INTRODUCTION —
Radiographic imaging is useful for diagnosing osteomyelitis, delineating the extent of disease, and planning therapy. Imaging findings cannot definitively diagnose osteomyelitis and must be interpreted in clinical context [1].
The benefits and limitations of plain radiographs, magnetic resonance imaging (MRI), computed tomography (CT), positron emission tomography (PET)-CT, other nuclear modalities, and ultrasonography for the diagnosis of osteomyelitis will be reviewed here.
An integrated diagnostic approach to evaluation of adults with suspected osteomyelitis is presented in detail separately. (See "Osteomyelitis in the absence of hardware: Approach to diagnosis in adults", section on 'Diagnosis'.)
PATHOPHYSIOLOGY —
Osteomyelitis can occur as a result of hematogenous seeding, contiguous spread of infection to bone from adjacent soft tissues and joints, or direct inoculation of infection into the bone as a result of trauma or surgery [2,3]. (See "Osteomyelitis in the absence of hardware: Approach to diagnosis in adults", section on 'Categories of osteomyelitis'.)
In the setting of osteomyelitis due to hematogenous spread or direct inoculation, inflammatory exudate in the marrow causes elevated medullary pressure, which compresses vascular channels, leading to ischemia and bone necrosis. If the areas of necrotic bone separate from the remaining viable bone, sequestra are formed. Surviving bone and periosteum may ultimately produce a sheath of bone surrounding the area of necrosis, which is referred to as an involucrum. Both the sequestra and involucrum may be apparent radiographically (image 1 and image 2) [2-4].
In contrast, contiguous osteomyelitis from a soft tissue source leads first to inflammation involving cortical bone; osteoclastic activity facilitates resorption of the infected bone, which may be radiographically visible as cortical erosions. Sequestra may form when periosteal vascular supply becomes impaired.
ROLE OF IMAGING —
Radiologic imaging is an integral component of the evaluation for suspected osteomyelitis. The optimal radiographic modality depends upon specific clinical circumstances and should be tailored accordingly.
Although imaging cannot definitively confirm the presence of osteomyelitis, imaging results consistent with osteomyelitis markedly increase the probability of osteomyelitis in patients with compatible clinical findings.
GENERAL APPROACH —
Although there is no standardized algorithm to imaging in patients suspected of having osteomyelitis, in general imaging should begin with conventional radiographs (x-rays) (algorithm 1). Even though plain radiographs have lower sensitivity than other options, they are more readily available and are useful for detecting alternative or concomitant conditions, as described below (algorithm 1). (See 'Plain radiograph' below.)
If further imaging is necessary, unless contraindicated or unavailable, noncontrasted MRI is usually the study of choice. If more definitive assessment of the tissues surrounding the area of suspected osteomyelitis is desired, additional images with gadolinium enhancement may be acquired concurrently. However, if MRI is not an option, we select one of the other modalities described below. (See 'MRI' below and 'CT' below and 'Role of nuclear imaging' below.)
Ultimately, our selection of imaging tests depends on many factors, including test availability, concern for concomitant soft tissue infection, and patient characteristics that may affect selection (eg, metallic foreign material, kidney disease) (table 1).
IMAGING MODALITIES —
Multiple imaging modalities are available to evaluate osteomyelitis.
Sensitivity and specificity vary substantially by the type of imaging test, the presence of confounding bony pathology, and whether the infection is acute or chronic. In patients with concomitant soft tissue infection, some tests are much better than others at detecting and defining the extent of soft tissue infection (table 1 and figure 1) [5].
Plain radiograph — Despite lower sensitivity than other imaging modalities, we typically obtain plain (ie, conventional) radiograph as an initial test for patients with suspected osteomyelitis (algorithm 1). Plain radiographs are more readily available than other imaging tests and are useful to screen for fracture, metallic foreign body, gas in soft tissue, and other alternative or concomitant diagnoses (image 3) [5-10].
In patients with negative or equivocal radiograph results and continued suspicion of osteomyelitis, advanced diagnostic imaging (eg, MRI) should be pursued. Some patients whose radiograph is consistent with osteomyelitis may not need any additional imaging, while others warrant advanced imaging to better delineate the extent of osteomyelitis or concomitant soft tissue infection (image 1 and image 2 and image 4 and image 5).
