EFFECT OF DOSE-RATE ON LOCAL-CONTROL AND COMPLICATIONS IN DEFINITIVE IRRADIATION OF T1-2 SQUAMOUS-CELL CARCINOMAS OF MOBILE TONGUE AND FLOOR OF MOUTH WITH INTERSTITIAL IR-192

From 1971 to 1988, 134 T1 and 145 T2 biopsy-proven squamous cell carcinomas of mobile tongue and floor of mouth were definitively managed by iridium-192. Implantations were performed using either guide gutters or afterloading plastic catheters. The prescribed dose at the reference isodose (85% of the basal dose rate, Paris system) was 60-70 Gy. Total dose was not adjusted to dose rate or tumor volume. Results of the 279 implants have been analysed to look for a possible influence of dose rate on local control and necrosis. Follow-up patients free of local recurrence is 1-180 months with average of 51 months. The 279 tumors were divided in four groups according to dose and dose rate: greater-than-or-equal-to 62.5 Gy and greater-than-or-equal-to 0.5 Gy/h (n = 130), greater-than-or-equal-to 62.5 Gy and < 0.5 Gy/h (n = 36), < 62.5 Gy and greater-than-or-equal-to 0.5 Gy/h (n = 81), < 62.5 Gy and < 0.5 Gy/h (n = 32). The four groups were comparable according to age, sex, tumor diameter and macroscopic aspect. At 5 years, the estimated local control (Kaplan Meier) was 93, 87, 79 and 52%, respectively (dose adjusted to dose rate: p < 0.001, dose rate adjusted to dose: p < 0.01, Log-rank); the necrosis rate was 44, 24, 37 and 5%, respectively (dose adjusted to dose rate: p = 0.08, dose rate adjusted to dose: p < 0.01). Univariate comparison of local results in these four groups has shown that: (1) Local control is significantly related to dose independently of dose rate; the local control dose effect curve reaches a plateau above 62.5 Gy approximately, while reducing the dose below 62.5 Gy leads to a rapid increase of local failure. (2) Necrosis is significantly related to dose when the dose rate is less than 0.5 Gy/h. (3) Local control is dose-rate-dependent only below 62.5 Gy. (4) Necrosis is related to dose rate independently of dose. (5) Dose correction may lead to an increase of local failure rate without reduction of necrosis. (6) Dose rate correction may reduce necrosis rate without reducing local control. Multivariate analysis revealed that dose, dose rate and tumor diameter significantly influenced local control, while only tumor diameter and tumor site influenced the risk of necrosis; there was a trend to a significant relationship between dose rate and necrosis (p = 0.058). To maximize local control and minimize necrosis, the authors recommend a dose of 65-70 Gy at a dose rate of 0.3-0.5 Gy/h.