Content uploaded by Alice JS Fox
Author content
All content in this area was uploaded by Alice JS Fox on Sep 01, 2014
Content may be subject to copyright.
http://ajs.sagepub.com/
Medicine
The American Journal of Sports
http://ajs.sagepub.com/content/early/2014/08/20/0363546514545858
The online version of this article can be found at:
DOI: 10.1177/0363546514545858
published online August 20, 2014Am J Sports Med
Susannah L. Gilbert, Joseph T. Nguyen, Salma Chaudhury, Russell F. Warren and Scott A. Rodeo
Alice J.S. Fox, Michael O. Schär, Florian Wanivenhaus, Tony Chen, Erik Attia, Nikolaus B. Binder, Miguel Otero,
Fluoroquinolones Impair Tendon Healing in a Rat Rotator Cuff Repair Model: A Preliminary Study
Published by:
http://www.sagepublications.com
On behalf of:
American Orthopaedic Society for Sports Medicine
can be found at:The American Journal of Sports MedicineAdditional services and information for
P<P Published online August 20, 2014 in advance of the print journal.
http://ajs.sagepub.com/cgi/alertsEmail Alerts:
http://ajs.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
What is This?
- Aug 20, 2014OnlineFirst Version of Record >>
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
Fluoroquinolones Impair Tendon Healing
in a Rat Rotator Cuff Repair Model
A Preliminary Study
Alice J.S. Fox,
*
y
MSc, Michael O. Scha
¨r,
y
MD, Florian Wanivenhaus,
y
MD,
Tony Chen,
z
PhD, Erik Attia,
y
BS, Nikolaus B. Binder,
y
MD, PhD, Miguel Otero,
y
PhD,
Susannah L. Gilbert,
§
MS, Joseph T. Nguyen,
||
MPH, Salma Chaudhury,
y
MD, PhD,
Russell F. Warren,
y
MD, and Scott A. Rodeo,
y
MD
Investigation performed at the Hospital for Special Surgery, New York, New York, USA
Background: Recent studies suggest that fluoroquinolone antibiotics predispose tendons to tendinopathy and/or rupture. How-
ever, no investigations on the reparative capacity of tendons exposed to fluoroquinolones have been conducted.
Hypothesis: Fluoroquinolone-treated animals will have inferior biochemical, histological, and biomechanical properties at the
healing tendon-bone enthesis compared with controls.
Study Design: Controlled laboratory study.
Methods: Ninety-two rats underwent rotator cuff repair and were randomly assigned to 1 of 4 groups: (1) preoperative (Preop),
whereby animals received fleroxacin for 1 week preoperatively; (2) pre- and postoperative (Pre/Postop), whereby animals
received fleroxacin for 1 week preoperatively and for 2 weeks postoperatively; (3) postoperative (Postop), whereby animals
received fleroxacin for 2 weeks postoperatively; and (4) control, whereby animals received vehicle for 1 week preoperatively
and for 2 weeks postoperatively. Rats were euthanized at 2 weeks postoperatively for biochemical, histological, and biomechan-
ical analysis. All data were expressed as mean 6standard error of the mean (SEM). Statistical comparisons were performed using
either 1-way or 2-way ANOVA, with P\.05 considered significant.
Results: Reverse transcriptase quantitative polymerase chain reaction (RTqPCR) analysis revealed a 30-fold increase in expres-
sion of matrix metalloproteinase (MMP)-3, a 7-fold increase in MMP-13, and a 4-fold increase in tissue inhibitor of metalloprotei-
nases (TIMP)-1 in the Pre/Postop group compared with the other groups. The appearance of the healing enthesis in all treated
animals was qualitatively different than that in controls. The tendons were friable and atrophic. All 3 treated groups showed sig-
nificantly less fibrocartilage and poorly organized collagen at the healing enthesis compared with control animals. There was a sig-
nificant difference in the mode of failure, with treated animals demonstrating an intrasubstance failure of the supraspinatus tendon
during testing. In contrast, only 1 of 10 control samples failed within the tendon substance. The healing enthesis of the Pre/Postop
group displayed significantly reduced ultimate load to failure compared with the Preop, Postop, and control groups. There was no
significant difference in load to failure in the Preop group compared with the Postop group. Pre/Postop animals demonstrated
significantly reduced cross-sectional area compared with the Postop and control groups. There was also a significant reduction
in area between the Preop and control groups.
Conclusion: In this preliminary study, fluoroquinolone treatment negatively influenced tendon healing.
Clinical Relevance: These findings indicate that there was an active but inadequate repair response that has potential clinical
implications for patients who are exposed to fluoroquinolones before tendon repair surgery.
Keywords: fluoroquinolone; tendon healing; rotator cuff repair; fleroxacin; tendinopathy
Fluoroquinolones (FQs) are an important class of antimi-
crobial agents commonly used to treat infections from
gram-negative organisms, gram-positive organisms, and
anaerobic bacteria. Because of their favorable
pharmacokinetic properties, excellent bactericidal activity,
and broad antimicrobial spectrum,
17,28
FQs are widely
administered for urinary tract, upper respiratory, and intes-
tinal infections and in the treatment of certain musculoskel-
etal infections such as osteomyelitis and septic arthritis.
Side effects associated with the use of FQs have been
reported in the literature, however most are not severe.
Gastrointestinal problems (nausea, vomiting, and diar-
rhea) are the most frequent side effects, followed by mild
The American Journal of Sports Medicine, Vol. XX, No. X
DOI: 10.1177/0363546514545858
!2014 The Author(s)
1
AJSM PreView, published on August 20, 2014 as doi:10.1177/0363546514545858
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
neurological disorders, skin reactions, myalgia, arthralgia,
and arthritis.
16,25
Since 1983, cases of FQ-induced tendin-
opathy have been reported in the literature.
{
Although the
exact incidence of FQ-induced tendinopathy is unknown,
some authors have
estimated it to range from 0.14% to 0.4%.
11,23,24,50
The inci-
dence rate increases to 12.2% to 15.6% in transplant recip-
ients.
