34

C OPYRIGHT  2012

BY

T HE J OURNAL

OF

B ONE

AND J OINT

S URGERY, I NCORPORATED

Pathologic Fractures in Children with Acute

Staphylococcus aureus Osteomyelitis
Mohan V. Belthur, MD, Sherri B. Birchansky, MD, Alejandro A. Verdugo, MD, Edward O. Mason Jr., PhD, Kristina G. Hulten, PhD,
Sheldon L. Kaplan, MD, E. OBrian Smith, PhD, William A. Phillips, MD, and Jacob Weinberg, MD
Investigation performed at Texas Childrens Hospital, Baylor College of Medicine, Houston, Texas

Background: Osteomyelitis is a common pediatric musculoskeletal infection. This infection can weaken the normal
bone structure, resulting in the risk of a pathologic fracture. The purpose of this study was to evaluate the risk factors for
pathologic fracture in children with Staphylococcus aureus osteomyelitis.
Methods: Seventeen children who were treated for a pathologic long-bone fracture secondary to Staphylococcus aureus
osteomyelitis between January 2001 and January 2009 at a tertiary-care pediatric hospital were identified. These patients
were compared with a control group consisting of forty-nine children with Staphylococcus aureus osteomyelitis without a
fracture who were matched for age, sex, and methicillin susceptibility. A retrospective review of the clinical records,
magnetic resonance imaging (MRI) studies, and microbiologic findings was performed.
Results: Patients who developed a fracture presented with osteomyelitis at a mean age of 8.8 years (range, two to
seventeen years). Fifteen of the seventeen patients had methicillin-resistant Staphylococcus aureus (MRSA) isolates, and two
had methicillin-susceptible Staphylococcus aureus (MSSA). The mean time from disease onset to fracture was 72.1 days
(range, twenty to 150 days). The duration of hospitalization, number of surgical procedures, duration of antibiotic treatment,
and total number of complications differed significantly between the two groups. MRI studies at the time of admission
demonstrated a significantly greater prevalence of subperiosteal abscess and greater circumferential size of such an abscess
in the patients with a fracture. A sharp zone of abnormally diminished enhancement of the marrow was also more common in
these patients. The USA300-0114 pulsotype was more commonly associated with an elevated likelihood of fracture.
Conclusions: Staphylococcus aureus osteomyelitis is a serious infection that may predispose children to pathologic
fractures. Protected weight-bearing and activity restriction are recommended in children with Staphylococcus aureus
osteomyelitis who have the risk factors demonstrated in this study.
Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.

cute hematogenous osteomyelitis is a frequently treated

infection in the pediatric population1-6. Its presence is
usually suspected on the basis of a suggestive clinical
history and physical examination, elevated inflammatory markers,
and radiographic findings1-6. Magnetic resonance imaging (MRI)
can confirm the diagnosis and determine the extent of bone and
soft-tissue involvement7,8.
Staphylococcus aureus has been reported to be responsible
for 67% to 89% of cases of acute hematogenous osteomyelitis
in the pediatric population1-6. In recent years, an increasing
proportion of community-acquired methicillin-resistant

Staphylococcus aureus (CA-MRSA) isolates has been detected

compared with methicillin-susceptible Staphylococcus aureus
(MSSA). Approximately 60% of the community-acquired
Staphylococcus aureus infections in some regions in the United
States are now resistant to methicillin1,2,4,8-13.
At our institution, most CA-MRSA isolates belong to the
USA300 clone14, which is the most prevalent CA-MRSA clone
in the United States and accounts for >90% of CA-MRSA
isolates in most studies13-16. The clonal designation is based on
the pattern formed by enzymatically cleaved Staphylococcus
aureus during pulsed-field gel electrophoresis. USA300-0114 is

Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any
aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this
work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has
had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this
work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.

J Bone Joint Surg Am. 2012;94:34-42

http://dx.doi.org/10.2106/JBJS.J.01915

35
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

the most common pulsotype pattern. Almost all USA300

Staphylococcus aureus isolates carry the genes that encode the
Panton-Valentine leukocidin (PVL) toxin, which destroys
white blood cells by creating pores in their membranes13-16.
Osteomyelitis caused by Staphylococcus aureus that carries
the genes encoding PVL (pvl1) typically has a more severe
clinical course than osteomyelitis caused by pvl Staphylococcus
aureus isolates7,14-21. Patients with pvl1 Staphylococcus aureus
osteomyelitis have been found to have a greater prevalence of
bacteremia, a longer duration of hospitalization, and a greater
number of surgical procedures17 compared with patients with
pvl isolates. The more virulent pvl1 form of osteomyelitis may
cause greater structural damage to the normal architecture of the
bone, potentially increasing the risk of a pathologic fracture17-20.
Pathologic fractures associated with acute bacterial hematogenous osteomyelitis are rare in all age groups. Most of the
reports are in the adult population, with few cases in children22-30.
We studied children with acute Staphylococcus aureus hematogenous osteomyelitis and pathologic fractures managed at a
tertiary-care pediatric center to identify risk factors for fracture
in this patient population.
Materials and Methods

e retrospectively reviewed our experience with acute osteomyelitis at a

tertiary-care pediatric hospital between January 2001 and January 2009.
Children who had been treated for Staphylococcus aureus osteomyelitis and who
had sustained a pathologic long-bone fracture were identified. This study group
was matched on the basis of age, sex, site of infection, and antibiotic sensitivity
with a control group of children treated for Staphylococcus aureus osteomyelitis
who did not sustain a fracture. Patients with osteomyelitis caused by other
organisms and patients with infections associated with decubitus ulcers or
sickle-cell anemia were excluded. The study was approved by the institutional
review board of our medical school.

