KEY POINTS
- This retrospective study from National Cancer Center Hospital East, Japan, analyzed respiratory motion in 41 patients with lung cancer undergoing VMAT planning with four-dimensional CT (4DCT). Investigators tracked 229 thoracic lymph nodes, including 130 upper nodes (stations 1–6) and 99 lower nodes (stations 7–14). Every patient had at least one upper and one lower thoracic node, allowing direct within-patient comparison, while linear mixed-effects models accounted for clustering of multiple nodes within individual patients.
- Lower thoracic lymph nodes moved significantly more than upper nodes during respiration. Mean cranio-caudal displacement was 5.2 ± 3.0 mm for lower nodes versus 3.3 ± 1.9 mm for upper nodes, corresponding to a within-patient difference of 1.9 mm (95% CI 1.21–2.69; p<0.001). Mean three-dimensional displacement magnitude was also greater for lower nodes (5.7 ± 2.9 vs 4.1 ± 2.0 mm), with a paired difference of 1.6 mm (95% CI 0.88–2.28; p<0.001).
- Motion exceeding 5 mm was considerably more frequent in lower thoracic nodes. Cranio-caudal displacement greater than 5 mm occurred in 49% of lower versus 24% of upper nodes, while three-dimensional displacement exceeded 5 mm in 51% versus 37%, respectively. In mixed-effects logistic regression accounting for patient clustering, lower-node location was associated with higher odds of exceeding 5 mm in the cranio-caudal direction (OR 7.81, 95% CI 3.09–19.74; p<0.001) and three-dimensional displacement (OR 4.65, 95% CI 1.94–11.11; p<0.001).
- Respiratory motion was anisotropic, with different regional displacement patterns. The patient-level 90th percentile of mean cranio-caudal displacement was 6.0 mm for upper nodes versus 8.5 mm for lower nodes, while the corresponding three-dimensional values were 6.4 and 8.7 mm. Upper nodes showed relatively greater lateral movement, whereas lower nodes demonstrated more prominent cranio-caudal displacement. These values describe observed motion and were not intended as directly applicable internal margin prescriptions.
- Exploratory station-level analysis showed substantial heterogeneity even within the upper and lower regions. Lower mediastinal stations 8 and 11 exhibited some of the largest displacements, with median three-dimensional motion of approximately 6–8 mm. Within the upper mediastinum, stations 4 and 6 had median vector displacements of approximately 4.1 and 4.3 mm, whereas several other upper stations showed median motion closer to 1–3 mm. This variation challenges the assumption that a single regional margin can adequately describe all thoracic nodal targets.
- Diaphragmatic excursion did not reliably predict nodal motion. Mean diaphragmatic excursion was 13.0 ± 4.1 mm, but it was not independently associated with mean nodal cranio-caudal displacement (β=0.005 mm per mm of diaphragm movement; p=0.962) or three-dimensional displacement (β=0.024; p=0.801). Interactions between anatomical region and diaphragmatic motion were also nonsignificant. Therefore, diaphragm displacement alone cannot be considered a reliable surrogate for individual mediastinal or hilar node motion.
- The analysis quantified simulation-phase displacement rather than delivered-dose consequences. Motion was measured using centroid-based tracking across respiratory phases, with expert visual verification, but changes in lymph node shape and deformation were not fully characterized. The requirement for both upper and lower nodal involvement selected a particular patient population, while 4DCT reconstruction uncertainty, irregular breathing, and inter- and intrafraction changes during actual treatment were not evaluated. No clinical target coverage, toxicity, or recurrence outcomes were compared.
CLINICAL TAKEAWAY
Respiratory motion of thoracic lymph nodes varies substantially with anatomical location, with lower mediastinal and hilar nodes exhibiting greater displacement than upper nodes within the same patient. These findings support patient-specific 4DCT assessment and caution against relying solely on uniform nodal margins or diaphragmatic motion as a surrogate, although prospective dosimetric validation is required before adopting specific station-based margin recommendations.