Controlled release of antibiotics from coated
orthopedic implants
J. S. Price,' A. F. Tencer,lf* D. M. Arm,' and G. A. Bohach'
'Biomechanics Laboratory, Harborview Medical Center, Department of Orthopedics, ZA-48, 325 Ninth Avenue,
University of Washington, Seattle, Washington 98104; 'Department of Microbiology, University of Idaho,
Moscow, Idaho 83843
Chronic osteomyelitis is one of the most serious complica-
tions of orthopedic open fracture treatment. The objective
of this study was to develop a biodegradable implant coat-
ing with impregnated antibiotics as an adjunct to current
therapy. We used a polylactic-co-glycolic acid copolymer
(PLGA) as the biodegradable carrier and gentamicin as the
antibiotic. Our objectives were to establish elution charac-
teristics of the antibiotic from the polymer, and determine if
the coated orthopedic implants would inhibit bacterial
growth in vitro. In the elution study, coated implants were
incubated in phosphate buffered saline (PBS) at 37°C and
sampled daily for gentamicin levels. The in vitro model con-
sisted of test tubes containing Mueller-Hinton culture
broth inoculated with 5 x lo6 cfu of Staphylococcus uureus
and incubated at 37°C. The implants were switched to a
INTRODUCTION
Infection remains one of the most serious compli-
cations of orthopedic open fracture treatment. Acute
infection or chronic osteomyelitis develop in as many
as 5 3 3 % of open fractures.'T2 Persistent osteomyelitis
can necessitate amputation if not successfully treated.
Standard open fracture treatment consists of exten-
sive debridement and irrigation, fracture reduction
and fixation, and prophylactic parenteral antibiot-
ics.
Orthopedic infections including those secondary to
open fractures are usually caused by Gram positive
staphylococci (S. aureus, S. epidermidis).lr2 Strepto-
cocci and Gram negative organisms such as Entero-
bacteriaceae and Pseudomonas aeruginosa as well as
some anaerobes are also commonly found.',' Antimi-
crobial treatment needs to be effective against these
pathogens unless other organisms have been specif-
ically isolated. Typically, systemic antibiotic prophy-
*To whom correspondence should be addressed. E-mail:
atencer@u.washington.edu
Journalof Biomedical Materials Research, Vol. 30, 281-286 (1996)
0 1996 John Wiley & Sons, Inc.
new set of inoculated tubes each day. Tubes were sampled
for colony counting to determine bactericidal effects. Im-
plant coatings consisted of 40 mg of gentamicin as a 20%
mixture with PLGA. The elution curve showed an average
level of 138 kg/mL over 15 days. This local concentration
would be more than adequate to kill susceptible organisms.
The in vitro study showed a significant reduction in bacte-
rial growth in the test tubes containing coated implants.
Control tubes averaged 2.5 x 10' cfu/mL of S. uureus over 24
days. Coated implant tubes averaged 0.9 cfu/mL. This was
a reduction of greater than 99.999%(p < 0.0001). This study
showed that a thin biodegradable implant coating can be
developed with bactericidal activity against the organisms
frequently associated with osteomyelitis in cases of open
fractures. 0 1996 John Wiley & Sons, Inc.
laxis includes a broad spectrum first generation ceph-
alosporin (cefazolin) for Gram positive coverage and
an aminoglycoside (gentamicin) for Gram negative
organisms, with inclusion of a penicillin G or a semi-
synthetic penicillin for farm-related or severe crush
injuries that frequently involve Streptococci.'-12 Cip-
rofloxacin is a fluoroquinolone approved for osteo-
myelitis treatment, but should only be used when
Gram negative aerobes are present due to resistance
of Gram positive organisms.' Some recent literature
recommends treating established osteomyelitis with
either nafcillin, clindamycin, or ampicillin/sulbactam
for Gram positive coverage, plus a third generation
cephalosporin.
Systemic antibiotic delivery entails certain draw-
backs such as systemic toxicity, poor penetration into
ischemic or necrotic tissues, and hospitalization to
monitor drug levels and effects. Local antibiotic ther-
apy has become an accepted and common adjunct to
systemic antibiotics for both prophylaxis and existing
infections.'3rs11 This offers the advantages of a high
local antibiotic concentration without systemic toxic-
ity, as well as improved dead space management,
CCC 0021-9304/96/030281-06
282 PRICE ET AL.

