Steffen et al Acquired: Aortic Valve

ACQ
Resistance to infection of long-term cryopreserved human aortic
valve allografts
Viola Steffen, MD,a,b Georg Marsch, MD,a Karin Burgwitz,a,b Christian Kuehn, MD, PhD,a,b and
Omke E. Teebken, MD, PhDa

ABSTRACT
Objective: To analyze the in vitro antimicrobial activity of 3 antibiotic regimens
(group A, gentamicin-piperacillin-vancomycin-metronidazole-amphotericin B;
group B, gentamicin-piperacillin-flucloxacillin-metronidazole-amphotericin B;
and group C, meropenem-vancomycin-tobramycin-colistin-amphotericin B)
used in the processing of cryopreserved human ascending aortic tissue and aortic
valves against Staphylococcus epidermidis and Staphylococcus aureus. The re-
sults were additionally compared with the infection resistance of cryopreserved
ascending aortic tissue against Escherichia coli and Pseudomonas aeruginosa.
Elongated Escherichia coli on cryopreserved aortic wall
Materials: Each of 10 cryopreserved human allografts (CHAs) was divided into allograft.
25 pieces (separating aortic wall and valve). Eighteen segments were microbiolog-
ically tested, and 7 pieces underwent scanning electron microscopy. A bacterial Central Message
solution (4 mL; optical density, 0.20
0.02) was used for contamination. After The infection resistance of CHAs depends on antibi-
incubation, the optical density of the solution was measured. CHAs underwent otic pretreatment and is superior in the aortic wall
compared with valve tissue.
sonication to release viable adherent bacteria. The number of attached bacteria
was quantified by the colony forming units per square centimeter of CHA surface.
Perspective
Results: Antibiotic regimen groups B and C were more efficient than group A in Cryopreserved human allografts (CHAs) can be rec-
ommended to patients with destructive infections or
eradicating gram-positive organisms adherent to the aortic wall (P<.001). Group high risk of (re)infection. Our in vitro study offers
C showed enhanced resistance against E coli compared with group A or B novel insight into the principles of antibacterial activ-
(P <.001), whereas group B appeared to be more effective against P aeruginosa ity of CHAs. Enhanced retention of antibiotic agents
(P < .001). With reference to each antibiotic regimen, ascending aortic tissue may improve the infection resistance of CHAs. The
risk of reinfection can be reduced by adapting the
showed significantly less bacterial contamination with staphylococcal bacteria CHA decontamination protocol to the spectrum of
than valve grafts (P  .01). infection-causing bacteria.

Conclusions: CHAs possess antibacterial activity despite long-term storage over

See Article page 1239.
5 years. Antibiotic combinations applied during CHA processing have a
significant influence on their infection resistance. Ascending aortic tissue shows
a significantly enhanced bacterial resistance against staphylococcal bacteria See Editorial Commentary page 1249.
compared with aortic valves. (J Thorac Cardiovasc Surg 2016;151:1251-9)
See Editorial page 1230.

Prosthetic graft infections are responsible for major The incidence of prosthetic valve endocarditis ranges
morbidity and drastically increased mortality rates from 0.3% to 1.2% per patient year, with a cumulative
(20%-80%) and therefore remain among the most risk of 5% at 10 years.1,2 Furthermore, 0.9% to 1.9% of
challenging complications in cardiothoracic surgery.1-3 patients develop a prosthetic graft infection following
surgery on the thoracic aorta.3
From the aDepartment of Cardiothoracic, Transplantation, and Vascular Surgery,
Most graft infections are caused by gram-positive bacte-
and bCrossBIT Research Center for Biocompatibility and Immunology of Medical ria, namely Staphylococcus aureus and Staphylococcus
Implants, Hannover Medical School, Hannover, Germany. epidermidis.2,4,5 The increased incidence of highly
Funding sources: Internal funds.
Received for publication July 12, 2015; revisions received Oct 18, 2015; accepted for
publication Nov 18, 2015; available ahead of print Dec 23, 2015.
Address for reprints: Omke E. Teebken, MD, PhD, Department of Cardiothoracic, Scanning this QR code will take
Transplantation, and Vascular Surgery, Hannover Medical School, Carl-Neuberg- you to the article title page.
Str 1, OE 6210, 30625 Hannover, Germany (E-mail: Teebken.Omke@
MH-Hannover.de).
0022-5223/$36.00
Copyright Ó 2016 by The American Association for Thoracic Surgery
http://dx.doi.org/10.1016/j.jtcvs.2015.11.029

The Journal of Thoracic and Cardiovascular Surgery c Volume 151, Number 5 1251
 Acquired: Aortic Valve Steffen et al
ACQ

