Fax +41 61 306 12 34E-Mail karger@karger.chwww.karger.com

Original Paper

Eur Surg Res 2005;37:137–143 DOI: 10.1159/000085960

Kangaroo vs. Porcine Aortic Valves: Calcifi cation Potential after Glutaraldehyde Fixation

K. Narine

a Cyrille C. Chéry

b Els Goetghebeur

c R. Forsyth

d E. Claeys

e Maria Cornelissen

f L. Moens

b G. Van Nooten

a

Departments of a Cardiac Surgery, b

Analytical Chemistry, c Applied Mathematics and Computer Science,

d Pathology, e

Animal Production, and f Histology and Human Anatomy, University of Ghent, Ghent , Belgium

cifi cation of valves treated with 0.6 or 2% glutaraldehyde within and between the two species. Using the subcuta-neous model, we did not detect a difference in calcifi ca-tion potential between kangaroo and porcine aortic valves treated with either high or low concentrations of glutaraldehyde.

Copyright © 2005 S. Karger AG, Basel

Introduction

The majority of commercially available biological valve prostheses are made from porcine aortic valves or bovine pericardium after fi xation in low concentrations of glutaraldehyde (1,5-pentane dialdehyde; CHO (CH2) 3

CHO) [1, 2] . Although the precise mechanism is still con-troversial, calcifi c deterioration is a major cause of con-temporary bioprosthetic heart valve failure [3, 4] .

Carpentier et al. [5] reported signifi cant differences amongst different donor and recipient species with regard to calcifi cation of biological tissue. Weinholdt et al. [6] and more recently Neethling et al. [7] reported on the po-tential of kangaroo aortic valves to calcify less than por-cine aortic valves in vivo, using the sheep model.

Subcutaneous implantation of bioprosthetic valves in the rat model is still commonly used to investigate calci-fi cation potential. Apart from its low cost and simplicity, this subcutaneous model is attractive because the bio-

Key Words Kangaroo � Porcine � Aortic valves � Calcifi cation

Abstract The aim of this study was to evaluate and compare the calcifi cation potential of kangaroo and porcine aortic valves after glutaraldehyde fi xation at both low (0.6%) and high (2.0%) concentrations of glutaraldehyde in the rat subcutaneous model. To our knowledge this is the fi rst report comparing the time-related, progressive cal-cifi cation of these two species in the rat subcutaneous model. Twenty-two Sprague-Dawley rats were each im-planted with two aortic valve leafl ets (porcine and kan-garoo) after fi xation in 0.6% glutaraldehyde and two aor-tic valve leafl ets (porcine and kangaroo) after fi xation in 2% glutaraldehyde respectively. Animals were sacrifi ced after 24 h and thereafter weekly for up to 10 weeks after implantation. Calcium content was determined using in-ductively coupled plasma-mass spectrometry and con-fi rmed histologically. Mean calcium content per milli-gram of tissue (dry weight) treated with 0.6 and 2% glutaraldehyde was 116.2 and 110.4 � g/mg tissue for kangaroo and 95.0 and 106.8 � g/mg tissue for porcine valves. Calcium content increased signifi cantly over time (8.8 � g/mg tissue per week) and was not signifi cantly different between groups. Regression analysis of calcifi -cation over time showed no signifi cant difference in cal-

Received: December 9, 2004 Accepted after revision: March 3, 2005

Dr. Kishan NarineDepartment of Cardiac Surgery – 5K12, University Hospital GhentDe Pintelaan 185, BE–9000 Ghent (Belgium)Tel. +32 9 240 4700, Fax +32 9 240 3339 E-Mail kishan.narine@ugent.be

© 2005 S. Karger AG, Basel0014–312X/05/0373–0137$22.00/0

Accessible online at:www.karger.com/esr

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM

Narine /Chéry /Goetghebeur /Forsyth /Claeys /Cornelissen /Moens /Van Nooten

Eur Surg Res 2005;37:137–143 138

logical and morphological features of intrinsic leafl et cal-cifi cation seen have been reported to be analogous to those seen in clinical explants but at an accelerated rate [8–10] .

