osteomyelitis review

Lebanese Medical Journal 2012 Volume 60 (1) 51

MMIISSEE AAUU PPOOIINNTT// IINN--DDEEPPTTHHRREEVVIIEEWWOSTEOMYELITIS : REVIEW OF PATHOPHYSIOLOGY, DIAGNOSTIC MODALITIESAND THERAPEUTIC OPTIONShttp://www.lebanesemedicaljournal.org/articles/60-1/review1.pdf

Albert J. EID1, Elie F. BERBARI2

INTRODUCTION

Osteomyelitis is an infection of the bone and bone mar-row. It can involve any bone of the human body and canoccur through a variety of mechanisms. The difference inpathophysiology of various types of osteomyelitis man-dates specific therapeutic strategies aimed at the eradica-tion of the infection while preserving bone integrity andfunction. Osteomyelitis can be acute or chronic and mayresult from hematogenous seeding of the bone, extensionof the infection from adjacent soft tissue, or direct inocu-lation of the bone through skin and soft tissue defects fol-lowing trauma or surgery. Once established, osteomyelitisleads to progressive softening and necrosis of the boneresulting into the formation of sequestra. At this stage ofthe disease, surgical debridement becomes a requirementfor cure. The presence of hardware at the site of infectionis a complicating factor that could compromise treatmentsuccess. Despite adequate treatment, failure is not uncom-mon. Therefore, in order to improve the outcome, clini-cians should understand the disease process and gainexperience manipulating the treatment tools. Hence, theimportance of this review.

Herein, we will address the pathogenesis of osteo-myelitis, the available diagnostic modalities, and varioustherapeutic strategies. Due to the difference in pathophys-iology, microbiology, and management, we will separate-ly discuss different entities including osteomyelitis due toopen fractures, vertebral osteomyelitis, acute hematoge-nous osteomyelitis, osteomyelitis in patients with diabetesmellitus and osteomyelitis associated with prosthetic jointinfection. A brief overview of osteomyelitis due to somespecific pathogens such as Brucella species, Salmonellaspecies, mycobacteria and fungi will also be reviewed.

PATHOGENESIS OF OSTEOMYELITIS

Experimental animal model of bone infection has reveal-ed that bone is resistant to infection [1]. Osteomyelitisoccurs only when a large inoculum of bacteria is intro-duced into the bone in conjunction with trauma, necrosis,or presence of foreign bodies [2]. Bacteria (i.e. Staphy-lococcus aureus) adhere to bone matrix via receptors tofibronectin, laminin, collagen and other structural proteins.Microorganisms elude the host defenses and antibioticsthrough a variety of mechanisms including surviving in adormant state inside osteoblasts, developing a biofilm, andacquiring a very slow metabolic rate [3-4]. Furthermore,studies in a guinea pig model of post-traumatic osteo-myelitis showed significant reduction of leukocyte loco-motion after trauma and infection with S. aureus for up to90 days [5]. The presence of prosthetic material contigu-ous to bone can lead to specific polymorphonuclear leu-

Eid AJ, Berbari EF. Osteomyelitis : Review of pathophysiology,diagnostic modalities and therapeutic options. J Med Liban 2012 ; 60 (1) : 51-60.

Eid AJ, Berbari EF. Lostomylite : tude approfondie de lapathophysiologie, des modalits diagnostiques et des options thrapeutiques. J Med Liban 2012 ; 60 (1) : 51-60.

1Division of Infectious Diseases, University of Kansas MedicalCenter, Kansas City, Kansas; 2Division of Infectious Diseases,College of Medicine, Mayo Clinic, Rochester, Minnesota, USA.

Correspondence: Albert J. Eid, MD. Division of InfectiousDiseases. University of Kansas Medical Center, 2067 Delp.3901 Rainbow Blvd. Kansas City. KS 66160 USA.

e-mail: [email protected] Tel.: +1 913 588 6035

ABSTRACT : Osteomyelitis can affect every boneand is heterogeneous in its pathophysiology and pre-sentation. When the diagnosis is clinically suspected,further studies such as serum inflammatory markersand imaging studies should be performed. Magneticresonance imaging can be very useful in establishingthe diagnosis and determining the extent of infection.When possible, bone specimens should be obtainedand cultured prior to the initiation of antimicrobialtherapy. Surgical debridement is often required forchronic or contiguous osteomyelitis for successfuleradication of the infection. The ultimate test-of-cureis the lack of clinical relapse after the discontinuationof antimicrobials.

RSUM : Lostomylite peut atteindre tous les os dusquelette humain et possde une pathophysiologie etune prsentation clinique trs htrognes. Lorsque lediagnostic est considr, certains examens doivent tre obtenus, notamment les marqueurs dinflammation etdes examens radiologiques. Limagerie par rsonancemagntique peut tre trs utile pour confirmer le diag-nostic et dterminer ltendue de linfection. Lorsquecela savre possible, des chantillons osseux doiventtre obtenus et cultivs avant le dbut de lantibioth-rapie. Un dbridement chirurgical est souvent nces-saire pour lradication complte de lostomylitechronique ou contigu. Le test ultime de gurison estlabsence de rechute clinique aprs larrt des antibio-tiques.

52 Lebanese Medical Journal 2012 Volume 60 (1) A.J. EID, E.F. BERBARI Osteomyelitis

kocytes defect and therefore protects bacteria from phago-cytosis. Inflammation, resulting from the interaction be-tween bacteria and leukocytes, results in the release ofcytokines and the development of osteolysis [3-4].

In patients with acute osteomyelitis of the long bones,it is believed that the metaphysis is the site of predilectiondue to slow blood flow in the metaphyseal vascular loopsand the lack of phagocytic lining cells [6]. Once the infec-tion is established in the metaphysis, the inflammatoryexudate leads to increased pressure in the bone and intra-medullary canal. Subsequently, the extension of this exu-date into the cortex will eventually lead to periosteal ele-vation or rupture, disrupting the blood flow and causingbone infarction and the formation of an abscess or seques-trum. In chronic osteomyelitis, the inflammation is mild tomoderate and little or no ischemic necrosis occurs [4].

The type and site of osteomyelitis is largely determinedby the mechanism of infection, the virulence of the infect-ing organism, and the immune status and comorbid condi-tions of the patient [6-7] (Table I). Osteomyelitis coulddevelop through hematogenous seeding of the bone from aremote source of infection, extension of the infection fromadjacent soft tissue overlying the involved bone; or direct-ly via inoculation of the bone following trauma or surgery.

S. aureus is the most commonly isolated organism inosteomyelitis. Other microorganisms such as coagulase-negative staphylococci, aerobic gram-negative bacilli, andanaerobes are frequently encountered as well (Table II).Patients with certain conditions such as immunosuppres-sion, immune diseases (i.e. rheumatoid arthritis), diabetesmellitus, smoking, malnutrition, malignancy, extremes ofage, chronic hypoxia, and renal or hepatic failure are atincreased risk for osteomyelitis. Other local factors areequally important. Those include chronic edema, peripher-al vascular disease, neuropathy, prior surgery, extensivescarring, and radiation fibrosis [6-8].