MATERIALS AND METHODS

Abbreviations and Acronyms Bacterial Strains
CHA ¼ cryopreserved human allograft The strains were 20044DSMZ for S epidermidis, 20231DSMZ for
S aureus, 1103DSMZ for E coli, and 0411208MDPA01 for P aeruginosa.
OD600nm ¼ optical density Isolates were subcultured overnight in tryptic soy broth (Oxoid, Hamp-
PBS ¼ phosphate-buffered saline shire, United Kingdom). The bacterial concentration measured by optical
SEM ¼ scanning electron microscopy density (OD600nm), as used for graft contamination, amounted to
0.20
0.02. This OD600nm corresponded to 6 3 107 CFU/mL for S epider-
midis, 8 3 107 CFU/mL for S aureus, 3 3 106 CFU/mL for E coli, and
virulent and antibiotic-resistant microorganisms such 3 3 106 CFU/mL for P aeruginosa.
as Pseudomonas aeruginosa and methicillin-resistant
Staphylococcus aureus (MRSA) additionally hinders a CHAs
successful eradication of infection, and consequently, Ten aortic valve allografts, including a segment of the ascending aorta,
were obtained from the Deutsche Gesellschaft f€ur Gewebetransplantation
investigations into this problem are currently ongoing.2,6
or its predecessor DSO-G. These allografts were rejected for transplantation
Various therapies for prosthetic graft infections involving because their storage time had exceeded 5 years. CHAs were subjected to the
the aortic valve and the ascending aorta have been standard procedure of processing, consisting of harvesting from multiorgan
described; for example, adaptive antibiotic treatment, donors, preparation, antibiotic treatment, cryopreservation, and thawing as
antiseptic irrigation, and perhaps most importantly, surgical described previously.6 The CHAs were assigned to three groups according
to their antibiotic treatment (Table 1): group A (gentamicin, piperacillin,
debridement.2,7 Because extra-anatomical rerouting is not
vancomycin, metronidazole, and amphotericin B), group B (gentamicin, pi-
possible at the site of the aortic root and the arch, surgical peracillin, flucloxacillin, metronidazole, and amphotericin B), and group C
options involve either salvaging the existing graft by (meropenem, vancomycin, tobramycin, colistin, and amphotericin B).
aggressive debridement, irrigation, and coverage with
healthy tissue8 or graft removal and renewed in situ Experimental Protocol
replacement with synthetic grafts or biological tissue; for Each CHA was divided into 25 pieces under sterile conditions as
example, cryopreserved human allografts (CHAs).7,9-12 illustrated in Figure 1. The valves were dissected into 7 samples (0.75 cm2
CHAs represent effective alternatives for the treatment of or approximately 0.4 cm2 for controls). The wall specimens (1 cm2) were
infections, with extremely low reinfection rates.13 Native contaminated with both gram-positive and gram-negative microorganisms,
whereas the cusps were incubated with the bacteria most frequently respon-
and prosthetic valve endocarditis,12,14,15 as well as sible for prosthetic valve endocarditis: S epidermidis and S aureus.4 In detail,
vascular prosthetic infections,6 show promising outcomes the piece of CHA (aortic wall or cusp) was placed into 4 mL subcultured bac-
after CHA replacement. Furthermore, CHAs appear to be terial suspension (OD600nm, 0.20
0.02) and incubated at 37 C and 5% car-
effective even in complex destructive infections involving bon dioxide for 24 hours (n ¼ 21). Two specimens of the aortic wall and 2 of
the aortic root or ascending aorta.5,16 CHAs are not the valve were incubated in sterile tryptic soy broth and served as controls.
recommended for routine aortic valve or ascending aortic Microbiologic tests. Following incubation, all grafts were washed 3
times in 50 mL phosphate-buffered saline (PBS) (Dulbecco; Biochrom,
surgery, but can be considered for patients with Berlin, Germany) to eliminate unattached bacteria. The OD600nm of the
destructive infections or high risk of (re)infection.17 bacterial suspension was measured to evaluate bacterial growth in the sur-
The reason why CHAs exhibit this clinically observable roundings of the specimens. The samples were transferred into sterile Fal-
bacterial resistance remains unclear, but may be related to con tubes containing 5 mL PBS, and the remaining viable adherent bacteria
their viability, which allows an improved transfer of were released by sonication at low power (37 C, 100%) for 20 minutes.20
100 mL of this PBS-solution containing the dislodged bacteria was plated
antibiotic agents and immunocompetent cells through the onto tryptic soy broth-agar plates. All plates were incubated (37 C; 5%
wall into the perigraft space.18 Moreover, their antimicro- carbon dioxide), and colonies were visually identified and counted with a
bial activity might be attributed to processing and, in Molecular Imager (Gel-Doc XR; Bio-Rad Laboratories, Munich, Ger-
particular, to the storage of CHAs in antibiotics.19 However, many). The results were used to determine the number of attached bacteria
this topic remains poorly investigated to date. per square centimeter of CHA surface.
The aim of this in vitro study was to evaluate the antimi- Scanning electron microscopy. After incubation, each sample
was washed three times in PBS to remove unattached microorganisms.
crobial activity of CHAs, taking a closer look at 3 clinically Samples were subsequently fixed in a 2.5% glutaraldehyde and
approved antibiotic regimens used in the processing of sodium-cacodylate buffer (0.1 M; pH 7.3), after which they were passed
human ascending aortic tissue and aortic valves with respect through an increasing concentration of acetone before dehydration in a
to S epidermidis and S aureus, as gram-positive pathogens. critical point dryer (CPD-030; Bal-Tec GmbH, Balzers, Liechtenstein),
sputtered with gold, and observed by scanning electron microscopy
Additionally, we compared these data with the infection
(SEM) (Quanta-400F; FEI Company, Hillsboro, Ore), at 20 kV and a reso-
resistance of cryopreserved ascending aortic tissue lution of 1 nm.
against gram-negative pathogens; for example, E coli and
P aeruginosa. We hypothesized that the choice of Statistical Analysis
antibiotics has significant influence on infection resistance, All data are expressed as mean
standard error of the mean. The
and that valve and aortic wall tissue behave in a different Mann-Whitney test was chosen to compare the different groups. SPSS
manner. version 21.0 (IBM-SPSS Inc, Armonk, NY) and GraphPad Prism version