early primary wound closure, and more complete acteristics of antibioticsreleased from the carrier poly-
eradication of infections. Various vehicles for local mer, and to determine if PLGA-gentamicin coated
drug delivery have been developed, including poly- implants would inhibit bacterial growth in vitro
methylmethacrylate (PMMA) beads, PMMA and bio-
degradable bone cements, calcium phosphate/
hydroxyapatite, collagen, adsorption of antibiotics
onto implant materials, cancellous bone grafts, plas- MATERIALS AND METHODS
ter of Paris, and biodegradable polymers such as
polylactide-co-glycolide (PLGA), polyanhydrides,
and polyorthoesters (POE).1111-20 Film coated implant fabrication
Antibiotic loaded PMMA beads have become ac-
cepted as a localized treatment for bone infection pro-
PLGA (Lot # 115-27-1, Birmingham Polymers, Inc.,
phylaxis and existing osteomyelitis. 12-15 However,
Birmingham, AL) consisted of a 50:50 weight ratio of
they are nonresorbable and require a subsequent op-
PLA:PGA. Gentamicin sulfate powder (Lot #
eration for removal. PMMA does not biodegrade and
4A6127, Paddock) had a potency of 632.1 kg/mg
therefore only releases antibiotic that is adsorbed to
(63.21%)as gentamicin. PLGA was mixed with gen-
the surface or that diffuses out of pores within the
tamicin sulfate in a methylene chloride solvent me-
matrix. This results in a very low fraction (157.5%)
dium. Stainless steel fracture plates (cat. # 224.09,
of the total incorporated antibiotic being released. Re-
4.5-mm narrow DCP nine hole1151 mm, Synthes, Pa-
duced porosity of the beads inhibits drug release by
oli, PA) were sectioned into 20 pieces, each 31 mm
2550%. I 2 Additional concerns exist over possible for-
long. Ten were coated with solutions of 10,20, and 30
eign body reactions once the antibiotics are depleted.
wt % gentamicin sulfate (Fig. 1)for the elution kinet-
Biodegradable implants offer the advantages of
ics study, and 10 were retained for the in vitro bacte-
bioresorption that obviates removal operations, re-
rial inhibition study (the in vitro study eventually
duced foreign body reactivity, increased total antibi-
utilized 20% gentamicin sulfate coated plates). The
otic release, and release kinetics that can be modified
solvent was evaporated at room temperature and
via degradation rates. Biodegradable drug delivery
pressure for 24 h. Plates underwent 24 h of ethylene
systems such as poly(propylenefumarate/methyl-
oxide sterilization to prevent bacterial contamination.
methacrylate) (PPF-MMA) cement, calcium phos-
Table I shows film thickness and weight, and genta-
phate/hydroxyapatite, collagen, cancellous bone
micin sulfate weight (also expressed as active genta-
grafts, plaster of Paris, and biodegradable polymers
micin) for fracture plates in the kinetics study. The
such as PLGA, polyanhydrides, and POE have been
weight of gentamicin sulfate depended on both the
used. However, these systems require placement of
total film weight and percent antibiotic in the film.
an additional material into the wound. We sought to
Values expressed as gentamicin in parentheses were
develop a biodegradable coating for fracture fixation
calculated based on 63.21% potency. The 20% film
implants, containing impregnated antibiotics to be
was selected for the bacterial inhibition study. Table
used for infection prophylaxis at the time of stabili-
I1 shows similar data for fracture plates used in the in
zation of the fracture. Recent research shows that
vitro bacterial inhibition study.
such coatings can be an effective drug delivery sys-
This could be used for patients at high risk
for soft tissue infection or osteomyelitis such as those
with open fractures, and patients with local tissue,
vascular, or immune system deficiencies.
We used PLGA as the biodegradable carrier and
gentamicin as the antibiotic. PLGA and polylactic
acid (PLA) have been used in several recent studies of
antibiotic delivery to prevent and treat osteomyeli-
ti^.'^'' Implants from these polymers produce high
levels of antibiotic in cortical bone and bone mar-
row.’* PLGA has also been shown to be an effective
delivery system for bone growth factors.” Gentami-
cin and other aminoglycosides are commonly admin-
istered for orthopedic infection control. Their low
therapeutic index and associated oto- and nephrotox-
icity make them good candidates for local delivery, so
the hazards of systemic treatment may be reduced. Figure 1. Examples of (left) PLGA-gentamicin coated and
Our study objectives were to establish elution char- (right) uncoated bone fracture fixation plates.
ANTIBIOTIC RELEASE FROM FRACTURE PLATES 283

