Biofilm formation and antimicrobial susceptibility of staphylococci and

enterococci from osteomyelitis associated with percutaneous

orthopaedic implants

Magdalena Zaborowska,1,2* Jonatan Tillander,1,3* Rickard Branemark,1,4,5 Lars Hagberg,1,3

Peter Thomsen,1,2 Margarita Trobos1,2
1
Biomatcell Vinn Excellence Center of Biomaterials and Cell Therapy, PO Box 412, 405 30 Gothenburg, Sweden
2
Department of Biomaterials, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg,
Sweden
3
Department of Infectious Diseases, Institute of Biomedicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg,
Sweden
4
Department of Orthopaedics, Sahlgrenska University Hospital, Gothenburg, Sweden
5
Department of Orthopaedics, International Center for Osseointegration Research Education and Surgery (iCORES),
University of California, San Francisco

Received 15 June 2016; revised 24 August 2016; accepted 2 October 2016

Published online 00 Month 2016 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.33803

Abstract: Staphylococci and enterococci account for most
deep infections associated with bone-anchored percutaneous
implants for amputation treatment. Implant-associated infec-
tions are difficult to treat; therefore, it is important to investi-
gate if these infections have a biofilm origin and to
determine the biofilm antimicrobial susceptibility to improve
treatment strategies. The aims were: (i) to test a novel combi-
nation of the Calgary biofilm device and a custom-made sus-
R
ceptibility MIC plate (SensititreV), (ii) to determine the biofilm
formation and antimicrobial resistance in clinical isolates
causing implant-associated osteomyelitis, and (iii) to describe
the associated clinical outcome. Enterococci and staphylococ-
ci were characterized by microtitre plate assay, Congo Red
Agar plate test, and PCR. Biofilm susceptibility to 10 antimi-
crobials and its relationship to treatment outcomes were
determined. The majority of the strains produced biofilm in
vitro showing inter- and intraspecies differences. Biofilms
showed a significantly increased antimicrobial resistance
compared with their planktonic counterparts. Slime-
producing strains tolerated significantly higher antimicrobial
concentrations compared with non-producers. All seven
staphylococcal strains carried ica genes, but two did not pro-
duce slime. The degree of biofilm formation and up-
regulated antibiotic resistance may translate into a variable
risk of treatment failure. This new method set-up allows for
the reproducible determination of minimum biofilm eradica-
tion concentration of antimicrobial agents, which may guide
future antimicrobial treatment decisions in orthopaedic
implant-associated infection. V C 2016 Wiley Periodicals, Inc. J
Biomed Mater Res Part B: Appl Biomater 00B: 000000, 2016.
Key Words: biofilm, MIC, MBEC, Staphylococcus, Enterococ-
cus, biomaterial-associated infections

How to cite this article: Zaborowska, M, Tillander, J, Branemark, R, Hagberg, L, Thomsen, P, Trobos, M 2016. Biofilm
formation and antimicrobial susceptibility of staphylococci and enterococci from osteomyelitis associated with percutaneous
orthopaedic implants. J Biomed Mater Res Part B 2016:00B:000000.

INTRODUCTION morbidity, costs, and treatment complications. Osseointegra-

Implanted medical devices have revolutionised the treat-
ment of musculoskeletal disorders. Nevertheless, by virtue
of its physiochemical properties and adsorbed host-derived
proteins, the material surface also promotes biolm devel-
opment. Subsequent bone tissue infections lead to increased
tion, the direct contact between bone and the implant sur-
face, provides a stability to the implant and might offer
resistance to device-associated infection by tight bone adap-
tation.1,2 Bone-anchored percutaneous implants for amputa-
tion treatment improves quality of life and decreases the

Additional Supporting Information may be found in the online version of this article.
*Both authors contributed equally to this work.
Correspondence to: M. Trobos; E-mail: margarita.trobos@biomaterials.gu.se
Contract grant sponsors: The BIOMATCELL VINN Excellence Center of Biomaterials and Cell Therapy; the Vastra Go
taland Region; the Swedish
Research Council (K2015-52X-09495-28-4); the LUA/ALF Research Grant Optimization of osseointegration for treatment of transfemoral
amputees (ALFGBG-448851); the IngaBritt and Arne Lundberg Foundation; the Wilhelm and Martina Lundgren Foundation; the Handlanden
Hjalmar Svensson Foundation; the Adlerbertska Foundation; Doctor Felix Neubergh Foundation; and the Area of Advance Materials of Chalmers
and GU Biomaterials within the Strategic Research Area initiative launched by the Swedish Government.

C 2016 WILEY PERIODICALS, INC.