DIAGNOSIS AND DIAGNOSTIC MODALITIES

Osteomyelitis should be suspected when patients presentwith pain, swelling, erythema or warmth of the skin andsoft tissue overlying bone. Subacute or chronic pain iscommonly the only manifestation. Systemic symptoms(i.e. fever and chills) are present in patients with acuteosteomyelitis but seldom present in patients with chronicosteomyelitis. Draining sinuses are typically seen in casesof chronic osteomyelitis. The presence of exposed bone or prosthetic material for a period of time is highly sug-

TABLE IETIOLOGY OF OSTEOMYELITIS AND UNDERLYING CONDITIONS

Mechanism Special groups Etiology Hematogenous osteomyelitis

Common in children and adults Staphylococcus aureusNeonates Enterobacteriaceae, Group B streptococciInfants and children* Haemophilus influenzae group BInjection drug users Staphylococcus aureus, Pseudomonas aeruginosa,

Candida spp.Vertebral osteomyelitis

Most common in adults Staphylococcus aureusUrinary tract infection Aerobic gram-negative bacilli, Enterococcus spp.Injection drug users Pseudomonas aeruginosa, Staphylococcus aureus,

Candida spp.Following spine surgery Coagulase-negative staphylococci, Staphylococcus

aureus, aerobic gram-negative bacilliInfections of intravascular devices Candida spp.In endemic regions Mycobacterium tuberculosis, Brucella spp.

Contiguous focus osteomyelitisDiabetes mellitus, vascular insufficiency, Polymicrobial: Staphylococcus aureus, beta-hemolyticor following a contaminated open fracture streptococci, Enterococcus spp., aerobic gram-negative

bacilliSoil contamination Clostridium spp., Bacillus spp., Stenotrophomonas

maltophilia, Nocardia spp., atypical mycobacteria,Aspergillus spp., Rhizopus spp., Mucor spp.

Orthopedic fixation devices Staphylococcus aureus, coagulase-negative staphylococciCat bites Pasteurella multocidaFollowing puncture injuries on the foot Pseudomonas aeruginosaby nails or other sharp objectsFollowing periodontal infection Actinomyces spp.

*The incidence of osteomyelitis due to Haemophilus influenza type B significantly decreased due to the routine vaccination of children.

A.J. EID, E.F. BERBARI Osteomyelitis Lebanese Medical Journal 2012 Volume 60 (1) 53

gestive of osteomyelitis. In fact, the probe-to-bone test iswidely used to diagnose osteomyelitis in patients with dia-betic foot ulcers and contiguous osteomyelitis. Grayson et al. found that this test has a sensitivity of 66% and apositive predictive value of 89% [9].

The confirmation of osteomyelitis requires the use of a variety of laboratory, microbiologic, radiographic andpathologic tests. Erythrocyte sedimentation rate (ESR)and C-reactive protein (CRP) are usually abnormal. Thewhite blood cell count is occasionally elevated. Theplatelet count can be elevated (inflammatory marker)while the hemoglobin concentration can be low (anemiaof chronic disease). Blood cultures may be positive inacute hematogenous and vertebral osteomyelitis. Culturesof superficial wounds or sinus tracts should be cautiouslyinterpreted and should not be used to guide antimicrobialtherapy unless S. aureus is isolated from those cultures[10-11]. In one study, only 44% of the sinus tract culturesgrew the same operative pathogen. The isolation of S. au-reus from the sinus tract correlated with the presence of S. aureus in the operative specimen. However, less thanhalf of the sinus tract cultures obtained from patients withS. aureus osteomyelitis contained this organism [11]. Bonetissue sampling via needle aspiration under radiologicguidance or surgical procedure allow the identification ofthe infectious organism and the determination of its in vitrosusceptibility profile [6]. The latter information is crucialfor the initiation of appropriate and effective antimicrobialtherapy (Table III). The bone tissue collected from theinfected site can also be submitted for histopathologicexamination which is considered the gold standard for thediagnosis of osteomyelitis.

Conventional radiography has little value in diagnosingacute osteomyelitis but it might be helpful in cases ofchronic osteomyelitis. At least 10-14 days are needed be-fore abnormalities consistent with osteomyelitis are seen.In one study, the sensitivity of plain radiographs in cases of diabetic foot osteomyelitis was found to be 54%, whilethe specificity was 68%. The radiographic signs that wereconsidered diagnostic included focal or geographic areasof marrow lucency, loss of cortex with bony erosion, new

bone formation, bone sclerosis with or without erosion,sequestration, involucrum, and periosteal elevation [12].

Nuclear bone scans using a variety of radiotracers(Technetium 99m methylene diphosphonate, Gallium-citrate 67, and Indium 111-labeled white blood cells) arecommonly used to diagnose osteomyelitis. The perfor-mance of those scans varies depending upon the clinicalsituation [13]. In adults with normal radiographs (nolesions that cause increased bone turnover), the three-phase bone scan has higher accuracy than the other scanswith 94% sensitivity and 95% specificity. However, whenbone remodeling is increased, the specificity of the testdecreases to 33% [13]. In presence of osteomyelitis, sig-nificant uptake is seen in all three phases of the study(immediately after injection, at 15 minutes and 4 hours) asopposed to increased uptake only in the first two phases inpatients with cellulitis. Gallium scans detect the areas ofinflammation due to the affinity of gallium-67 to acutephase reactants (i.e. lactoferrin and transferrin) [14]. Theresults of studies evaluating their sensitivity and specifi-city are not consistent. When studies are combined togeth-er, the overall sensitivity is approximately 81% and thespecificity is 69% [13]. Some authors believe that galliumscans have better performance for the evaluation of verte-bral osteomyelitis. This is supported by data from onestudy showing high sensitivity, specificity and accuracy[15]. Tagged white blood cell scans show accumulation oftagged white cells in the bone marrow and at the site ofinflammation or infection [14]. In a meta-analysis, the sen-sitivity of Indium 111-labeled white blood cells scan was84% for extra-axial chronic osteomyelitis as opposed toonly 21% for the axial skeleton. Similarly, the specificitywas 80% in the peripheral skeleton and 60% in the axialskeleton [16].

Positron emission tomography (PET) using 18-fluoro-deoxyglucose is increasingly being utilized in the diag-nosis of osteomyelitis. In a systematic review and meta-analysis, Termaat et al. found PET scans to have a sensi-tivity of 96% and a specificity of 91% for the diagnosis ofosteomyelitis. [16]. PET scanning is a cheaper modalitywhen compared with other nuclear bone scanning tech-niques and is typically performed in one day. Unfortu-nately, however, false positive results can be seen withbone healing. Computed tomography (CT) providesexcellent cortical bony details showing cortical bone ero-sion or destruction and periosteal reaction. It could alsoshow small foci of air within the medullary canal, tiny for-eign bodies serving as a nidus for infection and seques-trum formation [17]. Magnetic resonance imaging (MRI)is more sensitive than CT in detecting osteomyelitis andas sensitive as nuclear studies. The sensitivity and speci-ficity of MRI ranged from 82% to 100% and 75% to 96%,respectively. MRI is considered the imaging modality ofchoice in cases of established osteomyelitis because itallows accurate determination of the extent of the infec-tion, especially in case of vertebral osteomyelitis (iden-tifies epidural abscess, phlegmon, and cord compression)[17].

TABLE II MICROBIOLOGY OF OSTEOMYELITIS

Common Less common

Staphylococcus aureus Mycobacterium tuberculosisCoagulase-negative staphylococci Rapidly growing mycobacteriaStreptococci Mycobacterium avium complexEnterococci Endemic fungiPseudomonas spp. Candida spp.Enterobacter spp. Aspergillus spp.Proteus spp. Mycoplasma spp.Escherischia coli Brucella spp.Serratia spp. Salmonella spp.Anaerobes (Peptostreptococcus spp., Actinomycetes

Clostridium spp., B. fragilis group.) Tropheryma whipplei

THERAPEUTIC STRATEGIES

The optimal treatment of osteomyelitis is still not welldefined due to the lack of prospective randomized con-trolled clinical trials. Most of the data supporting treat-ment guidelines derive from studies in animal models,expert opinions, and retrospective studies in humans. Ingeneral, treatment consists of a combination of surgicaldebridement and antimicrobial therapy [6]. Both treat-ment components should be tailored to manage variousforms of osteomyelitis. The details of treatment will bediscussed in the following sections addressing specificosteomyelitis entities.

