Feasibility of cone beam (CB) CT/ planning CT co-registration to estimate delivered doses in head and neck (HN) simultaneous integrated boost (SIB) intensity modulated radiotherapy (IMRT) treatments


Session type:

Andriana Michaelidou1, Rebecca Lynn1, David Convery1, Mary Lei1, Teresa Guerrero Urbano1
1Guys & St Thomas' NHS Foundation Trust, London, UK


Pre-treatment (planning) CT scans offer a snapshot of the patient’s anatomy. Serial CB-CT scans obtained during HN-IMRT can be co-registered with the pre-treatment scan, to estimate delivered doses and help design adaptive radiotherapy (ART) strategies, without the need for serial planning scans.


In this feasibility study, CB-CT scans of 3 patients treated with IMRT (60-65Gy in 30 daily fractions), obtained after mean doses of 14.45Gy, 34.67Gy, 46.23Gy and 57.06Gy, were transferred to Focal® and co-registered with the planning scans. Pre-treatment CTVs and OARs were transferred to the CB-CT images. Per-treatment CTVs and OARs were obtained by manual modification. PTVs were generated with a 5mm isotropic margin, clipped 5mm to skin. Per-treatment outlines were transferred to the planning scan. The treatment plan was applied and calculated for each CB-CT body outline using Monaco TPS®, generating 4 per-treatment dose maps, used to calculate delivered doses.


All CTVs decreased in size (radical CTVs 7.40% mean decrease, range 1.99%-14.01%; elective CTVs mean 9.57%, range 7.97%-12.45%).

Delivered doses were lower than planned with mean decreases in D95% as follows: radical CTV 2.52Gy (4.09%, range 3.09%-4.60%); elective CTV 3.37Gy (6.52%, range1.85%-12.89%); radical PTV 2.99Gy (4.91%, range 2.95-7.05; elective PTV 2.87Gy (5.49%, range 3.51%-7.87%). We recorded D98%, D95%, D50%, D5% and D2% with decreases in delivered doses noted across all.

Reduction in the volume of the parotids was observed (left mean reduction 10.44%, right 18.34%). Maximum increase in mean parotid dose was 0.86Gy. Maximum spinal cord doses decreased in all patients.


This method, by allowing for variations in body contour, can characterize changes in delivered versus planned doses and represents a good approximation to the delivered dose, whilst avoiding the need for serial planning scans. Criteria to select patients who might benefit from adaptive strategies are required and a larger study is currently underway to identify these.