1252 The Journal of Thoracic and Cardiovascular Surgery c May 2016

Steffen et al Acquired: Aortic Valve

ACQ
TABLE 1. Characteristics of the cryopreserved human allografts of group A. This was higher (P<.001) than the number of
(CHAs) bacteria found on the specimens of group B
CHA antibiotic Donor Donor Donor Duration of (1.3 3 101
9.4 3 100 CFU/cm2) or on the tissue of group
group* age (y) sex AB0-type cryopreservation (y) C (8.8 3 101
4.4 3 101 CFU/cm2). The OD600nm of the
A 18 F B 7.42 supernatant showed comparable results (Table 2).
18 F AB 8.19 After incubation with S aureus, we found 9.4 3 102

36 F AB 8.79 1.6 3 102 CFU/cm2 on the tissue of group A. This was
B 62 F 0 9.09 significantly (P<.001) more than that found on the samples
12 M AB 9.17 of group B (1.6 3 102
4.1 3 101 CFU/cm2) or in group C
40 F 0 9.64
(8.3 3 100
5.3 3 100 CFU/cm2). No significant
14 F 0 9.81
difference was apparent between groups B and C
30 M 0 10.00
C 29 M AB 6.49
(P ¼ .083). Again, the OD600nm produced corresponding
12 M A 6.61 results (Table 2). Both gram-positive bacteria showed
*Group A includes gentamicin, piperacillin, vancomycin, metronidazole, and ampho- significantly increased colony forming units/centimeters2
tericin B. Group B includes gentamicin, piperacillin, flucloxacillin, metronidazole, and OD600nm for the aortic wall grafts stored in the
and amphotericin B. Group C includes meropenem, vancomycin, tobramycin, antibiotics of group A.
colistin, and amphotericin B.
Aortic valve, S epidermidis, and S aureus. As shown in
5.0 for Windows (GraphPad Software Inc., San Diego, Calif) software
Figure 2, we found 2.3 3 105
4.3 3 103 CFU/cm2
packages were used. The study was approved by the local ethics committee attached to the cusps of group A after incubation
(reference No. 1008/2011). with S epidermidis. Grafts of group B exhibited 2.6 3 105

5.7 3 103 CFU/cm2, whereas 1.6 3 105

RESULTS 7.4 3 103 CFU/cm2 were detected on the samples of group
The average storage time of CHAs was 8.52 years (range, C. The number of adherent bacteria in group C was signi-
6.49-10.0 years) (Table 1). Macroscopic examination ficantly smaller compared with group A (P ¼ .010) or B
revealed no apparent structural deficiencies. (P ¼ .002). After incubation with S aureus we found a
smaller amount of bacteria (8.2 3 104
2.2 3 104 CFU/
Microbiologic Tests cm2) attached to the grafts of group B in comparison with
Aortic wall, S epidermidis, and S aureus. As shown in group A (2.0 3 105
1.4 3 104 CFU/cm2; P <.001). No
Figures 2 and 3, after incubation with S epidermidis, significant difference was apparent between groups B and
1.0 3 105
2.4 3 104 CFU/cm2 were found on the samples C (1.1 3 105
9.0 3 103 CFU/cm2; P ¼ .945). Overall,

FIGURE 1. Partition of 1 cryopreserved human allograft for microbiologic experiments and tissue examination. S. epid, Staphylococcus epidermis; S. aur,
Staphylococcus aureus; E. coli, Escherichia coli; P. aeru, Pseudomonas aeruginosa; SEM, scanning electron microscopy.

