‘Watchful Waiting’ Tied to Poor Long-term Outcomes in Patients With Severe AR, Analysis Suggests

Waiting until symptoms appear that “trigger” the decision to undertake surgery may not be the best option for patients with severe, asymptomatic aortic regurgitation (AR), according to new research. The findings, coming amid growing calls to intervene earlier in this population, are at odds with American and European guidelines.

“AR is a generally a clinically benign condition,” said senior author Jean-Louis Vanoverschelde, MD, PhD (Cliniques Universitaires Saint-Luc, Brussels, Belgium), in an email. “However, our data indicate that there are some risk factors that should be considered in asymptomatic patients,” specifically an ejection fraction (EF) below 55% and LV end-systolic dimension (LVESD) > 20 mm/m².

“Asymptomatic patients should probably be operated [on] before reaching these thresholds. These are significantly different than those proposed in the guidelines,” Vanoverschelde advised.

Current recommendations from the American Heart Association/American College of Cardiology and the European Society of Cardiology recommend holding off on surgery for patients with AR until they become symptomatic or develop a degree of LV dysfunction (NYHA class >1 or EF < 50%) sufficient to justify AR replacement or repair. Class II triggers, which are primarily based on the magnitude of volume overload, are considered level C evidence for surgery and are thus open for interpretation and clinical judgment.

In their study published online January 16, 2019, in JACC: Cardiovascular Imaging, Vanoverschelde and colleagues led by Christophe de Meester, PhD (Université catholique de Louvain, Brussels, Belgium), found that patients with Class I or Class II triggers paid an “outcome penalty” in the form of a higher risk of dying or developing heart failure symptoms over long-term follow-up than those with no symptoms at the time of surgery. The difference persisted after adjustment for baseline characteristics, suggesting that the time to refer for surgery is likely “before the onset of LV dysfunction, or at least very shortly after its development,” the researchers write.

There should be a change, and the resistance of the guideline committee . . . is incomprehensible. Maurice E. Sarano

Commenting on the study for TCTMD, Maurice E. Sarano, MD (Mayo Clinic, Rochester, MN), agreed with the conclusions and with the authors’ contention about reconsideration of current recommendations.

“There should be a change, and the resistance of the guideline committee . . . is incomprehensible,” he said in an interview.

“The only thing you can do is have scientific work proving that we should change,” Sarano observed. “Whether or not it will change is a matter of the people on the committee [considering] their conscience and the evidence.”

Outcomes at Odds With Guideline-Based Triggers

For the study, de Meester el al examined data on 356 patients who underwent surgery between 1995 and 2014. Of these, 57% presented with Class I triggers, 21% with Class II triggers, and 22% with no triggers for surgery. Most patients (80%) had aortic valve (AV) repair, while the remaining patients had a Ross procedure (7%), bioprosthetic AV replacement (9%), or mechanical AV replacement (4%).

Operative mortality was similar among all patients regardless of trigger class. However, survival at 10 years was significantly better for those with no preoperative triggers (89%) compared with those with Class II triggers at time of surgery (85%) or those in whom Class I triggers were present (71%; P = 0.010 for all comparisons).

Analysis by LVEF showed similar results, with better 10-year survival in patients with a preoperative LVEF ≥ 50% vs a preoperative LVEF < 50% (80% vs 69%; P = 0.011). Body surface area (BSA) was also a powerful predictor, with a survival advantage seen among patients with preoperative LVESD/BSA < 25 mm/m² compared with LVESD/BSA ≥ 25 mm/m² (84% vs 62%; P < 0.001). To TCTMD, Sarano said that finding may be of particular importance when risk-stratifying women, who tend to be smaller than men.

The findings of de Meester and colleagues also diverge from the European and American guidelines with regards to the importance of end-diastolic dimensions as a trigger for surgery, showing no association between the severity of LV dilatation as measured by LV end-diastolic volume and outcomes at 10 years.

Finally, the paper also suggests that when patients develop NYHA class III/IV symptoms just prior to surgery, their odds of long-term survival decrease precipitously in comparison with patients who remain asymptomatic (49% vs 86%; P < 0.001).

Same Conclusion, Different Population = No Change

To TCTMD, Sarano said the young age of many patients with AR and the fact that their excess mortality risk is not “gigantic” as a consequence of waiting for surgery is likely a large part of why the guideline committees have been resistant to changing the recommendations despite data and the opinion of many physicians that denying patients earlier surgery is hurting patients in the long run.

“We have to accumulate work showing that overall the mortality is not extremely high, but it is still a penalty that you impose on patients by waiting and you have to be more proactive with the right use of the right trigger,” he said.

Sarano’s group conducted a similar study several years ago in mitral regurgitation (MR) patients, finding the those with Class I or II triggers had higher mortality and more progression to congestive heart failure than those who underwent early surgery. Like in the current paper, Sarano and colleagues called for a revision to the guidelines at that time. To date, nothing has changed.

"Even though patients improve after the operation, they still pay a price [for waiting]. That is true for MR patients and it's also true now as demonstrated by this work in aortic regurgitation," he said.

As for the specific changes called for by de Meester and colleagues, Sarano said he is in complete agreement. "The guidelines, in my mind, have to be corrected by elevating the threshold of EF to 55%, by using the end-systolic dimension corrected for body surface area, and by stopping the use of the end-diastolic diameter," he emphasized.

Sarano added that a forthcoming paper by his group adds more evidence for the importance of body surface area in treatment decisions, showing that LVESD/BSA is also a powerful predictor of survival for AR patients managed medically.