Session Item

Deep-inspiration breath-hold (DIBH) stereotactic body radiotherapy (SBRT) is an option for very mobile or very small lung tumors. Inter-fractional tumor position variation is minimized by daily cone-beam CT (CBCT), but the extent of intra-fractional tumor position variation is difficult to assess.

The aim of this study was to estimate the intra-fractional tumor position variation based on CBCTs acquired before and mid-treatment for patients receiving lung SBRT in DIBH and free breathing (FB). Additionally, investigate what impact two different patient cohorts have on the resulting setup margins.

Two patient cohorts (MM1 (from April 2018 to March 2019), and MM2 (from October 2020 to September 2021)), in total 84 consecutive patients scheduled for lung SBRT (67.5 Gy / 3 fractions, 45 Gy / 3 fractions, or 50 Gy / 5 fractions) were retrospectively included in this study. Both cohorts included 21 DIBH and 21 FB patients each. Patients were selected for DIBH treatment if peak-to-peak tumor motion in the longitudinal (LNG) direction exceeded 1 cm in FB. The patients were audio-visually guided during DIBH, to a comfortable patient-specific breath-hold level using a gating window of 2–3 mm. Both cohorts used a marker based optical tracking system at treatment (Respiratory gating for TrueBeam, Varian). Pre- and mid-treatment CBCT were acquired for all fractions to evaluate the intra-fractional tumor position variation. All CBCTs were registered to the planning CT by soft-tissue tumor match. Resulting offsets were used to calculate bi-directional setup margins based on van Herk formalism for lung (van Herk et al. 2000). The Wilcoxon rank-sum test was used for statistical testing regarding differences in tumor position variation between FB and DIBH, and between the two patient cohorts. The results were considered statistically significant for p < 0.05.

A statistically significantly larger intra-fractional tumor position variation in LNG was observed for DIBH compared to FB for the pooled data (MM1+MM2, Figure 1A) (p = 0.001), resulting in a 3-4 mm larger setup margin (Table 1). In the lateral (LAT) direction the observed difference was small, but statistically significant (p = 0.004) and resulted in setup margin being 1 mm larger in DIBH compared to FB. No statistically significant difference in intra-fractional tumor position variation was observed in the vertical (VRT) direction and the resulting setup margins were identical for FB and DIBH respectively. Variations between MM1 and MM2 were in general small. Calculating setup margins based on the pooled data were found to be clinically acceptable, within approximately 1 mm differences (Table 1). 

Large intra-fractional tumor position variations during DIBH SBRT for patients with tumors moving > 1cm in FB result in increased setup margins compared to FB SBRT for patients with tumors moving < 1cm. Pooling data from two patient cohorts were clinically acceptable.