The Journal of Thoracic and Cardiovascular Surgery c Volume 151, Number 5 1253
 Acquired: Aortic Valve Steffen et al
ACQ

FIGURE 2. Bacteria (colony forming units/centimeter2) adherent to aortic wall or valve tissue after antibiotic treatment according to protocol A through C
and after incubation with A, Staphylococcus epidermidis (S. epidermis) or B, Staphylococcus aureus (S. aureus).

the difference between the aortic wall and valve was the OD600nm of the supernatants from group A was lower
reflected in the significantly increased amounts of bacteria than that from the grafts of groups B and C (Table 2). In
adherent to the valve for S epidermidis and for S aureus both investigations, group B exhibited low bacterial growth
within each antibiotic group (both P values  .01). This and a small number of adherent bacteria.
difference was also detected in the OD600nm (both P
values < .01). SEM
Aortic wall, E coli, and P aeruginosa. As shown in Representative pictures of CHAs impregnated with
Figure 3, after incubation with E coli, we found a smaller antibiotic agents according to protocol A through C and af-
amount of bacteria (1.2 3 103
5.6 3 102 CFU/cm2) on ter bacterial incubation are illustrated in Figures 4 and 5.
the samples of group C in comparison with group A Concerning the controls of groups A through C, aortic ves-
(5.3 3 103
3.3 3 102 CFU/cm2) or group B sels were partially covered with cells that were arranged ac-
(9.6 3 103
9.2 3 102 CFU/cm2; P <.001, respectively). cording to blood flow. An intact endothelium was not
The OD600nm showed corresponding results. P aeruginosa detectable because wide areas presented irregularities in
was most abundant on the grafts of group A the morphology of cells, shrinkage with disruption of
(4.4 3 104
3.2 3 103 CFU/cm2) and showed a significant intercellular contacts, and observable stripping of the
difference from the amount found in group B endothelium, exposure of the basal lamina, and partially
(8.8 3 103
1.2 3 103 CFU/cm2; P <.001) or group C damaged subendothelial layers (Figure 4, A). On the aortic
(2.4 3 104
2.5 3 103 CFU/cm2; P ¼ .001). In contrast, valve, the spectrum of changes ranged from a smooth

1254 The Journal of Thoracic and Cardiovascular Surgery c May 2016

Steffen et al Acquired: Aortic Valve

ACQ
FIGURE 3. Bacteria (colony forming units/centimeters2) adherent to aortic wall tissue after antibiotic treatment according to protocol A through C and
after bacterial incubation. S. epidermis, Staphylococcus epidermis; S. aureus, Staphylococcus aureus; E. coli, Escherichia coli; P. aeruginosa, Pseudomonas
aeruginosa.

surface consisting of an exposed basal lamina to stripped DISCUSSION

endothelial cells with patches of damaged subendothelial CHAs are well-known for their excellent hemodynamic
layers. The continuity of endothelial coverage was lost on characteristics, their decreased thromboembolic rate while
cusps (Figure 5, A). avoiding the need for anticoagulation, and their infection
Concerning gram-positive microorganisms on the aortic resistance.13,14,21 In this study, the infection resistance of
wall specimens, only a few adherent bacteria were CHAs previously treated with various antibiotic solutions
observed, reflecting the results of the OD600nm and colony was investigated.
forming units/centimeters2 measurements (Figure 4, B Tissue banks rely on antibiotic agents during CHA
and C). Noticeably, and in concordance with the microbio- processing, but their protocols differ significantly.22 The
logic examinations, extensive adherence of S epidermidis main objective is the decontamination of tissue to prevent
was seen in group A. bacterial transmission to allograft recipients without
In contrast, the valves were completely colonized with compromising tissue integrity.22 Previous studies have
both staphylococcal bacteria, as seen in Figure 5, B indicated that residual concentrations of these antibiotics
and C. A comparison of the three antibiotic treatments are still present and effective in CHAs after preparation
nonetheless revealed differences in the density of bacterial for implantation.23,24 Jashari and colleagues24 investigated
contamination. residual antibiotic concentrations in cryopreserved heart
With regard to the gram-negative microorganisms, we valves previously treated with vancomycin, lincomycin,
observed a characteristic network composed of the and polymyxin B. Only residues of vancomycin and
microorganisms and partially altered bacterial growth. lincomycin were detected in the allografts, but the concen-
The rod-shaped E coli of 2 mm in size were considerably trations were largely below their therapeutic doses.24
elongated with a filamentous morphology, stretched Nevertheless, the authors did not measure the toxic effect
by > 10 mm (Figure 4, D). on microorganisms. Furthermore, Buzzi and colleagues23