V 1
frequency of skin complications compared to the socket
solution.3 However, the method involves a permanent skin
breach with a protruding titanium component, providing a
potential entry point for the local skin ora (e.g., Staphylo-
coccus spp.), and faecal microbiota (e.g., Enterococcus spp.)
(Lenneras et al. submitted for publication). Although the
extraction of the prosthesis is relatively uncommon, failures
irrespective of cause (aseptic or septic) should be evaluated
in order to understand the underlying mechanisms.
Although insufciently understood, the failure of the pros-
thesis due to microorganisms is likely to involve both
material-cell interactions and communication between host
defence cells and bacteria.4 The systematic method and
overall prospective outcome data in 51 patients suggest that
the 2-year risk of deep infection and septic implant removal
in these patients with femoral amputation prostheses is
approximately 8 and 2%, respectively.3 Clinical management
of prosthetic joint infection and osteomyelitis is aided by
working denitions5,6 and classications.7 Clinical deni-
tions of osteomyelitis around percutaneous osseointegrated
implants are yet to be established. Ultimately, a denitive
diagnosis relies on high accuracy sampling and analysis of
infected bone tissues proving the presence of the causative
organism(s). Staphylococcus aureus and coagulase-negative
staphylococci (CoNS), are responsible for 90% of all
biomaterial-associated infections (BAI).8 Other bacteria fre-
quently isolated are enteric bacilli and streptococcal spp.
including enterococci.9 In chronic bone infection including
prosthetic joint infection, antibiotics alone will almost uni-
formly prove inadequate10,11 and clinical resolution does
not always correspond to the elimination of all bacteria,
that is, microbiological resolution.12
The biolm growth mode is largely accepted to be cen-
tral in persistent infections associated with biomaterials.13,14
However, directly linking in situ biolm to pathogenesis and
treatment outcome is difcult.14,15 Surface-attached bacteria
secrete extracellular polymeric substances (EPS), such as
polysaccharides, extracellular DNA and supportive pro-
teins.16,17 The accumulation phase in staphylococcal biolms
is mainly regulated by the chromosomal ica operon,18,19
which encodes for polysaccharide intercellular adhesin
(PIA). In enterococci, no such primarily responsible gene(s)
has as yet been identied, where additional determinants
may contribute to biolm formation.20 The EPS prevent
immune cells13 from reaching the bacterial cells in the bio-
lm. Moreover, bacterial subpopulations (persister cells)
have a low metabolism and reduced antibiotic uptake, espe-
cially to cell wall active antibiotics, that is, beta-lactams and
glycopeptides.2124 At present, the treatment of BAI is rou-
tinely based on pathogen-specic antimicrobial disc diffu-
sion tests or minimum inhibitory concentration (MIC). MIC
reects the antimicrobial susceptibility of planktonic bacte-
ria and not that of bacterial biolm. The in vitro biolm
effect of antimicrobial agents yielding minimum biolm
eradication concentration (MBEC) can be assessed using the
Calgary Biolm Device assay.25 Several staining methods for
biolm quantication, including crystal violet and Congo
red, have been described.2628 Crystal violet is a basic dye
2 ZABOROWSKA ET AL.
TABLE I. Patient Demographics
Demographics
Number of patients (women/men) 11 (3/8)
Number of implants 11
Reason for amputation (trauma/tumor) 10/1
Femoral amputation level (high/mid/low) 3/7/1
Age (years) at time of diagnosis 42 (2271)
of osteomyelitis (median) (range)
Time (months) since implantation 47 (2143)
(median) (range)
Duration (months) of antibiotic 4 (1.58)
treatment (median) (range)
Number of patients with definite osteomyelitis 8
(1signs/symptoms, 1X-ray, 1culturesa)
Number of patients with probable osteomyelitis 2
(1 signs/symptoms, 6X-ray, 1 cultures)
Number of patients with possible osteomyelitis 1
(1 signs/symptoms, 6X-ray, 2 cultures)
a
Two or more positive bone and/or bone marrow cultures out of
five, yielding indistinguishable bacteria in routine identification.31
that binds to negatively charged cell surface molecules as
well as polysaccharides of the extracellular matrix, and is
considered a measure of total biolm biomass.27 The detec-
tion of the ica operon by PCR followed by gel electrophore-
sis has been used to demonstrate the potential of
staphylococcal strains to produce extracellular matrix.17,29
R
The SensititreV customisable microbroth dilution plates for
MIC determination follow the Clinical and Laboratory Stand-
ards Institute (CLSI) guidelines and are commonly used by
clinical laboratories and global surveillance programmes.30
The aims of the present retrospective study were: (i) to
test a novel combination of the Calgary Biolm Device
(CBD) and a commercial susceptibility MIC plate
R
(SensititreV) using clinical isolates causing osteomyelitis
associated with percutaneous bone-anchored prosthesis, (ii)
to characterize the virulence of these strains in terms of
their biolm formation capacity and antimicrobial resis-
tance, and (iii) to describe the clinical outcome of these
patients.
MATERIALS AND METHODS
Bacterial isolates, patients, and clinical denitions
Upon inventory, 13 isolates (4 S. aureus, 3 CoNS, and 6
Enterococcus faecalis) originally obtained from bone biop-
sies, marrow aspirations and inner implant components
remained at the Clinical Bacteriological Laboratory at Sahl-
grenska University Hospital (Gothenburg, Sweden) or the
Culture Collection University of Gothenburg (CCUG). These
strains were isolated from 11 patients suffering from osteo-
myelitis surrounding the femoral implant between March
2008 and April 2012 (for basic demographics see Table I).
In two patients, primary infection was caused by two differ-
ent bacterial species: CoNS CCUG 64518 and E. faecalis
CCUG 64517; S. aureus CCUG 64514 and E. faecalis CCUG
64515. The strains were cultured on 5% horse blood
Columbia agar (Media Department, Clinical Microbiology
Lab, Sahlgrenska University Hospital, Gothenburg, Sweden)
BIOFILM FORMATION AND ANTIMICROBIAL SUSCEPTIBILITY

