Necessity of manually adjusting automatic deformed CTVs on repeat CT scan to evaluate CTV coverage in IMPT for lung cancer

Background/Introduction: Manual CTV adjustment after deformable CTV registration on repeat-CT scans (rCT) is time consuming and may delay decisions on whether or not to perform plan adaptations. Our aim was to investigate if omitting such manual CTV adjustments would alter the decisions and/or impact the overall adequacy of treatment. Methodology: Seventeen lung cancer patients underwent a 4D planning-CT (pCT0) and 3-5 weekly 4D rCTs. In total 82 4D-average rCTs were available. Robustly optimized IMPT plans were established on the pCT0 aiming for a CTV D98 ≥ 57 Gy. The plans were evaluated on all rCTs including setup and range error scenarios. CTVs were deformed (intensity based deformable registration) from pCT0 to the rCTs (CTVauto) and manually adjusted by a clinician (CTVman). When rCT CTVauto and/or CTVman D98 < 56.5 Gy (worst case scenario dose) the IMPT plan was adapted. Results: On average, CTVman was smaller than CTVauto (Δ volume in weeks 1-5: -4%, -6%, -9%, -11% and -15% (p < 0.001), respectively. The average CTVauto D98 and CTVman D98 on all rCTs were 55.17 Gy and 55.13 Gy (p = 0.767), respectively. Omitting CTVman on all 82 rCTs would result in 3 ‘unnecessary' plan adjustments and 1 ‘skipped' plan adjustment. In the latter single case, the weekly rCTs dose deformed and summed in the CTV on pCT0 was adequate despite the skipped plan adaptation (58.1 Gy). Conclusions: In 17 lung cancer patients, omitting manual adjustments of deformed CTVs on weekly rCTs did not impact adequate CTV coverage of IMPT treatment.