The use of glutaraldehyde as the preserving agent of bioprosthetic heart valves became widespread three de-cades ago after it was reported that fi xation of porcine heart valves in this agent resulted in stable cross-links and rendered the tissue essentially non-immunogenic [11, 12]. Several reports, however, have implicated glutaral-dehyde in the calcifi cation process of bioprosthetic heart valves [13, 14] . In order to avoid excessive glutaralde-hyde, current bioprosthetic valves are fi xed in low con-centrations ( ! 1%) of glutaraldehyde. Despite the forego-ing and the use of antimineralization treatments in tissue valve manufacture, the long-term durability of glutaral-dehyde-treated bioprosthetic valves continue to be lim-ited by their propensity to calcify [15–17] .

The reported low calcifi cation potential of glutaralde-hyde-fi xed kangaroo valves in the sheep circulatory mod-el prompted us to compare the calcifi cation potential of kangaroo and porcine aortic valves after glutaraldehyde fi xation in the rat subcutaneous model. Moreover, to evaluate a possible infl uence of the concentration of glu-taraldehyde used, we evaluated valves after fi xation in low and high concentrations of glutaraldehyde, namely 0.6 and 2.0% phosphate-buffered glutaraldehyde solu-tions.

Materials and Methods

Animals Twenty-two Sprague-Dawley rats weighing approximately

150 g each and 6 weeks old were obtained from Harlan Laborato-ries (Horst, The Netherlands). Animals were allowed to acclimatize to the Animal Facilities at our institution for 1 week before the in-vestigations. All animal care complied with The Guide for the Care and Use of Laboratory Animals, 1996 [18] . All of our protocols in-volving animals were approved by the Ethical Commission for An-imal Experiments of the University and the University Hospital of Ghent (Project No. ECP 022).

Valve Leafl et Procurement and Fixation Porcine valves (n = 15) used in this study were obtained from

the slaughterhouse of the Department of Animal Production, Uni-versity of Ghent. Immediately after slaughter, hearts were retrieved and placed on wet ice for transportation to the Laboratory of Ex-perimental Cardiac Surgery, University Hospital, Ghent.

Aortic valves (n = 15) from Western Grey kangaroos (Macropo-sus fuliginosus) were harvested and supplied under law by a mem-ber of the Professional Shooters Association of Western Australia. After harvest, hearts were placed on ice and transported to the Fre-mantle Heart Institute, University of Western Australia, from

where they were shipped by air to our laboratory after glutaralde-hyde fi xation. Before fi xation, each heart was rinsed with cold saline (0.9% NaCl) before valves were dissected free and again rinsed in cold saline.

Methods A total of 44 kangaroo and 44 porcine leafl ets were implanted.

Valves in each group were randomized. Twenty-two leafl ets of each type were fi xed in phosphate (KH 2 PO 4 )-buffered (pH 7.4) 0.6% glutaraldehyde. The remaining 22 leafl ets of each type were fi xed in 2% glutaraldehyde in the same buffer. Fixing solutions were pre-pared from a stock solution of 25% glutaraldehyde (Merck, Darm-stadt, Germany). Potassium phosphate (KH 2 PO 4 ) was also pur-chased from Merck. The fi nal pH (7.4) of both concentrations of glutaraldehyde was achieved by titration using 0.5 M NaOH. Leaf-lets were fi xed for 7 days, after which they were transferred to a solution of 0.2% glutaraldehyde in the same buffer and stored at4 ° C until further use.