followed by sub-culturing to ensure purity. The CoNS strains
were further identied as S. epidermidis using the API Staph
(bioMerieux SA, Marcy-lEtoile, France). The antimicrobial
treatment administered to the patients was according to
disc diffusion susceptibility testing. Diagnosis, treatment and
outcome information was thoroughly extracted from patient
records. Infection was dened and graded by signs and
symptoms of deep infection, X-ray ndings and tissue cul-
turing, as previously described (Table I).31 A 03 patient
outcome score was used in order to group the patients
according to the number of complications (relapse, reinfec-
tion, and implant extraction). Treatment failure was dened
as either relapse within the study period or unresponsive-
ness to given antimicrobial treatment leading to implant
extraction. Reinfection, that is, the recovery of a different
organism after completed antimicrobial treatment with clini-
cal resolution, was not regarded as treatment failure. The
study protocol was approved by the Regional Ethics Review
Board of Gothenburg (Dnr 434-09).
Biolm-production ability
The total biolm biomass produced by each strain was
quantied using the crystal violet microtiter plate assay and
expressed as a biomass score. In addition, Syto9 uores-
cence staining was used to target cells within the biolm
biomass. Slime-production was determined by Congo red
agar (CRA) plate test and expressed as a slime score. The
presence of icaA and icaD genes in the staphylococcal
strains was evaluated by polymerase chain reaction (PCR).
An overall biolm production score (range 05), based on
the sum of the biomass and slime scores was assigned to
each strain in order to stratify the biolm categories.
Microtiter plate assay (crystal violet assay). The strains
were cultured on 5% horse blood Columbia agar overnight
at 378C. A solution of 108 colony-forming units (CFU 3
mL21) was created by suspending isolated colonies in tryp-
tic soy broth (TSB) (Trek Diagnostic, East Grinstead, UK)
until an optical density (OD, 546 nm) of 0.13 for S. aureus,
0.25 for S. epidermidis, and 0.27 for E. faecalis. Suspensions
were adjusted to an inoculum concentration of 105 CFU 3
mL21. The microtiter plate assay has been described else-
where,32 in brief: 200 mL of the inoculum of each strain was
dispensed into four wells of a at-bottom 96-well plate
(Nunc, Thermo Fisher, Gothenburg, Sweden) and incubated
at 378C for 24 h. The plate was inverted, rinsed (33), and
stained with 0.1% crystal violet solution (Scharlau, Sentme-
nat, Spain) for 10 min, then rinsed and eluted in 95% etha-
nol. The absorbance was measured at 600 nm in a plate
reader (FLUOstar Omega, BMG Labtech, Offenburg, Germa-
ny). The experiment was repeated three times. The cutoff
value (ODc) was dened as three standard deviations (SD)
above the mean OD of the blank (TSB): ODc 5 average OD of
blank 1 (33 SD of blank). The strains were classied as
previously described by Christensen et al.27 and further cat-
egorised by our own biolm biomass scoring (range 03).
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2016 VOL 00B, ISSUE 00
ORIGINAL RESEARCH REPORT
Fluorescence. The strains were cultured under the same
conditions and concentrations as for the microtiter plate
assay. From a working solution of Syto9 : saline (3:1000)
(Molecular Probes, Eugene, OR, USA), 200 mL was added to
each biolm grown on the bottom of a 96 microtiter well
plate and incubated in the dark at room temperature for 30
min. The wells were rinsed in a water bath and read in a
plate reader for uorescence top reading, using an excita-
tion lter of 485 nm and an emission lter of 520 nm.
Congo Red agar (CRA) plate test. CRA plates were pre-
pared by adding 0.8 g of Congo red (Sigma) and 36 g of sac-
charose (Sigma-Aldrich, St. Louis, MD, USA) to 1 L of brain
heart infusion agar (Oxoid, Basingstoke, Hampshire, UK). To
investigate the slime-producing ability, the strains were cul-
tured on CRA plates and incubated for 24 and 48 h at
378C.26 A previously described six-color reference scale was
adopted where intensely black, black, and almost black colo-
nies are formed on CRA by slime-producing strains and bor-
deaux, red, and intensely red colonies are formed by non-
slime-producing strains.29
DNA extraction, polymerase chain reaction (PCR), and gel
electrophoresis. As described above, isolated colonies from
each strain were suspended in MuellerHinton broth (MHB)
(Trek Diagnostics, East Grinstead, UK) until specic OD val-
ues equivalent to 108 CFU 3 mL21 were reached. From the
bacterial suspensions, 1.5 mL of each strain was centrifuged
at 16,000g for 1 min, and DNA was extracted from the pel-
let using the Gene Elute Bacterial Genomic DNA kit (Sigma,
St. Louis, MO, USA). The quality and concentration of the
isolated DNA were assessed by a nanospectrophotometer
(Pearl; Implen GmbH, Munich, Germany). The following
primers were used for the detection of icaA: 50 -
TCTCTTGCAGGAGCAATCAA as the forward primer and 50 -
TCAGGCACTAACATCCAGCA as the reverse primer.29 For the
detection of icaD, the following primers were used: 50 -
ATGGTCAAGCCCAGACAGAG as the forward primer and 50 -
CGTGTTTTCAACATTTAATGCAA as the reverse primer.29 A
multiplex PCR kit was used (Qiagen GmbH, Hilden, Germa-
ny). PCR products were then analysed by 2.2% agarose gel
electrophoresis (Lonza, Rockland, ME, USA). The S. epidermi-
dis ATCC 35984 and ATCC 12228 strains were used as posi-
tive and negative controls, respectively.
MBEC assay: method for susceptibility testing of
biolms
The Calgary Biolm Device for biolm growth and a micro-
R
broth dilution plate (SensititreV), which adheres strictly to
CLSI and ISO guidelines for MIC susceptibility testing, were
combined for the determination of the MBEC of all strains
(Figure 1).
Selection of antibiotics. Ten different antimicrobial agents
commonly used in staphylococcal or enterococcal infections
were chosen.33,34 A custom-made antimicrobial panel
R R
(SensititreV CML1FNUN, TREKV, Cleveland, OH, USA) was
designed (Supporting Information Table I) and included the
3
FIGURE 1. Flowchart of the methodological procedure for the susceptibility testing of clinical strains in planktonic and biofilm mode. CFU:
colony-forming units; MIC: minimum inhibitory concentration; MBEC: minimum biofilm eradication concentration.
following antimicrobial agents: gentamicin (GEN), clindamy-
cin (CLI), vancomycin (VAN), linezolid (LZD), ciprooxacin
(CIP), oxacillin (OXA), fusidic acid (FA), ampicillin (AMP),
trimethoprim/sulfamethoxazole (SXT), and rifampin (RIF).
Concentrations included the common range for MIC deter-
mination, and some extreme above breakpoint concentra-
tions. MIC and MBEC calculations were only performed for
the antimicrobial agents with a known effect against staphy-
lococcal and enterococcal strains, respectively. The MBEC/
MIC ratio was employed to illustrate the fold increase in the
antimicrobial resistance of biolms and to facilitate compar-
isons between different species.
Calgary Biolm Device (CBD). All strains were subcultured
and suspended in MHB, as described above, and an inocu-
lum (107 CFU 3 mL21) was added to each well of a 96-well
Calgary Biolm Device (CBD) (MBEC P&G Assay,
InnovotechV, Calgary, Canada).35,36 Each CBD was inoculated
R
R
with one strain. The plates were sealed with ParalmV
inside a humid chamber to limit evaporation and placed in
a rotary incubator shaker at 150 rpm at 378C for 24 h to
4 ZABOROWSKA ET AL.
allow the formation of biolms on the pegs of the
device lid.
Colony forming units per peg (CFU 3 peg21). Conical
shaped pegs (n 5 4) from each CBD microtiter lid were
removed aseptically using pliers, according to the manufac-
turers instructions, placed in saline, vortexed for 1 min, and
sonicated for 1 min at 40 Hz to detach the cells from pegs
into the saline (CFU 3 peg21 5 CFU 3 mL21) for quantica-
tion of viable CFU on blood agar plates.35,36
MBEC determination: susceptibility testing of biolm
cells. MBEC was determined according to the manufac-
turers protocol.25,36 In brief, biolms grown on peg lids
from the CBD were rinsed and placed on the antimicrobial
R
plates (SensititreV) and incubated for 20 h at 378C. Thereaf-
ter, each peg lid was rinsed and placed in a recovery plate
containing neutralising agents and sonicated at 40 Hz for 5
min to release the remaining biolm into the recovery plate.
The peg lids were replaced by regular sterile lids, and incu-
bated at 378C for additional 20 h. MBEC for the tested anti-
microbial agent was determined by ocular inspection using
BIOFILM FORMATION AND ANTIMICROBIAL SUSCEPTIBILITY
 TABLE II. Patient Clinical Data, Treatment Regimens, Susceptibility Testing (MIC Versus MBEC) and its Impact on Treatment Choice and Failure
Time of Treatment Prospective Retrospective
Infection After Sample of (Before and Susceptibility Susceptibility
Implantation Origin (Strain After Lab 1 Testing from Testing from Outcome Biofilm
Strain Osteomyelitis (years) Isolated from) Results) Lab 1 (MIC) Lab 2 (MBEC) Reinfection Relapse Extraction Score* Score*