Medical managementIn order to establish a microbiological diagnosis andunless the patient is hemodynamically unstable, everyeffort should be directed at obtaining bone specimens priorto the initiation of empiric antimicrobial therapy. The opti-mal duration of antimicrobial therapy for osteomyelitis isnot known. In an animal model, S. aureus grew from 78%of specimens after 14 days of therapy with clindamycin butonly from 16% of specimens after 28 days of therapy. Inthis model, surgical debridement was not part of the treat-ment, however [18]. In the clinical setting, relapse is notuncommon even after 4 weeks of antibiotic therapy. There-fore, many experts recommend 4 to 6 weeks of therapy [6].In some instances, when surgical debridement is subop-timal or lacking, treatment is extended to 8-10 weeks.During treatment, the inflammatory markers (ESR andCRP) are expected to go down when compared to baselinevalues. In one study of vertebral osteomyelitis, the rate ofclinical failure when ESR fell or failed to fall by more than25% of the baseline value during the first month of treat-ment was 12% and 50%, respectively [19].

Parenteral antimicrobial therapy is commonly used;however, oral antimicrobials with good bioavailability andbone penetration (e.g. trimethoprim-sulfamethoxazole,clindamycin, tetracyclines, fluoroquinolones, metronida-zole, and linezolid) can be used if compliance is ascer-tained. Betalactams (i.e. intravenous penicillins and cepha-losporins) are commonly used to treat osteomyelitis due to their efficacy and relative safety when given for a pro-longed period of time [6] (Table III). Vancomycin is usedto treat methicillin-resistant S. aureus (MRSA) and am-picillin-resistant Enterococcus species; however, vanco-mycin use is associated with higher treatment failure ratewhen compared with betalactams [20]. Data regarding theuse of newer drugs such as linezolid and daptomycin forthe treatment of osteomyelitis due to MRSA and vanco-mycin-resistant Enterococcus (VRE) are limited [21-23].

In recent years, osteomyelitis due to resistant Gram-negative bacilli (i.e. multidrug resistant Pseudomonasaeruginosa, extended-spectrum -lactamase producingKlebsiella pneumonia and Escherichia coli) has become aserious therapeutic challenge. In those cases, antimicro-bial therapy should always be guided by in vitro suscepti-bility studies. Intravenous colistin and carbapenems (i.e.doripenem) are frequently used and synergy studies couldalso be considered in difficult cases.

The data from animal models suggest that fluoro-

TAB

LE III

ANTI

MIC

RO

BIAL

THER

APY

FOR

O

STEO

MYE

LITI

S D

UE TO

SP

ECIF

IC M

ICR

OO

RG

ANIS

MS

Mic

roor

gani

sms

Firs

t cho

ice

Alte

rnat

ive

Cho

ice

Stap

hylo

cocc

i, ox

acill

in-s

ensi

tive

Naf

cillin

sod

ium

or o

xacil

lin s

odiu

m, 1

.5-2

.0 g

IV q

4 h

r for

4-6

wks

Vanc

omyc

in*,

20 m

g/kg

IV q

12

hr fo

r 4-6

wks

OR

Cef

azol

in 1

-2 g

IV q

8 h

r for

4-6

wks

Stap

hylo

cocc

i, ox

acill

in-re

sist

ant

Vanc

omyc

in*,

15 m

g/kg

IV q

12

hr fo

r 4-6

wks

Line

zolid

600

mg

PO/IV

q 1

2 hr

for 6

wks

O

RLe

voflo

xacin

**50

0-75

0 m

g PO

/IV d

aily

and

rifam

pin

600

mg

PO d

aily

for 6

wks

Peni

cilli

n-se

nsiti

ve s

trept

ococ

ciAq

ueou

s cr

ysta

lline

peni

cillin

G, 2

0 x

106

U/2

4 hr

IV, e

ither

con

tinuo

usly

orVa

ncom

ycin

*, 20

mg/

kg IV

q 1

2 hr

for 4

-6 w

ksin

6 e

qual

ly di

vided

dai

ly do

ses

for 4

-6 w

ksO

RC

eftri

axon

e 1-

2 g

IV o

r IM

q 2

4 hr

for 4

-6 w

ksO

RC

efaz

olin

1-2

g IV

eve

ry 8

hr f

or 4

-6 w

ksEn

tero

cocc

i or s

trept

ococ

ci w

ithAq

ueou

s cr

ysta

lline

peni

cillin

G, 2

0 x

106

U/2

4 hr

IV, e

ither

con

tinuo

usly

orVa

ncom

ycin

*, 20

mg/

kg IV

q 1

2 hr

for 4

-6 w

kspe

nici

llin

MIC

0.

5 g

/mL

orin

6 e

qual

ly di

vided

dai

ly do

ses

for 4

-6 w

ksO

ptio

nal:

Add

Gen

tam

icin

sulfa

te, 1

mg/

kg IV

or I

M e

very

8 h

r for

1-2

wks

Abi

otro

phia

def

ectiv

a,O

RAm

picil

lin s

odiu

m 1

2 g/

24 h

r IV

eith

er c

ontin

uous

ly or

in 6

equ

ally

divid

edG

ranu

licat

ella

spp.

, Gem

ella

spp.

daily

dos

es fo

r 4-6

wks

Opt

iona

l: Ad

d G

enta

mici

n su

lfate

, 1 m

g/kg

IV o

r IM

eve

ry 8

hr f

or 1

-2 w

ksEn

tero

bact

eria

ceae

Cef

triax

one

2 g

IV q

24

hr fo

r 4-6

wks

Cip

roflo

xacin

** 5

00-7

50 m

g PO

q 2

4 hr

for 4

-6 w

ksPs

eudo

mon

as a

erug

inos

aor

Cef

epim

e 2

g IV

q 1

2 hr

for 4

-6 w

ksC

ipro

floxa

cin**

750

mg

PO q

12

hr fo

r 4-6

wks

Ente

roba

cter

spp.

OR

Mer

open

em 1

g IV

q 8

hr f

or 4

-6 w

ksO

RC

efta

zidim

e 2

g IV

q 8

hr f

or 4

-6 w

ks

hr: h

our

IM

: int

ram

uscu

lar

IV:

intra

veno

us

wks

: wee

ks

PO

: per

os

*D

ose

base

d on

nor

mal

cre

atin

ine

clea

ranc

e; v

anco

myc

in tr

ough

leve

l sho

uld

be 1

5-20

.**

Shou

ld b

e av

oide

d if

poss

ible

, in

the

pedi

atric

pop

ulat

ion

and

in o

steo

mye

litis

ass

ocia

ted

with

frac

ture

s.


quinolones significantly impair fracture healing [24];therefore, experts recommend caution when using fluoro-quinolones for the treatment of osteomyelitis in the contextof fractured bone [6]. The use of rifampin for the treatmentof staphylococcal osteomyelitis could be beneficial due tobiofilm penetration and increased bactericidal activitywhen co-administered with other antibiotics [25-26], espe-cially in case of hardware-associated osteomyelitis [27-28]. Hyperbaric oxygen therapy (HBOT) can be used incases of chronic and refractory osteomyelitis but its useremains highly controversial [29].

