Effects of physical configuration and chemical structure of suture materials on bacterial adhesion

Effects of Physical Configuration and Chemical Structure Of Suture lvjaterials on Bacterial Adhesion A Possible Link to Wound Infection

Chih-Chang Chu, PhD, Liverpool, England David F. Wltllams, PhD, Liverpool, England

The development of infection in incisional wounds continues to be one of the most serious complications that can occur in surgical patients. It can have significant effects on both mortality and morbidity, influencing the final outcome of the operation. Surgical infection may not only severely retard normal healing but also may induce life-threatening clinical situations, particularly in patients with critical chronic illness. Almost all postoperative wound infections are initiated along and in the vicinity of the suture lines. It has been well-demonstrated that the presence of suture material in the wound increases the susceptibility of host tissue to infection [I-4]. It is known that the number of bacteria needed to establish infection can be reduced 10,000-fold by the presence of a silk suture [5,6]. In addition, suture materials can also serve as a vehicle for mechanical transport of bacteria into the surgical wounds. This physical process may be aided further by the capillary action of the suture materials. It is already known that the majority of infections start on the mucous membranes through microbial attachment with or without significant penetration [7). The degree of adherence by pathogens to the mucous membrane surface was found to be considerably selective. The specific interactions between surface components of bacteria and host are responaible for this selective adherence. For example,

Fran the oepamat of Dental Scbnce. School of Dental Swgsry, The UnivarsHy of Llvqool. Pmtmke PtncB. Lhwpool. Engbnd. Sgported In partbytheBmlrh~andEnglnsalngReeeerchCouncllV~FeC Iowatip &mt C/O7735 end the Brltlsh Science and En9lneerlng fWetmh Camll @ml B/34515. Reque~f(ar~l~~drouldbeaddrcwsedtoch~chu.phD,Depfrbnmof~mdEmdrAnnlysk. CamelI lhhrersny. lthaca, New Yak 14953.

Volume 147, Fe&uwy

1094

pathogenic Staphylococcus pyogenes attach to human pharyngeal cells better than Escherichia coli; Vibrio cholerae and E. coli adhere preferably to the epithelium of the upper rather than the lower bowel, whereas Shigella flexneri attach to colonic cells rather than to those of the upper bowel [8]. The question is whether or not this selective adherence is applicable to foreign materials, such as sutures, that may be present in the wound area. The question is particularly important in view of the wide range of surface finishes and chemical properties that suture materials can have. Furthermore, different microorganisms have different surface characteristics which could also contribute to the selective adherence of bacteria to foreign materials. Thus, the purpose of this study was to examine this preferential adherence of bacteria to suture materials quantitatively by radiolabelled cells and qualitatively by scanning electron microscopy. The study employed two types of microbes that differed in their surface characteristics, including a new synthetic, absorbable suture material. Material and Mc+ds Ten suture materials, absorbable and nonabsorbable, synthetic and natural origin, monofilament and braid, and costed and uncoated, were used: chromic catgut, braided polyglycslic acid (Dexon@l,braided polyglycolide &tide (Vicrile), polydioxanone (PDS), a new synthetic monofilament absorbable suture [S], braided polyethylene terephthslate. (Mersilenb), braided, silicone-coated polyethylene terephthalate (Tycron”), braided, polybutilatecoated polyethylene terephthalate (Ethibonfl), braided polyamide (Surgilone), polysniide nionofilament (Ethilon@),and polypropylene monofilament (Prolen@). Their physical and chemical characteristics are listed in Table I. All of the suture materials were 2-O.The bmterial strains teeted were Staphylococcus aureus and E. coli. They are

197

Chu and Williams

TABLE I

Physlcal and Chemical Characteristics of 2-O Suture Material Suture Material

Catgut Dexon Vicryl PDS Mersilene Tycron Ethibond Surgilon Ethilon Prolene

Chemical Structure

Physical Configuration

Surface Coating

Manufacturer

Protein Polyglycolic acid Polyglycolide lactide Polydioxanone Polyethylene terephthalate Polyethylene terephthalate Polyethylene terephthalate Polyamide (nylon 66) Polyamide (nylon 66) Polypropylene