TABLE I 100 pL of parent culture broth). Ten coated fracture

Characteristics of Coated Fracture Plate Films Used in plates and 10 uncoated control fracture plates were
Release Kinetics Study placed in tubes and incubated at 37°C on an orbital
~ ~~

Wt % Gentamicin shaker rotating at 50 rpm. A second control consisted

Gentamicin Film Thickness Film Weight Sulfate Weight of 10 inoculated tubes with no implants at all. Every
Sulfate * SD (pm) * SD (mg) 5 SD (mg)
24 h fracture plates were removed from their tubes,
10 167 & 12 377 zk 2 *
38 1 (24) rinsed with sterile PBS, and placed into a fresh set of
20 131 2 23 289 t 4 58 c 1 (37) inoculated tubes. Aliquots were removed from the
30 *
186 36 348 2 41 *
105 13 (66) previous day's tubes for microbiological testing. All
0 (Control) 195 398 0
work was done under sterile conditions.
Each value represents average of three plates, except sin- Optical density (OD) measurements and serial di-
gle control plate. Numbers in parentheses are weight of lution colony counts were performed to determine
gentamicin alone. bacterial concentrations in the tubes after incubation
with the fracture plates. OD values were measured
Release kinetics using a spectrophotometer (Spec-20, Bausch & Lomb)
at 610 nm. Colony counts were performed by serially
An elution study was performed to determine the diluting aliquots of broth with PBS and inoculating
release kinetics of gentamicin from the PLGA carrier. petri dishes of Mueller-Hinton agar. Petri dishes
Coated implants were prepared using 10, 20, and were incubated for 24 h at 37°C before counting and
30% gentamicin sulfate solutions. These were each only those dilutions producing between 10 and 300
incubated in 10 mL of pH 7.4 phosphate-buffered sa- counts were recorded. All counts were done in trip-
line (PBS) at 37°C. Aliquots were removed for genta- licate. Phage typing was not performed, although se-
micin assay at 6 h, and 1, 3, 6, 9, 12, 15, 20, and 26 lected colonies were Gram stained to look for Gram
days after the start of incubation. Samples were positive cocci. Bacterial concentrations were com-
stored at - 20°C until submitted for analysis. Assays pared between tubes with coated implants and those
were performed via fluorescence polarization immu- with uncoated controls to assess the films' bacteri-
noassay (TDx, Abbott Laboratories, Chicago, IL). Ev- cidal effects. Inoculated tubes without implants were
ery 24 h fracture plates were placed in a fresh solution compared with the first set of controls to determine
of PBS. The gentamicin sulfate concentration to be the effect of uncoated fracture plates on bacterial
used in the bacterial inhibition study needed to pro- growth. The experiment was maintained for 24 days.
vide the highest elution rates from 24 h through 14-21
days of use.
RESULTS

