ANTIBIOTIC COCKTAIL AND METHOD OF USE

FIELD OF THE INVENTION

This invention relates to a method of decontaminating tissues, such as heart valves, for transplantation while maintaining the viability of the tissue. The invention also relates to an antibiotic cocktail for use in the method.

BACKGROUND OF THE INVENTION

The clinical performance of heart valve allografts has been correlated with the viability of the valve leaflet fibroblasts at transplantation. See, O'Brien et al., J. Card. Surer. , 2 (Suppl), 153-167 (1987) and Stark, J. Thoracic Cardiovasc. Surq. , 97, 1-9 (1989) . Accordingly, the highest viability standards must be maintained throughout the entire processing procedure which includes procurement, transportation, decontamination, freezing, storage, thawing and transplantation.

With respect to the decontamination step, antibiotic cocktails for microbial decontamination of tissue are known. In particular, several antibiotic cocktails are known which contain a plurality of antibacterial agents and a single antifungal agent

(amphotericin B or, occasionally, nystatin) . See, e.g. , Watts et al. , An . Thorac. Surer. , 21, 230-36

(1976) ; Strickett et al., Pathology. 15, 457-62 (1983) ; Armiger et al. , Pathology, 15, 67-73 (1983); Kirklin and Barratt-Boyes, Cardiac Surgery, 421-22 (1986) ; Heacox et al . , in Cardiac Valve Allografts 1962-1987, 37-42 (Yankah et al . eds. 1988) ; Angell et al . , J. Thorac. Cardiovasc. Surg., 98, 48-56 (1989) ; Lange and

Hopkins, in Cardiac Reconstruction With Allograft Heart Valves, 37-63 (Hopkins ed. 1989) ; U.S. Patent No. 4,890,457, ; U.S. Patent No. 4,695,536; and PCT application WO 92/12632.

Amphotericin B is considered to be the most effective antifungal agent available. However, amphotericin B has been found to be toxic to the cells responsible for allograft longevity in patients and has, consequently, been removed from some heart valve decontamination solutions. See Hu et al. , Cardiovasc. Res . , 23 , 960-64 (1989) ; Lange and Hopkins, in Cardiac Reconstruction With Allograft Heart Valves. 37-63 (Hopkins ed. 1989) ; McNally and Brockbank, J. Med. Engineer. & Tech. , 16, 34-38 (1992) . As a result, yeast contamination currently leads to the rejection of a significant number of heart valves. See McNally and Brockbank, J. Med. Engineer. & Tech., 16, 34-38 (1992) .

SUMMARY OF THE INVENTION

The invention provides an antibiotic cocktail for decontaminating tissue. The cocktail comprises: 1) amphotericin B and fluconazole as antifungal agents; and 2) a plurality of antibacterial agents. The agents are present in the cocktail in amounts effective to substantially inhibit both yeast and bacterial growth while substantially maintaining the viability of the tissue.

The invention also provides a method of decontaminating a tissue comprising contacting the tissue with an antibiotic cocktail according to the invention. The tissue is contacted with the antibiotic cocktail at a temperature and for a period of time effective to substantially inhibit the growth of yeast and bacteria while substantially maintaining the viability of the tissue.

As a result of the use of the antibiotic cocktails of the invention, many tissues previously rejected

because of yeast or bacterial contamination can now be used for transplantation. This is a significant achievement in view of the scarcity of tissue available for transplantation.

DETAILED DESCRIPTION OF THE INVENTION

"Amounts effective" is used to mean that each of the agents is present at a sufficient concentration such that the cocktail substantially inhibits yeast and bacterial growth but does not substantially decrease the viability of the tissue being decontaminated. These amounts can be determined by dose response testing as is known in the art using standard microbiological tests and viability tests such as those described below. Preferably the agents are present in the cocktail in amounts which are cidal for yeasts and bacteria frequently isolated from tissue.

"Substantially inhibits" means that the cocktail completely inhibits yeast and bacterial growth in at least 90%, preferably at least 95%, most preferably at least 99%, of the tissues decontaminated with the cocktail. "Completely inhibits yeast and bacterial growth" means that yeast and bacterial growth are not detectable by standard microbiological assays after the tissue has been treated with the cocktail.

Viability can be measured in a number of ways. Preferably, however, the tissue is incubated with a radioactively-labeled amino acid, and the incorporation of the amino acid into proteins is monitored by counting disintegrations per minute (DPM) per unit of tissue. "Substantially maintaining the viability" is, therefore, defined to mean that tissue which has been treated with the cocktail incorporates at least about 85% of the DPM per unit tissue as does tissue not treated with the cocktail.

The antifungal agents used in the cocktails of the invention are amphotericin B and fluconazole. The use

of these two antifungal agents in combination substantially inhibits yeast growth while substantially maintaining the viability of tissue to be treated with the agents. This result is quite surprising in view of the reported toxicity of amphotericin B to tissue (see Background section above) .

Suitable concentrations of amphotericin B and fluconazole can be determined by dose response testing as is known in the art using standard microbiological tests and viability tests such as those described below. Neither amphotericin B, nor fluconazole, alone substantially inhibits yeast growth at doses which substantially maintain tissue viability, but the combination of these two antifungal agents does. Indeed, it has unexpectedly been found that the combination of amphotericin B and fluconazole has synergistic antiyeast activity (see Examples 14 and 15) .

A concentration of from about 0.3 μg/ml to about 2.0 μg/ml of amphotericin B is preferred for use in antibiotic cocktails for sterilizing heart valves. A concentration of fluconazole of from about 50 μg/ml to about 100 μg/ml is preferred for use in antibiotic cocktails for sterilizing heart valves.

The antibacterial agents useful in the practice of the invention are chosen so that the combination of antibacterial agents will be effective against a wide range of bacteria, including gram-negative, gram- positive, aerobic and anaerobic bacteria. In addition, the antibacterial agents are chosen so that the combination of agents is effective against bacteria commonly found to contaminate the tissue being treated. Many such bacteria are known (e.g. , staphylococci, streptococci and propionibacteria) and others can be identified by standard microbiological tests (see, e.g.. Example 19 below) . Thus, broad spectrum antibacterial agents from two or more families are

preferred. Finally, the combination of antibacterial agents must not substantially decrease the viability of the tissue being treated.

Preferably the plurality of antibacterial agents is chosen from the following families: cephalosporins, glycopep ides, aminoglycosides, lincosamides, beta- lactams and rifamycins. More preferably, the combination of antibacterial agents comprises vancomycin and imipenem, and most preferably vancomycin, imipenem and netilmicin. For the decontamination of heart valves, a combination of 50 μg/ml vancomycin, about 10-100 μg/ml imipenem and about 20-50 μg/ml netilmicin is preferred.

Imipenem is a beta-lactam antibiotic. It is active against most aerobic gram-positive and gram- negative bacteria and most anaerobic gram-positive and gram-negative bacteria.

Vancomycin is a tricyclic glycopeptide. It is active against many gram-positive organisms, including staphylococci, streptococci, enterococci, Clostridium and Corynebacterium. It is inactive against gram- negative bacteria.

Netilmicin is in the aminoglycoside family and is effective against many aerobic gram-negative bacteria and some aerobic gram-positive bacteria. Netilmicin is inactive against most streptococci and most anaerobic bacteria.

The concentrations of the antibacterial agents are chosen to be at least 4 to 8 times the minimum inhibitory concentrations for the targeted bacteria as determined by standard microbiological sensitivity assays. Within these parameters, the concentrations of antibacterial agents can be adjusted as a result of dose response testing on tissue using standard microbiological tests and viability tests such as those described below.

From the foregoing, it can be seen that preferred antibiotic cocktails according to the invention contain amphotericin B, fluconazole, vancomycin, imipenem and netilmicin. For sterilizing heart valves, antibiotic cocktails containing about

0.3-1.0 μg/ml amphotericin B, about 50-100 μg/ml fluconazole, about 50 μg/ml vancomycin, about 10-100 μg/ml imipenem and about 20-50 μg/ml netilmicin are preferred. More preferred is the following cocktail:

1.0 μg/ml amphotericin B, 100 μg/ml fluconazole, 50 μg/ml vancomycin, 96 μg/ml imipenem, and 50 μg/ml netilmicin.