Monofilament Braided multifilament Braided multifilament Monofilament Braided multifilament

Chromic salt None Glycolide, lactide Unknown None

Ethicon Davis & Geck Ethicon Ethicon Ethicon

Braided multifilament

Silicone

Davis 81Geck

Braided multifilament

Polybutilate

Ethicon

Braided multifilament Monofilament Monofilament

Silicone None None

Davis & Geck Ethicon Ethicon

not only the most common sources of surgical wound infection but also represent two different bacterial surface characteristics. Gram-negative bacteria like E. coli contain an additional structure of lipopolysaccharide which does not exist in gram-positive bacteria, such as Staph. aureus. This difference in surface properties was expected to result in different tendencies to adhere to suture material. These microorganisms were subcultured on agar slants and stored at 4°C without further passage. Adherence of radiolabelled bacteria: The bacteria were transfered by a sterile wire loop from the stock agar to a freshly prepared 25 ml brain-heart infusion broth containing 0.25 ml of 250 Ci radiolabelled [methyl-3H] thymidine (Radiochemical Center, Amersham, England) with a specific activity and radioactive concentration of 44 Ci/mmole and 1 mCi/ml, respectively. After incubating the radiolabelled broth for 18 hours at 37”C, the bacteria were collected by centrifugation and washing in 0.1 M sterilized phosphate-buffered saline solution (pH 7.4) 3 times at 5,006 rpm for 25 minutes. At the end of the third washing, the bacteria were resuspended in 25 ml of sterilized phosphate-buffered saline solution and subsequently diluted to 1 part per 10 in sterilized phosphate-buffered saline solution as the final testing solution. The sterilized suture specimens, each 3 cm long, were immersed in sterilized plastic vials, each containing 5 ml of final testing solution, for 20,60, 120, and 180 minutes. The suture specimens were then removed after the predetermined period of immersion and washed four times in sterilized phosphate-buffered saline solution to remove the nonadherent bacteria. Finally, each suture specimen was placed in 10 ml of 0.1 M, sodium hydroxide solution and shaked for 1 hour to remove adherent bacteria. This method has been used previously by other investigators

and found to be satisfactory. The sodium hydroxide solution was then adjusted to pH 7 by hydrochloric acid and triton (Triton@)X lOO/Toluene scintillant fluid was added to 3 ml of the neutralized sodium hydroxide for determination of radioactivity, measured as counts per minute in a liquid scintillation counter (Packard Tri-Carb 300). A vial containing only the scintillant fluid was used to monitor the background counts per minute. The counts per minute of a known quantity of bacteria (in number of bacteria per milliliter) was also counted for the purpose of converting the counts per minute into number of bacteria. To compare the relative affinity of suture materials

198

toward bacterial adherence, an adherence index was calculated as defined by Katz et al [IO]. A slightly different definition, however, was proposed by us: Adherence Index =

number of bacteria of a suture specimen number of bacteria of a reference suture

The reference suture w’ssthe one that showed the smallest number of adherent bacteria after 20 minutes of immersion. In this study, PDS was chosen as the reference suture.

Electron microscopy: Similar samples’of the s&e suture materials used in the radiolabelled study were incubated for 1 hour in 10 ml of bacterial suspension for a

scanning electron microscope morphclogic study. The bacterial suspension was obtained from a 1 part per 10 dilution of the stock unlabelled bacterial solution in Ringer’s solution. After 1 hour of immersion, the suture materials were taken out and washed three times in phosphate-buffered saline solution then agitated in a solution containing 2.5 percent glutaraldehyde per 0.1 M cacodylate buffer for, 1 hour. The suture specimens were washed again by 0.1 M buffered sodium cacodylate and dehydrated with increasing concentrations of alcohol. The suture specimens were then fixed to aluminum stubs, coated with gold, and examined in a scanning electron microscope (Joel 35~).

