LeukoScan for imaging infection in different clinical settings: a retrospective evaluation and extended review of the literature.

PURPOSE

The aim of the current study was to determine the overall diagnostic accuracy of Tc-99m-labeled antigranulocyte monoclonal antibody Fab' fragments (LeukoScan) for the routine detection of bone and soft tissue infections in a retrospective evaluation.

PATIENTS AND METHODS

138 patients (63 men, 75 women; mean age, 58.29 +/- 25.38 years) with fever of unknown origin and possible endocarditis (n = 59), infection of arthroplastic joints (n = 20), arthritis (n = 16), peripheral (n = 15) and central bone infections (n = 14), soft tissue infection (n = 6), appendicitis (n = 4), pericarditis (n = 2), or vascular graft infection (n = 2) underwent imaging after injection of 555 to 925 MBq (15 to 25 mCi) Tc-99m-labeled antigranulocyte monoclonal antibody Fab' fragments (LeukoScan).

RESULTS

True-positive results were found in 63 of 81 lesions. The overall sensitivity and specificity were 76% and 84%, respectively. In arthritis, seven of seven foci could be detected, whereas false-negative results were found in infections of the femoral bone in three of nine lesions and in periprosthetic infections of long bones in three of eight lesions. Good results were found in five of six soft-tissue infections, in four of six patients with endocarditis, in three of four atypical cases of appendicitis, in two of two infected vascular grafts, and in one of one patient with pericarditis. Subacute and chronic infections of the spine always showed photopenic areas in eight of eight patients. If photopenic lesions were included as diagnostic criteria, the sensitivity and specificity were 88% and 67%, respectively.

CONCLUSIONS

Tc-99m-labeled antigranulocyte monoclonal antibody Fab' fragments can be used for imaging acute infections of peripheral bones and soft tissues. False-negative results are likely in patients with chronic infections. Sensitivity can be increased while decreasing specificity by including photopenic lesions in the spine as diagnostic criteria for localizing disease.