Session Item

Traditional treatment planning for esophageal cancer was mainly concentrated on target coverage and lung dose reduction while accepting relatively high dose to the heart. Recently a normal tissue complication probability (NTCP) model based indication protocol for treatment modality selection (i.e. photon or proton therapy) has been implemented in the Netherlands, in which a predicted 2 year overall survival (OS) increase of 5% is required to select for proton therapy. The NTCP model uses the volume of the GTV and the neo-adjuvant therapy status as clinical factors. The mean heart dose (MHD) is the only dosimetric parameter in the NCTP model. Consequently, a lower MHD directly leads to a higher predicted overall survival. Lowering MHD leads to an increase in lung dose with a possible increase of pulmonary toxicity. In this study, we investigated the trade-off between heart and lung dose sparing in photon radiotherapy for esophageal cancer patients. 

At least five plans with varying heart and lung doses were optimized for seven esophageal cancer patients treated with volumetric arc therapy in 23 or 28 fractions of 1.8 Gy (patient characteristics in table 1). While maintaining adequate target coverage (planning target volume D95≥98%), the following plans were created: (1) optimal lung sparing; (2) optimal heart sparing; (3) a range of plans, each with a step-wise increase of approximately 1 Gy mean lung dose (MLD), while lowering the MHD as much as possible. For all plans the estimated predicted 2 year overall survival was derived from the NTCP model. Also, plan robustness was assessed by recalculation of the plan on the inhale and exhale phase of the 4DCT. 

In total, 41 plans were optimized with varying lung and heart dose levels. All plans except the maximum heart sparing plans were sufficiently robust for breathing motion. As compared to the maximum lung sparing plan, the MHD could be reduced by approximately 9.5 Gy for a 1 Gy increase in MLD. On average, approximately 3 Gy increase in MLD resulted in 15.8 Gy reduction of MHD. Increasing the MLD more than 3 Gy did not lead to substantial better heart sparing in our patients (figure 1). The mean predicted 2 year overall survival of the maximal lung sparing plans was 38% (range 23–60%). Increasing the MLD with approximately 3 Gy resulted in an average predicted 2 year overall survival of 49% (range 32–77%), while the predicted 2 year overall survival of the maximum heart sparing plan was very similar: 52% (range 34–80%). 

The mean heart dose can be reduced substantially with only limited increase of mean lung dose. Increasing the mean lung dose with approximately 3 Gy relative to a maximum lung sparing esophageal VMAT plan is recommended to optimally reduce the MHD without compromising plan robustness.