Skeletal Radiol (2002) 31:202–207 DOI 10.1007/s00256-001-0466-8

Leopoldo M. Gigena Christine B. Chung Nittaya Lektrakul Christian W. A. Pfirrmann Mi Sook Sung Donald Resnick

Received: 26 May 2001 Revised: 23 August 2001 Accepted: 6 November 2001 Published online: 7 February 2002 © ISS 2002

L.M. Gigena · C.B. Chung · N. Lektrakul C.W.A. Pfirrmann · M.S. Sung D. Resnick (✉) Department of Radiology, Veterans Affairs San Diego Healthcare Systems, University of California, San Diego, 3350 La Jolla Village Drive, San Diego, CA 92161, USA e-mail: [email protected] Tel.: +1-858-5528585 ext 3343 Fax: +1-858-5527565 D. Resnick Osteoradiology Section (114), Department of Radiology, Veterans Affairs San Diego Healthcare System, 3350 La Jolla Village Drive, San Diego, CA 92161, USA

A RT I C L E

Transient bone marrow edema of the talus: MR imaging findings in five patients

Abstract Objective: To describe the MR findings of transient bone marrow edema (TBME) of the talus and to address the differential diagnostic considerations. Design and patients: The imaging findings of TBME of six tali were retrospectively reviewed in five patients with a clinical history of pain without trauma. Inclusion criteria were MR imaging findings that, when compared with clinical data and results of follow-up assessment, allowed the diagnosis of TBME. MR imaging, standard radiography, and bone scintigraphy were performed. The images were reviewed with particular attention to the pattern and distribution of abnormal marrow signal intensity as well as associated findings. Results: In four cases the entire talus was involved, and in two cases only a portion of the bone was affected. No

Introduction Transient bone marrow edema (TBME) is an uncommon self-limited condition of unknown etiology that is characterized clinically by pain without a history of trauma. This entity can be difficult to distinguish from osteonecrosis, stress fracture, reflex sympathetic dystrophy syndrome (RSDS), infection, or tumor. The most commonly involved sites are the bones about the hip, followed in frequency by those about the knee, ankle, and foot. The affected bone often, but not invariably, shows regional osteopenia with conventional radiography and increased radionuclide uptake. The MR imaging findings are those of nonspecific marrow edema with a pattern of decreased

fractures were detected. MR imaging demonstrated diffuse decreased signal intensity of the marrow on T1weighted images with corresponding increased signal intensity on T2weighted images. In all six cases MR imaging detected associated findings, which included joint effusion and soft tissue edema. All patients improved clinically with conservative therapy over a period of 6 months to 1 year. Conclusions: Although unusual, TBME can involve the talus. Marrow edema without evidence of a fracture and in the absence of history of trauma is a characteristic MR imaging feature, allowing confident diagnosis and institution of conservative therapy. Keywords Transient bone marrow edema · Talus · MRI

signal intensity on T1-weighted images and increased signal intensity on T2-weighted images. As the name of this disorder implies, marrow edema is transient and resolves on follow-up imaging examinations. Early recognition of this syndrome is difficult because it is a diagnosis of exclusion. This entity must be distinguished from other causes of marrow abnormality such as osteonecrosis, stress fracture, RSDS, infection, and tumor. As the talus is a rare site of TBME syndrome, with few reported cases in the literature, our purpose is to describe the MR imaging findings of TBME in this site. In addition, the imaging evaluation and means to establish the diagnosis of TBME syndrome are revisited, emphasizing the importance of clinical and imaging follow-up.

203

Table 1 Summary of the MR imaging features of six tali with transient marrow edema Case no.

