Paik, Eugene MD; Wissman, Robert D. MD
Department of Radiology, University of Cincinnati, OH.
Address correspondence and reprint requests to Eugene Paik, MD, Department of Radiology, University of Cincinnati, 234 Goodman Street, Cincinnati, OH 45219. E-mail: firstname.lastname@example.org.
Abstract: Osteomyelitis and neuropathic arthropathy are common debilitating complications of the foot in diabetic patients. As foot ulcers are widely common in diabetic patients, their presence cannot reliably distinguish osteomyelitis from neuropathic arthropathy from a clinical perspective. Here, we discuss the role of magnetic resonance imaging in distinguishing the imaging characteristics of osteomyelitis and the neuropathic joint.
Complications of the foot frequently plague patients with diabetes. Among the more common include osteomyelitis and neuropathic osteoarthropathy. Distinguishing these 2 entities clinically is often difficult and frustrating for the attending clinician. Similarly, diagnostic imaging without an organized approach can potentially lead to nonspecific findings and subsequent confusion. Here, we discuss the role of magnetic resonance imaging (MRI) with a general approach to imaging patterns in distinguishing these two entities.
A 46-year-old posttraumatic paraplegic man with diabetes presented to the emergency department with a 3-week history of intermittent fever, decubitus ischial ulcers, and a left heel ulcer. Purulent drainage had been noted from several of the decubitus ulcers. Although he eventually underwent surgical debridement of the decubitus ulcers, management of his left heel ulcer remained in question. Radiographs of the left foot demonstrated nonspecific findings.
On MRI, bright T2 signal changes were noted in the soft tissues of the heel with a soft tissue defect, consistent with an ulcer and cellulitis. In addition, bright T2 signal changes (Figs. 1A and B) with low T1 signal changes (Fig. 1C) were demonstrated in the calcaneal bone marrow. These calcaneal signal changes were immediately subjacent to the ulcer, a pattern consistent with osteomyelitis.
A 60-year-old diabetic man presented to the emergency department with erythema and swelling of the left foot. He had recently been treated with antibiotics for cellulitis but had been noncompliant. Radiographic examination revealed a fracture of the second metatarsal, despite no history of trauma. Admission followed to evaluate for and treat a possible osteomyelitis.
In addition to the fracture of the second metatarsal, MRI demonstrated high T2 signal changes (Fig. 2A) and low T1 signal changes (Fig. 2B) involving the entire midfoot and hindfoot. No sinus tracts, ulcers, or fluid collections were present. These findings were consistent with neuropathic osteoarthropathy.
Patients with diabetes mellitus are prone to a myriad of common and potentially debilitating complications of the foot, including neuropathy, vascular disease, and infection. Foot ulcers frequently develop in areas of increased pressure which are often handicapped by poor circulation and neuropathy which can lead to poor wound healing. Ulcerations occur with an annual incidence of 2% to 3% in the diabetic population and lead to more than 50,000 amputations a year.1
Diabetic neuropathy, the most common complication of diabetes, results in decreased perception of even minor foot trauma.2,3 In the setting of repetitive microtrauma and decreased sensation, one is prone to the development of neuropathic osteoarthropathy or Charcot arthropathy. Features include instability, dislocation/subluxation, destruction, and deformity.4
Given the frequency of foot ulcers and concomitant neuropathy, the clinician is often faced with a diagnostic dilemma. Here, we discuss the MRI characteristics of osteomyelitis and neuropathic arthropathy of the foot, both common problems in the diabetic patient.
MRI is highly sensitive in characterizing the diabetic foot and has greater specificity and spatial resolution than bone scintigraphy.5 A concise detailed history is often of great significance to the radiologist in guiding the imaging protocol. For the purposes of imaging, the foot is typically divided into forefoot and hindfoot. This is important because coil selection as well as protocols may vary depending on the location of pathology. Therefore, the exact location of pathology, particularly the location of ulcers, is most helpful.
MRI allows for multiplanar acquisitions typically obtained in axial, sagittal, and coronal planes. T1-weighted images (T1WI), in which fat is bright, provides anatomic detail and can demonstrate marrow changes. Fat-suppressed T2-weighted images (T2WI), in which water is bright, are sensitive for edema within bone and soft tissues.5 Contrast infusion is most helpful in the detection of soft tissue changes such as abscesses and sinus tracks but also increases both sensitivity and specificity in the detection of infection.6
In contradistinction to typical types of nondiabetic osteomyelitis, the most common cause of diabetic osteomyelitis is direct extension from adjacent skin and soft tissue infection rather than hematogenous spread.7 Typical MRI characteristics of osteomyelitis include low bone marrow signal on T1WI (Figs. 1C and 3C) and high bone marrow signal on T2WI (Figs. 1A, 2A, and 3B). These findings, which are consistent with edema, are in and of themselves nonspecific. Such findings can be seen in many abnormalities, most important of which is reactive bone marrow edema from neuropathic arthropathy. Differentiating the 2 entities solely on signal characteristics can be difficult; however, certain guidelines are helpful. Osteomyelitis tends to have more abnormal signal changes on T1WI than reactive bone marrow edema and may show intense increased signal following contrast administration (Fig. 3A). Osteomyelitis tends to occur at pressure points such as the toes, metatarsophalangeal (MTP) joints, and calcaneus. Overlying soft tissue changes such as abscess and sinus tracks as well as periosteal reaction are secondary findings that help increase the accuracy of MRI.
Unlike the septic joint, neuropathic arthropathy is characterized by involvement of multiple joints as the disease results in regional instability.3 Erythema and swelling are common clinical presentations shared by osteomyelitis. From an imaging standpoint, the neuropathic joint can share many features of osteomyelitis.6 The location of the abnormalities may guide one to distinguishing between the 2 entities. Neuropathic arthropathy most commonly affects the midfoot, an uncommon location for osteomyelitis (Figs. 1A and B). The involved joint(s) are more likely to be away from any cutaneous findings.
The clinical and radiological distinction between osteomyelitis and neuropathic arthropathy in the diabetic foot is commonly problematic. Although several of the MRI characteristics of these 2 entities can be shared, osteomyelitis is most commonly caused by direct extension of infection, and the identification of overlying ulcers and sinus tracts is paramount. Reactive bone marrow changes will exhibit the same T2 signal abnormalities but will generally demonstrate less T1 signal abnormalities. In addition, the location of changes, whether involvement of the midfoot is observed or there is involvement of the typical pressure points of the foot, is important and useful in distinguishing features.
1. Culleton JL. Preventing diabetic foot complications. Postgrad Med. 1999;106:74-84.
2. Vinik AI. Management of neuropathy and foot problems in diabetic patients. Clin Cornerstone. 2003;5(2):38-55.
3. Morrison WB. Work-up of the diabetic foot. Radiol Clin North Am. 2002;40(5):1171-1192.
4. Jones EA, Manaster BJ, May DA, et al. Neuropathic osteoarthropathy: diagnostic dilemmas and differential diagnosis. Radiographics. 2000;20: S279-S293.
5. Rosenberg ZS, Beltran J, Bencardino JT. MR imaging of the ankle and foot. Radiographics. 2000;20:S153-S179.
6. Chatha DS, Cunningham PM, Schweitzer ME. MR imaging of the diabetic foot: diagnostic challenges. Radiol Clin North Am. 2005;43:747-759.
7. Schweitzer ME, Morrison WB. MR imaging of the diabetic foot. Radiol Clin North Am. 2004;42:61-71.
© 2005 Lippincott Williams & Wilkins, Inc.