Radiographic findings of osteomyelitis include osteopenia and bone resorption, cortical loss, bony destruction, and periosteal reaction [11-14]. In the setting of chronic osteomyelitis, reactive sclerosis, sequestra, and involucrum may be observed [11,12]. Often, there are associated soft tissue findings including swelling, loss of fascial planes, and ulceration; soft tissue gas may be observed.
False-negative results are not uncommon with plain radiograph, especially in patients whose symptom-duration is less than two weeks. Periosteal changes can be subtle and bony destructive changes on radiography lag at least two weeks behind clinical infection because 50 to 75 percent of the bone matrix must be destroyed before plain radiographs demonstrate abnormalities [3,5,15-17].
Conditions that can mimic or obscure the findings of osteomyelitis on plain radiographs include bone trauma, fracture, Charcot arthropathy, and the presence of orthopedic hardware [18,19].
MRI — MRI is useful for delineating the extent of osteomyelitis as well as any associated soft tissue infection (eg, abscess, myositis) (picture 1) [20,21].
For patients who have had plain radiograph and need further imaging, MRI is generally considered the study of choice (image 4 and image 6 and algorithm 1) [22-27]. Depending on the availability of MRI and the level of clinical suspicion for osteomyelitis, some clinicians forego plain radiograph and proceed directly to MRI for initial evaluation [28,29]. However, an expert panel on musculoskeletal imaging recommends plain radiographs before MRI as they may provide complementary information such as improved sensitivity of soft tissue air detection on radiographs (image 4) [10].
MRI has high sensitivity and specificity for diagnosis of osteomyelitis (table 1) [5,8,9,17,22,30-34]. In a 2019 meta-analysis including 81 studies, MRI had a pooled sensitivity and specificity of 96 and 81 percent, respectively [5]. Because MRI has such a high sensitivity, in patients suspected of having osteomyelitis, an MRI with no evidence of osteomyelitis is sufficient for exclusion of osteomyelitis if clinical signs or symptoms have been present for at least one week [1].
MRI findings of osteomyelitis include cortical destruction, bone marrow edema, and soft tissue inflammation (image 5 and table 2) [35,36]. Bone marrow edema is the earliest MRI finding of osteomyelitis but is also present in other infective and noninfective conditions (including reactive to adjacent soft tissue infection). Other conditions that can be similar to those of osteomyelitis on MRI include contusion, fracture, postsurgical change, arthritis, neoplasm, and Charcot arthropathy [37]. Lastly, bone marrow changes may persist for weeks to months after osteomyelitis resolves, so repeat MRIs are generally not performed without clear indication.
Intravenous contrast (eg, gadolinium) does not improve the detection of osteomyelitis on MRI. However, it does provide more detailed information about soft tissue infection and can distinguish between phlegmon, necrotic tissue, abscess, sinus tracts, and fistula (picture 1) [20,21,24,38,39]. In patients with kidney dysfunction, some clinicians avoid gadolinium due to the possibility of nephrogenic systemic fibrosis. However, multiple other agents are considered safe in renal insufficiency. (See "Nephrogenic systemic fibrosis/nephrogenic fibrosing dermopathy in advanced kidney disease".)
In some patients, the presence of metal foreign material is a contraindication to MRI. In others, MRI is safe, but metallic hardware at the site of infection may degrade MRI image quality and limit diagnostic capability. Depending upon the size and the composition of foreign material, altering image acquisition parameters and using metal artifact reduction sequences may allow for an accurate diagnostic study [40].
CT — CT is not as sensitive or specific as MRI but is more readily available in some facilities (table 1). In the aforementioned meta-analysis of 81 studies, CT had a pooled sensitivity and specificity of 70 and 90 percent, respectively [5].
A noncontrast CT provides more information than plain radiographs, including improved assessment of cortical and trabecular integrity, periosteal reaction, intraosseous gas, soft tissue gas, sinus tracts, and osseous sequestra and involucrum.
Although noncontrast CT can detect gas in the soft tissues, intravenous contrast is required for detection of other soft tissue abnormalities such as sinus tracts and distinguishing phlegmon from abscess (image 7) [1,13,18,24,27,41-44]. Use of intravenous contrast for CTs increases risk of acute kidney injury, especially in patients with other risk factors. (See "Contrast-associated and contrast-induced acute kidney injury: Clinical features, diagnosis, and management".)
Like MRI, the presence of metallic hardware at the site of infection can cause artifact that may degrade CT image quality and limit diagnostic capability (image 8). Unlike MRI, where the presence of metallic material may be an absolute contraindication, metal is never a contraindication to CT.
