peritoneal dialysis catheters by scanning and transmission

Vol. 18, No. 6JOURNAL OF CLINICAL MICROBIOLOGY, Dcc. 1983, p. 1388-13980095-1137/83/121388-11$02.00/0Copyright ©0 1983, American Society for Microbiology

Examination of the Morphology of Bacteria Adhering toPeritoneal Dialysis Catheters by Scanning and Transmission

Electron MicroscopyTHOMAS J. MARRIE,'* MICHAEL A. NOBLE.' AND J. WILLIAM COSTERTON2

Depatitnentis of lMedi(ine anIdI Microbiolotg'y, DIilIlOiasie UJlin'hersity td7( tlte V'ictoria Genieral Hospital,Htalifa.x, Nova Scotia, Caintada B3H I V8, ' ind tle Deparntmet of Biology, Unitersitv of Calgary, Calgary,

Albe-ta, Canada T2N IN42

Received 13 June 1983/Accepted 7 September 1983

We examined Tenckhoff peritoneal catheters by scanning and transmissionelectron microscopy to study the morphology of bacterial adherence. Twocatheters were removed from uninfected patients, three from patients with exitsite infections, four from patients with peritonitis, and one from a patient withboth exit site infection and peritonitis. Infecting organisms included three ofStaphvlooeus aureits and one each of Enterobacter sp., Staphvloco(e(eus epider-mnidis, Achrotnobaeter xvlosoxidans, Serratia sp., Klebsiella sp., and Calndidaalbica ns. Considerable morphological variation in adherence to the peritonealdialysis apparatus occurred. No inflammatory cells were ever seen in associationwith infected cuffs, only two of the five patients with peritonitis had inflammatorycells associated with their catheters. In both instances, these cells tended to occurin clumps and demonstrated no flattening when in contact with the surface.Colonization of the catheter was uneven-bacteria tended to occur in clusters.Extensive matrix formation was evident in several instances, and condensation ofthis matrix onto the bacteria during the dehydration process rendered clumps ofbacterial cells amorphous at times. Bacteria were adherent to the subcutaneouscuff in those patients with exit site infections. Gram-negative bacteria attached toindividual dacron fibers of the cuff, often several layers deep. Gram-positivebacteria tended to adhere in clusters.

The development of the Tenckhoff catheter(26) has led to the increasing use of continuousambulatory peritoneal dialysis as a means ofmanaging patients with renal failure (27). Themajor limitation to the usefulness of this tech-nique is the high incidence of infection (5, 6, 8,20, 27). Such infections may involve the sitewhere the catheter exits from the abdominal wall(exit site infection) or the peritoneal cavity (peri-tonitis) or both (27). The pathogenesis of exitsite infections is nearly always contaminationfrom adjacent skin (27) or from carriage sites ofpathogenic microorganisms such as the nose(22). Peritonitis can result from breaks in tech-nique, exit site infections, or migration of bacte-ria across the bowel wall (27). Bacterial adher-ence to the surfaces of mammalian cells hasbeen studied extensively, and a consensus isdeveloping that such adherence is important inthe establishment of infection (7, 21, 25). Re-cently, bacterial attachment to various devicesused in the treatment of patients has been stud-ied (17, 18, 23, 24). Differential adherence is

evident-Staphy1olo(o(e(us aluurelis adheres betterto gut sutures than to silk or nylon (24). Surfaceirregularities are preferential sites of attachment(18), and amorphous extracellular materialseems to mediate such bacterial attachment (17,18). Infected peritoneal catheters seemed to usto be ideal to study the morphology of bacterialattachment to a prosthetic device. We wonderedwhether attachment to the dacron cuff would bedifferent than to the silicone rubber tubing. Thedacron cuff is in a relatively dry environment,whereas the tubing is continuously bathed by thedialysate. Further, many patients received anti-biotics for considerable periods of time beforeremoval of the catheter, and we wonderedwhether this therapy would affect the morpholo-gy of the bacteria.

