DOI: 10.1148/radiol.2382041393
(Radiology 2006;238:622-631.)
© RSNA, 2006
Neuropathic Arthropathy of the Foot with and without Superimposed Osteomyelitis: MR Imaging Characteristics1
Mazyar E. Ahmadi, BS,
William B. Morrison, MD,
John A. Carrino, MD, MPH,
Mark E. Schweitzer, MD,
Steven M. Raikin, MD and
Hans P. Ledermann, MD
1 From Drexel University College of Medicine, Philadelphia, Pa (M.E.A.); Departments of Radiology (W.B.M.) and Orthopedic Surgery, Rothman Institute (S.M.R.), Thomas Jefferson University Hospital, 111 S 11th St, Suite 3390, Philadelphia, PA 19107; Department of Radiology, Brigham and Women's Hospital, Boston, Mass (J.A.C.); Department of Radiology, Hospital for Joint Disease, New York University, New York, NY (M.E.S.); and Radiologisches Institut, Universitatsspital Basel, Basel, Switzerland (H.P.L.). Received August 10, 2004; revision requested October 19; revision received February 14, 2005; accepted March 10; final version accepted April 28.
Address correspondence to W.B.M. (e-mail: William.Morrison{at}Jefferson.edu).
 |
ABSTRACT
|
|---|
Purpose: To determine retrospectively the magnetic resonance (MR) findings associated with pedal neuropathic arthropathy with and without superimposed osteomyelitis and to identify any useful discriminating features.
Materials and Methods: Investigational review board approval was obtained and allowed review of records and images without informed consent. HIPAA compliance was observed. Contrast-enhanced MR images in patients with diabetic neuropathic arthropathy of the foot were examined by two reviewers in consensus. Affected joints were examined for marrow, articular, periarticular, and soft-tissue findings. Presence of superimposed osteomyelitis was documented. A subgroup that had undergone MR before infection was evaluated for comparison; 
2 and t tests were used to evaluate the associations.
Results: Of 128 neuropathic joints in 63 patients (24 female, 39 male; aged 31–78 years), 43 had superimposed osteomyelitis. Effusion was common in all neuropathic joints, but thin rim enhancement was more common in noninfected joints (62% vs 21%, P < .001) and diffuse joint fluid enhancement was more common with infection (47% vs 26%, P = .052). Subluxation, bone proliferation, fragmentation, and erosion were seen in both groups, but intraarticular bodies were more common in noninfected joints (53% vs 12%, P < .001). In the periarticular soft tissues, edema, enhancement, and ulceration were common in both groups. Fluid collections in the soft tissues were more commonly associated with infected joints (95% vs 48%, P < .001) and, when present next to an infected joint, were larger than those next to noninfected neuropathic joints (2.6 cm2 [range, 0.3–8.6 cm2] vs 1.6 cm2 [range, 1.0–2.4 cm2]). Soft-tissue fat replacement (68% vs 36%, P = .002) and sinus tracts (84% vs 0%, P < .001) were also more common with infection. In the marrow, periarticular signal intensity abnormality was common in both groups, but the extent was greater with infection. Subchondral cysts were seen almost exclusively in noninfected joints (76% vs 2%, P < .001). Similar results were obtained in the subgroup of 21 joints (15 patients) with both pre- and postinfection MR images.
Conclusion: Sinus tract, replacement of soft-tissue fat, fluid collection, and extensive marrow abnormality are MR imaging features indicating superimposed infection. Thin rim enhancement of effusion, presence of subchondral cysts, or intraarticular bodies indicate absence of infection.
© RSNA, 2006
|
INTRODUCTION
|
|---|
In the United States, approximately 15 million persons are afflicted with diabetes mellitus (1). Diabetic patients experience considerable morbidity and mortality from microvascular (2,3) and macrovascular (4) disease, as well as peripheral sensorimotor (5) and autonomic (6,7) neuropathy, which occur in more than half of these patients. This combination often results in deformity and altered weight bearing (8) with subsequent osteoarthropathy (9), callus (10), ulceration (11–14), and infection (15,16), the end result of which is often amputation (17,18).
Magnetic resonance (MR) imaging has shown utility for the diagnosis of pedal osteomyelitis (19–21). However, neuropathic arthropathy can simulate infection in its clinical appearance, with swelling and erythema of the foot, and also at MR imaging, with bone marrow edema and enhancement (22–26). Since infection is often superimposed in patients with neuropathic disease, distinguishing between the two abnormalities with MR imaging is further complicated (27,28). We hypothesized that certain MR findings (quantitative or qualitative) could serve as useful discriminators between infection and noninfection. Thus, the purpose of this study was to determine retrospectively the MR imaging findings associated with pedal neuropathic arthropathy with and without superimposed osteomyelitis and to identify any useful discriminating features.
|
MATERIALS AND METHODS
|
|---|
Patients
Investigational review board approval was obtained at Thomas Jefferson University Hospital prior to performance of this study, and it allowed retrospective review of records and images without informed consent. Health Insurance Portability and Accountability Act compliance was observed. The study population consisted of consecutive diabetic patients with a clinical diagnosis of diabetic pedal neuropathy who were undergoing MR imaging of the foot with contrast material enhancement at our institution from January 1997 to August 2002. All patients had clinically documented neuropathic disease and established osteoarthropathy and were referred for MR imaging for clinically suspected infection of a particular region of the foot or ankle, as specified by history, skin marker, and/or ulceration. Any joint with an MR imaging abnormality within the region of clinically suspected infection was included in the analysis. Two of us (M.E.A. and H.P.L.) reviewed the medical records for 68 feet of 63 patients (24 female, 39 male; mean age, 53 years; range, 31–78 years), who comprised this study population. In 15 patients, preinfection MR image findings were available for comparison.