Initial Treatment Protocol

Patients who presented to the emergency center with symptoms of fever,
swelling, pain, or a limp were assessed for the presence of a musculoskeletal
infection. A thorough history, physical examination, baseline blood work including inflammatory markers, and radiography were performed in each case.
A clinical diagnosis of musculoskeletal infection was confirmed and staged with
use of advanced imaging, primarily MRI.
During the index procedure to treat the osteomyelitis, a small drill was
used to decompress the bone and obtain a bone sample for culture. The osseous
defect usually involved <20% of the bone diameter, as indicated by postoperative radiographs. Curets, Steinmann pins, or flexible nails were used to assist in
the debridement of the infected bone. Lavage with normal saline solution
containing 50 U/mL of bacitracin was performed in each case. Drainage of any
associated subperiosteal abscess, intraosseous abscess, intramuscular abscess,
and/or septic arthritis was also undertaken during the index procedure. Cast or
splint immobilization was used postoperatively in an affected lower extremity,
and an immobilizer was used if the shoulder was involved.
Empiric antibiotic treatment targeting MRSA and MSSA was begun once
blood and tissue samples had been obtained for culture. Once the causative organism and its antibiotic susceptibility were identified, the patient was transitioned to
the appropriate parenteral antibiotic. Clindamycin-sensitive MRSA was treated with
31,32
clindamycin , clindamycin-resistant MRSA was treated with vancomycin, and
MSSA was treated with nafcillin or cefazolin. Vancomycin trough levels were obtained in critically ill patients at the request of the pediatric infectious disease service.
The treating infectious disease physician determined the duration of the
parenteral and oral antibiotics. Patients were transitioned to oral antibiotics on the

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

basis of clinical, inflammatory, and imaging parameters. Clindamycin-sensitive

31,32
MRSA was treated with oral clindamycin
, clindamycin-resistant MRSA was
treated with linezolid or Bactrim (sulfamethoxazole and trimethoprim), and
MSSA was treated with cephalexin.
A return to the operating room was usually required when the patient
did not improve clinically or experienced an associated complication. When the
patient was ready for mobilization, the treating orthopaedic surgeon decided on
the postoperative weight-bearing protocol on the basis of the fracture location,
organism, and age of the patient.

Data Collection
Multiple parameters available at the time of hospital admission and patient
outcome parameters (including the length of hospitalization, number of surgical procedures, and duration of antibiotic treatment) were reviewed. Imaging
studies and microbiologic results of cultures from the patients blood and from
the operative sites were also reviewed.
The imaging findings for each patient were retrospectively reviewed by a
fellowship-trained musculoskeletal radiologist (S.B.B.) with use of the MRIs
acquired during the initial infection work up, including a combination of
multiplanar STIR (short tau inversion recovery) and/or fat-suppressed
T2-weighted, T1-weighted, and gadolinium-enhanced fat-suppressed T1-weighted
sequences. Specific scan parameters, such as the field of view, matrix size, and
slice thickness, varied depending on the region investigated. The MRI features
reviewed for each patient included the location and extent of the osteomyelitis;
the presence of a subperiosteal abscess and its characteristics, including length,
thickness, and maximum circumferential size relative to the bone (measured on
a selected axial image as a percentage of the bone circumference that was
encircled by the abscess); the presence of an intramuscular abscess; the integrity
of the cortex; the pattern of marrow signal enhancement; the presence of joint
7
effusion; and signs of deep vein thrombosis . Radiographs were also reviewed to
confirm a subsequent fracture.
All isolates underwent testing of their susceptibility to antibiotics including clindamycin, erythromycin, oxacillin, trimethoprim-sulfamethoxazole,
and vancomycin with use of the disk diffusion method, as specified by the
33
Clinical and Laboratory Standards Institute . Isolates were genotyped with use
of pulsed-field gel electrophoresis, and the presence of the PVL gene was de34
termined with use of the polymerase chain reaction (PCR) method .

Statistical Analysis
Comparisons between the two groups involved analysis of clinical, radiographic, and microbiologic parameters. The Mann-Whitney U test was used to
compare continuous variables (e.g., age and temperature), and the Fisher exact
test was used to compare categorical variables (e.g., sex). A p value of <0.05 was
considered significant.

Source of Funding
No external funding was received for the present study. An investigatorinitiated Staphylococcus aureus surveillance study at our institution has been
funded by Pfizer since 2001; this grant allowed for the prospective collection
of the isolates as well as the molecular testing.

Results
Fracture Group (see Appendix)
hree hundred and sixty-four patients with acute Staphylococcus aureus hematogenous osteomyelitis were evaluated
and treated at our tertiary-care pediatric hospital between
January 2001 and January 2009, and seventeen (4.7%) of these
patients were identified as having an associated pathologic
fracture.
The mean time from disease onset to fracture was 72.1
days (range, twenty to 150 days). Pathologic fractures were

36
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

Fig. 1-A

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

Fig. 1-B

Figs. 1-A through 1-D Distal femoral osteomyelitis in an otherwise healthy

three-year-old boy (Patient 2). Fig. 1-A Gadolinium-enhanced fat-suppressed
T1-weighted axial magnetic resonance image demonstrating a subperiosteal
abscess (arrows) associated with osteomyelitis of the distal aspect of the
left femur. The patient subsequently underwent two surgical procedures to
appropriately drain the abscess. Fig. 1-B Anteroposterior and lateral radiographs demonstrating a pathologic fracture of the femur six weeks after the
initial diagnosis of osteomyelitis. No history of associated trauma was reported. Fig. 1-C Anteroposterior and lateral radiographs of the femur demonstrating residual deformity one year after the fracture.

