Ausf. N . Z . J . Surg. 1990,
801
60, 801-803
SURGICAL RESEARCH ESTABLISHMENT OF A VASCULAR GRAFT INFECTION MODEL IN THE SHEEP CAROTID ARTERY J . P. FLETCHER,'" M.
DRYDEN,"'.
R.
M U N R 0 , 3 i - J.
H. XU"
AND
M. D. HEHIR*
Departments of +Surgery and ?Bacteriology, Wesimead Hospital, Westmead, New South Wales A study was performed to assess whether a model of vascular grali infection could be established in sheep carotid artery. Either a protein sealed Dacron or a polytetrafluoroethylene (PTFE) graft was used in 22 sheep while a control operation was performed on 4 animals. Staphylococcus aureus in concentrations of 10'. lo4, 10" or 10' colony-forming units (CFU) was inoculated into the wound before closure. No infection occurred with an inoculum of 10'. The lowest concentration of organism producing infection was lo4 for PTFE, 10' for Dacron and 10' for controls. Six of eleven DaCrQn grafts and seven o f ten PTFE grafts became infected. The sheep carotid artery has proven a satisfactory model loor studies of vascular graft infection. Key words: dacron, infection, polytetrafluoroethylene, staphylococcus aureus, vascular prosthesis.
Introduction Postoperative infection of a vascular prosthesis, although uncommon, is a severe complication of vascular surgery with high morbidity and mortality. Infection may occur by contamination of the operative site at the time of surgery, or may occur later due to bacteraemia with intravascular seeding of the prosthesis. Early infection of prosthetic vascular grafts is minimized by meticulous surgical technique and perioperative antibiotic prophylaxis. Grafts have been pretreated with antibiotics and endothelial seeding has been used in attempts to further minimize the risk of infection.'-' The aim of this study was to assess whether a model of vascular graft infection could be established in sheep carotid artery. The sheep was chosen as the experimental animal because of its ready availability. The carotid artery was chosen because of its reasonable size (4-5 mm) for acceptance of a vascular graft, ready surgical accessibility and ease of postoperative care.
Methods The study was approved by the Animal Care and Ethics Committee at Westmead Hospital. Merino sheep were used and were cared for and housed I MD. FKACS, FKCS. 'MA. MSc. ' MRCP, MRCPath. Dip. Bact.. FRCPA.
Corrcapondcncc. Asaoc. Prof. J . P. Fletchcr. Dcpdttment of Surgery. Wcstincad Hoap~tal.Westmead. NSW 2 145. Australia. Accepted for Qubllc~lllOii14 March I990
according to approved criteria. Anaesthesia was induced with Pentothal and maintained with nitrous oxide and halothane. The common carotid artery was exposed, a 1 c m segment was excised, and a 4 or 5 mm diameter, 2-3 cm length of vascular prosthesis inserted. Either a protein sealed knitted Dacron graft (Gelseal, Vascutek, Scotland) or a polytetrafluoroethylene (PTFE) graft (Gore-Tex, Flagstaff, Arizona) was inserted in 22 animals. The interposition graft was anastomosed end-to-end to proximal and distal artery with 6/0 polypropylene suture. In 4 animals a control operation was performed in which the carotid artery was divided and reanastomosed without the insertion of a prosthesis. Before wound closure 1 mL of 0.9% saline containing the appropriate concentration of Siaphylococcus aureus was infiltrated around the artery. The sheep were returned to grazing postoperatively. They were sacrificed 3 weeks after surgery or earlier if bacteraemia was suspected. The graft and surrounding tissue were harvested and submitted for bacteriological studies; patency of the graft, the presence of an abscess or tissue infection was noted and a swab taken of tissue surrounding the graft site. Four human blood-culture isolates of Staphylococcus aureus were used in the study. The strains were randomly allocated to each sheep experiment. On the day prior to operation the Staphylococcus aureus was cultured overnight in brain heart infusion broth which was then serially 10-fold diluted in 0.9% saline. With the overnight broth culture containing about 10' colony forming units (CFU) per mL, the appropriate dilution was chosen for inoculation around the vascular graft. Miles and
802
FLETCHER ET A L .