Surgical treatmentThe goal of surgery in patients with osteomyelitis is todrain all infected fluid at the site of infection, completelydebride all infected and devitalized soft tissue and bone,and achieve good wound coverage. Ideally all infectedhardware should be removed; however, this is not possi-ble when the fractured bone is not stable enough. Whensignificant soft tissue defect is present, musculocutaneousflaps can be used to eliminate the dead-space and provideadequate coverage of the wound. Patients presenting withsevere and recurrent osteomyelitis might require multiplesurgical debridements in order to insure complete removalof the infected tissue. In those situations, antibiotic impreg-nated polymethyl metacrylate (PMMA) beads can beplaced deep into the wound around the infected bone.Those beads allow the local delivery of higher concentra-tions of antibiotics; however, they should be ultimatelyremoved from the wound.

MANAGEMENT OF PATIENTS WITH VARIOUS TYPESOF OSTEOMYELITIS

Osteomyelitis in patients with contaminated openfractureOsteomyelitis can develop at the site of an open fracturein 3% to 25% of cases [30-31]. Factors determining therate of infection include the type of fracture, degree ofcontamination of the wound (fracture type IIIB), extent ofsoft tissue injury, the timing of surgical debridement, theuse of systemic, and local antimicrobials (i.e. PMMAbeads) [30-35]. Patients are usually young men presentingwith fractures of the lower extremities, mainly the tibiaand fibula. Short course of systemic antibiotic therapy (24 hours) in addition to the use of antibiotic-impregnatedPMMA beads is recommended [6] in order to minimizethe risk of subsequent osteomyelitis. Antibiotic-impreg-nated intramedullary nails are being used as well and the preliminary data from retrospective studies showedpromise [36-37]. The impact of osteomyelitis in thosecases could be devastating due to nonunion of the infect-ed bone and the need for amputation.

Patients typically present several months after the acci-dent with nonunion at the fracture site and poor woundhealing or draining sinuses. Local signs and systemicsymptoms are not common. Imaging studies are not help-ful in establishing the diagnosis of osteomyelitis due to

bony changes secondary to trauma and surgery and arti-facts due to prosthetic material. Following a course of par-enteral antimicrobial therapy, oral suppressive therapyshould be considered and continued until the fractureheals. In case of relapse of the infection after bone heal-ing, the hardware should be removed and another courseof parenteral antimicrobial therapy should be given [6].

Vertebral osteomyelitis and spondylodiskitis Vertebral osteomyelitis and intervertebral disk infectioncommonly result from hematogenous seeding via the seg-mental artery supplying two adjacent end plates of conti-guous vertebrae [38]. Less frequently, vertebral osteo-myelitis and disk space infection complicates spinalsurgery. The routine use of preoperative antibiotic pro-phylaxis further reduced the incidence of postoperativespinal infection. In one study, the rates of infection withand without antimicrobial prophylaxis were 0.2% and2.8%, respectively [39]. The infection involves most com-monly the lumbar spine followed by the thoracic and cer-vical spine [40-41]. Vertebral osteomyelitis and disk spaceinfection can be complicated with paravertebral and epi-dural abscesses. The latter could lead to spinal cord ornerve root compression and require urgent surgical inter-vention.

The majority of patients with vertebral osteomyelitispresent with back pain while less than half have fever.Neurological symptoms such as motor or sensory deficits,bladder and bowel incontinence are not uncommon andshould receive urgent attention. S. aureus is the most com-monly isolated organism while coagulase-negative staphy-lococcus and gram-negative bacilli are less commonlyencountered. Pseudomonas aeruginosa and Candida spp.are typically seen in intravenous drug users. Vertebralosteomyelitis due to Brucella spp. and Mycobacteriumtuberculosis is common in endemic areas. When bactere-mia is present, endocarditis should be ruled out [41-43].

Spine MRI is the test of choice when vertebral osteo-myelitis is suspected. Gallium or PET scans are alterna-tives when MRI cannot be performed (i.e. patients withclaustrophobia or patients with implantable cardiovasculardevices). If the diagnosis is supported by imaging studies,a microbiological diagnosis should be established. CT-guided percutaneous biopsies are easy to perform but havelimited sensitivity (52%) [44]. When the results of the firstpercutaneous biopsy are inconclusive, a second CT-guidedbiopsy should be attempted before resorting to an openbiopsy. Tissue should be submitted for pathology as well asbacterial, fungal and mycobacterial stains and cultures.When the patient has concomitant bacteremia, it is safe toassume that his vertebral osteomyelitis is caused by thesame organism isolated from the blood. Surgical debride-ment is necessary when patients present with an epiduralabscess compressing the spinal cord, large paravertebralabscess, spine instability or failing medical treatment [6].

Effective antimicrobial therapy should be given for atleast four to six weeks. When vertebral osteomyelitis istreated with surgical debridement and instrumented spine


fusion or when hardware-associated osteomyelitis devel-ops following spinal surgery, some experts recommendparenteral antibiotic therapy followed by oral suppressiveantibiotic therapy for up to two years to allow bone heal-ing [6].

Acute hematogenous osteomyelitisAcute hematogenous osteomyelitis occurs mainly in pre-pubertal children. Long tubular bones (i.e. femur, tibia, andfibula) are more commonly affected than flat bones andspine [45]. Multiple bones could be infected concomitant-ly. The infection originates in the metaphyseal region andcan extend to the cortex or the epiphysis and joint space. S. aureus and Streptococcus pneumoniae are responsiblefor most cases. Children present with pain in the metaphy-seal region of a long bone. Blood cultures are typicallypositive. Radiological findings such as periosteal newbone formation (periosteal elevation) or bone destructionsupport the diagnosis. If plain radiographs are not conclu-sive, MRI may establish the diagnosis. Blood culturesshould be obtained before administering antimicrobials.Bone, subperiosteal exudate, and synovial fluid should beobtained if possible in order to identify the infecting micro-organism. Acute hematogenous osteomyelitis in children isfrequently treated with antimicrobial therapy alone. Anti-biotics are typically given for three weeks. Sequential ther-apy with intravenous followed by oral antibiotics is com-monly used. Surgical treatment is considered in case of no clinical response after 48 hours of antibiotic therapy,septic arthritis, or failure to cure the infection despite ade-quate antibiotic therapy. Adequate and prompt treatmentleads to excellent outcome in newborns, infants and chil-dren; however, in cases associated with septic arthritis orcomplicated with the destruction of the growth plate, theprognosis is more reserved. In those situations, abnormalgrowth of the affected bone is likely to occur and signi-ficant damage of the infected joint might be irreversible[46].

Osteomyelitis in patients with diabetes mellitusPatients with diabetes mellitus are at increased risk ofdeveloping foot infections due to hyperglycemia, neuro-pathy, and peripheral vascular disease. During their life-time, 15% of diabetic patients develop foot ulcers and 6%are hospitalized because of those ulcers [47].

The infectious process usually originates at the site of anulcer and commonly progresses to reach contiguous bone.When the size of an ulcer is > 2 x 2 cm and when ESR is > 70 mm/hour, osteomyelitis is more likely to be present[48]. The presence of exposed bone or the ability to probethe bone is strongly suggestive of underlying osteomyelitis(positive predictive value of 89%) [9, 48-49]. Further test-ing is not required when the probe-to-bone test is positive.In other situations, plain radiographs, nuclear bone scans,or MRI could be helpful to establish the diagnosis. ESRand CRP should always be measured at baseline and moni-tored during therapy to assess response.