Results Tables II and III summarize the results from radiolabelled E. coli and Staph. aureus on both absorbable and nonabsorbable sutures over 3 hours. It was found that the amount of adhered bacteria depended on the type of suture material, the type of bacteria, and the duration of contact. The bacterial adherence on suture materials was also a dynamic rather than a static phenomenon. The affinity of suture materials toward E. coli were grouped into three categories depending on their kinetic adherence profiles. The first group, consisting of chromic catgut, Dexon, Mersilene, Surgilon, and Ethilon showed a relative decrease in the amount of adherent bacteria as the duration of contact increased from 20 to 60 minutes, and this was followed by an increase as the duration increased to 120 minutes. A further increase in contact time to 180 minutes resulted in a decrease in the number of adherent

The Amerkan Journal of Surgery

Suture Materials and Bacterial Adhedon

TABLE II

Amount ot Radidabelkd Escherichla coli Adhered on 2-O Suture Material (3 cm) ContactTime

Suture

20Min

Chromiccatgut

8.4 X 9.9 x 1.5 x 5.2 x 6.1 X 8.9 X 6.8 x 6.4 X 1.2 x 7.6 X

Dexon Vicryl PDS Mersilene Tycron Ethibond Surgilon Ethilon Prolene

lo3 103 104 103 lo3 lo3 103 lo3 104 lo3

60Min 8.8 x 6.9 X 7.3 x 7.6 X 5.4 x 7.7 x 7.5 x 5.3 x 5.3 x 6.3 X

120Min

103 lo3 103 lo3 103 103 103 103 103 lo3

8.2 X 9.2 x 5.1 x 5.5 x 7.4 x 4.2 X 2.0x 7.6 X 6.4 X 7.8 X

bacteria to below the level at 20 minutes. Thus, the adherence kinetic profiles exhibited a zig-zag shape. The magnitude of the change of adherent E. coli in time depended on the type of suture material. Mersilene and chromic catgut sutures showed the smallest fluctuation with time, whereas Surgilon and Ethilon showed the largest fluctuation. The second group consisting of Vicryl, Tycron, and Prolene, exhibited a continuous decrease in the number of adherent bacteria in time after the initial 20 minute exposure. The largest decrease in the number of adherent E. coli occurred from 20 to 60 minutes in Vicry1 sutures and from 60 to 120 minutes in Tycron sutures. The third group, consisting of PDS and Ethibond, showed the same zig-zag adherence kinetic profile as the first group but it was out of phase along the time scale. There was an increase in the number of adherent bacteria as the time increased from 20 to 60 minutes followed by a decrease at 120 minutes and an increase again at 180 minutes. In the case of Staph. aureus, chromic catgut showed a continuous increase in the number of adherent bacteria with time whereas PDS sutures exhibited the opposite phenomenon (that is, a continuous decrease with time). Vicryl showed a maximum adherence at 120 minutes whereas Dexon sutures exhibited a zig-zag adherence kinetic profile. Dexon sutures also had the highest average number of adherent Staph. aureus bacteria over the four contact periods. PDS, however, had the lowest average number of adhered bacteria over the same time period. There was’s fivefold difference between these two suture materials. To compare the relative affinity of the tested sutures toward both Staph. aureus and E. coli, an adherence index was used. The results are summarized TABLE III