Mizra colony counts were performed to determine the exact concentration of each inoculum.6 Bacteriology was performed on the tissue swab and on the sheep carotid artery and vascular prosthesis immediately after harvesting. The swab was inoculated onto sheep blood agar and a semiquantitative count of Staphylococcus aui-eus CFU was made. The graft was dissected free, the outer surface rolled over a sheep blood agar plate and a semiquantitative count made. The lumen of the graft was flushed three times with I mL of 0.9% saline, and quantitative culture was carried out. All Stciphylococcus aweus isolates from sheep were phage typed to ensure that the strains isolated were the same as the inoculated strains.
Results One sheep died immediately postoperatively, another became obviously infected at 1 week and was sacrificed. All other sheep were sacrificed at 3 wccks. The first 12 sheep were infected with Sfciphylococcus LiureuS concentrations of IO', lo4, IO", 1 0 ~ CFU. Each infecting dose was inoculated into a control sheep, a sheep with a Dacron prosthesis and a sheep with a PTFE prosthesis. Six inoculated with loJ CFU Striphylococciis ciureus or less did not become infected. The graft remained patent and no organisms were isolated from the wound or graft. Of six sheep with an infecting dose of 10" CFU or greater, all graft sites became infected. All infected grafts were occluded and an abscess was present in three. On the basis of these results, an inoculation of 10" or lo6 CFU was chosen for the next 14 sheep, seven receiving a Dacron graft and seven a PTFE graft. Thus there were I 1 sheep receiving a Dacron prosthesis, I 1 a PTFE prosthesis and 4 controls. One of 4 controls became infected, but only at the highest inoculuin concentration of 10' CFU. Six of 1 1 Dacron grafts became infected with an inoculum of 10" or lox CFU. Seven of 10 PTFE grafts became infected; all six of lo6 or 10' were infected while one of three at 10'' became infected. No infection occurrcd with an inoculuin of 10'. Results are summarized in Table 1. A total of nine Dacron and eight PTFE grafts with an inoculum of 104-10" CFU was tested (Table 2). Five Dacron and six PTFE prostheses became infected. If infection was present, organisms were isolated from surrounding tissue, the outer surface of the prosthesis and the lumen of the graft in all cases (Table 3). In 8 infected sheep, graft occlusion and abscess formation were present. In all cases the phage type of the St~iphyloc~occus uureus isolated at graft harvesting corresponded with that of the inoculated strain.
Table 1. Number of positive graft cultures per inoculum of Stcrphplococcus ciurrus ~~
Infecting dose (CFU)
Graft
10'
10'
10"
10'
Total no. inl'ected
Control Dacron PTFE
0/1
0/1 0/3 1/3
0/I 5/6 515
l/I l/I
1/4 611 I
1/1
7/10
0/1 0/1
Table 2. Infection in Dacron and PTFE prostheses following inoculation with 10'- 10' CFU of Sr~~p/iylococc.iis nureus
Dacron
PTFE
No. sheep Infected surrounding tissue Infected graft surface Infected graft lumen Graft occlusion Abscess present No. infected No. not infected and graft patent Table 3. Details of infection following inoculation with 10'- lo6 CFU of Stnphylococcus ~ I L I ~ C L I S