The invention also provides a method of sterilizing a tissue comprising contacting the tissue with an antibiotic cocktail according to the invention. A variety of tissues may be decontaminated in this manner, including heart valves, pericardium, vessels, and musculoskeletal connective tissue. As used herein, the term "musculoskeletal connective tissue" includes tissue such as tendons, ligaments and menisci, and excludes tissue such as bone.

The tissue is contacted with the antibiotic cocktail at a temperature and for a period of time effective to substantially inhibit yeast and bacterial growth while substantially maintaining the viability of the tissue. Such times and temperatures can be determined empirically as is known in the art . It has been found that heart valves can be effectively decontaminated by incubating them in an antibiotic cocktail according to the invention for 24-48 hours at a temperature of 35-39°C.

EXAMPLES EXAMPLE 1

This example describes the testing of four antibiotic cocktails on heart valve tissue.

A. Heart Valve Procurement And Preparation Heart valves found to be unsuitable for allograft transplantation were allocated for experimental use upon receipt of informed consent from the organ procurement organization responsible for mediating with the donor's family. The heart valves were dissected as described in U.S. Patent No. 4,890,457.

B. Sterilization Valve leaflets were cut in half, and each half was incubated in 4 ml of Dulbecco's Modified Eagle Medium (DMEM) (GIBCO) buffered with 25 mM 4- (2-hydroxyethyl) - 1-piperazine ethanesulfonic acid (HEPES) and containing 10% fetal bovine serum (hereinafter "Solution B") and antibiotics of the types and at the concentrations indicated below. The antibiotics included two combinations of antibacterial agents designated Cocktail B and Cocktail C having the following formulations :

Cocktail B:

62 μg/ml rifampin 57 μg/ml netilmicin 114 μg/ml vancomycin 136 μg/ml cefotaxime 136 μg/ml lincomycin

Cocktail C:

12 μg/ml imipenem 50 μg/ml vancomycin

In addition, two other cocktails were tested. These cocktails were the same as Cocktails B and C, except that they also contained 100 μg/ml fluconazole.

To prepare a stock solution of rifampin, 600 mg of sterile rifampin were dissolved in 10 ml of sterile

water. Rifampin was obtained from Merrell Dow (tradename RIFADIN) .

A solution of 100 mg/ml of netilmicin sulfate obtained from Schering-Plough (tradename NETROMYCIN) was used as the stock solution of this antibiotic.

A stock solution of vancomycin was made by dissolving 500 mg of sterile vancomycin hydrochloride in 10 ml of sterile water. Vancomycin hydrochloride was obtained from Eli Lilly & Co. (tradename VANCOCIN HCl) .

A cefotaxime stock solution was made by dissolving 2.0 grams of sterile sodium cefotaxime in 3 ml of sterile water. Sodium cefotaxime was obtained from Hoechst-Roussel (tradename CLAFORAN) .

A solution of 300 mg/ml of lincomycin hydrochloride obtained from Upjohn (tradename LINCOCIN) was used as the stock solution of that antibiotic.

A stock solution of imipenem was prepared by adding 20 ml sterile saline to a 250 mg bottle of imipenem obtained from Merck & Co. (tradename PRIMAXIN) .

Finally, a solution of 2 mg/ml of fluconazole obtained from Pfizer Roerig (tradename DIFLUCAN) was used as the stock solution.

An appropriate amount of each of the stock solutions was added to Solution B to give the final concentrations used in Cocktails B and C and Cocktails B and C plus fluconazole.

Experiment 1 : Paired halves of leaflets were treated with either Cocktail B or Cocktail B plus fluconazole. Similarly, paired halves of leaflets were treated with either Cocktail C or Cocktail C plus fluconazole. All of the valve leaflet halves were incubated for 48 hours at 37°C in the antibiotic cocktails and then cryopreserved.

Experiment #2 : Paired halves of leaflets were again used. One half was used as a fresh control and

immediately labelled with ³ H-glycine to determine leaflet fibroblast viability (see below) . The other half of each leaflet was incubated in Cocktail B or Cocktail B plus fluconazole. Similarly, one half of another pair was used as a fresh control and the other half was incubated in Cocktail C or Cocktail C plus fluconazole. The antibiotic-treated leaflets were incubated for 48 hours at 37°C in the antibiotic cocktails and then cryopreserved.

B. Crvopreservation After the treatments with the antibiotic cocktails were completed, the valve leaflet halves were packaged, cryopreserved and stored as described in U.S. Patent No. 4,890,457, with modifications to the freezing program so that individual valve leaflet halves could be preserved at a cooling rate of ~-l°C/minute in -1.6 ml of cryoprotectant solution. A 1.8 ml Nalge cryovial was used to contain the leaflet tissue. The freezing program takes the tissue from +4° to -80°C in approx 80 to 90 minutes.

C . Storage

The valve leaflet halves were stored in liquid nitrogen freezer for at least two days, but no more than three weeks, before being thawed.

D. Thawing

The valve leaflet halves were thawed by removing the valve tissue to be thawed from the liquid nitrogen freezer. As quickly as possible, the tissue was transferred to a 37°C waterbath. When all of the ice had dissolved (about 2 minutes) , the cryovial was transferred to a hood, and 4 ml of room temperature lactated Ringers plus 5% dextrose (D5LR) was dispensed into a 5-ml snap-cap tube (Falcon) for each cryovial being thawed. The exterior of the cryovial was rinsed with 70% isopropanol and any excess wiped away before opening. Using sterile forceps, the leaflet tissue was transferred into the tube containing the D5LR. The tissue was allowed to remain in D5LR for 5 minutes .

E. Viability Testing Tissue viability was assessed by the measurement of ³ H-glycine incorporation into proteins as follows. The thawed valve leaflet halves were placed in snap-cap tubes (Falcon) containing 16 μCi of tritiated-glycine (DuPont New England Nuclear) in 1 ml of DMEM supplemented with 15 μg/ml ascorbic acid. Radiolabeled glycine incorporation was determined after a 48 hour incubation at 37°C in an atmosphere consisting of 5% carbon dioxide in air. The tissue was then washed four times with phosphate-buffered saline, dehydrated with ethanol and ether, air dried and weighed. The tissue samples were then rehydrated with 200 μl of water and solubilized by the addition of 500 μl of 1 M NaOH. After incubation at 60°C for 1 hour, the samples were sonicated twice for 20-seconds. The homogenates were centrifuged at 12,000 x g, and 100 μl aliquots of the

supernatants were placed on glass fiber filter discs (Whatman No. 1822024; Whatman Chemical Separation Inc., Clifton, N.J.) . The filter discs were dried, and the proteins precipitated by addition of ice-cold 10% trichloroacetic acid for 30 minutes. This was followed by five ethanol rinses, two ether rinses, and drying. The discs were then placed in 10 ml of scintillation fluid (Cytoscint ES, ICN Biomedicals, Inc., Irvine, CA) . Tritium incorporation was measured by scintillation counting using a LS1701 (Beckman Instruments, Inc., Palo Alto, CA) .

In Experiment 1, viability was also assayed by labelling some of the leaflet halves with ³ H- hypoxanthine for autoradiography. To do so, a solution containing 15 μCi/ml of ³ H-hypoxanthine (DuPont/NEN) in serum free medium (DMEM) containing 5 mg/lOO ml gentamicin (Sigma) and 15 mg/100 ml ascorbic acid (Sigma) was prepared. Each leaflet half was placed in a 5 ml sterile snap-cap tube. Using a micropipettor, 0.75 ml of the ³ H-hypoxanthine solution was added to the tissue in the tube. The tubes were incubated at 37°C with 5% C0 ₂ for 48 hours. Then, the tritiated medium was decanted, and the tissue was washed twice with 3-4 ml of PBS (GIBCO) . Another 3-4 ml PBS was added, and the tubes were incubated for 30 minutes at room temperature before removing the liquid. Then, 3-4 ml of PBS was added to each tube, and the tubes were incubated at 4°C for 18 to 72 hours. The PBS was decanted, and the tissue transferred to a 7 ml glass scintillation vial. The vial was filled with 10% neutral buffered formalin (Fisher) , capped and stored at room temperature until the sample was analyzed by autoradiography performed as follows.