TABLE 2. Optical density in the supernatant of cryopreserved human allograft after antibiotic treatment according to protocol A through C and
after bacterial incubation
P value P value P value
Antibiotic group A A-B B B-C C A-C
Aortic wall
Staphylococcus epidermidis 1.54
0.10 < .001 0.07
0.01 < .001 0.13
0.01 < .001
Staphylococcus aureus 0.31
0.03 .030 0.20
0.02 .203 0.14
0.02 .001
Escherichia coli 1.79
0.02 .689 1.83
0.03 .001 1.46
0.09 .007
Pseudomonas aeruginosa 0.24
0.02 .639 0.25
0.02 < .001 0.54
0.03 < .001
Valve tissue
Staphylococcus epidermidis 2.07
0.03 .022 1.75
0.08 .240 1.56
0.10 .010
Staphylococcus aureus 1.52
0.17 .042 1.02
0.13 .002 1.62
0.03 1.000

The Journal of Thoracic and Cardiovascular Surgery c Volume 151, Number 5 1255
 Acquired: Aortic Valve Steffen et al
ACQ

FIGURE 4. A, Cryopreserved aortic wall tissue after antibiotic treatment according to protocol A through C and after incubation with A, control,
B, Staphylococcus epidermis (S. epidermidis), C, Staphylococcus aureus (S. aureus), D, Escherichia coli (E. coli), or E, Pseudomonas aeroginosa
(P. aeruginosa).

demonstrated the effectiveness of residual antibiotics in First, we found an important difference between the
cryopreserved human aortic heart valves and blood vessels infection resistance of aortic wall and valve tissue. Irrespec-
by microbiologic examination, namely an agar diffusion tive of the antibiotic treatment applied during processing,
test. After being thawed, CHA tissue revealed important significantly more staphylococcal bacteria were adherent
inhibition zones on plates seeded with S aureus and to the cusps in comparison with the aortic wall (Figure 2).
P aeruginosa.23 Nevertheless, the period of cryopreserva- This difference was also reflected in the OD600nm and in
tion remained unclear and the authors did not differentiate representative SEM images (Figures 4 and 5). The findings
between heart valve and vessel tissue. This study confirmed suggest that the retention of the antibiotics is insufficient to
indirectly the presence of residual antibiotics in cryopre- prevent bacterial colonization of cusps, whereas the
served tissue. Analyses of residual concentrations of remaining antibiotics in the aortic wall specimens offer
antibiotics in long-term CHAs are still lacking, but several increased infection resistance. We used the same inoculum
studies have shown that antibiotics are stable and potent for aortic wall samples and for the thinner samples of the
after long-term storage at 80 C for 180 days and valve. Therefore, the latter is likely to contain fewer
longer.25,26 antibiotics and thus exhibits limited regional antibiotic
Our results support the hypothesis that infection release.
resistance of CHAs depends on the antibiotic pretreatment Second, regarding the aortic wall specimens, we found a
during processing and their residual activity. significantly increased adherence of S epidermidis in group

1256 The Journal of Thoracic and Cardiovascular Surgery c May 2016

Steffen et al Acquired: Aortic Valve

ACQ
FIGURE 5. Cryopreserved aortic valve tissue after antibiotic treatment according to protocol A through C in A, control and after incubation with
B, Staphylococcus epidermis (S. epidermidis) or C, Staphylococcus aureus (S. aureus).

A compared with group B or group C. The antibiotics of particularly enhanced antimicrobial activity. In our study,
group A only differed from that of group B with respect the aortic vessels stored in the antibiotics in groups B and
to the use of vancomycin instead of flucloxacillin. K€onig C show significantly enhanced antimicrobial activity; that
and colleagues27 showed that high inocula of S epidermidis is, a reduced number of colony forming units against
(1 3 107 CFU/mL) only moderately affect the activity of staphylococcal bacteria. CHAs of group B appear to be
flucloxacillin, whereas a compromising effect on glycopep- more resistant to P aeruginosa, whereas group C showed
tides has been observed. Consequently, we presume that enhanced resistance against E coli (Figure 3). These
CHAs stored in antibiotics containing flucloxacillin instead significant differences in the quantity of adherent bacteria
of vancomycin (as a glycopeptide) have an enhanced within the same bacterial strain can predominantly be
antibacterial effect against the employed high inocula of explained on the basis of the residual activity of the
S epidermidis (6 3 107 CFU/mL). According to K€onig different antibiotic mixtures.
and colleagues,27 vancomycin has a decreased effect on Finally, we observed morphologic changes in the
high inocula of S epidermidis and thus results in an adhering microorganisms, especially of the rod-shaped E coli, which
biofilm, which entails a further loss in the antibiotic activity were considerably elongated. Justice and colleagues28
of vancomycin compared with flucloxacillin. Group C also described this filamentation as a response to environmental
contained vancomycin, but the amount of S epidermidis was stresses, particularly metabolic changes or antimicrobial
significantly smaller compared with the CHAs of group A. therapies, which improves bacterial survival. The inhibition
This may be related to the different antibiotic compositions. of septation during cell growth results in this elongation.
Third, our microbiologic data, apart from the peculiarity Several studies have attributed the filamentation to ß-lactam
of S epidermidis in group A, reveal a significant antibiotics (eg, piperacillin and meropenem) that bind to
antimicrobial activity of cryopreserved aortic vessels, penicillin-binding protein-3 (by inhibition of penicillin-
with a tendency toward reduced infection resistance against binding protein-3), particularly at low concentrations.29,30
P aeruginosa. Similar findings have been reported by The regional antibiotic release alters bacterial growth
Camiade and colleagues,19 who investigated the infection without eliminating all adherent bacteria. This could be
resistance of cryopreserved thoracic aortic allografts attributed to the high inocula of microorganisms, a
previously stored in gentamicin, lincomycin, and vancomy- limited antibiotic uptake during processing, or the
cin. In their study,19 P aeruginosa was the only bacterium elongated storage time.
not inhibited. Notably, CHAs without antibiotic treatment Because the procedure of processing of the 3 groups of
are highly susceptible to bacterial infection, whereas the CHAs only differed with regard to their antibiotic
reapplication of antibiotic agents after thawing also led to pretreatment, these results can predominantly be explained

