PARTNER at 5 Years: TAVR Shows Durable Benefit in Inoperable Patients
For inoperable patients with severe aortic stenosis, transcatheter aortic valve replacement (TAVR) provides sustained improvements in survival and function compared with standard treatment, according to 5-year follow-up from Cohort B of the PARTNER I trial published online March 15, 2015, ahead of print in the Lancet. Nonetheless, overall mortality is high in these older and sicker patients regardless of therapy.
The long-term findings were originally presented at the 2014 Transcatheter Cardiovascular Therapeutics conference in Washington, DC.
For PARTNER I, investigators randomized 358 patients (mean age 83 years; 54% female) with severe aortic stenosis who were deemed inoperable by a heart team to undergo TAVR with the balloon-expandable first-generation Sapien valve (Edwards Lifesciences; n = 179) or receive standard care (n = 179). TAVR was performed under general anesthesia via transfemoral access at 21 experienced international centers between May 11, 2007, and March 16, 2009.
Mean Society of Thoracic Surgeons Predicted Risk of Mortality (STS PROM) score was 11.7% overall, and 79% of those in the standard care group underwent valvuloplasty during the trial. In the main PARTNER findings, at 1 year TAVR reduced all-cause death compared with standard treatment (HR 0.55; 95% CI 0.40-0.74).
For the current analysis, investigators led by Samir R. Kapadia, MD, of the Cleveland Clinic (Cleveland, OH), looked at 5-year outcomes. Though mortality was high in both groups, it was markedly reduced following TAVR compared with standard treatment (HR 0.50; 95% CI 0.39-0.65). At 5 years, only 1 patient among the 6 in the standard treatment group who remained alive had not received a form of valve replacement, and that individual had undergone valvuloplasty during follow-up. Median survival was 31.0 months in the TAVR group compared with 11.7 months in the standard treatment group (P < .0001). Landmark analysis showed that the survival advantage of TAVR was maintained between years 3 and 5 (P = .028).
Cardiovascular Mortality Reduced the Most
The benefit of TAVR over standard care was greater for cardiovascular mortality (HR 0.41; 95% CI 0.31-0.55) than for all-cause mortality. However, 34% of deaths in the TAVR group and 17% in the standard care group were judged to be noncardiovascular, suggesting that morbidities unrelated to cardiovascular disease heavily contributed to death. Stroke rates were similar between the TAVR and standard treatment groups, and analysis of the competing risks of mortality and stroke showed no ongoing hazard of stroke with TAVR after the initial procedural risk. The risk of repeat hospitalization was lower with TAVR, while the proportions of patients in NYHA class 3 or 4 were similar (table 1).
Valve area and mean transvalvular gradient remained stable throughout follow-up, with the former measuring 1.52 cm2 and the latter 10.6 mm Hg at 5 years. Moderate or severe paravalvular leak was present in 14% of patients at first measurement after TAVR. None had structural valve deterioration requiring reintervention, and 1 patient underwent valve replacement for endocarditis after the index procedure.
STS PROM Scores Separate Patients
Interestingly, TAVR patients with an STS PROM score of at least 5% had higher all-cause mortality than those with lower scores, but there was no mortality difference between patients with scores of 5%-14.9% and those with higher scores. At 5 years, for patients with a score of less than 5% or of 5%-14.9%, mortality was lower in the TAVR group than the standard treatment group (P = .0012 and P = .0002, respectively). However, among patients with an STS PROM score of at least 15%, there was only a trend toward lower mortality in the TAVR group (P = .0749).
The mortality curves of the TAVR and standard treatment groups separated immediately in patients with an STS PROM score of less than 5%, at around 1 year in patients with a score of 5%-14.9%, and at around 2 years in those with a score higher than 15%. In contrast, cardiovascular mortality was lower with TAVR than with standard treatment across all score ranges.
At 5 years, although there was no difference in all-cause mortality according to the degree of paravalvular leak, cardiovascular mortality was higher in patients with moderate to severe paravalvular leak compared with no or mild leak (75% vs 51%; P = .043).
Multiple subgroups reaped a survival benefit with TAVR compared with controls; the only exception was patients with oxygen-dependent chronic obstructive pulmonary disease (COPD).
In addition, on multivariate analysis, predictors of mortality after TAVR were:
- BMI ≥ 26 kg/m2: OR 0.50 (95% CI 0.34-0.73)
- Oxygen-dependent COPD: OR 1.83 (95% CI 1.22-2.75)
- Peripheral vascular disease: OR 1.53 (95% CI 1.04-2.24)
Learning Patient Selection
“It is reassuring that, after 5 years, the [bioprosthetic] valve is still functioning with no deterioration,” Philippe Généreux, MD, of Columbia University Medical Center (New York, NY), told TCTMD in a telephone interview. “That being said, from a surgical point of view we need longer-term follow-up to ensure its integrity.”
The study authors observe that “[a]lthough the clinical outcomes are encouraging, better patient selection and reduction in procedural complications can help to make TAVR even more beneficial.”
Noting the adverse impact of comorbidities, Dr. Kapadia and colleagues say it is important to try to identify TAVR candidates who are likely to survive the high attrition of the postprocedural period. They point out that after 2 years the annual mortality rate of TAVR patients resembles that of an age- and sex-matched US population without stenosis or comorbidities.
Dr. Généreux echoed that theme. “We have learned from PARTNER B that we need to select patients very carefully” and not offer TAVR to the very sickest, he said. “The STS score can be used to risk stratify these patients, but most of the time there are other factors that influence our judgment.” For example, he noted, patients who are inoperable for anatomic reasons but have an STS score below 5 make good TAVR candidates, while those with higher STS scores and severe comorbidities do not.
TAVR Effective—But Listen to Patients
In an accompanying editorial, Arie P. Kappetein, MD, PhD, of Erasmus University Medical Center (Rotterdam, the Netherlands), warns that “[t]he more comorbidities, the greater the risk of overtreatment.”
Moreover, he observes, “Many older patients and their families have different ideas about what makes life worth sustaining.” In fact, quality of life may be even more important than mortality in this patient population, he asserts, and its assessment deserves further research.
For inoperable patients, Dr. Kappetein writes, “exploring their goals is necessary before pursuing an invasive treatment with its inherent complications, and a preoperative discussion between the heart team, patient, and other decision makers is an excellent way to ensure that care meets the patient’s wishes.”
“The main message of PARTNER B is that TAVR is the preferred treatment for inoperable patients,” Dr. Généreux concluded. “But there is a subgroup that is so sick that we should temper our [instinct] to treat them.”
Note: Several study coauthors are faculty members of the Cardiovascular Research Foundation, which owns and operates TCTMD.
1. Kapadia SR, Leon MB, Makkar RR, et al. 5-year outcomes of transcatheter aortic valve replacement compared with standard treatment for patients with inoperable aortic stenosis (PARTNER 1): a randomized controlled trial. Lancet. 2015;Epub ahead of print.
2. Kappetein AP. PARTNERs in the future of surgical aortic valve replacement [comment]. Lancet. 2015;Epub ahead of print.
- The PARTNER I trial was funded by Edwards Lifesciences.
- Drs. Kapadia and Kappetein report no relevant conflicts of interest.
- Dr. Généreux reports receiving speaker’s fees from Edwards Lifesciences.