MR Imaging
All MR imaging was performed with a 1.5-T magnet (Signa; GE Medical Systems, Milwaukee, Wis) and an extremity coil. T1-weighted spin-echo images were obtained with the following parameters: two signals acquired; 400–750/8–12 (repetition time msec/echo time msec); field of view, 15–20 cm; and matrix, 256 x 192 or 256 x 256. Fat-suppressed T2-weighted images were obtained with the following parameters: fast spin-echo sequence with two signals acquired; 2000–7800/75–108 (repetition time msec/effective echo time msec); echo train length, eight; field of view, 15–20 cm; and matrix, 256 x 192. Plane selection depended in part on the site of infection, but T1-weighted and T2-weighted images were acquired in at least two planes in all cases, with a section thickness of 4–5 mm and a 1-mm gap.
Pre- and postcontrast fat-suppressed T1-weighted images were obtained with a fast multiplanar spoiled gradient-recalled-echo (GRE) sequence at the following settings: 200–350/1.8–2.2 (repetition time msec/echo time msec); flip angle, 90°; field of view, 15–20 cm; matrix, 256 x 192; and section thickness, 3–4 mm with a 1-mm gap. Postcontrast images were acquired in at least two planes. Gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) was used as the intravenous contrast agent, at a dose of 0.1 mmol per kilogram of body weight. T1-weighted fast multiplanar spoiled GRE sequences were performed with fat suppression and identical parameters before and after contrast material administration. For all T2-weighted fast spin-echo and gadolinium-enhanced T1-weighted sequences, fat suppression was accomplished through selective presaturation of lipid resonance frequency. In all examinations, sagittal fast spin-echo inversion-recovery (IR) images were also obtained with the following parameters: echo train length, eight; 3000–6000/45–75/150 (repetition time msec/echo time msec/inversion time msec); matrix, 256 x 192; section thickness, 4 mm with a 1-mm gap; and field of view, 16–20 cm.
Image Interpretation
Two reviewers (W.B.M. and M.E.S.), who had 11 and 14 years of experience in interpreting MR images of the foot, respectively, analyzed all MR images in consensus, blinded to patient history, clinical examination findings, laboratory data, and the location of suspected osteomyelitis or neuropathic arthropathy. The reviewers did know, however, that all patients had diabetic peripheral neuropathy and that they had undergone imaging because of concern about infection based on clinical findings. If multiple postinfection MR images were available, only the initial postinfection images were used for analysis.
Images were examined, and findings were categorized as joint, periarticular soft-tissue, or periarticular bone marrow abnormalities. However, there are numerous articulations at the hindfoot and midfoot that may be affected by neuropathic disease or infection together; for practical purposes, if multiple joints in a localized region showed a similar abnormality, findings were aggregated into a single joint region.
Articular findings were subcategorized as effusion (graded subjectively as small, medium, or large), joint fluid enhancement on postcontrast images (subjectively graded as thin rim enhancement, thick rim enhancement, or diffuse joint fluid enhancement), subluxation (mild, moderate, or marked offset of the articular surfaces or dislocation on any images [complete incongruity of the joint]), erosion (focal defects at the margin of the joint), destruction (absence of articular surfaces), intraarticular bodies (intraarticular foci completely surrounded by joint fluid), fragmentation (articular surface broken apart), or bone proliferation (osteophytes or additional bone formation at the joint margin).
Findings in adjacent soft tissues were subcategorized as fluid collection (subjectively categorized as small, medium, or large foci of fluid signal intensity, measured in millimeters from sagittal short inversion time IR images, expressed as an area by multiplying the two dimensions), rim enhancement (of any fluid collection), fat replacement (loss of T1 signal intensity in the subcutaneous tissues), edema (ill-defined hyperintensity in the subcutaneous tissues on images obtained with T2-weighted or short inversion time IR sequences, hereafter referred to together as fluid-sensitive sequences), enhancement (periarticular only, subcutaneous only, or both), ulceration (interruption of the cutaneous signal intensity line), or sinus tract (linear fluid signal intensity seen with fluid-sensitive sequences or parallel lines of enhancement in a tram-track pattern on postcontrast images).
Findings in periarticular bone marrow were subcategorized as low T1 signal intensity, high T2 signal intensity, or gadolinium enhancement. The distance of signal intensity abnormalities from the articular surface (focal, expressed in millimeters from the articular surface, or diffuse if the whole bone was involved), as well as the degree of enhancement (mild, moderate, or marked), was recorded. The presence of subchondral cysts was also recorded. Prior comparison MR images acquired before infection, when available, were evaluated separately for the above findings according to the same criteria. This evaluation was not blinded. Changes in findings, along with the intervals between MR imaging examinations, were recorded.