11,24
Although the Achilles tendon is most commonly
affected after FQ treatment, inflammation and involve-
ment of other tendons such as supraspinatus and biceps
brachii also have been reported.
27
The symptoms after
FQ treatment range from mild pain around the affected
tendon to complete rupture requiring surgical interven-
tion. In 2008, the US Food and Drug Administration man-
dated that all FQ products have a ‘‘black-box’’ warning
indicating the increased risk of FQs in adverse events
such as tendinopathy and tendon rupture.
41
Direct toxic effects and ischemic injury,
23,39,46
or alter-
ations in the synthesis or breakdown of extracellular
matrix (ECM) components,
5,8,29,39,48
may contribute to
FQ-induced tendon damage. Alterations and dysfunction
of cellular components, combining localized matrix-
degrading activity and deficient ground substance
production, may lead to changes in the biomechanical
properties of tendon matrix, resulting in tendinopathy
and subsequent rupture.
48
Inflammation of the paratenon
and degenerative changes in tendon cells have also been
reported in FQ-treated animals,
21,38
and FQs have been
shown to have a number of effects on various mammalian
cell types in culture, including both increased and
decreased expression of inflammatory mediators,
32,49
reduced expression of ECM proteins,
4,48
reduced mito-
chondrial activity,
4
and noncytotoxic inhibition of canine
tendon cell proliferation.
48
These observations notwithstanding, the pathophysio-
logical mechanisms underlying FQ-induced tendinopathy
are still poorly understood, and very little is known about
the reparative capacity of tendons exposed to FQ. There-
fore, the purpose of this study was to investigate the effect
of the FQ fleroxacin on rotator cuff injury and repair in an
established rat model. Fleroxacin, as well as the dosage
used, was chosen because of its toxic potential to induce
tendinopathy, as reported by a previous study
20
and fur-
ther confirmed by a pilot (unpublished) study at our insti-
tution. We hypothesized that FQ-treated animals would
show inferior biochemical, histological, and biomechanical
properties at the healing enthesis compared with control
animals.
MATERIALS AND METHODS
This study was approved by our institutional animal care
and use committee.
Study Design
Because previous studies have demonstrated anatomic simi-
larities with the human shoulder, a rat model was selected to
study rotator cuff tendon healing after surgical repair.
6
A
total of 92 male Sprague-Dawley rats (obtained at 275-
300 g; Harlan Laboratories) underwent unilateral detach-
ment of the supraspinatus tendon from the greater tuberos-
ity followed by immediate anatomic repair with
transosseous fixation as described previously in detail.
3
Rats were treated with either fleroxacin or 13phosphate-
buffered saline (PBS) vehicle and were randomly assigned
to 1 of 4 groups: (1) preoperative (Preop), whereby animals
received fleroxacin for 1 week preoperatively; (2) pre- and
postoperative (Pre/Postop), whereby animals received flerox-
acin for 1 week preoperatively and for 2 weeks postopera-
tively; (3) postoperative (Postop), whereby animals received
fleroxacin for 2 weeks postoperatively; and (4) control,
whereby animals received 13PBS for 1 week preoperatively
and for 2 weeks postoperatively. Fleroxacin (900 mg/kg) or
13PBS was administered by oral gavage every 24 hours.
This particular FQ and dose was chosen based on its toxic
potential to induce tendinopathy.
20
All animals were eutha-
nized by overexposure to carbon dioxide at 2 weeks postoper-
atively for biochemical (n = 5), histological (n = 8), or
biomechanical analysis (n = 10).
Drug Administration
Fleroxacin (TCI America) was prepared daily in a solution
of 13PBS (Gibco; Invitrogen Life Technologies) and NaOH
(Sigma-Aldrich) to achieve a dose of 900 mg/kg. Then, 1-N
HCl was used to achieve a pH between 8 and 8.5. Rats
received the suspended fleroxacin solution or 13PBS daily
by orogastric gavage. A standard dose volume of 2.5 mL/kg
was used throughout. All animals were lightly anesthe-
tized with 2% isoflurane (Baxter Inc) for purposes of
administering orogastric treatments.
Surgical Technique
The rats were anesthetized with an intraperitoneal injec-
tion of ketamine (80 mg/kg; Fort Dodge Animal Health)
and xylazine (5 mg/kg; Akorn Inc). Anesthesia was main-
tained with 2% isoflurane (Baxter Inc). All operations
were performed by use of sterile technique with the rat
in the lateral decubitus position. A deltoid-splitting
*
Address correspondence to Alice J.S. Fox, MSc, Laboratory for Soft Tissue Research, Hospital for Special Surgery, 535 East 70th Street, New York,
NY 10021, USA (e-mail: foxa@hss.edu; ajsfox27@hotmail.com).
y
Laboratory for Soft Tissue Research, Hospital for Special Surgery, New York, New York, USA.
z
Laboratory for Soft Tissue Research, Department of Biomechanics, Hospital for Special Surgery, New York, New York, USA.
§
Department of Biomechanics, Hospital for Special Surgery, New York, New York, USA.
||
Healthcare Research Institute, Hospital for Special Surgery, New York, New York, USA.
One or more of the authors has declared the following potential conflict of interest or source of funding: R.F.W. receives royalties from Biomet and owns
stock in Ivy Sports. S.A.R. is a consultant for Smith and Nephew. The study was funded by the Russell F. Warren Research Chair Fund.
{
References 1, 12, 19, 29, 31, 35, 43, 44, 47, 50.
2Fox et al The American Journal of Sports Medicine
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
incision was made, and the acromioclavicular joint was
divided, allowing visualization of the rotator cuff tendons.
The supraspinatus tendon was isolated and a modified
Mason-Allen stitch was placed by use of 4-0 Ethibond
(Johnson & Johnson Inc) nonabsorbable suture. The ten-
don was then sharply detached from the greater tuberosity
and the footprint gently decorticated with a scalpel blade
to ensure complete debridement of the native enthesis.