MATERIALS AND METHODS

Catheters. Tenckhoff peritoneal catheters (Lifemed,Division of Verniton. Compton, Calif.) were removedfrom eight patients because of infection: exit siteinfection, three patients; recurrent peritonitis, four

1388

BACTERIA ADHERING TO DIALYSIS CATHETERS

TABLE 1. Characteristics of 10 continuous ambulatory peritoneal dialysis patients

Duration Duration of atntibioticPatient of cathe- Organism treatment before re- Organism isolated

ter in moval of catheter from caitheter or cuffsitu (mo) (days)

1 4 Peritonitis C. albicans 105 4 C. albicaonsCFU/ml

2 4 Exit site peritonitis S. aluzreius 18 S. aurteus. confluentgrowth

3 5 Peritonitis Enterobacter sp. 10 on 2 different oc- Enteohbaciter sp.casions 1 CFU

4 0.5 None evident clini- S. epidermidis 0.5 0 S. epidermil7idiscally, catheter re- CFU/ml 5 CFUmoved because ofpoor drainage

5 4 Exit site S. alureius 7 S. (aitreus (cuff)confluent growth

6 1.7 Peritonitis P. aieru}ginosta, Ser- 7 A. xV-l)osoX-idansratiti sp., A. xylo- (catheter + cuff)soxidans (latestepisode)

7 2 None-removed attime of transplan-tation

188 4.5 Exit site S. alureuis 3 S. aiurieus 2 CFU

19 1 Peritonitis second- Klebsiella sp., L. 6 Klebsiella sp.

ary to renal biop- plantitariume, B.sy firagilis

10 8 Exit site Serratia sp. 0 Se-rratia sp.

TABLE 2. Observations by scanning electron microscopy on the catheters removed from nine of the dialysispatients

ObservationsPatient" See figure:c

Cuff Inner surface of catheter Outer surface of catheter

1 Candidai Amorphous material l

2 Coccoid bacteria Occasional clusters of coc-coid bacteria

3 No bacteria seen Occasional clusters of rod- 7, toplike bacteria

4 No bacteria seen Occasional inflammatorycell

5 Thick layer cocci Microcolonies of cocci 6 and 7.bottom

6 Marked colonization of in- Sparsely colonized with Sparsely colonized with 4 and 5dividual fibers by rod- rod-shaped bacteria, rod-shaped bacilli, in-shaped bacteria some areas of dense col- flammatory cells

onization

7 No bacteria seen Considerable accretion, no Occasional rod-like bacte- 8bacteria seen ria

8 Gram positive cocci Amorphous accretions 3

9 No bacteria seen Sparse bacterial coloniza- Sparse bacterial coloniza-tion (rods). many in- tion. many inflammatoryflammatory cells cells

'Observation of the catheter of patient 10 showed nothing unusual.

VOL. 18, 1983 1389

1390 MARRIE, NOBLE, AND COSTERTON

patients. One patient had both infections, and in twopatients the catheters were removed for other reasons.The catheters were immediately cut into several por-tions and transported to the laboratory.

Culture technique: catheters. The cuff portion and alength of the intraperitoneal portion of each catheter (3to 4 cm) was rolled over the surface of a blood agarplate (Trypticase soy agar [BBL Microbiology Sys-tems, Cockeysville, Md.] containing 5% sheep blood)and a MacConkey agar plate. These plates were incu-bated aerobically for 18 h. All aerobic isolates wereidentified by conventional techniques. In addition, thecuff and catheter were rolled over the surface of a

J. CltIN. MICROBIOL.

prereduced Brucella blood agar plate containing 10 p.gof vitamin K, per ml, 100 pg of kanamycin per ml, and7.5 p.g of vancomycin per ml. These plates wereincubated anaerobically in GasPak jars (BBL) for 48 h.Disposable GasPak hydrogen-carbon dioxide genera-tors were used in the jars to supply a gas mixturecontaining hydrogen and carbon dioxide. Any anaero-bic microorganisms were identified by standard labo-ratory techniques.

Peritoneal fluid. Peritoneal fluid was cultured on allpatients by the technique of Vas et al. (28). In brief,100 ml of fluid was filtered through an 0.22-tm mem-brane filter (Millipore Corp., Bedford, Mass.), and the

Au'J A..

t t tM4 i'

*iSw'

IONWV

A'

..,w .It...,, .t

.~6

-6.