Implantation and Explantation of Leafl ets All animals were anesthetized using a combination of ketamine

(0.9 mg/g), xylazine (0.0075 mg/g) and atropine (0.0001 mg/g). Fol-lowing anesthesia, four subcutaneous pouches were created in each animal, one in each abdominal quadrant. Leafl ets were implanted in each animal as follows. In the right upper quadrant, one kanga-roo leafl et fi xed in 0.6% glutaraldehyde and in the right lower quad-rant, one kangaroo leafl et fi xed in 2% glutaraldehyde. In the cor-responding upper and lower quadrants on the left side, porcine leafl ets fi xed in 0.6 and 2% glutaraldehyde were implanted respec-tively. All leafl ets were thoroughly rinsed to remove excess glutar-aldehyde before implantation. After recovery from anesthesia, all animals were returned to the animal facilities and fed a standard rat diet. Two animals were sacrifi ced after 24 h, and then another 2 weekly for up to 10 weeks. After sacrifi ce, each leafl et was divid-ed into segments by cutting from the free edge down towards the base. One segment was taken at random, frozen at –80 ° C and kept for quantitative calcium determination. Another segment was fi xed in 4% formaldehyde for histological examination.

Histology Specimens of non-implanted (fresh) and implanted kangaroo

and porcine aortic valve leafl ets were examined histologically. Fresh specimens were fi xed in 4% phosphate-buffered formalde-hyde and stained with Masson’s trichrome to illustrate the gross histological structure. Explanted samples for light microscopy were fi xed immediately in 4% phosphate-buffered formaldehyde. Cal-cifi c deposits were identifi ed with von Kossa stain.

Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) In this work, a quadrupole-based ICP-MS [19–21] (Elan DRC

plus, PerkinElmer-SCIEX, Concord, Canada) was used. This equip-ment contains a dynamic reaction cell, in which selective ion-mol-ecule chemistry allows the reduction of interferences that the ana-lyte might suffer from. As such, the technique allows for the detec-tion of signifi cantly lower concentrations of calcium (approx. 1,000-fold) than classic atomic absorption spectrometry. In our ap-plication, this technology permits the elimination of the argon ions ( 40 Ar + ), which may interfere with the determination of calcium (Ca) via the signal of 40 Ca + .

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM

Calcifi cation of Kangaroo vs. Porcine Aortic Valves

Eur Surg Res 2005;37:137–143 139

Explanted samples were thawed in a clean room, Class 100. Every precaution was taken to avoid sample contamination. Sam-ples were washed twice with MilliQ water (doubly distilled water which was further purifi ed using a MilliQ water purifi cation system (Millipore, Bedford, Mass., USA)). Tissues were lyophilized and the dry weight determined before undergoing microwave oven- assisted acidic digestion (Microwave Digestion System: MLS-1200 MEGA Technology, Milestone, USA, with MDR (microwave di-gestion rotors), used with tetrafl uormethaxil vessels). Samples plus 1 ml 14 M HNO 3 (purifi ed by sub-boiling in quartz equipment), 0.2 ml H 2 O 2 , 3 ml H 2 O, 100 � l of 50 mg/l cobalt (Alfa, Karlsruhe, Germany). The samples in solution were diluted to 100 ml using MilliQ water.

Measurements were performed on the sample solutions after being diluted another 100 times. Cobalt is used as the internal ref-erence for all ICP-MS measurements. Blank solutions and external standards were prepared in an analogous manner to the samples. Calcium concentrations were expressed in micrograms of calcium per milligram of tissue (dry weight).

Statistical Analysis Calcium content over time was analyzed using regression meth-

ods and analysis of variance (ANOVA). Two groups of 11 animals yielded duplicate observations on calcium content in porcine and

kangaroo leafl ets treated with 0.6 and 2% glutaraldehyde as 1 ani-mal per group was sacrifi ced at 24 h after implantation and then weekly for 10 weeks. We fi tted regression models which allowed the mean response to vary with time, species and glutaraldehyde con-centration. We also looked at possible interaction effects between these prognostic factors. Time, and only the time variable, contrib-uted signifi cantly to the model. Once this was verifi ed, we adjusted all treatment comparisons for the time effect. This allowed us to analyze average outcomes as well as variations over the weeks. Hence, when we report mean differences, these differences are pres-ent at fi xed values of time. Calcium content was regressed on time for kangaroo and porcine valves fi xed in 0.6 and 2% glutaraldehyde, respectively. To compare calcium concentrations between kanga-roo and porcine valves, we considered per animal the difference in calcifi cation of its implanted kangaroo and porcine valves at each concentration of glutaraldehyde. These paired differences were then regressed on time, on the concentration of glutaraldehyde and on their interaction. The fi nal model was decided upon by perform-ing a backward stepwise regression procedure (with p = 0.05 for retention).

