Abstract
Background: The differentiation of bone infarction
from acute osteomyelitis in patients with sickle-cell disease is
challenging, as the clinical presentations of the two conditions are
similar and imaging and laboratory studies are of limited value.
Methods: A combination of radionuclide bone-marrow
and bone scans was performed sequentially within a twenty-four-hour
period (with one exception) to aid in the differentiation between
bone infarction and osteomyelitis in seventy-nine consecutive
episodes of acute bone pain in children with sickle-cell disease.
Results: Seventy cases of bone infarction were diagnosed
on the basis of decreased uptake on the bone-marrow scan and abnormal
uptake on the bone scan at the site of pain. Antibiotic administration
was discontinued in sixty-six of the seventy cases after the imaging
results were obtained, and the bone pain resolved. In four of the
seventy-nine cases, there was normal uptake on the bone-marrow scan
and abnormal uptake on the bone scan at the site of pain, findings
that were suggestive of acute osteomyelitis. In three of these cases,
osteomyelitis was proven by culture, and the symptoms in all four
resolved with antibiotic treatment. In five of the seventy-nine cases,
the bone-marrow and bone scans were normal and thought to indicate
neither osteomyelitis nor bone infarction; in all of these cases,
the symptoms resolved without the use of antibiotics.
Conclusions: These findings suggest that osteomyelitis
can be differentiated from bone infarction in children with sickle-cell anemia
and acute bone pain by a combination of sequential bone-marrow and
bone scintigraphy.
The differentiation of bone infarction from acute osteomyelitis
in patients with sickle-cell disease is challenging because the
clinical presentations of the two conditions are similar. In both
conditions, a child may have a painful, swollen, and tender limb
that has a limited range of motion. Fevers are common in both conditions,
and interpretation of the erythrocyte sedimentation rate is difficult.
Conventional radiographs are of little value in differentiating
bone infarction from acute osteomyelitis at the initial presentation,
as they may appear normal or show periosteal new bone along the
diaphysis in either condition. The use of a combination of radionuclide
bone-marrow scanning and bone scanning to differentiate bone infarction
from acute osteomyelitis has been recommended1,2.
Proponents of scintigraphic methods have reported that bone infarction
leads to decreased uptake on a bone-marrow scan, whereas acute osteomyelitis
results in normal uptake on a bone-marrow scan; the results on a
bone scan are variable for both infarction and osteomyelitis1,2. Keeley and Buchanan3,4 found that bone-marrow scanning
and bone scanning were not helpful in the differentiation of bone
infarction from acute osteomyelitis.
Most reports outside of Africa have indicated that acute osteomyelitis
is much less common than bone infarction in children with sickle-cell
disease presenting with acute musculoskeletal pain1,2,4,5. Treatment of bone infarction
consists simply of hydration and pain control, with symptoms usually
diminishing markedly in three to five days4,5.
Treatment of acute osteomyelitis requires weeks of antibiotic therapy.
Early differentiation of bone infarction from acute osteomyelitis
could minimize hospitalization and obviate unnecessary treatment
in the majority of children with sickle-cell disease who have bone
pain. Our purpose was to examine the usefulness of combined bone-marrow
scanning and bone scanning in the differentiation of these two conditions
in children with sickle-cell disease at one institution.
Our retrospective study included all children (nine months to
nineteen years of age) seen at our institution between 1988 and
1998 with known sickle-cell hemoglobinopathy and suspected of having
either bone infarction or acute osteomyelitis. Their symptoms included localized
swelling, tenderness, erythema, pain, and fever. A total of seventy-nine
studies were performed on forty-five children; twenty patients had
multiple scans because they had multiple hospital admissions for
bone pain.
The bone-marrow scan and bone scan were both performed within
a twenty-four-hour period in seventy-eight cases; in one case, the
scans were made four days apart. The average time between the onset
of symptoms and the completion of both scans was 8.24 days (range, one
to fifty-six days).
