[⁹⁰Y]Y-PSMA-617 for the treatment of metastatic castration-resistant prostate cancer - Post-therapeutic kidney and bone marrow dosimetry for individualized therapy

Objectives: The therapy with lutetium-177 labeled PSMA ligands (e.g. PSMA-617) has been established as a promising treatment option for patients suffering from metastatic castration-resistant prostate cancer (mCRPC). As proposed for treatment of neuroendocrine tumors it is hypothetically assumed that larger lesions could benefit from enhanced crossfire effects due to the higher penetration depth in tissue of the β-particles emitted by yttrium-90 compared to the low-energy β-energy of lutetium-177. However, to date Y-90 has not been intensively used for PSMA targeted therapy and dosimetry is only based on estimates based on previously performed [¹⁷⁷Lu]Lu-PSMA-617 cycles [1]. As known, intra-therapeutic imaging based on bremsstrahlung for dosimetry purposes is a challenge On the other hand, dosimetry accompanying therapy is important for these patients, as they have been heavily pre-treated and it is important to protect the organs at risk (OAR) to the best of our ability. Therefore, this study aimed to establish an imaging protocol and workflow allowing post-therapeutic dosimetry for kidneys and bone marrow in [⁹⁰Y]Y-PSMA-617 therapy.

Methods: 17 patients (mean age_ 72 yr (range: 59-84yrs)) suffering from mCRPC without or insufficient response to several cycles of [¹⁷⁷Lu]Lu-PSMA-617 treatment were selected for [⁹⁰Y]Y-PSMA-617 therapy. All therapies were conducted in accordance with the German Medicines Law (AMG, §13[2b]). For dosimetry, each patient underwent SPECT/CT imaging of the upper abdomen at approximately 2h, 24h, 48h and 72h p.i., using a Siemens Symbia T6 equipped with HEGP collimators (energy window: 125keV±20%; matrix: 128x128; 64 projections, 35 sec each). Quantitative SPECT/CT reconstruction was performed using a 3D-OSEM algorithm (5i15s) with corrections for attenuation, and scatter and detector blurring using a fully Monte Carlo based collimator modeling (Hermes Hybrid Recon v.3.0) [2]; no additional post filtering has been applied. A special calibration factor adapted to the general kidney geometry was determined. The phantoms required for this purpose were produced using a rapid prototyping technology which has been already presented [3]. It was placed inside of a NEMA-2007 torso phantom. Absorbed doses (AD) of the kidneys and of tumor lesions within the field of view were calculated based on a voxel level dosimetry scheme (Hermes Voxel Dosimetry, v.1.0.1) [4]. VOIs based on CT were applied to the calculated dose map to extract AD for each kidney and tumor lesions. Red bone marrow dosimetry was performed based on blood sampling according to EANM guideline. The calculated AD were compared with results for [¹⁷⁷Lu]Lu-PSMA-617 therapy from our own department and other published studies.

Results: The therapy was well tolerated by all patients and no relevant severe complications have been observed. Mean administered activity was 4.23 ± 0.54 GBq. The mean AD in the kidneys was 2.1 Gy ± 1.0 Gy/GBq and 0.19 ± 0.16 Gy/GBq for the red bone marrow. For tumor lesions (19 osseous and 15 soft tissue lesions) the mean AD was 16.4 ± 7.8 Gy/GBq. Compared to [¹⁷⁷Lu]Lu-PSMA-617, AD of the kidney and bone marrow are about 2.8 and 5 times higher, respectively. The following table shows a comparison of results of our department and already published data:

Conclusions: Regardless of the great success of PSMA-targeted therapy, there are patients without response despite PSMA expression and/or who have large tumor lesions. [⁹⁰Y]Y-PSMA-617 offers an alternative. AD of tumor lesions are higher, but also OAR are more severely affected. The proposed dosimetry scheme seams feasible in clinical routine. However, the quantification of the Y-90-SPECT/CT has to be validated and discussed critically. References: [1] Rathke H et al. J Nucl Med Nov 2, 2018 [2] Porter et al. EJNMMI Research (2018) 8:7 [3] Kurth J et al. JNM. 2013;54(suppl 2):2174 [4] Hippeläinen E et al, Nucl Med Commun, 38 (5), 357-365. 2017