Joint Bone Spine 78 (2011) 75–79

Original article

Osteomyelitis in adults: An underrecognized clinical entity in

immunocompetent hosts. A report of six cases
Cécile Gaujoux-Viala a , Valérie Zeller a,b,e,∗ , Philippe Leclerc b,e , Valérie Chicheportiche c,e ,
Patrick Mamoudy b,e , Nicole Desplaces b,d,e , Jean Marc Ziza a,e
a
Service de médecine interne et rhumatologie, groupe hospitalier Diaconesses Croix Saint-Simon, 125, rue d’Avron, 75020 Paris, France
b
Service de chirurgie osseuse et traumatologique, groupe hospitalier Diaconesses Croix Saint-Simon, 125, rue d’Avron, 75020 Paris, France
c
Service de radiologie, groupe hospitalier Diaconesses Croix Saint-Simon, 125, rue d’Avron, 75020 Paris, France
d
Laboratoire de biologie médicale, groupe hospitalier Diaconesses Croix Saint-Simon, 125, rue d’Avron, 75020 Paris, France
e
Centre de référence pour les infections ostéo-articulaires complexes, groupe hospitalier Diaconesses Croix Saint-Simon, 125, rue d’Avron, 75020 Paris, France

a r t i c l e i n f o a b s t r a c t

Article history: Objective: Osteomyelitis is rare in adults and typically occurs in patients with risk factors such as sickle
Accepted 24 February 2010 cell disease or immune deficiency. Cases in immunocompetent adults without sickle cell disease are
Available online 25 August 2010 extremely rare. The objective of this work was to describe the epidemiological, clinical, laboratory,
and radiological features and the management of long-bone osteomyelitis in immunocompetent adults
Keywords: without sickle cell disease.
Osteomyelitis Methods: We conducted a retrospective descriptive study of all immunocompetent adults without sickle
Long bones
cell disease who were admitted to our center between November 2002 and November 2008 for long-bone
Adult
Immunocompetent
osteomyelitis. In all patients, the clinical symptoms started in adulthood, in the absence of a childhood
history of osteomyelitis.
Results: We identified six patients meeting our inclusion criteria over the 6-year study period.
The causative microorganism was methicillin-susceptible Staphylococcus aureus in four patients and
Salmonella in two patients (wild-type S. typhi and S. enterica, respectively). In each patient, there was
a single focus of osteomyelitis and a single causative microorganism. The symptoms developed insidi-
ously and lacked specificity. At presentation, the patients had moderate pain with or without a swelling.
There was no fever initially in five patients, three of whom had major diagnostic delays as a result. Treat-
ment associated antibiotics and surgery in all patients and the initial outcome was consistently favorable
(median follow-up: 15 months; range: 8–72).
Conclusion: Osteomyelitis can occur even in immunocompetent adults. The protracted course and atypical
presentation of osteomyelitis in immunocompetent adults may lead to major diagnostic delays.
© 2010 Société française de rhumatologie. Published by Elsevier Masson SAS. All rights reserved.

1. Introduction ever, immunocompetent adults without sickle cell disease may be

Osteomyelitis is caused by hematogenous dissemination of one
or more microorganisms to the bone. Most cases occur in prepu-
bertal children, usually at the metaphyses of the long lower-limb
bones. In adults, bone infections are generally caused by direct
contamination (after surgery or a compound fracture) or by the
spread of an adjacent infection, and the diagnosis is therefore read-
ily made [1,2]. Osteomyelitis in adults is usually due to reactivation
of a lesion acquired in childhood. Onset after 18 years of age is
extraordinarily rare. Adult-onset osteomyelitis typically occurs in
patients with sickle cell disease or immune deficiencies [3]. How-
∗ Corresponding author. Phone: +33 144 641 694; fax: +33 144 643 337.
E-mail address: vzeller@hopital-dcss.org (V. Zeller).
affected, raising diagnostic and therapeutic challenges [4].
The objective of this work was to describe the epidemiological,
clinical, laboratory, and radiological features and the management
of hematogenous long-bone osteomyelitis in immunocompetent
adults without sickle cell disease, based on a retrospective series of
six patients.
2. Methods
We conducted a descriptive retrospective study at the Bone
Surgery and Rheumatology departments of the Diaconesses Croix
Saint-Simon Hospital in Paris, France. We included patients who
were admitted between November 2002 and November 2008
for osteomyelitis, defined as bone marrow infection (metaphy-
seal involvement with or without diaphyseal involvement) related

