User login
SAN FRANCISCO – It’s been barely half a year since U.S. cardiologists and cardiac surgeons first became able to routinely offer operable, high-risk patients with aortic stenosis the option of transcatheter valve replacement, yet in the first few months the transcatheter approach quickly rivaled open surgery.
But for the time being in U.S. practice, transcatheter aortic valve replacement (TAVR) remains boxed into the high-risk niche, along with the subgroup of patients who are not suitable for open surgery, king of a pair of relatively small hills.
And no matter how well TAVR performs in the current pair of trials that are comparing it with open surgical aortic valve replacement (SAVR) for intermediate-risk patients, it will remain relegated to niche status for years to come. That’s because roughly two-thirds of all operable patients with aortic stenosis who need valve replacement fall into the low-risk category, with a Society of Thoracic Surgeons (STS) risk score of less than 4%, which experts agree will remain SAVR’s exclusive territory for the foreseeable future.
The high-risk stratum of operable patients, which TAVR now dominates, constitutes about 10% of all patients who need a new aortic valve and can undergo open surgery, patients with an STS score greater than 8%. The intermediate-risk category – an STS score of 4%-8%, where TAVR now vies against open SAVR in two high-profile trials – makes up the final quarter of the operable-patient pie.
Within the high-risk and operable universe, TAVR’s rise has been meteoric, starting last October when the Food and Drug Administration gave Edwards, marketer of the SAPIEN valve system, approval for these patients. A small survey of operators from U.S. TAVR programs in March at the annual scientific session of the American College of Cardiology (ACC) revealed a uniform perception that by early 2013 a sizable majority of U.S. patients with severe aortic stenosis who are deemed operable and are at high surgical risk will wind up being treated by TAVR instead of SAVR. The cardiac surgeons who collaborate on TAVR seem to have fully conceded the advantages of TAVR for these patients.
"We generally go with TAVR. Most patients want it, and with the equivalence" in outcomes from the first PARTNER (Placement of Aortic Transcatheter Valves) cohort A (operable patients) trial (New Engl. J. Med. 2010:364:2187-98), "you usually go with the less invasive procedure," said Dr. Joseph E. Bavaria, professor of surgery and director of thoracic aortic surgery at the University of Pennsylvania in Philadelphia. The small number of high-risk patients who go to open surgery tend to be men, "because they do better with surgery, especially if their life expectancy is greater than 5-8 years," or patients with high stroke risk, Dr. Bavaria said in an interview at the annual meeting of the American College of Cardiology.
"Operable, high-risk patients get TAVR. I’m a surgeon saying that. I’ve already done the [PARTNER cohort A] trial, and I don’t want to do it again," said Dr. Michael Mack, a cardiothoracic surgeon at the Heart Hospital in Plano, Texas. "The results are the same [from TAVR and SAVR] at 30 days, 1 year, and 2 years, but boy do we beat up patients with open surgery. If it was my mom, she’d get TAVR," he said in an interview. "The tie goes to the less invasive treatment."
When high-risk, operable patients are seen by the heart team Dr. Mack works with, the only ones who go to SAVR are patients – generally men – who have a large aortic annulus and need a 29-mm-diameter valve, which is not available for the time being to U.S. TAVR patients; women with septal hypertrophy causing significant left ventricular outflow-tract obstruction; and the small percent of patients who opt for open surgery, usually because it’s the more established approach or because they fear a higher stroke risk from TAVR.
"The majority of high-risk, operable patients now go to TAVR; I think that’s pretty much true across the United States," said Dr. Jeffrey J. Popma, a cardiologist who does TAVR and is a professor of medicine at Harvard University in Boston.
Where TAVR stood in 2010 and 2011
How is TAVR performing? Performance can only be completely assessed months or years after the fact, so the impact that high-risk, operable U.S. patients received from TAVR’s use in routine practice in late 2012 and the first months of 2013 remains to be seen. The most recently treated patients now available for meaningful analysis in large numbers come from 2010 and 2011: new data reported at the ACC meeting from a U.S. program of continued TAVR access that began in late 2009 following the end of recruitment into the first PARTNER trial, and 1-year follow-up of nearly 14,000 patients who underwent TAVR in 2011 as part of routine practice in Germany and were entered into the country’s national TAVR registry.
Both databases had good news for high-risk or inoperable patients. TAVR outcomes in the first couple of years immediately following PARTNER in both the United States and Germany showed clinically meaningful improvements over the way TAVR performed during the first PARTNER trial. Less optimistic news for the low- and intermediate-risk patients who underwent TAVR in Germany, where wider device use is possible, was that in broad terms open surgery outperformed TAVR in these lower-risk patients, although TAVR’s defenders are quick to point out how tricky it is to make cross-treatment comparisons with registry data.
The nonrandomized, continued-access cohort that followed the first PARTNER trial at 22 U.S. centers, 3 sites in Canada, and 1 site in Germany included 1,017 inoperable or high-risk patients who had successful transfemoral TAVR between August 2009 and December 2011. When compared with the 415 inoperable or high-risk patients who underwent transfemoral TAVR in both cohorts of PARTNER, the more recently treated patients had a statistically significant decrease in all-cause 1-year mortality, from a 25% death rate in PARTNER to a 20% rate during the continued-access period, Dr. William F. Fearon reported at the meeting.
The continued-access patients also showed significant cuts in their rates of major vascular complications, which dropped from 15% of patients in PARTNER to 6% during continued access; and in rates of major bleeding complications, which fell from 15% in PARTNER to 7% during continued access. Strokes were also down during continued access, 5% compared with 7% during PARTNER, but this was not a statistically significant drop, reported Dr. Fearon, a cardiologist at Stanford (Calif.) University.
"Maybe these improvements are due to better patient selection, and maybe we have also gotten better at what we do," commented Dr. Popma.
The latest German experience
The 1-year German Aortic Valve Registry (GARY) results from 2011 show similar improvements compared with PARTNER. Among the 2,689 patients who underwent transvascular TAVR, the 1-year total mortality rate was 21%, and in the subset of patients with stroke data the combined rate of major and minor stroke was 5%.
But it was the way that transvascular TAVR (which includes both transfemoral and other vascular approaches but excludes the 1,181 patients who underwent transapical TAVR) stacked up against open surgical replacement that raised concern.
Of the 13,860 total patients entered into GARY during 2011, 9,985 underwent SAVR, with 6,523 of these patients undergoing an isolated procedure (the rest had valve replacement combined with coronary artery bypass). One-year mortality was 7% in patients who had isolated SAVR, dramatically below the 21% rate among the transvascular TAVR patients, Dr. Friedrich-Wilhelm Mohr reported at the meeting. The transapical TAVR patients had a 28% 1-year mortality rate.
To address the issue of between-treatment differences in patients’ underlying risk, Dr. Mohr presented two sets of analyses that stratified patients with two different risk-scoring systems, the EuroSCORE and the AKL (aortic valve surgery) score, also known as the German aortic valve score. The results showed how disparate the outcomes were when patients were subgrouped by their underlying risk. Among the patients who had lone SAVR, 80% fell into the lowest-risk level, with an AKL score of less than 3. Among the transvascular TAVR patients, only 17% had AKL scores below 3.