A similar analysis was performed to compare results at different concentration levels of glutaraldehyde in the same species of valve. Missing data were treated as missing completely at random.

Fig. 1. Representative histological prepara-tions of the body of the leafl et of fresh kan-garoo (K) and porcine (P) aortic valve leaf-lets (Masson’s trichrome stain). f = Fibrosa; s = spongiosa; v = ventricularis. Collagen stains blue. ! 100.

Fig. 2. Representative histological prepara-tions of kangaroo (K) and porcine (P) after explantation (von Kossa stain) at 24 h (K1 and P1), 7 weeks (K2 and P2) and 14 weeks (K3 and P3) showing the body or central portion of the leafl ets. Calcium stains black. ! 100.

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM

Narine /Chéry /Goetghebeur /Forsyth /Claeys /Cornelissen /Moens /Van Nooten

Eur Surg Res 2005;37:137–143 140

Results

Histology Figure 1 shows histological preparations of unimplant-

ed kangaroo and porcine leafl ets. In both species a fi brous layer or fi brosa is located on the aortic surface of the leaf-let and the less fi brous ventricularis layer is located on the ventricular or infl ow surface. Separating the fi brosa and ventricularis is a spongious layer or spongiosa which is thicker in the porcine leafl ets. The density of collagen bundles in the fi brosa is most pronounced in the kangaroo valve leafl ets when compared with porcine valve leafl ets. In addition, the ventricularis layer in kangaroo valves is thicker than in porcine valves and more staining of col-lagen was observed in this layer in kangaroo leafl ets.

Von Kossa stain of explants confi rmed calcifi cation in both porcine and kangaroo leafl ets. Morphologically, we observed no difference in calcifi cation in comparable ex-plants over time. In both types we observed calcifi cations in all three layers of the leafl et but more extensive in the spongiosa and ventricularis. Figure 2 shows representa-tive von Kossa preparations for porcine and kangaroo leafl ets explanted at 24 h, 7 weeks and 10 weeks.

Calcifi cation The raw data for calcium content in explanted leafl ets

are shown in table 1 a and are statistically summarized in table 1 b. Figure 3 shows the regression lines per type of valve for this raw calcium data; calcium content is ex-

pressed in micrograms of calcium per milligram of tissue (dry weight) ( � g Ca/mg tissue).

The difference in calcifi cation of valves treated with 2.0 and 0.6% glutaraldehyde was regressed on time, on the concentration of glutaraldehyde and on their interac-tion ( table 2 a). We found a signifi cant increase in calcifi -cation over time of 8.8 � g Ca/mg tissue per week (95% confi dence interval [5.5 and 12.1]) that did not differ sig-nifi cantly between the four groups of valves. A regression analysis of within-animal differences in calcifi cation be-tween valves treated with 0.6 and 2.0% glutaraldehyde solutions over time showed no signifi cant difference be-tween the different concentrations. Table 2 b shows the statistical summary for these paired differences. Figure 4 shows the raw data with the regression lines for this anal-

Fig. 3. Regression lines per type of valve for the raw calcium data shown in table 1a. K = Kangaroo, P = porcine, 0.6 and 2.0 represent glutaraldehyde concentrations respectively; AV = aortic valve.

