Page 62 - ARNM-2-4

Advances in Radiotherapy
& Nuclear Medicine Different approaches for the computation of BED

2 Gy/fraction, the percentage of the normal lung volume corresponding BED or EQD . The relationship between
2
receiving doses ≥20 Gy is employed to assess the risk of the latter quantities and NTCP is much more tenuous. In
radiation pneumonitis. 1 general, NTCP depends on the organization of the organ
Although DVPs and associated planning constraints at risk, which can be described using the concepts of
for the treatment target and normal organs are still used volume effects, functional subunits, and serial and parallel
6,16-18
most frequently for plan evaluation, the DVP-based tissues developed and employed in previous studies.
criteria are indirect measures of the radiobiological Because the computations of BEDs and/or EQD only
2
effects of the treatment. A potentially better approach involve parameters of the LQ model and are considered
is based on radiobiological parameters such as tumor the treatment regimen (i.e., total dose and dose per
control probability (TCP) and normal tissue complication fraction), neither BED nor EQD should be used as NTCP
2
probability (NTCP), which can be computed using different surrogates. 19
radiobiological models of dose response in tissues. The This study compares different approaches used to
2-4
existing dose-response models can be divided into two compute BED for a non-uniformly irradiated target.
groups, referred to as mechanistic or phenomenological Because EQD2 is proportional to BED (e.g., Jones and
models. The former models mathematically express dose Hoban ), the conclusions of this work are also applicable
19
response by considering the basic biological processes in to the computation of EQD2.
the irradiated tissues. Conversely, the latter models aim to
express the empirical relationships between the absorbed In the original framework, BED was defined under the
11
dose and the induced biological effects. assumption of a uniformly irradiated organ of interest. In
this case, BED depends on the total dose, dose per fraction,
In principle, mechanistic models are more advantageous and alpha/beta ratio. In practice, the absorbed dose in the
because they can be applied in various cases including treatment target is always non-uniform. Consequently, the
different irradiated tissues and/or treatment regimens. computation of BED generally requires knowledge of both
Unfortunately, the complex phenomenon of radiation- alpha and beta and dose distribution in the target. 19,20 The
induced tissue damage and underlying biological use of BED distribution was suggested in seminal studies
processes are not sufficiently understood. Consequently, by Lee et al. and Wheldon et al., with the expectation
22
21
mechanistic models are normally limited because they that the transition from the physical dose to BED can
employ several assumptions and consequently are not yet facilitate plan optimization and clinical research.
fully capable of faithfully predicting treatment outcomes.
Phenomenological models are utilized in clinical practice 21 The procedure for computing the BED distribution
because they are simpler than the mechanistic models. in requires voxel-by-voxel conversion of the dose to
20
Phenomenological models are typically derived using BED based on the LQ model. Conversely, Niemierko
empirical data acquired for specific tissues and the developed the concept of equivalent uniform dose (EUD),
conventional dose per fraction of 2 Gy. 5-10 This frequently which is computed by averaging the probability of survival
19
prevents the employment of phenomenological models in the organ of interest. Jones and Hoban utilized the
in clinical cases for which the parameters of these models concept of EUD (which is similar to BED and EQD )
nud
2
have not been validated. for external beam radiotherapy to perform radiobiological
comparisons of different treatment plans. O’Donoghue
23
Among the radiobiological parameters used for employed EUD to assess the radiobiological effect of
the evaluation and comparison of treatment plans, the non-uniform dose distributions on tumors treated with
biologically effective dose (BED) 11-13 and equivalent radionuclide therapy.
dose in 2 Gy fractions (EQD ) are used most often.
14
2
The popularity of BED and EQD is attributed to the 1.2. Main objective
2
following: (a) both quantities are normally defined using Modern TPSs include modules for computing the BED
the standard linear–quadratic (LQ) model for cell kill; (b) and EQD in each voxel and the mean values of these
the computation of BED and EQD in treatment planning quantities. 2 24-26 The mean BED (BED ) is obtained by
2
mean
systems (TPS) generally requires only the α:β ratio from averaging BEDs in different voxels. 25,26 This study aimed
the LQ model, without needing the separate values of α to (a) compare BED with the corresponding BED
mean
nud
and β; and (c) BED and/or EQD are frequently employed computed by averaging the probability of tumor cell
2
as TCP surrogates. survival in the target and (b) demonstrate that the use of
TCP monotonically increases with increasing BED BED mean can lead to incorrect ranking of the compared
and EQD . As a result, the intercomparison of treatment treatment plans. To compare BEDs, 51 cases of non-small
15
2
plans in terms of TCP can be performed by comparing the lung cancer treated with volumetric modulated arc therapy
Volume 2 Issue 4 (2024) 2 doi: 10.36922/arnm.4826