Among patients with AKL scores less than 3, those who underwent SAVR without bypass had a roughly 5% mortality rate after 1 year, compared with about a 15% mortality rate among the transvascular TAVR patients. Analyses in higher-risk patient subgroups showed that the survival gap between SAVR and TAVR patients progressively shrank, until in high-risk patients – as in those studied in PARTNER – the mortality rates were about the same in the SAVR and TAVR groups, said Dr. Mohr, professor and director of heart surgery at Leipzig University, Germany. He also reported similar findings when patients were stratified by their baseline EuroSCORE.
A big factor behind the worse survival among lower-risk TAVR patients is the problem of aortic regurgitation, Dr. Mohr said. Following valve replacement, 56% of the transvascular TAVR patients had grade 1 regurgitation, 7% had grade 2, and less than 1% had grade 3 leakage. These rates are way too high, he said. "Our major concern is the incidence of aortic regurgitation; more than half of the [TAVR] patients had some kind of regurgitation. Regurgitation matters whether it is mild or severe."
Despite his concern, TAVR use in Germany is accelerating. The total number of aortic valve replacements done in Germany jumped from just less than 14,000 in 2011 to more than 22,000 last year; TAVR cases more than doubled, from fewer than 4,000 in 2011 to nearly 10,000 in 2012, with a "move to intermediate-risk groups," said Dr. Mohr, who made it clear that he does not endorse this trend. "TAVR must be proven better than or equal to surgery before it’s used on a large scale," he said. The GARY results show "a clear difference [between TAVR and SAVR] in lower-risk patients. The GARY results do not support going into intermediate-risk patients."
Dr. Martin B. Leon agreed with Dr. Mohr that a definitive determination of whether TAVR is at least as good as SAVR in intermediate-risk patients must await results from the two major, ongoing multicenter trials testing this hypothesis, the PARTNER II trial, using the Edwards balloon-expandable SAPIEN XT valve system, and the SURTAVI (Safety and Efficacy Study of the Medtronic CoreValve System in the Treatment of Severe, Symptomatic Aortic Stenosis in Intermediate Risk Subjects Who Need Aortic Valve Replacement) trial. But Dr. Leon also cautioned that possible confounders might be distorting the GARY results, creating what he calls the "GARY Fallacy."
"The GARY Fallacy is the absurd notion that you can compare SAVR and TAVR in various risk strata without formal risk-adjustment methods to account for imbalances in baseline variables that are not captured in standard risk scores," he said in a talk at the meeting. "Aortic-stenosis patients were selected for TAVR based on their presumed increased risk, including many variables that are not represented in the risk algorithms, such as frailty, liver disease, porcelain aorta, a hostile chest, dementia, and severe chronic obstructive pulmonary disease. Therefore, the risk scores for TAVR patients underrepresent their true risk," Dr. Leon said.
But it’s also possible that TAVR needs more refinement before it completely catches up with SAVR, especially in patients who have the best outcomes from SAVR.
"I think SAVR outperforms TAVR in low-risk patients in the German registry because SAVR is a more mature procedure," said Dr. Raj R. Makkar, a TAVR operator and director of interventional cardiology at Cedars-Sinai Medical Center in Los Angeles. "Maybe this will change, as TAVR becomes safer with fewer valve leaks, strokes, and vascular complications."
Assessing patient risks
Experts also realize that the EuroSCORE, the AKL score, and the other risk-stratification tools now available have flaws when applied to TAVR patients. "Risk stratification for TAVR and SAVR is very problematic; the EuroSCORE has shown poor predictive value for mortality," Dr. Leon said. And while the AKL score was developed specifically for SAVR patients, its relevance to TAVR patients is suspect. In the first PARTNER trial, the multivariate predictors of mortality in the TAVR patients were "completely different" from the predictors in the SAVR patients, he noted.
"It’s tricky calculating scores," agreed Dr. Makkar. Both he and Dr. Mohr cited comorbidities such as pulmonary hypertension, cirrhosis, knee replacement producing impaired mobility, and porcelain aorta that each ratchet up a patient’s risk but have no effect whatsoever on a patient’s EuroSCORE or STS score.
The German cardiology and cardiac surgery societies recognize the limitations of current risk-scoring formulas and are developing a risk-stratification tool specifically designed for TAVR patients, Dr. Mohr said.
The way patients are assessed before, during, and after TAVR is receiving careful scrutiny from some investigators who reported their findings at the ACC meeting, with a particular focus on efforts to characterize and minimize aortic-valve regurgitation following TAVR.
One report, for example, reviewed 2,679 patients who underwent TAVR at any of 33 French centers and 1 in Monaco between January 2010 and October 2011, and were enrolled in the French Aortic National CoreValve and Edwards (FRANCE 2) Registry, established by the French cardiology and thoracic and cardiovascular surgery societies (N. Engl. J. Med. 2012;366:1705-15). FRANCE 2 includes nearly 1,900 patients who received the balloon-expandable Edwards SAPIEN valve, and nearly 900 treated with the self-expanding Medtronic CoreValve device.
Following TAVR, 60% of all patients in the registry had paravalvular aortic regurgitation: 45% with grade 1 regurgitation, 14% with grade 2, and 1% with grade 3 or 4. In a multivariate analysis, the self-expandable device was linked to twice the rate of higher-grade aortic regurgitation, grade 2 or higher, compared with the balloon-expandable valve; and TAVRs done via the femoral artery approach were also about twice as likely to result in higher-grade regurgitations compared with other catheterization routes, reported Dr. Eric Van Belle, a professor at the Cardiology Hospital in Lille, France.
Postprocedural paravalvular regurgitation of grade 2 or higher "was associated with a twofold increase in 1-year mortality, and was the strongest independent predictor of mortality," said Dr. Van Belle in his talk at the meeting. In addition, "annulus diameter and prosthesis diameter were major determinants of aortic regurgitation" in patients who received a balloon-expandable valve.
The FRANCE 2 results showed that postprocedural aortic regurgitation at grade 2 or higher "is a major issue and should be avoided, especially when there is no significant aortic regurgitation at baseline, or when a nonfemoral delivery approach is used." The link between nonfemoral delivery approaches and lower rates of aortic regurgitation suggests "good control of the depth of device delivery and improved catheter technology are key to reducing regurgitation rates," Dr. Van Belle said. "In addition, the prosthesis diameter relative to annulus diameter is key to preventing regurgitation with balloon-expandable devices. Prevention of aortic regurgitation is a major challenge for developing the next-generation device technology."
Minimizing aortic regurgitation
Two ways to cut aortic regurgitation rates following TAVR are to better match the valve to the annulus size, and when possible not finish a TAVR procedure until regurgitation has been minimized.
Two-dimensional transthoracic echocardiography had been the standard approach for annulus sizing as recently as 3 years ago, but it has been replaced with more accurate approaches, either three-dimensional transthoracic echo or CT. "Two-dimensional measurements are seriously limited due to variations and noncircular annular anatomy," Dr. Makkar said in a talk at the meeting. "CT provides the best overall assessment, because in addition to the cross-sectional measurement of the annulus it provides the best assessment of calcification," and three-dimensional transthoracic echo is now standard for intraprocedural assessments, he said.