Selection criteria ensured that all patients had neuropathic pedal disease with arthropathy. A reference standard for presence of infection was then obtained for each subject. Medical records were evaluated by two of us (M.E.A., H.P.L.) who were not involved in image interpretation and were blinded to data collected from image review. Records were reviewed to document the presence and location of infection. The diagnosis of osteomyelitis was based on tissue diagnosis obtained during amputation, open surgical débridement, surgical biopsy, or percutaneous biopsy of the suspected area(s). Biopsy was guided in all cases by information obtained from the MR examination with respect to marrow abnormality.
A positive culture or histologic diagnosis of osteomyelitis was considered positive for infection. If both culture and histologic findings were negative for infection, the site was categorized as negative for osteomyelitis. Findings at clinical follow-up were also reviewed. On the basis of this review, the presence and location of infection and the organism cultured, as well as the time of clinical onset of infection compared with that of MR imaging, were recorded. The interval between biopsy or culture and MR imaging examination was recorded. On the basis of this clinical information, abnormal joints were divided into neuropathic arthropathy without superimposed infection and neuropathic arthropathy with superimposed infection.
Statistical Analyses
Descriptive statistics of the MR imaging findings in the joints, adjacent soft tissues, and periarticular bone marrow were calculated for each of the two study groups (infected vs noninfected joints, preinfection vs postinfection images). For dichotomous variables (finding either present or absent), a 2 test was performed to compare differences in proportions of findings between the infected and noninfected subgroups. For continuous variables (measurements of extent and area), the subgroups were compared with a two-tailed unpaired t test. For the study group with preinfection MR images, paired analysis for changes between pre- and postinfection findings was performed with the McNemar test on a contingency table. For continuous variables, pre- and postinfection findings were compared by using a two-tailed paired t test. The statistical software used was Simple Interactive Statistical Analysis (Uitenbroek DG, 1997; available at http://home.clara.net/sisa). P values of less than .05 were considered to indicate significant differences.
While diabetic neuropathy is a systemic disorder, musculoskeletal foot manifestations are often asymmetric in distribution. A neuropathic foot is predisposed to infection because an insensate extremity is at risk for skin ulceration and contamination. Osteomyelitis results as a superimposed process from a transcutaneous spread of a soft-tissue infection, which is an independent event for each foot. Therefore, we used the assumption that the lesions were independent because of the small number of potentially clustered observations and the clinical context of the problem being investigated.
|
RESULTS
|
|---|
Overall, 68 feet of 63 patients with neuropathic arthropathy were examined. Chart and laboratory review revealed that 85 of the 128 joints evaluated were neuropathic only, and 43 were neuropathic with superimposed osteomyelitis. With regard to the 128 sites evaluated, diagnosis was documented by amputation in 75, by surgical excision or biopsy in 23, and by percutaneous biopsy in 32. Of the 43 joints positive for infection, 10 (23%) had Staphylococcus aureus, and 33 (77%) had multiorganism infections. All patients were diabetic.
Histopathologic evaluation and bone culture were performed an average of 9.6 days after imaging. Joints with neuropathic arthropathy were located in the forefoot in 14% (18 of 128 joints), in the midfoot in 55% (70 of 128 joints), and in the hindfoot in 31% (40 of 128 joints) of cases; 56% (24 of 43 joints) of all infections occurred in the forefoot, 44% (19 of 43 joints) occurred in the midfoot, and 0% occurred in the hindfoot. Fifteen patients (21 joints) with superimposed infection had for comparison an MR image obtained before clinical suspicion of infection in that region (average interval between examinations, 7 months; range, 1–42 months).
Tables 1–6 list results for all MR imaging signs evaluated. Results are listed in separate tables for the overall population (noninfected vs infected neuropathic joints) and for the subgroup with available prior comparison MR images (ie, pre- vs postinfection images). The following is a summary of the findings, divided into articular, periarticular soft-tissue, and marrow findings:
View this table:
[in this window]
[in a new window]
|
Table 3. Periarticular Soft-Tissue Findings on MR Images of Neuropathic Joints with and without Superimposed Infection
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Periarticular Soft-Tissue Findings on MR Images of Neuropathic Joints Obtained before and after Infection
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Periarticular Bone Marrow Findings on MR Images of Neuropathic Joints with and without Superimposed Infection
|
|
View this table:
[in this window]
[in a new window]
|
Table 6. Periarticular Bone Marrow Findings on MR Images of Neuropathic Joints Obtained before and after Infection
|
|
Articular Findings
Joint effusion and enhancement.—Effusion was seen in neuropathic joints with and in those without infection (Table 1); there was no significant difference between the groups, even when preinfection image findings were compared with postinfection image findings (Table 2). However, when intravenous contrast material was administered, the neuropathic joints without infection tended to show thin rim enhancement (62% vs 21% in infected joints, P < .001). With infected joints, thick rim enhancement (26%) or diffuse joint fluid enhancement (47%) predominated (Fig 1). In the population with available preinfection MR images, effusions became more frequent and larger after infection, although 48% had effusion initially; differences were not significant.