lo3 103 103 103 103 lo3 103 lo3 lo3 lo3

180Min 6.9 X 5.6 X 5.0 x 5.9 x 5.3 x 4.2 X 7.8 X 2.1 x 4.1 x 4.8 X

lo3 103 103 103 103 lo3 lo3 103 103 lo3

Averaae 7.6 X 8.0 X 8.1 X 6.1 X 8.1 X 6.3 X 6.0 X 5.4 x 7.0 x 7.1 x

lo3 lo3 lo3 lo3 lo3 lo3 lo3 103 103 103

in Figures 1, 2, and 3. An adherence index below 1 indicates that the number of adhered bacteria on the suture specimen is less than on the PDS reference suture at 20 minutes. Except for a few occasions, it was found that the sutures exhibited a higher bacteria adherence than PDS over the study period. Staph. aureus adherence was also higher than E. coli with a given suture. With chromic catgut sutures, the adherence index of E. coli remained relatively constant with time whereas the adherence index of Staph. aureus increased continuously with time. After 120 minutes, Vicryl sutures behaved similarly to PDS sutures in E. coli broth, but the affinity toward Staph. aureus was very different from PDS. Dexon sutures showed a higher bacterial affinity than PDS, particularly in Staph. aureus. PDS sutures, however, behaved quite differently from the other absorbable sutures, exhibiting fewer adherent Staph. aureus bacteria with time and maintaining a relatively unchanged E. coli adherence. With nonabsorbable sutures, the physical configuration of the sutures contributed to their ability to attract bacteria. For example, as shown in Figure 2, braided nylon sutures (Surgilon) had a higher adherence index with E. coli than monofilament nylon sutures (Ethilon) at 20 minutes, but this difference became too small to be significant at longer periods of contact. Different surface finishes also influenced the suture’s bacteria adherence but the effect was small as shown in Figure 3. Siliconized polyester sutures (Tycrona) had the highest initial adherence index among the three polyester sutures, but its adherence index decreased with time and reached a level slightly below that of PDS reference sutures at 120 and 180 minutes. The uncoated polyester (Mersilene@) adherence

Amount of Radiolabelled Staphylococcusaureus Adhered on 2-O Suture Material (3 cm) ContactTime 20Mln

Suture Chromic catgut

4.3 x 6.3 X 3.9 x 1.3 x

D8XOl-l

Vicryl PDS

Volume

147, F.&wary

1984

105 lo5 105 105

80Min 4.1 x 5.0 x 4.8 X 1.0 x

105 105 lo5 105

120Mln 5.5 x 1.6 X 9.1 x 5.8 X

105 lo8 105 lo4

180Mln 5.7 x 7.6 X 4.3 x 4.5 x

106 lo5 105 104

Average 4.9 x 8.8 x 5.5 x 0.8 X

105 105 105 lo5

199

Chu and Williams

Tlma

of Contact

hhutd Thm

a( Contrt

Iminuted

F&e 1. lbe bacterialadherencelndexd 2-O suturemateA& a&r varkus lncubatkn t&nes.Topkfl, ckomk cat@. M ckaed wanpke ~~~~aureus;~open~lhdlcatesE~~ooN.Top~,~x~~~~~~~ aureua; the open square Indkates Escherkhla toll. Bottomlett, Vkryl. lln? closed circle indkates Staphytocaccusauthe open circle indkates Escherlchla toll. Bottomright, PDS. lb closed inverted triangle lndkates Staphykcocces aurew the open lnverted trlangk indkates Escherkhla toll.

200

The American Journal ol Surgery

Suture Materials and Bacterial Adhesion

4 F&we 2,77m efhwt ofphyalcai conf&mWm of nybn 2-O suture wlthea#ulmcekrkxofEcon~v~~ t/mes.7Beopmdbmond/ndkateaSuqWon(&rairbd)andthe PartiaUyclosed dbmond Indkates Ethllon ( monofUament).

index was similar to that of PDS. The qualitative observations from the scanning electron micrographs are also consistent with the quantitative data. As shown in Figures 4 and 5, there were very few Staph. aureus or E. coli bacteria adhered on PDS sutures after 60 minutes of contact, whereas other sutures showed quite a high affinity to microbe adherence after the same contact period. From the scanning electron micrographs, the higher affinity of Staph. aureus to sutures compared with E. coli, as observed in the radiolabelled study, was further demonstrated. Not only did the microbes adhere on the rough surface of the interstices among the filaments, but also a major portion of them stayed on the smooth noncrevice parts of the sutures. Another interesting morphologic observation was that Staph. aureus adhered on the suture surface in clusters, whereas E. coli tended to adhere individually. comments The presence of foreign materials in a wound significantly enhances the susceptibility of surrounding tissues to infection. Among all these foreign materials, suture materials are the most important because

