infrequently implemented in cardiovascular research.”

Read it: Opportunities, Challenges, and Validation Strategies to Advance Cardiovascular Research Using Population Cohorts

Nonculprit Lesions in STEMI: Assess Now or Revisit Later?

The severity of nonculprit coronary lesions in hemodynamically stable patients with STEMI may inform physicians’ approach to the assessment and treatment of such lesions, suggested William F. Fearon, MD, in The New England Journal of Medicine.

The Context

• A recent trial found patients were more likely to receive percutaneous coronary intervention for nonculprit lesions if they were immediately assessed with a pressure wire using instantaneous wave-free ratio rather than if they underwent a delayed, noninvasive assessment using cardiac stress MRI (43% vs 19%).
• The rate of death, recurrent myocardial infarction, or hospitalization for heart failure at 3 years was similar regardless of the approach to assessing nonculprit lesions.
• The severity of nonculprit lesions in this trial — about two thirds had 50%-70% stenosis — was less than in some prior studies, Fearon noted.

In Their Own Words

“It may be that more severe non-culprit lesions cannot be safely deferred, because the chance for spontaneous myocardial infarction is more likely in severe lesions, especially if the [fractional flow reserve] is abnormal. Until we have more data, guidelines recommend that patients in hemodynamically stable condition with STEMI and severe nonculprit lesions should continue to undergo early angiography-guided complete revascularization, whereas those with less severe nonculprit lesions can undergo a more deliberate ischemia-guided approach.”

Read it: Assessing Nonculprit Coronary-Artery Lesions in STEMI

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.