Fig. 1-C

identified in the femur (10), fibula (4), tibia (1), and proximal
aspect of the humerus (2). The fracture involved the metaphysealdiaphyseal region of the long bone in sixteen patients; the
remaining patient (patient 12) had a segmental femoral
fracture, with one of the fractures also involving this region.
Only one child (patient 7) had sustained a traumatic injury at
the time of fracture occurrence; this patient fell in a hole
fourteen months after the initial debridement surgery and
fractured through pathologic bone. Another child (patient 16)
sustained a fracture at the biopsy site without a history of associated trauma.
The mean duration of follow-up in the fracture group
was 22.4 months (range, 6.5 to 41.5 months) from the time of
initial presentation in the emergency department. In fifteen of
the seventeen patients, the fracture was initially treated nonoperatively with use of cast immobilization for the lower limb
or a shoulder immobilizer for the upper limb. Two of the

children who were older (>12 years) and heavier (patients 13

and 15) underwent internal fixation with antegrade trochanteric nails. One of these children (patient 13) also underwent
prophylactic stabilization of the infected contralateral femur,
which was at risk for an impending fracture.
Osseous union was achieved in fifteen patients (88%).
The remaining two patients (12%) failed to achieve union at
the fracture site, requiring surgical intervention consisting of
either bone-grafting with internal fixation (patient 4) or bone
transport with external fixation (patient 11). A substantial
angular deformity of the femur was identified in patient 2 (Figs.
1-A through 1-D). This child subsequently underwent a realignment osteotomy with internal fixation at another institution after the family relocated. In an additional six patients,
the fracture healed with a mild to moderate angular deformity
(bringing the percentage of healed fractures with an angular
deformity to 41%). These six patients are currently being

37
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

TABLE I Clinical Characteristics According to the Presence of a Pathologic Fracture

Fracture Group (N = 17)

Non-Fracture Group (N = 49)

P Value

Body temperature at admission* (C)

39.4 (38.2-41.1)

39.1 (37.0-40.6)

0.238

Erythrocyte sedimentation rate at admission* (mm/hr)

52.8 (11-108)

63.5 (13-119)

0.108

C-reactive protein at admission* (mg/dL)

16.2 (1.8-39.3)

18.2 (2-50.4)

0.976

White blood-cell count at admission* (109/L)

16.1 (5.4-36.8)

14.2 (5.8-34)

0.430

Clinical Parameters

Time to presentation* (d)

Duration of hospitalization* (d)

5.1 (1-21)

3.9 (2-8)

0.286

25.6 (11-60)

12.5 (5-46)

<0.001

No. of patients who needed PICU admission#

6/17 (35%)

Duration of stay in PICU*# (d)

Duration of parenteral antibiotics* (wk)
Duration of oral antibiotics* (wk)

11/49 (22%)

0.324

9.3 (3-20)

5.1 (2-10)

0.388

7.1 (4-12)

5.4 (0-10)

0.005

42.1 (12-76)

11.5 (0-78)

<0.001

Total duration of antibiotics (parenteral and/or oral)* (wk)

49.2 (16-84)

16.9 (4-84)

<0.001

Duration of follow-up* (mo)

22.4 (6.5-41.5)

10.4 (3-34)

<0.001

*Values are given as the mean, with the range in parentheses. Normal range, 0-20 mm/hr. Normal range, <1.0 mg/dL. Normal range, 5-14.5
109 cells/L. #PICU = pediatric intensive-care unit.

followed to allow time for remodeling of the deformity and

may require future realignment surgery. One child (patient 12)
developed physeal arrest and may require surgery for limblength equalization. Two children (patients 9 and 17) with
successful fracture union developed sequestrum secondary
to chronic osteomyelitis; both underwent saucerization and
resection of the sequestrum.
Non-Fracture Group (see Appendix)
The non-fracture group consisted of forty-nine patients with
Staphylococcus aureus osteomyelitis who did not develop a fracture. These patients were matched with the patients in the fracture
group with regard to age, sex, site of infection, and antibiotic
sensitivity. The mean age at the time of presentation was 8.8 years
(range, two to seventeen years) in the fracture group and eight
years (range, 1.2 to 16.3 years) in the non-fracture group. MRSA
was the causative organism in 88% of the patients in each group.
The osteomyelitis was located in the femur (26), the tibia (17), the
fibula (4), or the humerus (2) of the patients in the non-fracture
group. The mean duration of follow-up in the non-fracture group
was 10.4 months (range, three to thirty-four months) from the
time of initial presentation.

Fig. 1-D

Clinical photograph demonstrating deformity and shortening in the

left thigh. No gross motion was detected at the fracture site. A
shoe lift allowed the child to place full weight on the left lower
extremity.

