PSMA- and FDG-PET mismatch assessment for optimized selection of PSMA radioligand therapy candidates

Ziel/Aim Patients with metastatic castration resistant prostate cancer routinely undergo PSMA-PET to evaluate the efficacy of PSMA targeting radioligand therapy (Lu-PSMA). Despite the presence of PSMA expressing metastases, PSMA-negative but FDG-positive metastases are associated with worse response to Lu-PSMA therapy and shorter overall survival. Current assessment is based on visual criteria with reduced reproducibility. Reproducible analysis of PSMA or FDG positive tumor volume is needed for an optimized patient selection. To this end, we used a semi-automated approach to quantify the metabolic tumor volume and mismatch in PSMA- and FDG-PET aquisitions prior to Lu-PSMA therapy.

Methodik/Methods Patients who were included in this retrospective analysis received both PSMA- and FDG-PET CTs to evaluate eligibility for Lu-PSMA therapy. The metabolic tumor volume was quantified both in the PSMA and FDG PET scan using the semi-automatic software MICIIS (Siemens), that assists in the exclusion of physiological uptake by neural networks. All metastases were segmented and labelled with the anatomical location.

Ergebnisse/Results A total of 44 patients were included in this analysis. 14 (32%) of these patients had a relevant PSMA- FDG+ mismatch which led to an exclusion from a Lu-PSMA therapy, whereas 30 (68%) patients showed no mismatch and received Lu-PSMA therapy as planned. Overall survival was significantly longer for patients without PSMA- FDG+ mismatch versus mismatch (median 15,1 (CI95: 7,5-22,7) vs. 6,8 (Ci95: 2,0-11,6) months; p = 0,027).

Schlussfolgerungen/Conclusions The semi-automated delineation of PSMA positive or FDG positive metastases is feasible and easy-to-use for PSMA-negative/FDG-positive mismatch quantification. Organ-wise quantification of PSMA and FDG tumor volume offers the potential for more detailed patient stratification. Further studies and analyses are needed to determine association with oncologic outcome.

Publication History

Article published online:
08 April 2021