It a X: *~

._p.., A-

FIG. 1. Scanning electron micrograph of the outer surface of the tip of a Tenchoff catheter removed frompatient 1 in Table 1 because of C. albicans peritonitis. Tissue (omentum '?) was adherent to the tip of the catheter.Note the amorphous background. Many 5-jtm spherical structures are embedded in this background. Only C.albicans was isolated in culture, and transmission electron microscopy showed typical yeast cells. The barrepresents 5 tLm.

T

"'I'll.".-tx

',!.'I.

Abo

0, :P"ik. .b--

W... ;

,wAw. .". -

I. *..' ,-P.. A.- 0.1 4 "'

i

1-

-41

VOL. 18, 1983 BACTEKIA ADHERING TO DIALYSIS CATHETERS 1391

'+^,;k4"~~~~~~t4ZM4..~~~~- ; ''S:.

614~~~X#4.~

44

tt e

.4,~~

\-**4it *

adeig to th cufmtra B ros.Br: inet 1 p: th ohrs5 pm.

A4 < - M b * .. .> h .W,t.IS ..-; -e vt2 % t3 9 $ v ;'-$v;~~~~~~~~~~~~~~~~~~~~~W

;'< . 4%,it. mX * <w w; ; ' ¢ X - w ;'- * t * hW , g , *

.A.,*..;.;g-t.;.,2§*> :. >- F .: t x .

_M ¢$42 ,>. tNj ;atiii;| F s + s;4 . s

S f f X t> Xj<triX-,AU*< < s + l!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~T-

%f 47 J 4 , .

r<43 ~~~94 £<j *N $: % , j4 ' ,~~~~~~*v ss4' % 't_ - ''. }mw ' M st ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4

. a5.-64s

v;\~~~~~reoe rmptin2inTbe1Spreclizto of the surac wa evdn (A). Inmn intne,; itwa

difficult to be sure that bacteria were present (A, inset). Occasional 1-,LLm coccold structures were evidentadhering to the cuff material (13 arrows). Bars: inset. I Fim: the others. 5 Vnm.

17- 10 itAoll


filter was put onto a blood agar plate. Another 100 mlof fluid was treated in a similar manner and culturedanaerobically. This was not carried out in an anaerobicchamber.

Scanning electron microscopy. Cuffs and pieces ofthe intraperitoneal portion of the catheter were placedin a fixative solution consisting of 5% glutaraldehydebuffer (0.067 M, pH 6.2) with 0.15% ruthenium red for1 h at room temperature =22°C. The preparationswere washed three times in the buffer and then "metal-lized" by using osmium tetroxide and thiocarbohydra-zide (13). This was followed by dehydration in ethanoland Freon 113 before critical point drying (3). Thespecimens were then examined with a Hitachi S450(Hitachi, Rexdale, Ontario) scanning electron micro-scope.

Transmission electron microscopy. Small pieces werecut from some of the cuffs and material was scrapedfrom the surface of some of the catheters. This materi-al was fixed as outlined above, then washed five timesin the buffer, and dehydrated through a graded acetoneseries. All of the solutions used in processing thesespecimens except the 90 and 100% acetone solutionswere made up to contain 0.5% ruthenium red. Afterfurther dehydrations in propylene oxide, the speci-mens were embedded in Vestopal (R) (Ladd Indus-tries, Burlington, Vt.), sectioned, stained with uranylacetate and lead citrate (19) reinforced with evapo-rated carbon, and examined with a model 801 electronmicroscope (Associated Electronic Industries, HarlowNew Town, England) at an accelerating voltage of 60kV.