The tissue was soaked in 15% sucrose in 0.1 M sodium phosphate buffer, pH 7.4, for 2 hours to overnight. The solution was at room temperature at first and then lowered to 4°C. The tissue was dissected, embedded in a

plastic mold with OCT, and then frozen in liquid nitrogen. Frozen sections were cut on a cryostat at 6 microns and thaw mounted on chrome alum coated slides (slides dipped twice in a 37°C solution containing 3.0 g gelatin-bloom 275 and 0.3 g chromium potassium sulfate in 1000 ml distilled H ₂ 0) . The slides were dried in 37°C oven overnight to 48 hours, and the OCT was removed from the sections. The sections were then hydrated in ethanol from 95%, 75% down to 50%. The slides were left in 50% ethanol for 10 minutes, and then rinsed in distilled water for 30 sec. The sections were defatted by dehydrating in ethanol to xylene. The slides were hydrated in the same ethanol solutions down to double distilled water. The slides were placed in 37°C oven to dry (1 hr. to overnight) . Liquid Emulsion Kodak NTB2 was melted in a 40°C waterbath for 30 minutes and then diluted 1:1 with distilled water. The slides were slowly dipped individually into this diluted emulsion 2 times (6 seconds) and allowed to stand 1 hour or until dry. The boxes were sealed tight and stored in a desiccant in coldroom (4°C) until the slides were developed.

To develop the slides they were allowed to come to about 14°C while still sealed (about 11 minutes at room temperature) . This avoids condensation on the slides, which reduces the signals. All steps were carried out at a temperature of 14°C. The slides were placed in a glass slide rack and immediately developed for 2.5 minutes in D-19 developer (Kodak Catalog # 146 4593) . The slides were then rinsed in distilled water for 30 seconds, after which they were fixed in Kodak fixer (Polymax T fixer, Catalog # 829-5321) for 5 minutes. The slides were rinsed in running tap water at 14°C, and the temperature of the water was slowly raised to room temperature, 21°C, for 10 minutes. Finally, the slides were counterstained with Harris Hematoxylin (Sigma Cat # HHS-32) and Eosin Y, alcoholic solution (Sigma, Cat. # HT 110-1-32) .

The results are shown in the following tables. The incorporation of ³ H-glycine into proteins is expressed in disintegrations per minute (DPM) per mg of dry tissue. The autoradiography results are expressed as a percentage of live cells counted per total number of cells counted.

Experiment 1

Cocktail C Cocktail C Pius Fluconazole

Valve DPM/m£ Autorad DPM/mp Autorad

B1126 2088 78% 3232 79%

B1127 5996 71% 6322 62.5%

Cocktail B Cocktail B Plus Fluconazole

Valve DPM/mE Autorad DPM/me Autorad

B1128 1430 18% 1526 11%

B1129 2754 28% 2816 27%

Experiment 2 (DPM/mg)

Cocktail B Avg % of

Valve Leaflet Fresh Cock. B Plus Flue. Cocktail B

B1134 1 1847 1589 (86%)* 2 2112 1882 (89%) 3 1826 2244 (123%)* 129%

B1135 1 1830 2114 (116%) 2 1536 810 (53%) 45% 3 2255 1091 (48%)

B1137 1 2179 2674 (123%) 2 2776 4375 (158%) 3 1724 2811 (163%) 80%

B1138 1 1934 3320 (172%)

2 1449 3412 (235%) 85% 3 1499 2250 (150%)

B1139 1 9765 7596 (78%) 2 5084 5276 (104%) 3 10519 3786 (36%) 59%

B1140 1 2154 3804 (177%) 2 5530 5483 (99%) 106% 3 4105 2597 (63%)

Cocktail C Avg. %

Valve Leaflet Fresh Cock. C Plus Flue. Cock. C

B1141 1 2410 4955 (206%) 2 2287 4714 (206%) 95% 3 3762 4661 (124%)

B1144 1 2736 4355 (159%)

2 5550 5128 (92%) 112% 3 4752 4611 (97%)

B1147 1 1526 2116 (139%)

2 1819 3959 (217%) 148% 3 1582 2287 (145%)

B1148 1 4448 9922 (223%) 2 3144 8075 (257%) 3 3708 8290 (224%) 92%

B1150 1 2484 5085 (205%)

2 2796 4999 (179%) 3 3139 7109 (226%) 141%

B1151 1 3938 3868 (98%)

2 5279 7160 (136%) 3 4155 6055 (146%) 110%

* Percent of fresh

Experiment 1 - The results indicate that t reatmen with Cocktail C containing fluconazole produced valve tissue which was at least as viable as that after treatment with Cocktail C without fluconazole. This is

seen in the scintillation results and the auto¬ radiography. Cocktail B containing fluconazole is also virtually equivalent to Cocktail B without fluconazole. With Cocktail B, however, the tissue is less viable to begin with.

Experiment 2 - On average the Cocktail B containing fluconazole gave results which were 84% of those obtained with Cocktail B not containing fluconazole. On average Cocktail C containing fluconazole gave results which were 116% of Cocktail C without fluconazole. When any of the antibiotic treatments is compared to the fresh tissue values, they are all higher than the fresh tissue.

From the results of Experiments 1 and 2, it can be concluded that 100 μg/ml fluconazole added to Cocktail B or C is not harmful to the tissue.

EXAMPLE 2

Cocktails containing the following antibiotics in the following concentrations were prepared and tested as described in Example 1.

Imipenem 12μg/ml Vancomycin 50μg/ml

Fluconazole lOOμg/ml

Rifampin 0, 10, 30 or 100 μg/ml.

Paired valve leaflet halves were used. One half of each leaflet was a control. It was incubated in the antibiotic cocktail which did not contain rifampin. The other half was treated with one of the antibiotic cocktails containing rifampin. The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg and percent of control .

Valve Leaflet Contro lO g/ml % 30μgml % lOOμg/ml % 1

B1154 1 2139 2053 96

2 2194 2065 94 3 5176 1673 32

B1156 1 4371 4816 110 2 9792 9238 94 3 8748 3583 41

B1157 1 6924 6304 91 2 4775 4685 98 3 3590 2575 72

B1159 1 4786 3566 75 2 6316 5774 91 3 5705 2891 51

B1161 1 4091 3002 73

2 5310 2034 38 3 3876 1170 30

B1162 1 3232 3690 114

2 2620 1693 65 3 2349 547 23

AVG 93 80 42%

% %

EXAMPLE 3

Antibiotic cocktails were prepared and tested as described in Example 1. Paired valve leaflet halves were used. One half of each leaflet was used as a control and treated with the following antibiotic cocktail :

Imipenem 12 μg/ml

Vancomycin 50μg/ml

Fluconazole lOOμg/ml.

The other half was treated with one of the

following test antibiotic cocktails:

Imipenem 12, 24 or 48 μg/ml Vancomycin 50μg/ml Fluconazole lOOμg/ml Rifampin 10 μg/ml.

The results of the assay for incorporation of ³ H- glycine into proteins are presented in the following table in DPM/mg and percent of control. The columns labeled 12, 24 and 48 μg indicate the amount of imipenem per ml used in the test (non-control) cocktails.

VALVE CONTRO 12μg % 24μg % 48μg % L CONTRL CONTR CONTRL L

B1163 3347 4516 135% 4927 3200 65% 2546 3602 141%

B1164 3557 3601 101% 3486 2169 62% 5031 4780 95%

B1165 3461 3789 109% 4014 2260 56% 3768 2944 78%

B1166 3569 1553 44% 3215 6539 203% 4012 6568 164%

B1168 2891 1565 54% 2193 1617 74% 2084 1316 63%

B1169 5202 3903 75% 3343 2775 83% 5030 4208 84%

VALVE CONTRO 12μg % 24μg % 48μg % j L CONTRL CONTR CONTRL L

AVG 86% 90% 104% 1

EXAMPLE 4

Antibiotic cocktails containing the following antibiotics in the following concentrations were prepared and tested as described in Example 1.