Volume 147, Fetmmry 1984

Figure 3. llw effect of coatingmate&b of dacron2-O sutureon theadhemnmin&xofEscherkhbcoUaUervarkusfncubatkn tkmrs.lWopen&cbbuMatesWnlknn MWOabd);fln~ ~-~tdllconbsd);theqpanhtmJbEth&umd(Mybutlbtecoated).

of the high frequency of use in wound closure and many distressing complications, such as infection, wound disruption, and chronic sinus formation, occur in sutured wounds [11-131. The fact that suture materials can predispose to wound infection can be explained by the formula developed by Altemeier and Culbertson [14]: Risk of Wound Infection = dose of bacterial contamination host resistance

X

virulence

This suggests that the inoculation of a certain number of bacteria into a wound does not necessarily result in the development of a wound infection. The key is the extent of contamination. Elek and Conen [5] demonstrated that 7.5 X 106viable staphylococci were normally required to induce an infection intradermally whereas as few as 300 bacteria were needed to elicit a similar infection in the presence of a silk suture. In this study, we have shown that a range of affinity of bacterial adherence on suture materials has been observed due to the difference of suture materials as well as the type of bacteria. The new synthetic absorbable suture PDS had the least affinity toward

201

Chu and Williams

Flpe

chmk

4. Scam&g ok&on

mkrograph of StapnvrococcuS aureus adherence on 2-O suture materiats after 80 mtmdes ImWbatkn. (a) catgut, (b) Dexon, (c) Vkryl, (d) PDS.

we 8. Scam&# ekotmn nrlcrogrrph of Eahwkfda m,(c)

202

cofl adherence on 2-O wlwv matutab aHer 60 mhuie# hwbatbn.

(a) Vki

Tycron,(d)-.

The Amarkan JournalO! SufWw

Suture Materials and B&cWrial Adhesion

both Staph. aureus and E. coli. This lower bacterial affinity is believed to be partially due to its monofilament configuration which has lower surface area than braided structures, although this cannot be the only factor. The chemical structure of PDS sutures may also contribute to this low affinity. The Dexon suture showed much higher bacterial affinity than PDS over the whole time period. The braided structure of the Dexon suture has an important influence on its high bacterial affinity, but the chemical structure of sutures which affect hydrophilicity of sutures also has a definite role. An examination of both PDS and Dexon chemical structures indicates that the former is expected to be more hydrophobic than the latter suture because of the lesser amount of the water affinity group (ester group). Hoget et al [15] demonstrated the concept of hydrophilicity and hydrophobicity of substrates and their influence on bacterial adhesion. The high bacterial affinity of Dexon sutures has also been observed in two other studies [10,16]. The present data, in addition to other data, clearly indicate that the chemical nature of the suture and its coating material are the major factors contributing to bacterial attachment. For example, chromic catgut, PDS, Ethilon, and Prolene sutures are all in monofilament form, but their bacterial affinity differs widely from one another, from as low as 5.2 X 103 E. coli bacteria for PDS to as high as 1.25 X 104 E. coli bacteria for Ethilon. Furthermore, braided sutures have long been expected to have more adherent bacteria than their monofilament counterparts from the point of view of surface area, but our results do not totally agree with this general expectation. Surgilon, a silicone-coated braided nylon suture had half of the bacterial affinity (Staph. aureus) of Ethilon, an uncoated monofilament nylon at 20 and 180 minutes of contact. This suggests that the coating material has a more important influence than the physical configuration. The reasons are not clear due to the lack of agreement between our findings and those of others. Our study also indicates that far more Staph. aureus adhered to sutures than did E. coli; the difference being as high as 100 times in some cases. This difference observed both quantitatively (radiolabelled) and qualitatively (scanning electron microscope) may be due to the different cell wall structures of the bacteria. A gram-positive cell wall, like Staph. aureus, consists of a single rigid layer of peptidoglycan, whereas a gram-negative cell wall, such as E. coli, is a multilayered, complex structure and consists of additional layers of lipopolysaccharide and protein. This difference in cell wall structure can be expected to result in different adherence affinity as observed in this and other studies. Another interesting observation of our study is that the attachment of bacteria on a suture surface