Crossed bone tunnels were drilled at the anterior and pos-
terior margins of the footprint and 2 mm lateral to the
articular surface by use of a 22-gauge needle (Becton
Dickinson). Suture ends were then passed through the
bone tunnels and firmly tied over the humeral metaphyseal
cortex, anatomically repairing the supraspinatus tendon to
its native footprint.
6
The deltoid split and wound were subse-
quently closed in a standard layered fashion with absorbable
sutures. Buprenorphine (0.05 mg/kg; Reckitt Benckiser Phar-
maceuticals Inc) was administered subcutaneously for
analgesia during the postoperative period. Ad libitum
weightbearing and cage activity were allowed postopera-
tively. The animals were observed daily for abnormal clinical
signs (lethargy, loss of appetite) and changes in body weight.
Reverse-Transcriptase Quantitative
Polymerase Chain Reaction Analysis
Previous studies have suggested that FQ-induced tendon
damage involves cytotoxicity, inflammatory-like reactions,
and alterations in matrix deposition and remodeling. Thus,
to better understand the mechanisms underlying FQ actions
in tendon repair, we performed reverse-transcriptase quanti-
tative polymerase chain reaction (RTqPCR) analysis on total
RNA isolated from supraspinatus tendon, assessing changes
in messenger RNA (mRNA) expression of genes involved in
these processes. For RT-qPCR analysis, total RNA was iso-
lated from the supraspinatus tendon (n = 5 per group) by
use of TRIzol reagent (Life Technologies) followed by DNase
Itreatmentandcolumnclean-up(Qiagen).TotalRNAwas
reverse transcribed with the QuantiTect Reverse Transcrip-
tion Kit (Qiagen) according to the manufacturer’s instruc-
tions. Amplifications were carried out with SYBR Green I–
based RT-PCR on the Opticon 2 Real-Time PCR Detector
System (BioRad). Amplificationefficiencieswerecalculated
for all primers by use of serial dilutions of pooled complemen-
tary DNA samples. The data were calculated as the ratio of
each gene to expression of b-actin (ACTB), and glyceralde-
hyde 3-phosphate dehydrogenase (GAPDH) was used as an
additional housekeeping gene control. Melting curves were
generated to ensure a single gene-specific peak, and no-
template controls were included for each run and each set
of primers to control for nonspecific amplifications. Data
are represented as relative expression of each gene compared
with ACTB mRNA expression.
Determination of Serum Fleroxacin Levels
Analysis of fleroxacin in serum was conducted by use of
reverse high-performance liquid chromatography (n = 23
per group; University of Tennessee, College of Veterinary
Medicine). Enrofloxacin, a structurally related compound,
was used as the internal standard. The compounds were
extracted from plasma using acetonitrile to precipitate
plasma proteins. The system consisted of a 2695-separation
module and a 2475-fluorescence detector (Waters Corp).
Separation was attained on a Waters Atlantis dC
18
3.9 3
150–mm (5 mm) column preceded by a 5-mmAtlantisguard
column. Standard curves for plasma analysis were prepared
by spiking untreated plasma with fleroxacin, which produced
alinearconcentrationrangeof50to7500ng/mL.Themethod
was accurate and precise across this concentration range.
Histological Analysis
Histological analysis of the healing enthesis was performed
at 2 weeks postoperatively (n = 8 per group). The right supra-
spinatus muscle, supraspinatus tendon, and humerus were
carefully dissected free of all remaining soft tissues immedi-
ately after euthanasia. Tissue samples were fixed in 10%
neutral-buffered formalin (Decal Chemical Corp) for 48
hours. After fixation, tissues were decalcified in EDTA
(Sigma-Aldrich) for 72 hours and washed in PBS solution.
The tissues were then dehydrated and embedded in paraffin;
5mm–thick coronal sections of the repaired supraspinatus
tendon and the greater tuberosity were mounted on silane-
coated slides and stained with safranin-O, hematoxylin and
eosin, and picrosirius red. The greater tuberosity, repaired
tendon-bone insertion, and midsubstance of the supraspina-
tus tendon were examined under light and polarized light
microscopy at 340 and 3100, respectively, to assess fibrocar-
tilage and collagen organization (Eclipse E800; Nikon).
All digital images were captured with a SPOT RT camera
(Diagnostic Instruments) and imported into ImageJ
(National Institutes of Health).Theareaofnewfibrocartilage
formation at the healing enthesis was determined by outlin-
ing the area of metachromasia with safranin-O staining.
Total area for each specimen was measured by use of ImageJ
software.
22
Collagen deposition and maturation at the heal-
ing enthesis were semiquantitatively assessed by measuring
brightness on picrosirius red–stained slides viewed under
polarized light microscopy. Measurements were obtained by
rotating the polarization plane until maximum brightness
was obtained to control for variations in specimen orientation
on the slide. To facilitate comparisons between groups, all tis-
sues were embedded and cut in exactly the same orientation,
and sections were cut to a uniform thickness. After digital
capture, images were imported into ImageJ, whereby the
images underwent 8-bit digitization with a resolution of
640 (horizontal) 3480 (vertical) pixels. This produced images
in which noncollagenous material was dark (zero) and collag-
enous material was depicted in gray scales from 1 to 255.
Five rectangular areas measuring 50 350 mmwereran-
domly selected at the tendon region adjacent to the healing
enthesis, and the gray scales were measured. The light inten-
sities were measured under exactly the same conditions of
illumination for all specimens.
