Diagnosing Osteomyelitis in Patients With Diabetic Neuropathic Osteoarthropathy

Approximately 15% of diabetic people will develop foot ulcers during their lifetime, and early detection of osteomyelitis is crucial to the management of diabetic foot ulcers (1). Differentiating osteomyelitis from neuropathic osteoarthropathy is clinically difficult, as the symptoms and signs are nonspecific. These patients all present with hot and erythematous feet. At presentation, there is often no change on plain radiographs (2). Many of the imaging findings are also similar, especially in rapidly progressing, noninfected neuro-osteoarthropathy. The most reliable method of establishing infection is to analyze microbiological samples of the lesion. However, this is not always practical and may lead to seeding of the infection or damage to the area biopsied. Magnetic resonance imaging (MRI) is a useful method of tissue localization and is currently the most sensitive method to detect osteomyelitis (3).

Technetium (Tc)-99 m Infecton consists of ciprofloxacin linked to Tc99 m. The antibiotic is taken up and bound specifically by living bacteria, where it inactivates DNA gyrase. As the antibiotic is chelated with 99 Tc, the area of bacterial infection should be identifiable during imaging (4).

A total of 16 diabetic patients with a hot swollen foot were studied prospectively using plain radiographs, MRI, Galium-67, and Tc99 m Infecton. The MRI and plain radiographs were reported independently, blinded from the radionuclide imaging, and vice versa. The definitive diagnosis was established by findings at surgery, microbiological results, or definitive imaging (e.g., plain radiograph to detect fractures).

In our prospective study, four (25%) patients had osteomyelitis, three (19%) had neuropathic fractures, and nine (56%) had soft tissue swelling. MRI accurately diagnosed all of the four cases with osteomyelitis. Tc99 m Infecton was only able to localize infection to bone in one of the four cases with osteomyelitis. In the rest of the cases, Infecton could not differentiate whether infection was confined to soft tissue or bone. Plain radiographs were able to diagnose two of the four cases with osteomyelitis. MRI correctly diagnosed fractures in all of the three patients who had evidence of fractures on plain radiograph. Infecton and Gallium scans reported bone or soft tissue as infected in all of the three cases. Therefore, the nuclear medicine scans can falsely indicate infection or inflammation in the presence of fractures.

Radionuclide imaging is not reliable to differentiate among infection, inflammation around fractures, or Charcot joint, even when infection is correctly identified. The limited spatial resolution in the forefoot does not allow accurate discrimination between soft tissue infection and osteomyelitis. Plain radiograph was essential in the initial work-up, as hot spots on Infecton scans and Gallium 67 scans can indicate fracture rather than infection. MRI is the imaging of choice to distinguish osteomyelitis from other conditions, such as cellulitis and neuropathic osteoarthropathy in diabetic patients with a hot swollen foot. Infecton scans are helpful when used in conjunction with MRI to localize an infected area before surgery but cannot be used independently as a diagnostic tool in the assessment of a hot swollen diabetic foot.