Graft PTFE PTFE PTFE PTFE PTFE PTFE Dacron Dacron
Dacron Dacron
Dacron
Graft Infecting Tissue outer dose infection surface 10' 10"
I O6 I O6 10"
o6
1
I 0" 10" 1 O6 I 06 I06
+++ +++ +++ +++ t ++ +++
+++ +++ +++ +++ + +++ +++ +++ +++ f ++ ++ +++ +++ +++
Graft inner
Appear-
surface
ance
5.3 x 10'
O/A O/A
6.2 X 10i 5.8 x 10'
O/A
4.5 x 10' 540 CFU
O/A P
1.2 x 10'
'? 0 O/A O/A
6.5 x loi 2.9 x 10' 2 CFU 6.9 X 10' 4.9 x 10"
P 0iA
O/A
0 : occluded graft. P. patent graft. A: ab\ce\a present
Discussion
In this sheep study a control operation of division and reanastomosis of the carotid artery did not cause infection until an inoculation of 10' CFU Staphylococcus (iureus was used. With insertion of a vascular graft, infection occurred at a lower concentration of organisms: 10' for Dacron and lo" for PTFE. Once infection was established, organisms were always recoverable from within the lumen of thc graft. The quantification of organisms from within the lumen is probably an underestimation. as other studies have shown sonication to be a more accu-
803
VASCULAR GRAFT INFECTION MODEL IN SHEEP
rate method of determining the number of organisms adherent to a vascular graft;’ however, this method was not used in our study because it does not allow differentiation between bacteria adherent to the outside and those adherent to the lumen of the graft. In 2 sheep, fewer than lo3 CFU were isolated from the lumen of an infected graft and in both cases the graft remained patent. In all other cases where organisms were isolated from the lumen the prosthesis was occluded. Breakdown of the graft to artery anastomosis was not observed in our infected grafts and this may be related to the high incidence of graft occlusion associated with infection in this model. Only 1 sheep died postoperatively, an acceptable mortality of 3 . 8 % . A great advantage with this animal model was the ease of postoperative care, the sheep being returned to grazing immediately after recovery from the anaesthetic. Smaller animal models such as rabbits have vessels less comparable in size to humans, and we have found increasing difficulty in obtaining and housing dogs which have been the most widely used animal model to date.3.”.8-y~10-’2 Sheep carotid artery has been used for studies of endothelialization of Dacron grafts but has not been used for studies of vascular graft infection. l 3 Several aspects of vascular graft infection could be studied further in such a model: for example, the effect of perioperative antibiotic administration, the effect of an intravenous rather than local bacterial challenge and the response of different graft material. In conclusion, this study has demonstrated that the sheep carotid artery is a suitable model for studies of prosthetic vascular graft infection.
2.
3.
4.
5.
6. 7.
8.
9.
10.
II .
Acknowledgements We thank Miss Alison Vickery, Department of Microbiology, Royal Prince Alfred Hospital, Sydney for performing the phage typing of the Stciph,/ncocC K S nut-eus isolates. We also thank Vascutek for supplying the Gelseal grafts and Gore for supplying the Gore-Tex grafts. References 1 . RICHARIXON R. L., PKiE J. W., WOLFI: R . Y . , LtlIES C. & HOPSONW. B. (1970) The outcome of antibiotic-
12.
13.