Clinical and Laboratory Findings (Table I)

The patients in the fracture group presented to the emergency
department with signs of osteomyelitis at a mean of 5.1 days
(range, one to twenty-one days) after the onset of symptoms
compared with 3.9 days (range, two to eight days) in the nonfracture group (p = 0.286).
The laboratory parameters and body temperature at the
time of presentation did not differ significantly between the two
groups. The mean duration of hospitalization was 25.6 days
(range, eleven to sixty days) in the fracture group compared with
12.5 days (range, five to forty-six days) in the non-fracture group

38
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

TABLE II Initial Surgical Management of the Infection According to the Presence of a Pathologic Fracture

No. of patients who needed multiple surgical episodes

No. of surgical episodes per patient*
Time between first and last surgical episode in patients
who needed multiple episodes* (d)

Fracture Group
(N = 17)

Non-Fracture Group
(N = 49)

P Value

11/17 (65%)

15/49 (31%)

0.006

2.3 (1-5)
16.1 (3-46)

1.4 (0-5)

0.002

8.1 (2-28)

0.153

No. of surgical procedures per patient*

2.9 (1-6)

1.5 (0-5)

<0.001

No. of complications (medical and surgical) per patient*

5.5 (2-14)

1.3 (0-8)

<0.001

No. of patients who will need future surgery to treat

delayed sequelae

7/17 (41%)

3/49 (6%)

<0.001

*The values are given as the mean, with the range in parentheses.

(p < 0.001). The percentage of patients who required admission

to the pediatric intensive care unit and the mean duration of the
stay in this unit were similar in the two groups.
Patients in the fracture group required a longer duration
of treatment with parenteral antibiotics (p = 0.005) and with
oral antibiotics (p < 0.001) compared with the patients in the
non-fracture group. The mean total duration of antibiotic
treatment (parenteral and/or oral) was 49.2 weeks (range,
sixteen to eighty-four weeks) in the fracture group compared
with 16.9 weeks (range, four to eighty-four weeks) in the nonfracture group (p < 0.001).
Surgical Management (Table II)
All of the patients in the fracture group and all but three of the
patients in the non-fracture group underwent initial surgical
treatment to address collections of purulent material associated
with Staphylococcus aureus osteomyelitis. The surgical procedures involved bone biopsy and drainage of purulent collections such as subperiosteal abscesses, intraosseous abscesses,
intramuscular abscesses, and septic arthritis. Extraskeletal foci
of infection were also addressed by pediatric general surgeons.
Following evacuation of purulent material, drains were inserted between the bone and the periosteum and inside septic
joints. Fifteen patients in the fracture group had drains inserted, and two had the wounds packed open because of the
severity of the infection. Forty-five of the children in the nonfracture group had drains inserted, and one had the wound
packed open.
Some patients underwent more than one procedure during a single surgical episode because they had multiple foci of
infection. A surgical episode was defined as a single anesthetization during which at least one surgical procedure was performed. Eleven (65%) of the seventeen patients in the fracture
group required multiple surgical episodes to treat the infection
compared with fifteen (31%) of forty-nine in the non-fracture
group (p = 0.006). A mean of 2.3 surgical episodes (range, one to
five) were required to control the infection in the fracture group
compared with 1.4 episodes (range, zero to five) in the nonfracture group. A mean of 2.9 surgical procedures (range, one to

six) were required during the initial hospitalization to manage

the infection in the fracture group compared with 1.5 surgical
procedures (range, zero to five) in the non-fracture group (p <
0.001). The mean time between the first and last surgical episodes was 16.1 days (range, three to forty-six days) in the fracture
group compared with 8.1 days (range, two to twenty-eight days)
in the non-fracture group (p = 0.153).
Surgical Complications (see Appendix)
The patients in the fracture group experienced a mean of 5.5
surgical or medical complications (range, two to fourteen)
compared with 1.3 complications (range, zero to eight) in the
non-fracture group (p < 0.001). The long-term sequelae of
osteomyelitis required or will require further surgery in 41%
(seven) of the seventeen patients in the fracture group compared with 6% (three) of the forty-nine patients in the nonfracture group (p < 0.001).
Magnetic Resonance Imaging (Table III)
MRIs were acquired at the time of admission for sixteen of
the seventeen patients in the fracture group. The remaining
patient underwent contrast-enhanced computed tomography (CT) scanning before MRI was readily available at our
institution. The findings of these MRI studies were compared
with those for the forty-nine patients in the non-fracture
group.
The MRI scans demonstrated no significant difference
between the two groups in the extent of osteomyelitis relative to
the length of the bone. A subperiosteal abscess was identified in
94% (fifteen) of sixteen patients in the fracture group compared with 57% (twenty-eight) of the forty-nine patients in the
non-fracture group (p = 0.007). The length and the thickness of
the subperiosteal abscess did not differ significantly between
the two groups. However, the maximum circumferential extent
of the subperiosteal abscess was 50% of the bone circumference in 93% (fourteen) of fifteen patients in the fracture group
compared with 50% (fourteen) of twenty-eight in the nonfracture group (p = 0.006) (Fig. 2-A). An intramuscular abscess
was present in 69% (eleven) of sixteen patients in the fracture

39
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

Fig. 2-A

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

Fig. 2-B

Figs. 2-A and 2-B Gadolinium-enhanced fat-suppressed T1-weighted magnetic resonance images demonstrating osteomyelitis of the distal femoral
metadiaphysis in an eight-year-old boy (Patient 8) who later developed a pathologic femoral facture. Fig. 2-A Axial image demonstrating a subperiosteal
abscess encircling >50% of the bone circumference (arrows). Fig. 2-B Coronal image demonstrating a sharp demarcation (arrows) between poorly
enhancing abnormal marrow distally and relatively normally enhancing marrow proximally. Note the subperiosteal and muscle abscess (asterisk).

group compared with 16% (eight) of forty-nine in the nonfracture group (p < 0.001).
Fifteen patients in the fracture group underwent
gadolinium-enhanced fat-suppressed MRI studies at admis-

sion. A sharp zone of abnormally diminished bone marrow

enhancement (Fig. 2-B) was identified in fourteen (93%) of
these fifteen patients compared with twelve (24%) of forty-nine
patients in the non-fracture group (p < 0.001).