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a: MR images show thick rim enhancement of joint fluid in a 55-year-old man with neuropathic arthropathy and surgically documented superimposed septic arthritis and osteomyelitis. (a) Sagittal fast spin echo IR image (4817/75/150) shows effusion of the fourth metatarsophalangeal joint (arrow), which is subluxed. (b) Corresponding sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (250/2) shows thick rim enhancement (arrow) of the joint fluid. Note the presence of an ulcer and sinus tract (arrowhead).
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b: MR images show thick rim enhancement of joint fluid in a 55-year-old man with neuropathic arthropathy and surgically documented superimposed septic arthritis and osteomyelitis. (a) Sagittal fast spin echo IR image (4817/75/150) shows effusion of the fourth metatarsophalangeal joint (arrow), which is subluxed. (b) Corresponding sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (250/2) shows thick rim enhancement (arrow) of the joint fluid. Note the presence of an ulcer and sinus tract (arrowhead).
|
|
Subluxation and dislocation.—Subluxation and dislocation (Fig 1) were common overall and slightly more common in infected than in noninfected neuropathic joints (79% vs 62%). Differences were not significant.
Bone fragmentation and proliferation.—The frequencies of bone fragmentation (51% vs 51%) and proliferation (45% vs 40%) were similar in neuropathic joints with and without infection. Differences were not significant.
Erosion and destruction.—Joint erosion (51% vs 39%) and destruction (49% vs 38%) were more common in infected than in noninfected neuropathic joints, but again, differences were not statistically significant. However, in the subgroup with preinfection MR images, the frequency of erosion (Fig 2) increased significantly after infection (from 19% to 57%, P = .008).
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a: MR images show joint erosion. (a) Sagittal fast spin-echo IR image (5100/45/150) of the midfoot of a 52-year-old man with neuropathic arthropathy of the Lisfranc joint shows mild subchondral edema (arrowhead) but preservation of the articular surfaces. (b) Follow-up sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE MR image (333/2) obtained in the same patient 16 months later shows erosion of the joint margin (arrowhead). Superimposed osteomyelitis was documented at biopsy of the first metatarsal base and cuneiform. Note also discrete marrow enhancement, extending 2.0 cm from the metatarsal articular surface (long arrow) and diffusely involving the cuneiform (short arrow).
|
|
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b: MR images show joint erosion. (a) Sagittal fast spin-echo IR image (5100/45/150) of the midfoot of a 52-year-old man with neuropathic arthropathy of the Lisfranc joint shows mild subchondral edema (arrowhead) but preservation of the articular surfaces. (b) Follow-up sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE MR image (333/2) obtained in the same patient 16 months later shows erosion of the joint margin (arrowhead). Superimposed osteomyelitis was documented at biopsy of the first metatarsal base and cuneiform. Note also discrete marrow enhancement, extending 2.0 cm from the metatarsal articular surface (long arrow) and diffusely involving the cuneiform (short arrow).
|
|
Intraarticular bodies.—Intraarticular bodies were seen in approximately half of the neuropathic joints without infection (53%). The frequency of these bodies was lower in infected neuropathic joints (12%, P < .001). This finding was confirmed in the subgroup with preinfection images, with bodies detected in 33% of joints at the initial examination and in 10% after infection, although the difference was of borderline significance (P = .06).
Periarticular Soft Tissues
Fluid collections.—Soft-tissue fluid collections were more common adjacent to infected joints (95% vs 48% next to noninfected neuropathic joints, P < .001) (Table 3) and, when present, were larger than those next to neuropathic joints without infection (mean, 2.6 cm2 [range, 0.3–8.6 cm2] vs 1.6 cm2 [range, 1.0–2.4 cm2], P < .001). Rim enhancement was almost universal in fluid collections adjacent to infected joints (98%), but this finding was also common in fluid collections next to noninfected neuropathic joints (78%). In the subgroup that underwent MR imaging prior to infection, periarticular fluid collections became more common after infection (100% vs 67%, P = .016) (Table 4), but the mean size of the fluid collections decreased (from 3.3 cm2 [range, 3.0–4.2 cm2] to 1.6 cm2 [range, 0.3–2.5 cm2]).
Soft-tissue signal intensity.—Soft-tissue edema was seen with similar frequencies next to infected and noninfected neuropathic joints (79% vs 68%). However, loss of subcutaneous fat signal intensity adjacent to the joint (Fig 3) was seen significantly more frequently in the setting of infection (67% vs 36%, P = .002). Soft-tissue enhancement was common in both groups (Fig 4) and was slightly more common adjacent to noninfected neuropathic joints (84% vs 65%).