The impact that a concerted effort to minimize aortic regurgitation can have on outcomes was examined in a single-center study reported by Dr. Jan-Malte Sinning, a cardiologist at University Hospital in Bonn, Germany. Last year, Dr. Sinning and his associates reported developing a quantitative measure of aortic regurgitation immediately following TAVR, the aortic regurgitation (AR) index, based on the difference between a patient’s diastolic blood pressure in the aorta and the left ventricular end-diastolic pressure (J. Am. Coll. Cardiol. 2012;59:1134-41). They then calculated the AR index immediately after TAVR in a prospective series of 167 patients, and set themselves the goal of immediately taking whatever steps were needed to bring the AR index above 25, a cutoff that seemed to correspond to no worse than mild regurgitation.
In their series, 62 patients underwent immediate post-TAVR corrective steps to reduce aortic regurgitation and bring their AR index above 25, most commonly additional balloon dilatation of the valve. The result was that while 16 patients had severe regurgitation before these steps, no patient had severe regurgitation following the corrective measures. The adjustments also cut the incidence of moderate regurgitations from 41 patients to 10, Dr. Sinning reported.
The consequence was that the 30-day stroke rate in the new cohort was 1%, compared with a 6% rate in a historical TAVR cohort at University Hospital in Bonn. The need for pacemaker implants was also cut in half by the intervention compared with the historical group, and 30-day mortality was 3% with these interventions, compared with 7% in the historical controls.
The results suggest that using the AR index as a trigger for taking corrective measures can help improve survival, and that post-TAVR dilatation can successfully decrease regurgitation without increasing patients’ stroke risk, Dr. Sinning concluded.
U.S. operators who perform TAVR say they take similar steps these days to deal with aortic regurgitation after TAVR. "We all understand that you don’t want patients to leave with a leak," said Dr. Popma. "We do everything we can to minimize leaks. When a patient leaves the lab with a moderate or severe leak, there wasn’t anything we could do about it."
Finding TAVR’s limits
Even as TAVR has become the go-to method for high-risk patients, operators have also tried to define the procedure’s outer limit, the point of disease severity when a patient is dying with aortic stenosis rather than because of aortic stenosis, and performing TAVR doesn’t make sense.
Dr. Makkar analyzed 369 inoperable patients who underwent TAVR in both the PARTNER I trial and the nonrandomized continued-access phase that followed. The patients fell into three groups: those who were inoperable because of technical reasons, such as a porcelain aorta or prior chest irradiation; patients who were inoperable because of comorbidities, such as frailty or severe lung disease; and those with both limitations. The 85 patients who were inoperable only because of a technical limitation had a 2-year mortality rate of 23% following TAVR; the other 284 patients who all had comorbidities had a 2-year mortality rate of about 43%.
Last year, Dr. Makkar reported that analysis of these 369 inoperable patients showed a 2-year mortality rate of 20% in the subgroup with an STS score of less than 5%, a mortality rate of 40% among those with a score of 5%-14.5% (but still significantly below the 60% 2-year mortality among similar inoperable patients managed by standard medical therapy only), and a 60% mortality rate among patients with a baseline STS score of 15% or more, an outcome that was no better than that of patients managed without TAVR (N. Engl. J. Med. 2012;366:1696-704).
In a multivariate analysis he recently ran on the data from these 369 patients, the likelihood of 2-year mortality rose by a statistically significant 3% for every 1% increase in the patient’s baseline STS risk score.
"It is important to remember that there are diminishing returns from TAVR in patients with more comorbidities, especially when their STS risk score is greater than 15%," said Dr. Makkar.
Dr. Sinning and Dr. Van Belle had no relevant disclosures. Dr. Mohr is a PARTNER investigator but had no other disclosures. Dr. Bavaria has been a speaker for Edwards and is a PARTNER investigator. Dr. Mack has received travel support from Edwards and is a PARTNER investigator. Dr. Popma has been a consultant to Boston Scientific, Abbott Vascular, and Covidien, and received research support from Boston Scientific, Abbott Vascular, Abiomed, Medtronic, and Cordis. Dr. Fearon has been a consultant to Heart Flow and received research support from St. Jude Medical. Dr. Leon has been a consultant to Symetis; has a major equity stake in Sadra, Claret, Valve Medical, and Apica; has received research support from Boston Scientific, Edwards, and Medtronic; and is a PARTNER investigator. Dr. Makkar has been a consultant to Cordis, Medtronic, Abbott, Entourage Medical, and Abiomed; has been a speaker for Lilly; has received research and travel support from Edwards; and is a PARTNER investigator.
On Twitter @mitchelzoler
SAN FRANCISCO – It’s been barely half a year since U.S. cardiologists and cardiac surgeons first became able to routinely offer operable, high-risk patients with aortic stenosis the option of transcatheter valve replacement, yet in the first few months the transcatheter approach quickly rivaled open surgery.
But for the time being in U.S. practice, transcatheter aortic valve replacement (TAVR) remains boxed into the high-risk niche, along with the subgroup of patients who are not suitable for open surgery, king of a pair of relatively small hills.
And no matter how well TAVR performs in the current pair of trials that are comparing it with open surgical aortic valve replacement (SAVR) for intermediate-risk patients, it will remain relegated to niche status for years to come. That’s because roughly two-thirds of all operable patients with aortic stenosis who need valve replacement fall into the low-risk category, with a Society of Thoracic Surgeons (STS) risk score of less than 4%, which experts agree will remain SAVR’s exclusive territory for the foreseeable future.
The high-risk stratum of operable patients, which TAVR now dominates, constitutes about 10% of all patients who need a new aortic valve and can undergo open surgery, patients with an STS score greater than 8%. The intermediate-risk category – an STS score of 4%-8%, where TAVR now vies against open SAVR in two high-profile trials – makes up the final quarter of the operable-patient pie.
Within the high-risk and operable universe, TAVR’s rise has been meteoric, starting last October when the Food and Drug Administration gave Edwards, marketer of the SAPIEN valve system, approval for these patients. A small survey of operators from U.S. TAVR programs in March at the annual scientific session of the American College of Cardiology (ACC) revealed a uniform perception that by early 2013 a sizable majority of U.S. patients with severe aortic stenosis who are deemed operable and are at high surgical risk will wind up being treated by TAVR instead of SAVR. The cardiac surgeons who collaborate on TAVR seem to have fully conceded the advantages of TAVR for these patients.
"We generally go with TAVR. Most patients want it, and with the equivalence" in outcomes from the first PARTNER (Placement of Aortic Transcatheter Valves) cohort A (operable patients) trial (New Engl. J. Med. 2010:364:2187-98), "you usually go with the less invasive procedure," said Dr. Joseph E. Bavaria, professor of surgery and director of thoracic aortic surgery at the University of Pennsylvania in Philadelphia. The small number of high-risk patients who go to open surgery tend to be men, "because they do better with surgery, especially if their life expectancy is greater than 5-8 years," or patients with high stroke risk, Dr. Bavaria said in an interview at the annual meeting of the American College of Cardiology.
