UTEROSACRAL SPACE INVOLVEMENT IN LOCALLY ADVANCED-CARCINOMA OF THE UTERINE CERVIX

Purpose: Radiation therapy is the standard management for locally adva nced cervical cancer, but it has not yielded fully satisfactory result s; a relatively high incidence of local failure remains. Standard radi ation therapy techniques combine external beam radiation and brachythe rapy generating a homogeneously composite dose distribution covering t he lateral parametria but may not be adequate in the uterosacral and p erirectal areas due to the spatial arrangement of intracavitary system and the constraints of rectal tolerance. We hypothesize that these do simetric characteristics might lead to a higher incidence of central/m arginal failures when the uterosacral space is involved by locally adv anced carcinoma of uterine cervix. Methods and Materials: Between Janu ary 1970 and December 1989, 343 patients with clinical Stage IIIB cerv ical cancer were treated at the Mallinckrodt Institute of Radiology wi th radiation therapy alone. We identified 83 patients with clinical ev idence of tumor in the uterosacral region; the remaining 260 patients either did not have uterosacral involvement or were unspecified. The d ose of external beam irradiation ranged from 18.02 to 33.20 Gy to the central pelvis and 48.22 to 59.40 Gy to the lateral parametrium. The a verage total dose, including brachytherapy contribution, to point A an d the lateral pelvis was 80.30 to 86.46 Gy and 60.50 to 73.40 Gy, resp ectively. External beam dose to the lateral parametria was, on average , 10 Gy higher in patients with uterosacral involvement. Results: We c ategorized the patterns of pelvic failure into central/marginal (inclu ding medial parametrium) and lateral parametria. The cumulative incide nce of central/marginal failure at 5 years was significantly higher in the group of patients with uterosacral involvement (36% compared with 21% for patients without uterosacral involvement or unspecified) (p = 0.002). Lateral parametrial failure was similar for patients with and without uterosacral involvement (39% and 38% at 5 years, respectively ) (p = 0.42). The actuarial incidence of distant metastasis was identi cal in the two groups: 46% at 5 years. Multivariate analysis confirmed that uterosacral space involvement increased the risk of pelvic recur rence (p = 0.044) and was the most significant factor that influenced the central/marginal pelvic failure (p = 0.002). Conclusions: Uterosac ral involvement by locally advanced carcinoma of the uterine cervix si gnificantly increased overall pelvic failure and was the most signific ant prognosticator of central/marginal pelvic failure. This is the res ult of the spatial constraints of the standard intracavitary geometry that deliver inadequate dose posteriorly to encompass the uterosacral space. Plausible ways to compensate the underdose in the uterosacral s pace include increasing whole pelvis dose without compromising the int racavitary brachytherapy dose, using a supplemental interstitial impla nt or adding a posterior oblique external beam boost. (C) 1998 Elsevie r Science Inc.