Comparison of intensity modulated radiation therapy (IMRT) treatment techniques for nasopharyngeal carcinoma

We studied target volume coverage and normal tissue sparing of serial tomot herapy intensity modulated radiation therapy (IMRT) and fixed-field IMRT fo r nasopharyngeal carcinoma (NPC), as compared with those of conventional be am arrangements. Twelve patients with NPC (T2-4N1-3M0) at Mallinckrodt Inst itute of Radiology underwent computed tomography simulation. Images were th en transferred to a virtual simulation workstation computer for target cont ouring. Target gross tumor volumes (GTV) were primary nasopharyngeal tumor (GTV(NP)) with a prescription of 70 Gy, grossly enlarged cervical nodes (GT V(LN)) with a prescription of 70 Gy, and the uninvolved cervical lymphatics [designated as the clinical tumor volume (CTV)] with a prescription of 60 Gy. Critical organs, including the parotid gland, spinal cord, brain stem, mandible, and pituitary gland, were also delineated. Conventional beam arra ngements were designed following the guidelines of Intergroup (SWOG, RTOG, ECOG) NPC Study 0099 in which the dose was prescribed to the central axis a nd the target volumes were aimed to receive the prescribed dose +/- 10%. Si milar dosimetric criteria were used to assess the target volume coverage ca pability of IMRT. Serial tomotherapy IMRT was planned using a 0.86-cm wide multivane collimator, while a dynamic multileaf collimator system with five equally spaced fixed gantry angles was designated for fixed-beam IMRT. The fractional volume of each critical organ that received a certain predefine d threshold dose was obtained from dose-volume histograms of each organ in either the three-dimensional or IMRT treatment planning computer systems. S tatistical analysis (paired t-test) was used to examine statistical signifi cance. We found that serial tomotherapy achieved similar target volume cove rage as conventional techniques (97.8 +/- 2.3% vs. 98.9 +/- 1.3%). The stat ic-field IMRT technique (five equally spaced fields) was inferior, with 92. 1 +/- 8.6% fractional GTV(NP) receiving 70 Gy m 10% dose (P < 0.05). Howeve r, GTV,, coverage of 70 Gy was significantly better with both IMRT techniqu es (96.1 <plus/minus> 3.2%, 87.7 +/- 10.6%, and 42.2 +/- 21% for tomotherap y, fixed-field IMRT, and conventional therapy, respectively). CTV coverage of 60 Gy was also significantly better with the IMRT techniques. Parotid gl and sparing was quantified by evaluating the fractional volume of parotid g land receiving more than 30 Gy; 66.6 +/- 15%, 48.3 +/- 4%, and 93 +/- 10% o f the parotid volume received more than 30 Gy using tomotherapy, fixed-fiel d IMRT, and conventional therapy, respectively (P c 0.05). Fixed-field IMRT technique had the best parotid-sparing effect despite less desirable targe t coverage. The pituitary gland, mandible, spinal cord, and brain stem were also better spared by both IMRT techniques. These encouraging dosimetric r esults substantiate the theoretical advantage of inverse-planning IMRT in t he management of NPC. We showed that target coverage of the primary tumor w as maintained and nodal coverage was improved, as compared with conventiona l beam arrangements. The ability of IMRT to spare the parotid glands is exc iting, and a prospective clinical study is currently underway at our instit ution to address the optimal parotid dose-volume needs to be spared to prev ent xerostomia and to improve the quality of life in patients with NPC, (C) 2001 Wiley-Liss, Inc.