RESULTS

Table 1 shows the duration that each peritone-al catheter was in place, the site of infection, the

microorganisms isolated, and the duration ofantibiotic therapy before removal of the cathe-ter. In Table 2, the scanning electron microsco-py findings are summarized. The morphology ofadherence of the various microorganisms to thecatheters and cuffs varied considerably. Onlyone patient had fungal peritonitis (patient 1 inTable 1). The Candida cells (Fig. 1) were embed-ded in an amorphous background. This probablyrepresented the tissue that was adherent to thetip of the catheter at the time of the removal.The morphology of bacterial adherence to thecatheter cuffs is shown in Fig. 2B, 3, and 4. Rod-like bacteria were seen adherent to individualdacron fibers in Fig. 4B. The bacteria wereadherent in layers. Bacterial adherence to theinner and outer surfaces of the catheters variedconsiderably from one area to another (Fig. SAand B). In general, inflammatory cells wereuncommonly observed; however, when presentthey often occurred in large clumps such as inFig. 5 (B). Some microorganisms were associat-ed with an extensive matrix (Fig. 6B). All thesites of this catheter that were colonized showedbacteria embedded in a matrix. This matrixcondensed down around bacteria during thedehydration process, often rendering identifica-tion of adherent structures as clumps of microor-ganisms difficult (Fig. 7A and inset). Figure 7B isa dramatic illustration of this phenomenon; ma-trix enclosed cells and bare cells are shown sideby side.One catheter from a patient who had no

FIG. 3. Transmission electron micrograph of material shaved from the cuff of a peritoneal catheter removedfrom a patient (patient 8, Table 1) who had a S. (aureuts exit site infection. S. iureius was isolated from the cuff.Note the typical gram-positive cell wall of the two bacteria shown. The bar represents 0.1 p.m.

J. CLIN. MICROBIOL.


clinical or bacteriological evidence of infectionhad a few rod-like bacteria adherent to a cleancatheter surface. These structures varied in sizeand shape (Fig. 8). No obvious effects of antibi-otic therapy were noted (Fig. 3 through 6).

DISCUSSIONIn vitro studies of the adherence of coagulase-

negative staphylococci to the surfaces of intra-venous catheters have shown that within a shorttime (5 to 30 min) single bacterial cells attach,

t~o. F t - w

4.,rv;*{, i t ,

A>

'\tSs, ^-L\N>"o

7', &e-t*, ,~t'tf4:t:.~ ~

J;e~~~~~~~

FIG. 4. Scanning electron micrographs of the cuff of a Tenchoff peritoneal catheter removed from patient 6 inTable 1. Aclhrombacterxylosovidans was isolated from this material. (A) A low power view showing the fibrousnature of dacron cuff. (B) Many rod-shaped structures adhering to an individual fiber. It is evident that thebacteria are piled in layers and interconnected by a fibrous matrix. Bar (A). 500 pLm: bar (B). 5 pLm.

VOL. 18, 1983 1393

1394 MARRIE. NOBLE, AND COSTERTON

and microcolonies are evident after 40 to 60 min,culminating in heavy colonization by 6 to 12 h(18). Subsequent events (48 to 96 h) includeerosion of layers of the catheter surface andencasement of the bacterial cells by a slimymaterial (18). Production of this slime-like mate-rial correlates with "clinical significance" ofsuch isolates and is enhanced by casamino acidsand glucose (2).

In marine environments, growth and divisionof bacteria at a solid-liquid interface occurs

quickly and at nutrient concentrations too low topermit growth in the aqueous phase (10). Such afinding explains the earlier observations of Zo-bell (30) of the pronounced tendency of bacteriato grow in close association with surfaces inaquatic environments.

In this and in other studies (14-16), we haveobserved that bacteria adhere to various foreigndevices in the human body in a manner similar tothat shown in both in vitro experiments (2, 18)and in aquatic environments (10). The infected_' 4t 'Pt".-.,.'1 X.4 '' X

t ..*}IN.

4~~~~~~~~~~~e

* .,-. . - 4ANN:wzyr; s.

4-

r,,..........-k...'. :- ....a-~~ V,-; !r,...4,

a' s6? fq > s

;;> # srt; . '^;

I.,

..*,-

. X,

.. , ,,|,,9~~'

v3Se .A

gS,.. . - ^~~~~~~~~~~4,,~~. AL..04¾

FIG. 5. Scanning electron micrographs of two portions of the peritoneal catheter from the same patient as inFig. 4. (A) Outer aspect of the surface. Note the colonization by rod-shaped bacteria. In places, the outline of thebacterial cell is obscured, and towards the upper left hand cornei the bactel-ia are covered completely by anamorphous matrix. Only inflammatory cells are seen covering the inner aspect of another portion of the catheter(B). Both bars repr-esent 5 Lin.

