EP-1521 A HYBRID METHOD TO REDUCE LOW DOSE VOLUME OF CRITICAL ORGANS FOR BREAST IMRT: IMRT PLUS 3DCRT

Abstract

Materials and Methods: Ten LANPC patients, previously treated with conventional 3D conformal radiotherapy were selected and two additional plans created for each patient using IMRT and Varian’s VMAT solution, RapidArc® v8.6.14. The conventional plan was reoptimised using the same Varian EclipseTM Analytical Anisotropic Algorithm (AAA) as used for the IMRT and VMAT plans. The IMRT plans consisted of five fields with angles selected based on the position of the tumour and OARs. Two coplanar arcs were used for the VMAT plans, while the conventional plans contained either three or four static fields. The tumour position had a direct impact on the ability to meet the constraints. PTV conformity indices and dose-volume histograms (DVH) were used to compare techniques. OARs included kidneys, spinal cord, liver, duodenum and stomach. Results: Mean conformity indices of 1.04, 1.05 and 1.52 for VMAT, IMRT and conventional planning, respectively, suggest that VMAT and IMRT give superior coverage of the PTV. DVH data shows dramatic reduction in ipsilateral kidney doses using both VMAT and IMRT. However, there can be an increase in the dose to the contralateral kidney which is often spared using conventional methods. Of the patients analysed, 50% met the V20Gy constraint of 15% for the ipsilateral kidney using VMAT. Conventional and IMRT techniques only achieved this in 10% of patients. IMRT reduced the kidney doses compared with conventional methods but still exceeded 15%. The mean ipsilateral kidney doses for conventional, IMRT and VMAT were recorded as 17.2Gy, 14.3Gy and 15.1Gy respectively. Conclusions: VMAT and IMRT allowed better conformity of the PTV whilst minimising dose to the OARs. In some cases IMRT provided better contralateral kidney sparing than VMAT. However, the effect of dose bathing on the kidneys and other tissues is not well-known and therefore toxicity is difficult to predict. This technique is more time consuming to plan because the position of each beam is dependant on tumour location, which can result in more complex optimisation objectives. VMAT uses a greater number of angles and therefore a class solution is more feasible. This work will be continued using RapidArcTM v10.0.28 which will facilitate enhanced plan optimisation and hence, improvements relating to the VMAT statistics are anticipated. Future objectives include the use of 4DCT to individualise PTV margins with the intention of investigating dose escalation, and possibly hypo-fractionated treatment regimes for this site.