Table 1. Raw data of the average calcium content in explanted leaf-lets (a) and statistical summary of the raw data shown (b)

a

Explants Average calcium content, �g Ca/mg dry weight of tissue

P0.6 K0.6 P2.0 K2.0

24 h 1.75 1.7 1.9 2.11 week 46.1 28.2 40.2 100.92 weeks 104.6 102.9 126.7 96.03 weeks 65.6 184.1 137.0 145.34 weeks 106.8 89.95 103.0 128.05 weeks 112.7 181.5 110.0 108.56 weeks 186.0 161.5 116.0 151.07 weeks 127.5 141.0 98.5 127.08 weeks 75.0 115.0 130.0 129.09 weeks 120.0 105.8 103.5 131.0

10 weeks 138.5 166.0 122.0 155.0

b

P0.6 K0.6 P2.0 K2.0

Number 19 22 19 20Mean 94.9842 116.1500 106.7947 110.3500Standard error of

mean 12.48984 15.58478 11.83288 7.53473Standard deviation 54.44196 73.09911 51.57835 33.69633Valid percentiles25 65.6000 46.5000 87.0000 104.000050 112.7000 128.4500 125.0000 111.450075 138.0000 167.7500 144.1000 128.0000

P = Porcine, K = kangaroo, 0.6 and 2.0 represent different glu-taraldehyde concentrations.

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM

Calcifi cation of Kangaroo vs. Porcine Aortic Valves

Eur Surg Res 2005;37:137–143 141

ysis of porcine and kangaroo valves treated with 2 and 0.6% glutaraldehyde, respectively. The stepwise back-ward regression (with p = 0.05 for retention) retained no predictors in this model. A similar analysis for differ-ences in calcifi cation between the porcine and kangaroo valves gave similar results: no signifi cant predictors re-mained in the model and the global average difference

shown in table 2 b is not signifi cantly different from zero. As the mean difference in calcifi cation between the por-cine and kangaroo valves (adjusted for time and level of glutaraldehyde) has an estimated error of 10.6, the non-signifi cant difference is not surprising. Most striking in-deed are the large variations seen in outcomes between and within animals.

Fig. 4. Regression lines of the raw data of the paired differences in calcium concentra-tion in porcine (left) and kangaroo (right) valves treated with 2 and 0.6% glutaralde-hyde over time.

a

Model Unstandardized coeffi cients

Signifi -cance

95% confi dence interval for B

B standarderror

lowerboundary

upperboundary

1 (Constant) 66.750 19.073 0.001 28.762 104.736Time 8.790 1.638 0.000 5.527 12.052Porcine –11.960 10.641 0.265 –33.154 9.234Dose 0.972 10.635 0.927 –20.209 22.154

b

P2.0–P0.6 K2.0–K0.6 P0.6–K0.6 P2.0–K2.0

Number 16 20 19 18Mean 9.9687 –11.8300 –15.8789 6.2333Standard error of mean 11.34243 12.34426 14.36324 8.71913Median 4.5000 –5.5000 –8.0000 9.5500Standard deviation 45.36972 55.20523 62.60793 36.99213Valid percentiles 25 9.6750 –58.2500 –59.9000 –1.7500

50 4.5000 –5.5000 –8.0000 9.550075 30.1250 30.2500 29.0000 26.7500

K = Kangaroo valves; P = porcine valves; 2.0 and 0.6 represent glutaraldehyde concen-trations.

Table 2. Linear regression model of the raw data (a) and summary statistics for the paired differences in calcium concen-tration of valves treated with 2% and 0.6% glutaraldehyde (b)