soaked arterial gral’ts in guinea pig wounds contaniinated with E . c d i or S. aureus. J . Tlioruc. Curdiovcisc. Surg. 71, 635-7. WEBBL. X . , MYERS R . T., CORDELL A . R . , HOBGOOD C. D., COSTEKTON J . W. & GRlSTlNA A. G . (1986) Inhibition of bacterial adhesion by antibacterial surface pretreatment of vascular prostheses. J . Vtrsc. Surg. 4 , 16-21. GRECOR . S., TROOSKIN S. Z . , DONEILA. P. & HARVEY R . A . (1985) The application of antibiotic bonding to the treatment of established vascular prosthetic infection. Arch. Surg. 120, 71-5. MODAKs. M., SAMPATH L . , FOX C . L., BENVENISTKY A , , NOWYGROII R . & REEMSTMAU K. (1987) A new method for the direct incorporation of antibiotic in prosthetic vascular grafts. Surx. C ~ ~ / i e c oOhsrct. l. 164, 143-7. BlRlNYl L. K . , D0UVILL.E E. C . , LEWIS s. A., B J O R N ~ SON H. s. & KtMPCZlNSKl R. F. (1987) Increased resistance to bacteremic graft infection after endothelial cell seeding. J . Vtrsc. Surg. 5 . 193-7. CRUICKSHANK R . , DuGulD J . P . , MARMION B. P. & SWAIN R . H. A . (1975) Mediccrl Microbiology, 12th edn. p. 307. Churchill Livingstone, Edinburgh. SCHMI’ITD. D., BANDYK D. F., PEQUET A . J . & TOWNE J . B. (1986) Bacterial adherence to vascular prostheses. J . Vcisc. Surg. 3, 732-40. FREEDMAN L. R . & VALONEJ . (1979) Experimental infective endocarditis. Pros. Cordiovcrsc. Dis. 22. 715-8. BUCKELS J . A. C . , N O K I X S T G A A K D A. G. &WILSON S. E. ( 1987) Microporous vascular grafts do not require neointima for resistance to bacterial infection. J . Vusc. Surg. 5 ; 198-202. BENNUNR . s., WILLIAMS R. A . & WILSON S. E. (1984) Comparison of infectibility of vascular prosthetic materials by quantitation of median infective dose. Surg. 95, 22-6. DRIES D. J.. ALAZRAKI N . , LAWRENCE P. F. el rrl. (1985) Detection of acute synthetic vascular graft infection with I I ‘In-labeled leukocvte scanning: An animal study. AJR 145. 1053-6. McDANlEL M . D.. SCIINI-D A . R. & BECKJ . R . (1986) Efficacy of silver sulphadiazine in the treatment of prosthetic arterial bypass graft infection. Sirrg. Gynecol. Ohstri. 162, 169-73. PASQUlNELLl G . , PREDA P . , C U R 1 1 T . , D’ADDAIOM. & Lusciii R . (1987) Endothelialization of a new Dacron graft in an experimental model: Light microscopy/ electron microscopy and inimunocytocheniistry. Scrriiiii/ig Microscopy 1, 1327-8.
801
60, 801-803
SURGICAL RESEARCH ESTABLISHMENT OF A VASCULAR GRAFT INFECTION MODEL IN THE SHEEP CAROTID ARTERY J . P. FLETCHER,'" M.
DRYDEN,"'.
R.
M U N R 0 , 3 i - J.
H. XU"
AND
M. D. HEHIR*
Departments of +Surgery and ?Bacteriology, Wesimead Hospital, Westmead, New South Wales A study was performed to assess whether a model of vascular grali infection could be established in sheep carotid artery. Either a protein sealed Dacron or a polytetrafluoroethylene (PTFE) graft was used in 22 sheep while a control operation was performed on 4 animals. Staphylococcus aureus in concentrations of 10'. lo4, 10" or 10' colony-forming units (CFU) was inoculated into the wound before closure. No infection occurred with an inoculum of 10'. The lowest concentration of organism producing infection was lo4 for PTFE, 10' for Dacron and 10' for controls. Six of eleven DaCrQn grafts and seven o f ten PTFE grafts became infected. The sheep carotid artery has proven a satisfactory model loor studies of vascular graft infection. Key words: dacron, infection, polytetrafluoroethylene, staphylococcus aureus, vascular prosthesis.
Introduction Postoperative infection of a vascular prosthesis, although uncommon, is a severe complication of vascular surgery with high morbidity and mortality. Infection may occur by contamination of the operative site at the time of surgery, or may occur later due to bacteraemia with intravascular seeding of the prosthesis. Early infection of prosthetic vascular grafts is minimized by meticulous surgical technique and perioperative antibiotic prophylaxis. Grafts have been pretreated with antibiotics and endothelial seeding has been used in attempts to further minimize the risk of infection.'-' The aim of this study was to assess whether a model of vascular graft infection could be established in sheep carotid artery. The sheep was chosen as the experimental animal because of its ready availability. The carotid artery was chosen because of its reasonable size (4-5 mm) for acceptance of a vascular graft, ready surgical accessibility and ease of postoperative care.
