Internal target volume determined with expansion margins beyond composite gross tumor volume in three-dimensional...

Gross tumor volume (GTV) of lung cancer defined by fast helical CT scan represents an image of moving tumor captured at a point in active respiratory movement. However, the method for defining internal margins beyond GTV to account for its expected physiologic movement and all variations in size and shape during the administration of radiation has not been established. The goal of this study was to determine the internal margins with expansion margins beyond individual GTVs defined with (1) fast scan at shallow free breathing, (2) breath-hold scans at the end of tidal volume inspiration and expiration, and (3) 4-s slow scan to approximate the composite GTV of all scans.

A series of sequential CT scans were acquired with (1) a fast helical scan at shallow free breathing and (2) breath-hold scans at the end of tidal volume expiration and inspiration for the first 6 patients, and (3) a 4-s slow scan at quiet free breathing, which was added for the latter 7 patients. We fused breath-hold scans and the 4-s slow scan to the fast scan at shallow free breathing to generate the composite GTV. Margins necessary to encompass the composite GTV beyond individual GTVs defined by either fast scan at quiet free breathing, breath-hold scans, or the 4-s slow scan at quiet free breathing were defined as expansion or internal margins and termed the internal target volumes. The centroid of the tumor volume was also used as another reference for tumor movement.

Thirteen patients with 14 tumors were enrolled into the study. Substantial tumor movement was noted by either the extent of internal margins beyond each GTV or the movement of the centroid. Internal margins varied significantly according to the method of CT scanning for determination of GTV. Even for tumors in the same lobe of the lung, a wide range of internal margins and significant variation in the centroid movement in all directions (x, y, and z) were observed. The GTV of a single fast helical scan at free breathing (n = 14) required the largest internal margin (mean, 3.5 mm; maximum, 18 mm; standard deviation [SD], 4.2 mm) to match the composite GTV, compared with those of the 4-s slow scan (mean 2.7 mm, maximum 14 mm, SD 3.5 mm) or combined breath-hold scans (mean 1.1 mm, maximum 9 mm, SD 1.9 mm). Internal margins (expansion margins) required to approximate the composite GTV in 95% of cases were 13 mm, 10 mm, and 5 mm for the GTVs of a single fast scan, 4-s slow scan, and breath-hold scans at the end of tidal volume inspiration and expiration, respectively.

The internal margins required to account for the internal tumor motion in three-dimensional conformal radiotherapy are substantial. For the use of symmetric and population-based margins to account for internal tumor motion, GTV defined with breath-hold scans at the end of tidal volume inspiration and expiration has a narrower range of internal margins in all directions than that of either a single fast scan or 4-s slow scan.