CARCINOMA OF THE UTERINE CERVIX .1. IMPACT OF PROLONGATION OF OVERALLTREATMENT TIME AND TIMING OF BRACHYTHERAPY ON OUTCOME OF RADIATION-THERAPY

Purpose: Some studies have described decreased pelvic tumor control an d survival rates in invasive carcinoma of uterine cervix when the over all time in a course of definitive irradiation is prolonged. We attemp t to confirm or deny these observations and evaluate the impact of tim ing of brachytherapy on outcome. We also explore the hypothesis that m ore extensive tumors technically require prolongation of the course of irradiation; thus, decreased tumor control and survival in these pati ents may not necessarily be the result of time/dose factor. Methods an d Materials: Records of 1224 patients (Stage IB to III) treated with d efinitive irradiation (combination of external beam and two intracavit ary insertions to deliver doses of 70 to 90 Gy to point A) were review ed. Follow-up was obtained in 97% of the patients (median, 12 years; m inimum, 3 years; maximum, 28 years). The relationship between outcome and overall treatment time and time of intracavitary insertions was an alyzed in each stage and according to tumor size/extent. Results: Ther e was strong correlation between overall treatment time (OTT) and tumo r stage (less than or equal to 7 weeks: 81% for Stage IB; 74% for Stag e IIA; 52% for Stage IIB; and 47% for Stage III). Interruptions of the rapy accounting for prolongation of treatment time occurred in 25-30% of patients, most frequently because of holidays and weekends and side effects of therapy. Overall treatment time had a major impact on pelv ic tumor control in Stages IB, IIA, and IIB; in Stage IB 10-year actua rial pelvic failure rates were 7% with OTT less than or equal to 7 wee ks, 22% with 7.1 to 9 weeks, and 36% with >9 weeks (p less than or equ al to 0.01). For Stage IIA the corresponding values were 14%, 27%, and 36% (p=0.08), and in Stage IIB pelvic failure rates were 20%, 28%, an d 34%, respectively (p=0.09). In Stage III, pelvic failure was 30%, 40 %, and 50%, respectively (p = 0.08). There was also a strong correlati on between OTT and 10-year cause-specific survival (CSS); in Stage IB rates were 86% with OTT of less than or equal to 7 weeks, 78% for 7.1 to 9 weeks, and 55% for greater than or equal to 9 weeks (p<0.01). The corresponding rates in Stage IIA were 73%, 41%, and 48% (p less than or equal to 0.01). For patients with Stage IIB, CSS rates were 72% for OTT less than or equal to 7 weeks, 60% for 7.1 to 9 weeks, and 70% fo r >9 weeks (p=0.01). Patients with Stage III disease had 45% 10-year C SS when treatment was delivered in 9 weeks or less and 36% for longer overall times (p = 0.16). In multivariate analysis of patients with St age IB and IIA, OTT and clinical stage were the most important prognos tic factors for pelvic tumor control, disease-free survival, and CSS. Tumor size was a prognostic factor for CSS. In Stages IIB and III, OTT , clinical stage, unilateral or bilateral parametrial invasion, and do se to point A were significant prognostic factors for pelvic tumor con trol, disease-free survival, and CSS. Prolongation of time had a signi ficant impact on pelvic tumor control and CSS regardless of tumor size , except in Stage IB tumors less than or equal to 3 cm. Regression ana lysis confirms previous reports that prolongation of OTT results in de creased pelvic tumor control rate of 0.85% per day for all patients, 0 .37% per day in Stages IB and IIA, 0.68% per day in Stage IIB, and 0.5 4% for Stage III patients treated with greater than or equal to 85 Gy to point A. Performance of all intracavitary insertions within 4.5 wee ks from initiation of irradiation yielded decreased pelvic failure rat es in some groups of patients (8.8 vs. 18% in Stage IB and IIA tumors less than or equal to 4 cm and 12.3 vs. 35% in Stage IIB) (p less than or equal to 0.01). Conclusions: Prolongation of treatment time in pat ients with Stage IB, IIA, IIB, and III carcinoma of the uterine cervix has a significant impact on pelvic tumor control and CSS. The effect of OTT was present regardless of tumor size except in Stage IB tumors less than or equal to 3 cm. This may be related to biologic factors su ch as cell repopulation and increased proliferation resulting from tre atment interruptions, in addition to initial clonogenic cell burden. I rradiation for patients with invasive carcinoma of the cervix should b e delivered in the shortest possible overall time.