1297-319X/$ – see front matter © 2010 Société française de rhumatologie. Published by Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.jbspin.2010.03.001
76 C. Gaujoux-Viala et al. / Joint Bone Spine 78 (2011) 75–79

to hematogenous seeding, that is, with no history of surgery or

compound fracture and with recovery of a microorganism con-
sistent with hematogenous dissemination (Staphylococcus aureus,
streptococci, enterobacteria. . .). The microbiological diagnosis was
established in all patients by at least two intraoperative bone
samples growing the same microorganism. All microbiological
specimens were collected using a standardized procedure for bone
and joint infections [5]. The other inclusion criteria were as fol-
lows: infection confined to the long bones (femur, tibia, fibula,
humerus, radius, or ulna); age 18 years or older at symptom
onset; and no history of osteomyelitis in childhood, no sickle cell
disease, and no immune deficiency (e.g., HIV infection, diabetes,
malignancy, immunosuppressive treatment, hepatic cirrhosis, or
advanced renal failure). All patients underwent the following tests:
HIV serology, fasting blood glucose, hemoglobin electrophoresis,
serum protein electrophoresis, liver function tests, serum elec-
trolytes, and serum urea and creatinine.
The following data were collected for each included patient:
age, sex, and ethnicity; clinical manifestations (fever, pain, edema, Fig. 1. Radiograph of the femur showing medullary osteolysis (thick arrow), thick-
skin lesions, function), whether a portal of entry was identified, the ening of the cortex, and a periosteal reaction (slender arrow).
causative organism, the bone site involved, the time to diagnosis,
laboratory test results (leukocyte count, C-reactive protein [CRP],
and erythrocyte sedimentation rate [ESR]), and findings suggestive
of infection on radiographs (medullary osteolysis and periosteal
reaction) or magnetic resonance imaging (MRI) scans (medullary
abscess with or without soft tissue involvement).

3. Results

Between November 2002 and November 2008, 145 patients

were managed at our center for infection of a long bone. Among
them, 28 (19%) had osteomyelitis, including 16 with a recurrence
of childhood osteomyelitis and 12 with adult-onset osteomyeli-
tis. Among these 12 patients, three had sickle cell disease and
three had immune deficiencies (malignancy, long-term glucocor-
ticoid therapy, and systemic lupus erythematosus, respectively).
Thus, over a 6-year period, only six cases of adult-onset osteomyeli-
tis in immunocompetent patients without sickle cell disease were
seen. Table 1 reports the main features in these six patients. All six
patients had a single site of infection and a single causative organ-
ism. Symptom onset was insidious in three patients. Pain was a
consistent feature, being moderate in four patients and severe in
two patients. Only three patients reported an inflammatory time
pattern of the pain. In four patients, pain onset was followed by
the development of edema. A single patient had a fever initially,
and in two others a fever developed later in the course of the infec-
tion. None of the patients had draining sinus tracts at admission
but one patient had a healed sinus tract. No portal of entry was
identified in any of the patients. Laboratory tests showed moderate
leukocytosis in three patients and systemic inflammation in four
patients. Medullary osteolysis was a consistent radiographic find-
ing, whereas a periosteal reaction was visible in only two patients
(Fig. 1). MRI showed an abscess in all six patients (Fig. 2). In three
patients, the time to diagnosis was extremely long, ranging from
4 months to 10 years. In one patient (#2), the diagnostic delay
was ascribable to a misleading clinical picture with no systemic
or local signs of infection and intermittent pain suggesting a spinal
disorder. This patient sought advice from a variety of healthcare
professionals (general practitioner, rheumatologist, osteopath, chi- Fig. 2. a: magnetic resonance imaging of the thigh, T1-weighted sequence: note the
ropractor, homeopath); received several treatments (nonsteroidal low-signal medullary image on the left and the soft tissue edema (solid arrow); b:
anti-inflammatory drugs, aspirin, paracetamol), with little effect; magnetic resonance imaging of the thigh, T1-weighted sequence after gadolinium
and was finally categorized as having a personality disorder with injection: note on the left the low-signal medullary image with postgadolinium rim
enhancement and the soft tissue edema with postgadolinium enhancement (solid
nonorganic symptoms. Two other patients (#3 and #5) experienced
arrow).
intermittent flares responsible for functional impairment and mild-
to-moderate pain; one of these patients (#3) had no fever at all
Table 1
Main features of the six immunocompetent patients with adult-onset osteomyelitis and no sickle cell disease.