The bone-marrow scan was performed after the intravenous administration
of 0.280 mCi/kg of technetium-99m sulfur colloid;
the minimum dose was 2.0 mCi, and the maximum dose was 18.0 mCi.
The bone scan was subsequently performed after the intravenous administration of
0.280 mCi/kg of technetium-99m methylene diphosphonate;
the minimum dose was 2.0 mCi, and the maximum dose was 18.0 mCi.
The bone scan was performed as a triple-phase examination. Phase
one consisted of dynamic scintigraphy to demonstrate blood-flow
patterns; phase two, tissue-phase scintigraphy to reveal soft-tissue
hyperemia associated with inflammation; and phase three, delayed static
scintigraphy to reveal sites of abnormal tracer uptake in the bone
itself.
A camera with a large field of view and high-resolution collimators
was used to make 500,000-count images; 150,000-count pinhole-collimator
images were obtained when needed. For both studies, images were
obtained on film and on a computer to achieve optimum image intensity.
If the bone-marrow scan showed decreased uptake, thought to be
indicative of decreased blood flow in the bone marrow, and the bone
scan showed abnormal uptake at the site of acute pain, a diagnosis
of bone infarction was made. If the bone-marrow scan showed normal
uptake and the bone scan showed abnormal uptake, a diagnosis of
acute osteomyelitis was made. If neither the bone-marrow scan nor
the bone scan showed abnormal uptake, neither diagnosis was made.
None of the symptomatic sites had decreased uptake on the bone-marrow
scan and normal uptake on the bone scan. However, this combination
of findings was noted in asymptomatic locations and was thought
to be consistent with the presence of old infarctions.
Seventy episodes of pain were diagnosed as being caused by acute
bone infarction on the basis of decreased radionuclide uptake on
the bone-marrow scan and abnormal uptake on the bone scan at the
site of pain. On the basis of these results, administration of antibiotics
that had been started empirically was stopped by the treating physician
in sixty-six of these seventy cases. In the remaining four
cases, antibiotic treatment was continued at the discretion of the
treating physician. Osteomyelitis did not develop in any of the seventy
cases, and all episodes of bone pain resolved.
In four of the seventy cases with nuclear imaging findings consistent
with bone infarction, blood cultures were positive. Antibiotic administration
was stopped in two of these patients after the nuclear imaging results
were determined, and the symptoms completely resolved with no sign
of infection. In one of these children, alpha-hemolytic group-D
Streptococcus had grown on culture of a blood specimen taken seven
days before imaging and coagulase-negative Staphylococcus had grown
on culture of another blood specimen taken four days prior to imaging.
Coagulase-negative Staphylococcus had grown on culture of a single
blood specimen taken from the second child.
The other two patients who had positive blood cultures and imaging
results consistent with bone infarction were believed to have concomitant
infection. Staphylococcus aureus grew on culture
of a blood specimen from one patient; however, these organisms were
believed to have come from a dental abscess, and clinically the
child was not thought to have osteomyelitis. Candida grew on culture of
a blood specimen from the second patient, but these organisms were
thought to be from an infection around an intravenous line. Again,
the patient was not thought to have osteomyelitis on the basis of
the clinical findings as the hospital course progressed. Both children
continued to be treated with antibiotics.
Normal uptake on the bone-marrow scan and increased uptake on
the bone scan at the site of pain suggested acute osteomyelitis
in four of the seventy-nine cases. In three cases, Staphylococcus
aureus as well as other bacteria (coagulase-negative Staphylococcus
in one, Pseudomonas aeruginosa in one, and both
Bacteroides and coagulase-negative Staphylococcus in one) grew on
culture of blood specimens. No organisms grew on culture of blood
specimens from the fourth patient, and a biopsy was not performed. Thus,
three patients had culture-proven osteomyelitis and one had osteomyelitis
diagnosed on clinical grounds only. All four patients were treated
with at least six weeks of intravenous antibiotics, with resolution
of the osteomyelitis.
