Modeling Variability for at Risk Lung Functionality by Location of Esophageal Cancer Radiation Therapy Fields

Abstract

Dose volume histogram (DVH) parameters are used to assess lung tolerance for esophageal cancer radiotherapy. Although lung functionality is heterogeneous, DVH evaluation treats all lungs as equivalent. We modeled variability of ventilation in lung exposed to radiotherapy by site of esophageal cancer. We developed simulated 3D conformal 4-field radiation treatment plans for gastroesophageal junction and mid-thoracic esophagus patients from 9 individual 4D CT scans. Standard expansions were used (4 cm superiorly/inferiorly and 1cm radially for clinical target volume and then 5mm in all directions for planning target volume (PTV)). Radiation dose was 5,040 cGy in 28 fractions to the PTV. We then determined points in the lung that were at the superior and inferior edges of the radiation fields for each plan and measured the displacement of those points from full exhalation to full inhalation as a surrogate for lung expansion. P-values were determined through the Wilcoxon Rank Sum test and t-tests. Lung displacement as measured by points at the superior and inferior edges of the gastroesophageal and mid-thoracic fields were 12.1 mm (± 6.7), 19.2 mm (± 5.9), 3.8 mm (± 1.3), and 7.1 mm (± 2.5), respectively, between full exhalation and full inhalation. Points in the superior aspect of the gastroesophageal fields moved significantly more than points in the superior aspect of the mid-thoracic fields (p=0.0078). Displacement was generally lateral and caudal in direction. There was a difference in displacement in the simulated tumors as well, with the mid-thoracic tumors moving a mean distance of 6.7 mm caudally and the gastroesophageal tumors moving a mean distance of 10.0 mm ventrally and caudally (p=0.039). There was also a difference in dose to the inferior aspects of the lung that displaced the most during the breathing cycle between the gastroesophageal and mid-thoracic fields, 3,541 cGy and 580 cGy, respectively (p=0.0001). This difference in dose to the most expansive part of the lung was not well expressed through the mean lung dose, V5, V10, or V30. The V5 to the total lung volume was higher for mid-thoracic plans than for gastroesophageal plans (63% vs 43%; p=0.0078) as was V10 (52.2% vs 36.9%; p=0.0039), while V30 was higher for gastroesophageal plans (17% vs 11%; p=0.039). The mean lung dose was numerically higher at 1,540 cGy for mid-thoracic plans versus 1,320 cGy for gastroesophageal plans although not statistically significant (p=0.18). Gastroesophageal junction fields have superior lung DVH parameters but are treating more ventilated lung versus mid-thoracic esophagus fields. Further study of differences in pulmonary outcomes based on location of tumor and ventilation of treated lung are warranted. In conjunction with tumor movement results, gastroesophageal junction tumors would appear to be more amenable to advanced planning techniques managing movement and functional lung avoidance.