Patient #1 Patient #2 Patient #3 Patient #4 Patient #5 Patient #6

Age (years)/sex 21/male 20/female 36/female 23/male 32/male 34/male

Ethnicity Caucasian Caucasian African North African Caucasian African
Site of involvement Radius Femur Ulna Tibia Femur Femur

C. Gaujoux-Viala et al. / Joint Bone Spine 78 (2011) 75–79

Previous treatment of the infection No No No No Yes Yes
Referral Orthopedic Internet Usual physician Rheumatologist Rheumatologist Emergency room
surgeon
Time to diagnosis 3 weeks 10 months 10 years 2 weeks 4 months 2 weeks
Pain Moderate Moderate Moderate Marked Moderate Moderate
Inflammatory Intermittent then Intermittent Inflammatory pattern Intermittent Knee effusion
pattern permanent with an
inflammatory
pattern
Functional impairment Moderate None None Marked (limp) Moderate Moderate
Edema Marked None Moderate and intermittent Marked Marked No
Healed sinus tract
Sinus tract No No No No No
Fever After 10 days No No No After 94 days Yes
Organism Salmonella enterica Methicillin- Salmonella typhi Methicillin-susceptible S. aureus Methicillin- Methicillin-
subsp. enterica susceptible S. susceptible S. susceptible S.
aureus aureus aureus
Leukocytes/mm3 10 590 10 980 5410 10 300 11 280 7320
CRP (mg/L)/ESR (mm/h) 91/NA 136/105 12/33 6/36 160/100 259/54
Radiographs Medullary Medullary Medullary osteolysis Medullary osteolysis Medullary Medullary
osteolysis osteolysis osteolysis osteolysis
Periosteal reaction Periosteal
reaction
MRI Medullary and soft Medullary abscess Medullary abscess Medullary abscess Medullary and soft Medullary and
tissue abscess tissue abscess soft tissue abscess

NA: not available; CRP: C-reactive protein; ESR: erythrocyte sedimentation rate; MRI: magnetic resonance imaging.

77
78 C. Gaujoux-Viala et al. / Joint Bone Spine 78 (2011) 75–79

and the other (#5) had no fever for the first 3 months. In patient

Gentamicin (1 week)
#3, who sought advice from his general practitioner 10 years after

Apparent recovery
symptom onset, the extremely insidious course of the symptoms

Clindamycin

Clindamycin
and African ethnicity suggested tuberculous osteitis [6]. He under-

Patient #6

Rifampin

Rifampin
went two bone biopsies (one percutaneous needle biopsy and one
surgical biopsy), which failed to confirm the diagnosis of bacte-
rial osteomyelitis, despite prolonged culturing. Histology showed

8
a nonspecific inflammatory granuloma. Examination of multiple
intraoperative specimens recovered Salmonella typhi (Table 1).

Gentamicin (1 week)
All patients received both antibiotics and surgical therapy. The

Apparent recovery
surgical procedure consisted of curettage, excision of infected and
necrotized tissues, and removal of sequestra. In three patients,
screws were used to close and secure the bony flap. Antibiotic ther-

Cefazoline
Patient #5

Pefloxacin
Rifampin

Rifampin
apy was started intraoperatively, immediately after the collection
of microbiological specimens. In five patients, two antibiotics were

8
given for 6 weeks, including at least one via the intravenous route,
after which the treatment was continued orally for 6 additional
weeks (12 weeks of antibiotics in all). The remaining patient (#3)

Gentamicin (1 week)
received intravenous antibiotics for 10 days then oral antibiotics

Apparent recovery
for 10 weeks. The antibiotics were selected based on previously
published guidelines [5].

Cefazoline
Patient #4

Pefloxacin
Rifampin

Rifampin
All patients were reevaluated on an outpatient basis after 6
weeks, 3 months, 6 months, 12 months, and 24 months. At each

24
visit, a physical examination was performed to assess the pres-
ence of pain and the appearance of the scar, and a radiograph
was taken. Blood cell counts and a C-reactive protein assay were

Pefloxacin (10 weeks)

Ceftriaxone (10 days)
performed after 6 weeks and 3 months. Apparent recovery was

Apparent recovery
defined as absence of local signs of inflammation (erythema, sinus
tract, inflammatory pain), absence of a fever, and normal blood
cell counts and C-reactive protein level. Median follow-up was 15

Patient #3
months (range: 8–72). At last follow-up, all patients met our criteria
for apparent recovery (Table 2).