Five patients had normal bone-marrow scans and normal bone scans
and were thought to have neither osteomyelitis nor bone infarction.
All of these patients’ symptoms resolved without the use
of antibiotics, and none were diagnosed with osteomyelitis on clinical grounds.
The mean erythrocyte sedimentation rate was 40 mm/hr for
the children with infarction and 57 mm/hr for those with
osteomyelitis. The mean temperature was 37.8°C for the children
diagnosed with infarction and 37.2°C for those with osteomyelitis.
In fifty-three of the seventy-nine cases in the
study, the children were taking prophylactic penicillin VK and continued
taking it. Of the four patients diagnosed with osteomyelitis, two
were taking penicillin.
It is often difficult to distinguish clinically between bone infarction
and osteomyelitis in patients with sickle-cell disease who are experiencing
a painful crisis. Although a positive blood culture is a strong
indication of osteomyelitis, bacteria are isolated in the cultures
of only about 50% of patients with an acute untreated infection6. Other clinical indications do not
reliably differentiate between the two conditions. Therefore, patients
suspected of having osteomyelitis are often treated empirically
with analgesic medication, intravenous antibiotics, and hydration,
even though osteomyelitis is actually relatively uncommon1,4.
In our study, bone infarction during a vasoocclusive crisis led
to reduced activity of the radionuclide on bone-marrow scans and
corresponding abnormal activity on bone scans. In contrast, acute
osteomyelitis resulted in normal activity on bone-marrow scans and
abnormal activity on bone scans. These findings indicate that the combination
of sequential bone-marrow and bone scans within a twenty-four-hour
period is a useful tool for early differentiation between osteomyelitis
and bone infarction. It should be remembered that acute osteomyelitis
may present with increased or decreased uptake on a bone scan, although,
in this series, all four patients with acute osteomyelitis had increased
uptake. Bone-marrow scanning targets the reticuloendothelial system
of bone marrow and resident white blood cells in the marrow. In contrast,
bone scanning reflects reparative osteoblastic response. There may
be an increase in activity on the bone-marrow scans of sites of
acute osteomyelitis, but the increase is subtle and is below our
threshold of reliable detection. Bone infarction, on the other hand,
is readily identifiable as a "cold" area of decreased uptake.
Plain radiographs are of limited value in differentiating bone
infarction from acute osteomyelitis, and other imaging methods have
also met with limited success7.
In a recent study from Saudi Arabia, ultrasonography correctly identified
seventeen patients with acute osteomyelitis in a group of fifty-three
patients with sickle-cell disease and a painful crisis8. The authors reported no false-positive
or false-negative results, and the diagnosis of acute osteomyelitis
was confirmed by a positive culture for all seventeen patients.
The authors stated that they "do not hesitate to state
categorically that ultrasonography can differentiate acute osteomyelitis
from vasoocclusive crisis." Other studies demonstrated
more modest success with the use of ultrasound. Howard et al.9, in a study of fifty-nine
children, reported that ultrasonography provided a correct diagnosis
for twenty-six of twenty-nine children with acute
osteomyelitis but gave a false-negative result for the other three
patients.
Magnetic resonance imaging has been quite useful for the diagnosis
of acute osteomyelitis. In a series of thirty-five patients
without sickle-cell disease, magnetic resonance imaging was shown
to be 92% sensitive and 96% specific for the diagnosis
of acute osteomyelitis10. One
recent article on nine patients with sickle-cell disease concluded
that the pattern of contrast enhancement on magnetic resonance imaging
may allow accurate distinction between acute infarction and osteomyelitis11. More evidence is needed with regard
to the accuracy of magnetic resonance imaging in the differentiation between
infarction and osteomyelitis. In addition, radionuclide imaging
may prove to be more cost-effective than magnetic resonance
imaging, especially in children with multiple sites of abnormality.