The Journal of Thoracic and Cardiovascular Surgery c Volume 151, Number 5 1257
 Acquired: Aortic Valve Steffen et al
ACQ

on the basis of the various antibiotic mixtures and their acceptable preoperative mortality and reinfection rates of
residual activity in CHAs. 0% to 0.04%.5,16
In contrast, infection resistance was hypothesized to be However, CHAs also have distinct disadvantages such
attributable to the viability of CHAs. An improved transfer as their limited availability13 and their accelerated degener-
of immunocompetent cells and antibiotics through the wall ation.12,14,15 Therefore, alternative grafts are used for the
might lead to an immune-mediated eradication of management of per annular expansion of infectious
infection.18 However, the level of viability is a matter of endocarditis. Several authors indicate that the freedom
debate.21 Mirabet and colleagues31 demonstrated well- from reinfection or reoperation and late mortality is
preserved viable cells in cardiac valve allografts despite similar among mechanical composite grafts, biological
cryopreservation and long-term storage. Endothelial and valve conduits, and CHAs. They conclude that an
especially interstitial cells retain the capacity to proliferate aggressive debridement and prolonged antibiotic
in vitro.31 treatment, and not the type of substitute material, are the
However, several studies show conflicting results. critical factors in determining clinical outcome.10,11 With
Koolbergen and colleagues32 found that a reduction in regard to radical surgical techniques, CHAs might allow a
cellularity and a loss of endothelial cells led to almost more conservative debridement, which is essential in the
acellular tissue 1 year after the implantation of CHAs. mediastinum because of the surrounding vital structures,5
Moreover, nonimplanted valves exhibited a loss of particularly because the tested aortic wall specimens of
endothelial cells. Mitchell and colleagues33 showed similar groups B and C possess an enhanced infection resistance
results indicating that explanted cryopreserved heart valves against the frequent infection-causing staphylococcal
are morphologically nonviable, even after only 1 day of bacteria. Nevertheless, the low incidence of graft infections,
implantation. Thus, they are unlikely to have the ability to heterogeneous study populations, diverse therapeutic
grow, remodel, or exhibit metabolic functions.33 In our approaches, and various protocols for CHA processing
study, SEM showed subendothelial fibers in groups A hinder the comparison of outcomes.
through C, indicating a loss of endothelial cells. This could
be related to several factors during the processing and/or our Limitations
experimental setup irrespective of the used antibiotics. The The number of CHAs analyzed in this study was rather
loss is presumably more pronounced in the cusps than in the small. However, legal requirements restrict the use of
aortic wall and might predispose to bacterial colonization, CHAs for research. Consequently, we had to employ
possibly also explaining the increased susceptibility of CHAs that had been stored for>5 years and were therefore
cusps to staphylococcal bacteria. An intact endothelium is no longer considered suitable for transplantation. For ethical
resistant to bacterial colonization34; however, the loss of reasons, we were unable to investigate untreated and fresh
endothelial coverage in groups A through C suggests that human tissue to compare our results. Another shortcoming
this protective layer is not the relevant factor in avoiding is that an analysis of residual concentrations of antibiotics
bacterial adherence to CHAs (Figures 4 and 5). in long-term CHAs is still missing. Furthermore, infection
An improved transfer of immunocompetent cells and resistance was assessed in a static instead of dynamic flow
antibiotics through the wall might also be responsible for model, neglecting the potential influence of circulation.
the bacterial resistance, possibly because of at least partially We are aware that the influence of the whole living organism,
viable tissue. However, our data comparing the CHAs of and particularly the immune-mediated eradication of infec-
groups A through C suggest that the bacterial resistance is tion, cannot be evaluated using our in vitro model.
dependent on the type of antibiotic pretreatment and is
not primarily an intrinsic ability of CHAs. CONCLUSIONS
The clinical use of CHAs in aortic prosthetic valve CHAs offer a preventive and reliable treatment option
endocarditis represents a generally accepted treatment.1,2 against reinfection. Because their infection resistance
In particular, CHAs are believed to be the best conduit for depends on the antibiotic combination selected during
a destroyed annulus, a frequent complication in 56% to processing, further investigations concerning this treatment
100% of cases associated with prosthetic valve are necessary to improve the antimicrobial activity against
endocarditis predominantly caused by staphylococci35 frequent and highly virulent infection-causing bacteria.
because of their infection resistance. In the presence of a Additional impregnation with antibiotic agents during
root abscess, the reconstruction of the aortic root can be implantation might offer a further protective procedure for
achieved with allograft aortic valves, because the adhering avoiding bacterial adherence to the graft.
anterior mitral leaflet can be used to patch defects created
by the resection of the abscess.15 Concerning the treatment Conflict of Interest Statement
of prosthetic graft infections involving the ascending aorta, Authors have nothing to disclose with regard to commercial
CHAs have demonstrated encouraging outcomes with support.