Imipenem 12, 24, 48 or 96 μg/ml Vancomycin 50μg/ml Fluconazole lOOμg/ml

Paired valve leaflet halves were used. One half of each leaflet was used as a control and was treated with the antibiotic cocktail containing 12 μg/ml imipenem. The other half was treated with one of the test antibiotic cocktails containing 24, 48 or 96 μg/ml imipenem. The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg and percent of control. The columns labeled 24, 48 and 96 μg indicate the amount of imipenem per ml used in the test antibiotic cocktails.

VALVE CONTRO 24μg % CONTRL 48 ιg % CONTRL 96μg % L CONTRL

B1178 1700 4458 262% 3267 1944 60% 2944 1907 65%

B1180 7227 9336 129%

11035 11321 103%

9493 12004 126%

B1181 3806 4802 126% 7091 4942 70% 3651 3322 91%

B1182 6375 6372 100% 6050 8826 146% 9135 6524 71%

B1184 6621 8093 122% 6219 9415 151% 7324 9272 127%

B1185 8312 5054 95% 6444 6755 105% 6422 5851 91%

AVERAG 139% 106% 90% E

EXAMPLE 5

Antibiotic cocktails containing the following antibiotics in the following concentrations were prepared and tested as described in Example 1.

Imipenem 12μg/ml Vancomycin 50μg/ml Fluconazole lOOμg/ml Rifampin lOμg/ml. Amphotericin B 0, 0.3, 1.0 or 3.0 μg/ml.

The amphotericin B dilutions were made as follows. A bottle containing 50 mg amphotericin B (E.R. Squibb &

Sons, Inc., Princeton, NJ; trade name FUNGIZONE) was reconstituted with 10 ml of sterile water. One ml of this solution was removed and further diluted with 9 ml of Solution B to give a concentration of 500 μg/ml. Further dilutions were made from this stock solution in Solution B.

Paired valve leaflet halves were used. One half of each leaflet was a control and was treated with the antibiotic cocktail without any amphotericin B. The other half was treated with one of the test antibiotic cocktails containing amphotericin B. The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg and percent of control. In the table, the concentrations refer to the concentration of amphotericin B used in the test antibiotic cocktails.

VALVE CONTRO 3μg/ml % lμg/ml % 0.3μg/ml % L CONTR CONTR CONTRL L L

B1170 6026 5799 96% 4401 3397 77% 7638 5281 69%

B1171 2428 470 19% 1666 619 37%

873 1692 194%

B1173 6729 5390 80% 8068 4619 57% 7206 8148 113%

B1174 5894 4694 80% 4841 4616 95% 4040 6724 166%

B1175 12453 7584 61%

6666 6130 92%

6482 6704 103%

B1177 6281 6182 98% 3682 3294 89% 2159 4074 189%

AVERAGE 72% 74% 139%

A concentration of 0.3 μg/ml amphotericin B in the antibiotic mixture yielded tissue which has an average

viability greater than the control. Tissue viability was reduced by 1 and 3 μg/ml of amphotericin B to 74% and 72% of control, respectively.

EXAMPLE 6

Antibiotic cocktails containing the following antibiotics in the following concentrations were prepared and tested as described in Examples l and 5, except that the valve leaflet halves were only incubated in the antibiotic cocktails for 24 hours, instead of 48 hours.

Imipenem 48μg/ml Vancomycin 50μg/ml Fluconazole lOOμg/ml Rifampin 0, 10, 30 or 60 μg/ml Amphotericin B 0.3 μg/ml.

Paired valve leaflet halves were used. One half of each leaflet was a control and was treated with the antibiotic cocktail without any rifampin. The other half was treated with one of the test antibiotic cocktails containing rifampin. The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg and percent of control. In the table, the concentrations refer to the concentration of rifampin used in the test antibiotic cocktails.

VALVE CONTROL lOμg/ml % 30μg/ml % 60μg/ml % CONTRL CONTRL CONTRL

B1199 4098 2552 62% 58% 2311 1348 2846 2099 74%

B1200 1974 3512 178% 1216 2279 187% 2286 630 28%

B1202 2238 2704 121% 2365 2310 98% 3470 1805 52%

B1203 3540 3001 85% 4239 4241 100% 4035 2643 66%

B1204 3709 2640 64% 3368 2025 56%

B1207 5970 4773 80% 7433 3980 54% 6174 3829 62%

B1209 6029 3597 60% 2752 5750 209% 3875 1487 38%

B1210 11545 5436 47% 11317 7136 63% 7415 6281 85%

B1211 5637 3180 56% 6182 3336 54% 7903 677 9%

B1213 5724 4597 80% 8390 4418 53% 4721 3879 82%

AVERAGE 85% 94% 55%

EXAMPLE 7

An experiment was performed to compare ³ H-glycine incorporation by fresh tissue with ³ H-glycine incorporation by tissue processed in various ways. All procedures were the same as those described in Examples 1 and 5, except as otherwise indicated below.

The three leaflets were removed from each of eight valves, and each leaflet was cut in half. The six leaflet halves were treated as follows:

Leaflet IA Fresh control, no antibiotic treatment, no cryopreservation. Leaflet IB Incubated with Solution B for 24 hours at

37°C and cryopreserved. Leaflet 2A Fresh control, no antibiotic treatment, no cryopreservation.

Leaflet 2B Incubated with antibiotic cocktail at 37°C for 24 hours and cryopreserved.

Leaflet 3A Incubated with Solution B for 24 hours at 37°C and cryopreserved.

Leaflet 3B Incubated with antibiotic cocktail at 37°C for 24 hours and cryopreserved.

All fresh control valve leaflet halves were immediately labeled with ³ H-glycine and processed for scintillation counting. After thawing, the Solution B- treated and antibiotic-treated valve leaflet halves were also labeled with ³ H-glycine and processed for scintillation counting.

The antibiotic cocktail consisted of:

Imipenem 48 μg/ml

Vancomycin 50 μg/ml

Fluconazole 100 μg/ml

Amphotericin B 0.3 μg/ml

The results are presented in the following table. The numbers are DPM/mg.

VALVE LEAFLE17 FRESH SOL. B ANTIBIOTICS

B1222 1 5097 7110

2 4155 9311

3 6610 6439

B1223 1 4684 9400

2 5582 8131

3 8364 6841

B1226 1 3670 8153

2 3987 11425

3 7817 7757

B1229 1 6107 8177

2 8470 9683

3 7945 6297

B1230 1 6067 7412

2 4098 6935

3 4258 5523

B1232 1 3022 5474

2 2534 7574

3 6138 4051

B1233 1 2451 4436

2 3701 10861

3 6364 5402

B1234 1 5666 2870

2 5006 3476

3 4185 3233

MEAN 4643 6545 7059

STD DEV 1544 1852 2450

SEM 386 463 612

There is a statistically significant difference between the fresh control group and the two groups incubated at 37°C before cryopreservation. There is no statistical difference between the Solution B-treated and the antibiotic-treated groups. Statistical analysis was performed using the Student- Newman-Keuls method.

EXAMPLE 8

Antibiotic cocktails containing the following antibiotics in the following concentrations were prepared and tested as described in Examples 1 and 5.

Imipenem 48, 72 or 96 μg/ml

Vancomycin 50 μg/ml

Fluconazole 100 μg/ml

Amphotericin B 0.3 μg/ml.

The three leaflets from each valve were cut in half. One half of each leaflet was incubated with an antibiotic cocktail containing a particular concentration of imipenem for 24 hours, and the other half was incubated with the same antibiotic cocktail for 48 hours as follows:

Leaflet IA - 48 μg/ml imipenem, 24 hours

IB - 48 μg/ml imipenem, 48 hours

2A - 72 μg/ml imipenem, 24 hours

2B - 72 μg/ml imipenem, 48 hours

3A - 96 μg/ml imipenem, 24 hours

3B - 96 μg/ml imipenem, 48 hours

The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg. In the table, the concentrations refer to the concentration of imipenem used in the antibiotic cocktails.

