KEY POINTS
- This post hoc secondary analysis compared two prospective multicenter spine SBRT cohorts treated using different high-dose target definitions. The segment-based cohort included 57 patients with 63 vertebral metastases, while the GTV-based cohort included 177 patients with 197 metastases, giving 234 patients and 260 lesions overall. The source trials were conducted between 2012–2015 and 2016–2023, respectively; patients were not randomized between the two target-volume strategies.
- Both approaches used simultaneous integrated boosts with elective lower-dose vertebral coverage. Segment-based high-dose PTVs encompassed the GTV and adjacent vertebral compartments at risk of microscopic spread, while GTV-based high-dose PTVs were generated using a 2-mm expansion of the GTV. Prescription schedules included 35–40 Gy in five fractions or 48.5 Gy in ten fractions to the high-dose volume, with 20 Gy in five fractions or 30 Gy in ten fractions to the remaining elective vertebral volume. Median high-dose PTV decreased from 58 cm³ with segment-based planning to 11 cm³ with GTV-based planning (p<0.001).
- Kaplan–Meier estimates favored GTV-based boosting for freedom from local progression. At two years, freedom from local progression was 95.3% with GTV-based planning versus 86.4% with segment-based planning (p<0.001). However, when death was treated as a competing event, two-year cumulative local failure incidence was 2.7% versus 3.3% (p=0.42), respectively. This discrepancy is important: the study does not establish that the smaller high-dose volume independently improves local control.
- Vertebral compression fracture estimates were lower in the GTV-based cohort. At two years, Kaplan–Meier estimates of progressive or new vertebral compression fracture were 10.1% versus 34.6% (p<0.001) for GTV-based versus segment-based treatment. On univariable Cox analysis, segment-based planning was associated with greater fracture hazard (HR 5.21, 95% CI 2.53–10.72; p<0.001). Yet competing-risk cumulative fracture incidence was 4.5% versus 5.7% (p=0.37), substantially reducing confidence in the apparent difference when deaths were accounted for.
- The cohorts differed substantially in baseline fracture risk and disease characteristics. Baseline vertebral fractures were present in 24% of segment-based lesions versus 5% of GTV-based lesions, osteolytic metastases in 71% versus 43%, and epidural involvement in 70% versus 21%. The later GTV-based cohort also had better performance status, more oligometastatic disease, and no prior spine surgery, compared with 16 previously operated patients in the segment-based cohort. These imbalances could account for part of the observed outcome differences.
- Severe neurological toxicity was uncommon in both cohorts. No radiation-induced myelopathy or plexopathy was reported during available follow-up, and no grade ≥4 late adverse events occurred. Two patients in the segment-based cohort (3.5%) and three in the GTV-based cohort (1.7%) experienced grade 3 pain. Three segment-based patients required surgery for vertebral compression fractures, whereas no GTV-based patients underwent fracture-related surgery during the reported observation period.
- The post hoc comparison cannot establish superiority of one targeting strategy. The trials used different eligibility criteria, dose schedules, treatment eras, and patient populations, and local progression and fractures were assessed without centralized radiological review. The large divergence between Kaplan–Meier probabilities and competing-risk cumulative incidence further complicates interpretation, particularly in patients with substantial mortality. Prospective randomized comparison would be required to determine whether smaller boost volumes independently reduce vertebral fractures without compromising disease control.
CLINICAL TAKEAWAY
Restricting high-dose spine SBRT escalation to the GTV while maintaining lower-dose elective vertebral coverage was associated with favorable observed local control and fewer fractures in a historical prospective-cohort comparison. However, substantial baseline imbalances and nonsignificant competing-risk comparisons prevent concluding that target-volume reduction itself caused these differences. The approach is promising for dose-escalation planning but requires more rigorous comparative validation before replacing established contouring recommendations.
SOURCE
International Journal of Radiation Oncology, Biology, Physics