Chronic osteomyelitis in patients with diabetes mellitus

is usually polymicrobial. Multiple aerobic and anaerobicbacteria can be involved. When possible, bone tissueshould be submitted for cultures prior to the initiation ofantimicrobial therapy. When hemodynamically significantstenosis is present, revascularization should be consideredbefore surgical debridement of the infected foot. Intra-venous antimicrobials such as piperacillin-tazobactam,ticarcillin-clavulanate, ampicillin-sulbactam, Ertapenem,or other -lactams in combination with metronidazole arecommonly used. An oral regimen consisting of levofloxa-cin or moxifloxacin with metronidazole or clindamycincan also be used [50]. HBOT might be beneficial in thiscontext due to limited oxygenation of the infected tissueand the likelihood of anaerobic infection; however, datafrom randomized, controlled studies are lacking. Limit-ed amputation of the limb should be considered in caseswith persistent or recurrent osteomyelitis despite optimaltherapy.

Osteomyelitis associated with prosthetic joint infectionJoint arthroplasty represent a true success story due to itsability to restore function. The projected number of pa-tients who will undergo arthroplasty in the United States in2030 is 3-4 millions [51]. It is estimated that 0.8 to 1.9% ofknee arthroplasties and 0.3 to 1.7% of hip arthroplasties getinfected [52]. The treatment also consists of a combinationof surgical and antimicrobial therapy. Three different ther-apeutic strategies are available:

1. Surgical debridement and retention of the prosthesis isan option when certain criteria are met. Those includeduration of symptoms < 3 weeks, infection develop-ing within 3 months of arthroplasty or presentationconsistent with hematogenous seeding, well-fixedprosthesis, adequate soft tissue envelop, absence ofsinus tract, and established microbiologic diagnosisand favorable susceptibility profile [52-53].

2. One-stage exchange entails resection of the infectedprosthesis, debridement of the infected tissue and im-mediate implantation of a new prosthesis. Antibiotic-impregnated PMMA cement is frequently used dur-ing reimplantation. This strategy is not advisablewhen the infection is caused by virulent organisms(i.e. S. aureus, Gram-negative bacteria) or in presenceof draining sinus or significant purulence.

3. Two-stage exchange consists of resection of the in-fected prosthesis followed by parenteral antimicro-bial therapy for 4-6 weeks and reimplantation of anew prosthesis 2-6 weeks after finishing antimicro-bial therapy. The use of a spacer impregnated with an-timicrobial(s) (i.e. vancomycin, gentamicin or tobra-mycin) is a common practice currently and wasshown to reduce the rate of recurrent infection [54].

Antimicrobial therapy should be guided by in vitro sus-ceptibilities (Table III). Following debridement and reten-tion of the prosthesis, Rifampin was successfully used incombination with fluoroquinolones. The duration of treat-ment was 3-6 months for prosthetic knee infection and 6 months for prosthetic hip infection [27, 55]. Shorter


course (4 weeks) of parenteral antibiotic therapy can begiven followed by suppressive oral antibiotic therapy [56].Rifampin-based regimen is also recommended after one-stage exchange.

Osteomyelitis in injection drug usersIntravenous drug use is associated with hematogenousosteomyelitis due to the introduction of bacteria and fungiinto the bloodstream. However, direct inoculation or con-tiguous spread is also possible [6]. Multifocal infectionsmay occur. The spine, sterno-clavicular, sterno-chondral,and sacro-iliac joints are commonly involved [57-58]. Themost common organisms responsible for osteomyelitis ininjection drug users (IDU) are S. aureus, Pseudomonasspecies, and Candida species. Eikenella corrodens, whichis part of the normal oral flora, is seen in IDU who licktheir needles or skin prior to injection (needle lickersosteomyelitis) [59]. M. tuberculosis can cause skeletalinfection in IDU and most cases involve the spine [60].Treatment of osteomyelitis in IDU is similar to treatmentof other patients; however, an oral regimen should beused, if at all possible, to avoid using long-term intra-venous access (e.g. peripherally inserted intravenouscatheter).

Osteomyelitis due to Brucella speciesBrucellosis is endemic in the Mediterranean basin and theArabian Gulf; therefore, osteomyelitis due to Brucellaspecies should be considered in this setting. Brucellosis iscommonly transmitted via ingestion of contaminated foodsuch as raw milk, cheese made from unpasteurized milk, orraw meat. Infection could also result from direct skin inoc-ulation through skin abrasions or cuts during contact withanimal carcasses, placentas, and animal vaginal secretions.Patients present with fever, night sweats, arthralgias,anorexia and fatigue. When the diagnosis is delayed, local-ized infection occurs. Osteoarticular infections, main-ly sacroiliitis, represent 20-30% of those localized infec-tions [61-62]. Specific cultures (Ruiz-Castaneda or lysiscentrifugation) or prolonged incubation of cultures can be used to establish a diagnosis. Serological studies usingserum agglutination (standard tube agglutination) andenzyme-linked immunosorbent assay (ELISA) can be very helpful, especially in case of negative cultures. Thecombination of doxycycline-streptomycin or doxycycline-streptomycin-rifampin seem to be more effective thandoxycycline-rifampim [63-64]. Ciprofloxacin-rifampincan be considered as an alternative regimen for doxy-cycline-streptomycin when patients experience drug toxic-ity [65]. Osteoarticular disease, similar to other localizingmanifestations, carries with it higher risk of relapse com-pared to brucellosis without localizing disease. Therefore,antibiotic therapy should be administered for 3-6 months[61, 64] in order to minimize the risk of relapse.

Osteomyelitis due to Salmonella speciesOsteomyelitis due to Salmonella species is uncommonaccounting for 0.45% of all osteomyelitis cases and is seen

in only 0.8% of all Salmonella infections [66]. Patientswith sickle cell disease, however, have much higher risk of developing this infection. Other risk factors includecomorbidities such as diabetes mellitus, systemic lupuserythematosus, lymphoma, liver disease, previous surgeryor trauma, treatment with steroids, and extremes of age[66]. Patients are usually treated with third-generationcephalosporins or fluoroquinolones. In anecdotal reports,the duration of successful antibiotic therapy ranged from6-12 weeks [66-69].

Osteomyelitis due to mycobacteriaOsteomyelitis can result from infection due to M. tubercu-losis or non-tuberculous mycobacteria (NTBM). Myco-bacterial osteomyelitis should be suspected when bacterialcultures remain negative or when granulomatous inflam-mation is identified in bone tissue [70]. Tuberculosis (TB)of the musculoskeletal system represents 1-5% of all TBcases [6]. Osteomyelitis is attributed to hematogenousseeding of the bone at the time of primary infection andshould be considered in patients with history of treated oruntreated tuberculosis, known exposure to M. tuberculosis,positive tuberculin skin test, and evidence of old or activeTB on chest X-ray. Potts disease or vertebral osteomyeli-tis due to M. tuberculosis represents around 50% of allmusculoskeletal TB cases [71]. The lumbar and lower thoracic spine is more commonly involved than the upperthoracic and cervical spine [72-73]. In 50% of cases, spineMRI shows evidence of paravertebral abscesses and thoseare frequently bilateral. Extraspinal tuberculosis presentswith swelling, pain, and mild erythema. A cold abscessis usually present and spontaneous drainage through asinus tract could occur. The most common manifestationsof spinal osteomyelitis are pain and stiffness. Due to de-layed diagnosis, neurological deficit due to spinal cordcompression is present in 42-76% of patients [74-75].Treatment of tuberculous osteomyelitis is for the most part medical. The regimens of antituberculous medicationsused to treat osteoarticular TB are similar to those used totreat other forms of TB, but treatment should be given fora longer period of time (12-18 months) [76]. Surgical ther-apy might be necessary for abscess drainage, cord decom-pression, and spinal stabilization [74].