Biomechanical Testing
Biomechanical testing of the repaired tendon-bone inter-
face was performed at 2 weeks postoperatively (n = 10
Vol. XX, No. X, XXXX Fluoroquinolones Impair Tendon Healing 3
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
per group). On the day of testing, each shoulder was thawed
at room temperature, and the humerus with attached
supraspinatus tendon was meticulously dissected under
magnification. All dissections were performed in a blinded
fashion with respect to treatment group. At this time, the
qualitative appearance (tendon swelling, thickening, discol-
oration) of the supraspinatus tendon and its humeral inser-
tion was evaluated, and both were digitally photographed
by 2 blinded observers (A.J.S.F. and M.O.S.). The dimen-
sions of each supraspinatus tendon were measured with
a digital micrometer, and the cross-sectional area was deter-
mined. The specimen was then placed into a custom-
designed uniaxial testing system. The tendon was secured
in a screw grip by use of 280-grit silicon carbide sandpaper,
Insta-set accelerator, and Insta-cure1cyanoacrylate
(Bob Smith Industries). The humerus was secured into
a custom-designed vise grip that prevented fracture through
the humeral physis. The supraspinatus tendon was secured
to a 111-N load cell attached to a linear bearing that allowed
alignment of the tendon in the direction of its pull. The
humeral jig was secured to the linear stage, and grip-to-
grip distance was standardized across all specimens. The
specimen was preloaded to 0.1 N and then loaded to failure
at a rate of 14 mm/s, corresponding to approximately 0.4%
strain. The ultimate load to failure, stiffness, the Young
modulus, maximum stress, energy to failure, and mode of
failure were calculated from tensile tests. Displacement
was measured by use of a 1 mm–resolution micrometer sys-
tem attached to the linear stage. The linear region of the
load-displacement curve was used to calculate the stiffness
for each specimen.
Statistical Analysis
Statistical analysis was performed with SPSS Statistics for
Windows v 20.0 (IBM Corp), with P\.05 considered signif-
icant. One-way analysis of variance (ANOVA) was used to
compare serum fleroxacin levels. Two-way ANOVAs were
used to compare RTqPCR activity, histomorphometric
measures, and biomechanical data between the 4 groups.
All data are expressed as mean 6standard error of the
mean (SEM).
RESULTS
Serum Fleroxacin Levels
High-performance liquid chromatography levels of both
experimental and control groups provided a quantitative
index of the fleroxacin concentration. Mean fleroxacin lev-
els of the Pre/Postop and Postop groups were 64,132 6
4974 ng/mL and 67,124 69010 ng/mL, respectively.
Mean fleroxacin level of the Preop and control groups
was 0 ng/mL.
Gross Examination
Five animals receiving fleroxacin died and were subse-
quently replaced. Necropsy revealed that these deaths
were not due to the fleroxacin treatment itself but rather
to accidental endotracheal intubation, a result of techni-
cian error. The Pre/Postop and Postop animals lost weight
over the duration of the study, while the Preop and control
animals gained weight. Analysis of body weight revealed
that all FQ-treated groups gained significantly less weight
than did control animals, with the most significant
decrease in body weight observed in the Postop and Pre/
Postop groups (2.4 63.2 g [Preop] vs 265.5 63 g [Pre/
Postop] vs 277.4 63.7 g [Postop] vs 20.4 64.2 g [control];
P\.01).
Our initial macroscopic evaluation showed that all rota-
tor cuff repairs were grossly intact at the time of collection.
No failed repairs, proximal humeral physeal fractures, or
suture pullouts from the transosseous tunnels were
encountered. However, the healing enthesis of the Pre/
Postop animals was qualitatively different from the other
groups, and the supraspinatus tendon was friable and atro-
phic. No observable differences were seen between the
other treatment groups.
RTqPCR Analysis
Our results showed a significant upregulation of interleu-
kin (IL)-1bmRNA in the Pre/Postop group compared
with control and Preop groups (Figure 1A), while tumor
necrosis factor (TNF)-aexpression was significantly upre-
gulated in the FQ-treated Postop and Pre/Postop groups
when compared with vehicle-treated controls (Figure 1B).
Further, the mRNA levels of the matrix-degrading
enzymes matrix metalloproteinase (MMP)-3 and MMP-13
were significantly upregulated in the Pre/Postop group
when compared with all other conditions (Figure 1, C
and D). Reduced expression of tissue inhibitor of metallo-
proteinases (TIMP)-2 levels was accompanied by an
increased expression of MMP-3 and MMP-13 levels in all
FQ-treated groups, with significant differences in Postop
versus control groups (Figure 1F). However, the TIMP-1
levels were significantly increased in the Pre/Postop condi-
tion when compared with all other groups (Figure 1E). No
relevant alterations in the expression levels of the apopto-
sis markers p53 and caspase 3 were detected in our analy-
sis (data not shown), and we found no conclusive or
significant results when evaluating the expression levels
of scleraxis and of various collagens (type I a
2
, type II a
1
,
and type III a
1
) (data not shown).
Histological Analysis
Quantitative histomorphometry revealed that the fleroxacin-
treated animals had significantly reduced fibrocartilage at
the healing enthesis compared with the control animals. Ani-
mals in the Preop, Pre/Postop, and Postop fleroxacin-treated
groups had a mean area of new fibrocartilage of 323,946 6
49,139 mm
2
,316,984642,819 mm
2
,and350,7716
68,899 mm
2
,respectively,whereascontrolanimalshad
ameanfibrocartilageareaof628,691697,163 mm
2
(P\
.05) (Figure 2A). Analysis of collagen birefringence with
polarized light microscopy revealed significantly less orga-
nized collagen at the healing enthesis in all 3 fleroxacin-
4Fox et al The American Journal of Sports Medicine
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
treated groups (Preop, Pre/Postop, and Postop) compared
with control animals (65.09 64.05 vs 78.91 610.91 vs
68.90 69.16 vs 99.13 611.67, respectively; P\.05) (Figure
2, B-D).
Biomechanical Testing
There were significant differences in the mode of failure
between the experimental and control animals. The FQ-
treated animals in the Pre/Postop group displayed intra-
substance failure of the supraspinatus tendon during bio-
mechanical testing, whereas only 1 of 10 control samples
failed within the tendon substance, with the remaining
control samples failing at the healing enthesis. In the
Preop group, 6 of 10 samples failed within the tendon
substance, while the remaining 4 samples avulsed from
the supraspinatus muscle. In the Postop group, 5 of 10
samples failed within the tendon substance, while 4
avulsed from the supraspinatus muscle and 1 failed at
the healing enthesis.