Volume 147, February 1884

was not time independent but a dynamic phenomenon. This indicates that the attachment of bacteria on a suture surface is a reversible process. Sugarman and Mosher [16] reported a similar time-dependent attachment phenomenon but the pattern differed from that in the present study. A nonlinear continuous increase in adherent bacteria with the duration of contact was found in their study, and the magnitude of the increase was the largest during the early period of contact. Finally, a saturation stage was achieved at a later period. With few exceptions, most of the sutures in the present study showed a zig-zag attachment pattern. The mechanism of this type of reversible attachment has been proposed by Marshall et al [17] and was interpreted in terms of the balance between the electrical double-layer repulsion energies at different electrolyte concentrations and the van der Waals attractive energies. Reversible sorption occurs more frequently in a motile bacteria and hence the highly motile bacteria would be expected to have a more frequent adsorption and desorption than a less motile bacteria. In this study, the more steady adsorption of Staph. aureus on a suture surface when compared with E. coli illustrates this point of view. The high motility of E. coli results from the presence of a special organelle of motility, the flagellum. It has a velocity of 16.5 Fm/s which is equivalent to eight times its own cell length moved per second [18]. The bacteria could occassionally break away from the surface if the momentum of rotation was greater than the adhesion force and go to the broth medium. The process of reversible attachment of bacteria onto the surface of suture material could turn into an irreversible process if there is a firm adhesion of bacteria to the suture’s surface by way of an extracellular adhesive medium. This adhesive medium was first reported by ZoBell [19] and has also been observed in numerous other studies [20-221. This adhesive medium has the appearance of tangled polymeric fibrils of polysaccharides or branching sugar molecules that extend from the bacterial surface and form a feltlike glycocalyx surrounding an individual cell or a colony of cells and is secreted by bacteria. Thus, irreversible bacterial adherence would not occur until their synthesis of this extracellular adhesive material was complete. This indicates that there will be a time delay before irreversible attachment can occur. From the energy conservation point of view, bacteria must expend energy to generate and maintain a glycocalyx adhesive material, and it is a metabolically expensive luxury in the protected environment of a pure bacterial culture [22]. This may be the reason for the discrepancy between the present and other results. Morphologic observations by the scanning electron microscope indicate that Staph. aureus adhered on the suture surface in clusters whereas E. coli adhered in an individual fashion. This difference in mor-

203

Chu and Williams

phologic attachment may be correlated with the production of glycocalyx adhesive matrix. Summary The purpose of this study was to examine the effects of physical configuration and the chemical nature of suture materials on the preferential adherence of bacteria. Ten suture materials of 2-O (chromic catgut, Dexon, Vicryl, PDS, Mersilene, Tycron, Ethibond, Surgilon, Ethilon, and Prolene) were used. The bacterial strains tested were Staph. aureus and E. coli. The level of bacterial adherence was determined quantitatively by radiolabelled cells and qualitatively by scarming electron microscopy. It was found that the amount of adhered bacteria depended on the type of suture material, the type of bacteria, and the duration of contact. In the group of absorbable sutures, the new PDS sutures exhibited the smallest affinity toward the adherence of both E. coli and Staph. aureus. Dexon sutures had the highest affinity toward these two bacteria. With nonabsorbable sutures, the physical configuration of the sutures contributed more to their ability to attract bacteria than the surface finish. The bacterial adherence on suture materials was also time dependent. Scanning electron microscope morphologic observation also indicated that Staph. aureus adhered on the suture surface in clusters whereas E. coli tended to adhere individually. Acknowledgment: We thank Dr. G. Jayson of the Liverpool Polytechnic for his assistance in &radiating the suture samples and the technical help of J. Dwyer, J. Chesters, and H. Dunn. We also thank Ethicon, Ltd. in Scotland for their cooperation in supplying the PDS sutures.

