KEY POINTS
- This prospective single-centre study included 108 patients with non-metastatic prostate cancer treated between July 2023 and July 2025. Median age was 74 years; 62% had intermediate-risk disease, 14% had high-risk disease, and 35% received androgen-deprivation therapy.
- Patients received online adaptive treatment on a 0.35-T MR-linac using a uniform 2-mm prostate PTV margin. Prescriptions were 36.25 Gy in five fractions for low-risk disease, 36.25–40 Gy for intermediate-risk disease, and 40 Gy for high- or very-high-risk disease.
- A new MRI and adaptive plan were generated before every fraction, with treatment delivered on alternate days. Cine MRI monitored motion at four frames per second, and the beam stopped when more than 5% of the prostate moved outside the 2-mm gating boundary.
- Grade ≥2 genitourinary toxicity occurred in 22% at two weeks, but declined to 3% at three months and 3% at six months. Grade ≥2 gastrointestinal toxicity was 1% at two weeks and 0% at three and six months.
- One grade 3 urinary-retention event was reported at 12 months; no grade 3 gastrointestinal event occurred. Rectal spacers were not routinely used, although the absence of a non-adaptive comparison prevents attribution of toxicity outcomes specifically to daily adaptation.
- Higher baseline urinary symptom burden independently predicted grade ≥2 genitourinary toxicity: each one-point increase in IPSS increased the odds by approximately 9% (OR 1.09, 95% CI 1.01–1.18; p=0.02). Prostate volume above 50 cm³ and dose level were not independently associated with toxicity.
- At six months, clinically relevant deterioration occurred in 9% for IPSS, 21% for urinary irritative symptoms, and 12.5% for bowel function. Median PSA declined from 8.56 ng/mL to 1.03 ng/mL at six months and 0.76 ng/mL at 12 months, with no biochemical failure reported during the first year.
CLINICAL TAKEAWAY
Daily adaptive MR-guided prostate SBRT with 2-mm margins produced limited persistent urinary morbidity and almost no clinically significant gastrointestinal toxicity in this prospective series. The results support feasibility but do not prove that adaptation adds benefit beyond MRI guidance, gating, and margin reduction.