The healing enthesis of the Pre/Postop group displayed
significantly reduced ultimate load to failure compared
with the Preop, Postop, and control groups (3.60 60.72 N
vs 9.39 60.64 N vs 8.82 61.61 N vs 7.72 60.84 N, respec-
tively; P\.05) (Figure 3A). There was no significant differ-
ence in load to failure in the Preop group compared with the
Postop group. Pre/Postop animals demonstrated signifi-
cantly reduced cross-sectional area compared with the
Postop group (1.91 60.41 vs 3.40 60.45 mm
2
, respectively)
(P\.05) and control group (4.49 60.38 mm
2
)(P\.001).
Figure 1. Results of reverse-transcriptase quantitative polymerase chain reaction analyses on total RNA isolated from supraspi-
natus tendon: (A) Interleukin (IL)-1bexpression. The Pre/Postop group showed a significant upregulation of IL-1bmessenger RNA
(mRNA) compared with control and Preop animals. (B) Tumor necrosis factor (TNF)-aexpression. Postop and Pre/Postop groups
showed a significant upregulation of TNF-amRNA compared with control and Preop animals. (C) Matrix metalloproteinase
(MMP)-3 expression. The Pre/Postop group showed a significant upregulation of MMP-3 mRNA compared with all other groups.
(D) MMP-13 expression. The Pre/Postop group showed a significant upregulation of MMP-13 mRNA compared with all other
groups. (E) Tissue inhibitor of metalloproteinases (TIMP)-1 expression. The Pre/Postop group showed a significant increase in
TIMP-1 mRNA expression compared with all other groups. (F) TIMP-2 expression. The Postop group showed a significant
decrease in TIMP-2 expression levels compared with the control group. Data are shown as mean 6standard error of the
mean (SEM) (error bars). ACTB, b-actin. *P\.05. **P\.01. ***P\.001.
Vol. XX, No. X, XXXX Fluoroquinolones Impair Tendon Healing 5
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
There was also a significant reduction in cross-sectional
area between the Preop group (2.31 60.31 mm
2
) and con-
trol group (P\.01) (Figure 3B). There were no significant
differences between groups for stiffness, Young modulus,
maximum stress, or energy to failure (data not shown).
DISCUSSION
Fluoroquinolones are commonly administered antimicrobial
agents that target bacterial DNA gyrase (topoisomerase II)
activity
17
and are characterized by good tissue penetration,
broad antimicrobial spectrum, and a relatively low inci-
dence of serious side effects. Deleterious effects of FQ on
tendons have been documented since the 1980s,
1
and while
studies have explored the cellular and tissue responses of
FQ-induced tendinopathy, there have been no investiga-
tions on the reparative capacity of tendons exposed to
FQs. Therefore, the purpose of this study was to determine
the effect of FQ on tendon healing by using an established
rodent rotator cuff repair model.
6
Although the underlying mechanism of FQ-induced ten-
dinopathy and tendon rupture remains unclear, recent
studies suggest that it is likely multifactorial. Inflammation
Figure 2. Histological results. (A) Area of fibrocartilage: All 3 fluoroquinolone (FQ)-treated groups showed significantly less fibro-
cartilage compared with control rats. (B) Collagen organization: All 3 FQ-treated groups showed significantly less organized col-
lagen at the healing enthesis compared with control rats. Safranin-O staining of the (C) control and (D) Pre/Postop groups. There
was a significantly reduced area of fibrocartilage in the Pre/Postop group compared with control rats (340). All data expressed as
mean 6standard error of the mean (SEM) (error bars). *P\.05.
AB
Figure 3. Biomechanical results. (A) Load to failure: The healing enthesis of the Pre/Postop group displayed significantly reduced
ultimate load-to-failure compared with the Preop (P\.01), Postop (P\.01), and control groups (P\.05). (B) Cross-sectional
area measurement: Pre/Postop animals demonstrated significantly reduced cross-sectional area compared with the Postop
and control groups (P\.05 and P\.001, respectively). There was also a significant reduction in area between the Preop
and control groups (P\.01). Data expressed as mean 6standard error of the mean (SEM) (error bars).
6Fox et al The American Journal of Sports Medicine
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
of the paratenon, degenerative changes in tenocytes, and
alterations in collagen were reported in studies of FQ-
treated animals.
21,29,38
Fluoroquinolones have been shown
to have a number of effects on various mammalian cell types
in culture, including both increased and decreased expres-
sion of inflammatory mediators,
32,49
reduced expression of
some ECM proteins (type I collagen, elastin, fibronectin,
and b
1
integrin),
4,8,36,37,48
reduced mitochondrial activity,
4
direct toxicity on collagen,
23
elevated levels of activated cas-
pase 3 (an apoptosis marker),
36
and noncytotoxic inhibition
of canine tendon cell proliferation.
48
Jorgensen et al
19
attributed the histological changes seen in FQ-damaged
Achilles tendons to an ischemic process. The investigators
hypothesized that tendon rupture may be due to a vascular
phenomenon leading to ischemia. However, the plausibility
of this theory is called into question as the majority of ten-
don ruptures occur at sites that are relatively avascular.
15
Tendon pain and degeneration have been associated with
an increase in the normal turnover of matrix proteins.
2,18,33
It has been reported that FQs alter soft tissue structures
and MMP expression in both in vitro
7-10,42,48
and in vivo
models.
37,39
In agreement with these studies, our biochemi-
cal findings suggest that alterations in inflammatory-driven
abnormal ECM remodeling may contribute to FQ-induced
tendon damage. Our findings also suggest that FQs have
no effect on ECM production but rather promote the produc-
tion of a lower quality matrix (eg, type III collagen instead of
type I collagen). The increased IL-1band TNF-aexpression
could contribute to the upregulation in MMP-3 and MMP-13
levels in the Pre/Postop FQ-treated group, indicating a pro-
nounced alteration in the tendon ECM homeostasis. A
recent study of 24 patients with full-thickness rotator cuff
tears reported an upregulation of TIMP-1 and MMP-3
expression, suggesting an association with MMP expression
and rupture and retear of the rotator cuff tendons.
14
These
abnormal patterns in gene expression could account for the
changes in histological and biomechanical properties seen in
the FQ-treated specimens. Furthermore, this model of FQ-
induced tendinopathy may provide scientists and clinicians
with a novel paradigm for evaluating effects of MMPs and
TIMPs on tendon.
