KEY POINTS
- This retrospective treatment-planning study from Japan evaluated 46 patients with non-small cell lung cancer (NSCLC) using paired 4DCT-derived ventilation and SPECT perfusion imaging. Forty-seven patients were initially considered, but one was excluded because no planning strategy met the predefined dose constraints. Four VMAT plans were generated for each patient: anatomical planning without functional imaging (A), ventilation-guided planning (V), perfusion-guided planning (Q), and combined ventilation–perfusion planning (VQ).
- All plans prescribed 60 Gy in 30 fractions, using two full VMAT arcs with 6-MV photons and Acuros XB dose calculation. Ventilation was estimated from deformable registration of respiratory-phase CT images, while perfusion was measured using technetium-99m macroaggregated albumin SPECT. Both functional datasets were converted into percentile maps, with the highest-functioning third of lung tissue designated as high ventilated lung (HVL) or high perfused lung (HPL). Plans were optimized to spare these regions while satisfying RTOG 0617-based constraints.
- Combined VQ planning achieved the lowest mean doses to high-functioning lung regions. Mean dose to HVL decreased from 12.4 Gy with anatomical planning to 11.5 Gy with VQ planning, an absolute reduction of 0.9 Gy. Mean dose to HPL decreased from 11.1 to 10.0 Gy, a reduction of 1.1 Gy. All three function-guided strategies significantly reduced high-functioning lung exposure compared with anatomical planning (p<0.001), with VQ providing the most consistent sparing across both functional compartments.
- Optimizing ventilation or perfusion alone did not protect both functional compartments equally. For HVL, ventilation-guided planning produced lower mean dose than perfusion-guided planning (11.7 vs 11.8 Gy; p=0.007) and lower V20Gy (19.1% vs 19.9%; p<0.001). Conversely, perfusion-guided planning better protected HPL than ventilation-guided planning (10.2 vs 10.5 Gy; p=0.005). Combined VQ plans achieved an additional mean-dose reduction of approximately 0.2–0.5 Gy relative to single-modality approaches, indicating modest but consistent complementary benefits.
- Whole-lung dosimetry improved, but functional optimization introduced trade-offs in other organs. Mean whole-lung V20Gy decreased from 22.9% with anatomical planning to 21.5% with VQ planning, while mean lung dose declined from 12.8 to 12.4 Gy (both p<0.001). However, heart V40Gy increased from 2.8% to 3.6% (p=0.001), and spinal cord maximum dose increased from 39.7 to 42.4 Gy (p<0.001). All plans remained within the predefined constraints, but target conformity also deteriorated slightly, with conformity index decreasing from 0.88 to 0.85.
- The benefit of functional avoidance depended more on tumor proximity to preserved lung than on target volume. For HVL, overlap with a 25-mm-expanded PTV correlated with reductions in mean dose (Spearman ρ=0.50) and V20Gy (ρ=0.66). For HPL, the strongest correlations occurred around 35–45 mm, including ρ=0.44 for V20Gy at a 35-mm expansion. In contrast, correlations between PTV volume and functional dose reduction were weak, with maximum absolute ρ=0.20. These exploratory findings suggest that the spatial relationship between tumor and functional lung may help identify patients most likely to benefit.
- Clinical benefit was not assessed. This was an in-silico comparison rather than a prospective treatment intervention, and no radiation pneumonitis, pulmonary function, quality-of-life, or survival outcomes were measured. In addition, 36/46 patients (78.3%) had upper-lobe tumors, limiting generalizability to other tumor locations. Functional imaging also introduces uncertainty from deformable registration, differences in spatial resolution, and the need for additional perfusion imaging. The study did not establish validated functional dose constraints or demonstrate that the modest dosimetric gains translate into lower pulmonary toxicity.
CLINICAL TAKEAWAY
Combining 4DCT ventilation with SPECT perfusion imaging may improve functional lung avoidance during NSCLC radiotherapy compared with anatomical or single-modality planning, particularly when highly functional lung lies close to the target. However, the additional lung sparing was relatively modest and accompanied by increased heart and spinal cord exposure. Prospective clinical validation is required before this approach can be considered a proven strategy for reducing radiation pneumonitis.