Valve 48μg/ml 72μg/ml 96μg/ml

24 h 48 h 24 h 48 h 24 h 48 h

B1245 7053 5875 9219 9933 4977 6236

B1246 5526 5758 7670 6128 5375 5805

B1247 3252 6582 6713 5808 7105 8708

B1256 2643 3459 2796 2876 2677 2602

B1257 3188 4685 5661 5724 5982 1792

B1258 6322 4947 6503 7195 7025 12875

B1259 9798 7731 14088 14544 7409 12861

B1260 3700 3584 3214 5586 5053 4185

B1266 3671 4887 4022 5273 4016 4551

B1267 7535 5655 9086 8917 9233 5647

MEAN 5269 5316 6897 7198 5885 6526

SEM 748 411 1069 1026 595 1219

Statistical analysis of this data by the student's T-test shows that there is no statistical difference between any of the treatment groups.

EXAMPLE 9

Antibiotic cocktails containing the following antibiotics in the following concentrations were prepared and tested as described in Examples 1 and 5, except that the valve leaflet halves were incubated in the antibiotic cocktails for 24 hours only.

Imipenem 96 μg/ml, Vancomycin 50 μg/ml, Fluconazole 100 μg/ml,

and one of the following:

0.3 μg/ml amphotericin B Control 0.6 μg/ml amphotericin B I 1.0 μg/ml amphotericin B II 0.6 μg/ml amphotericin B plus

50 μg/ml netilmicin III

The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table (units in DPM/mg) . Note that valves B1307, B1310, and B1311 had an incorrect concentration of netilmicin added in the cocktail. These data were therefore not included.

VALVE CONTROL I II III B1307 4438 7264 6370 5770

B1310 1773 2522 2107 4091

B1311 2712 1617 1560 1495

B1313 3940 3790 4644 3058 2841 1598

B1315 3678 3588 4727 4941 3570 4476

B1316 7255 9624 8867 10954 5510 7924

B1317 4039 4261 7180 5096 2703 3240

B1318 5461 7025 7704 9013 7838 7346

B1319 3178 3435 2983 2963 3190 2847

B1320 9820 8535

5527 5256 7530 7866 B1321 4528 2870 5439 4486 3872 3763 B1322 3815 3544 3511 2269 4205 3249 B1324 6268 7065 4692 4438 7847 6492 B1325 4839 5273 4516 4462 5922 6478

Stat ist ical analys is : Mean Std . Dev SEM

Control 4887 1997 320

0 . 6 Amp B 5030 2451 655

1 . 0 Amp B 4878 2502 669

. 6 Amp B + Neti l 5025 2255 680

Using the Kruskal-Wallis one way anova on ranks, there is no statistical difference between the groups (P = 0.994) . Therefore, the antibiotic regimens tested provide tissue which is equally viable.

EXAMPLE 10

Antibiotic cocktails were prepared and tested as described in Examples 1 and 5, except that the valve leaflet halves were incubated in the antibiotic cocktails for 24 hours only. Paired valve leaflet halves were used. One half of each leaflet was used as a control and treated with the following antibiotic cocktail :

Imipenem 96 μg/ml

Vancomycin 50μg/ml

Fluconazole lOOμg/ml

Amphotericin B 0.6 μg/ml.

The other half was treated with one of the following test antibiotic cocktails:

Imipenem 96 μg/ml

Vancomycin 50μg/ml

Fluconazole lOOμg/ml

Amphotericin B 0.6 μg/ml

Netilmicin 20, 30 or 50 μg/ml.

The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg and percent of control . The columns labeled 20, 30 and 50 μg/ml indicate the concentration of netilmicin used in the test (non-control) cocktails.

VALV CONTRO 20μg ml % contr 30μg/ml % contr 50μ.g/ml % E L contr

B1326 6762 4218 62% 5259 4944 94% 3701 4309 116%

B1327 4433 4506 102% 7519 9040 120% 6210 6714 108%

B1328 6504 4648 71% 6862 7312 107% 4225 3514 83%

B1329 4233 4193 99% 6058 6569 108% 5331 5188 97%

B1331 3646 3530 97% 3682 5611 152% 2988 4516 151%

B1332 3910 5831 149% 6968 6736 97% 7404 6046 82%

B1333 1327 1929 145% 1160 2660 229% 472 433 91%

B1334 7628 4994 65% 9000 10401 116% 8553 11484 134%

B1337 6624 5048 76% 6290 6440 102% 6766 6349 94%

B1338 2843 3233 114% 9114 8342 92% 4908 4218 86%

B1340 values too low

AVG 5514 4213 98% 6805 122% 5815 104%

EXAMPLE 11

Antibiotic cocktails were prepared and tested as described in Examples 1 and 5, except that the valve leaflet halves were incubated in the antibiotic cocktails for 24 hours only. Paired valve leaflet halves were used. One half of each leaflet was used as a control and treated with the following antibiotic cocktail :

Imipenem 96 μg/ml Vancomycin 50μg/ml Fluconazole lOOμg/ml Netilmicin 50 μg/ml.

The other half was treated with one of the following test antibiotic cocktails:

Imipenem 96 μg/ml Vancomycin 50μg/ml Fluconazole lOOμg/ml Netilmicin 50 μg/ml

Amphotericin B 0.3, 0.6, 1.0, 2.0 or 3.0 μg/ml

The results of the assay for incorporation of ³ H-glycine into proteins are presented in the following table in DPM/mg. The results for the ³ H-hypoxanthine autoradiography are presented in the second table as percent live cells compared to total number of cells. The columns labeled 0.3, 0.6, 1.0, 2.0 or 3.0 μg/ml indicate the concentration of amphotericin B used in the test (non-control) cocktails.

VALVE CONTROL 0.3μg/ml 0.6 .g/ml l.Oμg/ml 2.0 _/ xg/ml 3.0μg/ml

B1341 3658 4388 4643 4937 4248 2569

B1342 6198 4655 2795 3218

B1343 4293 3368 4450 4777 2744 2937

B1347 4302 4938 4238 5217 3744 3443

B1348 4488 5632 5979 5136 5804 3653

B1349 8769 8540 7646 4377 8094 6033

B1351 5806 5413 7156 5833 3498 5743

B1352 7242 5254 3777 4888 3921 5177

B1353 3740 4147 3291 3657 2964 3212

B1355 1813 1064 1565 2083 1923 2028

B1358 4974 6902 4854 5284 6209 7139

B1360 3388 3319 2796 2448 2945 2867

B1361 10903 7264 6773 7783 6793 9171

B1363 7340 4628 7321 7463 3620 5134

MEAN 5494 4965 4960 4914 4236 4452

SEM 645 493 526 452 480 543 avg % 90% 90% 89% 77% 81% control

VALVE CONTROL 0.3μg/ml 0.6μg/ml l.Oμg/ml 2.0μg/ml 3.0μgml

B1350 76.5% 75.7% 67.5% 69% 70.5% 76.5%

B1354 82% 61.5% 82.8% 76.3% 68.3% 75.3%

B1356 81% 86% 84% 64% 68% 55.5%

B1359 91% 58% 66% 59% 71% 70%

MEAN 82.6% 70.3% 75.1% 67.1% 69.5% 69.3% avg % 85% 91% 81% 84% 84% control

EXAMPLE 12

An experiment was performed to determine the viability of valve leaflets processed using Cocktail C, Cocktail D or Cocktail B plus Cocktail C. See Example 1 for the formulations of Cocktails B and C. Cocktail D has the following formulation:

Imipenem 96 μg/ml

Vancomycin 50 μg/ml

Fluconazole 100 μg/ml

Amphotericin B 0.3 μg/ml.

Unless otherwise specified, the cocktails were prepared and tested as described in Examples 1 and 5.

