KEY POINTS
- This retrospective single-center study evaluated 25 peripheral lung cancer cases using VMAT-based SBRT to investigate the interpretation of PlanIQ feasibility dose–volume histogram (FDVH) f-values. Median tumor diameter was 1.7 cm, median GTV volume 2.5 cm³, and median PTV volume 12.4 cm³. Four central tumors were excluded from the original 29-case dataset to maintain a consistent peripheral lung SBRT planning scenario.
- All plans prescribed 50 Gy in four fractions, normalized so that PTV D95% equaled the prescription dose, while allowing intentional intratumoral dose heterogeneity with PTV D2% approximately 135–145% of prescription. The resulting mean PTV D2% was 70.0 ± 2.05 Gy, while D98% was 48.54 ± 0.60 Gy. Plans were generated using two partial-arc VMAT fields with 6-MV FFF beams, and dose was calculated using Acuros XB.
- PlanIQ analysis was performed under two conditions. Condition A used the conventional 50-Gy prescription dose as the reference, while Condition B incorporated adjusted reference doses to better represent heterogeneous target irradiation: 61.4 Gy for GTV and 43.2 Gy for PTV minus GTV. These values were derived from the mean minimum doses across the study cohort and applied consistently to all 25 cases.
- Under standard Condition A, PlanIQ classified 6/25 plans (24%) as theoretically impossible for normal-lung V10, 17/25 (68%) for V20, and 9/25 (36%) for mean lung dose. These classifications occurred despite the corresponding dose distributions having been generated clinically, illustrating potential disagreement between FDVH feasibility estimates and actual achievable SBRT dose distributions.
- Adjusting the target reference condition improved some results but did not resolve the discrepancy for normal lung. Under Condition B, the number of plans classified as impossible decreased to 5 for V10, 14 for V20, and remained 9 for mean lung dose. The reduction in impossible classifications for V20 from 17 to 14 patients was not statistically significant (p=0.250), while numerical f-value changes were also nonsignificant for V10 (p=0.705) and V20 (p=0.832).
- The impact of reference-dose adjustment was much greater for optimization structures near the target. For the innermost 2-mm ring (Ring 1), 19/25 cases were classified as impossible under Condition A, whereas all 25 became evaluable under Condition B. Significant differences in f-values were observed for Ring 1, Ring 2, and mean lung dose (all p<0.001), and for Ring 3 (p=0.007). Interestingly, the highest median f-values occurred in the intermediate ring rather than consistently decreasing with distance from the PTV.
- Exploratory analysis suggested that FDVH–clinical plan disagreement for lung V20 occurred more frequently with smaller tumor diameter, GTV volume, PTV volume, and PTV-to-ipsilateral-lung volume ratio, with corresponding p-values of 0.016, 0.008, 0.008, and 0.007. However, these associations were unadjusted and hypothesis-generating. The study included only one planning platform and PlanIQ implementation, with no independent validation or clinically established f-value thresholds for heterogeneous lung SBRT.
CLINICAL TAKEAWAY
PlanIQ-derived feasibility scores may not accurately represent achievable normal-lung doses when applied directly to SBRT plans with intentional central target hotspots and steep dose gradients. Adjusting reference-dose assumptions can improve interpretation near the PTV but does not fully resolve discrepancies in lung dose metrics; therefore, f-values should complement rather than replace direct DVH assessment, plan-quality evaluation, and clinical judgment.