TABLE III Magnetic Resonance Imaging Findings According to the Presence of a Pathologic Fracture
Fracture Group
(N = 17*)

Non-Fracture Group
(N = 49)

P Value

Osteomyelitis relative to the length of the bone (%)

50.6 (25-75)

50.6 (10-75)

0.937

Presence of subperiosteal abscess

15/16 (94%)

28/49 (57%)

0.007

Length of subperiosteal abscess (mm)

92.1 (40-190)

90.9 (20-200)

0.794

Thickness of subperiosteal abscess (mm)

10.8 (5-27)

8.1 (2-20)

0.179

Extent of subperiosteal abscess 50% of the bone

circumference

14/15 (93%)

14/28 (50%)

0.006

Sharp zone of abnormal bone marrow enhancement

14/15 (93%)

12/49 (24%)

<0.001

Presence of intramuscular abscess

11/16 (69%)

8/49 (16%)

<0.001

*Sixteen patients underwent magnetic resonance imaging at the time of admission, and this included gadolinium-enhanced fat-suppressed
imaging in fifteen. The values are given as the mean, with the range in parentheses.

40
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

TABLE IV Microbiologic Findings According to the Presence of a Pathologic Fracture

Fracture Group
(N = 17)

Non-Fracture Group
(N = 49)

P Value

Presence of genes encoding PVL virulence factor*

17/17 (100%)

45/49 (92%)

0.565

Presence of USA300 pulsotype

16/17 (94%)

43/49 (88%)

0.667

Presence of USA3000114 pulsotype

14/17 (82%)

27/49 (55%)

0.034

*PVL = Panton-Valentine leukocidin.

Microbiologic Characteristics (Table IV)

The prevalence of the PVL virulence factor and the prevalence of
the USA300 clone of Staphylococcus aureus did not differ significantly between the two groups. However, the USA300-0114
pulsotype was found in 82% (fourteen) of seventeen patients in
the fracture group compared with 55% (twenty-seven) of fortynine patients in the non-fracture group (p = 0.034).
Discussion
estructive processes can cause defects in the architecture
of bone. These defects reduce the strength of the bone by
causing uneven stresses when the bone is loaded, which can
result in a fracture35. The healthy bone of a child has greater
plasticity than that of an adult, and a greater loss of normal
architecture and mineral content is therefore necessary to lead
to a fracture36. Pathologic fractures occur as a result of defects in
the bone caused by malignancy, underlying bone disease, or
infection35-37. Chronic osteomyelitis is a recognized, although
infrequent, cause of pathologic fractures, which can occur
during the process of bone resorption and formation of new
bone23. Most cases of chronic osteomyelitis reported in the
modern medical literature have occurred in children26. Although Staphylococcus aureus infection is the most common
cause of osteomyelitis, cases of pathologic fracture secondary
to Staphylococcus aureus osteomyelitis have been reported
rarely 19,22,24,27-30 compared with those caused by other
microorganisms 25,26,38,39.
Approximately 60% of acute hematogenous osteomyelitis cases in the United States are culture-positive, with
Staphylococcus aureus being the most commonly isolated organism (in 70% of the culture-positive cases)1-4,8-12. CA-MRSA
has emerged as a major pathogen in many areas of the United
States and around the world, complicating the empiric approach to antibiotic treatment of suspected Staphylococcus
aureus osteomyelitis8-13. Osteomyelitis caused by CA-MRSA
isolates that carry the genes encoding PVL is typically associated with a more aggressive disease process and a poorer patient
outcome than that associated with other Staphylococcus aureus
isolates or other types of organisms7,14-21.
Multiple virulence factors have been linked to the severity of CA-MRSA infection14,17-20. It has been speculated that
the epidemic clone of CA-MRSA (USA300) can lead to a more
aggressive manifestation of disease in part because of its expression of PVL15, which is a cytotoxin that destroys leukocytes