E. faecalis Definite 12 MB 6/6, S 1/1 AMX, FLX S: AMP, LZD S: none No Yes No 1 4
CCUG64515 thereafter LZD
E. faecalis Possible 12 BT 3/3, S 1/1 VAN S: VAN, AMP S: none No No Yes 1 3
CCUG64517
E. faecalis Definite 5.5 MB 6/6, S 1/1 AMP, LZD S: AMP, LZD S: none Yes Yes Yes 3 5
CCUG64519 thereafter AMX
E. faecalis Definite 6.5 MB 3/3, S 1/1 CIP, AMX S: AMP, S: none No No Yes 1 4
CCUG64524 R: CIP
E. faecalis Definite 4.8 MB 2/2, S 1/1 AMX thereafter S: AMP S: none Yes No Yes 2 4
CCUG64526 RIF/CLI
E. faecalis Probable 8.3 MB 4/4, S 1/1 AMX S: AMP S: none No No No 0 3
CCUG64527
S. epidermidis Possible 12 BT 3/3, S 1/1 VAN thereafter S: VAN, S: VAN, RIF No No Yes 1 4
CCUG64518 FA and RIF FA, RIF
S. epidermidis Probable 3 S 1/1 RIF, FA S: FA; S: VAN, LZD Yes No No 1 5
CCUG64521 RIF unknown
S. epidermidis Definite 0 BT 7/8 VAN, RIF, FA 2 S: GEN, VAN, CIP, No No No 0 2
CCUG64523 OXA, SXT, RIF
S. aureus Definite 12 MB 6/6, S 1/1 AMX, FLX S: for Isoxa-pc, S: CLI, GEN, VAN, No Yes No 1 4
CCUG64514 thereafter LZD LZD CIP, OXA, SXT, RIF
S. aureus Definite 5 BT 3/9 CDX S: CDX S: CLI, GEN, CIP, OXA, No No Yes 1 4

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2016 VOL 00B, ISSUE 00
CCUG64516 FA, AMP, SXT, RIF
S. aureus Definite 3 MB 3/3, S 1/1 RIF, CIP S: RIF; S: VAN, CIP Yes No No 1 2
CCUG64520 CIP unknown
S. aureus Definite 3.5 MB 12/12 FLX thereafter S for Isoxa-pc, S: none Yes Yes No 2 3
CCUG64522 RIF, CLI CLI, RIF

MB 5 marrow blood, S 5 screw, BT 5 bone tissue; AMX 5 amoxicillin, LZD 5 linezolid, AMP 5 ampicillin, VAN 5 vancomycin, CIP 5 ciprofloxacin, FLX 5 flucloxacillin, RIF 5 rifampin, CLI 5 clinda-
mycin, FA 5 fusidic acid, GEN 5 gentamycin, OXA 5 oxacillin, SXT 5 trimethoprim/sulfamethoxazole, CDX 5 cefadroxil, isoxa-pc 5 isoxazolyl-penicillin; MIC 5 minimum inhibitory concentration;
MBEC 5 minimum biofilm eradication concentration.
a
Biofilm production score (BS) versus outcome score: nonsignificant trend, strains with the highest biofilm production score (5) were isolated from patients with higher outcome scores than
strains with lower biofilm production scores (43) and than non-biofilm producing strains (score 2) (R2 5 0.472; p 5 0.103).

5
ORIGINAL RESEARCH REPORT
a Manual Viewer (Trek Diagnostic, East Grinstead, UK) and
further conrmed by OD measurements in a plate reader.
The antimicrobial plates included consecutive doubling
doses of the antimicrobials, separated by several doubling
steps from a few extreme high concentrations (Supporting
Information Table 1). In the occasion when MIC or MBEC
was above the highest concentration (Y) present on the
plate, the arbitrary value of 2 3 Y was selected. When MIC
or MBEC was between the last two doses present in the
plate (>X and <Y), which were distant by more than one
doubling step, the MIC or MBEC 5 Y. The experiment was
repeated three times for all the strains. After antimicrobial
testing, the biolm-free pegs did not produce growth in the
MBEC assay and the positive control wells with biolms on
pegs showed turbidity.
MIC determination: susceptibility testing of planktonic
cells
Antimicrobial susceptibility measurements were also per-
formed on planktonic cultures of the strains with equal nal
concentrations (CFU 3 mL21) as for the biolm susceptibili-
ty measurements. Incubation and analysis of the plates
were performed according to CLSI recommendations. In
brief:37 100 lL of the inoculum was added to all wells of
the Sensititre plate. The plates were incubated for 20 h at
378C and MIC values were determined as described above
for MBEC values.
In addition, the strains were subjected to standard MIC
susceptibility testing following the CLSI recommendations,
in relation to the recommended nal inoculum concentra-
tions (5 3 105 CFU 3 mL21) and culture conditions. The
strains were categorised as susceptible, intermediate or
resistant according to CLSI breakpoints.37 The experiment
was repeated three times for each strain. The S. aureus
ATCC 29213 strain was used as a control strain.37
Statistics
All comparative statistics were performed per species and
all species together. One-way ANOVA was performed on MIC
versus MBEC per species. MBEC/MIC ratios versus biolm
production score, biomass score and slime score, respective-
ly, were compared using a nonparametric analysis of vari-
ance (KruskalWallis), followed by MannWhitney.
Spearmans bivariate correlation was performed between:
biolm biomass score and slime score; crystal violet OD and
Syto9 values; and all the comparisons between biolm pro-
duction/biomass/slime scores versus total patient outcome
score, and reinfection/relapse/extraction cases, respectively.
For all tests, differences were considered signicant at
p < 0.05. Statistical analyses were performed using SPSS Sta-
tistics 21 (IBM Corporation, USA).
RESULTS
Clinical aspects and outcome
A total of, two possible, two probable and nine denitive
cases of osteomyelitis were determined (Tables I and II).
Marrow blood was cultured and yielded positive cultures in
eight cases, bone tissue cultures in four cases and screw
6 ZABOROWSKA ET AL.
FIGURE 2. Flowchart of patient treatment outcome and its relationship
to given treatment and MBEC level. Low MBEC: <4 3 MIC for at least
one given antibiotic with monotherapeutic efficacy to causative
organism in cases treated with combinations.
cultures in ten cases. The median length of antibiotic treat-
ment was 4 months (Table I). There were a total of four
relapses, six implant extractions and ve cases of reinfection
within the study period. Two strains were assigned the out-
come score 0, eight strains the outcome score 1, two strains
the outcome score 2 and one strain was assigned outcome
score 3 (Table II). Five of seven staphylococcal infections
were treated with a rifampin-based regimen. The only
relapse among these was caused by a weak biolm-
producing S. aureus strain with high MBEC for rifampin
(16 mg 3 mL21) and clindamycin (>128 mg 3 mL21).
Although not signicant, a trend was observed where the
strains with the highest biolm production (biolm score 5)
were isolated from patients that experienced higher out-
come scores (more adverse events) than the strains with
lower biolm production (biolm scores 43) and than the
non-biolm-producing strains (biolm score 2). Particularly,
higher slime production (slime score 2) was associated with
more relapse cases than non- or weak slime production
(slime score 01) (R2 5 0.512; p 5 0.073). In six out of
seven patients that experienced treatment failure, the
strains exhibited high MBECs against the antibiotics used
for treatment (Figure 2).
Biolm-forming capacity
The results of the determination of biolm biomass (crystal
violet assay) and slime-production ability (CRA plate test)
of each strain, providing an overall biolm production score
and classication, are presented in Table III. Four strains
were categorized with biomass score 3 (ODmean 5 1.19), sev-
en with biomass score 2 (ODmean 5 0.26), and two with bio-
mass score 1 (ODmean 5 0.16). The results of the two
staining techniques, crystal violet and Syto9 (data not
shown), demonstrated a signicant correlation (R2 5 0.836,
p 5 0.0003). The CRA plate test indicated normal slime pro-
duction after the test period of 24 and 48 h in eight strains
(4 E. faecalis, 1 S. epidermidis, 3 S. aureus), weak slime pro-
duction in three strains (2 E. faecalis, 1 S. epidermidis), and
two strains were non-slime producers (1 S. aureus, 1 S. epi-
dermidis). No signicant correlation was found between bio-
mass score and slime score. In six strains, there was strong
agreement between total biolm biomass and slime produc-
tion; in contrast, two non-slime producers were categorized
with biomass score 2. Among the 13 analyzed strains, two
BIOFILM FORMATION AND ANTIMICROBIAL SUSCEPTIBILITY
 ORIGINAL RESEARCH REPORT

