Background: It was the aim of the quality control program of the randomized trial HIT '91 (intensive chemotherapy before irradiation versus maintenance chemotherapy after irradiation) to assess prospectively the quality of neuroaxis irradiation with respect to the protocol guidelines and to evaluate acute toxicity with respect to treatment arm.
Patients, materials and methods: Data of 134 patients undergoing irradiation of the craniospinal axis were available. Positioning aids, shielding techniques, treatment machines, choice of energy, total dose and fractionation were evaluated. A total of 651 simulation and verification films were analyzed to assess the coverage of the clinical target volume (whole brain, posterior fossa, sacral nerve roots) and deviations of field alignment between simulation and verification of first treatment. Field matching between whole brain and adjacent cranial spinal fields was analyzed with respect to site and width of junction. Acute maximal side effects were evaluated according to a modified WHO score for neurotoxicity, infections, skin, mucosa and myelotoxicity.
Results: In 91.3% of patients contemporary positioning aids and individualized shielding techniques were used to assure a reproducible treatment. In 98 patients (73.1%) linear accelerators and in 36 patients (26.8%) Cobalt machines were used. Single and total dose were administered according to the protocol guidelines in more than 90% of patients. In 20.2% of patients the cribriform plate, in 1.4% the middle cranial fossa and in 21.1% the posterior fossa and in 4.5% the 2nd sacral segment were incompletely encompassed by the treatment portals. Ninety-five percent of deviations of field alignment were less than 13.0 mm (whole brain) and 12 mm (cranial spinal field) with a random error between 4.9 and 7.6 mm (whole brain) and 6.9 mm and 9.9 mm (spinal canal), respectively. In 77.5% of patients the junctions between whole brain and cranial spinal fields were placed without a gap. A gap between 5 and 10 mm was left in 15 patients (18.7%), exceeding 10 mm in 3 patients. Acute neurotoxicity and skin reactions were mild, the rate of infections was low in both treatment arms. However, myelotoxicity resulted in interruptions of radiotherapy in 31.9% after intensive chemotherapy as compared to 20.0% without preceding chemotherapy.
Conclusions: In the HIT '91 trial a precise radiotherapy of craniospinal axis has been performed in the majority of patients. Our findings indicate that the high quality is possibly an important contributing factor for the therapeutic outcome. However, preceding intensive chemotherapy caused marked toxicity of subsequent irradiation leading to a high rate of interruptions. Our database is subject to a future analysis of recurrences.