Intensity of 18F-FDG PET Uptake in Culture-Negative and Culture-Positive Cases of Chronic Osteomyelitis

Abstract

Microbiologic cultures are not infrequently negative in patients with a histopathologic diagnosis of chronic osteomyelitis. Culture-negative cases may represent low-grade infections with a lower metabolic activity than culture-positive cases. ¹⁸F-FDG PET could potentially detect such a difference. We determined whether the level of ¹⁸F-FDG PET uptake differs in patients with culture-negative and culture-positive osteomyelitis. We reviewed the clinical charts of 40 consecutive patients, who had diagnostic ¹⁸F-FDG PET for a suspected bone infection. Twenty-six patients were eligible with a confirmed diagnosis based on microbiologic cultures and/or histopathologic examination. Sixteen of 26 patients had chronic osteomyelitis. Eight of them had positive cultures, seven had negative cultures, and one patient had no cultures of the biopsy specimen. The patients with histologically and/or microbiologically proven osteomyelitis were correctly interpreted as true positive in the routine clinical reading of ¹⁸F-FDG PET images. There was no relationship between the level of ¹⁸F-FDG PET uptake and the presence of positive or negative bacterial cultures. The result favors the concept that that culture-negative cases of osteomyelitis are false-negative infections due to nonculturable microbes. ¹⁸F-FDG PET may help to confirm the presence of metabolically active infection in these patients and guide their appropriate treatment.

Figure 1

Comparison of SUV_mean (a), SUV_max (b), SUV_mean ratio (c), and SUV_max ratio (d) values measured in ¹⁸F-FDG PET imaging of osteomyelitis patients. The differences between culture-negative (n = 7) and culture-positive (n = 8) cases were not statistically significant. Box plots are showing median, 1st and 3rd quartiles, minimum and maximum values, and outliers (white circles and asterisks).

Figure 2

A 25-year-old man with an indolent Brodie's abscess in the proximal tibia (case #8). The patient had been hospitalized for knee pain 10 years earlier, but no specific diagnosis was made. He now suffered a sports related ACL ligament rupture of his left knee. As an incidental finding, anterior-posterior and lateral radiographs (a) showed cystic lesion with surrounding sclerosis in the proximal tibia. Coronal and transaxial MR-images (b) demonstrated a 2 cm sclerotic osseous lesion with contrast medium enhancement and oedema of the surrounding tissues. Coronal and transaxial ¹⁸F-FDG PET images (c) showed an increased local uptake of the tracer. Compared with the corresponding ROI of the contralateral tibia, SUV_mean ratio was 10.1 and SUV_max ratio 20.51. The lesion was correctly characterized with infection scintigraphy with labeled antibody fragments (LeukoScan) (d) and three-phase bone scintigraphy (e). Based on percutaneous biopsy samples taken under fluoroscopy, the final histological diagnosis was Brodie's abscess and the microbiologic culture revealed S. aureus as the causative pathogen.