1258 The Journal of Thoracic and Cardiovascular Surgery c May 2016

Steffen et al Acquired: Aortic Valve

ACQ
The authors thank Nina McGuinness for providing language 18. Vogt PR, Brunner-LaRocca HP, Lachat M, Ruef C, Turina MI. Technical details
editing and Reiner Gebauer for providing technical assistance in with the use of cryopreserved arterial allografts for aortic infection: influence on
early and midterm mortality. J Vasc Surg. 2002;35:80-6.
electron microscopy. 19. Camiade C, Goldschmidt P, Koskas F, Ricco JB, Jarraya M, Gerota J, et al.
Optimization of the resistance of arterial allografts to infection: comparative
References study with synthetic prostheses. Ann Vasc Surg. 2001;15:186-96.
1. Habib G, Thuny F, Avierinos JF. Prosthetic valve endocarditis: current approach 20. Kuehn C, Graf K, Mashaqi B, Pichlmaier M, Heuer W, Hilfiker A, et al.
and therapeutic options. Prog Cardiovasc Dis. 2008;50:274-81. Prevention of early vascular graft infection using regional antibiotic release. J
2. Mahesh B, Angelini G, Caputo M, Jin XY, Bryan A. Prosthetic valve Surg Res. 2010;164:185-91.
endocarditis. Ann Thorac Surg. 2005;80:1151-8. 21. Yap CH, Yii M. Allograft aortic valve replacement in the adult: a review. Heart
3. Coselli JS, Koksoy C, LeMaire SA. Management of thoracic aortic graft Lung Circ. 2004;13:41-51.
infections. Ann Thorac Surg. 1999;67:1990-3. 22. Heng WL, Albrecht H, Chiappini P, Lim YP, Manning L. International heart
4. Nataloni M, Pergolini M, Rescigno G, Mocchegiani R. Prosthetic valve valve bank survey: a review of processing practices and activity outcomes. J
endocarditis. J Cardiovasc Med (Hagerstown). 2010;11:869-83. Transplant. 2013;2013:163150.
5. Khaladj N, Pichlmaier U, Stachmann A, Peterss S, Reichelt A, Hagl C, et al. 23. Buzzi M, Guarino A, Gatto C, Manara S, Dainese L, Polvani G, et al. Residual
Cryopreserved human allografts (homografts) for the management of graft antibiotics in decontaminated human cardiovascular tissues intended for
infections in the ascending aortic position extending to the arch. Eur J transplantation and risk of falsely negative microbiological analyses. PLoS
Cardiothorac Surg. 2013;43:1170-5. One. 2014;9:e112679.
6. Bisdas T, Bredt M, Pichlmaier M, Aper T, Wilhelmi M, Bisdas S, et al. Eight-year 24. Jashari R, Faucon F, Hoeck BV, Gelas SD, Fan Y, Vandenbulcke S.
experience with cryopreserved arterial homografts for the in situ reconstruction Determination of residual antibiotics in cryopreserved heart valve allografts.
of abdominal aortic infections. J Vasc Surg. 2010;52:323-30. Transfus Med Hemother. 2011;38:379-86.
7. LeMaire SA, Coselli JS. Options for managing infected ascending aortic grafts. J 25. Zander J, Maier B, Zoller M, Dobbeler G, Frey L, Teupser D, et al. Effects of
Thorac Cardiovasc Surg. 2007;134:839-43. biobanking conditions on six antibiotic substances in human serum assessed by a
8. Tossios P, Karatzopoulos A, Tsagakis K, Sapalidis K, Grosomanidis V, novel evaluation protocol. Clin Chem Lab Med. July 21, 2015 [Epub ahead of print].
Kalogera A, et al. Treatment of infected thoracic aortic prosthetic grafts with 26. Nickolai DJ, Lammel CJ, Byford BA, Morris JH, Kaplan EB, Hadley WK, et al.
the in situ preservation strategy: a review of its history, surgical technique, and Effects of storage temperature and pH on the stability of eleven beta-lactam
results. Heart Lung Circ. 2014;23:24-31. antibiotics in MIC trays. J Clin Microbiol. 1985;21:366-70.
9. Czerny M, von Allmen R, Opfermann P, Sodeck G, Dick F, Stellmes A, et al. 27. K€onig C, Schwank S, Blaser J. Factors compromising antibiotic activity against