Eleven human valves were used. One leaflet from each valve was incubated with Cocktail D for 24 hours (Method G) . Another leaflet was incubated in Cocktail C for 24 hours (Method D) or in Cocktail C for 24 hours followed by incubation in Cocktail B for 24 hours (Method E) . After thawing, each leaflet was cut in half. One half was labelled with ³ H-glycine, and the other half was labelled with ³ H-hypoxanthine. The results for the incorporation of ³ H-glycine into protein are presented in the following table (units of DPM/mg) .

VALVE Method Method Method G

D/E

B1272 D 2008 1576

B1274 D 1484 3858

B1276 D 2527 2110

B1277 D 3578 8432

MEAN 2399 3994

SD 894 3115

B1273 E 742 2012

B1275 E 4449 6267

B1278 E 2317 2383

B1279 E 388 410

B1280 E 842 1047

B1281 E 589 2122

B1282 E 470 1326

MEAN 1400 2224

SD 1496 1911

Thus, treatment with Cocktail D containing antifungal agents fluconazole and amphotericin B (Method G) resulted in viability of valve leaflets comparable to or better than treatment with Cocktails B and C not containing antifungal agents (Methods D and E) .

EXAMPLE 13

An experiment was performed to compare the viability of heart valve tissue processed in one of the following ways:

Method F: Tissue with longer ischemic time (potentially lower cell viability) which was incubated with Cocktail C for 24 hours, optionally followed by

incubation for 24 hours with Cocktail B (formulations given in Example 1) . The decision to treat tissue with Cocktail B was made based on the results of microbiological cultures of tissue taken prior to antibiotic treatment with Cocktail C as described in McNally and Brockbank, J. Med. Engineer. & Technol.. 16, 34-38 (1992) and PCT application WO 92/12631.

Method H: Incubation with the following antibiotic cocktail, which will be designated Cocktail E, for 24 hours:

Imipenem 94-100 μg/ml

Vancomycin 50 μg/ml

Netilmicin 50 μg/ml

Fluconazole 100 μg/ml

Amphotericin B 1.0 μg/ml

Cocktails B, C and E were prepared and tested as described in Examples 1 and 5, except as otherwise provided herein.

One leaflet was removed from each valve at dissection and cut in half. One half was immediately labeled with ³ H-glycine, and the other half was immediately labeled with ³ H-hypoxanthine. These were the fresh controls.

The remainder of the valve was cut in half with one whole leaflet left intact in each half. One of the valve halves was treated by Method F (Cocktail C, optionally followed by treatment with Cocktail B) . The other valve half was treated by Method H (Cocktail E) . The frozen valve halves were thawed and cut in half. One half was labeled with ³ H-glycine, and one half was labeled ³ H-hypoxanthine.

The results are presented in the following table (units of DPM/mg for ³ H-glycine incorporation into

proteins and units of percent live cells compared to total cells for ³ H-hypoxanthine autoradiography) .

Comparisons between treatment groups were made with the nonpara etric Mann-Whitney rank sum test . When ³ H-glycine was used to label the tissue there was no significant difference between the treatments (P=0.655) . Likewise, when ³ H-hypoxanthine was used to label the tissue there was no significant difference between the two treated groups (P=0.684) . Using the Kruskal-Wallis one way anova on ranks there was no significant difference between any of the 3 groups when ³ H-glycine was used.

VALVE GLYCINE HYPOXAN GLYCINE HYPOXAN GLYCINE HYPOXAN FRESH FRESH METHOD F METHOD F METHOD H METHOD

H

B1385 10557 70% 2137 43% 4802 52.6%

B1386 4120 75.8% 2719 32% 6742 61%

B1387 2493 72.3% 2385 41.5% 3645 41.4%

B1388 3366 41.2% 3547 31.2% 946 10.3%

B1389 6569 60.6% 2066 30.6% 1111 18.4%

B1390 2474 48.2% 3253 26.8% 5848 22.8%

B1391 2451 86.5% 1420 26.3% 2942 52%

B1392a 1236 52.2% 658 12.3% 381 5.8%

B1392b ND* ND 3196 39.8% 5937 38.8%

B1393 5774 78.2% 2798 31.4% 4286 44.2%

B1394 3544 63.2% 5840 42.4% 6798 49.6%

B1395 4676 84.4% 18298 55.3% 13986 37.2%

B1396 4792 60.9% 2778 27.4% 2119 32.2%

B1397 4524 85.6% 3639 39.8% 3825 30.8%

B1399 4209 75.9% 1698 10.9% 1819 22.9%

B1400 3179 56% 1533 29.8% 9316 16.4%

B1401 3181 33.3% 2402 30.2% 2317 13.2%

B1402 1962 88.4% 1716 18.5% 2889 39.4%

B1403 1167 36.4% 1730 54.7% 1289 51.0%

B1406 2504 43.6% 9662 43.6% 2978 21.8%

B1407 3160 82.9% 3816 28.3% 2884 28.2%

B1408 11194 79.4% 11264 40.0% 7406 70.9%

B1409 896 81.6% 4665 46.4% 1822** 59.6%

B1410 2044 68.6% 7453 51.3% 3804 44.4%

B1412 4248 21.8% 2038 29.0% 1652 50.0%

MEAN 4062 64.5% 4085 34.5% 4155 36.6%

SD 2535 19.1% 3991 11.6% 3104 17.3%

SEM 529 3.9% 815 2.3% 634 3.5%

* ND = not done

** very heavy plaque

EXAMPLE 14

The Candida albicans used in these studies was ATCC strain 14053 (American Type Culture Collection) . The purity of the strain was verified before and after cultivation.

The test organism was cultured on a Sabouraud Agar Plate, streaked for isolation, and incubated at 35°C for 24 hours. A single colony was removed using a sterile cotton swab and placed in 5.0 ml of sterile saline. The mixture was vortexed, and dilutions were made and compared to a standard turbidity chart until the turbidity matched a 0.5 turbidity standard (a 0.5 MacFarland turbidity standard suspension) . Serial dilutions were made as follows: (1) 0.5 ml of the 0.5 MacFarland turbidity standard was added to 4.5ml of sterile normal saline (Tube A; expected population 10 ⁵ cfu/ml) ; (2) 0.1 ml of the Tube A suspension was added to 9.9 ml of sterile normal saline (Tube B; expected population 10 ³ cfu/ml) ; and (3) 0.1 ml of the Tube B suspension was added to 9. ml of sterile normal saline (Tube C; expected population 10 cfu/ml) . Each serial dilution was vortexed for 30 seconds before the next transfer.

Solutions of fluconazole and amphotericin B having the concentrations given in the following table were made in Solution B.

Cocktail Fluconazole Amphotericin B

Control 0 0

I lOOμg/ml 0

II lOOμg/ml 0.3μg/ml

III 0 0.6μg/ml

Tubes B and C were challenged with these cocktails. Each cocktail (24.0 ml) was placed into a sterile 115 ml plastic cup, and 1.0 ml of the yeast population from Tube B or C was added. The cups were mixed gently, capped and placed in the incubator for 24 hours at

35°C. After the prescribed incubation period, 1.0 ml of the yeast-cocktail mixture was removed from each tube and processed to obtain an enumeration of the surviving yeast organisms. The remaining mixture was placed back into the incubator and incubated for an additional 24 hours. After this incubation period, the cups were removed, swirled gently and 1.0 ml was removed, and the above process was repeated.

Enumeration of the yeast was performed as follows; each sample was tested in duplicate. One milliliter (1.0 ml) of yeat suspension was added to the top assembly of a sterile disposable filter apparatus which contained 99 ml of sterile normal saline. The contents were mixed gently and filtered. The filter pad was removed aseptically using sterile disposable forceps and placed on Sabouraud Agar plates. The plates were incubated at 35°C for 24 hours, and the colonies were then counted.

The results are presented in the following tables. Tube B contained 734 colony forming units (CFU) of C. alj icans/ml at time zero, and Tube C contained 13 cfu/ml at time zero.