Age (years)/Sex

Side

Talus marrow edemaa Head

Neck

Body

Joint effusionb

Soft tissuesc

Ankle and foot bones involved

Other joints involved

Calcaneus, navicular, cuboid, metatarsal Calcaneus, distal tibia N/A N/A Calcaneus, navicular, cuboid, distal tibia Calcaneus, cuboid, navicular, metatarsal

Contralateral talus

1

51/M

L

X

+/–

+/–

++

++

2 3 4 5

56/M 70/M 61/M 71/F

R L R L

X X +/– X

X X X X

X X X X

++ ++ ++ ++

+++ N/A ++ ++

6

38/F

R

X

X

X

++

+

Contralateral talus N/A N/A Both knees and hips Knee, contralateral ankle

a (X) total involvement, (+/–) partial involvement, (–) no involvement b (++) moderate effusion

c (+) mild increased signal intensity, (++) moderate increased signal intensity, (+++) high increased signal intensity N/A, not applicable

Material and methods

Results

MR imaging examinations of the foot and ankle with bone marrow edema derived from four hospitals over an 8-year period were retrospectively reviewed by consensus of two musculoskeletal radiologists. Five patients, including three men (one with bilateral involvement) and two women who ranged in age from 38 to 71 years (mean age 54.5 years), were found to have findings consistent with TBME of the talus. Inclusion criteria for the study included diffuse or focal marrow edema of the talus as established on MR imaging, with clinical or imaging follow-up examinations to document resolution of this finding. Exclusion criteria included the presence of a fracture line, marrow signal abnormality centered around an articulation suggesting arthritis, or evidence of loss of cortical margin or joint space that might indicate infection. Medical records were reviewed to obtain patient demographics and document the clinical course of the disorder. Due to the retrospective, multi-institution nature of the study, some variability in the MR imaging protocol existed. MR imaging studies in five of the six cases were performed on a 1.5-T magnet (Signa, General Electric Medical System, Milwaukee, Wis.), with a single study performed on a 0.5-T magnet (Gyroscan T5 NT, Philips, Best, The Netherlands). Protocols, although varied, consisted of T1- and T2-weighted spin echo sequences and a short inversion time recovery (STIR) sequence, as well as intermediateand T2-weighted fat-suppressed fast spin echo sequences, supplemented in one case with a T1-weighted fat-suppressed spin echo sequence obtained after intravenous gadolinium administration. The field of view varied between 12 and 26 cm, slice thickness ranged from 3 to 6 mm, and the interslice gap was 1 mm. The number of acquisitions was 1 or 2. The imaging matrix ranged from 256×128 to 256×256. The images of these five patients were analyzed by the consensus of two musculoskeletal radiologists. The following parameters were evaluated with MR imaging: (1) anatomic distribution of marrow edema in the talus (head, neck and/or body); (2) patterns of abnormal bone marrow edema signal intensity (diffuse versus focal; heterogeneous versus homogeneous); (3) integrity of surrounding structures such as bones, ligaments, and tendons; and (4) associated findings including other sites of osseous involvement, joint effusion, and soft tissue edema. Clinical follow-up evaluation to complete resolution of symptoms was available in all patients. Follow-up imaging studies were available in three cases and included MR imaging (n=2), conventional radiography (n=3), and bone scintigraphy (n=1).

The results are summarized in Table 1. Presenting symptoms in all five patients included pain and variable soft tissue swelling in the foot and ankle for a 6- to 9-month period with no history of trauma or repetitive sporting or occupational activity. In two cases, patients had previous episodes of TBME in locations other than the talus or the bones about the hip or knee. Initial imaging examination consisted of MR imaging (n=6), conventional radiography (n=4), and bone scintigraphy (n=4). MR imaging studies demonstrated abnormal marrow signal intensity of the talus in all cases consistent with marrow edema (Figs. 1, 2, 3). In three cases, the signal abnormality was diffuse, involving the entire talus. In the remaining three cases the signal abnormality was focal, involving only a portion of the talus. MR imaging characteristics of marrow edema consisted uniformly of decreased signal intensity on T1-weighted images and increased signal intensity on T2-weighted images. The signal abnormality was best demonstrated on the STIR sequences. MR imaging studies showed a diffuse homogeneous pattern of signal abnormality in the talus marrow in four patients and a diffuse heterogeneous pattern in two patients. Associated MR imaging findings included marrow edema in adjacent osseous structures in four cases (Fig. 1), specifically in the calcaneus (n=4), navicular (n=3), and cuboid (n=3) bones, distal tibia (n=2), and metatarsal bones (n=2). All the patients presented with joint effusion. Five cases had evidence of soft tissue edema. No signs of fractures (acute or stress-related) or injuries of ligaments, tendons, or muscles were detected. Conventional radiography revealed diffuse osteopenia in four cases. Bone scintigraphy (n=4) demonstrated intense increased uptake of the radiopharmaceutical agent