J. Cl-IN. MICROBIOL.

,S.A-

AT

VOL. 18, 1983 BACTERIA ADHERING TO DIALYSIS CATHETERS 1395

44

4t,~~ ~ ~ ~ ~ ~ j

Aa~~~~~~,R; - 41;Rr76 , t ;s ATiX V

X s

I S

Un

VEtI........... ^ \

s;S\s~r '* 4''Y,

~~~~~~~~'4~ ~ ~~'4,

-49~~~~~~~~~~~~-4~~~~~W .1."_ -A,t

, V 4r^iaX

v 2~~~~~~~~~~~~~~~~~~~~~~~'-t~~ ~a{"''$h * C itf,; .' '-

FIG. 6. Scanning electron micrographs of the outer surface of the subcutaneous portion of a peritonealcatheter removed from patient 5 in l'able 1. S. aiur-euis was isolated. A low power view (A) shows that the surfaceof the catheter is obscured. At higher magnification (B) myriads of cocci are evident. An amorphous matrixcovers many of these cells. Bar (A). `' 'imn bar (B). 5 p.m.


peritoneal catheters have a relatively dry surface(cuff) and an aquatic environment (intraperitone-al portion of the catheter). We did not observeany major morphological differences in the ad-herence of bacteria to these two components ofthe catheter nor did we observe any differencesbetween the exterior and interior surfaces of thecatheters.

It is known that S. arelis (27) and Pseiudoino-nas aeruoginosa (8, 11) infections complicatingperitoneal dialysis are associated with a low cure

A~~~~~~~~~

.: ,t, 4*?. b % * *nisJ"SA Wea ,,'tY',-.S, - * .5

.,'..'; ., ,'' 54.. .,:

s; -'

''- 't:. -,,j,aB ls 'E-'

- ... s; -S > S-.A ,. , . , i_ .. .. ,B ;vb -<**9. .' .>,<. tj s rate and frequently necessitate removal of the

catheter. The apparent mechanisms are that S.au-reius predisposes to intraabdominal abscesses(27), and P. aeruiginosa infection is frequentlyassociated with sinus tract infection (11).Whether differences in exopolymer productionby microorganisms adherent to such foreignbodies are also important in this regard remainsto be determined. That such differences exist isevident from an examination of Fig. 4 through 8.Host defense factors are impaired in the pres-

V.

it

Ak'a,

4

FIG. 7. Scanning electron micrographs of portions of peritoneal catheters removed from patient 3 in Table 1(A) and patient 5 in Table 1 (B). These photomicrographs illustrate the effect of condensation of bacterial matrixduring the dehydration process. The clump (A, arrows) which is shown at higher magnification in inset isrecognizable as bacterial cells, with difficulty. The bottom micrograph (B) shows naked cells side by side withones covered by condensed matrix. The bars represent 5 p.m.

J. CLIN. MICROBIOL.

I

,..15:0

m

r


FIG. 8. Scanning electron micrograph of the outer surface of a peritoneal catheter removed from patient 7 inTable 1 shortly after he had undergone successful renal transplantation. The cultures of this catheter werenegative. In only one of many areas examined did we find the rod-like structures shown here. Note theconsiderable variation in morphology. Also note the clean surface of the catheter. The bar represents 5 prm.

ence of a foreign body. It appears that opsoniza-tion is inadequate and that polymorphonuclearleukocytes degranulate on contact with someforeign materials leading to a decrease in theirphagocytic function (29). In addition, humanmonocytes remain spherical when in contactwith foreign materials, a factor which may havesome effect on their efficiency as phagocyticcells (12). The decreased pH and increasedosmolality of the solutions used in peritonealdialysis result in a decrease in activity of bloodleukocytes (4). We observed inflammatory cellson only three catheters (patients 4, 6, and 9;Table 1 and Fig. 5). The appearance of theinflammatory cells in Fig. SB is representative ofthose seen in patients 4 and 9 as well. Note therounded appearance of these cells. Confluentlayers of bacteria or fungi were frequently seenand not one inflammatory cell was evident (Fig.1. 