Aims: The aim of this study was to evaluate the number of CT slices require
d to produce satisfactory dose distribution for tangential field irradiatio
n of the chest wall and breast and to assess correlation of this with the v
olume of breast tissue treated. Forty-six patients underwent a CT scan of t
he thorax. An optimized plan was produced by assessing dose distribution on
the central axis (CAX) slice only. This plan was then recalculated using t
he entire CT data set without any changes to the beam parameters. A separat
e optimized plan was generated using the CAX slice and two slices indicativ
e of the upper and lower level of the field. This three-slice plan was then
calculated using the entire CT data set. Finally an optimized 3D plan was
generated using the entire CT data set. The different planning methods were
compared using dose-volume histograms (DVH). Dose inhomogeneity was define
d as any treatment volume outside the ICRU 50 dose distribution recommendat
ions.
Results: Fifty-two percent of single-slice plans and 21% of three-slice pla
ns (when assessed volumetrically) had greater volumes of breast tissue outs
ide the ICRU 50 report guidelines suggesting that better homogeneity could
be achieved by assessing a greater number of slices. Seventy-nine percent o
f three-slice plans showed no homogeneity improvement if the plan was calcu
lated with the entire 3D data set.
Conclusions: We conclude that a single-slice plan is unsatisfactory in prov
iding sufficient information about the dose variation across the treatment
volume and that ideally a 3D plan with DVHs should be produced. If the requ
ired data is unavailable then a minimum of three slices should be used as a
n approximation. We also propose a software tool for treatment planning sys
tems, which calculates the percentage of the total PTV having dose outside
the ICRU 50 radiation dose distribution homogeneity guideline range. (C) 19
99 Elsevier Science Ireland Ltd. All rights reserved.