Tube B 24hr 48hr

I 31 colonies 45 colonies

II 0 colonies 0

III 0 colonies 0

Tube C 24hr 48hr

I 0 colonies 1 colony

II 0 colonies 0 colonies

III 0 colonies 0 colonies

The results for Tube B show that 100 μg/ml fluconazole was effective in reducing the C. albicans bioburden from 734 colonies to 31 colonies at 24 hours, At 48 hours the colony count (45) suggests a tolerance or static phenomenon. Amphotericin B at 0.6 μg/ml and the combination of 100 μg/ml fluconazole and 0.3 μg/ml amphotericin B were fungicidal .

EXAMPLE 15

The following antibiotic cocktails were prepared as described in Examples 1 and 5 and tested against ten yeast isolates :

Cocktail Content

I Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Fluconazole - 100 μg/ml Amphotericin B - 0.3 μg/ml

II Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Fluconazole - 100 μg/ml Amphotericin B - 0.6 μg/ml

III Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Fluconazole - 100 μg/ml Amphotericin B - 1.0 μg/ml

IV Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Amphotericin B - 0.3 μg/ml

V Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Amphotericin B - 0.6 μg/ml

VI Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Amphotericin B - 1.0 μg/ml

The ten yeast isolates used were chosen because they are considered to be clinically significant human yeast isolates. The ten yeast isolates tested were:

1. Rhodotorula rubra*

2. Candida lusitaniae*

3. Saccharomyces cervisae*

4. Cryptococcus laurentii*

5. Candida guilliermondii*

6. Torulopsis glabrata*

7. Candida tropicalis*

8. Candida albicans®

9. Trichosporon beigeli*

10. Candida parasilosis*

* Provided by the American College of Pathologists .

@ The Candida albican isolate was obtained from donor tissue in which the yeast isolate had survived an antibacterial cocktail incubation. Identification was confirmed by using the Germ Tube Test for the confirmation of Candida albicans . The Germ Tube Test involves making a suspension of an isolate in a commercially prepared solution of bovine serum (REMEL) . The mixture is incubated for 2-4 hour at 35°C and examined microscopically. Confirmation of a positive germ tube is the formation of pseudohyphae budding (germination) off of the yeast bud cell.

The antibiotic coctails were challenged with ten individual predetermined bioburden yeast suspensions as described in the previous example . The Sabouraud Agar plates were incubated at 35°C for 5 days before enumeration and checked again at day 7 for any additional growth of yeast colonies. The 5 to 7 day incubation/enumeration period was chosen to provide the optimum recovery of any surviving yeast isolates which

may grow after recovery from injury due to the presence of antifungal agents.

The results are presented in the following tables. Cocktail III exhibited the greatest fungicidal activity on all yeast isolates.

RESULTS WITH YEAST ISOLATES AT 24 HOURS (CFU/ml)

Yeast Initial Positive Cock Cock Cock Cock Cock Cock isolate Bio¬ Control tail tail tail tail tail tail burden I II III IV V VI

Rhodo. 1.2 x 6.2 x rubra 10* IO ⁴ 460 106 10 469 140 10

Candida 4.3 x 1.7 x luistan 10' IO ⁵ 440 70 59 801 736 233

Sacchar 1.8 x 9.4 x cervis 10 ⁴ IO ⁴ 530 151 11 155 15 12

Crypto. NG N/A N/A N/A N/A N/A N/A N/A lauren

Candida 5.8 x 2.1 x guillie 10* 10 ^s 401 9 10 300 70 0

Torulop 3.7 x 2.0 x glabra. IO ⁴ 10 ⁶ 20 0 0 10 0 0

Candida 3.1 x 4.7 x tropica IO ⁴ IO ⁶ 860 337 40 610 301 110

Candida 2.0 x 3.9 x albican 10 ⁴ 10' 263 22 10 120 10 10 ^'

Tricho 1.1 x 4.1 x beigeii 10 ⁴ IO ⁴ 0 0 0 810 167 154

Candida 2.8 x 8.7 x parapsi 10 ⁴ IO ⁴ 20 0 510 380 170

13

RESULTS WITH YEAST ISOLATES AT 48 HOURS (CFU/ml)

Yeast Initial Positive Cock Cock Cock Cock Cock Cock isolate Bio¬ control tail tail tail tail tail tail burden I II III IV V VI

Rhode 1.2 x 8.6 x rubra IO ⁴ IO ⁵ 706 2150 388 4108 2100 660

Candida 4.3 x 1.1 x luistan IO ⁴ 10' 1180 1076 245 1978 1190 1011

Sacchar 1.8 x 1.2 x cervis IO ⁴ 10 ^s 150 40 16 5010 87 0

Crypto. NG N/A N/A N/A N/A N/A N/A N/A lauren

Candida 5.8 x 1.3 x guillie IO ⁴ 10" 860 466 11 1987 454 71

Torulop 3.7 x 2.0 x glabra. IO ⁴ 10' 20 0 0 560 0 0

Candida 3.1 x 1.7 x tropica IO ⁴ 10' 2130 1208 528 1817 869 105

Candida 2.0 x 1.2 x albican IO ⁴ 10 ⁷ 370 26 13 3640 2019 49

Tricho 1.1 x 2.3 x beigeii IO ⁴ 10 ^s 0 0 0 9466 2019 610

Candida 2.8 x 3.6 x par psi IO ⁴ 10 ⁷ 56 15 1006 380 177

33

Comparison of Cocktail I (0.3 μg Amphotericin B) after 24 and 48 hours

A comparison of Cocktail I after 24 hours with Cocktail I after 48 hours indicates the usual tolerance phenomenon for yeast .

Comparison of Cocktails I and IV

(0.3 μg amphotericin B with and without fluconazole) after 24 hours

YEAST ISOLATES COCKTAIL COCKTAIL Net I IV result

Rhodoturla 460 469 0 rubra cfu/ml cfu/ml

Candida 440 801 -1 lusitaniae cfu/ml cfu/ml

Saccharomyces 530 155 +1 cervis cfu/ml cfu/ml

Cryptococcus NG NG NA laurenti

Candida 401 300 0 guilliermondii cfu/ml cfu/ml

Troulopsis 20 10 0 glabrata cfu/ml cfu/ml

Candida 860 610 0 tropicalis cfu/ml cfu/ml

Candida 263 120 0 albicans cfu/ml cfu/ml

Trichosporon 0 810 -1 beigeii cfu/ml cfu/ml

Candida 20 510 -1 parasilosis cfu/ml cfu/ml

Cocktail IV without fluconazole was not as effective as Cocktail I with fluconazole after 24 hours. The data indicate that the two antifungal agents give synergistic results when used together. See also the previous example

Comparison of Cocktails II and V

;o.6 μg amphotericin B with and without fluconazole) after 24 hours

YEAST ISOLATES COCKTAIL COCKTAIL Net II V result

Rhodoturla 106 140 0 rubra cfu/ml cfu/ml

Candida 70 736 -1 lusitaniae cfu/ml cfu/ml

Saccharomyces 151 15 +1 cervis cfu/ml cfu/ml

Cryptococcus NG NG NA laurenti

Candida 9 70 0 guilliermondii cfu/ml cfu/ml

Troulopsis 0 0 0 glabrata cfu/ml cfu/ml

Candida 337 " 301 0 tropicalis cfu/ml cfu/ml

Candida 22 10 0 albicans cfu/ml cfu/ml

Trichosporon 0 167 -1 beigeii cfu/ml cfu/ml

Candida 13 380 -1 parasilosis cfu/ml cfu/ml

Cocktail V without fluconazole was not as effective as Cocktail II with fluconazole after 24 hours. The data indicate that the two antifungal agents give synergistic results when used together. See also the previous example,