Role of nuclear imaging — These imaging techniques may be helpful in the following situations:
●MRI is not an option
●CT is not an option or reveals inconclusive results
●Biopsy is not feasible
For patients who do not meet the above criteria, nuclear imaging studies have a limited role (table 1 and algorithm 1).
Nuclear imaging includes positron emission tomography (PET; usually combined with CT), single photon emission CT (SPECT-CT), and gamma (scintillation or Anger) studies. These studies require radiotracers, which may be used alone (eg, Gallium-67) or "tagged" to (combined with) other molecules or cells that target the areas of infection [45].
PET — 18F-FDG PET/CT has excellent sensitivity and specificity for detection of osteomyelitis (table 1 and algorithm 1) [22,30,32,46-51]. In a meta-analysis that included 16 studies that assessed PET scans, the pooled sensitivity and specificity were 85 and 93 percent, respectively [5].
18F-FDG PET uses 18F-FDG, a glucose analog that accumulates in leukocytes, which in turn accumulate at the site of infection [48]. These studies are able to detect both acute and chronic osteomyelitis, and can also distinguish osteomyelitis from adjacent soft tissue infection. Preliminary studies suggest PET/CT may be useful distinguishing infection from Charcot arthropathy [46,47]. It also has a relatively quick turn-around time compared with other nuclear studies and is relatively insensitive to metallic hardware. Disadvantages include inability to discriminate tumor and inflammation from infection [46,47,52].
Scintigraphy — Scintigraphic studies generate two-dimensional images, in contrast to other nuclear imaging modalities (eg, PET, SPECT). The main scintigraphic studies that have been used to evaluate osteomyelitis are bone scans, tagged white blood cell (WBC) scans, and gallium scans (table 1 and algorithm 1).
In general, scintigraphic studies have high sensitivity for detecting evidence of inflammation, but may not be able to differentiate osteomyelitis from other conditions, such as recent trauma or surgery, recently healed osteomyelitis, septic arthritis, degenerative joint disease, bone tumors, and Paget disease [53-58]. False-negative results are possible in areas of relative ischemia (a common comorbidity in osteomyelitis) since radiotracer may not be adequately delivered to the target site. Additionally, these studies are often more time consuming than other modalities.
A meta-analysis including 34 publications of scintigraphic studies found a pooled sensitivity and specificity of 84 and 71 percent, respectively [5].
Three-phase bone scan — Three-phase bone scans use a radionuclide tracer (typically technetium-99m bound to a phosphorus-containing compound [eg, Tc-99m MDP]) that accumulates in areas of bone turnover and increased osteoblast activity [57]. The scans are performed using a gamma camera at three points in time following tracer injection: immediately after injection (blood flow phase), 15 minutes after injection (blood pool phase), and four hours after injection (osseous phase). In the setting of osteomyelitis, there is intense uptake in all three phases; in the setting of cellulitis, there is increased activity only in the first two phases and normal or mild diffuse increased activity in the third phase (image 9) [13].
The sensitivity and specificity of three-phase bone scan for detection of osteomyelitis depends on findings from concomitant plain radiographs. If radiographs are normal, three-phase bone scan has a sensitivity and specificity of approximately 95 percent [1,57,59]. If radiographs are abnormal, sensitivity remains high, but specificity declines. Causes of false-positive results include fracture, presence of orthopedic hardware, arthritis, Charcot arthropathy, soft tissue infection, and neoplasm [7,45,47,58,60].
Tagged wbc scan — For a tagged WBC scan, also known as a tagged leukocyte scan, the tagged white cells accumulate in the bone marrow and at sites of inflammation or infection; they are not specific for bone (image 9). Reported sensitivities and specificities are similar to bone scans and range from 72 to 100 percent [30,48,61].
Tagged WBC scans use autologous WBCs and require a circulating WBC concentration of at least 2000 per microliter [26,47]. The WBCs are labeled with indium-111 oxyquinoline, gallium-67, or technetium-99m [1], which vary in their availability, stability, and discriminatory ability [48]. To perform the study, blood is drawn from the patient, and the white blood cells are separated for labeling with tracer; after a few hours, the tagged white cells are returned to the patient's circulation via intravenous injection. The time of imaging is dependent upon the isotope used and may be longer than 24 hours [24].
At most sites, specific isotope selection is dependent upon availability.