Furthermore, radionuclide imaging permits an assessment of the entire appendicular
and axial skeleton, allowing for identification of multiple sites
of involvement. Finally, radionuclide imaging may be more accurate
than magnetic resonance imaging in children because of the high
signal intensity of hematopoietic bone, in contrast with the increased
fat content of adult bone marrow.
It is essential to consider the clinical presentation when interpreting
nuclear imaging studies. In our study, 217 asymptomatic sites were
noted to have decreased uptake on bone-marrow scans. We speculated
that these sites may represent old infarctions. Sometimes an infarction can
be so extensive it entirely destroys an area of bone marrow so that
repopulation of the marrow is impossible. This seems to be more
common in older children, who have less capacity for recovery of
marrow activity at the site of the infarction.
It is important to remember that there is no gold standard for
the diagnosis of acute osteomyelitis; in many series, even cultures
of biopsy specimens have revealed negative findings at up to 40% of
sites6. This lack of a gold standard
makes the interpretation of any diagnostic test difficult. Given
this limitation, we believe that it is notable that, in our series,
sixty-six of sixty-six pain crises, including
two in children with positive blood cultures, that were diagnosed
as being caused by bone infarction on bone-marrow and bone scans resolved
clinically after antibiotic administration was stopped. There were
no false-negative results in this series. In addition, of four cases
of acute osteomyelitis diagnosed on the basis of nuclear imaging,
three ultimately were proved by the results of cultures and one was
diagnosed clinically.
Kim HC, Alavi A, Russell MO,Schwartz E. Differentiation of bone and bone marrow infarcts from osteomyelitis
in sickle cell disorders. Clin Nucl Med,1989;14: 249-54. 14249
1989
[PubMed][CrossRef]
Rao S, Solomon N, Miller S,Dunn E. Scintigraphic differentiation of bone infarction from osteomyelitis
in children with sickle cell disease. J Pediatr,1985;107: 685-8. 107685
1985
[PubMed][CrossRef]
Buchanan GR. Differentiation of bone infarct from infection in a child in
sickle cell disease. Pediatr Infect Dis J,1996;15: 725. 15725
1996
[PubMed]
Keeley K,Buchanan GR. Acute infarction of long bones in children with sickle cell
anemia. J Pediatr,1982;101: 170-5. 101170
1982
[PubMed][CrossRef]
Dalton GP, Drummond DS, Davidson RS,Robertson WW Jr. Bone infarction versus infection in sickle cell disease
in children. J Pediatr Orthop,1996;16: 540-4.. 16540
1996
[PubMed][CrossRef]
Waldvogel FA,Papageorgiou PS. Osteomyelitis: the past decade. N Engl J Med,1980;303: 360-70 . 303360
1980
[PubMed][CrossRef]
Rifai A,Nyman R. Scintigraphy and ultrasonography in differentiating osteomyelitis
from bone infarction in sickle cell disease. Acta Radiol,1997;38: 139-43. 38139
1997
[PubMed]
Sadat-Ali M, al-Umran K, al-Habdan I,al-Mulhim F. Ultrasonography: can it differentiate between vasoocclusive
crisis and acute osteomyelitis in sickle cell disease?. J Pediatr Orthop,1998;18: 552-4. 18552
1998
[PubMed][CrossRef]
Howard CB, Einhorn M, Dagan R,Nyska M. Ultrasound in diagnosis and management of acute haematogenous
osteomyelitis in children. J Bone Joint Surg Br,1993;75: 79-82.. 7579
1993
[PubMed]
Unger E, Moldofsky P, Gatenby R, Hartz W,Broder G. Diagnosis of osteomyelitis by MR Imaging. AJR Am J Roentgenol,1988;150: 605-10.. 150605
1988
[PubMed]
Umans H, Haramati N,Flusser G. The diagnostic role of gadolinium enhanced MRI in distinguishing
between acute medullary bone infarct and osteomyelitis. Magn Reson Imaging,2000;18: 255-62. 18255
2000
[PubMed][CrossRef]