Comparison of Cocktails III and VI

(1.0 μg Amphotericin B with and without fluconazole) after 24 hours

YEAST ISOLATES COCKTAIL COCKTAIL Net

III VI result

Rhodoturla 10 10 0 rubra cfu/ml cfu/ml

Candida 59 233 -1 lusitaniae cfu/ml cfu/ml

Saccharomyces 11 12 0 cervis cfu/ml cfu/ml

Cryptococcus NG NG NA laurenti

Candida 11 0 0 guilliermondii cfu/ml cfu/ml

Troulopsis 0 0 0 glabrata cfu/ml cfu/ml

Candida 40 110 0 tropicalis cfu/ml cfu/ml

Candida 10 10 0 albicans cfu/ml cfu/ml

Trichosporon 0 154 -1 beigeii cfu/ml cfu/ml

Candida 0 170 -1 parasilosis cfu/ml cfu/ml

Cocktail VI without fluconazole was not as effective as Cocktail III with fluconazole after 24 hours. The data indicate that the two antifungal agents gave synergistic results when used together. See also the previous example,

EXAMPLE 16

Example 15 was repeated using the following antibiotic cocktails :

Cocktail Content

I Imipenem - 50 μg/ml

Vancomycin - 50 μg/ml Fluconazole - 100 μg/ml Amphotericin B - 0.3 μg/ml

VII Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Netilmicin - 50 μg/ml Fluconazole - 50 μg/ml Amphotericin B - 1.0 μg/ml

VIII Imipenem - 50 μg/ml Vancomycin - 50 μg/ml Netilmicin - 50 μg/ml Fluconazole - 100 μg/ml Amphotericin B - 1.0 μg/ml

Ten different yeast isolates were used. These organisms were chosen because they had been recovered from cardiovascular donor tissue. They were:

11. Candida albicans

12. Candida albicans

13. Candida albicans

14. Candida albicans

15. Candida albicans

16. Yeast, not Candida albicans

17. Yeast, not Candida albicans

18. Candida albicans

19. Yeast, not Candida albicans

20. Yeast, not Candida albicans

Identification was made using the Germ Tube Test. The results after 24 hours of incubation in the antibiotic

cocktails are presented in the following table (units are CFU/ml) .

Ten yeast Initial Control Cocktail Cocktail Cocktail Isolates Yeast at 24 I VII VIII recovered Bio¬ hours from donor burden tissue

Candida 1.6xl0 ³ 2.6x10* 700 cfu 0 cfu 0 cfu albicans

Candida 1.2X10 ³ 6.3x10* 886 cfu 10 cfu 0 cfu albicans

Candida 1.8x10* 6.8x10* 501 cfu 0 cfu 0 cfu albicans

Candida 3.9xl0 ³ 1.6x10* 921 cfu 0 cfu 0 cfu albicans

Candida 1.3X10 ³ β.lxlO ³ 586 cfu 0 cfu 0 cfu albicans

Yeast, not 4.8xl0 ³ 2.7x10* 70 Cfu 0 cfu 0 cfu C. alb.

Yeast, not 7.0xl0 ³ 4.6x10* 47 cfu 0 cfu 0 cfu C. alb.

Candida 1.6X10 ³ 7.3xl0 ³ 320 cfu 40 cfu 0 fu albicans

Yeast, not 9.8xl0 ^J 2.3X10 ³ 689 cfu 0 cfu 0 cfu C. alb.

Yeast, not 5.9xl0 ³ 3.3x10* l.lxlO ³ 0 cfu 0 cfu C. alb.

The addition of netilmicin at 50 μg/ml (Cocktails VII and VIII) did not have any adverse effects on the fungicidal activity of fluoconazole and amphotericin B. The increase of amphotericin B from 0.3 μg/ml (Cocktail I) to 1.0 μg/ml (Cocktails VII and VIII) shows greater fungicidal activity. The reduction of fluconazole from 100 μg/ml to 50 μg/ml in the presence of 1.0 μg/ml amphotericin B (Cocktail VII versus Cocktail VIII) did not affect the fungicidal activity greatly.

EXAMPLE 17

Heart valves from 223 donors were dissected as described in U.S. Patent No. 4,890,457. The dissected valves were decontaminated using one of the methods listed below.

Duplicate sets of specimens consisting of approximately one gram of tissue were prepared at dissection. One set of specimens was treated by prior art processing method D, E, or F (described in Examples 12 and 13) . The remaining set of samples was treated using a single antibiotic mixture, Cocktail E according to the invention.

METHOD ANTIBIOTIC COCKTAIL TIME/TEMPERATURE

D C

12 μg/ml imipenem 22-26 Hr/37°C

50 μg/ml vancomycin

E C

12 μg/ml imipinem 22-26 Hr/37°C

50 μg/ml vancomycin

B

62 μg/ml rifampin 22/26 Hr/37°C

57 μg/ml netilmicin

114 μg/ml vancomycin

136 μg/ml cefotaxime

136 μg/ml lincomycin

C 12 μg/ml imipenem 22-26 Hr/37°C

50 μg/ml vancomycin

+ optionally

B 62 μg/ml rifampin 22-26 Hr/37°C

57 μg/ml netilmicin 114 μg/ml vancomycin 136 μg/ml cefotaxime 136 μg/ml lincomycin

H E

94-100 μg/ml imipenem 22-26 Hr/37°C

50 μg/ml vancomycin

57 μg/ml netilmicin

100 μg/ml fluconazole

1.0 μg/ml amphotericin B

Microbiological analyses were performed as follows. For each dissected heart, three sets of tissue samples were prepared at dissection to represent the product prior to treatment with antibiotics (pretreatment) and subsequent to antibiotic treatment (post treatment) . The samples taken were pieces of external heart muscle tissue, pieces of the mitral or tricuspid valve, pieces of muscle tissue proximal to the aortic or pulmonary valve, and several milliliters of solution in which the heart was received.

After dissection, the pretreatment set of samples was submitted to the microbiology laboratory for microbial analysis of organisms present. Of the remaining duplicate sets of specimens one was treated along with the heart valves by the method D, E or F. The other set was treated by method Z.

At the end of each treatment period the specimens were asceptically collected and sent to the microbiology lab for analysis. Each set of specimens was assayed separately.

Tissue samples were homogenized by pouring the sample into a sterile plastic bag and placing into a stomacher for 30-60 seconds. The homogenized tissue samples were cultured on the following media: blood agar plate; chocolate agar plate; blood agar plate (anaerobic) ; enriched thioglycollate broth; Sabouraud' s agar with gentamicin; and Sabouraud' s heart infusion agar with blood.

Both the blood agar and the chocolate agar plates were incubated in 5-10% C0 ₂ in air at 35°C for 4 days and observed on the 2nd and 4th days. The thioglycollate broth tubes were incubated at 35°C for 14 days and observed daily for the first 7 days and periodically thereafter. The Sabouraud' s agar plates were incubated at 30°C, and the plates were observed 2 times each the first week and again at 14 days.

The cultures were observed visually for growth. Gram stains were prepared of all positive cultures and of any negative thioglycollate broths to confirm "negative" visual

observations . Accepted state-of -the-art microbiological tests were utilized to identify positive cultures .

The results of the microbiological analyses were presented below :

METHOD D VS . METHOD H

134 Heart Donors Method D No growth on 132 ;

Propionibacterium acnes isolated on 2 ; 98% effective

Method H _No growth on 134; 100% effective

METHOD E vs . METHOD H

64 Heart Donors Method E No growth on 60; Yeast isolated on 4; 94% effective

Method H No growth on 64; 100% effective

METHOD F vs . METHOD H

24 Heart Donors Method F No growth on 23; Group B streptococcus and Enterobacter species isolated on 1 , 96% effective

Method H No growth on 24; 100% effective

CUMULATIVE RESULTS:

Methods D, E, F Method H

216 No growth, 97% effective 223 No growth, 100% effective 7 Growth

No bacterial or fungal growth was observed in any Method H specimens. Thus, Cocktail E is superior to Cocktails B and C (used singly or in combination in Methods D, E and F) in inhibiting bacterial growth, as well as growth of yeast . No bacterial or fungal growth was observed in any Method H specimens. Thus, Cocktail E is superior to Cocktails B and C (used singly or in combination in Methods D, E and F) in inhibiting bacterial growth, as well as growth of yeast.

It is intended that the foregoing detailed description be regarded as illustrative, rather than limiting, and that it be understood that it is the following claims, including all

equivalents, which are intended to define the scope of thi. invention.