The rat supraspinatus tendon has a unidirectional
arrangement of perforating collagen (Sharpey) fibers at
the fibrocartilaginous enthesis. Successful tendon repair
depends, in part, on the integrity of the enthesis and is rec-
ognized as the earliest sign of osseous integration.
34
Composed mainly of proteoglycans and collagen, the enthe-
sis facilitates the transmission and dissipation of tensile,
compressive, and shear forces. Our histological results
indicate that fibrocartilage formation and collagen organi-
zation were deleteriously altered by the administration of
FQ, regardless of whether it was administered preopera-
tively, postoperatively, or both. The FQ-treated animals
had a significant reduction in fibrocartilage and Sharpey
collagen fibers at the healing enthesis, indicating inferior
osseous integration of the tendon to bone. Calcified fibro-
cartilage protects the bone from excessive shear, and colla-
gen protects the tendon from compression. Without a well-
established enthesis, the structural integrity of the repair
will be compromised, resulting in reduced biomechanical
properties. With all FQ-treated groups displaying inferior
material properties, this may explain the anecdotal evi-
dence in the literature of tendinopathy and tendon rupture
with exposure to FQ.
In the present study, we found that the FQ-treated ten-
dons had a significantly reduced load to failure and cross-
sectional area compared with control specimens. We also
observed a predilection for midsubstance tendon rupture
in the FQ-treated animals, which is in contrast to the con-
trol specimens that failed at the healing enthesis. These
findings indicate that FQ alters the microstructural prop-
erties of tendons and their enthesis and may predispose
them to degeneration, findings that are supported in the
literature.
#
Not surprisingly, the Pre/Postop group had
the most significantly altered biomechanical properties.
With a significantly reduced ultimate failure force and
cross-sectional area, our data suggest that FQ may have
a negative effect on tendon ECM degradation, which is
exacerbated after repair. Sustained exposure (ie, pre- and
postoperatively) to FQ is likely to be detrimental, as it
may lead to increased susceptibility to injury or failure
as a result of a degenerated tendon and decreased visco-
elasticity. Interestingly, despite the significant differences
in load to failure, mode of failure, and cross-sectional area
between groups, we did not detect any differences in stiff-
ness, Young modulus, maximum stress, or energy to fail-
ure. The influence of geometry may account for the lack
of observable differences in the Young modulus between
groups, indicating that structural properties are impaired
but material properties are not. Longer time periods are
required to further examine these biomechanical findings.
Although the Pre/Postop group demonstrated the most
dramatic changes histologically, biomechanically, and bio-
chemically, the findings of the Preop and Postop groups are
important to note. Despite the absence of any detectable
serum fleroxacin level at time of harvest, the Preop FQ-
treated group demonstrated a statistically significant
reduction in both tendon cross-sectional area and the
area of fibrocartilage at the healing enthesis as well as
a predilection for intrasubstance failure during biome-
chanical testing. Such findings suggest a structural and
compositional degeneration of the tendon ECM. Interest-
ingly, the changes in the Postop group were more pro-
nounced, with a statistically significant reduction in
collagen organization, fibrocartilage, and TIMP-2; a signif-
icant increase in TNF-a; and a range of failure locations
during biomechanical testing. Down-regulated TIMP-2
has been reported to be associated with tendinopathy.
18,26
These observations suggest that the effects of FQ adminis-
tration are much more pronounced in the postoperative
period, when the sequence of inflammation, repair, and
remodeling occurs, compared with the preoperative period,
when no injury has yet occurred.
Our results have important clinical implications.
Awareness of the association between tendon disorders
and FQ should lead to a careful assessment of patients.
Our results indicate that the risk of structural failure after
#
References 1, 12, 19, 29-31, 35, 43, 44, 47, 50.
Vol. XX, No. X, XXXX Fluoroquinolones Impair Tendon Healing 7
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
soft tissue repair or reconstructive procedures in patients
who have taken FQ may be higher than in patients who
have not taken FQ. The risk-benefit ratio of FQ should
be considered, and patients should be carefully advised of
the possibility of tendon disorders, during or after FQ
treatment, and counseled to immediately suspend treat-
ment at the earliest suspicion of tendinopathy.
13,27
We rec-
ommend delaying surgery (by at least 6 weeks) if a patient
has recently taken an FQ antibiotic. If a patient is taking
FQ preoperatively and surgery is absolutely necessary,
we suggest using a different antibiotic postoperatively.
These recommendations may appear somewhat arbitrary;
however, due to the paucity of data regarding the underly-
ing cause of FQ-induced tendinopathy and tendon rupture,
no absolute recommendations can be provided. Future clin-
ical studies, in addition to focusing on postsurgery drug
protocols, should provide more definitive guidelines for
clinicians on how to treat patients with FQ-induced tendin-
opathy and/or tendon rupture.
The limitations of this study are important. First, fler-
oxacin was selected because of its toxic potential to induce
tendinopathy, as reported by a previous study and con-
firmed by a pilot study at our institution.
20
Nonetheless,
fleroxacin is not representative of the most commonly
administered FQ, nor is it the most commonly reported
FQ associated with inducing tendinopathy. In this regard,
we recognize that this study examines a worst-case sce-
nario of acute toxicity. Second, fleroxacin induced a number
of metabolic alterations throughout the body that led to
a less robust condition in the treated animals compared
with the controls. The loss of weight and increased leth-
argy displayed by the FQ-treated rats may be caused by
a generalized catabolic effect and/or an inflammation-
related systemic effect, which could influence the enthesis
in a nonspecific way such as reduced mechanical load or
altered drug metabolism. This finding of poor health
when FQs are used in animal models, however, is not
unique to our study.
40,45
Third, although the rat is an
appropriate animal model for studying the rotator cuff
based on anatomic considerations,
6
the model approxi-
mates an acute rotator cuff injury and does not represent
the chronic degenerative effects of long-standing FQ ther-
apy. Longer time periods may provide valuable informa-
tion regarding the effect of sustained FQ use on the
tendon and healing enthesis. Fourth, our study may have
benefited from immunohistochemical analysis of various
cytokines, which could help to further define the underly-
ing mechanism of disrupted tendon-bone healing.