References 1. Edlich

RF. Panek F’H, Roc@heaver GT. et al. Physical and chemical configuration of sutures in ths development of surgical infection. Ann Surg 1973;177:879-87. 2. Alexander JW, Kaplan JZ, Aitemeler WA. Role of suture materials in the development of wound infection. Ann Surg 1987;185:192-9. 3. Bfomstedt B, Dsterberg B. Sutue rr&rIats and wound infection.

204

An expe&ntal study. Acta Chir Stand 1978;144:28974. 4. Dsterberg B, Blomstedl B. Effect of sutue r+riais cm bacteriai survival in infected wounds. Acta Chir Stand 1979;145: 431-4. 5. Elek SD, Conen PE. The virulence of S. pyogenes for man. A study of the problems of wound infection. Br J Exp Pathol 1957;38:573-88. 8. Howe CW. Marston AT. A study on sauces of postoperative staphylococcal infection. Surg Gynecol Dbstet 1982;115: 288. 7. Smith H. Microbial surfaces in relation to pathogenicity. Bacteriol Rev 1977;41:475-500. . 8. Gibbons FtJ. Spinell DM, Skobe Z. Selective adherence as a determinant of the host tropisms of certain indigenous and pathogenic bacteria. Infect lmmuil 1978;13:238-48. 9. Ray JA, Doddi N, Regula D. et al. Polydloxanone (PDS) a novel monofilament synthetic absc+bable suhrre. Surg Gyrrecol Dbstet 1981;153:497. 10. Katz S, lzhar M, Mirelman D. Bacterial adherence to surgical sutures. A possible factor in suture induced infection. Ann surg 1981;194:35. 11. Everett WG. Sutures, incisions, and anastomoses. Ann R Coil Surg Engl 1974;55:31-8. 12. Brunius U, Zerderfekft B. Suture materials in general surgery, a comment. Prog Surg 1970;8:38-44. 13. Everett WG. Suture materials in general surgery. Prog Surg 1970;8:14. 14. Altemeier WA, Culbertson WR. Sqrgical infection. In: Rhoads J, Moyer C. eds. Surgery, principles and practice. 4th ed. Philadelphia: JB Lipplncott. 1970:48-74. 15. l-bgl AH. Feijen J, Dankert J, et al. A&e&on of staphylowccus epkfermidis and safxcphyticusonto rrp-teRon and oellubse acetate. Proceedings of the International Conference on Biomedical Polymers. Dufham, England: Biological Engineering Society, 1982:39-47. 18. Sugarman B, Musher D. Adherence of bacteria to suture materials. Proc Sot Exp Biol Med 1981:187:158. 17. h4arshall KC, Stcut R, Mitchell R. f@chantsm of the inittal events in the sorption of marine bacteria to surfaces. J Gen Microbiol 1971;88:337-48. 18. Vaituzis Z, Doe&h RN. Motility tracks: technique for quantitative strMy of bacterial movement. Appl Microbial 1989; 17:584-a. ) 19. ZoBell CE. The effect of solid surfaces upan bacterial ectivity. J Bacterial 1943;48:39-59. 20. Corpe WA. An acid polysaccharide produced by a primary film-forming marine bacterium. Dev Must Microbial 1970;11:402-12. 21. Corpe WA. Attachment of marine bacteria to solid surfaces. In:hlsntyRS. ed. A&esfons in biological systems. New Yo+k: Academic Press, 1970 22. Costertcn JW. Ingram JM, Cheng KJ. Structwe and function of the cell envelope of gram-negative bacteria. Bacterial Rev 1974;38:87-110.