72
4. Discussion
Gentamicin (1 week)

Apparent recovery
The incidence of osteomyelitis seems to be declining. Accord-
Antibiotic therapy and outcomes in the 6 immunocompetent patients with adult-onset osteomyelitis.

ing to a study of 275 pediatric osteomyelitis cases in Scotland [7]

Cefazoline
Patient #2

Pefloxacin
Rifampin

Rifampin
reported in 2001, the incidence decreased from 87 to 42/100 000
person-years over a 25-year period. Cases involving the long bones
18

diminished, whereas cases at other sites remained unchanged. The

prevalence of S. aureus osteomyelitis declined from 55% to 31%. In
contrast, cases due to direct inoculation or spread from adjacent
Gentamicin (1 week)

sites are on the rise [8].

Apparent recovery

We are aware of a single published study that focuses specif-

Pefloxacin per os

ically on adult-onset hematogenous osteomyelitis [9]. In this

Ceftriaxone
Patient #1

Pefloxacin

retrospective study, 39 cases were identified between 1960 and

1985. Of these 39 patients, 19 had a history of infection or a pre-
disposing disease (immune deficiency or sickle cell disease). The
12

insidious course was a distinctive feature compared to childhood

osteomyelitis [9]. In a case-series of 25 inpatients with osteomyeli-
tis in the Congo [10] being 15 years of age or older was an
inclusion criterion but the proportion of patients older than 18
years was not reported. Of these 25 patients, 14 had sickle cell
disease and three had HIV infection. Finally, the authors did not
distinguish childhood-onset and adult-onset cases. Another ret-
Oral antibiotic therapy (6 weeks)

rospective study identified 102 cases of chronic osteomyelitis in

IV antibiotic therapy (6 weeks)

Burkina-Faso [11] but failed to separate the patients into age

groups. Of 454 patients with bone infections in the Washington
area of the US, only 27 (6%) had hematogenous osteomyelitis [12].
Follow-up (months)

Over a 6-year period, 12 patients with adult-onset osteomyelitis

NA: not available.

were managed at our center. All these patients were young adults
(age ≤ 36 years) and half of them had immune deficiencies, in keep-
Outcome