"Operable, high-risk patients get TAVR. I’m a surgeon saying that. I’ve already done the [PARTNER cohort A] trial, and I don’t want to do it again," said Dr. Michael Mack, a cardiothoracic surgeon at the Heart Hospital in Plano, Texas. "The results are the same [from TAVR and SAVR] at 30 days, 1 year, and 2 years, but boy do we beat up patients with open surgery. If it was my mom, she’d get TAVR," he said in an interview. "The tie goes to the less invasive treatment."
When high-risk, operable patients are seen by the heart team Dr. Mack works with, the only ones who go to SAVR are patients – generally men – who have a large aortic annulus and need a 29-mm-diameter valve, which is not available for the time being to U.S. TAVR patients; women with septal hypertrophy causing significant left ventricular outflow-tract obstruction; and the small percent of patients who opt for open surgery, usually because it’s the more established approach or because they fear a higher stroke risk from TAVR.
"The majority of high-risk, operable patients now go to TAVR; I think that’s pretty much true across the United States," said Dr. Jeffrey J. Popma, a cardiologist who does TAVR and is a professor of medicine at Harvard University in Boston.
Where TAVR stood in 2010 and 2011
How is TAVR performing? Performance can only be completely assessed months or years after the fact, so the impact that high-risk, operable U.S. patients received from TAVR’s use in routine practice in late 2012 and the first months of 2013 remains to be seen. The most recently treated patients now available for meaningful analysis in large numbers come from 2010 and 2011: new data reported at the ACC meeting from a U.S. program of continued TAVR access that began in late 2009 following the end of recruitment into the first PARTNER trial, and 1-year follow-up of nearly 14,000 patients who underwent TAVR in 2011 as part of routine practice in Germany and were entered into the country’s national TAVR registry.
Both databases had good news for high-risk or inoperable patients. TAVR outcomes in the first couple of years immediately following PARTNER in both the United States and Germany showed clinically meaningful improvements over the way TAVR performed during the first PARTNER trial. Less optimistic news for the low- and intermediate-risk patients who underwent TAVR in Germany, where wider device use is possible, was that in broad terms open surgery outperformed TAVR in these lower-risk patients, although TAVR’s defenders are quick to point out how tricky it is to make cross-treatment comparisons with registry data.
The nonrandomized, continued-access cohort that followed the first PARTNER trial at 22 U.S. centers, 3 sites in Canada, and 1 site in Germany included 1,017 inoperable or high-risk patients who had successful transfemoral TAVR between August 2009 and December 2011. When compared with the 415 inoperable or high-risk patients who underwent transfemoral TAVR in both cohorts of PARTNER, the more recently treated patients had a statistically significant decrease in all-cause 1-year mortality, from a 25% death rate in PARTNER to a 20% rate during the continued-access period, Dr. William F. Fearon reported at the meeting.
The continued-access patients also showed significant cuts in their rates of major vascular complications, which dropped from 15% of patients in PARTNER to 6% during continued access; and in rates of major bleeding complications, which fell from 15% in PARTNER to 7% during continued access. Strokes were also down during continued access, 5% compared with 7% during PARTNER, but this was not a statistically significant drop, reported Dr. Fearon, a cardiologist at Stanford (Calif.) University.
"Maybe these improvements are due to better patient selection, and maybe we have also gotten better at what we do," commented Dr. Popma.
The latest German experience
The 1-year German Aortic Valve Registry (GARY) results from 2011 show similar improvements compared with PARTNER. Among the 2,689 patients who underwent transvascular TAVR, the 1-year total mortality rate was 21%, and in the subset of patients with stroke data the combined rate of major and minor stroke was 5%.
But it was the way that transvascular TAVR (which includes both transfemoral and other vascular approaches but excludes the 1,181 patients who underwent transapical TAVR) stacked up against open surgical replacement that raised concern.
Of the 13,860 total patients entered into GARY during 2011, 9,985 underwent SAVR, with 6,523 of these patients undergoing an isolated procedure (the rest had valve replacement combined with coronary artery bypass). One-year mortality was 7% in patients who had isolated SAVR, dramatically below the 21% rate among the transvascular TAVR patients, Dr. Friedrich-Wilhelm Mohr reported at the meeting. The transapical TAVR patients had a 28% 1-year mortality rate.
To address the issue of between-treatment differences in patients’ underlying risk, Dr. Mohr presented two sets of analyses that stratified patients with two different risk-scoring systems, the EuroSCORE and the AKL (aortic valve surgery) score, also known as the German aortic valve score. The results showed how disparate the outcomes were when patients were subgrouped by their underlying risk. Among the patients who had lone SAVR, 80% fell into the lowest-risk level, with an AKL score of less than 3. Among the transvascular TAVR patients, only 17% had AKL scores below 3.
Among patients with AKL scores less than 3, those who underwent SAVR without bypass had a roughly 5% mortality rate after 1 year, compared with about a 15% mortality rate among the transvascular TAVR patients. Analyses in higher-risk patient subgroups showed that the survival gap between SAVR and TAVR patients progressively shrank, until in high-risk patients – as in those studied in PARTNER – the mortality rates were about the same in the SAVR and TAVR groups, said Dr. Mohr, professor and director of heart surgery at Leipzig University, Germany. He also reported similar findings when patients were stratified by their baseline EuroSCORE.
A big factor behind the worse survival among lower-risk TAVR patients is the problem of aortic regurgitation, Dr. Mohr said. Following valve replacement, 56% of the transvascular TAVR patients had grade 1 regurgitation, 7% had grade 2, and less than 1% had grade 3 leakage. These rates are way too high, he said. "Our major concern is the incidence of aortic regurgitation; more than half of the [TAVR] patients had some kind of regurgitation. Regurgitation matters whether it is mild or severe."
Despite his concern, TAVR use in Germany is accelerating. The total number of aortic valve replacements done in Germany jumped from just less than 14,000 in 2011 to more than 22,000 last year; TAVR cases more than doubled, from fewer than 4,000 in 2011 to nearly 10,000 in 2012, with a "move to intermediate-risk groups," said Dr. Mohr, who made it clear that he does not endorse this trend. "TAVR must be proven better than or equal to surgery before it’s used on a large scale," he said. The GARY results show "a clear difference [between TAVR and SAVR] in lower-risk patients. The GARY results do not support going into intermediate-risk patients."
Dr. Martin B. Leon agreed with Dr. Mohr that a definitive determination of whether TAVR is at least as good as SAVR in intermediate-risk patients must await results from the two major, ongoing multicenter trials testing this hypothesis, the PARTNER II trial, using the Edwards balloon-expandable SAPIEN XT valve system, and the SURTAVI (Safety and Efficacy Study of the Medtronic CoreValve System in the Treatment of Severe, Symptomatic Aortic Stenosis in Intermediate Risk Subjects Who Need Aortic Valve Replacement) trial. But Dr. Leon also cautioned that possible confounders might be distorting the GARY results, creating what he calls the "GARY Fallacy."