Methods The study was approved by the Animal Care and Ethics Committee at Westmead Hospital. Merino sheep were used and were cared for and housed I MD. FKACS, FKCS. 'MA. MSc. ' MRCP, MRCPath. Dip. Bact.. FRCPA.
Corrcapondcncc. Asaoc. Prof. J . P. Fletchcr. Dcpdttment of Surgery. Wcstincad Hoap~tal.Westmead. NSW 2 145. Australia. Accepted for Qubllc~lllOii14 March I990
according to approved criteria. Anaesthesia was induced with Pentothal and maintained with nitrous oxide and halothane. The common carotid artery was exposed, a 1 c m segment was excised, and a 4 or 5 mm diameter, 2-3 cm length of vascular prosthesis inserted. Either a protein sealed knitted Dacron graft (Gelseal, Vascutek, Scotland) or a polytetrafluoroethylene (PTFE) graft (Gore-Tex, Flagstaff, Arizona) was inserted in 22 animals. The interposition graft was anastomosed end-to-end to proximal and distal artery with 6/0 polypropylene suture. In 4 animals a control operation was performed in which the carotid artery was divided and reanastomosed without the insertion of a prosthesis. Before wound closure 1 mL of 0.9% saline containing the appropriate concentration of Siaphylococcus aureus was infiltrated around the artery. The sheep were returned to grazing postoperatively. They were sacrificed 3 weeks after surgery or earlier if bacteraemia was suspected. The graft and surrounding tissue were harvested and submitted for bacteriological studies; patency of the graft, the presence of an abscess or tissue infection was noted and a swab taken of tissue surrounding the graft site. Four human blood-culture isolates of Staphylococcus aureus were used in the study. The strains were randomly allocated to each sheep experiment. On the day prior to operation the Staphylococcus aureus was cultured overnight in brain heart infusion broth which was then serially 10-fold diluted in 0.9% saline. With the overnight broth culture containing about 10' colony forming units (CFU) per mL, the appropriate dilution was chosen for inoculation around the vascular graft. Miles and
802
FLETCHER ET A L .
Mizra colony counts were performed to determine the exact concentration of each inoculum.6 Bacteriology was performed on the tissue swab and on the sheep carotid artery and vascular prosthesis immediately after harvesting. The swab was inoculated onto sheep blood agar and a semiquantitative count of Staphylococcus aui-eus CFU was made. The graft was dissected free, the outer surface rolled over a sheep blood agar plate and a semiquantitative count made. The lumen of the graft was flushed three times with I mL of 0.9% saline, and quantitative culture was carried out. All Stciphylococcus aweus isolates from sheep were phage typed to ensure that the strains isolated were the same as the inoculated strains.
Results One sheep died immediately postoperatively, another became obviously infected at 1 week and was sacrificed. All other sheep were sacrificed at 3 wccks. The first 12 sheep were infected with Sfciphylococcus LiureuS concentrations of IO', lo4, IO", 1 0 ~ CFU. Each infecting dose was inoculated into a control sheep, a sheep with a Dacron prosthesis and a sheep with a PTFE prosthesis. Six inoculated with loJ CFU Striphylococciis ciureus or less did not become infected. The graft remained patent and no organisms were isolated from the wound or graft. Of six sheep with an infecting dose of 10" CFU or greater, all graft sites became infected. All infected grafts were occluded and an abscess was present in three. On the basis of these results, an inoculation of 10" or lo6 CFU was chosen for the next 14 sheep, seven receiving a Dacron graft and seven a PTFE graft. Thus there were I 1 sheep receiving a Dacron prosthesis, I 1 a PTFE prosthesis and 4 controls. One of 4 controls became infected, but only at the highest inoculuin concentration of 10' CFU. Six of 1 1 Dacron grafts became infected with an inoculum of 10" or lox CFU. Seven of 10 PTFE grafts became infected; all six of lo6 or 10' were infected while one of three at 10'' became infected. No infection occurrcd with an inoculuin of 10'. Results are summarized in Table 1. A total of nine Dacron and eight PTFE grafts with an inoculum of 104-10" CFU was tested (Table 2). Five Dacron and six PTFE prostheses became infected. If infection was present, organisms were isolated from surrounding tissue, the outer surface of the prosthesis and the lumen of the graft in all cases (Table 3). In 8 infected sheep, graft occlusion and abscess formation were present. In all cases the phage type of the St~iphyloc~occus uureus isolated at graft harvesting corresponded with that of the inoculated strain.