TABLE III. Biofilm Formation Ability of Clinical Isolates from Osteomyelitis in Patients with Osseointegrated Amputation
Prostheses
Biofilm Biofilm
Crystal Biomass Congo Red Slime Production Production
Strain code Violet (OD)a,d Scorea (Classification)b,d Scoreb Scorec Classificationc

E. faecalis strains
CCUG 64515 0.205 2 1 (p) 2 4 Moderate
CCUG 64517 0.215 2 1 (wp) 1 3 Weak
CCUG 64519 0.627 3 1 (p) 2 5 Strong
CCUG 64524 0.407 3 1 (wp) 1 4 Moderate
CCUG 64526 0.291 2 1 (p) 2 4 Moderate
CCUG 64527 0.158 1 1 (p) 2 3 Weak
S. epidermidis strains
CCUG 64518 2.342 3 1 (wp) 1 4 Moderate
CCUG 64521 1.390 3 1 (p) 2 5 Strong
CCUG 64523 0.357 2 2 (np) 0 2 Non-producer
S. aureus strains
CCUG 64514 0.351 2 1 (p) 2 4 Moderate
CCUG 64516 0.220 2 1 (p) 2 4 Moderate
CCUG 64520 0.192 2 2 (np) 0 2 Non-producer
CCUG 64522 0.161 1 1 (p) 2 3 Weak
S. aureuscontrol strains
ATCC 29213 0.232 2 1 (p) 2 4 Moderate
ATCC 25923 0.113 1 1 (p) 2 3 Weak
S. epidermidiscontrol strains
ATCC 35984 3.379 3 1 (p) 2 5 Strong
ATCC 35983 0.636 3 1 (wp) 1 4 Moderate
ATCC 12228 0.338 2 2 (np) 0 2 Non-producer
a
Phenotypic classification of the biofilm biomass of the strains by crystal violet: the cutoff value (ODc) using tryptic soy broth as blank was
0.094. OD  0.094 (biomass score 5 0); 0.094 < OD  0.188 (biomass score 5 1); 0.188 < OD  0.376 (biomass score 5 2); OD > 0.376 (biomass
score 5 3).
b
Phenotypic classification of slime production ability onto Congo red agar: np 5 non-slime producer (slime score 5 0); wp 5 weak slime pro-
ducer (slime score 5 1); p 5 normal slime producer (slime score 5 2).
c
Biofilm production classification: biofilm production score 5 biomass score 1 slime score. Non-biofilm producer (biofilm production score-
5 02); weak biofilm producer (biofilm production score 5 3); moderate biofilm producer (biofilm production score 5 4); strong biofilm producer
(biofilm production score 5 5).
d
Biomass score versus slime score (R2 5 0.127, p 5 0.680).

strains were categorized as strong biolm producers, six
strains as moderate producers, three as weak producers and
two were designated as non-producers. All staphylococcal
strains were positive for both icaA and icaD genes (Figure 3).
Antimicrobial susceptibility testing of strains as
planktonic (MIC) versus as biolm (MBEC)
All strains adhered to the polystyrene pegs of the MBEC
assay. The number of viable CFU on the pegs ranged from
2.2 3 103 to 7.9 3 105 CFU per peg (Table IV) and the
inocula were on average 10 CFU 3 mL21 lower in all MBEC
trials compared with MIC trials. Details on the strain sus-
ceptibilities according to MIC and MBEC assays are pre-
sented in Supporting Information Table II. The results were
closely reproduced in the three separate trials.
When comparing MIC versus MBEC values for the three
different species: E. faecalis showed signicantly higher
MBEC than MIC against VAN, LZD, CIP, AMP, and RIF
(p < 0.001); S. epidermidis strains showed a signicantly
higher MBEC than MIC for LZD and CIP (p < 0.05); and S.
aureus MBECs were signicantly higher for all tested antimi-
crobials (p < 0.05) with the exemption of AMP and SXT.
The MBEC/MIC ratios demonstrated differences between
planktonic and adherent bacteria in terms of antimicrobial
performance (Table IV). Most antimicrobials were effective
in killing or inhibiting planktonic bacterial growth at low
concentrations, while only a few biolms were eradicated at
low antimicrobial concentrations. All the strains with overall
biolm production scores of 3, 4, and 5 (weak, moderate,
and strong biolm producers, respectively) were associated
with signicantly higher MBEC/MIC ratios than non-biolm
producing strains (biolm score 2) (p < 0.0001) for VAN,
LZD, CIP, AMP, and RIF. Furthermore, slime-producing
strains showed signicantly higher MBEC/MIC ratios than
FIGURE 3. PCR detection of icaA and icaD genes in the staphylococcal
strains. All tested strains were positive for both genes.

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2016 VOL 00B, ISSUE 00 7
 TABLE IV. Ratios Between the MBEC and MIC Values for Ten Antimicrobial Agents
Biofilm
production
Strain code CLI GEN VANa LZDa CIPa OXA FA AMPa SXT RIFa CFU/peg classificationa

E. faecalis strains
CCUG 64515 N/A N/A 512 128 512 N/A N/A 1024 N/A 256 9.8 3 103 Moderate
CCUG 64517 N/A N/A 1024 128 512 N/A N/A 1024 N/A 256 4.6 3 103 Weak
CCUG 64519 N/A N/A 1024 128 256 N/A N/A 2048 N/A 256 7.5 3 104 Strong
CCUG 64524 N/A N/A 512 128 512 N/A N/A 1024 N/A 64 4.7 3 104 Moderate
CCUG 64526 N/A N/A 1024 128 512 N/A N/A 1024 N/A 512 1.4 3 104 Moderate
CCUG 64527 N/A N/A 1024 128 512 N/A N/A 2048 N/A 256 2.7 3 104 Weak
S. epidermidis strains
CCUG 64518 1 4 1 16 8 0.5 4 2 4 0.12 1.2 3 104 Moderate
CCUG 64521 1 1 0.5 2 2 0.125 64 1 0.125 4 8.8 3 103 Strong
CCUG 64523 0.016 0.016 2 4 0.125 0.25 4 0.25 0.5 0.0038 8.8 3 103 Non-producer
S. aureus strains
CCUG 64514 2.08 1 1 256 2 1 2048 1 1 2 2.2 3 103 Moderate
CCUG 64516 1 1 16 128 1 1 2 1 1 1 4.8 3 104 Moderate
CCUG 64520 64 32 1 32 2 128 2048 1 32 533 2.7 3 105 Non-producer
CCUG 64522 1024 1024 1024 128 32 1024 2048 2 32 533 7.9 3 105 Weak