View larger version (71K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a: MR images show subcutaneous fat replacement. (a) Sagittal T1-weighted image (500/10) of the midfoot of a 54-year-old man shows preservation of fat signal intensity in the subcutaneous tissues. (b) Sagittal T1-weighted image (600/10) of the foot of the same patient 12 months later shows replacement of the subcutaneous fat signal intensity (arrows); at this time osteomyelitis was confirmed at biopsy of the cuboid.
|
|
View larger version (58K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b: MR images show subcutaneous fat replacement. (a) Sagittal T1-weighted image (500/10) of the midfoot of a 54-year-old man shows preservation of fat signal intensity in the subcutaneous tissues. (b) Sagittal T1-weighted image (600/10) of the foot of the same patient 12 months later shows replacement of the subcutaneous fat signal intensity (arrows); at this time osteomyelitis was confirmed at biopsy of the cuboid.
|
|
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a: MR images show soft-tissue enhancement. (a) Coronal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (250/2) obtained in a 52-year-old man with neuropathic arthropathy of the midfoot shows a small cutaneous ulcer (arrow) and enhancement of the surrounding subcutaneous tissues (arrowheads). The patient's condition responded to conservative management. (b) Another coronal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (280/2) obtained 17 months later for suspicion of osteomyelitis shows recurrent ulceration (white arrow) and both subcutaneous and periarticular enhancement (arrowheads). Biopsy of the medial cuneiform subsequently proved the presence of osteomyelitis. Note diffuse enhancement of the first and second cuneiforms (black arrows).
|
|
View larger version (161K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b: MR images show soft-tissue enhancement. (a) Coronal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (250/2) obtained in a 52-year-old man with neuropathic arthropathy of the midfoot shows a small cutaneous ulcer (arrow) and enhancement of the surrounding subcutaneous tissues (arrowheads). The patient's condition responded to conservative management. (b) Another coronal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (280/2) obtained 17 months later for suspicion of osteomyelitis shows recurrent ulceration (white arrow) and both subcutaneous and periarticular enhancement (arrowheads). Biopsy of the medial cuneiform subsequently proved the presence of osteomyelitis. Note diffuse enhancement of the first and second cuneiforms (black arrows).
|
|
Ulceration and sinus tract.—Ulceration (Figs 1, 4) was common next to both infected (72%) and noninfected joints (78%); these results are compatible with the high incidence of this finding in diabetics with clinically suspected infection, who formed our study population. However, the presence of a sinus tract (Figs 1, 5) in the soft tissues next to the joint was seen only in infected neuropathic joints (84% vs 0%, P < .001). This finding was confirmed in the subgroup with preinfection images; a sinus tract was seen in 0% of joints on the initial examination and in 90% on the postinfection image (P < .001).
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a: MR images show development of a sinus tract. (a) Sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (230/2) obtained in a 60-year-old man with neuropathic arthropathy of the midfoot and hindfoot shows multiple joint subluxations with subcutaneous enhancement only (arrow). (b) Twelve months later the patient presented with a draining plantar ulcer; another sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (280/2) shows a tram-track pattern of soft-tissue enhancement representing a sinus tract (arrowheads) leading to the cuboid (arrow), which demonstrates enhancement proved to represent osteomyelitis at surgery.
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b: MR images show development of a sinus tract. (a) Sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (230/2) obtained in a 60-year-old man with neuropathic arthropathy of the midfoot and hindfoot shows multiple joint subluxations with subcutaneous enhancement only (arrow). (b) Twelve months later the patient presented with a draining plantar ulcer; another sagittal T1-weighted fat-suppressed postcontrast fast multiplanar spoiled GRE image (280/2) shows a tram-track pattern of soft-tissue enhancement representing a sinus tract (arrowheads) leading to the cuboid (arrow), which demonstrates enhancement proved to represent osteomyelitis at surgery.
|
|
Periarticular Bone Marrow
Marrow signal intensity.—Evaluation of separate sequences revealed that the presence of periarticular marrow signal intensity abnormality did not differ significantly between neuropathic joints with and those without infection (Table 5). The mean extent of signal intensity abnormality from the articular surface was significantly greater in the setting of infection for all sequences (1.9 vs 1.1 cm for low T1 signal intensity, P = .007; 2.6 vs 1.5 cm for high T2 signal intensity, P = .003; 2.4 vs 1.3 cm for gadolinium enhancement, P < .001). Although the average extent differed between the two groups, the ranges overlapped such that no cutoff could be proposed. However, diffuse marrow involvement (eg, no normal marrow signal intensity in the entire bone) was found to be significantly more common in the setting of infection (Figs 2, 4, 6) over all sequences (41% vs 12% for low T1 signal intensity, P = .003; 48% vs 12% for high T2 signal intensity, P < .001; 42% vs 9% for gadolinium enhancement, P < .001).
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a: MR images show disappearance of subchondral cysts. (a) Sagittal fast spin-echo IR image (5000/70/150) obtained in a 55-year-old man with neuropathic arthropathy of the Lisfranc joint shows subchondral cysts (arrowheads); at this time there was no clinical suspicion of infection. (b) MR imaging was performed again in the same patient after 11 months, after development of a draining sinus tract originating from the midfoot; sagittal fast spin-echo IR image (4817/75/150) of the same area shows disappearance of the subchondral cysts, erosion and destruction of the joint (arrows), and diffuse marrow edema in the bones of the midfoot and hindfoot, which was subsequently proved to represent osteomyelitis.
|
|
View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b: MR images show disappearance of subchondral cysts. (a) Sagittal fast spin-echo IR image (5000/70/150) obtained in a 55-year-old man with neuropathic arthropathy of the Lisfranc joint shows subchondral cysts (arrowheads); at this time there was no clinical suspicion of infection. (b) MR imaging was performed again in the same patient after 11 months, after development of a draining sinus tract originating from the midfoot; sagittal fast spin-echo IR image (4817/75/150) of the same area shows disappearance of the subchondral cysts, erosion and destruction of the joint (arrows), and diffuse marrow edema in the bones of the midfoot and hindfoot, which was subsequently proved to represent osteomyelitis.
