Beispiel für die gestaltung der abstracts für die tagung 2006 (14 pt, fett)

Comparison of breast cancer treatment plan properties using two different
dose calculation algorithms

M. Sassowsky, D. Frauchiger, E. Born, R. Mini
Division of Medical Radiation Physics, Inselspital Berne and University of Berne, Switzerland
mail:
Introduction
Today different dose calculation algorithms are available in treatment planning systems (TPS). Two
examples are the analytical anisotropic algorithm (AAA) and the pencil beam algorithm (PBC). In the
presence of pronounced tissue inhomogeneities, the dose distribution calculated by AAA is generally
considered closer to the reality than the one calculated by PBC, as it includes a modelling of the lateral
energy transport. From the physicist’s point of view, this would imply using AAA instead of PBC.
However, past clinical experience was gathered with PBC, and a change in the calculation algorithm
might lead to a different tumor control probability (TCP) and/or normal tissue complication
probability (NTCP). It is thus advisable to compare the properties of treatment plans calculated with
AAA and PBC – preferably for a larger set of plans instead of a single plan. In the treatment of breast
cancer, the lung is an important organ at risk. Breast cancer treatment plans involve the boundary
between low density lung tissue and other tissue of higher density. The aim of this study is to
investigate possible consequences when changing the dose calculation algorithm from PBC to AAA.
Material and Methods
72 clinically applied breast cancer plans calculated with PBC were recalculated with AAA. The plans
were stratified into 4 groups (left/right; with/without suprax fields). For each patient the difference of
the DVHs calculated with AAA and PBC, respectively, was determined. Average DVHs were
calculated to obtain information representative for each group of plans. The shapes of the averaged
DVHs were compared between the four groups of plans, and between AAA and PBC. The number of
MUs to obtain the prescribed PTV D50% was compared between AAA and PBC. Furthermore, the
deviation between the lung V20Gy and the lung Dmean, calculated with PBC and AAA, respectively, was
calculated.
Results
From the set of the 72 plans, the average changes of the abovementioned quantities have been
determined. Compared to AAA:
PBC underestimates the number of MU to obtain the prescribed PTV D50% by: (uncertainties are given as 1 standard deviation)
Discussion
When changing the calculation algorithm from PBC to AAA, one could either adapt the dose
prescription and the accepted limit on V20Gy. Or, given the fact that the magnitude of observed
differences in the abovementioned quantities appears small, one might prefer to keep the prescription
dose and limit of V20Gy. This approach would lead to a marginally increased TCP and a more cautious
evaluation of the plans with respect to the lung exposure.

Source: http://www.sgsmp.ch/ann2010/PDF/a23.pdf