Dosimetric assessment of the field abutment region in head and neck treatments using a multileaf collimator

Abstract

Background and purpose: The use of conventional asymmetric collimators for junctioning of abutted fields can lead to significant dose inhomogeneity, due to jaw misalignment. However, recent technologic advances enable us to fabricate much finer leafpositioning accuracy. Consequently, it is anticipated that the use of multileaf collimator (MLC) will potentially improve dose homogeneity at the junction of abutted fields. In this work, we evaluated the dose inhomogeneities at the match-plane in monoisocentric three-field head and neck setups, using MLC for field abutment.

Material and methods: To define either the anterior or the lateral fields, the MLC was used with either the longitudinal (0 degree angle) or the transverse (90 degree angle) settings. For 0 degree setting, each leaf moves in a direction perpendicular to the gantry rotation axis, hence the "tongue and groove" (T&G) design can effect matching-area dose at the side of the leaf (Figure 1a). For 90 degree setting, the rounded shape of the leaf produces its effect at the leaf end (Figure 1b). Four combinations of abutted anterior field and abutted lateral field defined by MLC, i.e., abutted using MLC side-by-side, side-by-end, end-by-side and end-by-end, were compared (Table 1). Dose inhomogeneity was measured at the junction of the two abutted fields with films in a solid water phantom (Figure 2). The effect of jaw settings as a backup diaphragm on the dose distribution was also studied. Reproducibility of the results was confirmed by repeated measurements over a 1-year period.

Results: Abutted fields using MLC side-by-side caused underdose of approximately 15% (Figure 3a). Abutted fields using MLC side-by-end produced > 10% overdose that could be improved to +/- 1% for 0.5 mm overlap of the leaf end from the lateral portals (Figure 4). When using end-by-side, an overdose of approximately 15% was observed. However, the dose improved to a homogeneous dose for 0.8 mm overlap of leaf end from the anterior portal (Figure 5). End-by-end showed an overdose of > 20% (Figure 3b). This overdose could be smoothed out by overlaps of both leaf ends by 0.8 mm from both lateral and anterior portals (Figure 6). The ideal jaw position was found to be at 1 mm away from the beam central axis in any combination (Table 2).

Conclusion: The use of MLCs for photon field junction matching is appropriate and represents an alternative approach to the problem of field matching using the asymmetric jaws in head and neck treatments.