Evaluation of Intrafraction Shifts in Lung SBRT Targets over the Course of a Lengthy Treatment

Immobilization-free image-guided lung SBRT has been the standard of care at our institution for over five years.  The technique provides patient comfort through absence of rigid immobilization devices and shorter setup time, relying on 4D-CT simulation and CBCT imaging for accurate target setup at treatment.  The planning target volume (PTV) is the internal target volume (ITV), with an additional margin of 5mm to account for systematic setup uncertainties.  Systematic errors were evaluated at SBRT commissioning and factored into the treatment verification process.

As part of ongoing quality and process improvement, we evaluated the intrafraction motion during long SBRT treatments, as experienced by patients receiving 34, 30 and 16 Gy per fraction.  The study involved a comparison of pre- and immediate post-treatment CBCT images and assessment of potential displacement of the visible lung lesion on the post-treatment CBCT image.  Acceptable displacement was defined as the visible lesion falling well within the margins of the PTV.

Thirty-six cases were reviewed, receiving 30Gy (n=19) or 34Gy (n=6) per fraction and 48 Gy in 3 fractions (n=10), without interruption.  All cases were planned with 6MV energy beams, double or triple arc VMAT technique, and delivered on the Varian TrueBeam STx (with HDMLC) unit with a maximum dose rate of 600 MU/sec.  The beam-on time was, on average, 7.6 min, 15.4 min and 18.1 min, respectively.   

In three cases, one receiving 34Gy and two receiving 30Gy, the post-treatment CBCT indicated intrafraction motion where part of the visible lung lesion was out of the PTV.  All three cases were upper left lung lesions, and the significant displacement was in the posterior direction (11, 9 and 6mm displacement of the GTV center of mass).  The displacement was partially due to general change in patient position that occurred over the course of treatment.

This analysis suggests that: 1) long beam-on time for lung SBRT has a risk of unacceptable intrafraction displacement of the tumor outside the PTV; and 2) the moving tumor visible on 4D-CT images and reflected in the ITV may not fully characterize the entire range of tumor motion.  This may be more important for treatments with long beam-on time duration.   Possible solutions are periodic monitoring of lung tumor position during treatment, as well as the use of higher dose rate beams to shorten the treatment time thereby reducing the potential for intrafraction motion.