5A, and 6). Inflammatory cells were neverobserved on cuffs (Fig. 4). Certainly inflamma-tion of the exit site was evident in several ofthese patients, and all the patients with peritoni-tis had -500 leukocytes per mm3 of peritonealeffluent. In other studies of such patients withperitonitis, the leukocyte counts have rangedfrom 720 to 3,290 per mm3 (9). More than 50% ofthese cells were monocytes or macrophages (9).An alternate approach to the problem of infec-

tion in continuous ambulatory peritoneal dialy-

sis patients would be to identify the mechanismof binding of bacterial exopolymers to the sur-face and then alter the surface to prevent adher-ence. Incorporation of isothiazalone biocides,which penetrate the polymeric matrix of bacte-ria, into the wall of the catheter may be anotherapproach (1) to the prevention of bacterial ad-herence to such prosthetic devices.

ACKNOWLEDGMENTS

This research was supported by a grant from the MedicalResearch and National Science and Engineering ResearchCouncil of Canada.We thank Joyce Nelligan and Carol Kwan for technical

assistance and Daureen Stover for typing the manuscript.

LITERATURE CITED

1. Characklis, W. G. 1973. Attached microbial growths II.Frictional resistance due to microbial slimes. Water Res.7:1249-1258.

2. Christensen, G. D., W. A. Simpson, A. L. Bisno, and E. H.Beachey. 1982. Adherence of sline-producing strains ofStophylococcus epidermnidis to smooth surfaces. Infect.Immun. 37:318-326.

3. Cohen, A., D. P. Marlow, and G. E. Garner. 1968. A rapidcritical point using fluorocarbon ("Freon') as intermedi-ate and transition fluids. J. Microscop. Paris 7:331-342.

4. Duwe, A. K., S. I. Vas, and J. W. Weatherhead. 1981.Effects of the composition of peritoneal dialysis fluid onchemiluminescence, phagocytosis. and bactericidal activi-ty in vitro. Infect. Immun. 33:130-135.

5. Fenton, P. 1982. Laboratory diagnosis of peritonitis inpatients undergoing continuous ambulatory peritonealdialysis. J. Clin. Pathol. 35:1181-1184.

6. Fenton, S., G. Wu, D. Cattran, A. Wodgymar, and A. F.

VOL. 18. 1983 1397

1398 MARRIE. NOBLE, AND COSTERTON

Allen. 1981. Clinical aspects of peritonitis in patients oncontinuous ambulatory peritoneal dialysis. Periton. Dialy-sis Bull. I(Suppl):54-57.

7. Gibbons, R. J., and J. Van Houte. 1975. Bacterial adher-ence in oral microbial ecology. Annu. Rev. Microbiol.29:19-44.

8. Gokal, R. 1982. Peritonitis in continuous ambulatoryperitoneal dialysis. J. Antimicrob. Chemother. 9:417-422.

9. Hurley, R. M., D. Muogbo, G. W. Wilson, and M. Ali.1977. Cellular composition of peritoneal effluent: re-sponse to bacterial peritonitis. Can. Med. Assoc. J. 117:1060-1062.

10. Kjelleberg, S., B. A. Humphrey, and K. C. Marshall. 1982.Effect of interfaces on small starved marine bacteria.Appl. Environ. Microbiol. 43:1166-1172.

11. Krothapalli, R., B. Duffy, C. Locke, W. Payne, H. Patel,V. Perez, and H. 0. Senekjian. 1982. Pselidwontiotits perito-nitis and continuous ambulatory peritoneal dialysis. Arch.Int. Med. 142:1862-1863.