Table 1. Number of positive graft cultures per inoculum of Stcrphplococcus ciurrus ~~
Infecting dose (CFU)
Graft
10'
10'
10"
10'
Total no. inl'ected
Control Dacron PTFE
0/1
0/1 0/3 1/3
0/I 5/6 515
l/I l/I
1/4 611 I
1/1
7/10
0/1 0/1
Table 2. Infection in Dacron and PTFE prostheses following inoculation with 10'- 10' CFU of Sr~~p/iylococc.iis nureus
Dacron
PTFE
No. sheep Infected surrounding tissue Infected graft surface Infected graft lumen Graft occlusion Abscess present No. infected No. not infected and graft patent Table 3. Details of infection following inoculation with 10'- lo6 CFU of Stnphylococcus ~ I L I ~ C L I S
Graft PTFE PTFE PTFE PTFE PTFE PTFE Dacron Dacron
Dacron Dacron
Dacron
Graft Infecting Tissue outer dose infection surface 10' 10"
I O6 I O6 10"
o6
1
I 0" 10" 1 O6 I 06 I06
+++ +++ +++ +++ t ++ +++
+++ +++ +++ +++ + +++ +++ +++ +++ f ++ ++ +++ +++ +++
Graft inner
Appear-
surface
ance
5.3 x 10'
O/A O/A
6.2 X 10i 5.8 x 10'
O/A
4.5 x 10' 540 CFU
O/A P
1.2 x 10'
'? 0 O/A O/A
6.5 x loi 2.9 x 10' 2 CFU 6.9 X 10' 4.9 x 10"
P 0iA
O/A
0 : occluded graft. P. patent graft. A: ab\ce\a present
Discussion
In this sheep study a control operation of division and reanastomosis of the carotid artery did not cause infection until an inoculation of 10' CFU Staphylococcus (iureus was used. With insertion of a vascular graft, infection occurred at a lower concentration of organisms: 10' for Dacron and lo" for PTFE. Once infection was established, organisms were always recoverable from within the lumen of thc graft. The quantification of organisms from within the lumen is probably an underestimation. as other studies have shown sonication to be a more accu-
803
VASCULAR GRAFT INFECTION MODEL IN SHEEP
rate method of determining the number of organisms adherent to a vascular graft;’ however, this method was not used in our study because it does not allow differentiation between bacteria adherent to the outside and those adherent to the lumen of the graft. In 2 sheep, fewer than lo3 CFU were isolated from the lumen of an infected graft and in both cases the graft remained patent. In all other cases where organisms were isolated from the lumen the prosthesis was occluded. Breakdown of the graft to artery anastomosis was not observed in our infected grafts and this may be related to the high incidence of graft occlusion associated with infection in this model. Only 1 sheep died postoperatively, an acceptable mortality of 3 . 8 % . A great advantage with this animal model was the ease of postoperative care, the sheep being returned to grazing immediately after recovery from the anaesthetic. Smaller animal models such as rabbits have vessels less comparable in size to humans, and we have found increasing difficulty in obtaining and housing dogs which have been the most widely used animal model to date.3.”.8-y~10-’2 Sheep carotid artery has been used for studies of endothelialization of Dacron grafts but has not been used for studies of vascular graft infection. l 3 Several aspects of vascular graft infection could be studied further in such a model: for example, the effect of perioperative antibiotic administration, the effect of an intravenous rather than local bacterial challenge and the response of different graft material. In conclusion, this study has demonstrated that the sheep carotid artery is a suitable model for studies of prosthetic vascular graft infection.