|
|
Subchondral cysts.—Subchondral cysts were common in the neuropathic joints without infection (76%) and rare in the neuropathic joints with superimposed infection (2%, P < .001). In the subgroup with images obtained before joint infection, subchondral cysts present at the initial examination (62%) were no longer apparent after the onset of joint infection (Fig 6).
|
DISCUSSION
|
|---|
Various studies have shown the utility of MR imaging for detection of osteomyelitis in the foot and ankle (19,22,23,25,29–31). However, the presence of concurrent neuropathic disease can lower the specificity of MR imaging findings for osteomyelitis (21,24–26,28,31). Neuropathic disease, especially when acute, can simulate infection, manifesting at clinical examination as warm, swollen, erythematous tissue (32) and at MR imaging as diffuse soft-tissue edema (31,33), soft-tissue fluid collection, effusion, and marrow abnormalities, with marrow enhancement and periarticular soft-tissue enhancement on postcontrast images (29).
The frequencies of many articular MR signs, including effusion, subluxation, bone proliferation, fragmentation, and destruction, were similar in the infected and noninfected neuropathic joints. The high incidence of rim enhancement and joint marginal erosion in infected joints is also predictable, owing to the inflammatory nature of the disease process. Interestingly, the frequency of intraarticular bodies was lower in infected neuropathic joints. The reason for this finding is unclear; it may be due to dissolution of the bodies after superinfection of the joint. Alternatively, it could be that the bodies are obscured by surrounding inflammation or that their signal intensity characteristics change such that they more closely approximate those of fluid. The dissolution hypothesis could be tested by using correlation with computed tomographic findings, which were not available for our investigation. The low incidence of subchondral cysts in the group with infection (and disappearance on follow-up images of neuropathic joints that became superinfected) has not been previously reported, to our knowledge. This is potentially a useful finding, particularly since neuropathic joints commonly undergo cystic change. The cysts might be merely obscured by the surrounding subchondral signal intensity in infection, or they could possibly be decompressed or destroyed by articular erosion. Nevertheless, the presence of subchondral cysts seems to be an excellent discriminator, suggesting the absence of superimposed infection.
Since osteomyelitis and septic arthritis of the foot in diabetics usually result from transdermal inoculation and soft-tissue spread, periarticular soft-tissue signs have been suggested as good MR imaging discriminators for deep infection (20). However, the utility of these findings has not been tested in a pure neuropathic population of patients who are particularly susceptible to callus and ulcer formation with associated soft-tissue alteration. In fact, the incidences of skin ulceration were similar in patients with and those without infection. However, we found that in the neuropathic foot, replacement of soft-tissue fat signal intensity adjacent to the joint in question did suggest underlying infection. Similarly, soft-tissue fluid collections were more common adjacent to infected joints, and when present they were larger than those next to neuropathic joints without infection. However, we also found evidence that sinus tracts may drain these fluid collections, resulting in paradoxically decreased size after development of superimposed infection. Rim enhancement was almost universal in fluid collections adjacent to infected joints (98%), but this sign was also common in the fluid collections next to noninfected neuropathic joints. Soft-tissue enhancement can be influenced to a certain extent by the degree of soft-tissue ischemia and could lead to a false-negative finding in the setting of infection. Also, hyperemia in acute neuropathic disease could result in false-positive enhancement, simulating infection. Our findings suggest that soft-tissue enhancement or the lack thereof is not a good discriminator for distinguishing between infected and noninfected neuropathic joints.
Although on average there was more extensive marrow abnormality adjacent to infected neuropathic joints than adjacent to noninfected joints, there was too much overlap in measurements to propose a discriminatory cutoff point. However, diffuse marrow involvement (eg, no normal marrow signal intensity in the entire bone with any sequence) was seen commonly in the setting of infection and rarely without infection and should be considered a potentially useful MR imaging sign. This finding is consistent with the pathologic etiology of neuropathic osteoarthropathy; since it is a joint-based process, reactive changes would be expected to predominate in the subchondral bone. Although marrow changes associated with pedal osteomyelitis on MR images have been discussed by numerous previous authors, the finding of diffuse marrow abnormality as a sign of infection in the context of underlying neuropathic arthropathy has not been evaluated, to our knowledge.
We acknowledge certain limitations to our methodology. The MR imaging findings evaluated are often adjacent to each other on individual images and cannot be blocked out. This is a source of potential bias. The fact that referring clinicians were not blinded to original MR imaging interpretations likely influenced therapy and certainly guided biopsy locations; this was another unavoidable source of bias. However, it is also likely that the influence of MR findings on the planning of biopsy sites reduced the incidence of false-negative results. Radiographic correlation was not consistently available; therefore, radiographic findings were not incorporated into the inclusion criteria. However, any joint in the region of clinical concern on radiographs would undoubtedly have had MR imaging abnormalities, so we are confident that no abnormal joints were missed owing to lack of radiographic correlation. Conversely, normal radiographs do not exclude early neuropathic osteoarthropathy, so use of radiographs as a reference standard for inclusion would limit analysis to chronic disease only.
