Lung tumours, especially those in the lower lobes, can move a lot during respiration; this motion needs to be accounted for during radiotherapy. In cases where 4D CT simulation scans are not performed, the current protocol at our centre is to apply a generic (internal motion + setup) margin of 0·70 cm in the axial plane and 1·20 cm in the longitudinal plane to all lung tumours, regardless of location. We analyse the tumour motions of a cohort of our local patients and categorise them into different locations in the lung. We seek to assess the adequacy of the current margins and to derive a more accurate set of standard margins which are specific for lung tumour locations.

All cases of lung tumours treated with stereotactic ablative radiotherapy between 2012 and 2016 were identified retrospectively and 4D CT scan data analysed. These tumours were grouped into the following locations: upper zone (UZ), middle zone (MZ) and lower zone (LZ). The treatment planning system was used to generate the displacements of the centre of mass of the tumours in the right–left, anterior–posterior and superior–inferior axes; these were compared with the current generic margins. Median displacements were calculated for each axis in each location. New planning target volume (PTV) margins were derived by summing the median displacement, median absolute deviation (MAD) and 0·5 cm (for setup error).

Sixty-three cases were eligible for analyses. Motion in the superior–inferior direction was the greatest for all tumour locations, ranging from a median of 0·17 cm (MAD 0·12 cm) in UZ to 0·77 cm (MAD 0·27 cm) in LZ. Median tumour displacements in the anterior–posterior and right–left axes were similar for all locations, <0·30 and 0·20 cm, respectively. The current generic margins were adequate for only one-third of the cases in this study. A new PTV margin of 2·10 cm in the superior–inferior axis may be required for LZ tumours, while an additional 1–2 mm should be added to the current radial margins.

The current generic margins are inadequate for the majority of cases. Tumour motion is the greatest in LZ in the superior–inferior axis. Motion mitigation strategies are essential for large LZ tumours.