SUMMARY
The results of this preliminary study suggest that there is
a significant association between FQ use and impaired ten-
don healing. All FQ regimens, whether administered pre-
operatively, postoperatively, or both, have the potential
to cause a detrimental effect on tendon structure and its
ability to heal. Therefore, we advise that an alternative
antibiotic be administered, particularly when soft tissue
repair is required. Additional studies are necessary to
elucidate the histochemical, histological, and biomechani-
cal changes reported here and to provide potential targets
for therapeutic intervention.
ACKNOWLEDGMENT
The authors thank the Center for Laboratory Animal Serv-
ices staff and Orla O’Shea, MSc, for their assistance with
animal care and histology, respectively.
REFERENCES
1. Bailey RR, Kirk JA, Peddie BA. Norfloxacin-induced rheumatic dis-
ease. N Z Med J. 1983;96:590.
2. Bank RA, TeKoppele JM, Oostingh G, Hazleman BL, Riley GP. Lysyl-
hydroxylation and non-reducible crosslinking of human supraspina-
tus tendon collagen: changes with age and in chronic rotator cuff
tendinitis. Ann Rheum Dis. 1999;58:35-41.
3. Bedi A, Fox AJS, Kovacevic D, Deng X-H, Warren RF, Rodeo SA.
Doxycycline-mediated inhibition of matrix metalloproteinases
improves healing after rotator cuff repair. Am J Sports Med. 2010;
38:308-317.
4. Bernard-Beaubois K, Hecquet C, Hayem G, Rat P, Adolphe M. In
vitro study of cytotoxicity of quinolones on rabbit tenocytes. Cell
Biol Toxicol. 1998;14:283-292.
5. Burkhardt JE, Hill MA, Lamar CH, Smith GN, Carlton WW. Effects of
difloxacin on the metabolism of glycosaminoglycans and collagen in
organ cultures of articular cartilage. Fundam Appl Toxicol. 1993;20:
257-263.
6. Carpenter JE, Thomopoulos S, Flanagan CL, DeBano CM, Soslow-
sky LJ. Rotator cuff defect healing: a biomechanical and histologic
analysis in an animal model. J Shoulder Elbow Surg. 1998;7:599-605.
7. Chang H-N, Pang J-HS, Chen CPC, et al. The effect of aging on
migration, proliferation, and collagen expression of tenocytes in
response to ciprofloxacin. J Orthop Res. 2012;30:764-768.
8. Corps AN, Harrall RL, Curry VA, Fenwick SA, Hazleman BL, Riley GP.
Ciprofloxacin enhances the stimulation of matrix metalloproteinase 3
expression by interleukin-1beta in human tendon-derived cells:
a potential mechanism of fluoroquinolone-induced tendinopathy.
Arthritis Rheum. 2002;46:3034-3040.
9. Corps AN, Harrall RL, Curry VA, Hazleman BL, Riley GP. Contrasting
effects of fluoroquinolone antibiotics on the expression of the colla-
genases, matrix metalloproteinases (MMP)-1 and -13, in human ten-
don-derived cells. Rheumatology (Oxford). 2005;44:1514-1517.
10. Deng Y, Chen B, Qi Y, Magdalou J, Wang H, Chen L. The effects of
levofloxacin on rabbit anterior cruciate ligament cells in vitro. Toxicol
Appl Pharmacol. 2011;257:67-73.
11. Donck JB, Segaert MF, Vanrenterghem YF. Fluoroquinolones and
Achilles tendinopathy in renal transplant recipients. Transplantation.
1994;58:736-737.
12. Franck JL, Bouteiller G, Chagnaud P, Sapene M, Gautier D. Achilles ten-
don rupture in 2 adults treated with pefloxacin, one of the cases with
bilateral involvement [in French]. Rev Rhum Mal Osteoartic.1991;58:904.
13. Gabutti L, Stoller R, Marti HP. Fluoroquinolones as etiology of tendin-
opathy [in German]. Ther Umsch. 1998;55:558-561.
14. Gotoh M, Mitsui Y, Shibata H, et al. Increased matrix metallopro-
tease-3 gene expression in ruptured rotator cuff tendons is associ-
ated with postoperative tendon retear. Knee Surg Sports Traumatol
Arthrosc. 2013;21:1807-1812.
15. Harrell RM. Fluoroquinolone-induced tendinopathy: what do we
know? South Med J. 1999;92:622-625.
16. Hayem G, Carbon C. A reappraisal of quinolone tolerability: the expe-
rience of their musculoskeletal adverse effects. Drug Saf. 1995;13:
338-342.
17. Hooper DC, Wolfson JS. Fluoroquinolone antimicrobial agents.
N Engl J Med. 1991;324:384-394.
8Fox et al The American Journal of Sports Medicine
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from
18. Ireland D, Harrall R, Curry V, et al. Multiple changes in gene expression
in chronic human Achilles tendinopathy. Matrix Biol. 2001;20:159-169.
19. Jorgensen C, Anaya JM, Didry C, et al. Arthropathy with Achilles ten-
don involvement induced by pefloxacin. Apropos of a case [in
French]. Rev Rhum Mal Osteoartic. 1991;58:623-625.
20. Kashida Y, Kato M. Toxic effects of quinolone antibacterial agents on
the musculoskeletal system in juvenile rats. Toxicol Pathol.
1997;25:635-643.
21. Kato M, Takada S, Kashida Y, Nomura M. Histological examination
on Achilles tendon lesions induced by quinolone antibacterial agents
in juvenile rats. Toxicol Pathol. 1995;23:385-392.
22. Koike Y, Trudel G, Curran D, Uhthoff HK. Delay of supraspinatus
repair by up to 12 weeks does not impair enthesis formation: a quan-
titative histologic study in rabbits. J Orthop Res. 2006;24:202-210.
