The present role of bone marrow scintigraphy.

Abstract

An experimental study in animals performed in 1973 by Ito et al. [1] showed that bone marrow scans applying radiocolloids reveal skeletal involvement by malignant tumours earlier than do planar X-rays and radionuclide bone scans. The authors stated that “radiocolloid marrow scanning appears to be one of the methods of choice in the early diagnosis of metastatic skeletal malignancies”. With the practical limitations of the method in mind, the authors called for “development of more excellent radiopharmaceuticals suitable for bone marrow imaging, improvement of resolution, and advance in the handling of scan data”. What happened to bone marrow scintigraphy (BMS) in the past 25 years regarding these three requirements demanded by Ito et al.? Radiocolloids initially used for imaging of the liver were adopted for BMS throughout the 1980s and also in the 1990s. Nanosized colloid proved to be superior to other colloids with regard to bone marrow/spleen and bone marrow/background ratios [2]. Immunoscintigraphy with a technetium-99m-labelled murine monoclonal IgG antibody against NCA-95 was introduced as a novel radiopharmaceutical with improved imaging characteristics in 1988 [3]. Use of this antibody, equally directed against granulopoietic bone marrow cells and granulocytes in the peripheral blood, resulted in a bone marrow uptake 2–4 times higher than that of microcolloid [4]. The image quality was clearly superior; uptake in overlying ribs and thoracic vertebrae was usually not significantly obscured by that of the liver and spleen [5]. 99mTclabelled monoclonal anti-NCA95 antibodies are now regarded as the radiopharmaceutical of choice for BMS [6]. A disadvantage of anti-NCA95 immuno-scintigraphy using complete IgG antibodies is the induction of human antimurine antibodies (HAMA) in approximately 5% of patients, resulting in altered biodistribution, increased hepatic uptake and non-diagnostic imaging in the case of repeated application [7]. 99mTc-labelled anti-NCA-90 immunoscintigraphy applying Fab’ fragments, now used for imaging of infection, is not suitable for bone marrow imaging, and a fragmented antibody which would combine high bone marrow uptake and low antigenicity is not yet available. Resolution of gamma cameras has not changed basically during the last two decades, although thinner crystals optimised for 99mTc as the leading radionuclide may have improved image quality slightly. The use of singlephoton emission tomography (SPET) has been shown to provide additional valuable information in bone imaging of the spine [8]. However, in BMS the use of SPET has not yet been exploited in larger studies, although optimal resolution and delineation of vertebral structures are as crucial as in bone scintigraphy. Sagittal reconstruction of the SPET data set may give the best overview of the spine and is easily comparable to magnetic resonance imaging (MRI) or fluorine-18 fluorodeoxyglucose positron emission tomography (FDG-PET) images. Volumeand surface-rendered images provide a 3D image of the functional bone marrow mass, and overlay techniques can be employed to fuse image data with those of MRI or FDG-PET.