Purpose: To determine independent prognostic factors in a group of 1875 patients with invasive carcinoma of the intact uterine cervix treated with radiotherapy alone in a French cooperative study from 1970 to 1993.
Materials and methods: Patients were staged according to the UICC-FIGO and MDAH substaging. The distribution per FIGO stage was Ia-Ib: 25.5%; IIa: 12%; IIb: 29%; IIIa: 5%; IIIb: 25%, and IV: 3.5%. Ninety-two percent had squamous cell carcinoma. The maximum diameter of the clinically detectable cervical disease was less than 3 cm in 24.5% of Stages I-II and in 10% of Stages III-IV, more than 5 cm in 13.5% of Stages I-II, and in 16% of Stages III-IV. Nodal involvement was shown on lymphangiogram in 16% of Stages I-II and in 32.5% of Stages III-IV.
Results: 1) Univariate analysis of Stages I and II: stage, cervical disease diameter, and nodal involvement are significant prognostic factors. Five-year specific survival rate (5ySS) is 83.5% in Stage Ib, 81% in IIa and 71% in IIb. Five-year disease-free survival rate (5yDFS) is 86% in tumors less of 3 cm, 76% in tumors of 3 to 5 cm, and 61.5% in tumor larger than 5 cm. Lymphangiogram strongly influences the 5-year pelvic disease-free survival rate (5yPDFS): respectively, 90% in nonpositive lymphangiogram vs. 65% when positive. A significant drop in specific and disease-free survival is observed (10 and 14%, respectively (p = 0.04) when comparing adenocarcinoma and squamous cell carcinoma. Age is a significant prognostic factor for specific survival because patients aged less than 30 years old have 91% vs. about 75% for patients over 30 years (p = 0.03). 2) Univariate analysis of Stages III-IV: Stage and positive lymphangiogram are predictive factors for relapse and death. The MDAH substaging is more reliable to predict the probability of pelvic disease-free survival in Stage III. At 5 years, the FIGO Stages IIIa and IIIb have a rather similar PDFS (65% vs. 59%). Conversely, the difference of survival rates between MDAH Stage IIIA and Stage IIIB is more demonstrative (69% vs. 47.5%). 3) Multivariate analysis (Cox P. H. R. model). Nodal involvement and stage remain significant for all three models in all stages (p < 0.0001). Age above 70 years influences specific survival for Stage I-II (p = 0.01). Tumors larger than 5 cm and adenocarcinoma also appear to be independent prognostic factors for specific and disease-free survival in Stage I-II (p = 0.05 and p = 0.005, respectively).
Conclusions: The relevance of tumor size (less or greater than 4 cm) is now recognized in the 1995 revised FIGO staging in Stage Ib but unfortunately not in other stages. Tumor size per stage and nodal status should be systematically recorded to allow a better prediction of failure rates and to compare literature reports.