23. Le Huec JC, Schaeverbeke T, Chauveaux D, Rivel J, Dehais J, Le
Rebeller A. Epicondylitis after treatment with fluoroquinolone antibi-
otics. J Bone Joint Surg Br. 1995;77:293-295.
24. Leray H, Mourad G, Chong G, Marcelli C, Borderie P, Mion C. [Spon-
taneous ruptures of the Achilles tendon after kidney transplantation:
use of fluoroquinolones]. Presse Med. 1993;22:1834.
25. Lietman PS. Fluoroquinolone toxicities. An update. Drugs. 1995;
49(suppl 2):159-163.
26. Lo IK, Marchuk LL, Hollinshead R, Hart DA, Frank CB. Matrix metal-
loproteinase and tissue inhibitor of matrix metalloproteinase mRNA
levels are specifically altered in torn rotator cuff tendons. Am J Sports
Med. 2004;32:1223-1229.
27. McGarvey WC, Singh D, Trevino SG. Partial Achilles tendon ruptures
associated with fluoroquinolone antibiotics: a case report and litera-
ture review. Foot Ankle Int. 1996;17:496-498.
28. Moellering RC. The place of quinolones in everyday clinical practice.
Chemotherapy. 1996;42(suppl 1):54-61.
29. Movin T, Gad A, Guntner P, Foldhazy Z, Rolf C. Pathology of the
Achilles tendon in association with ciprofloxacin treatment. Foot
Ankle Int. 1997;18:297-299.
30. Ng WF, Naughton M. Fluoroquinolone-associated tendinopathy:
a case report. J Med Case Rep. 2007;1:55.
31. Pierfitte C, Royer RJ. Tendon disorders with fluoroquinolones. Ther-
apie, 1996;51:419-420.
32. Riesbeck K, Sigvardsson M, Leanderson T, Forsgren A. Superinduc-
tion of cytokine gene transcription by ciprofloxacin. J Immunol.
1994;153:343-352.
33. Riley GP, Curry V, DeGroot J, et al. Matrix metalloproteinase activi-
ties and their relationship with collagen remodelling in tendon pathol-
ogy. Matrix Biol. 2002;21:185-195.
34. Rodeo SA, Arnoczky SP, Torzilli PA, Hidaka C, Warren RF. Tendon-
healing in a bone tunnel: a biomechanical and histological study in
the dog. J Bone Joint Surg Am. 1993;75:1795-1803.
35. Royer RJ, Pierfitte C, Netter P. Features of tendon disorders with flu-
oroquinolones. Therapie. 1994;49:75-76.
36. Sendzik J, Shakibaei M, Schafer-Korting M, Stahlmann R. Fluoroqui-
nolones cause changes in extracellular matrix, signalling proteins,
metalloproteinases and caspase-3 in cultured human tendon cells.
Toxicology. 2005;212:24-36.
37. Shakibaei M, de Souza P, van Sickle D, Stahlmann R. Biochemical
changes in Achilles tendon from juvenile dogs after treatment with
ciprofloxacin or feeding a magnesium-deficient diet. Arch Toxicol.
2001;75:369-374.
38. Shakibaei M, Stahlmann R. Ultrastructure of Achilles tendon from
rats after treatment with fleroxacin. Arch Toxicol. 2001;75:97-102.
39. Simonin MA, Gegout-Pottie P, Minn A, Gillet P, Netter P, Terlain B.
Pefloxacin-induced Achilles tendon toxicity in rodents: biochemical
changes in proteoglycan synthesis and oxidative damage to colla-
gen. Antimicrob Agents Chemother. 2000;44:867-872.
40. Takizawa T, Hasimoto K, Itoh N, Yamashita S, Owen K. A compara-
tive study of the repeat dose toxicity of grepafloxacin and a number
of other fluoroquinolones in rats. Human Exp Toxicol. 1999;18:38-45.
41. Tanne JH. Virginity pledge ineffective against teen sex despite gov-
ernment funding, US study finds. BMJ. 2008;337:a3168.
42. Tsai W-C, Hsu C-C, Chen CPC, et al. Ciprofloxacin up-regulates ten-
don cells to express matrix metalloproteinase-2 with degradation of
type I collagen. J Orthop Res. 2011;29:67-73.
43. van der Linden PD, Sturkenboom MCJM, Herings RMC, Leufkens
HGM, Stricker BHC. Fluoroquinolones and risk of Achilles tendon
disorders: case-control study. BMJ. 2002;324:1306-1307.
44. van der Linden PD, van Puijenbroek EP, Feenstra J, et al. Tendon dis-
orders attributed to fluoroquinolones: a study on 42 spontaneous
reports in the period 1988 to 1998. Arthritis Rheum. 2001;45:235-
239.
45. von Keutz E, Schluter G. Preclinical safety evaluation of moxifloxacin,
a novel fluoroquinolone. J Antimicrob Chemother. 1999;43(suppl
B):91-100.
46. Waterston SW, Maffulli N, Ewen SW. Subcutaneous rupture of the
Achilles tendon: basic science and some aspects of clinical practice.
Br J Sports Med. 1997;31:285-298.
47. West MB, Gow P. Ciprofloxacin, bilateral Achilles tendonitis and uni-
lateral tendon rupture—a case report. N Z Med J. 1998;111:18-19.
48. Williams RJ, Attia E, Wickiewicz TL, Hannafin JA. The effect of cipro-
floxacin on tendon, paratenon, and capsular fibroblast metabolism.
Am J Sports Med. 2000;28:364-369.
49. Yoshimura T, Kurita C, Usami E, et al. Immunomodulatory action of
levofloxacin on cytokine production by human peripheral blood
mononuclear cells. Chemotherapy. 1996;42:459-464.
50. Zabraniecki L, Negrier I, Vergne P, et al. Fluoroquinolone induced
tendinopathy: report of 6 cases. J Rheumatol. 1996;23:516-520.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav
Vol. XX, No. X, XXXX Fluoroquinolones Impair Tendon Healing 9
at CORNELL UNIV WEILL MED COLG on August 30, 2014ajs.sagepub.comDownloaded from