CHICAGO – Despite a steady rise in the number of patients presenting with aortic stenosis, many patients are not being referred for surgery because of the perception that the surgical risk is too high, according to Dr. Leonard Girardi.

The challenge for surgeons is to get the message out to their cardiology colleagues and to an increasingly sophisticated patient population what the real risks of aortic valve replacement are, Dr. Girardi said at Heart Valve Summit 2011.

Asymptomatic severe aortic stenosis has generally been considered a benign prognosis, and thus the American College of Cardiology/American Heart Association guidelines have not recommended aortic valve replacement (AVR) for patients with isolated asymptomatic severe stenosis in the absence of concomitant cardiac or aortic surgery. Data from three large trials, however, show that the probability of patients with severe aortic stenosis remaining symptom- or event-free at 2 years is 21%-67%.

Moreover, in a study of 740 patients with severe aortic stenosis, mean survival more than doubled from 3.9 years without AVR to 8.2 years with surgery among asymptomatic patients, and from 2.9 years to 6.5 years among symptomatic patients (Annals Thor. Surg. 2006;82:2116-22). While 10% of patients in the cohort opted not to have surgery, a full 50% were not even referred for surgery because their cardiologist deemed them too high risk, said Dr. Girardi, the O. Wayne Isom Professor of Cardiothoracic Surgery at Weill Cornell Medical College in New York.

Given that most centers have an operating mortality of 3% overall and 1% or less among low-risk patients, there is ever-increasing evidence that surgery is the right thing to do, he said.

Accurate risk assessment for AVR is problematic, however, because all of the risk stratification algorithms are flawed and overestimate mortality risk. The Society of Thoracic Surgeons (STS) PROM algorithm is the most accurate, but it has no way to report ejection fraction (EF) – a variable that would clearly stratify a patient with an EF of 20% from one with an EF of 60%, he said.

The logistic and additive EuroSCORES do not take into account cerebral vascular accident or presence of a previous stroke. Instead, they use the vague classification of "neurological dysfunction," which simply means neurologic activity affecting daily function.

Finally, the Ambler risk score, which was recently developed for valve surgery, takes into account only 13 variables, Dr. Girardi said at the meeting, cosponsored by the American Association for Thoracic Surgery and the American College of Cardiology Foundation.

Researchers at Cornell used the four different algorithms to calculate the risk of mortality among 638 patients requiring AVR, in whom actual mortality was 3.76%. The STS PROM score was the closest at 4.26%, while estimated mortality came in at a whopping 13.21% using the logistic EuroSCORE, 7.46% with the additive EuroSCORE, and 6.99% with the Ambler score, Dr. Girardi said. The ability of these algorithms to estimate mortality risk in high-risk patients – even those in the upper 10th percentile of risk – proved even worse, he added.

So what’s a surgeon to do?

The answer may be homegrown. "You have to look at your local and regional databases to see what surgeons are doing, because I think they’re certainly doing a lot better job than these scoring systems are suggesting they are," Dr. Girardi said.

New York State has one of the nation’s oldest databases – all New York cardiac surgery centers are required to participate in it, and the State Department of Health routinely audits it. The 40,000-patient database puts the mortality risk for AVR among all comers at 2.6%, increasing to 4.5% when AVR is combined with coronary artery bypass grafting (CABG). The outcomes are similar at 2.7% and 4.2%, respectively, based on data from the latest STS database, Dr. Girardi said.

Mortality risk estimates must also be coupled with complication rates, particularly given the increased risk of stroke with transcatheter aortic valve implantation. At Cornell, the risk of AVR and AVR/CABG was 1.1% in 2009 and 0.4% in 2010. During the same period, the stroke rate for these patients was 0.7%, acute renal failure 0.5%, and respiratory failure rate 5%, but that includes intubation for more than 72 hours, he pointed out.

Data should also be analyzed for the increasingly important population of octogenarians. Over the past 5 years, there was a 30% increase in the overall number of AVR patients in New York State but a 70% increase in the number of octogenarians undergoing AVR with or without CABG, Dr. Girardi observed. For 2009 and 2010, the mortality rate at Cornell for octogenarians – of whom 25% were older than 90 years – was 1.1% among all comers, with a stroke rate of 1% and a respiratory failure rate of 7%, largely because of increased patient frailty.

"High-risk cohorts clearly need even more analysis to figure out what we can do with them," he said. "But I think we’re going to have to get better at this as we compare our therapies with less invasive therapies coming along."

Dr. Girardi reported no relevant conflicts of interest.

Perception of Aortic Valve Surgery Risk May Be Exaggerated