Gallium and dual tracer studies — Gallium scans utilize the affinity of gallium-67 to acute phase reactants (lactoferrin, transferrin, and others) to demonstrate areas of inflammation that may be related to infection [61]. If negative, gallium scans effectively exclude the diagnosis of osteomyelitis [1,62,63].
Despite good sensitivity, these scans are used less commonly than other nuclear modalities because they have suboptimal specificity, require one to three days for completion, and expose patients to a relatively high radiation dose. At present, gallium imaging is limited primarily to the spine [48].
A gallium scan can be performed concurrently with a technetium-labeled three-phase bone scan by injecting both radionuclides at the same time; this combination of modalities is called a "dual tracer" study. The result is positive if gallium tracer uptake is greater than that of the bone scan.
SPECT and SPECT-CT — SPECT (single-photon emission CT) uses radioisotopes imaged with a gamma camera that is rotated around the patient, producing cross sectional images that together result in three-dimensional imaging. CT is often included as a component of SPECT studies.
SPECT can easily be added as a complement to any two-dimensional gamma imaging study (eg, three-phase bone scan, tagged WBC scan, gallium scan) to provide anatomic cross sectional images, improving the distinction between soft tissue infection and osteomyelitis (image 9) [64]. Although SPECT can also be used as a stand-alone study, it is often coupled with a CT acquired in the same sitting to more definitively locate tracer accumulation.
Sensitivity and specificity of SPECT and SPECT-CT rival MRI and PET (table 1 and algorithm 1). In a meta-analysis including 14 studies of SPECT or SPECT-CT for osteomyelitis, pooled sensitivity was 95 percent and specificity was 82 percent [5].
Ultrasound — Ultrasound may be a useful diagnostic tool for circumstances in which other modalities are not readily available. However, bone is not well depicted by ultrasound, and data on sensitivity and specificity of ultrasound for diagnosis of osteomyelitis are scarce.
Ultrasound has been used more frequently in pediatric patients and is an excellent tool for guiding aspiration of abscesses and other fluid collections [1,24,65-68].
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Osteomyelitis and prosthetic joint infection in adults".)
SUMMARY AND RECOMMENDATIONS
Role of imaging – Radiologic imaging is an integral component of the evaluation for suspected osteomyelitis. Although imaging cannot definitively confirm the presence of osteomyelitis, imaging results consistent with osteomyelitis markedly increase the probability of osteomyelitis in patients with compatible clinical findings. (See 'Role of imaging' above.)
General approach – In general, imaging should begin with conventional radiographs (x-rays). If further imaging is necessary, unless contraindicated or unavailable, noncontrasted MRI is usually the study of choice unless it is contraindicated or unavailable (algorithm 1). (See 'General approach' above.)
Specific studies – Sensitivity and specificity vary substantially by the type of imaging test. In patients with concomitant soft tissue infection, some tests are much better than others at detecting and defining the extent of soft tissue infection (table 1). (See 'Imaging modalities' above.)
●Plain radiographs – Despite lower sensitivity than other imaging modalities, we typically obtain plain (ie, conventional) radiograph as an initial test. These tests are more readily available than other imaging tests and are useful to screen for fracture, metallic foreign body, gas in soft tissue, and other alternative or concomitant diagnoses (algorithm 1). (See 'Plain radiograph' above.)
●MRI – For patients who have had plain radiograph and need further imaging, MRI is generally considered the study of choice (algorithm 1). MRI has high sensitivity and specificity for diagnosis of osteomyelitis and any associated soft tissue infection (table 1). (See 'MRI' above.)
●CT scan – CT scans are not as sensitive or specific as MRI but are more readily available in some facilities (algorithm 1). Intravenous contrast is generally required for detection of associated soft tissue infection. (See 'CT' above.)
●Limited role of nuclear imaging – Nuclear imaging studies are less commonly used than other imaging modalities but may be helpful if MRI or CT is not an option, CT reveals inconclusive results, and/or biopsy is not feasible.
Nuclear studies include positron emission tomography (PET) scans, single-photon emission CT (SPECT), and scintigraphic studies (eg, bone scan, tagged white blood cell scan, gallium scan) (algorithm 1). (See 'Role of nuclear imaging' above.)
ACKNOWLEDGMENT —
The editorial staff at UpToDate acknowledge Dr. Perry Horwich and Dr. Mary Hochman, who contributed to earlier versions of this topic review.