Zachary Ells, Linus Hempel, Julia Ells, Muyu Wu, Maximilian Scheifele, Magdalena Schoell, Sophie C. Siegmund, Mathias Zacherl, Rudolf A. Werner, Jeremie Calais, Magnus Dahlbom, Guido Böning, Adrian Holzgreve, Astrid Delker, Lena M. Unterrainer
- Introduction
Administration activities of 177Lu-PSMA were chosen cautiously in the treatment of advanced prostate cancer. The primary aim was to investigate the association between increased injected activities, and the calculated absorbed doses (AD).
Methods
In this single center, retrospective analysis, patients with PSMA-avid metastatic castration-resistant prostate cancer treated with [177Lu]Lu-PSMA-I&T were included. Patients underwent multiple timepoint SPECT/CT imaging at 24, 48, and 72 h post-injection. Patients included were injected with 7.4, or 9.1 GBq. Segmentations were completed manually for the kidneys (CT-based), while a 3.0 SUV threshold-based approach was used for lesions. Dosimetry calculations were then performed using a multiple-timepoint protocol. Manual verification of the individual segmentations coregistered between timepoints was completed. All time activity data was fit using a mono-exponential function. Time-integrated activity was then convolved withIntroduction
Administration activities of 177Lu-PSMA were chosen cautiously in the treatment of advanced prostate cancer. The primary aim was to investigate the association between increased injected activities, and the calculated absorbed doses (AD).
Methods
In this single center, retrospective analysis, patients with PSMA-avid metastatic castration-resistant prostate cancer treated with [177Lu]Lu-PSMA-I&T were included. Patients underwent multiple timepoint SPECT/CT imaging at 24, 48, and 72 h post-injection. Patients included were injected with 7.4, or 9.1 GBq. Segmentations were completed manually for the kidneys (CT-based), while a 3.0 SUV threshold-based approach was used for lesions. Dosimetry calculations were then performed using a multiple-timepoint protocol. Manual verification of the individual segmentations coregistered between timepoints was completed. All time activity data was fit using a mono-exponential function. Time-integrated activity was then convolved with voxel-s-value using a predefined 177Lu kernel resulting the AD in Gy. No further partial volume corrections were made.
Results
76 patients treated were included: 22 and 54 were injected with 9.1 ± 0.3 and 7.4 ± 0.3 GBq respectively. The AD/GBq to the kidney was comparable, 0.29 ± 0.08 vs. 0.33 ± 0.13 Gy/GBq [p=0.193] (9.1 vs. 7.4 GBq). The tumor AD/GBq was comparable between 9.1 and 7.4 GBq cohorts: total tumor burden (0.95 ± 0.42 vs. 0.92 ± 0.52 Gy/GBq) [p=0.302]; bone lesions 0.78 ± 0.42 vs. 0.74 ± 0.46 Gy/GBq [p=<0.001]; hottest lesion in each patient 1.37 ± 0.61 vs. 1.30 ± 0.74 [p=0.421]; coldest lesion in each patient 0.40 ± 0.16 vs. 0.50 ± 0.32 Gy/GBq [p=0.930].
Conclusion
The AD/GBq across the disease burden, individual lesions, and kidney remained comparable even with increased injected activity. No evidence of the tumor sink effect was found. No evidence of a clinically relevant tumor sink effect in patients with <2000 mL tumor volume was found but patients with substantially high TTV (> 2 L) lacks.…
