LED, which occurs when there is an imbalance of electron energy entering or leaving a small tissue volume, perturbs the dose distribution within the tumour and nearby tissues. If not accounted in dose calculations, LED can result in the delivered dose differing drastically from the planned treatment. Researchers from the University of Western Ontario (London, ON) have now performed a detailed simulation study of the beam and lung density parameters that can cause LED (Phys. Med. Biol. 57 1543).

"Our primary objective was to explore the combination of parameters that could produce LED, so that clinicians can be more aware of this effect, confirm that a valid dose calculation algorithm is being used, potentially avoid the deleterious dosimetric effects of LED through optimal selection of treatment conditions, and design future clinical trials that are less sensitive to the unexpected effects of LED," explained Brandon Disher, from the University's Department of Medical Biophysics.

Critical densities

Disher and colleagues used Monte Carlo code to simulate four water–lung–water slab phantoms, three of which contained small tumours. Dose calculations were performed using six beam energies (1.25 to 18 MV), five field sizes (1 x 1 to 15 x 15 cm2) and 12 lung densities (0.001 to 1 g/cm3) – a total of 1440 simulations.

The researchers first examined the lung-only phantom. Depth-dose profiles revealed that, for a given beam energy and field size, dose within lung tissue generally increased as lung density decreased. Below a critical density, however, LED occurred, along with a sudden decrease in lung dose. For example, for a 1.25 MV, 10 × 10 cm2 field, the critical lung density was 0.1 g/cm3. At 0.001 g/cm3, the central-axis dose decreased by up to 46% (relative to dose in water).

To quantify LED severity, the team developed the relative depth-dose factor (RDDF). LED occurs when RDDF is less than 1, while a RDDF of 0.7 or less indicates severe LED, in which both lung and tumour dose are drastically reduced.

RDDF isocontour maps revealed that, at 1.25 MV, regions of severe LED occurred primarily for ultralow lung densities (below 0.1 g/cm3) and field sizes of less than 5 x 5 cm2. At 6 MV, low RDDF values were seen for larger field sizes and higher densities. With 18 MV photon energies, LED even occurred in adipose-like tissues (lung density of 0.7 to 1.0 g/cm3) and field sizes of up to 15 x 15 cm2.

The authors also present a map displaying combinations of treatment parameters that cause LED in a lung phantom. They suggest that these data can be used as an aid for clinicians to select lung radiotherapy parameters.

Tumour tests

The researchers next examined the tumour-containing phantoms. Depth-dose profiles showed that LED caused a dose reduction (compared to water) within the lung tissue, and variable dose levels within the tumour. Dose reductions were worse (up to 80%) for smaller tumours irradiated with high-energy beams, small field sizes and low lung density.

For the 3 x 3 x 3 cm3 tumour, using 6 MV, 5 × 5 cm2 fields, dose to the proximal and distal tumour surfaces were reduced by 11% and 15% for the severe LED (RDDF of 0.3) seen at a lung density of 0.001 g/cm3. In contrast, the dose at the tumour centre was enhanced by 8%.

For a 1 x 1 x 1 cm3 tumour and 3 × 3 cm2 photon fields, reduced dose occurred at higher lung densities due to the smaller field size. At a density of 0.001 g/cm3 (RDDF of 0.2), for example, dose in the tumour centre was reduced by 10% and 20%, for 6 and 18 MV fields, respectively. Tumour surfaces were also greatly under-dosed.

The authors note that when LED occurs, the central tumour dose distribution is highly variable, and depends upon factors including LED level, beam energy, field size, lung density and tumour size. "In general, if full electron equilibrium conditions are maintained, dose will increase in the lung tumour due to greater X-ray beam penetration in low-density lung," Disher explained. "However, if equilibrium is disrupted, dose will generally be reduced within lung tumours. This decrease will be more severe for smaller tumours."

The researchers conclude by recommending that future SBRT clinical trials should only use dose calculation methods that account for electron scatter. These "category 2" algorithms include the Collapsed Cone Convolution and Analytic Anisotropic Algorithm, and more preferably, Monte Carlo techniques that explicitly track electron transport.

They emphasize that the detrimental effects of LED are most common for lung densities below 0.1 g/cm3. Thus, special attention should be given to lung-cancer patients with emphysema, which results in extremely low lung density.

"We have now characterized the radiation therapy conditions that create LED using a simple phantom and a single field of radiation," Disher told medicalphysicsweb. "We would like to extend this work to clinical examples using multiple radiation fields, patient geometries and tissue heterogeneity."

Related articles in PMB
An in-depth Monte Carlo study of lateral electron disequilibrium for small fields in ultra-low density lung: implications for modern radiation therapy
Brandon Disher et al Phys. Med. Biol. 57 1543
The effects of field-of-view and patient size on CT numbers from cone-beam computed tomography
Katrina Y T Seet et al Phys. Med. Biol. 54 6251
A 3D pencil-beam-based superposition algorithm for photon dose calculation in heterogeneous media
L Tillikainen et al Phys. Med. Biol. 53 3821