Bacterial inhibition study Release kinetics

Staph!!'ococcus aUreuS (American Type '01- The 10, 20, and 30% gentamicin sulfate:PLGAfilms
lection # 25923, antibiotic susceptibility testing mas- produced distinctly different pharmacokinetics, Fig-
ter culture, Harborview Medical Center Microbiology ure illustrates the elution for the three mix-
Laboratories, WA) was for the bat- tures. The 10% films showed an average daily level of
terial inhibition study due to its presence in the ma- 21 pg/mL Over 2o days. The 20% films eluted an av-
jority of orthopedic infections. S. aureus was main- erage of 133 pg/mL. The 30% films produced a mean
tained on Mueller I' agar (cat' # 11438f BBLt level of 374 pg/mL, Each film was significantly differ-
Becton Dickinson, Cockeysville, MD) and used to in- ent than the other two films (p < 0.05). ~ 1 three1
oculate a parent culture broth. The parent culture in mixtures exceeded the minimum inhibitory concen-
Mueller-Hinton broth (cat. # 11443, BBL, Becton
Dickinson) was replaced every 24 h and maintained
TABLE I1
at a concentration of 8.2 x lo8 cfu/mL. The minimum Characteristics of 20% Gentamicin Sulfate Coated
inhibitory concentration Of gentamicin for s. Fracture Plate Films Used in Bacterial Inhibition Study
aureus is 2 4 pg/mL. Implants coated for this study
Gentamicin
had PLGA-gentamicin sulfate films that were 169 pm Film Thickness Film Weight Sulfate Weight
thick and weighed 318 mg with 40 mg of active gen- SD (pm) * SD (mg) +- SD (mg)
tamicin.
The in uitro model consisted of 20 test tubes, each 169 k 11 318 5 12 64 * 2 (40)
containing 10 mL of h'heller-Hinton broth. Tubes All values are averages of 10 plates. Number in paren-
were inoculated with 9.6 x lo6 cfu of S. aureus (about theses are values as gentamicin.
284 PRICE ET AL.

Coated implants were selected to have a 20% genta-

=
.
E
3 0 '3'0 ",,,
30% coat = 2077 pglml.
micin sulfate film that contained 40 mg of active gen-
tamicin per implant.

Bacterial inhibition
2001 20%
The in vitro study showed a significant reduction in
100
1 bacterial growth in the test tubes containing coated
implants over both uncoated control implants and
-0
n
3 6 9 12 15 18
control tubes with no implants (Fig. 4). Coated im-
plant tubes averaged only 1.2 cfulml. This was a re-
Time (days) duction of greater than 99.999% ( p < 0.0001) from the
inoculating dose of 9.6 X lo5 cfu/mL (9.6 x lo6 cfu
Figure 2. Mean daily gentamicin concentration in elution
sample. total). A reduction of at least 99.9% is considered
proof of minimum bactericidal concentration (MBC)
in antibiotic susceptibility tests." The coated im-
tration of 1-4 p&mL needed to inhibit organisms
plants eluted gentamicin to levels at or above the
commonly causing orthopedic infections.
MBC on each day of the 24-day study.
Figure 3 shows the antibiotic remaining on the
Control tubes with uncoated implants averaged 2.5
coatings as a percentage of the total. This was calcu-
x 10' cfu/mL of S. uureus over 24 days. Control tubes
lated based on subtracting weighted average daily
with no implants averaged 2.2 X 10' cfu/mL when
elutions from the initial loading weight of antibiotic.
tested on day 13. Uncoated fracture plates did not
Films with higher antibiotic concentrations eluted the
inhibit bacterial growth and actually supported
antibiotics at a disproportionately faster rate. This re-
growth, as concentrations were significantly higher ( p
sulted in early depletion of the 30% coatings, which
< 0.01) in tubes with uncoated fracture plates than in
may explain why their elution curve dropped beIow
control tubes without any implants. It was also noted
the 20% coating after 12 days.
that by day 3, test tubes could be inoculated with S.
The 20% film was selected for the bacterial inhibi-
uureus simply by transferring control fracture plates
tion study because it developed a significant initial
into fresh broth. Fracture plates were triple rinsed
burst release and maintained high concentrations
with PBS but not sterilized at the time of transfer. A
through the end of the time period. The 10% coating
biofilm on the implants was not visibly obvious, but
was not selected because its antibiotic levels were too
was not investigated histologically.
low to be effective in a larger expected in viuo volume.
The 30% coating eluted with a high level burst effect
at 2-4 days (peak concentration of 2077 pg/mL), but
by day 12 had fallen below the 20% coating, with DISCUSSION
most of the antibiotic apparently eluted. It lacked suf-
ficiently high concentrations at the end of the study Orthopedic soft tissue infections and osteomyelitis
to be effective for 3 weeks in a large in uivo volume. are serious complications of open fracture injuries.
los I