The ratios were calculated dividing the MBEC value by the MIC value for each strain and antimicrobial agent. The ratios show the
increased/equal resistance factor of the biofilm susceptibility versus its equivalent planktonic susceptibility (approximately equal final amount of
CFU in the planktonic culture as in the biofilm).
MIC (minimum inhibitory concentration), MBEC (minimum biofilm eradication concentration), CLI (clindamycin), GEN (gentamycin), VAN
(vancomycin), LZD (linezolid), OXA (oxacillin), FA (fusidic acid), AMP (ampicillin), SXT (trimethoprim/sulfamethoxazole), RIF (rifampin).
a
Biofilm production versus MBEC/MIC ratio: MBEC/MIC ratios from strains with biofilm production scores of 3, 4, and 5 (weak, moderate, and
strong biofilm producers, respectively) were significantly higher than the MBEC/MIC ratios of non-biofilm producing strains (score 2) (p < 0.0001)
for VAN, LZD, CIP, AMP, and RIF.

non-slime-producing strains (p < 0.0001). In contrast, for
the same antimicrobials the CV biolm biomass scores
showed contradictory results, where biolm biomass score
1 showed higher MBEC/MIC ratios than biomass score 2
and 3 (p < 0.05).
The MBECs were high and uniform for all enterococcal
strains showing the highest MBEC/MIC ratios compared to
staphylococci, which showed more interstrain variability
depending on the antimicrobial agent. The MBEC/MIC ratios
for E. faecalis ranged from 642,048 (median 512) for VAN,
CIP, LZD, AMP, and RIF (Table IV). The range of the MBEC/
MIC ratios for the ten antibiotics in S. aureus was 12048
(median 9) and for the S. epidermidis strains was 0.003864
(median 1). Furthermore, the S. aureus strain that caused a
FIGURE 4. Distribution of the susceptibility testing to four common
antimicrobial agents: vancomycin, linezolid, ciprofloxacin, and rifam-
pin according to (a) MIC 5 minimum inhibitory concentration and (b)
MBEC 5 minimum biofilm eradication concentration for the 13 clinical
strains.
relapse in a patient exerted a 459-fold increase in MBEC
versus a 63-fold MBEC by the nonrelapse causing staphylo-
cocci. Nineteen percent of all the strains were susceptible to
VAN, CIP, LZD, and RIF according to MIC whereas 77% were
resistant according to MBEC evaluation (Figure 4).
DISCUSSION
This study represents the rst report on the biolm proper-
ties of bacteria causing deep infections associated with bone-
anchored femoral amputation prostheses. In the present
study, the biolm-forming ability of bacterial isolates from
orthopaedic implant-associated osteomyelitis was determined
with both phenotypic and genotypic methods. A major obser-
vation was that the majority of the clinical strains (11 out of
13) formed biolm in vitro and most of them were classied
as moderate or strong biolm producers.
In agreement with observations on prosthetic joint infec-
tions,8 biolm-producing S. aureus and S. epidermidis strains
from infected percutaneous osseointegrated amputation pros-
theses showed signicantly higher MBEC/MIC ratios than the
non-biolm-producing strains. Since also the slime-producing
strains had higher MBEC/MIC ratios compared with the non-
slime-producing ones, the ability to produce slime may have
contributed to the observed higher tolerance to antimicro-
bials. However, biolm tolerance to antimicrobials is multifac-
torial and not only caused by slime.38,39 Restricted
antimicrobial diffusion, slow growing persister cells, and
altered physiological activity of the bacteria constitute addi-
tional alternative mechanism for the observed ndings.38,40
Planktonic and biolm antibiograms differed for VAN,
LZD, CIP and RIF, commonly used for the treatment of

