DISCUSSION
This single-center propensity-matched study of over 1,500 isolated bioprosthetic surgical aortic valve replacements demonstrates comparable short-term and long-term outcomes of bovine and porcine prostheses. There were no significant differences in postoperative clinical outcomes, echocardiographic data, readmission rates, reintervention rates, or long-term survival between the propensity-matched groups.
Notable differences between bovine and porcine valves have been hitherto established. Importantly, we know that the mode of structural valve deterioration (SVD) differs between these two valve types; porcine valves tend to fail via leaflet tears, while bovine pericardial valves tend to degenerate via fibrosis/calcification.8Interestingly, despite these respective modes of SVD, previous prospective and/or randomized studies have demonstrated higher transprosthetic gradients for porcine valves, with smaller indexed effective orifice areas (despite larger implanted valve sizes) when compared to bovine prostheses.9,10,11 Bovine valves have been found to have more favorable hemodynamics in general across the majority of these studies. Our study, on the other hand, found comparable gradients between the two xenograft materials, though this finding may not have persisted at 5 or 10 years.
How bovine and porcine valves compare with one another regarding short-term and long-term clinical outcomes remains unclear. In an effort to answer these questions, several comparative studies have been performed to-date, though most have included concomitant operations and have not accounted for differences in baseline characteristics with propensity matching. Moreover, data from these studies have yielded conflicting results.
One such study, involving roughly 13,000 patients in Sweden, found a long-term survival benefit with porcine prostheses, though with a higher need for reoperation when compared to bovine prostheses.6 A study of almost 40,000 patients in England and Wales demonstrated no difference in long-term survival or need for reintervention among patients who received bovine or porcine aortic valves, as did another study from Duke.4,12Importantly, all of these studies included patients who received concomitant CABG, and none of them utilized propensity matching to account for baseline differences. A recent meta-analysis of studies published from 2010-2015 also demonstrated no difference in long-term survival in patients who underwent AVR with a bovine or porcine prosthesis, though these studies also included patients who underwent concomitant operations and did not match on baseline characteristics.13 Our study, which included only isolated AVRs and which incorporated propensity matching to adjust for confounders, may help to clarify and adjudicate prior findings.
As valve-in-valve TAVR becomes more popular, suitability for this subsequent intervention, should it be necessary, also becomes an important consideration when choosing a surgical prosthesis. Prior studies have not found specific differences in feasibility or outcomes of valve-in-valve TAVR according to use of bovine or porcine surgical valves, though it has been demonstrated that residual stenosis following valve-in-valve TAVR is more common with surgical valves that have an inner diameter of < 20 mm.14 Thus, it is not so much the type of surgical valve, but rather the size of the implanted valve, that seems to carry the most importance.
Another important question that has been investigated is the comparison between stented and stentless bioprosthetic aortic valves.15,16,17,18 A randomized controlled trial was performed to compare clinical outcomes and hemodynamic performance between these two valve types. In comparison to stented prostheses, this trial found stentless valves to have greater effective orifice areas, as well as greater improvement in postoperative left ventricular function when used in patients with small annuli or left ventricular dysfunction.19 A recent meta-analysis has adjudicated this finding of superior short-term hemodynamic outcomes with stentless valves in patients who have small annuli.20
Outcomes of valve-in-valve TAVR following stented versus stentless prostheses have also been investigated. In a study using the Valve-in-Valve International Data registry, stentless valve-in-valve TAVR was associated with more periprocedural complications such as device malposition, coronary obstruction, and paravalvular leak.21 Thus, while long-term outcomes of stented versus stentless AVR appear to be comparable, and stentless valves confer better hemodynamics in patients with small annuli, these valves may also complicate the performance of future valve-in-valve TAVR.
This study is inherently limited by its retrospective, observational design. Moreover, multiple valve types were included, and the performance of a particular valve model could have impacted outcomes. We did not include data on indexed effective orifice area; however, we matched on implanted aortic valve size to adjust for this potential confounder. Serial, longitudinal echocardiographic data at set time-points was not available but would have been very useful in evaluating changes in hemodynamic performance of these two valves over time. The study also incorporated longitudinal data with varying follow-up times, with some patients being lost to follow-up. Finally, the data is from a single high-volume center, which may limit generalizability of the findings.
The use of either bovine or porcine bioprosthetic aortic valves yields comparable postoperative outcomes, long-term survival, freedom from reintervention, and freedom from readmission.