Evaluation of patient positional reproducibility on the treatment couch and its impact on dose distribution using rotating gantry system in scanned carbon-ion beam therapy.

PURPOSE

The daily variations in patient setup may cause beam range uncertainties. We evaluated the reproducibility of relative position between the patient and the treatment couch throughout the treatment course and assessed its effects on dose distributions when a beam passes through treatment couch using rotating gantry system.

METHODS

We enrolled 1023 patients (=13072 fractions) treated by carbon-ion pencil beam scanning therapy. Seven treatment sites including prostate, head and neck, bone and soft tissue, rectum, liver, lung, and pancreas were investigated. Inter-fractional changes in couch position relative to the patient were defined as translational errors. Changes in couch rotation were defined as rotational errors. Treatment planning was performed for 4 patients in each of the treatment sites. Dose distributions were then re-calculated after the couch was shifted according to average, 95th percentile, and maximum values of translational error.

RESULTS

Large positional errors (>1.5 cm) were observed in 5% of treatment fractions. Positional errors were largest in prostate and pancreas patients, while smallest in head and neck and lung patients. There were no or only small changes in PTV-D95 and CTV-D95 values for almost all treatment sites. Clinically significant changes were observed in the duodenum (difference in D2cc values ranged from -55% to 28% with maximum couch shift) in pancreas treatment.

CONCLUSIONS

Although underdosage to the PTV or CTV was limited, significant overdoses to organs at risk were found. The improvement of immobilization technique and appropriate selection of gantry angles could reduce the uncertainties due to changes in patient position.