Initial multicentre experience of ⁶⁸ gallium-PSMA PET/CT guided robot-assisted salvage lymphadenectomy: acceptable safety profile but oncological benefit appears limited

Objectives: To evaluate the safety and short-term oncological outcomes of ⁶⁸ gallium-labelled prostate-specific membrane antigen (⁶⁸ Ga-PSMA) positron-emission tomography (PET)/computed tomography (CT)-directed robot-assisted salvage node dissection (RASND) for prostate cancer oligometastatic nodal recurrence.

Materials and methods: Between February 2014 and April 2016, 35 patients across two centres underwent RASND for ⁶⁸ Ga-PSMA PET/CT-detected oligometastatic nodal recurrence. RASND was performed using targeted pelvic dissection, unilateral extended pelvic template or bilateral extended pelvic template dissection, depending on previous pelvic treatment and extent/location of nodal disease. Complications were reported using the Clavien-Dindo classification system. Definitions of prostate-specific antigen (PSA) treatment response to RASND were defined as 6-week PSA <0.2 ng/mL (broad definition) or PSA <0.05 ng/mL (strict definition) in those who had undergone primary prostatectomy, and 6-week PSA level < post-radiotherapy nadir in those who had undergone primary radiotherapy. Biochemical recurrence (BCR) after RASND was defined as a PSA >0.2 ng/mL or PSA > nadir, for those who had undergone primary prostatectomy and primary radiotherapy, respectively. Predictors of treatment response were analysed using univariate binary logistic regression.

Results: A total of 58 lesions suspicious for lymph node metastases (LNM) in 35 patients were detected on ⁶⁸ Ga-PSMA imaging. A total of 32 patients (91%) had histopathologically proven LNM at RASND, with a total of 87 LNM and a median (interquartile range) of 2 (1-3) LNM per patient. In all, eight patients (23%) experienced complications, all Clavien-Dindo grade ≤2. Treatment response was seen in 15 (43%) and 11 patients (31%), using the broad and strict definitions, respectively. BCR-free survival and clinical recurrence-free survival at a median follow-up of 12 months were 23% and 66%, respectively, for the entire cohort. Bilateral template dissection was the only significant univariate predictor of treatment response in our cohort.

Conclusions: Although RASND appears safe and feasible, less than half of our cohort had a treatment response, and less than a quarter experienced BCR-free survival at 12-month median follow-up. ⁶⁸ Ga-PSMA imaging underestimates micro-metastatic disease, therefore RASND will rarely be curative. Strict patient selection and restricting RASND to clinical trials is recommended. Long-term follow-up from such trials is required to further assess potential quality of life and mortality benefits.