Maarten van der Doelen

General discussion and future perspectives

no OS benefit was seen in patients harboring DDR alterations compared to DDR proficient patients. Thus, although studies were small, the reason for the heterogenicity in response to PSMA-targeting RLT is still largely unclear. Therefore, until better biomarkers are available, pretherapeutic clinicopathologic and imaging parameters, such as ECOG performance status, baseline ALP, LDH and PSA levels, immunohistochemical PSMA expression, PSMA uptake (avidity and heterogeneity) on PSMA-targeting PET/CT, the extent and localization of metastases and prior chemotherapy or RLT remain best predictive and prognostic markers for patients receiving PSMA-targeting RLT. (17, 20, 78-81) In the near future, large prospective cohort studies including translational research are warranted to explore (circulating) molecular biomarkers and their association with the differential responses to PSMA-targeting RLT and the OS of patients. FUTURE PERSPECTIVES Radiological evaluation prior to initiation of radium-223 therapy Novel imaging techniques such as 68 Ga-PSMA PET/CT show higher sensitivity than conventional imaging with CT for the detection of minute visceral metastases and non osteoblasticbone lesions. Recent study results stress that PSMA-targetingPET/CTmaybe more useful than CT alone to rule out extraskeletal disease and bone marrow infiltration before radium-223 therapy initiation, and commonly leads to a change in therapeutic management. (82, 83) Furthermore, according to a small retrospective cohort study, the implementation of baseline PSMA-targeting PET in addition to routine staging using CT may result in a higher percentage of patients with PSA decline during radium-223 therapy. In this study, PSA declines correlated with reduction of PSMA expression of target lesions on follow-up PSMA-targeting PET/CT scans. (83) The additional value of the acquisition of a baseline PSMA PET/CT over CT will become clearer in the near future. Another key point of attention is the high incidence of spinal cord compression during radium-223 in post-registration reports; in the ALSYMPCA trial only 4% of patients were diagnosed with spinal cord compression during the treatment period, while in real world series it is reported to be up to 16%. (84-86) This difference might be explained by the stricter patient management and radiological assessment in a trial population whereas in daily practice this appears to be overlooked. Because patients with baseline epidural disease frequently progress to spinal cord compression, this will lead to early

8

211