Musculoskeletal infections due to NTBM seem to haveincreased in frequency over the past few years, possiblydue to better recognition of this infection [77]. Multiplespecies can be responsible for osteomyelitis and those include M. abscessus, M. chelonae, M. fortuitum, M. mari-num, M. avium-intracellulare, M. kansasii, M. xenopiand M. terrae complex [78-81]. The infection devel-ops in bones, joints, bursae, or tendon sheaths followingdirect inoculation at the time of surgery, trauma, puncturewounds, or injections (intraarticular or extraarticular).Patients present with non-healing wounds associated withchronic drainage and sinus formation. Axial bone orextremities could also be infected through hematogenousseeding. Mycobacterial cultures of tissue specimens arerequired to establish a diagnosis; however, rapidly-grow-


ing mycobacteria grow on routine cultures as well. Theidentification of the infecting species and its susceptibilityprofile are crucial to guide antimicrobial therapy. A combi-nation of two or three antibiotics (depending upon thedegree of severity of the infection) should be used to treatthe infection. The antibiotics should be determined basedon the susceptibility profile. Osteomyelitis due to rapidly-growing mycobacteria is treated for six months whileosteomyelitis due to other mycobacteria is treated for atleast 12 months [82].

Osteomyelitis due to fungiFungal osteomyelitis is more likely to result from dissemi-nated fungal infection than direct inoculation of the bone.Osteomyelitis is commonly seen with disseminated en-demic fungal infections such as blastomycosis, coccidioi-domycosis and sporotrichosis. Up to 25% of patients withdisseminated blastomycosis develop osteomyelitis [83].Candida osteomyelitis can complicate candidemia or resultfrom surgical site infection. Even though uncommon, dis-seminated Aspergillus and Cryptococcus infections in im-munocompromised patients can lead to osteomyelitis.Fungal osteomyelitis due to molds such as Pseudollas-cheria boydii, Scedosporium prolificans, and Fusariumspecies is usually seen with open contaminated fractures.The clinical context is often helpful to identify the fungusresponsible for osteomyelitis. The role of surgical treat-ment is relatively limited and the medical treatmentdepends on the etiology.

Recurrent osteomyelitisPatients presenting with recurrent osteomyelitis shouldundergo comprehensive evaluation in order to determinethe underlying etiology. Recurrence of osteomyelitiscould be attributed to a variety of factors. Those includeincorrect microbiological diagnosis, inability to culturethe true pathogen (e.g. Mycobacteria, Coxiella, Myco-plasma, and Chlamydia species), use of inappropriate orsuboptimal dose of antimicrobial(s), development ofresistance of the infecting organism, and inadequate sur-gical debridement of the infected bone. Occasionally, fail-ure of treatment is attributed solely to the virulence of theinfecting bacteria (e.g. Brucella species, MRSA). Anti-biotic therapy should be held for at least two weeks priorto repeat surgical debridement. Tissue specimens must besubmitted for aerobic and anaerobic bacterial, as well asfungal and mycobacterial cultures. Following surgery,antimicrobial therapy should be guided by culture resultsand in vitro susceptibilities.

CONCLUSION

Osteomyelitis continues to be a serious infection associat-ed with high relapse rate and significant morbidity and lossof function. Preventive strategies should be emphasizedand widely implemented in order to reduce the incidence,especially in high risk patients such as those with diabetesmellitus. When osteomyelitis develops, the patient should

be treated by a multidisciplinary team including an ortho-pedic surgeon, vascular surgeon, plastic surgeon, interven-tional radiologist, endocrinologist and an infectious dis-eases specialist. Such an approach provides the best chancefor cure and restoration of function.

REFERENCES

1. Belmatoug N, Cremieux AC, Bleton R et al. A new modelof experimental prosthetic joint infection due to methi-cillin-resistant Staphylococcus aureus : a microbiologic,histopathologic, and magnetic resonance imaging charac-terization. J Infect Dis 1996 Aug ; 174 (2) : 414-17.

2. Lew DP, Waldvogel FA. Osteomyelitis. N Engl J Med1997 Apr 3 ; 336 (14) : 999-1007.

3. Norden CW. Lessons learned from animal models ofosteomyelitis. Rev Infect Dis 1988 Jan-Feb ; 10 (1) :103-10.

4. Ciampolini J, Harding KG. Pathophysiology of chronicbacterial osteomyelitis. Why do antibiotics fail so often ?Postgrad Med J 2000 Aug ; 76 (898) : 479-83.

5. Muller C, Zielinski CC, Passl R, Eibl MM. Divergentpatterns of leucocyte locomotion in experimental post-traumatic osteomyelitis. Br J Exp Pathol 1984 Jun ; 65(3) : 299-303.

6. Berbari EF, Steckelberg JM, Osmon DR. Osteomyelitis.In : Mandell, Douglas and Bennetts Principles and Prac-tice of Infectious Diseases, 7th edition, Volume 1. Edin-burgh, UK : Churchill Livingstone, 2009.

7. Chihara S, Segreti J. Osteomyelitis. Dis Mon 2010 Jan ;56 (1) : 5-31.

8. Mader JT, Shirtliff M, Calhoun JH. Staging and stagingapplication in osteomyelitis. Clin Infect Dis 1997 Dec ;25 (6) : 1303-9.

9. Grayson ML, Gibbons GW, Balogh K, Levin E,Karchmer AW. Probing to bone in infected pedal ulcers.A clinical sign of underlying osteomyelitis in diabeticpatients. JAMA 1995 Mar 1 ; 273 (9) : 721-3.

10. Perry CR, Pearson RL, Miller GA. Accuracy of culturesof material from swabbing of the superficial aspect of thewound and needle biopsy in the preoperative assessmentof osteomyelitis. J Bone Joint Surg Am 1991 Jun ; 73 (5) : 745-9.

11. Mackowiak PA, Jones SR, Smith JW. Diagnostic value ofsinus-tract cultures in chronic osteomyelitis. JAMA 1978Jun 30 ; 239 (26) : 2772-5.

12. Dinh MT, Abad CL, Safdar N. Diagnostic accuracy of thephysical examination and imaging tests for osteomyelitisunderlying diabetic foot ulcers : meta-analysis. Clin InfectDis 2008 Aug 15 ; 47 (4) : 519-27.

13. Schauwecker DS. The scintigraphic diagnosis of osteo-myelitis. Am J Roentgenol 1992 Jan ; 158 (1) : 9-18.

14. Gotthardt M, Bleeker-Rovers CP, Boerman OC, Oyen WJ.Imaging of inflammation by PET, conventional scintigra-phy, and other imaging techniques. J Nucl Med Dec ; 51(12) : 1937-49.

15. Hadjipavlou AG, Cesani-Vazquez F, Villaneuva-Meyer J,et al. The effectiveness of gallium citrate Ga 67 radionu-clide imaging in vertebral osteomyelitis revisited. Am JOrthop (Belle Mead NJ) 1998 Mar ; 27 (3) : 179-83.

16. Termaat MF, Raijmakers PG, Scholten HJ, Bakker FC,Patka P, Haarman HJ. The accuracy of diagnostic imag-ing for the assessment of chronic osteomyelitis : a sys-


tematic review and meta-analysis. J Bone Joint Surg Am2005 Nov ; 87 (11) : 2464-71.