204

205

Fig. 2A, B The same patient as in Fig 1, 4 years later, with pain and swelling in his right ankle and foot. A Sagittal intermediateweighted (TR/TE 3316/45) fat saturation fast spin echo image. Note the diffuse marrow edema involving the entire talus (straight arrow). This image also shows moderate effusion in the posterior aspect of joint (small white arrows) and soft tissue edema in the anterior aspect of the ankle (curved arrow). B Lateral bone scintigraphy. Increased uptake of radiopharmaceutical is seen in the right ankle (straight arrow). Note that normal uptake is demonstrated in the left ankle (curved arrow)

in the involved bones. Conventional radiography and bone scintigraphy were less useful than MR imaging in demonstrating the different portions of the talus that were affected. Follow-up MR imaging examinations were available in two cases and showed gradual resolution of the marrow edema, joint effusion, and soft tissue edema in both. Osteopenia also resolved on subsequent conventional ra-

diographs during a 1-year period in all cases. A biopsy of the talus was performed in one patient. Results indicated normal bone and marrow tissue without signs of malignancy, infection, or osteonecrosis. Three patients presented with TBME in other osseous structures simultaneously with or after involvement of the talus, suggesting a migratory nature of the process. In all cases, osseous structures of the lower extremities were affected. One patient had TBME about the left knee 1 year prior to presentation of talar TBME and also had edema about the right hip and knee at the same time as involvement of the talus. A second patient had involvement about the bones of the right knee, and the bones of contralateral ankle and foot were involved 6 months later. A third patient had involvement of the other talus 4 years prior to presentation (Fig. 2). All patients were treated conservatively with resolution of symptoms.

▲

Discussion Fig. 1A–F A 51-year-old man with pain and swelling in his left ankle and foot. A Sagittal T1-weighted (TR/TE 400/20) spin echo MR image. Diffuse bone marrow edema involving the distal aspect of the talus (arrow) is demonstrated. B Axial intermediateweighted (TR/TE 2500/40) fat saturation fast spin echo MR image. Note the diffuse bone marrow edema in the talus (straight white arrow), moderate joint effusion (curved black arrow), and soft tissue edema (curved white arrow). C Lateral bone scintigraphy image of the ankle and foot. Diffuse uptake of radiopharmaceutical in the left ankle and foot is noted, most pronounced in the region of the talus (white arrow). D Sagittal intermediate-weighted (TR/TE 2600/40) fat saturation fast spin echo MR image 3 months after the original study. Marrow edema is now seen diffusely in the anterior aspect of the talus (long white arrow), navicular (curved black arrow), cuneiform (open black arrow), metatarsal bone (solid black arrow), and calcaneus (short white arrow). Note the moderate joint effusion (white arrowhead). E, F Sagittal T1-weighted (TR/TE 400/17) spin echo (E); and sagittal T2weighted (TR/TE 5400/56) spin echo (F) MR images performed 1 year after the original study; incomplete resolution of the bone marrow edema and soft tissue edema are noted

In 1959 Curtis and Kincaid [1] described a syndrome characterized by transitory osteopenia of the hip in pregnancy. Shortly thereafter, [2] the term transient osteoporosis (TO) was introduced to describe a self-limited syndrome characterized by pain and loss of radiodensity followed by spontaneous recovery. Not long after [3, 4,5], a migratory and recurrent pattern of TO was described, designated regional migratory osteoporosis (RMO). Subsequently, it was suggested that rather than being distinct entities, RMO and TO likely represent examples of the same spectrum of disease [6, 7]. Wilson et al [8] in 1988 reported MR imaging characteristics and clinical findings of TO about the bones in the hip and the knee. The MR imaging findings in their study consisted of subtle regional decreased bone marrow signal intensity on T1-weighted images and a corresponding increased signal intensity on T2-weighted im-

