Proton PBS reduces low-dose lung exposure but raises robustness concerns in central SBRT

In ten central lung tumors, proton PBS reduced low-dose lung exposure versus VMAT, but some critical-organ doses were less robust.

KEY POINTS

  • This retrospective Japanese planning study evaluated ten patients with centrally located early-stage or oligometastatic lung tumors, defined as lesions within 2 cm of the proximal bronchial tree. Three treatment approaches were compared: proton pencil beam scanning (PBS) using multi-field optimization (MFO), PBS using single-field optimization (SFO), and photon volumetric modulated arc therapy (VMAT). All plans prescribed 60 Gy(RBE) or 60 Gy in eight fractions, normalized to an 80% prescription isodose line relative to the maximum dose, allowing higher intratumoral dose while maintaining the prescribed target margin dose.
  • The proton planning strategy incorporated robust optimization and respiratory motion management. Four individually selected coplanar proton fields were used with ±5-mm setup and ±3.5% range uncertainty, four-layer repainting, and respiratory-gated imaging at maximum exhalation. The ITV incorporated a 1-mm expansion around the GTV, increased to 6 mm inferiorly, followed by a 5-mm PTV expansion. The prescription was normalized to PTVeval D95%, and proton plans were required to achieve GTV maximum dose >75 Gy(RBE). A constant RBE of 1.1 was assumed.
  • MFO provided the highest near-minimum gross tumor dose. Mean GTV D98% was 65.2 ± 5.1 Gy(RBE) with MFO, compared with 63.4 ± 5.4 Gy(RBE) using SFO and 62.7 ± 3.9 Gy using VMAT. MFO outperformed SFO (p<0.01) and VMAT (p<0.05). Proton plans also achieved higher average GTV doses than VMAT, although VMAT demonstrated better PTVeval conformity and dose-gradient indices, indicating that greater intratumoral dose escalation did not translate into universally superior plan quality.
  • Lung sparing differed markedly by dose range. Both proton approaches significantly reduced normal-lung volumes receiving 5–15 Gy compared with VMAT (p<0.05). However, VMAT achieved significantly lower V25Gy–V60Gy than either PBS strategy. These competing advantages demonstrate that the reduction in low-dose exposure obtained through proton therapy was accompanied by greater intermediate- and high-dose exposure in selected lung regions, preventing a simple conclusion that one technique was uniformly superior.
  • Mean OAR doses favored PBS, but near-maximum doses revealed important limitations. Both proton approaches significantly reduced mean doses to the evaluated thoracic OARs compared with VMAT (p<0.05), including several critical structures adjacent to central tumors. Nevertheless, PBS produced higher chest-wall near-maximum dose metrics, while MFO increased skin dose. In one patient, the 64-Gy proximal bronchial tree maximum-dose reference constraint was exceeded with all three techniques. Contemporary bronchial tree and esophageal maximum-dose criteria were retrospectively evaluated rather than explicitly incorporated into the original planning objectives.
  • Target robustness was generally acceptable, but proton OAR robustness was less favorable. Under worst-case perturbations, MFO showed the smallest GTV D98% deviation, significantly outperforming SFO (p<0.05) but not VMAT. Among perturbed scenarios for plans with nominal GTV D98% above 60 Gy, coverage fell below prescription in 29.6% with MFO, 34.3% with SFO, and 33.3% with VMAT. For several normal tissues, PBS produced larger worst-case dose increases than VMAT, with significant differences involving seven OARs for MFO and five for SFO, emphasizing that nominal dose reduction does not guarantee improved robustness.
  • The study establishes dosimetric feasibility rather than clinical superiority. Only ten retrospectively selected patients were included, all using the same 80% prescription isodose level. Patient-specific respiratory interplay effects and dynamic four-dimensional accumulated dose were not explicitly evaluated, and treatment delivery efficiency was not assessed. Differences in planning systems, optimization objectives, and fixed proton RBE assumptions further limit direct translation. No comparative local control, pneumonitis, airway injury, or other clinical outcomes were available.

CLINICAL TAKEAWAY

Isodose-prescribed PBS SBRT can reduce low-dose lung exposure and mean dose to several critical thoracic organs while maintaining high intratumoral dose in central lung tumors. However, VMAT provided better sparing at higher lung doses, and PBS demonstrated greater uncertainty-related variation for selected OARs. Robust, patient-specific four-dimensional evaluation and contemporary serial-organ constraints are essential before applying this strategy clinically.

SOURCE

International Journal of Particle Therapy

Browse more research Suggest a correction