by creating pores in the cell membrane17,19,20. Children with

pvl1 Staphylococcus aureus osteomyelitis have been found to
experience a greater number of febrile days and have a greater
complication rate than children with pvl isolates14,19,20. Local
complications were more frequent and often required repeated
surgical drainage14,17,19,20. Extraosseous complications such as
pyomyositis, subperiosteal abscess, deep vein thrombosis, and
septic arthritis were identified in 60% to 70% of children with
CA-MRSA osteomyelitis14,17,19,20.
Dohin et al. reported that pvl1 Staphylococcus aureus
bone and joint infections were more severe than pvl infections
and require more prolonged treatment, and they described
three pathologic fractures among fourteen patients following
acute hematogenous osteomyelitis caused by pvl1 Staphylococcus aureus19. Gelfand et al. reported two cases of pathologic
fracture in adults following acute CA-MRSA hematogenous
osteomyelitis27.
CA-MRSA musculoskeletal infections require prompt
recognition and treatment, including aggressive surgical
drainage, debridement, and appropriate long-term antibiotics.
All of the patients in one previous series required surgical intervention, and sixteen (59%) of the twenty-seven patients
required multiple surgical debridements40. Nevertheless, there
is a substantial potential for long-term morbidity despite aggressive management40.
To our knowledge, the present study represents the
largest published report on pathologic fractures in children
following Staphylococcus aureus osteomyelitis. Sixteen (94%) of
the seventeen patients with a fracture in our study had a
Staphylococcus aureus isolate with the USA300 pulsotype. All
seventeen isolates (100%) were positive for the genes encoding
the PVL toxin. However, a high percentage of the patients in the
non-fracture group also had Staphylococcus aureus isolates that
were of the USA300 pulsotype (88%) and were pvl1 (92%).
Nevertheless, isolates from fourteen (82%) of seventeen patients in the fracture group were of the USA300-0114 pulsotype
compared with only twenty-seven (55%) of forty-nine patients
in the non-fracture group (p = 0.034). It is unclear why the
USA300-0114 pulsotype was associated with a greater risk of
pathologic fracture. We speculate that one or more virulence
factors other than PVL may also be contributing to the greater
fracture risk in these patients.
The patients in the fracture group had more complex
infections, which may have been the primary risk factor for

41
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

fracture. The patients in the fracture group had a more prolonged hospital stay, a longer duration of parenteral and oral
antibiotic treatment, and more complications, and surgical
treatment of the infection was more difficult. This suggests that
these patients had a more severe infection and inflammatory
response, and that the pathologic fracture was a consequence of
the severity of the infection.
MRI is recommended in the setting of a suspected
musculoskeletal infection associated with abnormal inflammatory markers (erythrocyte sedimentation rate [ESR] > 22
mm/hr and C-reactive protein [CRP] level > 3.6 mg/dL)7-10. In
our study, MRI findings such as the presence of a subperiosteal
abscess or an intramuscular abscess as well as a large relative
circumferential size of a subperiosteal abscess were associated
with a greater likelihood of developing a fracture. A sharp zone
of diminished marrow enhancement on MRI scans was also
identified as a risk factor. We speculate that a combination of
purulent marrow and a subperiosteal abscess with a large circumferential extent may impede both the endosteal and the
periosteal blood supply to the bone, causing necrosis of the
bone and placing it at a greater risk of fracture.
The radiographic appearance of many of the pathologic
fractures was consistent with osteopenia. We speculate that the
osteopenia was potentially related to the osseous destruction
caused by the osteomyelitis and possibly also related to disuse
of the extremity. CT-based structural analysis may ultimately
allow the treating physician to identify additional risk factors
for fracture on the basis of quantification of bone density in
patients with osteomyelitis41,42.
Although this retrospective study was performed at a
single institution and had a small sample size, the small
number of patients was matched for age, sex, site of infection,
and methicillin susceptibility with patients who did not develop a fracture. Nevertheless, the study cohort may not be
entirely representative of the general population. Our results
may also not apply to osteomyelitis caused by other staphylococcal clones or by other organisms.
We recognize that the short duration of follow-up in the
non-fracture group (mean, 10.4 months; minimum, three
months) is a potential limitation. A patient in this group was
discharged at the final office visit with the infectious disease
specialist after the laboratory markers had normalized, the
patient had no pain, and the radiographic appearance of the
affected bone was normal. Administration of antibiotics was
discontinued at that time, and the patient was told to obtain
follow-up as necessary. On the basis on these criteria, the
probability of a pathologic fracture in a patient in this group
after the end of the follow-up period would therefore be low.
Six pediatric orthopaedic surgeons (including J.W. and
W.A.P.) participated in this study. Each surgeon tailored the
treatment protocol to address medical and surgical complications, and the surgical protocol and the postoperative weightbearing and activity restriction protocols could also have differed
between surgeons. The sample size of this study was too small to
determine whether there were surgeon-related factors that affected the fracture risk.

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

In summary, our experience with seventeen children

with a pathologic long-bone fracture secondary to acute
Staphylococcus aureus hematogenous osteomyelitis demonstrated that this group of patients had a greater risk of developing sequelae of osteomyelitis compared with children
who did not develop a fracture and that many would require
further surgery for management of these sequelae. We identified multiple risk factors that will potentially aid in avoiding
these fractures. A prolonged hospital stay and multiple surgical procedures may be associated with increased fracture
risk. Careful evaluation of imaging studies is also required,
specifically including evaluation of the circumferential extent
of a subperiosteal abscess and the presence of an ischemic
transition zone in the marrow. The USA300-0114 pulsotype
appeared to be associated with a greater propensity to cause
pathologic fractures. Physicians caring for children with CAMRSA osteomyelitis must be aware of this potential complication of pediatric hematogenous osteomyelitis.
Staphylococcus aureus osteomyelitis is a particularly difficult infection to treat in children. The risk of pathologic
fractures and their sequelae may be reduced by recommending
protected weight-bearing and activity restriction in pediatric
patients with the risk factors identified in this study. Full
weight-bearing can begin once inflammatory markers have
returned to normal levels, the involved bone appears normal on
radiographs, and the child no longer has pain.
Appendix
Tables describing the characteristics and management of
the pathologic fractures, comparing the demographic
characteristics of the children with and without a pathologic
fracture, and describing the complications in the patients with
a pathologic fracture are available with the online version of
this article as a data supplement at jbjs.org. n

