Abstract
1788
Objectives: SPECT/CT is increasingly used for quantification of tumor activity implementing tracer uptake parameters originally developed and tested for PET, such as the standardized uptake value. Goal of this project was to apply quantitative SPECT/CT to imaging studies of neuroendocrine tumors with In-111 octreotide and investigate the effect of imaging parameters on estimated tumor activity (measured as SUVpeak) and detectability (measured as CNR).
Methods: Twenty subjects were imaged using a Siemens/Symbia T16 dual headed camera. Of these, 13 subjects demonstrated tumor uptake of In-111 octreotide in either a primary tumor or metastasis (In+ subjects). The dual head SPECT/CT camera was calibrated using a NIST-standardized selenium-75 point source. Images were obtained in list mode and reconstructed in 128x128 and 256x256 matrix with the conjugate gradient technique using one subset, three updates (24, 36, 48) and three smoothing FWHM (8, 10, 12 mm). Peak standardized uptake values (SUVpeak) and contrast to noise ratios (CNR) were obtained using spherical volumes of interest (5 mL for lesions, 30 mL for background and liver). Dependence of SUVpeak and CNR on imaging time post injection, reconstruction parameters, tumor presence, patient age and gender was tested by multivariate regression analysis. For lesion characterization the SUVpeak was obtained at a combination of matrix size, update number and filter parameter which resulted in maximal SUVpeak x CNR product. Results were reported as median values. Results: The liver SUVpeak values depended on time post injection, age and gender but not on tumor presence, reconstruction matrix, updates and filter. The following median values were measured in the liver: SUVbw=1.97; SUVlbw=1.23; SUVbsa=0.98. The following SUVpeak median (first quartile, third quartile) values were obtained in tumors: a) metastases in the chest (n=8): 2.71 (1.37, 12.39), b) primary or metastatic lesions in abdomen excluding liver (n=10): 19.9 (9.56, 33.97); metastases in the liver (n=9): 8.15 (5.99, 16.32). Pooled SUVpeak showed significant positive regression on time after injection and number of updates but no regression on age, gender, filter and matrix. On an individual bases SUVpeak showed significant regression on matrix size (p=4.66E-02) and particularly on updates (p=1.01E-93) and filter (p=6.52E-35). Time post injection demonstrated significant effect on pooled CNR.Conclusion: Quantification of In-111 octreotide uptake with SPECT is feasible in both tumor tissues and in liver as reference organ. Time after injection and number of updates affect the pooled SUVpeak value while pooled CNR depends on time post injection. On an individual basis SUVpeak also strongly depends on matrix size, updates and filter. The presented preliminary data lay the groundwork for further investigation of the role of quantitative In-111 octreotide SPECT studies. Potential clinical applications include lesion characterization and measurement of treatment response.