Diagnostic value of quantitative SPECT/CT integrated parameters for differentiating spinal tuberculosis from spinal metastases.

OBJECTIVES

To evaluate the diagnostic value of quantitative 99mTc-MDP SPECT/CT metabolic, CT attenuation, and adjacent vertebra integrated parameters for differentiating spinal tuberculosis (STB) from spinal metastases (SM).

MATERIALS AND

METHODS

This retrospective study included 212 patients who underwent quantitative 99mTc-MDP SPECT/CT, including 94 with STB and 118 with SM. Volumes of interest (VOIs) were delineated on fused SPECT/CT images, from which standardized uptake value (SUV), total bone uptake (TBU), metabolic bone volume (MBV), and CT Hounsfield unit (HU) parameters were obtained. One target metastatic vertebra was analyzed in each SM patient, whereas paired adjacent involved vertebrae were used to construct integrated parameters in STB patients. Diagnostic performance was assessed using ROC analysis, multivariable logistic regression, DeLong's test, and decision curve analysis.

RESULTS

SUVmax, SUVmean, metabolic bone volume, tracer bone uptake, CT HUmax, and CT HUmean were significantly higher in SM target vertebrae than in STB involved vertebrae, all p < 0.001. Among single integrated parameters, the lesion burden index based on SUVmax (LBISUVmax) showed the best performance, with an AUC of 0.854, sensitivity of 95.7%, and specificity of 73.7%. The AUCs of the metabolic, CT attenuation, and combined models were 0.938, 0.810, and 0.943, respectively. DeLong's test showed that the combined model outperformed the CT attenuation model, LBISUVmax, but not the metabolic model, p = 0.422.

CONCLUSION

Quantitative 99mTc-MDP SPECT/CT provides objective information for noninvasive differentiation of STB and SM, and metabolic integrated parameters may offer useful diagnostic value.