2.
3.
4.
5.
6. 7.
8.
9.
10.
II .
Acknowledgements We thank Miss Alison Vickery, Department of Microbiology, Royal Prince Alfred Hospital, Sydney for performing the phage typing of the Stciph,/ncocC K S nut-eus isolates. We also thank Vascutek for supplying the Gelseal grafts and Gore for supplying the Gore-Tex grafts. References 1 . RICHARIXON R. L., PKiE J. W., WOLFI: R . Y . , LtlIES C. & HOPSONW. B. (1970) The outcome of antibiotic-
12.
13.
soaked arterial gral’ts in guinea pig wounds contaniinated with E . c d i or S. aureus. J . Tlioruc. Curdiovcisc. Surg. 71, 635-7. WEBBL. X . , MYERS R . T., CORDELL A . R . , HOBGOOD C. D., COSTEKTON J . W. & GRlSTlNA A. G . (1986) Inhibition of bacterial adhesion by antibacterial surface pretreatment of vascular prostheses. J . Vtrsc. Surg. 4 , 16-21. GRECOR . S., TROOSKIN S. Z . , DONEILA. P. & HARVEY R . A . (1985) The application of antibiotic bonding to the treatment of established vascular prosthetic infection. Arch. Surg. 120, 71-5. MODAKs. M., SAMPATH L . , FOX C . L., BENVENISTKY A , , NOWYGROII R . & REEMSTMAU K. (1987) A new method for the direct incorporation of antibiotic in prosthetic vascular grafts. Surx. C ~ ~ / i e c oOhsrct. l. 164, 143-7. BlRlNYl L. K . , D0UVILL.E E. C . , LEWIS s. A., B J O R N ~ SON H. s. & KtMPCZlNSKl R. F. (1987) Increased resistance to bacteremic graft infection after endothelial cell seeding. J . Vtrsc. Surg. 5 . 193-7. CRUICKSHANK R . , DuGulD J . P . , MARMION B. P. & SWAIN R . H. A . (1975) Mediccrl Microbiology, 12th edn. p. 307. Churchill Livingstone, Edinburgh. SCHMI’ITD. D., BANDYK D. F., PEQUET A . J . & TOWNE J . B. (1986) Bacterial adherence to vascular prostheses. J . Vcisc. Surg. 3, 732-40. FREEDMAN L. R . & VALONEJ . (1979) Experimental infective endocarditis. Pros. Cordiovcrsc. Dis. 22. 715-8. BUCKELS J . A. C . , N O K I X S T G A A K D A. G. &WILSON S. E. ( 1987) Microporous vascular grafts do not require neointima for resistance to bacterial infection. J . Vusc. Surg. 5 ; 198-202. BENNUNR . s., WILLIAMS R. A . & WILSON S. E. (1984) Comparison of infectibility of vascular prosthetic materials by quantitation of median infective dose. Surg. 95, 22-6. DRIES D. J.. ALAZRAKI N . , LAWRENCE P. F. el rrl. (1985) Detection of acute synthetic vascular graft infection with I I ‘In-labeled leukocvte scanning: An animal study. AJR 145. 1053-6. McDANlEL M . D.. SCIINI-D A . R. & BECKJ . R . (1986) Efficacy of silver sulphadiazine in the treatment of prosthetic arterial bypass graft infection. Sirrg. Gynecol. Ohstri. 162, 169-73. PASQUlNELLl G . , PREDA P . , C U R 1 1 T . , D’ADDAIOM. & Lusciii R . (1987) Endothelialization of a new Dacron graft in an experimental model: Light microscopy/ electron microscopy and inimunocytocheniistry. Scrriiiii/ig Microscopy 1, 1327-8.