-
I
E

;1 0 ' - =
innoculum
I

'=0E 106-
c
5 lo5-
104-
0
u) lo3,
5
m
10'-

vj 10' -
gentamicin coated plates
: loo- 0

"
0 3 6 9 12 15 18 0 5 10 15 20 25
(Cfu value of 10"-1
Time (days) Time (days) represents zero)

Figure 3. Antibiotic remaining on coating as a percentage Figure 4. Mean bacterial concentration on uncoated frac-
of the total applied. ture plates and gentamicin coated plates.
ANTIBIOTIC RELEASE FROM FRACTURE PLATES
Such infections occur in up to 33% of treated injuries.
S. uureus is the most common causative organism,
with Streptococci and Gram negative species being
isolated with increasing frequency. Grade I11 open
fractures show the highest incidence of such compli-
cations. Prompt surgical debridement, fracture fix-
'r2
ation, and prophylactic antibiotics are important
treatment modalities. However, systemic toxicity and
poor penetration into ischemic tissues reduce the ef-
fectiveness of oral or parenteral therapy. The use of
local antibiotics to supplement systemic therapy has
been shown to be effective in reducing infectious
complications. This offers the advantages of high lo-
cal and low serum antibiotic levels, improved dead
space management, early primary wound closure,
and more complete elimination of infections.
PMMA-antibiotic beads are commonly used as a
local drug delivery device. While offering the benefits
of local therapy, they do require an additional re-
moval procedure, and large amounts of antibiotic re-
main trapped within the bead matrix where it is in-
effective. Foreign body reactions may also occur after
drug depletion. Several biodegradable delivery sys-
tems have been developed to address the shortcom-
ings of PMMA and other nonresorbable materials.
This study utilized PLGA as a biodegradable carrier
polymer for gentamicin sulfate. PLGA was selected
for its well-characterized degradation into common
metabolites and because its in v i m use is common in
resorbable sutures. Several researchers have recently
investigated this polymer for its potential as an anti-
biotic delivery method.16-19
We selected stainless steel fracture plates to coat
with the PLGA-gentamicin sulfate film. Bacteria are
known to colonize metal orthopedic implants and
form adherent biofilms that retard the enetration of
antibiotics to the underlying infection.'-z6 Addition-
ally, osteomyelitis is known to develop preferentially
beneath the surface of fixation devices adjacent to
b ~ n e . ' ~ , Coating
'~ implants with a bactericidal film
would inhibit bacteria from colonizing the implant
surface and provide a high antibiotic concentration in
a region commonly found as a nidus for infection.
Similarly high levels would be expected in the sur-
rounding soft tissues.
Our objectives were to establish elution character-
istics of antibiotics released from the carrier polymer,
and to determine if PLGA-gentamicin sulfate coated
implants would inhibit bacterial growth in uitro. Ten,
20, and 30% gentamicin sulfate films were fabricated
onto fracture plates. The kinetics study showed sig-
nificant differences between the three films. The 10%
mixture did not elute enough antibiotic to be effective
in v i m . The 30% film eluted with an initially high
burst effect, but soon dropped below the level of the