8 ZABOROWSKA ET AL. BIOFILM FORMATION AND ANTIMICROBIAL SUSCEPTIBILITY


periprosthetic joint and implant infections. This observation
indicates that these antimicrobial agents may have a lower
efcacy in chronic biolm infections.
MIC-guided treatment alone may have been insufcient
in these infections caused by biolm-producing strains.
Therefore, the current standard culture techniques may fail
in predicting antimicrobial susceptibilities, partially explain-
ing the clinical difculty to eradicate biomaterials-associated
infection. Biolm-based susceptibility testing simulate better
the pathophysiology of chronic biolm infections,41 although
additional clinical studies are necessary to provide more
evidence and demonstrate that it predicts better the
response to antimicrobial therapy than conventional plank-
tonic testing. The combined method (CBD and Sensititre
plate) used in this study for the susceptibility testing of bio-
lms should be feasible to implement in clinical microbiolo-
gy laboratories: it represents a standardized option for
growing biolms in vitro, the antimicrobial plates follow
standards, it provides results within the relative short time
of 5 days and it is economically sustainable.
When qualitatively relating the MBECs to treatment out-
come of the patients, treatment failure appears associated
with high MBECs against the antibiotics used. Assessment of
each patient case led to the following observations: (i) the
two patients who did not experience post-treatment compli-
cations were infected by a non-biolm-producing and a
weak biolm-producing strain, respectively; (ii) weak or
moderate biolm production was characteristic of patient
outcomes involving one to two complications; and (iii) the
only case experiencing all three categories of complications
(relapse, reinfection, and extraction) was caused by an E.
faecalis strain with strong biolm-production ability. The
present results support the hypothesis that strains that are
able to form biolm may persist within a host causing
chronic, relapsing and/or metastatic infection.42,43 Even
though statistical evidence cannot be obtained due to the
small patient cohort in this study, it appears that post-
treatment complications are more common in infections
caused by isolates with increased biolm abilities, and that
the strains belonging to the highest slime score 2 associated
with higher number of relapses than the non-slime or weak
slime producers.
Biolm formation was heterogeneous across species and
strains. MBECs were overall signicantly higher than MICs
for almost all of the 10 antimicrobials. Differences in inocu-
lum sizes between MIC and MBEC methods were too small
to explain this observation, hence supporting a biolm phe-
notype in most of the strains tested. Three staphylococcal
strains displayed surprisingly low MBECs toward some anti-
microbials. One of these strains did not produce biolm,
likely explaining the low MBECs. In contrast, the enterococ-
cal strains, independent of the biolm production degree,
uniformly exhibited high MBECs for all tested antimicrobials
compared with the staphylococcal strains in this study. This
is in accordance with previously reported data showing pro-
nounced antimicrobial tolerance in E. faecalis biolms.44,45
Such MBECs could only be achieved by local drug adminis-
tration, otherwise eliciting systemic toxicity.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2016 VOL 00B, ISSUE 00
ORIGINAL RESEARCH REPORT
All the isolates from patients infected with staphylococci
were icaA and icaD positive; however, two strains did not
produce slime according to CRA plate test. This phenotypic
inability has been previously reported,46 demonstrating var-
iability in genetic expression. Syto9 stains nucleic acids, ren-
dering both live and dead cells detectable inside the biolm,
although staining of the eDNA in the matrix cannot be ruled
out.47 The biomass (cells and slime) quantied by crystal
violet was therefore complemented with CRA plate test and
the presence of the ica operon (for staphylococci), in order
to assess separately the phenotypic and genotypic slime
production within the biolm. It is interesting to note that
the two non-biolm-producing strains and the PIA-negative
control strain (ATCC 12228) showed quantiable biomass
by the crystal violet microtitre plate assay, indicating that
the quantied biomass was mainly consisting of adhered
cells and not of EPS. When available, it is advisable to use
fully characterised negative control strains when performing
the microtitre plate assay for a specic bacterial species.
A limitation of the study is the small number of patients
and the heterogeneity of sampled tissues and treatments.
Furthermore, in no case was the diagnosis supported by
determination of bone tissue PMNs (polymorphonuclear leu-
kocytes), which is a commonly recommended diagnostic cri-
terion for osteomyelitis.48 However, the patients had a
clinical history strongly suggestive of infection, and although
it is not common practice in osteomyelitis diagnostics,49
bone marrow culturing has a high diagnostic value.50 Evi-
dently, in this subset of orthopaedic implant-associated oste-
omyelitis, better diagnostics will translate into more reliable
results when relating treatment outcome to the virulence of
the causative strain. Although this method involves more
steps than conventional MIC testing, its implementation
would not delay treatment decisions regarding prolonged
oral antimicrobials following initial intravenous treatment in
selected cases. For future prospective studies, a standard-
ized sampling procedure should be followed48,51 and treat-
ment outcomes compared between MBEC and conventional
MIC testing groups.
CONCLUSION
Overall, this study demonstrated that the majority of iso-
lates, causing osteomyelitis in patients with percutaneous
bone-anchored implants, are biolm producers showing an
increased antibiotic resistance compared with their plank-
tonic counterparts. Slime producing strains required higher
antimicrobial concentrations compared with non-producers.
Biolm susceptibility testing is therefore likely to provide a
more relevant basis for the treatment of osteomyelitis asso-
ciated with this type of orthopedic implant. The clinical use-
fulness of this novel combination of the Calgary biolm
device and Sensititre susceptibility plate should be further
optimised and evaluated in a prospective study of orthopae-
dic implant-associated infections.
ACKNOWLEDGMENTS
The assistance of Mrs Maria Hoffman and Mrs Katarzyna
Kulbacka-Ortiz throughout this study is gratefully
9
acknowledged. The sponsors were not involved in the study
design, data acquisition, interpretation, writing and submis-
sion of the article. Dr Rickard Branemark and his family are the
owners of the medical device company Integrum AB, which
supplies all the components in this implant system. All other
authors: none to declare.
REFERENCES
1. Gristina AG. Biomaterial-centered infection: Microbial adhesion
versus tissue integration. Science 1987;237:15881595.
2. Palmquist A, Omar OM, Esposito M, Lausmaa J, Thomsen P. Tita-
nium oral implants: Surface characteristics, interface biology and
clinical outcome. J R Soc 2010;7 Suppl 5:S515S527.
3. Branemark R, Berlin O, Hagberg K, Bergh P, Gunterberg B,
Rydevik B. A novel osseointegrated percutaneous prosthetic sys-
tem for the treatment of patients with transfemoral amputation: A
prospective study of 51 patients. Bone Joint J 2014;96B:106113.
4. Svensson S, Trobos M, Hoffman M, Norlindh B, Petronis S,
Lausmaa J, Suska F, Thomsen P. A novel soft tissue model for
biomaterial-associated infection and inflammationBacteriologi-
cal, morphological and molecular observations. Biomaterials
2015;41:106121.
5. Parvizi J. New definition for periprosthetic joint infection. Am J
Orthoped 2011;40:614615.
6. Osmon DR. Diagnosis and management of prosthetic joint infec-
tion: Clinical practice guidelines by infectious diseases society of
America. Clin Infect Dis 2013;2013:e1e25.
7. LCM. A modified staging system for chronic osteomyelitis.
J Orthoped 2015;12:184192.
8. Molina-Manso D, del Prado G, Ortiz-Perez A, Manrubia-Cobo M,
Gomez-Barrena E, Cordero-Ampuero J, Esteban J. In vitro suscep-
tibility to antibiotics of staphylococci in biofilms isolated from
orthopaedic infections. Int J Antimicrob Agents 2013;41:521523.
9. Arciola CR. Etiology of implant orthopedic infections: A survey on
1027 clinical isolates. Int J Artif Org 2005;28:10911100.
10. Lazzarini L, Lipsky BA, Mader JT. Antibiotic treatment of osteomy-
elitis: What have we learned from 30 years of clinical trials? Int J
Infect Dis 2005;9:127138.
11. Stengel D, Bauwens K, Sehouli J, Ekkernkamp A, Porzsolt F. Sys-
tematic review and meta-analysis of antibiotic therapy for bone
and joint infections. Lancet Infect Dis 2001;1:175188.
12. Lew DP, Waldvogel FA. Osteomyelitis. New Engl J Med 1997;336:
9991007.
13. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: A
common cause of persistent infections. Science 1999;284:1318
1322.
14. Lebeaux D, Chauhan A, Rendueles O, Beloin C. From in vitro to in
vivo models of bacterial biofilm-related infections. Pathogens
2013;2:288356.
15. Hall-Stoodley L, Stoodley P, Kathju S, Hoiby N, Moser C,
Costerton JW, Moter A, Bjarnsholt T. Towards diagnostic guide-
lines for biofilm-associated infections. FEMS Immunol Med
Microbiol 2012;65:127145.
16. Donlan RM. Biofilms: Microbial life on surfaces. Emerg Infect Dis
2002;8:881890.
17. Zimmerli W, Moser C. Pathogenesis and treatment concepts of
orthopaedic biofilm infections. FEMS Immunol Med Microbiol
2012;65:158168.
18. Cramton SE, Gerke C, Schnell NF, Nichols WW, Gotz F. The intercel-
lular adhesion (ica) locus is present in Staphylococcus aureus and
is required for biofilm formation. Infect Immun 1999;67:54275433.
19. Go tz F. Staphylococcus and biofilms. Mol Microbiol 2002;43:
13671378.
20. Kristich CJ, Li YH, Cvitkovitch DG, Dunny GM. Esp-independent
biofilm formation by Enterococcus faecalis. J Bacteriol 2003;186:
154163.
21. Brady RA, Leid JG, Calhoun JH, Costerton JW, Shirtliff ME. Oste-
omyelitis and the role of biofilms in chronic infection. FEMS
Immunol Med Microbiol 2008;52:1322.
22. Donlan RM. Biofilm formation: A clinically relevant microbiologi-
cal process. Healthcare Epidemiol 2001;33:13871392.
10 ZABOROWSKA ET AL.
23. Francolini I, Donelli G. Prevention and control of biofilm-based
medical-device-related infections. FEMS Immunol Med Microbiol
2010;59:227238.
24. Hoiby N, Bjarnsholt T, Givskov M, Molin S, Ciofu O. Antibiotic
resistance of bacterial biofilms. Int J Antimicrob Agents 2010;35:
322332.
25. Harrison JJ, Turner RJ, Ceri H. High-throughput metal susceptibil-
ity testing of microbial biofilms. MBC Microbiol 2005;5:53.
26. Arcola CR, Campoccia D, Gamberini S, Cervellati M. Detection of
slime production by means of an optimised Congo red agar plate test
based on a coloumetric scale in Staphylococcus epidermidis clinical
isolates genotyped for ica locus. Biomaterials 2002;23:42334239.
27. Christensen GD, Simpson WA, Younger JJ, Baddour LM, Barrett
FF, Melton DM, Beachey EH. Adherence of coagulase-negative
staphylococci to plastic tissue culture plates: A quantitative model
for the adherence of staphylococci to medical devices. J Clin
Microbiol 1985;22:9961006.
28. Peeters E, Nelis HJ, Coenye T. Comparison of multiple methods
for quantification of microbial biofilms grown in microtiter plates.
J Microbiol Methods 2008;72:157165.
29. Arciola CR, Collamati S, Donati E, Montanaro L. A rapid PCR
method for the detection of slime-producing strain of Staphylo-
coccus epidermidis and S. aureus in periprosthesis infections.
Diag Mol Pathol 2001;10:130137.
30. Varma JK, Greene KD, Ovitt J, Barrett TJ, Medalla F, Angulo FJ.
Hospitalization and Antimicrobial resistance in Salmonella out-
breaks, 19842002. Emerg Infect Dis 2005;11:943946.
31. Tillander J, Hagberg K, Hagberg L, Branemark R. Osseointegrated
titanium implants for limb prostheses attachments: infectious
complications. Clin Orthop Relat Res 2010;468:27812788.
32. Merritt JH, Kadouri DE, OToole GA. Growing and analyzing static
biofilms. Curr Protoc Microbiol 2011;1B.1.11B.1.18.
33. Zimmerli W, Widmer AF, Blatter M, Frei R, Ochsner PE. Role of
rifampin for treatment of orthopedic implant-related staphylococ-
cal infections a randomized controlled trial. J Am Med Assoc
1998;279:15371541.
34. Saginur R, St. Denis M, Ferris W, Aaron SD, Chan F, Lee C,
Ramotar K. Multiple combination bactericidal testing of staphylo-
coccal biofilms from implant-associated infections. Antimicrob
Agents Chemother 2006;50:5561.
35. Harrison JJ, Stremick CA, Turner RJ, Allan ND, Olson ME, Ceri H.
Microtiter susceptibility testing of microbes growing on peg lids:
A miniaturized biofilm model for high-throughput screening. Nat
Protoc 2010;5:12361254.
36. Ceri H, Olson ME, Stremick C, Read RR, Morck D, Buret A. The
Calgary biofilm device: New technology for rapid determination
of antibiotic susceptibilities of bacterial biofilms. J Clin Microbiol
1999;37:17711776.
37. Clinical and Laboratory Standards Institute (CLSI). Performance
Standards for Antimicrobial Susceptibility Testing, 26th ed. CLSI
Document M100-S26. Wayne, USA: Centers for Disease Control
and Prevention; 2015.
38. Percival SL, Hill KE, Malic S, Thomas DW, Williams DW. Antimi-
crobial tolerance and the significance of persister cells in recalci-
trant chronic wound biofilms. Wound Repair Regen 2011;19:19.
39. Bjarnsholt T, Ciofu O, Molin S, Givskov M, Hoiby N. Applying
insights from biofilm biology to drug developmentCan a new
approach be developed? Nat Rev Drug Discov 2013;12:791808.
40. Lewis K. Riddle of biofilm resistance. J Antimicrob Chemother
2001;45:9991007.
41. Aaron SD, Ferris W, Ramotar K, Vandemheen K, Chan F, Saginur
R. Single and Combination antibiotic susceptibilities of plankton-
ic, adherent, and biofilm-grown Pseudomonas aeruginosa iso-
lates cultured from sputa of adults with cystic fibrosis. J Clin
Microbiol 2002;40:41724179.
42. Sanchez CJ, Jr., Mende K, Beckius ML, Akers KS, Romano DR,
Wenke JC, Murray CK. Biofilm formation by clinical isolates and
the implications in chronic infections. BMC Infect Dis 2013;13:47.
43. Hall-Stoodley L, Costerton JW, Stoodley P. Bacterial biofilms:
From the natural environment to infectious diseases. Nat Rev
Microbiol 2004;2:95108.
44. Holmberg A, Mo rgelin M, Rasmussen M. Effectiveness of cipro-
floxacin or linezolid in combination with rifampicin against
BIOFILM FORMATION AND ANTIMICROBIAL SUSCEPTIBILITY

