Objective: New predictive factors for bladder tumor progression have been analyzed in many publications, often with contradictory results. Very few papers have referred specifically to T1G3 tumors. Our objective was to find new, clinically useful markers which either alone or in association with classical prognostic factors would allow the early selection of the correct therapeutic approach.
Material and methods: This was a retrospective study of 83 patients with T1G3 bladder tumors who were initially treated with transurethral resection + bacillus Calmette-Guérin therapy, with a minimum follow-up period of 3 years. We analyzed eight variables. New factors considered were: the level of submucosal invasion; microvessel density; and immunostaining for Ki-67 and p53. Independent prognostic variables for progression were established using logistic regression analysis, and risk groups were created from mathematical models.
Results: Five variables were determined as unfavorable: tumor multiplicity; tumor size >3 cm; carcinoma in situ; T1b substage; and p53 positivity. The first three factors predicted progression in only 32% of cases, while the addition of the new prognostic factors (T1b substage and p53 positivity) increased this rate to 65%. We established four risk groups, with rates of progression of 67% and 100% in the high-risk and very high-risk groups, respectively. For inclusion in these groups, both new predictive factors had to be unfavorable; if either one were absent then the three classical factors had to be present.
Conclusions: Microstaging and p53 positivity have a prognostic value for predicting progression in T1G3 tumors, providing 33% more information than that obtained with classical prognostic factors alone. The application of mathematical models identifies risk groups and allows the use of an early and more aggressive treatment.