A. Guhlmann et al., Fluorine-18-FDG PET and technetium-99m antigranulocyte antibody scintigraphy in chronic osteomyelitis, J NUCL MED, 39(12), 1998, pp. 2145-2152
Citations number
47
Categorie Soggetti
Radiology ,Nuclear Medicine & Imaging","Medical Research Diagnosis & Treatment
The aim of this study was to assess the usefulness of PET with 2-F-18-fluor
o-2-deoxy-D-glucose (FDG), as compared to immunoscintigraphy (IS) with Tc-9
9m-labeled monoclonal antigranulocyte antibodies (AGAbs), in the detection
of chronic osteomyelitis. Methods: Fifty-one patients suspected of having c
hronic osteomyelitis in the peripheral (n = 36) or central (n = 15) skeleto
n were evaluated prospectively with static FDG PET imaging and combined Tc-
99m-AGAb/Tc-99m-methylene diphosphonate (MDP) bone scanning within 5 days.
FDG PET and IS were evaluated in a blinded and independent manner by visual
interpretation, which was graded on a five-point scale of two observers' c
onfident diagnosis of osteomyelitis. Receiver operating characteristic (ROC
) curve analysis was performed for both imaging modalities. The final diagn
osis was established by means of bacteriologic culture of surgical specimen
s and histopathologic analysis (n = 31) or by biopsy and clinical follow-up
over 2 yr (n = 20). Results: Of 51 patients, 28 had osteomyelitis and 23 d
id not. According to the unanimous evaluation of both readers, FDG PET corr
ectly identified 27 of the 28 positives and 22 of the 23 negatives (IS iden
tified 15 of 28 positives and 17 of 23 negatives, respectively). The area u
nder the ROC curve was 0.97/0.97 (reader 1/reader 2) for FDG PET and 0.87/0
.90 for IS, with a high degree of interobserver concordance (kappa-values w
ere 0.96 for FDG PET and 0.91 for IS). In the central skeleton, the ROC cur
ve area was 0.98/1.00 for FDG PET and 0.71/0.77 for IS (p < 0.05). On the b
asis of ROC analysis, the overall accuracies of FDG PET and IS in the detec
tion of chronic osteomyelitis were 96%/96% and 82%/88%, respectively. With
regard to the optimal threshold values, sensitivity and specificity were 10
0%/97% and 95%/95% with FDG PET, compared to 86%/92% and 77%/82% with IS, r
espectively. Conclusion: In the peripheral skeleton, both FDG PET and combi
ned Tc-99m-AGAb/(TC)-T-99m-MDP scanning are appropriate imaging modalities
to diagnose chronic osteomyelitis. FDG PET additionally allows reliable dif
ferentiation between osteomyelitis and infection of the surrounding soft ti
ssue. In the central skeleton within active bone marrow, FDG PET is highly
accurate and superior to AGAb imaging in the diagnosis of chronic osteomyel
itis, which frequently presents as a nonspecific photopenic lesion at scint
igraphy with labeled white blood cells.