"The GARY Fallacy is the absurd notion that you can compare SAVR and TAVR in various risk strata without formal risk-adjustment methods to account for imbalances in baseline variables that are not captured in standard risk scores," he said in a talk at the meeting. "Aortic-stenosis patients were selected for TAVR based on their presumed increased risk, including many variables that are not represented in the risk algorithms, such as frailty, liver disease, porcelain aorta, a hostile chest, dementia, and severe chronic obstructive pulmonary disease. Therefore, the risk scores for TAVR patients underrepresent their true risk," Dr. Leon said.
But it’s also possible that TAVR needs more refinement before it completely catches up with SAVR, especially in patients who have the best outcomes from SAVR.
"I think SAVR outperforms TAVR in low-risk patients in the German registry because SAVR is a more mature procedure," said Dr. Raj R. Makkar, a TAVR operator and director of interventional cardiology at Cedars-Sinai Medical Center in Los Angeles. "Maybe this will change, as TAVR becomes safer with fewer valve leaks, strokes, and vascular complications."
Assessing patient risks
Experts also realize that the EuroSCORE, the AKL score, and the other risk-stratification tools now available have flaws when applied to TAVR patients. "Risk stratification for TAVR and SAVR is very problematic; the EuroSCORE has shown poor predictive value for mortality," Dr. Leon said. And while the AKL score was developed specifically for SAVR patients, its relevance to TAVR patients is suspect. In the first PARTNER trial, the multivariate predictors of mortality in the TAVR patients were "completely different" from the predictors in the SAVR patients, he noted.
"It’s tricky calculating scores," agreed Dr. Makkar. Both he and Dr. Mohr cited comorbidities such as pulmonary hypertension, cirrhosis, knee replacement producing impaired mobility, and porcelain aorta that each ratchet up a patient’s risk but have no effect whatsoever on a patient’s EuroSCORE or STS score.
The German cardiology and cardiac surgery societies recognize the limitations of current risk-scoring formulas and are developing a risk-stratification tool specifically designed for TAVR patients, Dr. Mohr said.
The way patients are assessed before, during, and after TAVR is receiving careful scrutiny from some investigators who reported their findings at the ACC meeting, with a particular focus on efforts to characterize and minimize aortic-valve regurgitation following TAVR.
One report, for example, reviewed 2,679 patients who underwent TAVR at any of 33 French centers and 1 in Monaco between January 2010 and October 2011, and were enrolled in the French Aortic National CoreValve and Edwards (FRANCE 2) Registry, established by the French cardiology and thoracic and cardiovascular surgery societies (N. Engl. J. Med. 2012;366:1705-15). FRANCE 2 includes nearly 1,900 patients who received the balloon-expandable Edwards SAPIEN valve, and nearly 900 treated with the self-expanding Medtronic CoreValve device.
Following TAVR, 60% of all patients in the registry had paravalvular aortic regurgitation: 45% with grade 1 regurgitation, 14% with grade 2, and 1% with grade 3 or 4. In a multivariate analysis, the self-expandable device was linked to twice the rate of higher-grade aortic regurgitation, grade 2 or higher, compared with the balloon-expandable valve; and TAVRs done via the femoral artery approach were also about twice as likely to result in higher-grade regurgitations compared with other catheterization routes, reported Dr. Eric Van Belle, a professor at the Cardiology Hospital in Lille, France.
Postprocedural paravalvular regurgitation of grade 2 or higher "was associated with a twofold increase in 1-year mortality, and was the strongest independent predictor of mortality," said Dr. Van Belle in his talk at the meeting. In addition, "annulus diameter and prosthesis diameter were major determinants of aortic regurgitation" in patients who received a balloon-expandable valve.
The FRANCE 2 results showed that postprocedural aortic regurgitation at grade 2 or higher "is a major issue and should be avoided, especially when there is no significant aortic regurgitation at baseline, or when a nonfemoral delivery approach is used." The link between nonfemoral delivery approaches and lower rates of aortic regurgitation suggests "good control of the depth of device delivery and improved catheter technology are key to reducing regurgitation rates," Dr. Van Belle said. "In addition, the prosthesis diameter relative to annulus diameter is key to preventing regurgitation with balloon-expandable devices. Prevention of aortic regurgitation is a major challenge for developing the next-generation device technology."
Minimizing aortic regurgitation
Two ways to cut aortic regurgitation rates following TAVR are to better match the valve to the annulus size, and when possible not finish a TAVR procedure until regurgitation has been minimized.
Two-dimensional transthoracic echocardiography had been the standard approach for annulus sizing as recently as 3 years ago, but it has been replaced with more accurate approaches, either three-dimensional transthoracic echo or CT. "Two-dimensional measurements are seriously limited due to variations and noncircular annular anatomy," Dr. Makkar said in a talk at the meeting. "CT provides the best overall assessment, because in addition to the cross-sectional measurement of the annulus it provides the best assessment of calcification," and three-dimensional transthoracic echo is now standard for intraprocedural assessments, he said.
The impact that a concerted effort to minimize aortic regurgitation can have on outcomes was examined in a single-center study reported by Dr. Jan-Malte Sinning, a cardiologist at University Hospital in Bonn, Germany. Last year, Dr. Sinning and his associates reported developing a quantitative measure of aortic regurgitation immediately following TAVR, the aortic regurgitation (AR) index, based on the difference between a patient’s diastolic blood pressure in the aorta and the left ventricular end-diastolic pressure (J. Am. Coll. Cardiol. 2012;59:1134-41). They then calculated the AR index immediately after TAVR in a prospective series of 167 patients, and set themselves the goal of immediately taking whatever steps were needed to bring the AR index above 25, a cutoff that seemed to correspond to no worse than mild regurgitation.
In their series, 62 patients underwent immediate post-TAVR corrective steps to reduce aortic regurgitation and bring their AR index above 25, most commonly additional balloon dilatation of the valve. The result was that while 16 patients had severe regurgitation before these steps, no patient had severe regurgitation following the corrective measures. The adjustments also cut the incidence of moderate regurgitations from 41 patients to 10, Dr. Sinning reported.
The consequence was that the 30-day stroke rate in the new cohort was 1%, compared with a 6% rate in a historical TAVR cohort at University Hospital in Bonn. The need for pacemaker implants was also cut in half by the intervention compared with the historical group, and 30-day mortality was 3% with these interventions, compared with 7% in the historical controls.
The results suggest that using the AR index as a trigger for taking corrective measures can help improve survival, and that post-TAVR dilatation can successfully decrease regurgitation without increasing patients’ stroke risk, Dr. Sinning concluded.
U.S. operators who perform TAVR say they take similar steps these days to deal with aortic regurgitation after TAVR. "We all understand that you don’t want patients to leave with a leak," said Dr. Popma. "We do everything we can to minimize leaks. When a patient leaves the lab with a moderate or severe leak, there wasn’t anything we could do about it."
Finding TAVR’s limits
Even as TAVR has become the go-to method for high-risk patients, operators have also tried to define the procedure’s outer limit, the point of disease severity when a patient is dying with aortic stenosis rather than because of aortic stenosis, and performing TAVR doesn’t make sense.
