Importance of protocol target definition on the ability to spare normal tissue: An IMRT and 3D-CRT planning comparison for intraorbital tumors

Abstract

We selected five intraorbital tumor sites that are frequently found in clinical practice in children diagnosed with orbital rhabdomyosarcoma and performed three-dimensional conformal radiotherapy (3D-CRT) and intensity-modulated photon radiotherapy (IMRT) planning. Results of target coverage and doses to critical structures were compared. The goal of this study was to evaluate and to document realistic expectations as to organ-sparing capabilities of modern radiation therapy planning technologies with a focus on lens-sparing irradiation. Furthermore, we investigated potential added benefits of IMRT compared with 3D-CRT and the influence of protocol volume criteria definitions on the ability to obtain normal tissue dose sparing using the orbit as an example of a complex anatomic site. The five intraorbital tumor sites were placed retrobulbar, temporal, nasal, in the upper inner and upper outer quadrant, the latter two more complex in shape. Gross tumor volume (GTV), clinical target volume (CTV), and planning target volume (PTV) were defined in image-fused computed tomography and magnetic resonance data sets. 3D-CRT and IMRT photon plans, using equal beam angles and collimation for direct comparison, were designed to 45 Gy prescription dose according to Intergroup Rhabdomyosarcoma Study Group-D9602 (IRSG-D9602) protocol (Intergroup Rhabdomyosarcoma Study V [IRS-V] protocol) for Stage I, Clinical Group 3 orbital rhabdomyosarcoma. To compare the impact of changed target definitions in IMRT planning, additional IMRT plans were generated using modified volume and dose coverage criteria. The minimum dose constraint (95%) of the PTV was substituted by a required minimum volume coverage (95%) with the prescribed dose. Dose-volume histograms (DVHs) were obtained, including target volumes, lens, optic nerves, optic chiasm, lacrimal gland, bony orbit, pituitary gland, frontal and temporal lobes. Protocol target volume coverage criteria were fulfilled in all cases (5/5) with 3D-CRT and IMRT. Using the protocol criteria, lens sparing was achieved only for two tumor sites (retrobulbar and lateral position) with either planning technique. Mean lens doses were 8.5 and 10.4 Gy for 3D-CRT and 7.5 and 13.2 Gy for IMRT, respectively. The mean lens doses for the other three tumor locations averaged 26.8 Gy. IMRT plans reduced the lens dose in four of five cases by an average of 2.6 Gy compared with 3D-CRT. Modified target protocol prescription markedly reduced mean lens doses by 23-50% and by as much as 18 Gy. Recorded mean lens doses after protocol modification were 26% lower using IMRT plans compared with 3D-CRT. The cold spot as a result of the relaxed volume coverage requirements was within 2% of the original protocol criteria and located at the edge of the PTV, outside the CTV. Compared with 3D-CRT, IMRT resulted in an increase of brain volume receiving 10% (V10) and 20% (V20) of the prescribed dose. Strict adherence to IRS-V protocol criteria prohibits at present lens sparing within compliance criteria for the majority of intraorbital tumor locations because of protocol-specific CTV and PTV target definitions. Changing protocol definitions by prescribing to the volume rather than to a dose constraint, IMRT planning significantly reduced lens doses. This was not accomplished to the same degree with 3D-CRT. Our study underlines the importance of appropriate selection of planning objectives to maximize the specific capabilities and advantages of IMRT in terms of sufficient target coverage and simultaneous sparing of critical structures. Our results can add to the ongoing discussion in the design of future 3D-CRT/IMRT protocols.