Mohan V. Belthur, MD
Sinai Hospital of Baltimore,
2401 West Belvedere Avenue,
Baltimore, MD 21215
Sherri B. Birchansky, MD
Alejandro A. Verdugo, MD
Edward O. Mason Jr., PhD
Kristina G. Hulten, PhD
Sheldon L. Kaplan, MD
E. OBrian Smith, PhD
William A. Phillips, MD
Pediatric Orthopedics and Scoliosis Surgery (A.A.V. and W.A.P.),
Pediatric Radiology (S.B.B.),
Pediatric Infectious Diseases (E.O.M. Jr., K.G.H., and S.L.K.),
and Childrens Research Nutrition Center (E.O.S.),
Texas Childrens Hospital,
6701 Fannin Street, Houston, TX 77030
Jacob Weinberg, MD
Orthopaedic Surgery, 204 Wyatt Earp Loop, Nolanville, TX 76559

42
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
V O L U M E 94-A N U M B E R 1 J A N UA RY 4, 2 012
d

P AT H O L O G I C F R A C T U R E S I N C H I L D R E N W I T H A C U T E
S TA P H Y L O C O C C U S AU R E U S O S T E O M Y E L I T I S

References
1. Copley LAB. Musculoskeletal infection in children. Curr Opin Orthop. 2005;16:
445-450.
2. Frank G, Mahoney HM, Eppes SC. Musculoskeletal infections in children. Pediatr
Clin North Am. 2005;52:1083-106, ix.
3. Krogstad P. Osteomyelitis. In: Feigin RD, Cherry JD, Demmler GJ, Kaplan SL,
editors. Feigin and Cherrys textbook of pediatric infectious diseases. 6th ed. Philadelphia: Saunders Elsevier; 2009. p 725-42.
4. McCarthy JJ, Dormans JP, Kozin SC, Pizzutillo P. Musculoskeletal infections in
children: basic treatment principles and recent advancements. J Bone Joint Surg
Am. 2004;86:850-63.
5. Nade S. Acute haematogenous osteomyelitis in infancy and childhood. J Bone
Joint Surg Br. 1983;65:109-19.
6. Waldvogel FA, Medoff G, Swartz MN. Osteomyelitis: a review of clinical features,
therapeutic considerations and unusual aspects. N Engl J Med. 1970;282:198-206.
7. Browne LP, Mason EO, Kaplan SL, Cassady CI, Krishnamurthy R, Guillerman RP.
Optimal imaging strategy for community-acquired Staphylococcus aureus musculoskeletal infections in children. Pediatr Radiol. 2008;38:841-7.
8. Copley LA. Pediatric musculoskeletal infection: trends and antibiotic recommendations. J Am Acad Orthop Surg. 2009;17:618-26.
9. Arnold SR, Elias D, Buckingham SC, Thomas ED, Novais E, Arkader A, Howard C.
Changing patterns of acute hematogenous osteomyelitis and septic arthritis:
emergence of community-associated methicillin-resistant Staphylococcus aureus.
J Pediatr Orthop. 2006;6:703-8.
10. Gafur OA, Copley LA, Hollmig ST, Browne RH, Thornton LA, Crawford SE. The
impact of the current epidemiology of pediatric musculoskeletal infection on evaluation and treatment guidelines. J Pediatr Orthop. 2008;28:777-85.
11. Gwynne-Jones DP, Stott NS. Community-acquired methicillin-resistant Staphylococcus aureus: a cause of musculoskeletal sepsis in children. J Pediatr Orthop.
1999;19:413-6.
12. Saavedra-Lozano J, Mejas A, Ahmad N, Peromingo E, Ardura MI, Guillen S, Syed
A, Cavuoti D, Ramilo O. Changing trends in acute osteomyelitis in children: impact of
methicillin-resistant Staphylococcus aureus infections. J Pediatr Orthop.
2008;28:569-75.
13. Zetola N, Francis JS, Nuermberger EL, Bishai WR. Community-acquired
meticillin-resistant Staphylococcus aureus: an emerging threat. Lancet Infect Dis.
2005;5:275-86.
14. Martnez-Aguilar G, Avalos-Mishaan A, Hulten K, Hammerman W, Mason EO Jr,
Kaplan SL. Community-acquired, methicillin-resistant and methicillin-susceptible
Staphylococcus aureus musculoskeletal infections in children. Pediatr Infec Dis J.
2004;23:701-6.
15. McCaskill ML, Mason EO Jr, Kaplan SL, Hammerman W, Lamberth LB, Hulten
KG. Increase of the USA300 clone among community-acquired methicillinsusceptible Staphylococcus aureus causing invasive infections. Pediatr Infec
Dis J. 2007;26:1122-7.
16. Sattler CA, Mason EO Jr, Kaplan SL. Prospective comparison of risk factors and
demographic and clinical characteristics of community-acquired, methicillinresistant versus methicillin-susceptible Staphylococcus aureus infection in children.
Pediatr Infect Dis J. 2002;21:910-7.
17. Bocchini CE, Hulten KG, Mason EO Jr, Gonzalez BE, Hammerman WA, Kaplan
SL. Panton-Valentine leukocidin genes are associated with enhanced inflammatory
response and local disease in acute hematogenous Staphylococcus aureus osteomyelitis in children. Pediatrics. 2006;117:433-40.
18. Cremieux AC, Dumitrescu O, Lina G, Vallee C, C
ote JF, Muffat-Joly M, Lilin T,
Etienne J, Vandenesch F, Saleh-Mghir A. Panton-valentine leukocidin enhances the
severity of community-associated methicillin-resistant Staphylococcus aureus rabbit
osteomyelitis. PLoS ONE. 2009;4:e7204.
19. Dohin B, Gillet Y, Kohler R, Lina G, Vandenesch F, Vanhems P, Floret D, Etienne
J. Pediatric bone and joint infections caused by Panton-Valentine leukocidin-positive
Staphylococcus aureus. Pediatr Infect Dis J. 2007;26:1042-8.
20. Gillet Y, Dohin B, Dumitrescu O, Lina G, Vandenesch F, Etienne J, Floret D.
[Osteoarticular infections with staphylococcus aureus secreting Panton-Valentine
leucocidin]. Arch Pediatr. 2007;14 Suppl 2:S102-7French.