20% film and could not maintain sufficiently high
concentrations for the desired 20-day period. The
285
20% film performed best, creating an immediate burst
effect of about 200 kg/mL while maintaining levels of
at least 40 @mL through the end of the study (av-
erage of 133 pg/mL). It is expected that this profile
will provide immediate prophylaxis for wound con-
tamination while continuing to provide benefit until
wound closure is achieved.
During the in uitro study the antibiotic-coated frac-
ture plates achieved a highly significant reduction in
bacterial growth compared to controls. Test tubes
with coated implants averaged only 1.2 cfu/mL of S.
aureus that was a reduction of more than 99 -999%(p <
0.0001) from the inoculating dose. This far exceeded
the results needed to demonstrate that an MBC had
been attained in ~itro.'~ Control tubes with uncoated
implants averaged 2.5 X 10' cfu/mL, or a 25-fold in-
crease over the inoculating dose in 24 h. Control
tubes without any implants had slightly less growth,
and substantial growth occurred after day 3 by trans-
ferring uncoated fracture plates that had been in
growing cultures to new medium. This supported the
idea that bacteria colonize stainless steel surfaces as
well as autoinoculate a sterile growth medium.
In summary, this study showed that a thin biode-
gradable implant coating can be developed with bac-
tericidal activity against the organisms frequently as-
sociated with osteomyelitis. These implants could be
used prophylactically to reduce the incidence of soft
tissue and bone infections that frequently complicate
open fractures. They should offer the additional ben-
efit of reducing implant colonization rates that neces-
sitate costly implant removal or replacement.
References
1. P. A. W. Ostermann, S. L. Henry, and D. Seligson,
"The role of local antibiotic therapy in the manage-
ment of compound fractures," Clin. Ovthop. Rel. Res.,
295, 102-111 (1993).
2. P. A. W. Ostermann, S. L. Henry, and D. Seligson,
"Timing of wound closure in severe compound frac-
tures," Orthopedics, 17, 397-399 (1994).
3. W. J. Leach and N. I. L. Wilson, "Trends in infection
prophylaxis in orthopedics," 1. R. Coll. Suvgeons (Ed-
inburgh), 37, 265-266 (1992).
4. J. T. Mader, G. C. Landon, and J. Calhoun, "Antimi-
crobial treatment of osteomyelitis, Clin. Ortkop. Rel.
Res., 295, 87-95 (1993).
5. P. J. Sanderson, "Preventing infection in orthopedic
implants," 1. Antimicvob. Chernother., 24, 277-280
(1989).
6 . J. H. Calhoun, S. L. Henry, D. M. Anger, J. A. Co-
bos, and J. T. Mader, "The treatment of infected non-
unions with gentamicin-polymethylmethacrylateanti-
biotic beads," Clin.Orthop. Rel. Res., 295, 2 S 2 7 (1993).
7. J. H. Calhoun, K. Klemm, D. M. Anger, J. A. Cobbs,
and J. T. Mader, "Use of antibiotic-PMMA beads in
the ischemic foot," Orthopedics, 17, 453-458 (1994).
8. J. T. Mader and G. Ciemy 111, "The principles of the
286
use of preventive antibiotics,” Clin. Orthop. Rel. Res.,
190, 75-82 (1984).
9. M. J. Patzakis and J. H. Calhoun, “Symposium: Cur-
rent concepts in the management of osteomyelitis,”
Contemp. Orthop. 28, 157-185 (1994).
10. J. T. Mader and J. Calhoun, “Long-bone osteomyeli-
tis: Diagnosis and management,” Hosp. Pract., Octo-