We also counted small adjacent joints, such as those of the midfoot, together. This was done for two reasons: First, owing to volume-averaging effects, it can be difficult and impractical to separate findings in these joints. Second, similar MR imaging findings are seen in small adjacent joints in both infection and neuropathic disease when multiple such joints are involved in the same patient. Counting them separately would skew the data toward patients and joint regions with more extensive disease of either type; for example, data from one patient with a midfoot infection could count 14 times more than data from another patient with a metatarsophalangeal infection.
Finally, the reference standard used, tissue sampling with microbiologic and histologic confirmation, is optimal but remains limited. Since bone is the tissue sampled, a reading of osteomyelitis or no osteomyelitis is rendered. However, the MR imaging findings and areas of clinical concern are centered around joints, since the underlying disease affecting all subjects is neuropathic osteoarthropathy. This leads to an apparent disparity of terminology, with joint-centric descriptors used for MR imaging findings and a reference standard documenting the presence or absence of osteomyelitis. Although it is possible that septic arthritis could have been classified as "no osteomyelitis" by using the reference standard, we consider this possibility unlikely, since most cases were sampled by means of amputation or open surgical biopsy.
There has been very little prior work involving the systematic analysis of MR imaging differences and similarities in pedal neuropathic osteoarthropathy and osteomyelitis (33). To our knowledge, in no prior study have investigators examined each separate sign as we did in a large, purely neuropathic diabetic population, which included many patients in whom multiple examinations revealed the development of infection in joints with neuropathic osteoarthropathy. We have defined the relative utility of separate MR imaging findings for the diagnosis of osteomyelitis in the setting of underlying neuropathic disease. This is potentially more useful than testing the overall accuracy of a gestalt impression of experienced investigators, which less experienced readers may not be able to reproduce. It can also give the practicing radiologist an idea of which findings are more important when presented with patterns of disease that do not correspond to the "textbook" standard. This format can also be used to generate prediction models, in which separate findings are used in a weighted fashion to arrive at a probability of disease.
In conclusion, while many MR imaging features coexist in neuropathic arthropathy with and without superimposed osteomyelitis, several findings may serve as useful discriminators. The presence of sinus tract, replacement of soft-tissue fat, extensive or diffuse marrow abnormality, thick rim enhancement or diffuse joint fluid enhancement, and joint erosion support superimposed osteomyelitis; conversely, at an abnormal joint in a diabetic foot, the preservation of subcutaneous fat, absence of soft-tissue fluid collections, presence of subchondral cysts, and presence of intraarticular bodies support neuropathic arthropathy alone, without infection.
|
FOOTNOTES
|
|---|
Abbreviations: GRE = gradient-recalled echo • IR = inversion recovery
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, M.E.A., W.B.M., H.P.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, M.E.A., W.B.M.; clinical studies, M.E.A., W.B.M., M.E.S., S.M.R., H.P.L.; statistical analysis, M.E.A., J.A.C., M.E.S., H.P.L.; and manuscript editing, M.E.A., W.B.M., J.A.C., H.P.L.
|
References
|
|---|
- Boulton AJ, Vileikyte L. The diabetic foot: the scope of the problem. J Fam Pract 2000;49:S3–S8.[Medline]
- Ierardi RP, Shuman CR. Control of vascular disease in patients with diabetes mellitus. Surg Clin North Am 1998;78:385–392.[CrossRef][Medline]
- Tooke JE. Microvascular function in human diabetes: a physiological perspective. Diabetes 1995;44:721–726.[Abstract]
- Arora S, LoGerfo FW. Lower extremity macrovascular disease in diabetes. J Am Podiatr Med Assoc 1997;87:327–331.[Abstract]
- O'Brien SP, Schwedler M, Kerstein MD. Peripheral neuropathies in diabetes. Surg Clin North Am 1998;78:393–408.[CrossRef][Medline]
- Goodwin DW, Salonen DC, Yu JS, Brossmann J, Trudell DJ, Resnick DL. Plantar compartments of the foot: MR appearance in cadavers and diabetic patients. Radiology 1995;196:623–630.[Abstract/Free Full Text]
- Moore TE, Yuh WT, el-Khoury GY. Plantar compartments of the foot: MR appearance in cadavers and diabetic patients [letter]. Radiology 1996;198:909.
- Cavanagh PR, Ulbrecht JS, Caputo GM. Biomechanical aspects of diabetic foot disease: aetiology, treatment, and prevention. Diabet Med 1996;13(suppl 1):S17–S22.