21. Hawkshead JJ 3rd, Patel NB, Steele RW, Heinrich SD. Comparative severity of
pediatric osteomyelitis attributable to methicillin-resistant versus methicillinsensitive Staphylococcus aureus. J Pediatr Orthop. 2009;29:85-90.
22. Barczynski A, Gazdzik TS, Pasierbek M. Atypical course of hematogenous osteomyelitis of the hip: a case report and clinical considerations. Orthop Traumatol
Rehabil. 2003;5:92-9.
23. Daoud A, Saighi-Bouaouina A. Treatment of sequestra, pseudarthroses, and
defects in the long bones of children who have chronic hematogenous osteomyelitis.
J Bone Joint Surg Am. 1989;71:1448-68.
24. Devnani AS. Management of pathological fracture neck of the femur following
recent osteomyelitis in a child. Singapore Med J. 2002;43:205-7.
25. Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 26-1995. A previously well, 29 year-old woman with a pathologic
fracture of the femur. N Engl J Med. 1995;333:507-11.
26. Ebong WW. Pathological fracture complicating long bone osteomyelitis in patients with sickle cell disease. J Pediatr Orthop. 1986;6:177-81.
27. Gelfand MS, Cleveland KO, Heck RK, Goswami R. Pathological fracture in acute
osteomyelitis of long bones secondary to community-acquired methicillin-resistant
Staphylococcus aureus: two cases and review of the literature. Am J Med Sci.
2006;322:357-60.
28. Pick RY, Segal D. Case report. Pathological subtrochanteric fracture in an infant:
an unusual disease complex and a successful result. J Pediatr Orthop. 1981;1:20913.
29. Taylor MN, Chaudhuri R, Davis J, Novelli V, Jaswon MS. Childhood osteomyelitis
presenting as a pathological fracture. Clin Radiol. 2008;63:348-51.
30. Walmsley BH. Pathological fracture in acute osteomyelitis in an adult. J R Soc
Med. 1983;76:77-8.
31. Herigon JC, Hersh AL, Gerber JS, Zaoutis TE, Newland JG. Antibiotic management of Staphylococcus aureus infections in US childrens hospitals, 1999-2008.
Pediatrics. 2010;125:e1294-300.
32. Liu C, Bayer A, Cosgrove SE, Daum RS, Fridkin SK, Gorwitz RJ, Kaplan SL,
Karchmer AW, Levine DP, Murray BE, J Rybak M, Talan DA, Chambers HF. Clinical
practice guidelines by the infectious diseases society of america for the treatment of
methicillin-resistant Staphylococcus aureus infections in adults and children: executive summary. Clin Infect Dis. 2011;52:285-92.
33. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing: 19th informational supplement M100-S19. Clinical
and Laboratory Standards Institute: Wayne, PA; 2009.
34. Hulten KG, Kaplan SL, Gonzalez BE, Hammerman WA, Lamberth LB, Versalovic
J, Mason EO Jr. Three-year surveillance of community onset health care-associated
staphylococcus aureus infections in children. Pediatr Infect Dis J. 2006;25:349-53.
35. Pugh J, Sherry HS, Futterman B, Frankel VH. Biomechanics of pathologic fractures. Clin Orthop Relat Res. 1982;169:109-14.
36. Ortiz EJ, Isler MH, Navia JE, Canosa R. Pathologic fractures in children. Clin
Orthop Relat Res. 2005;432:116-26.
37. Damron T, Morgan H, Prakash D, Grant W, Aronowitz J, Heiner J. Metastatic
disease of long bones. Clin Orthop Relat Res. 2003;415:201-7.
38. Quraishi NA, Davidson RN, Steele N, Grand F. Histoplasmosis as the cause of a
pathological fracture. J Bone Joint Surg Br. 2003;85:732-3.
39. Seddon DJ, Thanabalasingham T, Weinberg J. Spontaneous fracture of the ulna
complicating tuberculous osteomyelitis. Postgrad Med J. 1989;65:939-40.
40. Vander Have KL, Karmazyn B, Verma M, Caird MS, Hensinger RN, Farley FA,
Lubicky JP. Community-associated methicillin-resistant Staphylococcus aureus in
acute musculoskeletal infection in children: a game changer. J Pediatr Orthop.
2009;29:927-31.
41. Leong NL, Anderson ME, Gebhardt MC, Snyder BD. Computed tomographybased structural analysis for predicting fracture risk in children with benign skeletal
neoplasms: comparison of specificity with that of plain radiographs. J Bone Joint
Surg Am. 2010;92:1827-33.
42. Snyder BD, Hauser-Kara DA, Hipp JA, Zurakowski D, Hecht AC, Gebhardt MC.
Predicting fracture through benign skeletal lesions with quantitative computed tomography. J Bone Joint Surg Am. 2006;88:55-70.