Enterococcus faecalis in biofilms. J Antimicrob Chemother 2012;
67:433439.
45. Sandoe JA, Wysome J, West AP, Heritage J, Wilcox MH. Mea-
surement of ampicillin, vancomycin, linezolid and gentamicin
activity against enterococcal biofilms. J Antimicrob Chemother
2006;57:767770.
46. Arciola CR, Campoccia D, Gamberini S, Rizzi S, Donati ME,
Baldassarri L, Montanaro L. Search for the insertion element
IS256 within the ICA locus of Staphylococcus epidermidis clinical
isolates collected from biomaterial-associated infections. Biomate-
rials 2004;25:41174125.
47. Ratner BD, Hoffman AS, Schoen FJ, Lemons JE. Biomaterials
scienceAn introduction to materials in medicine 2013;565583.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2016 VOL 00B, ISSUE 00
ORIGINAL RESEARCH REPORT
48. Widmer A. New developments in diagnosis and treatment of
infection in orthopedic implants. Clin Infect Dis 2001;33(Suppl 2):
S94S106.
49. Lew DP, Waldvogel FA. Osteomyelitis. Lancet 2004;364:369379.
50. Guerra-Caceres JG, Gotuzzo-Herencia E, Crosby-Dagnino E, Miro-
Quesada M, Carrillo-Parodi C. Diagnostic value of bone marrow
culture in typhoid fever. Transaction of the royal society of tropi-
cal medicine and hygiene 1979;73:680683.
51. Nilsdotter-Augustinsson A, Koskela A, Ohman derquist B.
L, So
Characterization of coagulase-negative staphylococci isolated
from patients with infected hip prostheses: Use of phenotypic
and genotypic analyses, including tests for the presence of the
ICA operon. Eur J Clin Microbiol 2007;26:255265.
11