Institutional Audit of Adaptive Re-Planning With Pencil Beam Scanning Proton Beam Therapy

Abstract

Pencil beam scanning proton beam therapy (PBS-PBT) dose distributions are extremely sensitive to tissue deformation. We conducted a retrospective audit of 190 consecutive patients treated with image guided PBS-PBT over a period of 1.5 years, to analyze the incidence, causes and trigger for adaptive re-planning.Patients were treated with daily image guidance with either KV X-rays or Cone beam CT scan (CBCT) or both. Patients underwent quality assurance imaging either with a CT (QA-CT) or MRI after every 5-10 fractions or earlier based on CBCT. The nominal plans were overlaid on the QA-CT and the doses were recalculated to assess dose perturbations and need for adaptive re-planning. The thresholds for re-planning were based on the physician's choice. The trigger for re-planning was also recorded and classified as elective QA-CT, QA-CT based on CBCT trigger or based on clinical triggers. After the initial 6-9 months of the study several measures were implemented with regards to simulation and planning such as skin rind evaluation for head neck patients, modified planning techniques, avoidance of certain beam angles and changes to class solutions.At least 3 QA images were acquired for each patient. A total of58 patients (30.52 %) required adaptive re-planning, of which 9 (15%) underwent re-planning more than once. Of the patients undergoing re-planning, 44.8%were of head neck,15.5%brain, 10.3%pelvic,5.2%skull base, 5.2% CSI, and 12.3% were of thoracic subsite.68% of Head neck and 66% of thoracic patients underwent adaptive re-planning. The number of patients undergoing adaptive re-planning in week 1,2,3,4,5&6 were 11, 10, 18, 8, 7 and 4 respectively. Most common cause for adaptive planning was deformation of target (23.07%) and beam path changes (23.07%) followed by persistent set up errors (9.23%), OAR deformation (7.69%) and combination of either of the factors in 26.15%. 59% of plans were adapted due to overdose in OARs, 29.7% were due to under-coverage of target while 10.8% had both. The average dose perturbation in the high dose target was higher than the intermediate dose target. Of the first 70 patients 43% underwent plan adaptation while after implementation of measures to reduce the adaptive re-planning, in next 120 patients only 23.3% required the same. Trigger for adaptive re-planning was elective QA-CT in 15%, CBCT trigger in 78%, and clinical triggers in 7% of patients.Our retrospective audit showed that the need for adaptive re-planning for patients treated with PBS-PBT is significantly high especially for head neck and thoracic tumors. The most common causes for adaptive re-planning were target deformations and beam path changes which in a majority of cases are picked up by CBCT imaging. Innovative simulation and planning approaches could potentially reduce the need for adaptive re-planning.