Quality Assurance of Stereotactic Radiotherapy Techniques for a Randomized Phase III Trial Comparing Surgery to Radiosurgery (ROSEL) for Stage Ia NSCLC

Purpose/Objective(s): To evaluate the current techniques for performing stereotactic radiotherapy for NSCLC within centers participating
in the multicenter Dutch Radiosurgery Or Surgery for Early stage Lung cancer (ROSEL) trial.
Materials/Methods: A questionnaire was send to all 10 institutes that expressed their interest to join the trial. All centers had to
send a dry run patient plan for review. 4D-CT scanning and irradiation of the Quasar phantom was performed. The lung cylinder of
the Quasar phantom was adapted to contain both a 15 mm and a 30 mm diameter plastic sphere to mimic a tumor (GTV). Cylinder
motion was performed as previously described (Lujan, Med. Phys 1999) mimicking actual breathing motion with 15 and 25 mm
amplitudes and 3 and 6 seconds breathing periods. Gafchromic EBT films were used to measure dose in the stationary situation and
with 15 mm, 3 seconds period movement.
Results: Nine out of 10 centers perform 4D-CT scanning while 1 center performs 7 consecutive fast 3D CT scans. Seven institutes
use Maximum Intensity Projection for generation of the internal target volume, while 3 centers use the mid-ventilation concept.
Multiple non-coplanar beams are used in 6 centers, multiple co-planar beams in 2 centers, dynamic conformal arcs in a single centre,
and volumetric modulated arcs in another single centre. Five institutes use a standard (Type A), and the other five use a more
advanced (Type B) dose algorithm for heterogeneity corrections. Measurements on 4D-CT overestimated the GTV on average 0.3
cc (range: 0 to 1.8 cc) for the ø15 mm (1.77 cc) sphere and 1.0 cc (range:-1 to 5.5 cc) for the ø30 mm (14.1 cc) sphere. Measured
amplitudes were on average 0.5 mm, 1.5 mm and 2.0 mm smaller than expected for the 15mm and 3 sec, 15 mm and 6 sec and 25
mm and 3 sec movements, respectively. Although the mid-ventilation position generally could be determined within 3 mm of the
theoretical position, one deviation of 6.1 mm was observed due to an erroneous mid-phase determination. All centers verify the
patient position by means of online volumetric imaging. Gafchromic film dose measurements are being performed and will
also be presented.
Conclusions: All institutions could comply with the ROSEL guidelines. The use of advanced dose algorithms and motion-specific
CT scan protocols has been stimulated. The quality assurance data collec