The moving junction is used in craniospinal irradiation (CSI) to smooth out
any dose inhomogeneity across the head/spine junction. The aim of this stu
dy was to demonstrate the extent of the head/spine junction inhomogeneity i
n treatment plans of actual patients and to compare stationary and moving j
unction data. The radiotherapy plans, prescriptions and case notes of 18 pa
tients (12 medulloblastomas, 3 supratentorial primitive neuroectodermal tum
ours, 2 pineoblastomas and 1 pineal germinoma) treated with CSI (35 Gy in 2
1 fractions over 29 days) were examined. At 16 months median follow-up (ran
ge 1.5-35.5 months), no junctional recurrences or myelopathy were observed.
Using the moving junction technique the mean maximum anterior cord dose, f
rom 5 cm caudal to 3 cm cephalad of the Day 1 junction, was 36.3 Gy, and th
e mean minimum anterior cord dose was 32.9 Gy, with a mean within-patient v
ariation of 3.4 Gy (9.7% of 35 Gy). In four patients, comparison of dose va
riation across the field junction was made between the original plans and a
re-plan using a stationary junction. The effect of a matched junction, a 2
mm overlap and a 3 mm gap were studied both for moving junction and statio
nary junction techniques. Dose variations were similar in all cases for exa
ctly-matched fields. but for a 2 mm overlap or gap the dose variation was s
maller in all but one case for the moving junction technique. These data su
ggest that the moving junction is important to minimize the risk of overdos
e or underdose across the spine/head junction in CSI.