Self-made pericardial tube graft: a new surgical concept for treatment of graft biofilms of staphylococcus epidermidis. Eur J Clin Microbiol Infect Dis. 2001;
infections after thoracic and abdominal aortic procedures. Ann Thorac Surg. 20:20-6.
2011;92:1657-62. 28. Justice SS, Hunstad DA, Cegelski L, Hultgren SJ. Morphological plasticity as a
10. Leyh RG, Knobloch K, Hagl C, Ruhparwar A, Fischer S, Kofidis T, et al. bacterial survival strategy. Nat Rev Microbiol. 2008;6:162-8.
Replacement of the aortic root for acute prosthetic valve endocarditis: prosthetic 29. Gottfredsson M, Erlendsdottir H, Sigfusson A, Gudmundsson S. Characteristics and
composite versus aortic allograft root replacement. J Thorac Cardiovasc Surg. dynamics of bacterial populations during postantibiotic effect determined by flow
2004;127:1416-20. cytometry. Antimicrob Agents Chemother. 1998;42:1005-11.
11. Jassar AS, Bavaria JE, Szeto WY, Moeller PJ, Maniaci J, Milewski RK, et al. 30. Crosby HA, Bion JF, Penn CW, Elliott TS. Antibiotic-induced release of
Graft selection for aortic root replacement in complex active endocarditis: endotoxin from bacteria in vitro. J Med Microbiol. 1994;40:23-30.
does it matter? Ann Thorac Surg. 2012;93:480-7. 31. Mirabet V, Carda C, Solves P, Novella-Maestre E, Carbonell-Uberos F,
12. Fukushima S, Tesar PJ, Pearse B, Jalali H, Sparks L, Fraser JF, et al. Long-term Caffarena JM, et al. Long-term storage in liquid nitrogen does not affect cell
clinical outcomes after aortic valve replacement using cryopreserved aortic viability in cardiac valve allografts. Cryobiology. 2008;57:113-21.
allograft. J Thorac Cardiovasc Surg. 2014;148:65-72. 32. Koolbergen DR, Hazekamp MG, de Heer E, Bruggemans EF, Huysmans HA,
13. Vogt PR. Arterial allografts in treating aortic graft infections: something old, Dion RA, et al. The pathology of fresh and cryopreserved homograft heart valves:
something new. Semin Vasc Surg. 2011;24:227-33. an analysis of forty explanted homograft valves. J Thorac Cardiovasc Surg.
14. O’Brien MF, Harrocks S, Stafford EG, Gardner MA, Pohlner PG, Tesar PJ, et al. 2002;124:689-97.
The homograft aortic valve: a 29-year, 99.3% follow up of 1,022 valve 33. Mitchell RN, Jonas RA, Schoen FJ. Pathology of explanted cryopreserved
replacements. J Heart Valve Dis. 2001;10:334-44. allograft heart valves: comparison with aortic valves from orthotopic heart
15. Musci M, Weng Y, Hubler M, Amiri A, Pasic M, Kosky S, et al. Homograft transplants. J Thorac Cardiovasc Surg. 1998;115:118-27.
aortic root replacement in native or prosthetic active infective endocarditis: 34. Thiene G, Basso C. Pathology and pathogenesis of infective endocarditis in
twenty-year single-center experience. J Thorac Cardiovasc Surg. 2010;139: native heart valves. Cardiovasc Pathol. 2006;15:256-63.
665-73. 35. Anguera I, Miro JM, San Roman JA, de Alarcon A, Anguita M, Almirante B,
16. Lytle BW, Sabik JF, Blackstone EH, Svensson LG, Pettersson GB, et al. Periannular complications in infective endocarditis involving prosthetic
Cosgrove DM III. Reoperative cryopreserved root and ascending aorta aortic valves. Am J Cardiol. 2006;98:1261-8.
replacement for acute aortic prosthetic valve endocarditis. Ann Thorac Surg.
2002;74:1754-7. Key Words: ascending aorta, aortic valve, bacterial infec-
17. Svensson LG, Adams DH, Bonow RO, Kouchoukos NT, Miller DC, O’Gara PT,
et al. Aortic valve and ascending aorta guidelines for management and quality tion, cryopreserved human allograft, infective endocarditis,
measures. Ann Thorac Surg. 2013;95:1-66. prosthetic graft infection

The Journal of Thoracic and Cardiovascular Surgery c Volume 151, Number 5 1259