Dr. Makkar analyzed 369 inoperable patients who underwent TAVR in both the PARTNER I trial and the nonrandomized continued-access phase that followed. The patients fell into three groups: those who were inoperable because of technical reasons, such as a porcelain aorta or prior chest irradiation; patients who were inoperable because of comorbidities, such as frailty or severe lung disease; and those with both limitations. The 85 patients who were inoperable only because of a technical limitation had a 2-year mortality rate of 23% following TAVR; the other 284 patients who all had comorbidities had a 2-year mortality rate of about 43%.
Last year, Dr. Makkar reported that analysis of these 369 inoperable patients showed a 2-year mortality rate of 20% in the subgroup with an STS score of less than 5%, a mortality rate of 40% among those with a score of 5%-14.5% (but still significantly below the 60% 2-year mortality among similar inoperable patients managed by standard medical therapy only), and a 60% mortality rate among patients with a baseline STS score of 15% or more, an outcome that was no better than that of patients managed without TAVR (N. Engl. J. Med. 2012;366:1696-704).
In a multivariate analysis he recently ran on the data from these 369 patients, the likelihood of 2-year mortality rose by a statistically significant 3% for every 1% increase in the patient’s baseline STS risk score.
"It is important to remember that there are diminishing returns from TAVR in patients with more comorbidities, especially when their STS risk score is greater than 15%," said Dr. Makkar.
Dr. Sinning and Dr. Van Belle had no relevant disclosures. Dr. Mohr is a PARTNER investigator but had no other disclosures. Dr. Bavaria has been a speaker for Edwards and is a PARTNER investigator. Dr. Mack has received travel support from Edwards and is a PARTNER investigator. Dr. Popma has been a consultant to Boston Scientific, Abbott Vascular, and Covidien, and received research support from Boston Scientific, Abbott Vascular, Abiomed, Medtronic, and Cordis. Dr. Fearon has been a consultant to Heart Flow and received research support from St. Jude Medical. Dr. Leon has been a consultant to Symetis; has a major equity stake in Sadra, Claret, Valve Medical, and Apica; has received research support from Boston Scientific, Edwards, and Medtronic; and is a PARTNER investigator. Dr. Makkar has been a consultant to Cordis, Medtronic, Abbott, Entourage Medical, and Abiomed; has been a speaker for Lilly; has received research and travel support from Edwards; and is a PARTNER investigator.
On Twitter @mitchelzoler
SAN FRANCISCO – It’s been barely half a year since U.S. cardiologists and cardiac surgeons first became able to routinely offer operable, high-risk patients with aortic stenosis the option of transcatheter valve replacement, yet in the first few months the transcatheter approach quickly rivaled open surgery.
But for the time being in U.S. practice, transcatheter aortic valve replacement (TAVR) remains boxed into the high-risk niche, along with the subgroup of patients who are not suitable for open surgery, king of a pair of relatively small hills.
And no matter how well TAVR performs in the current pair of trials that are comparing it with open surgical aortic valve replacement (SAVR) for intermediate-risk patients, it will remain relegated to niche status for years to come. That’s because roughly two-thirds of all operable patients with aortic stenosis who need valve replacement fall into the low-risk category, with a Society of Thoracic Surgeons (STS) risk score of less than 4%, which experts agree will remain SAVR’s exclusive territory for the foreseeable future.
The high-risk stratum of operable patients, which TAVR now dominates, constitutes about 10% of all patients who need a new aortic valve and can undergo open surgery, patients with an STS score greater than 8%. The intermediate-risk category – an STS score of 4%-8%, where TAVR now vies against open SAVR in two high-profile trials – makes up the final quarter of the operable-patient pie.
Within the high-risk and operable universe, TAVR’s rise has been meteoric, starting last October when the Food and Drug Administration gave Edwards, marketer of the SAPIEN valve system, approval for these patients. A small survey of operators from U.S. TAVR programs in March at the annual scientific session of the American College of Cardiology (ACC) revealed a uniform perception that by early 2013 a sizable majority of U.S. patients with severe aortic stenosis who are deemed operable and are at high surgical risk will wind up being treated by TAVR instead of SAVR. The cardiac surgeons who collaborate on TAVR seem to have fully conceded the advantages of TAVR for these patients.
"We generally go with TAVR. Most patients want it, and with the equivalence" in outcomes from the first PARTNER (Placement of Aortic Transcatheter Valves) cohort A (operable patients) trial (New Engl. J. Med. 2010:364:2187-98), "you usually go with the less invasive procedure," said Dr. Joseph E. Bavaria, professor of surgery and director of thoracic aortic surgery at the University of Pennsylvania in Philadelphia. The small number of high-risk patients who go to open surgery tend to be men, "because they do better with surgery, especially if their life expectancy is greater than 5-8 years," or patients with high stroke risk, Dr. Bavaria said in an interview at the annual meeting of the American College of Cardiology.
"Operable, high-risk patients get TAVR. I’m a surgeon saying that. I’ve already done the [PARTNER cohort A] trial, and I don’t want to do it again," said Dr. Michael Mack, a cardiothoracic surgeon at the Heart Hospital in Plano, Texas. "The results are the same [from TAVR and SAVR] at 30 days, 1 year, and 2 years, but boy do we beat up patients with open surgery. If it was my mom, she’d get TAVR," he said in an interview. "The tie goes to the less invasive treatment."
When high-risk, operable patients are seen by the heart team Dr. Mack works with, the only ones who go to SAVR are patients – generally men – who have a large aortic annulus and need a 29-mm-diameter valve, which is not available for the time being to U.S. TAVR patients; women with septal hypertrophy causing significant left ventricular outflow-tract obstruction; and the small percent of patients who opt for open surgery, usually because it’s the more established approach or because they fear a higher stroke risk from TAVR.
"The majority of high-risk, operable patients now go to TAVR; I think that’s pretty much true across the United States," said Dr. Jeffrey J. Popma, a cardiologist who does TAVR and is a professor of medicine at Harvard University in Boston.
Where TAVR stood in 2010 and 2011
How is TAVR performing? Performance can only be completely assessed months or years after the fact, so the impact that high-risk, operable U.S. patients received from TAVR’s use in routine practice in late 2012 and the first months of 2013 remains to be seen. The most recently treated patients now available for meaningful analysis in large numbers come from 2010 and 2011: new data reported at the ACC meeting from a U.S. program of continued TAVR access that began in late 2009 following the end of recruitment into the first PARTNER trial, and 1-year follow-up of nearly 14,000 patients who underwent TAVR in 2011 as part of routine practice in Germany and were entered into the country’s national TAVR registry.
Both databases had good news for high-risk or inoperable patients. TAVR outcomes in the first couple of years immediately following PARTNER in both the United States and Germany showed clinically meaningful improvements over the way TAVR performed during the first PARTNER trial. Less optimistic news for the low- and intermediate-risk patients who underwent TAVR in Germany, where wider device use is possible, was that in broad terms open surgery outperformed TAVR in these lower-risk patients, although TAVR’s defenders are quick to point out how tricky it is to make cross-treatment comparisons with registry data.