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM

Narine /Chéry /Goetghebeur /Forsyth /Claeys /Cornelissen /Moens /Van Nooten

Eur Surg Res 2005;37:137–143 142

Discussion

In our subcutaneous model, kangaroo and porcine valves show an increase in calcifi cation with time and confi rm a trend which has been seen in porcine aortic valve leafl ets [9] . There was, however, no signifi cant dif-ference in the amount of calcifi cation seen in porcine and kangaroo leafl ets treated with 0.6 or 2% glutaraldehyde over time. Neethling et al. [7] reported lower calcifi cation in kangaroo aortic valve leafl ets implanted subcutane-ously in rats and explanted after 8 weeks. However, from our observations, while the calcium content in explanted leafl ets may differ at individual time points, there was no signifi cant difference in the progression of calcifi cation over time in the two species. Considering the small sam-ple sizes involved, we must realize when interpreting the p values that the data show a large variation in outcomes and little suggestion of an important possible difference between kangaroo and porcine valves in this setting. From our observations, 2% glutaraldehyde fi xation did not result in signifi cantly more calcifi cation than 0.6% glutaraldehyde in leafl et tissue in either porcine or kan-garoo aortic valves. While this study did not evaluate glu-taraldehyde concentrations 1 2%, it does indicate that at least at a concentration of 2%, glutaraldehyde did not re-sult in signifi cantly more calcifi cation in the subcutane-ous model than a concentration of 0.6%. Indeed, Zilla et al. [22] , who investigated the effect of high glutaraldehyde concentrations on calcifi cation of aortic wall tissue, sug-gested that high glutaraldehyde concentrations of glutar-aldehyde reduce rather than increase tissue calcifi cation potential. Although our study investigated aortic leafl et tissue, it is in keeping with the fi ndings of the latter au-thors in that there was no signifi cant increase in tissue calcifi cation with higher glutaraldehyde concentration in our model.

In the sheep circulatory model, Weinholdt et al. [6] as well as Neethling et al. [7] reported less calcifi cation of the kangaroo aortic valve. Interestingly, our subcutane-ous static model did not refl ect the fi ndings reported in the circulatory model. The correlation between calcifi ca-tion in the rat subcutaneous and circulatory models has been questioned by several authors [10] . A possible ex-planation of the different observations in the two models is that subcutaneously implanted tissues are not in a cir-culatory system and are thus not subjected to the same mechanical forces. This argument was reinforced by Car-pentier et al. [11] who demonstrated that iron pretreat-ment mitigated calcifi cation of aortic valves in the rat subcutaneous model, but appeared to promote calcifi ca-

tion when tested in the sheep circulatory model. Several authors have suggested that mechanical forces associated with the circulatory model might signifi cantly contribute to tissue valve calcifi cation [23–25] . Vyavahare et al. [24] suggested that fatigue induced damage to type 1 collagen and loss of glycosaminoglycans are major contributing factors to material degeneration in bioprosthetic valves . In our group, Van Nooten et al. [25] emphasized the role of mechanical stresses by implanting bioprosthetic valves asymmetrically in the sheep model. These authors con-cluded that abnormal mechanical stresses created by the implantation technique adversely affected calcifi cation potential. It is thus possible that mechanical stresses, not tested for in the subcutaneous model, can signifi cantly infl uence calcifi cation of leafl ets in the circulatory model. Furthermore, Meuris et al. [26] who studied the effects of recipient species, environmental factors and cellularity on the calcifi cation of aortic wall tissue in the rat subcu-taneous and sheep models suggested that blood contact in the sheep circulatory model may also infl uence calcifi -cation.

The mechanical behavior of tissue in a dynamic mod-el has been suggested to infl uence tissue deterioration. Vesely et al. [27] investigated the bending properties of glutaraldehyde-fi xed bovine pericardium and porcine aortic valves subjected to mechanical forces. These au-thors concluded that bovine pericardium demonstrated less compressive buckling than porcine aortic valves when bent. As a possible explanation of this observation, the latter authors suggested that bovine pericardium re-quires larger forces to bend after glutaraldehyde fi xation due to its dense and tightly layered structure without a spongiosa layer. In contrast, porcine valves are looser and can undergo compressive collapse of the spongiosa layer after fi xation. Such a structure requires less force to bend and its layers are more likely to undergo compressive buckling than bovine pericardium. Repetitive buckling could lead to structural deterioration and subsequent cal-cifi cation. Kangaroo valves have a denser arrangement of their collagen fi bers, with a thicker ventricularis, and a thinner spongiosa layer than their porcine counterparts. As such, one possible contributing factor to less calcifi ca-tion in kangaroo valves in the circulatory model com-pared to porcine aortic valves might be less compressive buckling in the denser kangaroo tissue ( fi g. 1 ).

