Room: Exhibit Hall | Forum 4
Purpose: The need to spare functional lung tissue significantly complicates treatment planning for malignant pleural mesothelioma for radiation after a pleurectomy/decortication. Correlations between patient-specific volume and dosimetric parameters were explored to predict achievable dose constraints and guide future treatment planning.
Methods: Fifty-two consecutive mesothelioma patients (Right:32, Left:20) receiving 45Gy in 25 fractions on a Tomotherapy platform were included in this study. Thirty dose and twenty volume parameters were extracted from the original clinical plans, including the planning target volume (PTV), ipsilateral, contralateral and total lung with and without PTV, heart, esophagus, liver, cord, kidneys, and stomach. LASSO regression analysis, followed by a polishing step without L1 regularization was performed to select the top features for a model correlating patient-specific volume parameters to each dose metric. To examine whether there is significant difference in dosimetric outcome between left- and right-sided treatments, student t-test was performed between the two groups for all dosimetric parameters.
Results: Total lungâ€“PTV mean dose is the best predicted dose parameter for all three tested patient subgroups, achieving coefficients-of-determination (RÂ²) of 0.5256, 0.5609, and 0.5109 for all left-, and right-sided patient subgroups, respectively. LASSO regression resulted in dose prediction model: Total Lung-PTV mean dose = 0.005 Ã— Ipsilaeral Lung-PTV+3.264E-4 Ã— Liver +10.049 from all patients. For the left-sided subgroup, cord volume dominated over liver volume due to the reduction in liver involvement for left-sided cases. As expected, the ipsilateral lung V20Gy is strongly correlated with heart volume for the left-sided cohort. Statistically significant differences (p<0.05) between left- and right-sided cohorts were found for the mean and V20Gy for Ipsilateral Lung and Ipsilateral Lung-PTV, and the mean dose and V30Gy of the heart and liver.
Conclusion: Robust correlations between PTV/lung volumes and achievable dosimetric parameters in mesothelioma cases are identified. The proposed predictive models could guide future treatment planning.