Session Item

To evaluate tumor motion and setup errors using MV portal imaging (PI) & KV three-dimensional Cone Beam CT (CBCT) in patients of carcinoma oesophagus receiving radical intent radiotherapy (RT).

Twenty patients of carcinoma oesophagus, planned for radical intent RT, were positioned using thoracic wing board device and planning CECT scans were acquired in free-breathing (FB) and 4 Dimensional CT (4DCT) using Varian Real-time Position Management (RPM) system. Patients were subdivided into Grp1: upper ± middle third (12 patients), and Grp2: lower third ± gastro-esophageal junction (GEJ) involvement (8 patients). Contouring was done on FB and 10 phases of 4DCT to generate ITV for motion assessment. Motion was assessed in superio-inferior (SI), anterio-posterior (AP) and medio-lateral (ML) directions manually using 1.0 mm grid and GTV of free-breathing scan as reference for determination of internal margin (IM). During treatment, verification MV Portal Imaging (PI) & KV Cone Beam CT (CBCT) were acquired on alternate days, with a minimum of three images in the first week and weekly thereafter. Off-line matching was done using carina and vertebral bodies as surrogates for PI while for CBCT, bone auto-match using clip box for vertebral bodies was used. Setup margins (SM) were calculated using van Herk’s formula (2.5 Σ + 0.7 σ) for PI and CBCT. PTV margins were calculated using SM+IM in all directions. 

For tumor motion assessment, 20 patients with 200 image sets corresponding to phases of respiration were assessed using 4DCT data. Internal Margins, mean ± SD (mm) in SI, AP, and ML directions for Grp1 were  4 ± 2, 2 ± 1, and 2 ± 1 while for Grp2 were 6 ± 2, 3 ± 1, and 3 ± 2 respectively. 20 patients with 164 Portal images and 112 CBCT images were assessed for set-up errors. The setup margins in SI (Y), AP (Z), and ML (X) axes for all the 20 patients were from PI were 13mm, 7mm, and 12mm respectively, and from CBCT’s were 11mm, 5mm, and 9mm respectively. The PTV margins (IM+SM) using PI for Grp1 were 17mm, 9mm, and 14mm and for Grp2 were 19mm, 10mm, and 15mm in SI, AP, and ML directions respectively. Whereas PTV margins using CBCT as verification modality for Grp1 were 15mm, 7mm, and 11mm, and for Grp2 were 17mm, 8mm, and 12mm in SI, AP, and ML directions respectively.

Both setup errors and motion were maximal in the SI direction and for lower 1/3rd ±GEJ tumors. CBCT verification can further lead to reduction in PTV margins.