The nonrandomized, continued-access cohort that followed the first PARTNER trial at 22 U.S. centers, 3 sites in Canada, and 1 site in Germany included 1,017 inoperable or high-risk patients who had successful transfemoral TAVR between August 2009 and December 2011. When compared with the 415 inoperable or high-risk patients who underwent transfemoral TAVR in both cohorts of PARTNER, the more recently treated patients had a statistically significant decrease in all-cause 1-year mortality, from a 25% death rate in PARTNER to a 20% rate during the continued-access period, Dr. William F. Fearon reported at the meeting.
The continued-access patients also showed significant cuts in their rates of major vascular complications, which dropped from 15% of patients in PARTNER to 6% during continued access; and in rates of major bleeding complications, which fell from 15% in PARTNER to 7% during continued access. Strokes were also down during continued access, 5% compared with 7% during PARTNER, but this was not a statistically significant drop, reported Dr. Fearon, a cardiologist at Stanford (Calif.) University.
"Maybe these improvements are due to better patient selection, and maybe we have also gotten better at what we do," commented Dr. Popma.
The latest German experience
The 1-year German Aortic Valve Registry (GARY) results from 2011 show similar improvements compared with PARTNER. Among the 2,689 patients who underwent transvascular TAVR, the 1-year total mortality rate was 21%, and in the subset of patients with stroke data the combined rate of major and minor stroke was 5%.
But it was the way that transvascular TAVR (which includes both transfemoral and other vascular approaches but excludes the 1,181 patients who underwent transapical TAVR) stacked up against open surgical replacement that raised concern.
Of the 13,860 total patients entered into GARY during 2011, 9,985 underwent SAVR, with 6,523 of these patients undergoing an isolated procedure (the rest had valve replacement combined with coronary artery bypass). One-year mortality was 7% in patients who had isolated SAVR, dramatically below the 21% rate among the transvascular TAVR patients, Dr. Friedrich-Wilhelm Mohr reported at the meeting. The transapical TAVR patients had a 28% 1-year mortality rate.
To address the issue of between-treatment differences in patients’ underlying risk, Dr. Mohr presented two sets of analyses that stratified patients with two different risk-scoring systems, the EuroSCORE and the AKL (aortic valve surgery) score, also known as the German aortic valve score. The results showed how disparate the outcomes were when patients were subgrouped by their underlying risk. Among the patients who had lone SAVR, 80% fell into the lowest-risk level, with an AKL score of less than 3. Among the transvascular TAVR patients, only 17% had AKL scores below 3.
Among patients with AKL scores less than 3, those who underwent SAVR without bypass had a roughly 5% mortality rate after 1 year, compared with about a 15% mortality rate among the transvascular TAVR patients. Analyses in higher-risk patient subgroups showed that the survival gap between SAVR and TAVR patients progressively shrank, until in high-risk patients – as in those studied in PARTNER – the mortality rates were about the same in the SAVR and TAVR groups, said Dr. Mohr, professor and director of heart surgery at Leipzig University, Germany. He also reported similar findings when patients were stratified by their baseline EuroSCORE.
A big factor behind the worse survival among lower-risk TAVR patients is the problem of aortic regurgitation, Dr. Mohr said. Following valve replacement, 56% of the transvascular TAVR patients had grade 1 regurgitation, 7% had grade 2, and less than 1% had grade 3 leakage. These rates are way too high, he said. "Our major concern is the incidence of aortic regurgitation; more than half of the [TAVR] patients had some kind of regurgitation. Regurgitation matters whether it is mild or severe."
Despite his concern, TAVR use in Germany is accelerating. The total number of aortic valve replacements done in Germany jumped from just less than 14,000 in 2011 to more than 22,000 last year; TAVR cases more than doubled, from fewer than 4,000 in 2011 to nearly 10,000 in 2012, with a "move to intermediate-risk groups," said Dr. Mohr, who made it clear that he does not endorse this trend. "TAVR must be proven better than or equal to surgery before it’s used on a large scale," he said. The GARY results show "a clear difference [between TAVR and SAVR] in lower-risk patients. The GARY results do not support going into intermediate-risk patients."
Dr. Martin B. Leon agreed with Dr. Mohr that a definitive determination of whether TAVR is at least as good as SAVR in intermediate-risk patients must await results from the two major, ongoing multicenter trials testing this hypothesis, the PARTNER II trial, using the Edwards balloon-expandable SAPIEN XT valve system, and the SURTAVI (Safety and Efficacy Study of the Medtronic CoreValve System in the Treatment of Severe, Symptomatic Aortic Stenosis in Intermediate Risk Subjects Who Need Aortic Valve Replacement) trial. But Dr. Leon also cautioned that possible confounders might be distorting the GARY results, creating what he calls the "GARY Fallacy."
"The GARY Fallacy is the absurd notion that you can compare SAVR and TAVR in various risk strata without formal risk-adjustment methods to account for imbalances in baseline variables that are not captured in standard risk scores," he said in a talk at the meeting. "Aortic-stenosis patients were selected for TAVR based on their presumed increased risk, including many variables that are not represented in the risk algorithms, such as frailty, liver disease, porcelain aorta, a hostile chest, dementia, and severe chronic obstructive pulmonary disease. Therefore, the risk scores for TAVR patients underrepresent their true risk," Dr. Leon said.
But it’s also possible that TAVR needs more refinement before it completely catches up with SAVR, especially in patients who have the best outcomes from SAVR.
"I think SAVR outperforms TAVR in low-risk patients in the German registry because SAVR is a more mature procedure," said Dr. Raj R. Makkar, a TAVR operator and director of interventional cardiology at Cedars-Sinai Medical Center in Los Angeles. "Maybe this will change, as TAVR becomes safer with fewer valve leaks, strokes, and vascular complications."
Assessing patient risks
Experts also realize that the EuroSCORE, the AKL score, and the other risk-stratification tools now available have flaws when applied to TAVR patients. "Risk stratification for TAVR and SAVR is very problematic; the EuroSCORE has shown poor predictive value for mortality," Dr. Leon said. And while the AKL score was developed specifically for SAVR patients, its relevance to TAVR patients is suspect. In the first PARTNER trial, the multivariate predictors of mortality in the TAVR patients were "completely different" from the predictors in the SAVR patients, he noted.
"It’s tricky calculating scores," agreed Dr. Makkar. Both he and Dr. Mohr cited comorbidities such as pulmonary hypertension, cirrhosis, knee replacement producing impaired mobility, and porcelain aorta that each ratchet up a patient’s risk but have no effect whatsoever on a patient’s EuroSCORE or STS score.
The German cardiology and cardiac surgery societies recognize the limitations of current risk-scoring formulas and are developing a risk-stratification tool specifically designed for TAVR patients, Dr. Mohr said.
The way patients are assessed before, during, and after TAVR is receiving careful scrutiny from some investigators who reported their findings at the ACC meeting, with a particular focus on efforts to characterize and minimize aortic-valve regurgitation following TAVR.
