Background: Primary tissue failure, which is mainly caused by calcification
, is still the limiting factor in the long-term outcome of heart valve biop
rostheses. Even though the precise nature of this process is not fully unde
rstood, in vitro tests have been developed to reproduce and predict calcifi
cation for individual bioprostheses. Methods: In vitro calcification testin
g was performed by using an accelerated pulsatile valve tester which was ad
apted for testing stented as well as stentless bioprostheses with physiolog
ical fluid dynamics. A total of 84 bioprostheses (porcine, pericardial and
stentless porcine of different manufacturers) were cyclically loaded at a t
est rate of 300/min at 37 degreesC within a rapid calcification fluid with
CaxP=130 (mg/d1)(2) at pH 7.4. Calcification was assessed by microradiograp
hy after 12x10(6) cycles. In a previous step, holographic interferometry wa
s performed to identify irregularities of valve leaflets in order to predic
t later calcification. Selected specimens of calcified bioprostheses underw
ent histology, transmission (TEM) and scanning (SEM) electron microscopy. T
issue mineralization was investigated by coupling SEM, electron microprobe
analysis (EM PA) and X-ray powder diffraction (XRPD) methods. Results: For
all tested bioprostheses, a significant calcification was achieved within 4
to 6 weeks of ongoing testing, and the degree of calcification increased w
ith time. A significant correlation between calcification and leaflet irreg
ularities (detected by holographic interferometry) was found (r=0.80, p=0.0
01). Calcification varied between individual bioprostheses, and significant
differences were detected for different groups (calculated as percentage o
f total leaflet area, mean+/-SD): porcine stented (37.3+/-12.0%), bovine st
ented (23.0+/-8.9 %), porcine stentless (16.2+/-7.6%). Histological and ult
rastructural investigation showed intrinsic calcification involving both th
e spongiosa and fibrosa with collagen fibrils, interfibrillar spaces and ce
lls as early sites of calcification. There was clear evidence of apatite cr
ystallization, and observations made with in vitro calcification were quite
similar to those occurring with in vivo implanted bioprostheses. Conclusio
n: In vitro tests can reproduce intrinsic calcification of bioprostheses ev
en in the absence of viable biologic host factors. Moreover, degree and sit
es of calcification have become predictable. This enables the development a
nd evaluation of bioprostheses with reduction of animal experiments. From o
ur results obtained with a broad range of available bioprostheses, stented
bovine and stentless porcine valves seem to be superior to conventional ste
nted porcine bioprostheses with regard to leaflet calcification.