206

Fig. 3A, B A 61-year-old man who presented with pain and swelling in his right ankle and foot. A Sagittal T1-weighted (TR/TE 366/16) spin echo MR image. It shows diffuse bone marrow edema involving the proximal and medial aspect of the talus (white arrow). B Sagittal intermediate-weighted (TR/TE 3000/27) fat saturation fast spin echo MR image. Diffuse marrow edema in the talus (straight arrow), moderate joint effusion (curved arrow), and soft tissue edema (open arrow) are demonstrated

ages. Because of the MR imaging characteristics, Wilson and associates concluded that the findings represent edema in the marrow. In addition, three of their patients did not have osteoporosis when material derived from bone biopsies was examined histologically. For this reason, they suggested that the term TO be replaced by TBME. TBME has been defined as a clinical syndrome of unknown etiology characterized by pain about one or more joints with spontaneous resolution after a period of time.

It is well recognized in the femoral head and the bones about the knee, but TBME of the talus has received little attention. There is still controversy about the cause and pathophysiology of TBME [9]. It has been suggested that the development of edema of the marrow is related to increased intramedullary pressure, which may be secondary to disturbed venous outflow rather than an interrupted arterial supply [10, 11]. Microtrauma has also been suggested as a possible cause of TBME. In addition, an association with pregnancy [1, 11] and disorders such as cirrhosis [12] and type IV hyperlipoproteinemia [13] and with electromyographic abnormality [3] has been reported. There are several entities to be distinguished from TBME. The combined clinical, radiographic, and MR imaging findings are different from those of avascular necrosis (AVN), stress fractures, osteomyelitis, infections, arthritis, neoplasm, and RSDS. AVN, RSDS, and fractures emerge as the entities most difficult to distinguish from TBME. Although AVN of the talus has been described in the absence of trauma [14, 15, 16], it has a characteristic clinical progression with focal and permanent bone sclerosis and collapse as radiographic findings, which differ markedly from those present in TBME. RSDS presents with characteristic clinical findings: skin changes, cutaneous vasomotor changes (hyperthermia), a history of trauma or surgery, muscle atrophy, and permanent disability. Further, RSDS is frequently confined to the upper extremities. One study [17] of 17 patients with proven RSDS showed no MR imaging abnormalities. These clinical and radiographic findings are not consistent with the diagnosis of TBME. Algodystrophy is a term used more often in the European than North American literature to describe a condition similar to RSDS with prominent autonomic instability and a frequent history of trauma [7]. Fractures represent another differential diagnostic consideration. The clinical history of injury is clearly important. Similarly, although stress fractures are not often accompanied by a history of trauma, their findings can be distinguished from TBME. On MR imaging, stress fractures manifest as irregular linear bands that have low signal intensity with all pulse sequences and variable surrounding bone marrow edema [18, 19]. Routine laboratory results in patients with TBME are usually normal, although a slightly elevated erythrocyte sedimentation rate has been reported [2, 3, 6]. When positive, conventional radiographs typically show osteopenia in the affected bone that becomes apparent within weeks or months after the onset of clinical manifestations. With both computed tomography (CT) and conventional radiography, however, a large amount of trabecular bone must be lost before osteopenia is recognized, which decreases their utility for early diagnosis [4, 20]. Bone scintigraphy is reported to demonstrate ab-

207

normal uptake of the radionuclide early in the course of the disease [5]. In our study, MR imaging findings were identical to those reported by Wilson et al. [8], consisting of regional subtle decreased bone marrow signal intensity on T1weighted images and a corresponding increased signal intensity on T2-weighted images. In addition, in our study, STIR images showed high sensitivity in the detection of the marrow edema, joint effusions, and soft tissue edema. One limitation of this study is the small number of patients, likely reflecting the uncommon localization of TBME in the talus. Second, owing to the nature of the disease process, the diagnosis in our patients was based on initial and follow-up clinical and imaging examinations. Histologic findings, available in one patient, sup-