ber 15, 71-86 (1994).
31. T. N. Gerhart, R. D. Roux, P. A. Hanff, G . L. Horow-
itz, A. A. Renshaw, W. C. Hayes, “Antibiotic-loaded
biodegradable bone cement for prophylaxis and treat-
ment of experimental osteomyelitis in rats,” J . Orthop.
Res., 11, 25C-255 (1993).
12. C. L. Nelson, S. G. Hickmon, and B. H. Harrison,
”Elution characteristics of gentamicin-PMMA beads
after implantation in humans,” Orthopedics, 17, 415-
416 (1994).
13. F. R. DiMaio, J. J. OHalloran, and J. M. Quale, ”In
vitro elution of ciprofloxacin from polymethyl-
methacrylate cement beads,” J. Orthop. Res., 12, 79-82
(1994).
14. D. K. Kuechle, D. M. Musher, G. C. Landon, and
P. C. Noble, ”Release of vancomycin, amikacin, and
daptomycin from acylic bone cement: Elution dynam-
ics and effects of porosity,” Trans. 36th meeting Orthop.
Res. SOC.,15, 462 (1990).
15. T. J. Nalty, P. C. Noble, G. C. Landon, andH. S. Tul-
los, “Factors affecting the pharmacokinetics of drug
delivery from PMMAiantibiotic composites,” Trans.
36th meeting Orthop. Res. SOC., 15, 295 (1990).
16. W. Chao, S. S. Lin, W. H. Charleton, S. A. Shah,
R. S. Tuan, “Biodegradable drug delivery system ap-
plied onto fracture fixation plates and intramedullary
rods: In viva comparative analysis,“ Trans. 41st meeting
Orthop. Res. SOC.,20, 299, 1995.
17. C. Teupe, R. Meffert, and S. Winckler, “Ciprofloxa-
cin-impregnated poly-L-lactic acid drug carrier,” Arch.
Orthop. Trauma Surg., 112, 33-35 (1992).
18. J. A. Miyano, M. Ramchandani, D. H. Robinson,
K. L. Garvin, “Bone penetration of ciprofloxacin de-
PRICE ET AL.
livered by biodegradable PLGA implants,” Trans. 40th
meeting Orthop. Res. SOC., 19, 774 (1994).
19. G. Wei, Y. Kotoura, M. Oda, et al., “A bioabsorbable
delivery system for antibiotic treatment of osteomyeli-
tis: The use of lactic acid oligomer as a carrier,” J. Bone
Joint Surg., 73B, 246-252 (1991).
20. D. M. Arm and A. F. Tencer, “Effect of cyclic mechan-
ical stress on the controlled release of proteins from a
biodegradable polymer implant,“ Trans. 21st meeting
Soc. Biomater., 18, 130, 1995.
21. A. G. Gristina and J. W. Costerton, ”Bacterial adher-
ence and the glycocalyx and their role in musculoskel-
etal infection,” Orthop. Clin. North Am., 15, 517-535
(1984).
22. K. Merritt, S. A. Brown, J. H. Payer, D. H. Ryerson,
“Influence of bacteria on corrosion of metallic bioma-
terials,” Trans. 17th meeting SOC.Biomater., 14, 106,
1991.
23. A. J. Barton, R. D. Sagers, W. G. Pitt, ”Bacterial ad-
hesion and growth on orthopedic polymers,“ Trans.
20th meeting Soc. Biomater., 17, 348, 1994.
24. A. B. Prewett, J. M. Domenick, N. Tsang, R. K.
O’Leary, A. U. Daniels,” Differential adherence of
three clinical isolates of Staphylococcus to various
metal surfaces,” Trans. 27th meeting SOC.for Biomater.,
14, 105, 1991.
25. L. X. Webb, J. Holman, B. de Araujo, D. J. Zaccaro,
E. S. Gordon, “Antibiotic resistance in staphlococci
adherent to cortical bone,” J. Orthop. Trauma, 8,28-33,
1994.
26. C. C. Chang, K. Merritt, ”Infection at the site of im-
planted materials with and without preadhered bac-
teria,” J. Orthop. Res., 12, 526-531, 1994.
27. J. C. Sherris, ed., Medical Microbiology, Appleton &
Lange, Norwalk, CT, 1990.
Received February 13, 1995
Accepted June 20, 1995