ing with previously published data. The other six patients were
Table 2

healthy adults in whom no portal of entry was found. The clin-

ical course was insidious and the picture sometimes misleading,
 C. Gaujoux-Viala et al. / Joint Bone Spine 78 (2011) 75–79
causing major diagnostic delays in three patients. Leukocytosis was
mild or nonexistent. Systemic inflammation, although consistently
found, was mild in two patients. The radiographs were more help-
ful, as they consistently showed medullary osteolysis. In contrast,
a periosteal reaction was an inconsistent finding. It is worth noting
that 50% to 75% of the bone matrix must be destroyed before the
lesion becomes visible on plain radiographs [13]. MRI contributed
meaningfully to the diagnosis by consistently visualizing medullary
abscesses.
Of our six patients, four had osteomyelitis due to methicillin-
susceptible S. aureus, the leading cause of osteomyelitis [1,2].
More surprising is the recovery of Salmonella in the other two
patients. These two patients were the only cases of Salmonella
infection among the 28 patients with osteomyelitis managed dur-
ing the study period. Data in the literature, in contrast, indicate
that Salmonella infections occur chiefly in immunocompromised
patients [14]. In a case-series study from Spain [15], 13 (93%)
of 14 patients with osteoarticular infections due to Salmonella
had immune deficiencies. Similarly, in a case-series of chronic
osteomyelitis patients in Burkina-Faso [11], all six patients with
Salmonella infection had sickle cell disease.
Genetic susceptibility factors for osteomyelitis were described
recently. Toll-like receptor 4 polymorphisms may predispose to
osteomyelitis caused by Gram-negative bacilli [16]. The NO53 poly-
morphism was associated with an increased risk of osteomyelitis
[17]. Similarly, the Bax gene G(-248)A promoter polymorphism
is associated with prolonged neutrophil survival in patients with
osteomyelitis, a characteristic that might contribute to the genesis
and perpetuation of the infection by preventing neutrophil apop-
tosis and prolonging the release of proteases [18]. These highly
specific genetic variants are still difficult to identify in everyday
clinical practice. Current knowledge of genetic factors does not yet
provide an explanation to the clinical variability of osteomyelitis.
Apart from the retrospective design, the main limitation of our
study is the small number of patients, which precludes defini-
tive conclusions about this rare infection. However, our study is
useful for raising awareness among clinicians of the existence
of osteomyelitis in immunocompetent adults. The clinical man-
ifestations were moderate in five of our six patients, a marked
departure from the acute clinical picture characteristic of childhood
osteomyelitis. In three patients, the atypical presentation resulted
in prolonged diagnostic wanderings that delayed the treatment.
The stronger defense mechanisms in adults than in children may
explain the less prominent symptoms. Thus, it seems that the infec-
tion may be confined to the bone, producing only limited systemic
disturbances. Furthermore, follow-up was too short in our patients
to determine whether recovery was complete. Thus, follow-up was
longer than 2 years in only two of the six patients.
79
Although rare, adult-onset osteomyelitis should be considered
in a young adult with suggestive clinical and radiological find-
ings. MRI is the best investigation for diagnosing osteomyelitis.
The microbiological diagnosis rests on the collection of multi-
ple intraoperative specimens. A bone biopsy is not appropriate
as surgical treatment is mandatory and a negative percutaneous
biopsy does not rule out the diagnosis. Tests should be done
routinely to look for immune dysfunction and sickle cell dis-
ease.
Conflict of interest statement
The authors have no conflict of interest to declare.
References
[1] Lazzarini L, Mader JT, Calhoun JH. Osteomyelitis in long bones. J Bone Joint Surg
Am 2004;86:2305–18.
[2] Lew DP, Waldvogel FA. Osteomyelitis. Lancet 2004;364:369–79.
[3] Haas DW, McAndrew MP. Bacterial osteomyelitis in adults: evolving consider-
ations in diagnosis and treatment. Am J Med 1996;101:550–61.
[4] Seybold U, Talati NJ, Kizilbash Q, et al. Hematogenous osteomyelitis mimicking
osteosarcoma due to community associated methicillin-resistant Staphylococ-
cus aureus. Infection 2007;35:190–3.
[5] Ea K, Zeller V, Lhotellier L, et al. Ostéites chronique de l’adulte. Diagnostic et
prise en charge. Antibiotiques 2007;9:120–9.
[6] Laloum E, Zeller V, Graff W, et al. Salmonella typhi osteitis can mimic tubercu-
losis. A report of three cases. Joint Bone Spine 2005;72:171–4.
[7] Blyth MJ, Kincaid R, Craigen MA, et al. The changing epidemiology of acute
and subacute haematogenous osteomyelitis in children. J Bone Joint Surg Br
2001;83:99–102.
[8] Craigen MA, Watters J, Hackett JS. The changing epidemiology of osteomyelitis
in children. J Bone Joint Surg Br 1992 Jul;74:541–5.
[9] Evrard J. Hematogenous osteomyelitis in the adult. Rev Chir Orthop Reparatrice
Appar Mot 1986;72:31–9.
[10] Moyikoua A, Pena-Pitra B, Ondzoto JM, et al. Ostéomyélites de l’adulte à propos
de 25 cas. Med Afr Noire 1992:39.
[11] Nacoulma SI, Ouedraogo DD, Nacoulma EW, et al. Chronic osteomyelitis at the
Ouagadougou teaching hospital (Burkina Faso). A retrospective study of 102
cases (1996–2000). Bull Soc Pathol Exot 2007;100:264–8.
[12] Tice AD, Hoaglund PA, Shoultz DA. Risk factors and treatment outcomes in
osteomyelitis. J Antimicrob Chemother 2003;51:1261–8.
[13] Butt WP. The radiology of infection. Clin Orthop Relat Res 1973;96:20–30.
[14] Medina F, Fuentes M, Jara LJ, et al. Salmonella pyomyositis in patients
with the human immunodeficiency virus. Br J Rheumatol 1995;34:
568–71.
[15] Ramos JM, Garcia-Corbeira P, Aguado JM, et al. Enferm Infecc Microbiol Clin
1995;13:406–10.
[16] Montes AH, Asensi V, Alvarez V, et al. The Toll-like receptor 4 (Asp299Gly) poly-
morphism is a risk factor for Gram-negative and haematogenous osteomyelitis.
Clin Exp Immunol 2006;143:404–13.
[17] Asensi V, Montes AH, Valle E, et al. The NOS3 (27-bp repeat, intron 4) poly-
morphism is associated with susceptibility to osteomyelitis. Nitric Oxide
2007;16:44–53.
[18] Ocana MG, Valle-Garay E, Montes AH, et al. Bax gene G(-248)A promoter poly-
morphism is associated with increased lifespan of the neutrophils of patients
with osteomyelitis. Genet Med 2007;9:249–55.