Table 5 CTV to PTV margin recommendations in various series, with image guidance

Present series (n = 46)

As in table 4

As in table 4, with correction of isocenter placement errors 3 mm or greater in size on R-L and S-I axes, 2 mm or greater on A-P axis. No post-correction EPI.

van der Heide [5] (n = 453)

Supine, knee cushion. Empty bladder, no bowel instructions.

2 - 4 fiducial markers. Daily aSi EPI. Correction of all errors prior to treatment. Standard deviations were provided, from which we calculated margins required to give 95% probability of CTV coverage (CTV - PTV margin calculated as SD × 1.65 [6]).

Cheung [7] (n = 33)

Supine, vacuum lock bag. Empty bladder and rectum.

3 fiducial markers. Pre- and post-RT EPI days 1-9 to calculate individualized CTV-PTV margins (averages shown), which were used during the IMRT boost phase, during which daily on-line correction was performed according to fiducial marker position. A 2 mm factor was added in quadrature to the total error, to account for inaccuracies in the on-line correction process. Average individualized CTV to PTV margins are shown, although several patients had margins larger than 0.7 cm along the A-P axis.

J Wu [10] (n = 13)

Supine, alpha cradle, soft foam support for lower legs. Empty rectum and partially full bladder (drink 500 mL water 45 mins before) for CT and treatment

3 fiducial markers. Daily pre-treatment portal images 3× per week over the course of treatment. CTV to PTV margin required to give 99% probability of CTV coverage by 95% isodose line. Margins calculated according to movement of center of mass.

Litzenberg [13] (n = 11)

Supine, flat couch, knee support. No bowel or bladder instructions.

As in table 4, with the inclusion of intra-fraction motion.

Meijer [20] (n = 30)

As in table 4

4 fiducial markers. Simulation study based on 8 CT scans spaced over the course of treatment. Set-up to skin markers then daily on-line imaging, with correction of all set-up errors. Margins calculated using a dose warping technique to give 90% probability of covering the CTV with at least 95% of the prescribed dose.

Beltran [21] (n = 40)

As in table 4

As in table 4, with daily correction of all errors.

Nairz [22] (n = 27)

As in table 4

As in table 4, with daily correction of all errors.

Graf [23] (n = 23)

As in table 4

As in table 4, with daily correction of all errors.

Q Wu [24] (n = 28)

Not stated

15 CT scans obtained during the course of treatment and registered with respect to bony anatomy with planning CT. Evaluation of both image-based and contour-based registration methods. Analysis based on both geometric and dosimetric parameters. Estimated CTV to PTV margins required to allow a dose reduction on the prostate (D99) of not more than 2% for 90% of patients.