We also observed large variations in leafl et calcifi cation between animals. Variations in the calcium content of aor-tic valves implanted in the rat subcutaneous model are also evident in the literature. In particular, mean calcifi ca-tion levels of 130.0 and 177.8 � g/g were reported by Schoen

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM

Calcifi cation of Kangaroo vs. Porcine Aortic Valves

Eur Surg Res 2005;37:137–143 143

et al. [9] and Hirsch et al. [28] in similar studies using glu-taraldehyde-fi xed porcine aortic valve leafl ets. Mako and Vesely [10] who examined in vivo and in vitro models of calcifi cation suggested that in the rat subcutaneous model the highest variability of calcifi cation likely resulted from variability in animals or the initial condition producing calcifi cation . In addition to rat variability, we have also observed within-animal variations in this model.

Conclusion

This study compares for the fi rst time the rate of cal-cifi cation in kangaroo and porcine aortic valve leafl ets in the rat subcutaneous model. We could not establish a sig-nifi cant difference in progressive calcifi cation in kanga-roo and porcine aortic valve leafl ets in our study after

glutaraldehyde fi xation or in the histological pattern of calcifi cation. Fixation of both kangaroo and porcine aor-tic valves in 2% glutaraldehyde did not result in signifi -cantly more calcifi cation compared to valves fi xed in 0.6% glutaraldehyde. Finally, in experimental work using the rat subcutaneous model, it should be emphasized that this model might not amply represent the in vivo circula-tory situation.

Acknowledgements

The authors wish to thank Mrs. Maria Olieslagers, Research Nurse, for her administrative and technical support; Mr. Valentijn Van Parys, Laboratory of Experimental Cardiac Surgery, Univer-sity Hospital Ghent, for his technical support in this study, and Dr. L. Neethling, Fremantle Heart Institute, for providing the kangaroo valves.

References

1 Schoen FJ: Approach to the analysis of cardiac valve prostheses as surgical pathology or au-topsy specimens. Cardiovasc Pathol 1995; 241–247.

2 Edmunds LH, Clarck RE, Cohn LH, Grunke-maier GL, Miller DC, Weisel RD: Guidelines for reporting morbidity and mortality after cardiac valvular operations. Eur J Cardiotho-rac Surg 1996; 10: 812–816.

3 Kim KM, Herrera GA, Battarbee HD: Role of glutaraldehyde in calcifi cation of porcine aor-tic valve fi broblasts. Am J Pathol 1999; 154: 171.

4 Bortolotti U, Milano A, Mazzucco A: Results of reoperation for primary tissue failure of por-cine bioprostheses. J Thorac Cardiovasc Surg 1985; 90: 564–568.

5 Carpentier SM, Monier M, Shen M, et al: Do donor or recipient species infl uence calcifi ca-tion of bioprosthetic tissues. Ann Thorac Surg 1995; 60:S328–S331.

6 Weinholdt C, Adt M, Weingartner J, et al: In vivo investigation of kangaroo aortic valve bioprostheses: An experimental animal model; in Bodnar E, Yacoub M (eds): Biological and Bioprosthetic Valves. Proc 3rd Int Symp, ed 1. London, Yorke Medical Books, 1986, pp 669–676.

7 Neethling WM, van Riet S, van den Heever JJ, Hough J, Hodge AJ: Evaluation of kangaroo aortic valves conduits in a juvenile sheep mod-el. J Card Surg 2001; 16: 392–399.

8 Fishbein MC, Levy RJ, Ferrans VJ, Dearden LC, Nashef A, Goodman AP, Carpentier AP: Calcifi cation of cardiac valve bioprostheses: Biochemical, histologic and ultrastructural ob-servations in a subcutaneous implantation model system. J Thorac Cardiovasc Surg 1982; 83: 602–609.

9 Schoen FJ, Levy RJ, Nelson AC, Bernard WF, Nashef A, Hawley M: Onset and progression of experimental bioprosthetic heart valve cal-cifi cation. Lab Invest 1985; 52 523–532.

10 Mako WJ, Vesely I: In vivo and in vitro mod-els of calcifi cation in porcine aortic valve cusps. J Heart Valve Dis 1997; 6: 316–323.