12. Leake, E. S., M. J. Wright, and A. G. Gristina. 1981.Comparative study of the adherence of alveolar andperitoneal macrophages and of blood monocytes to meth-yl methacrylate, polyethylene, stainless steel and vitalli-um. J. Reticuloendothel. Soc. 30:403-414.

13. Malick, L. E., and B. W. Wilson. 1975. Modified thiocar-bohydrazide procedure for scanning electron microscopy:routine use for normal, pathological or experimental tis-sues. Stain Tech. 50:265-269.

14. Marrie, T. J., and J. W. Costerton. 1983. A scanningelectronmicroscopic study of urine droppers and urinecollecting systems. Arch. Int. Med. 143:1135-1141.

15. Marrie, T. J., and J. W. Costerton. 1983. A scanning andtransmission electron-microscopic study of the surface ofintrauterine contraceptive devices. Am. J. Obs. Gyn.146:384-393.

16. Marrie, T. J., J. Nelligan, and J. W. Costerton. 1982. Ascanning and transmission electron microscopic stuIdy ofan infected endocardial pacemaker lead. Circulation 66:1339-1341.

17. Peters, G., R. Locci, and G. Pulverer. 1981. Microbialcolonization of prosthetic devices. II Scanning electronmicroscopy of naturally infected intravenous catheters.Zbl. Bakt. Hyg. I Abt. Orig. B 173:292-299.

18. Peters, G., R. Locci, and G. Pulverer. 1982. Adherence of

coagulase-negative staphylococci on surfaces of intrave-nous catheters. J. Infect. Dis. 146:479-482.

19. Reynolds, E. S. 1963. The use of lead citiate at high pH aselectron-opaque stain in electron microscopy. J. Cell.Biol. 17:298-342.

20. Rubin, J., W. A. Rogers, H. M. Taylor, D. Everett, B. F.Prowant, L. V. Fruto, and K. D. Nolph. 1980. Peritonitisduring continuous ambulatory peritoneal dialysis. Ann.Int. Med. 92:7-13.

21. Rutter, J. M., M. R. Burrows, R. Sellwood, and R. A.Gibbons. 1975. A genetic basis for resistance to entericdisease caused by E.vcherichia co/i. Nature (London)257:135-136.

22. Sewell, C. M., J. Clarridge, C. Lacke, E. J. Weinman, andE. J. Young. 1982. Staphylococcal nasal carri'age andsubsequent infection in peritoneal dialysis patients. J.Am. Med. Assoc. 284:1493-1495.

23. Sugarman, B. 1982. In vitro adherence of bacteria toprosthetic vascuIlar grafts. Infection 10:9-12.

24. Sugarman, B., and D. Musher. 1981. Adher-ence of bacte-ria to suture materials. Proc. Soc. Exptl. Biol. Med.167:156-160.

25. Svanborg-Eden, C., B. Erikson, U. Jodal, B. Kaijser, G.Lidin, U. Lindberg, L. A. Hanson, and S. Oiling. 1978.Adhesion to normal uroepithelial cells of Escherichia co/ifrom children with various forms of urinary tract infec-tions. J. Pediatrics 93:398-403.

26. Tenckhoff, H., and H. Schechter. 1968. A bacteriiologicallysafe peritoneal access device. Trans. Am. Soc. Artif.Intern. Organs 14:181-186.

27. Vas, S. I. 1983. Microbiologic aspects of chronic aimbuIla-tory peritoneal dialysis. Kid Inter-nat. 23:83-92.

28. Vas, S. I., I). E. Low, S. Layne, R. Khanna, and N.Dombros. 1981. Microbiological diagnostic appr-oach toperitonitis in continuous ambulatory peritoneal dialysispatients. p. 264-267. In R. C. Atkins, N. M. Thomson.and P. C. Fairrell (ed.), Peritoneal diailysis. ChurchillLivingston, New York.

29. Zimmerli, W., F. A. Waldvogel, P. Vaudaux, and U. E.Nydegger. 1982. Pathogenesis of foreign body infection:description and characteristics of an animal model. J.Infect. Dis. 146:487-497.

30. Zobell, C. E. 1943. The effect of solid surfiaces on bacterialactivity. J. Bacteriol. 46:39-56.

J. Cl.lN. MICROBIOL.

peritoneal dialysis catheters by scanning and transmission

Documents