Diagnosing spinal osteomyelitis: a comparison of bone and Ga-67 scintigraphy and magnetic resonance imaging.

PURPOSE: The objective of this investigation was to compare the accuracies of bone and Ga-67 scintigraphy and magnetic resonance imaging (MRI) for diagnosing spinal osteomyelitis and to determine the optimal radionuclide approach to this disorder.
METHODS: Twenty-two patients, with 24 sites of possible spinal osteomyelitis, who underwent three-phase bone scintigraphy with SPECT, Ga-67 scintigraphy with SPECT, and MRI with and without contrast were included in this retrospective review. Bone scans were interpreted as three-phase studies, delayed planar images alone, delayed planar plus SPECT, and SPECT alone (to identify uptake patterns). Sequential bone/ Ga-67 images were interpreted as planar and as SPECT studies. Planar and SPECT Ga-67 images were also interpreted alone. Precontrast MRI studies were used to identify osteomyelitis, whereas postcontrast images were used to identify soft tissue infection.
RESULTS: Eleven sites of spinal osteomyelitis were identified. Tracer uptake in two contiguous vertebrae, as noted on SPECT, was the most accurate bone scan criterion for detecting spinal osteomyelitis (71 %). SPECT bone/Ga-67 was significantly more accurate (92%) than both planar bone/Ga-67 (75%) and bone SPECT (P = 0.15 and P = 0.2, respectively). SPECT Ga-67 was as accurate as SPECT bone/Ga-67 and as sensitive as MRI (91 %); the radionuclide study was slightly but not significantly more specific (92% vs. 77%) than MRI. Of 11 sites of extraosseous infection, 10 were identified on MRI, 9 on SPECT Ga-67, 7 on planar Ga-67, and none on bone scintigraphy.
CONCLUSIONS: Spinal osteomyelitis and accompanying soft tissue infection can be diagnosed accurately with a single radionuclide procedure: SPECT Ga-67. This procedure can be used as a reliable alternative when MRI cannot be performed and as an adjunct in patients in whom the diagnosis is uncertain.