Dosimetric methods in the optimization of radiotherapy for carcinoma of the uterine cervix

Pelvic failures and late radiation sequelae were analyzed using the dosimetric parameters of ICRU Report 38 for 338 patients with Stage I-III carcinoma of the uterine cervix treated by radiation alone and followed for a minimum of 2 years. The pelvic recurrence rates were: Stage IB 5.1% (N = 118, 1% pelvis alone), Stage IIA 15.1% (N = 53, 9.4% pelvis alone), Stage IIB 15.8% (N = 76, 9.2% pelvis alone) and Stage IIIB 28.9% (N = 76, 17.1% pelvis alone). For Stages I and II pelvic failure was unrelated to cumulated lateral parametrial dose (CDPW) or reference volumes, but for Stage IIIB was higher for CDPW above 65 Gy. Overall complication rates were: grade 3-10.1% and grade 2-18.1% but were much lower for 176 patients treated with stem and ovoids (S + O: grade 3-5.7%, grade 2-15.7%) than for 43 receiving vaginal cylinders (grade 3-37.2%, grade 2-28%). Grade 3 rectal complications associated with cylinders were related to a maximal vaginal application over 1.50 cGy X m2 of total reference air kerma (or 2080 mgh) and cumulated rectal reference doses (CDRref) above 75 Gy. For the S + O group, grade 2 and 3 rectal complications increased with increasing reference volumes (hwt and HWT) and showed dose thresholds for CDRref and CDRmean (grade 3: 75 Gy). Prospective use of zones of risk defined graphically on a dose-volume plot (CDRref vs HWT) has reduced our severe complication rate without reducing local control. This technique requires individualization of patient therapy, rapid access to computerized dosimetry and the establishment of center- and applicator-specific risks of complications.