Scan speed reduction in post-therapy [177Lu]LuPSMA SPECT/CT: How fast can we go without afecting accuracy of dosimetry calculations?

Aim/Introduction: Accurate dose estimation in [177Lu]LuPSMA therapy is crucial for organs at risk such as the kidneys. However, the current method involves lengthy scans, causing
discomfort for patients with bone metastases common in prostate cancer, fnancial pressure and image quality problems due to the motion. Our research aims to optimise this process by
investigating acceptable scan reduction while maintaining the accuracy of the calculate dose to organs at risk. Materials and
Methods: A total of 11 post-therapy imaging sequences from patients undergoing [177Lu]Lu-PSMA therapy were selected. For each, two post-therapy SPECT/CT scans, performed at 1- and
7-days post-administration of 7.4 GBq of [177Lu]Lu-PSMA were used. The SPECT projection data was acquired using a dual-head camera with a 3/8” NaI detector, medium-energy collimator, an energy window centred at 208 keV, and a lower scatter window, 64 projections per detector, 14 seconds per projection, covering two feld of views (FOVs). To simulate shorter scan durations, the original projections were downgraded to 75%, 50%, and 25% of the counts using Hermes Hybrid reconstruction software. Subsequently, each set of projections was reconstructed using Monte Carlo-based OSEM algorithms, incorporating calibration
factors for 177Lu quantifcation. Kidney dosimetry was performed using a mono-exponential model using the ICRP-103 phantom with the Olinda 2.0. The resulting absorbed kidney doses, mean activity and noise level in the kidneys was calculated for all cases.
Results: The relative diference in the calculated absorbed dose for kidneys, when compared to the initial acquisition, was -1.2±3.6, 0.5±3.0, and 0.6±3.7%, for 75%, 50%, and 25% scans, respectively. Activity concentration values in the kidney have changed by an average of 2.0±2.5, 2.4±3.1 and 3.5±4.4%, for 75%, 50% and 25% scans on day 1, and 3.3±6.7, 6.8±11.2, 7.4±12.0%, for day 7 scans. At the same time, the noise propagation of reduced scans on
day 7 (9.1, 25.7 and 63.85% increase for scans with 75%, 50% and 25% of counts) was signifcantly higher than that of day 1 (2.1, 6.4 and 8.9%).*with respect of the original scan with 64 projections/ detector, 14s/projection scan ** average ± standard deviation for
11 post-therapy imaging sequences
Conclusion: Our fndings demonstrate the feasibility of reducing the scan duration of
post-therapy [177Lu]Lu-PSMA scans to 3.7 (25%) and 11.2 (75%) minutes per bed position for day 1 and day 7, respectively, without signifcantly compromising the accuracy and uncertainty of the post-therapy dose calculation for kidneys.