17. Pineda C, Vargas A, Rodriguez AV. Imaging of osteo-myelitis : current concepts. Infect Dis Clin North Am2006 Dec ; 20 (4) : 789-825.

18. Norden CW, Shinners E, Niederriter K. Clindamycin treat-ment of experimental chronic osteomyelitis due to Staphy-lococcus aureus. J Infect Dis 1986 May ; 153 (5) : 956-9.

19. Carragee EJ, Kim D, van der Vlugt T, Vittum D. Theclinical use of erythrocyte sedimentation rate in pyogenicvertebral osteomyelitis. Spine (Phila Pa 1976) 1997 Sep15 ; 22 (18) : 2089-93.

20. Tice AD, Hoaglund PA, Shoultz DA. Outcomes of osteo-myelitis among patients treated with outpatient parenter-al antimicrobial therapy. Am J Med 2003 Jun 15 ; 114 (9) : 723-8.

21. Razonable RR, Osmon DR, Steckelberg JM. Linezolidtherapy for orthopedic infections. Mayo Clin Proc 2004Sep ; 79 (9) : 1137-44.

22. Rao N, Hamilton CW. Efficacy and safety of linezolid forGram-positive orthopedic infections : a prospective caseseries. Diagn Microbiol Infect Dis 2007 Oct ; 59 (2) :173-9.

23. Rao N, Regalla DM. Uncertain efficacy of daptomycinfor prosthetic joint infections : a prospective case series.Clin Orthop Relat Res 2006 Oct ; 451 : 34-7.

24. Huddleston PM, Steckelberg JM, Hanssen AD, Rouse MS,Bolander ME, Patel R. Ciprofloxacin inhibition of experi-mental fracture healing. J Bone Joint Surg Am 2000 Feb ;82 (2) : 161-73.

25. Norden CW. Experimental chronic staphylococcalosteomyelitis in rabbits : treatment with rifampin aloneand in combination with other antimicrobial agents. RevInfect Dis 1983 Jul-Aug ; 5 (Suppl 3) : S491-S494.

26. Shirtliff ME, Mader JT, Calhoun J. Oral rifampin plusazithromycin or clarithromycin to treat osteomyelitisin rabbits. Clin Orthop Relat Res 1999 Feb (359) : 229-

36.27. Zimmerli W, Widmer AF, Blatter M, Frei R, Ochsner PE.

Role of rifampin for treatment of orthopedic implant-related staphylococcal infections : a randomized con-trolled trial. Foreign-Body Infection (FBI) Study Group.JAMA 1998 May 20 ; 279 (19) : 1537-41.

28. Perlroth J, Kuo M, Tan J, Bayer AS, Miller LG.Adjunctive use of rifampin for the treatment of Staphylo-coccus aureus infections : a systematic review of the lit-erature. Arch Intern Med 2008 Apr 28 ; 168 (8) : 805-19.

29. Kranke P, Bennett M, Roeckl-Wiedmann I, Debus S.Hyperbaric oxygen therapy for chronic wounds. CochraneDatabase Syst Rev 2004 (2) : CD004123.

30. Ostermann PA, Seligson D, Henry SL. Local antibiotictherapy for severe open fractures. A review of 1085 con-secutive cases. J Bone Joint Surg Br 1995 Jan ; 77 (1) :93-7.

31. Sterett WI, Ertl JP, Chapman MW, Moehring HD. Opentibia fractures in the splenectomized trauma patient :results of treatment with locking, intramedullary fixa-tion. J Trauma 1995 Apr ; 38 (4) : 639-41.

32. DeLong WG Jr., Born CT, Wei SY, Petrik ME, Ponzio R,Schwab CW. Aggressive treatment of 119 open fracturewounds. J Trauma 1999 Jun ; 46 (6) : 1049-54.

33. Holcombe SJ, Schneider RK, Bramlage LR, EmbertsonRM. Use of antibiotic-impregnated polymethyl metha-crylate in horses with open or infected fractures or joints :

19 cases (1987-1995). J Am Vet Med Assoc 1997 Oct 1 ;211 (7) : 889-93.

34. Henry SL, Ostermann PA, Seligson D. The prophylacticuse of antibiotic impregnated beads in open fractures. J Trauma 1990 Oct ; 30 (10) : 1231-8.

35. Kindsfater K, Jonassen EA. Osteomyelitis in grade II andIII open tibia fractures with late debridement. J OrthopTrauma 1995 Apr ; 9 (2) : 121-7.

36. Shyam AK, Sancheti PK, Patel SK, Rocha S, Pradhan C,Patil A. Use of antibiotic cement-impregnated intra-medullary nail in treatment of infected non-union of longbones. Indian J Orthop 2009 Oct ; 43 (4) : 396-402.

37. Qiang Z, Jun PZ, Jie XJ, Hang L, Bing LJ, Cai LF. Useof antibiotic cement rod to treat intramedullary infectionafter nailing : preliminary study in 19 patients. ArchOrthop Trauma Surg 2007 Dec ; 127 (10) : 945-51.

38. Wiley AM, Trueta J. The vascular anatomy of the spineand its relationship to pyogenic vertebral osteomyelitis. J Bone Joint Surg Br 1959 Nov ; 41-B : 796-809.

39. Schnoring M, Brock M. [Prophylactic antibiotics in lum-bar disc surgery : analysis of 1,030 procedures]. ZentralblNeurochir 2003 ; 64 (1) : 24-9.

40. Torda AJ, Gottlieb T, Bradbury R. Pyogenic vertebralosteomyelitis : analysis of 20 cases and review. ClinInfect Dis 1995 Feb ; 20 (2) : 320-8.

41. Hopkinson N, Stevenson J, Benjamin S. A case ascer-tainment study of septic discitis : clinical, microbiologi-cal and radiological features. QJM 2001 Sep ; 94 (9) :465-70.

42. Pigrau C, Almirante B, Flores X et al. Spontaneous pyo-genic vertebral osteomyelitis and endocarditis : inci-dence, risk factors, and outcome. Am J Med 2005 Nov ;118 (11) : 1287.

43. Vlahakis NE, Temesgen Z, Berbari EF, Steckelberg JM.Osteoarticular infection complicating enterococcal endo-carditis. Mayo Clin Proc 2003 May ; 78 (5) : 623-8.

44. Nolla JM, Ariza J, Gomez-Vaquero C et al. Spontaneouspyogenic vertebral osteomyelitis in nondrug users.Semin Arthritis Rheum 2002 Feb ; 31 (4) : 271-8.

45. Vazquez M. Osteomyelitis in children. Curr Opin Pediatr2002 Feb ; 14 (1) : 112-15.

46. De Boeck H. Osteomyelitis and septic arthritis in chil-dren. Acta Orthop Belg 2005 Oct ; 71 (5) : 505-15.

47. Ramsey SD, Newton K, Blough D et al. Incidence, out-comes, and cost of foot ulcers in patients with diabetes.Diabetes Care 1999 Mar ; 22 (3) : 382-7.

48. Butalia S, Palda VA, Sargeant RJ, Detsky AS, Mourad O.Does this patient with diabetes have osteomyelitis of thelower extremity ? JAMA 2008 Feb 20 ; 299 (7) : 806-13.

49. Newman LG, Waller J, Palestro CJ et al. Unsuspectedosteomyelitis in diabetic foot ulcers. Diagnosis and mon-itoring by leukocyte scanning with indium in 111 oxy-quinoline. JAMA 1991 Sep 4 ; 266 (9) : 1246-51.

