Performance evaluation and quantitative accuracy of a multi-pinhole small animal SPECT/CT scanner for theranostic Lu-177 imaging

Objectives: SPECT plays a fundamental role in targeted radionuclide therapy using theranostic radionuclides like Lu-177. However, both pre- and post-therapy SPECT studies must meet special requirements regarding the quantitative accuracy of generated images because the reliable assessment of tumor uptake and tumor-to-normal tissue ratios can only be performed using quantitatively accurate images. Hence, it is important to evaluate performance parameters and quantitative accuracy of preclinical SPECT systems for therapeutic radioisotopes before conducting pre- and post-therapy SPECT imaging or dosimetry studies. Phantom studies were performed by different researchers to evaluate the performance of multi-pinhole NanoSPECT camera using diagnostic radioisotopes, such as Tc-99m. However, no such study has been reported using therapeutic radionuclides. In this study, we performed phantom experiments to characterize the tomographic performance and quantitative accuracy of NanoSPECT/CT scanner for Lu-177 and compared with Tc-99m.

Methods: A mouse aperture (APT1, Mouse STD-WB) of NanoSPECT/CT camera (Bioscan, Inc.) having nine pinholes of 1.4 mm diameter was used for this study. The Lu-177 and Tc-99m point sources were prepared using glass capillary tube (internal diameter of 1.1-1.2 mm). We measured the reconstructed SPECT spatial resolution (FWHM) and system sensitivity (cps/MBq) for both Lu-177 and Tc-99m point sources positioned at the center and other locations (at every 2mm from the center) along the axial and transaxial FOV of NanoSPECT camera. The FWHM was also measured after adding uniform background to the raw projection images of point source before reconstruction. The uniform background was added in such a way that the ratio of the reconstructed pixel intensity of point to background was 10:1. We also measured the reconstructed SPECT uniformity and calibration factor of this system using uniform phantoms containing Lu-177 and Tc-99m sources. The measured spatial resolution (with and without adding uniform background), sensitivity and uniformity of the system for Lu-177 and Tc-99m were compared.

Results: The FWHM values at the center of FOV in axial, radial and tangential direction were measured to be 0.98, 0.95 and 1.16 mm for Lu-177 and 1, 1.03 and 1.07 mm for Tc-99m respectively. However, the spatial resolution for both radionuclides degraded when uniform background was added to the point image. FWHM values with background in axial, radial and tangential direction increased up to 14%, 23% and 31% for Lu-177 and 7%, 22% and 22% for Tc-99m, respectively. The sensitivity measured at the center of axial and radial FOV was 1,425 and 1,489 cps/MBq for Tc-99m and 389 and 401 cps/MBq for Lu-177, respectively. The decreasing trends in sensitivity were also observed when sources were moved either side from the center of FOV. Reconstructed SPECT uniformity was poor for Lu-177 (21.19% SD) when compared with Tc-99m (11.11% SD). The measured calibration factor of the system for Lu-177 source was 10.85 Bq/counts. Conclusion: The reconstructed spatial resolution of mouse aperture for Lu-177 was comparable to that of Tc-99m however, it degraded when uniform background was added. The measured sensitivity and uniformity of the system were low for Lu-177 source when compared to Tc-99m for mouse imaging.