Impact of post-correction validation scans on intrafraction tumour misalignment in lung stereotactic body...

Tumour position may change after the acquisition of an initial cone-beam computed-tomography for setup correction (CBCT^PreCor). In this study, the impact of post-correction validation (CBCT^Val) and intra-arc scans (CBCT^IA) on intrafraction tumour misalignment during volumetric-modulated-arc-therapy (VMAT) lung stereotactic-body radiotherapy (SBRT) was evaluated.

Early stage lung cancer patients (n = 167) treated with VMAT SBRT had the following image guided (IG)RT protocol: (1) CBCT^PreCor, (2) CBCT^Val to verify tumour alignment, (3) an intra-arc scan (CBCT^IA) during both VMAT arcs. Additional corrections were made for residual misalignments ≥0.3 cm. The actual and two simulated protocols were analysed: (1) clinical (CBCT^Clin), (2) simulations without validation scans (CBCT^No_Val), (3) simulations without repeat CBCTs (CBCT^No_IFMM). Grand-mean (GM), systematic (∑) and random (σ) tumour misalignment in CBCT^IA-1 and CBCT^IA-2 were calculated. Patient characteristics were evaluated for association with extra validation scans.

CBCT^Val triggered a second correction in 20.4 % of fractions in 47 % of patients and CBCT^IA-1 in 14.4 % of fractions in 40 % of patients. Omitting CBCT^Val increased ∑ and σ ranging from 27-30 % and 20-45 % for the different directions. Omitting also CBCT^IA further increased ∑ ranging from 55-90 %. Omitting CBCT^Val and CBCT^IA would require a 1-2 mm planning target volume margin increase. Respiratory amplitude and body mass index (BMI) were significantly associated with extra corrections (area under the curve: 0.75).

This study demonstrates that CBCT^Val and CBCT^IA-1 reduce geometric uncertainties in VMAT lung SBRT. Respiratory amplitude and BMI were significantly associated with extra corrections but cannot reliably identify patients requiring extra validation scans.