Purpose: To determine the influence of observer variation and treatment planner variation on the dose delivered to the target and normal structures when irradiating paranasal sinus carcinomas.
Patients and methods: Nine patients with paranasal sinus tumors underwent debulking surgery and subsequent radiation therapy. Two observers from two different institutions delineated the clinical target volumes (CTVs) for the elective and the boost volumes. These volumes were expanded in three dimensions with a 5-mm margin. At both institutions, a three-dimensional conformal treatment plan of 46 Gy to the elective volumes, plus 20 Gy to the boost target volumes, was designed. The delineated volumes and treatment plans were compared.
Results: The mean volume ratio between institutions of the elective CTVs was 0.9 (standard error = 0.05). The differences were located mainly at the bottom of the nasal cavity and at the frontal border of the target areas. The differences in boost CTVs were large; the mean volume ratio was 2.6 (standard error = 0.58). After expansion of the CTV, the mean distance between the planning target volume (PTV) and the chiasm differed by 0.5 cm between the two institutions. Cases with smaller distances between the PTV and the chiasm had more underdosage to the PTV. This effect was less pronounced for institution A (1 vol.%/cm) than for institution B (10 vol.%/cm) treatment plans, which were less conformal. When the treatment plan was designed for the PTV of institution B, 23 volume % of the PTV of institution A received <95% of the prescribed dose. If the treatment plan was designed for the (on average larger) PTV of institution A, the underdosed volume of PTV at institution B was 17%. The relative underdosage to the "other" PTV was larger when the original treatment plan was more conformal.
Conclusion: In the irradiation of paranasal sinus cancer, both the treatment planner and the observer have a significant influence on the dose to the target and organs at risk. Both effects are similar in magnitude. The observer effect increases with more conformal treatment plans. Minimizing the observer variation is important for adequate irradiation of paranasal sinus cancer.