PROGNOSTIC FACTORS FOR LOCAL AND DISTANT RECURRENCE IN STAGE-I AND STAGE-II CERVICAL-CARCINOMA

Purpose: The effects of tumor size, parametrial involvement, and other variables on treatment outcome for patients with Federation Internati onale de Gynecologic et d'Obstetrique (PICO) Stage I or II cervical ca rcinoma, as well as treatment complications, were analyzed retrospecti vely. Methods and Materials: Records of 125 patients with FIGO Stage I or II carcinoma of the uterine cervix selected for curative radiother apy between January 1980 and December 1990 were reviewed. Twelve patie nts (9.9%) underwent adjuvant extrafascial hysterectomy and 8 patients (6.4%) received chemotherapy. Median age was 55 years. Median follow- up time was 40 months, and minimum follow-up time was 24 months. The d ata were analyzed for site of first relapse, survival, overall inciden ce of complications, and incidence of grade 4 complications. Results: The overall 5-year survival was: Stage IA: 100%, Stage IB: 72%, Stage IIA: 90%, and Stage IIB: 72%. The 5-year survival with no evidence of disease (NED) was: Stage IA: 100%, Stage IB: 67%, Stage IIA: 90%, and Stage IIB: 50%. Patients with bulky (> 5 cm) tumors had a shorter over all and NED survival than patients with nonbulky tumors (53% vs. 83%; p = 0.0008 and 44% vs. 78%; p = 0.0001, respectively). Thirty-nine tum or recurrences (39 out of 125 = 31%) occurred and were scored as local (23 out of 125 = 18.3%), if initial failure had a local component, or distant (16 out of 125 = 12.7%), if initial failure was distant only, Patients with bulky (more than 5 cm) tumors (32 out of 125) were more likely to experience a recurrence (18 out of 32 = 56%) than patients with nonbulky tumors (21 out of 93 = 22%; p = 0.0004). The initial sit e of recurrence was more likely to be local for bulky tumors (14 out o f 18 = 78%) than for nonbulky tumors (9 out of 21 = 43%;p = 0.03). The probability of a recurrence increased with the number of involved par ametria (none: 20 out of 78 = 25%; one: 12 out of 34 = 35%; two: 7 out of 13 = 54%;p = 0.04 for linear trend), as did the probability that t he initial failure was distant rather than local (none: 4 out of 20 = 20%; one: 7 out of 12 = 58%; two: 5 out of 7 = 71%; p = 0.01 for linea r trend). Positive lymph nodes, vessel invasion, and low :hemoglobin l evel all correlated with an increased risk of a recurrence (RR 2.41, p = 0.004; RR 2.20, p = 0.01; OR 2.02, p = 0.01, respectively). There w ere 46 complications among 37 (29%) patients. The incidence of grade 4 complications was 8.8% (11 out of 125). History of pelvic surgery and bulky tumor were significant predictors of a grade 4 complication (p < 0.0001 and 0.021, respectively). Also, a dose rate to point A of > 0 .6 Gy/h increased the chance of a grade 4 complication (p = 0.007). Co nclusion: For patients with PICO Stage I or II cervical carcinoma, tum or size was more predictive of local recurrence than was overall stage , and the extent of parametrial involvement was strongly predictive of distant recurrence, as was the stage. These findings suggest that tum or size and extent of parametrial involvement should be incorporated i nto the staging system. Patients with bulky tumors had a shorter survi val and were more likely to experience a grade 4 toxicity of therapy. Dose rate to point A of > 0.6Gy/h was associated with the increased ri sk of grade 4 complications.