KEY POINTS
- Ten kidney-tumor patients previously treated with five-fraction proton therapy were retrospectively replanned for photon VMAT. Their 50 daily CBCTs were converted to virtual CT datasets, producing 80 simulated treatment plans.
- All photon plans prescribed 50 Gy in 5 fractions to 99% of the internal CTV. Investigators compared no added margin, uniform 0.5-mm and 1.0-mm margins, and fully reoptimized daily adaptive plans.
- With no additional margin, daily anatomy caused D99 to fall below the 10-Gy prescription in 30/50 fractions (60%). Prescription-level D99 was maintained in 92% with a 0.5-mm margin, 98% with a 1-mm margin and 100% with daily adaptation.
- The magnitude of the actual underdose was small: the worst single-fraction D99 deficit without a margin was only 2.2%, with a minimum D99 of 978 cGy versus the prescribed 1000 cGy. All strategies maintained D95 at or above prescription.
- Adding margins increased normal kidney exposure. Statistically significant differences were observed in healthy-kidney V40, V50 and mean dose, whereas lower dose-volume metrics showed smaller and more variable differences.
- Kidney V20 increased by >10% in 4/10 patients, reaching a maximum 22.8% increase, when the adaptive strategy was compared with nonadaptive planning using a 1-mm margin, suggesting that OAR benefit may be much more patient-specific than target-coverage benefit.
- Generalizability is limited: 9/10 tumors were posterior, virtual CTs relied on deformable registration, and no adaptive photon plans were actually delivered. The study also did not model all conventional setup, motion or delivery uncertainties.
CLINICAL TAKEAWAY
For kidney SABR in this cohort, daily anatomy alone caused only small target-dose perturbations, and a tiny margin reproduced most of the coverage benefit of online adaptation. The more compelling role for adaptation may therefore be selective OAR sparing—particularly for bowel- or kidney-constrained anatomy—rather than routine rescue of target coverage.