KEY POINTS
- This prospective cohort included 31 patients with biochemical recurrence after radical prostatectomy, ECOG performance status 0–1, no distant metastases, and no previous pelvic radiotherapy.
- All patients received MR-guided online adaptive radiotherapy using a 0.35-T MR-linac. The prostate bed received 32.5 Gy in five fractions; 13 patients (41.9%) also received 25 Gy to elective pelvic nodes, and three (9.7%) received a 35-Gy simultaneous boost to macroscopic disease.
- Daily adaptation used 2-mm margins for the prostate bed and gross disease and 4-mm margins for pelvic nodes, with cine-MRI gating based on a rectal surrogate. Median room time was 30 minutes per fraction, and median overall treatment duration was nine days.
- At a median follow-up of 8.8 months, no grade ≥3 treatment-related adverse events occurred. Grade ≥2 genitourinary toxicity was 3.3% at treatment completion and 4.3% at six months, while no grade ≥2 gastrointestinal toxicity was reported.
- EPIC-26 urinary scores remained stable from 79.6 at baseline to 81.5 at six months, and median bowel scores remained 100 at every assessment. Median IPSS increased modestly from three to six, without severe urinary symptoms.
- Median PSA declined from 0.31 ng/mL before treatment to 0.02 ng/mL at six months and 0.04 ng/mL at 12 months among 11 evaluable patients. All 31 patients met the study’s biochemical-response definition, but 77.4% received androgen-deprivation therapy, substantially confounding interpretation.
- Median planned prostate-bed V95% was only 84.3%, ranging from 62.5% to 98.2%, because organ-at-risk constraints were allowed to take priority. The small sample, heterogeneous nodal and systemic treatment, absent control group, and short follow-up prevent assessment of late safety or comparative efficacy.
CLINICAL TAKEAWAY
Five-fraction adaptive MR-guided salvage treatment appears technically feasible and was well tolerated during early follow-up, even when pelvic nodes were included. Post-prostatectomy SBRT remains investigational, and the relatively low target coverage and lack of mature late-toxicity data require particular caution.