11 Carpentier SM, Carpentier AF, Cen L, Shen M, Quintero LJ, Witzel TH: Calcium mitiga-tion in bioprosthetic tissues by iron pre-treat-ment: The challenge of iron leaching. Ann Thorac Surg 1995; 60:S332–S338.

12 Zuhdi N, Hancock W, Hawley W, Carey J, Greer A: Porcine aortic valves for replacement of human heart valves. J Cardiovasc Surg 1973; 8: 425–429.

13 Grabenwoger M, Sider J, Fitzal F, et al: Impact of glutaraldehyde on calcifi cation of pericar-dial bioprosthetic heart valve material. Ann Thorac Surg 1996; 62: 772–777.

14 Grabenwoger M, Grimm M, Eybl E, et al: New aspects of the degeneration of bioprosthetic heart valves after long-term implantation. J Thorac Cardovasc Surg 1992; 104: 14–21.

15 Schoen FJ, Levy RJ: Bioprosthetic heart valve failure: Pathology and pathogenesis. Cardiol Clin 1984; 2: 717–739.

16 Levy RJ, Schoen FJ, Golomb G: Bioprosthetic heart valve calcifi cation: Clinical features, pathobiology, and prospects for prevention. CRC Crit Rev Biocompat 1986; 2: 147–187.

17 Vyavahare N, Ogle M, Schoen FJ, Levy RJ: Elastin calcifi cation and its prevention with aluminium chloride pretreatment. Am J Pathol 1999; 155: 973–982.

18 Guide for the Care and Use of Laboratory An-imals. Bethesda, National Institutes of Health, Publ 85: 23, modifi ed 85.

19 Vandecasteele C, Block CB: Modern Methods for Trace Element Determination. Chichester, Wiley, 1997.

20 Montaser A (ed): Inductively Coupled Mass Spectrometry. New York, Wiley-VCH, 1998.

21 Tan SH, Horlick G: Background spectral fea-tures in inductively coupled plasma mass spec-trometry. Appl Spectrosc 1986; 40: 445–460.

22 Zilla P, Weissenstein C, Bracher M, Zhang Y, Koen W, Human P, von Oppel U: High glutar-aldehyde concentrations reduce rather than increase the calcifi cation of aortic wall tissue. J Heart Valve Dis 1997; 6: 502–509.

23 Sabbah HN, Hamid MS, Stein PD: Mechani-cal stresses on closed cusps of porcine biopros-thetic valves: Correlation with sites of calcifi -cation. Ann Thorac Surg 1986; 42: 93–96.

24 Vyavahare N, Ogle M, Schoen F, Zand R, Gloeckner D. Claire, Sacks M, Levy R: Mech-anisms of bioprosthetic heart valve failure: Fa-tigue causes collagen denaturation and glycos-aminoglycan loss. J Biomed Mater Res 1999; 46: 44–50.

25 Van Nooten G, Ozaki S, Herrijgers P, Segers P, Verdonck P, Flameng W: Distortion of the stentless porcine valve induces accelerated leafl et fi brosis and calcifi cation in juvenile sheep. J Heart Valve Dis 1999; 8: 34–41.

26 Meuris B, Ozaki S, Herijgers P, Verbeken E, Flameng W: Infl uence of species, environmen-tal factors, and tissue cellularity on calcifi ca-tion of porcine aortic wall tissue. Semin Tho-rac Cardiovasc Surg 2001; 13(suppl 1):99–105.

27 Vesely I, Mako WJ: Comparison of the com-pressive buckling of porcine aortic valve cusps and bovine pericardium. J Heart Valve Dis 1998: 34–39.

28 Hirsch D, Drader J, Thomas TJ, Schoen FJ, Levy JT, Levy RJ: Inhibition of glutaralde-hyde pretreated porcine aortic valve cusps with sodium dodecyl sulfate: Preincubation and controlled release studies. J Biomed Mater Res 1993; 27: 1477–1484.

Dow

nloa

ded

by:

Bio

med

isch

e B

iblio

thee

k

157.

193.

181.

143

- 5/

4/20

15 1

1:07

:13

AM