- Younger DS, Bronfin L. Overview of diabetic neuropathy. Semin Neurol 1996;16:107–113.[Medline]
- Lavery L, Gazewood JD. Assessing the feet of patients with diabetes. J Fam Pract 2000;49(11 suppl):S9–S16.[Medline]
- Hill SL, Holtzman GI, Buse R. The effects of peripheral vascular disease with osteomyelitis in the diabetic foot. Am J Surg 1999;177:282–286.[CrossRef][Medline]
- Lavery LA, Armstrong DG, Vela SA, Quebedeaux TL, Fleischli JG. Practical criteria for screening patients at high risk for diabetic foot ulceration. Arch Intern Med 1998;158:157–162.[Abstract/Free Full Text]
- Mueller MJ, Minor SD, Diamond JE, Blair VP 3rd. Relationship of foot deformity to ulcer location in patients with diabetes mellitus. Phys Ther 1990;70:356–362.[Abstract/Free Full Text]
- Reiber GE, Vileikyte L, Boyko EJ, et al. Causal pathways for incident lower-extremity ulcers in patients with diabetes from two settings. Diabetes Care 1999;22:157–162.[Abstract/Free Full Text]
- Fry DE, Marek JM, Langsfeld M. Infection in the ischemic lower extremity. Surg Clin North Am 1998;78:465–479.[CrossRef][Medline]
- Joshi N, Caputo GM, Weitekamp MR, Karchmer AW. Infections in patients with diabetes mellitus. N Engl J Med 1999;341:1906–1912.[Free Full Text]
- Most RS, Sinnock P. The epidemiology of lower extremity amputations in diabetic individuals. Diabetes Care 1983;6:87–91.[Abstract]
- Humphrey LL, Palumbo PJ, Butters MA, et al. The contribution of non-insulin-dependent diabetes to lower-extremity amputation in the community. Arch Intern Med 1994;154:885–892.[Abstract]
- Wrobel JS, Connolly JE. Making the diagnosis of osteomyelitis: the role of prevalence. J Am Podiatr Med Assoc 1998;88:337–343.[Abstract]
- Morrison WB, Schweitzer ME, Batte WG, Radack DP, Russel KM. Osteomyelitis of the foot: relative importance of primary and secondary MR imaging signs. Radiology 1998;207:625–632.[Abstract/Free Full Text]
- Morrison WB, Schweitzer ME, Bock GW, et al. Diagnosis of osteomyelitis: utility of fat-suppressed contrast-enhanced MR imaging. Radiology 1993;189:251–257.[Abstract/Free Full Text]
- Cook TA, Rahim N, Simpson HC, Galland RB. Magnetic resonance imaging in the management of diabetic foot infection. Br J Surg 1996;83:245–248.[CrossRef][Medline]
- Craig JG, Amin MB, Wu K, et al. Osteomyelitis of the diabetic foot: MR imaging-pathologic correlation. Radiology 1997;203:849–855.[Abstract/Free Full Text]
- Beltran J, Campanini DS, Knight C, McCalla M. The diabetic foot: magnetic resonance imaging evaluation. Skeletal Radiol 1990;19:37–41.[Medline]
- Gold RH, Tong DJ, Crim JR, Seeger LL. Imaging the diabetic foot. Skeletal Radiol 1995;24:563–571.[Medline]
- Moore TE, Yuh WT, Kathol MH el-Khoury GY, Corson JD. Abnormalities of the foot in patients with diabetes mellitus: findings on MR imaging. AJR Am J Roentgenol 1991;157:813–816.[Abstract/Free Full Text]
- Levine SE, Neagle CE, Esterhai JL, Wright DG, Dalinka MK. Magnetic resonance imaging for the diagnosis of osteomyelitis in the diabetic patient with a foot ulcer. Foot Ankle Int 1994;15:151–156.[Medline]
- Seabold JE, Flickinger FW, Kao SC, et al. Indium-111-leukocyte/technetium-99m-MDP bone and magnetic resonance imaging: difficulty of diagnosing osteomyelitis in patients with neuropathic osteoarthropathy. J Nucl Med 1990;31:549–556.[Abstract/Free Full Text]
- Yuh WT, Corson JD, Baraniewski HM, et al. Osteomyelitis of the foot in diabetic patients: evaluation with plain film, 99mTc-MDP bone scintigraphy, and MR imaging. AJR Am J Roentgenol 1989;152:795–800.[Abstract/Free Full Text]
- Weinstein D, Wang A, Chambers R, Stewart CA, Motz HA. Evaluation of magnetic resonance imaging in the diagnosis of osteomyelitis in diabetic foot infections. Foot Ankle 1993;14:18–22.[Medline]
- Morrison WB, Ledermann HP. Work-up of the diabetic foot. Radiol Clin North Am 2002;40:1171–1192.[CrossRef][Medline]
- Sella EJ, Barrette C. Staging of Charcot neuroarthropathy along the medial column of the foot in the diabetic patient. J Foot Ankle Surg 1999;38:34–40.[Medline]
- Marcus CD, Ladam-Marcus VJ, Leone J, Malgrange D, Bonnet-Gausserand FM, Menanteau BP. MR imaging of osteomyelitis and neuropathic osteoarthropathy in the feet of diabetics. RadioGraphics 1996;16:1337–1348.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
C. C. Roberts, W. B. Morrison, and P. T. Liu
Imaging Evaluation of Foot and Ankle Pathology: Self-Assessment Module
Am. J. Roentgenol.,
March 1, 2008;
190(3_Supplement):
S18 - S22.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P L Tan and J Teh
MRI of the diabetic foot: differentiation of infection from neuropathic change
Br. J. Radiol.,
November 1, 2007;
80(959):
939 - 948.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Kapoor, S. Page, M. LaValley, D. R. Gale, and D. T. Felson
Magnetic Resonance Imaging for Diagnosing Foot Osteomyelitis: A Meta-analysis
Arch Intern Med,
January 22, 2007;
167(2):
125 - 132.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