One report, for example, reviewed 2,679 patients who underwent TAVR at any of 33 French centers and 1 in Monaco between January 2010 and October 2011, and were enrolled in the French Aortic National CoreValve and Edwards (FRANCE 2) Registry, established by the French cardiology and thoracic and cardiovascular surgery societies (N. Engl. J. Med. 2012;366:1705-15). FRANCE 2 includes nearly 1,900 patients who received the balloon-expandable Edwards SAPIEN valve, and nearly 900 treated with the self-expanding Medtronic CoreValve device.
Following TAVR, 60% of all patients in the registry had paravalvular aortic regurgitation: 45% with grade 1 regurgitation, 14% with grade 2, and 1% with grade 3 or 4. In a multivariate analysis, the self-expandable device was linked to twice the rate of higher-grade aortic regurgitation, grade 2 or higher, compared with the balloon-expandable valve; and TAVRs done via the femoral artery approach were also about twice as likely to result in higher-grade regurgitations compared with other catheterization routes, reported Dr. Eric Van Belle, a professor at the Cardiology Hospital in Lille, France.
Postprocedural paravalvular regurgitation of grade 2 or higher "was associated with a twofold increase in 1-year mortality, and was the strongest independent predictor of mortality," said Dr. Van Belle in his talk at the meeting. In addition, "annulus diameter and prosthesis diameter were major determinants of aortic regurgitation" in patients who received a balloon-expandable valve.
The FRANCE 2 results showed that postprocedural aortic regurgitation at grade 2 or higher "is a major issue and should be avoided, especially when there is no significant aortic regurgitation at baseline, or when a nonfemoral delivery approach is used." The link between nonfemoral delivery approaches and lower rates of aortic regurgitation suggests "good control of the depth of device delivery and improved catheter technology are key to reducing regurgitation rates," Dr. Van Belle said. "In addition, the prosthesis diameter relative to annulus diameter is key to preventing regurgitation with balloon-expandable devices. Prevention of aortic regurgitation is a major challenge for developing the next-generation device technology."
Minimizing aortic regurgitation
Two ways to cut aortic regurgitation rates following TAVR are to better match the valve to the annulus size, and when possible not finish a TAVR procedure until regurgitation has been minimized.
Two-dimensional transthoracic echocardiography had been the standard approach for annulus sizing as recently as 3 years ago, but it has been replaced with more accurate approaches, either three-dimensional transthoracic echo or CT. "Two-dimensional measurements are seriously limited due to variations and noncircular annular anatomy," Dr. Makkar said in a talk at the meeting. "CT provides the best overall assessment, because in addition to the cross-sectional measurement of the annulus it provides the best assessment of calcification," and three-dimensional transthoracic echo is now standard for intraprocedural assessments, he said.
The impact that a concerted effort to minimize aortic regurgitation can have on outcomes was examined in a single-center study reported by Dr. Jan-Malte Sinning, a cardiologist at University Hospital in Bonn, Germany. Last year, Dr. Sinning and his associates reported developing a quantitative measure of aortic regurgitation immediately following TAVR, the aortic regurgitation (AR) index, based on the difference between a patient’s diastolic blood pressure in the aorta and the left ventricular end-diastolic pressure (J. Am. Coll. Cardiol. 2012;59:1134-41). They then calculated the AR index immediately after TAVR in a prospective series of 167 patients, and set themselves the goal of immediately taking whatever steps were needed to bring the AR index above 25, a cutoff that seemed to correspond to no worse than mild regurgitation.
In their series, 62 patients underwent immediate post-TAVR corrective steps to reduce aortic regurgitation and bring their AR index above 25, most commonly additional balloon dilatation of the valve. The result was that while 16 patients had severe regurgitation before these steps, no patient had severe regurgitation following the corrective measures. The adjustments also cut the incidence of moderate regurgitations from 41 patients to 10, Dr. Sinning reported.
The consequence was that the 30-day stroke rate in the new cohort was 1%, compared with a 6% rate in a historical TAVR cohort at University Hospital in Bonn. The need for pacemaker implants was also cut in half by the intervention compared with the historical group, and 30-day mortality was 3% with these interventions, compared with 7% in the historical controls.
The results suggest that using the AR index as a trigger for taking corrective measures can help improve survival, and that post-TAVR dilatation can successfully decrease regurgitation without increasing patients’ stroke risk, Dr. Sinning concluded.
U.S. operators who perform TAVR say they take similar steps these days to deal with aortic regurgitation after TAVR. "We all understand that you don’t want patients to leave with a leak," said Dr. Popma. "We do everything we can to minimize leaks. When a patient leaves the lab with a moderate or severe leak, there wasn’t anything we could do about it."
Finding TAVR’s limits
Even as TAVR has become the go-to method for high-risk patients, operators have also tried to define the procedure’s outer limit, the point of disease severity when a patient is dying with aortic stenosis rather than because of aortic stenosis, and performing TAVR doesn’t make sense.
Dr. Makkar analyzed 369 inoperable patients who underwent TAVR in both the PARTNER I trial and the nonrandomized continued-access phase that followed. The patients fell into three groups: those who were inoperable because of technical reasons, such as a porcelain aorta or prior chest irradiation; patients who were inoperable because of comorbidities, such as frailty or severe lung disease; and those with both limitations. The 85 patients who were inoperable only because of a technical limitation had a 2-year mortality rate of 23% following TAVR; the other 284 patients who all had comorbidities had a 2-year mortality rate of about 43%.
Last year, Dr. Makkar reported that analysis of these 369 inoperable patients showed a 2-year mortality rate of 20% in the subgroup with an STS score of less than 5%, a mortality rate of 40% among those with a score of 5%-14.5% (but still significantly below the 60% 2-year mortality among similar inoperable patients managed by standard medical therapy only), and a 60% mortality rate among patients with a baseline STS score of 15% or more, an outcome that was no better than that of patients managed without TAVR (N. Engl. J. Med. 2012;366:1696-704).
In a multivariate analysis he recently ran on the data from these 369 patients, the likelihood of 2-year mortality rose by a statistically significant 3% for every 1% increase in the patient’s baseline STS risk score.
"It is important to remember that there are diminishing returns from TAVR in patients with more comorbidities, especially when their STS risk score is greater than 15%," said Dr. Makkar.
Dr. Sinning and Dr. Van Belle had no relevant disclosures. Dr. Mohr is a PARTNER investigator but had no other disclosures. Dr. Bavaria has been a speaker for Edwards and is a PARTNER investigator. Dr. Mack has received travel support from Edwards and is a PARTNER investigator. Dr. Popma has been a consultant to Boston Scientific, Abbott Vascular, and Covidien, and received research support from Boston Scientific, Abbott Vascular, Abiomed, Medtronic, and Cordis. Dr. Fearon has been a consultant to Heart Flow and received research support from St. Jude Medical. Dr. Leon has been a consultant to Symetis; has a major equity stake in Sadra, Claret, Valve Medical, and Apica; has received research support from Boston Scientific, Edwards, and Medtronic; and is a PARTNER investigator. Dr. Makkar has been a consultant to Cordis, Medtronic, Abbott, Entourage Medical, and Abiomed; has been a speaker for Lilly; has received research and travel support from Edwards; and is a PARTNER investigator.
On Twitter @mitchelzoler
AT ACC 13