ported the diagnosis of TBME. Third, the retrospective multi-institution nature of the study precluded control over specific imaging protocols. Finally, we did not include proven cases of stress fracture of the talus (although they are rare) that might have enabled us to determine specific MR imaging features allowing their earlier differentiation from TBME. We believe, however, that the documented MR imaging findings are most compatible with the diagnosis of TBME. Furthermore the migratory nature of the process noted in our study supports the concept that RMO and TO represent a spectrum of disease. The presence of associated findings such as joint effusion and soft tissue edema appears to represent a part of the TBME syndrome and should not dissuade the radiologist from making the correct diagnosis.

References 1. Curtis PH Jr, Kincaid WE. Transitory demineralization of the hip in pregnancy. J Bone Joint Surg Am 1959; 41:1327–1333. 2. Hunder GG, Kelly PJ. Roentgenologic transient osteoporosis of the hip. A clinical syndrome? Ann Intern Med 1968; 68:539–552. 3. Tannenbaum H, Esdaile J, Rosenthal L. Joint imaging in regional migratory osteoporosis. J Rheumatol 1980; 7:237–244. 4. O’Mara RE, Pinals RS. Bone scanning in regional migratory osteoporosis: case report. Radiology 1970; 97:579– 581. 5. Bray ST, Partain CL, Teates DC, Guilford WB, Williamson BRJ, McLaughlin RC. The value of the bone scan in idiopathic regional migratory osteoporosis. J Nucl Med 1979; 20:1268–1271. 6. Lakhanpal S, Ginsburg WW, Luthra HS, Hunder GG. Transient regional osteoporosis. Ann Intern Med 1987; 106:444–450.

7. Naides SJ, Resnick D, Zuaifler N. Idiopathic regional osteoporosis; a clinical spectrum. J Rheumatol 1985; 12:763– 768. 8. Wilson AJ, Murphy WA, Hardly DC, Totty WG. Transient osteoporosis: transient bone marrow edema? Radiology 1988; 167:757–760. 9. Hofmann S, Engel A, Neuhold A, Leder K, Kramer J, Plenk H Jr. Bone marrow edema syndrome and transient osteoporosis of the hip. J Bone Joint Surg Br 1993; 75:210–216. 10. Rutishauer E, Rhoner A, Held D. Experimentelle Untersuchungen über die Wirkung der Ischämie auf den Knochen und das Mark. Virchows Arch [Pathol Anat] 1960; 333:101–118. 11. Beaulieu JG, Razzano CD, Levine RB. Transient osteoporosis of the hip in pregnancy: review of the literature and a case report. Clin Orthop 1976; 115:165–168. 12. Rozenbaum M, Zinman C, Nagler Am Pollak S. Transient osteoporosis of hip joint with liver cirrhosis. J Rheumatol 1984; 11:241–243. 13. Pinals RS, Jabbs JM. Type-IV hyperlipoproteinemia and transient osteoporosis. Lancet 1972; II:929.

14. Schmidt DM, Romash MM. Atraumatic avascular necrosis of the head of the talus: a case report. Foot Ankle 1988; 8:4. 15. Harris RD, Silver RA. Atraumatic aseptic necrosis of the talus. Radiology 1973; 106:81. 16. Delanois RE, Mont MA, Yoon TR, Mizell M, Hungerford DS. Atraumatic osteonecrosis of the talus. J Bone Joint Surg Am 1998; 80:529–536. 17. Koch E, Hofer HO, Sialer G, Marincek B, von Schulthess GK. Failure of MR imaging to detect reflex sympathetic dystrophy of the extremities. AJR Am J Roentgenol 1991; 156:113–155. 18. Lee JK, Yao L. Stress fractures: MR imaging. Radiology 1988; 169:217– 220. 19. Umans H, Pavlov H. Insufficiency fractures of the talus: diagnosis with MR imaging. Radiology 1995; 197:439–442. 20. Calvo E, Alvarez L, FernandezYruegas D, Vallejo C. Transient osteoporosis of the foot. Acta Orthop Scand 1997; 68:577–580.