50. Lipsky BA, Berendt AR, Deery HG et al. Diagnosis andtreatment of diabetic foot infections. Clin Infect Dis 2004Oct 1 ; 39 (7) : 885-910.

51. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projectionsof primary and revision hip and knee arthroplasty in theUnited States from 2005 to 2030. J Bone Joint Surg Am2007 Apr ; 89 (4) : 780-5.

52. Del Pozo JL, Patel R. Clinical practice. Infection associ-ated with prosthetic joints. N Engl J Med 2009 Aug 20 ;361 (8) : 787-94.

53. Brause B. Infections with prostheses in bones and joints.


In : Mandell, Douglas and Bennetts Principles and Prac-tice of Infectious Diseases, 7th edition, Volume 1. ChurchillLivingstone, 2009.

54. Cabrita HB, Croci AT, Camargo OP, Lima AL. Prospectivestudy of the treatment of infected hip arthroplasties with or without the use of an antibiotic-loaded cement spacer.Clinics (Sao Paulo) 2007 Apr ; 62 (2) : 99-108.

55. Barberan J, Aguilar L, Carroquino G et al. Conservativetreatment of staphylococcal prosthetic joint infections inelderly patients. Am J Med 2006 Nov ; 119 (11) : 993.e7-10.

56. Marculescu CE, Berbari EF, Hanssen AD et al. Outcomeof prosthetic joint infections treated with debridementand retention of components. Clin Infect Dis 2006 Feb15 ; 42 (4) : 471-8.

57. Kak V, Chandrasekar PH. Bone and joint infections ininjection drug users. Infect Dis Clin North Am 2002 Sep ;16 (3) : 681-95.

58. Chandrasekar PH, Narula AP. Bone and joint infectionsin intravenous drug abusers. Rev Infect Dis 1986 Nov-Dec ; 8 (6) : 904-11.

59. Swisher LA, Roberts JR, Glynn MJ. Needle lickers osteo-myelitis. Am J Emerg Med 1994 May ; 12 (3) : 343-6.

60. Firooznia H, Seliger G, Abrams RM, Valensi V, ShamounJ. Disseminated extrapulmonary tuberculosis in asso-ciation with heroin addiction. Radiology 1973 Nov ; 109 (2) : 291-6.

61. Colmenero JD, Reguera JM, Martos F et al. Compli-cations associated with Brucella melitensis infection : astudy of 530 cases. Medicine (Baltimore) 1996 Jul ; 75(4) : 195-211.

62. Ariza J, Pujol M, Valverde J et al. Brucellar sacroiliitis:findings in 63 episodes and current relevance. Clin InfectDis 1993 Jun ; 16 (6) : 761-5.

63. Ariza J, Gudiol F, Pallares R et al. Treatment of humanbrucellosis with doxycycline plus rifampin or doxycy-cline plus streptomycin. A randomized, double-blindstudy. Ann Intern Med 1992 Jul 1 ; 117 (1) : 25-30.

64. El Miedany YM, El Gaafary M, Baddour M, Ahmed I.Human brucellosis : do we need to revise our therapeuticpolicy ? J Rheumatol 2003 Dec ; 30 (12) : 2666-72.

65. Alp E, Koc RK, Durak AC et al. Doxycycline plus strep-tomycin versus ciprofloxacin plus rifampicin in spinalbrucellosis [ISRCTN31053647]. BMC Infect Dis 2006 ;6 : 72.

66. Arora A, Singh S, Aggarwal A, Aggarwal PK.Salmonella osteomyelitis in an otherwise healthy adultmale successful management with conservative treat-ment : a case report. J Orthop Surg (Hong Kong) 2003Dec ; 11 (2) : 217-20.

67. van Cappelle HG, Veenendaal D, de Vogel PL. Salmo-nella panama osteomyelitis in an otherwise healthypatient. A case report. Clin Orthop Relat Res 1995 Dec(321) : 235-8.

68. Santos EM, Sapico FL. Vertebral osteomyelitis due tosalmonellae : report of two cases and review. Clin InfectDis 1998 Aug ; 27 (2) : 287-95.

69. Agustsson AI, Olafsson K, Thorisdottir AS. Salmonellaosteomyelitis in pregnancy. Acta Obstet Gynecol Scand2009 ; 88 (10) : 1171-3.

70. Ikem IC, Bamgboye EA, Olasinde AA. Spinal tuberculo-sis : a 15-year review at OAUTHC Ile-Ife. Niger Post-grad Med J 2001 Mar ; 8 (1) : 22-5.

71. Pertuiset E, Beaudreuil J, Liote F et al. Spinal tuberculo-sis in adults. A study of 103 cases in a developed coun-try, 1980-1994. Medicine (Baltimore) 1999 Sep ; 78 (5) :309-20.

72. Lifeso RM, Weaver P, Harder EH. Tuberculous spondyli-tis in adults. J Bone Joint Surg Am 1985 Dec ; 67 (9) :1405-13.

73. Weaver P, Lifeso RM. The radiological diagnosis oftuberculosis of the adult spine. Skeletal Radiol 1984 ; 12(3) : 178-86.

74. Nussbaum ES, Rockswold GL, Bergman TA, Erickson DL,Seljeskog EL. Spinal tuberculosis : a diagnostic and man-agement challenge. J Neurosurg 1995 Aug ; 83 (2) : 243-7.

75. Hsu LC, Leong JC. Tuberculosis of the lower cervicalspine (C2 to C7). A report on 40 cases. J Bone Joint SurgBr 1984 Jan ; 66 (1) : 1-5.

76. Treatment of tuberculosis. MMWR Recomm Rep 2003Jun 20 ; 52 (RR-11) : 1-77.

77. Eid AJ, Berbari EF, Sia IG, Wengenack NL, Osmon DR,Razonable RR. Prosthetic joint infection due to rapidlygrowing mycobacteria : report of 8 cases and review of theliterature. Clin Infect Dis 2007 Sep 15 ; 45 (6) : 687-94.

78. Falkinham JO 3rd. Epidemiology of infection by nontu-berculous mycobacteria. Clin Microbiol Rev 1996 Apr ;9 (2) : 177-215.

79. Lahey T. Invasive Mycobacterium marinum infections.Emerg Infect Dis 2003 Nov ; 9 (11) : 1496-8.

80. Chan ED, Kong PM, Fennelly K, Dwyer AP, IsemanMD. Vertebral osteomyelitis due to infection with nontu-berculous Mycobacterium species after blunt trauma tothe back : 3 examples of the principle of locus minorisresistentiae. Clin Infect Dis 2001 May 15 ; 32 (10) :1506-10.

81. Hellinger WC, Smilack JD, Greider JL Jr et al. Localizedsoft-tissue infections with Mycobacterium avium/Myco-bacterium intracellulare complex in immunocompetentpatients : granulomatous tenosynovitis of the hand orwrist. Clin Infect Dis 1995 Jul ; 21 (1) : 65-9.

82. Griffith DE, Aksamit T, Brown-Elliott BA et al. An offi-cial ATS/IDSA statement : diagnosis, treatment, and pre-vention of nontuberculous mycobacterial diseases. Am JRespir Crit Care Med 2007 Feb 15 ; 175 (4) : 367-416.

83. Moore RM, Green NE. Blastomycosis of bone. A reportof six cases. J Bone Joint Surg Am 1982 Sep ; 64 (7) :1097-101.

osteomyelitis review

Documents