RESUMO
BACKGROUND: Although Hodgkin's lymphoma (HL) is a curable cancer, current treatment strategies based on risk stratification and response modulation are not precise enough. The predictive power of biological and morphological parameters is controversial, with prognostic models not reaching wide acceptance. PATIENTS AND METHODS: We analyzed the prognostic relevance of 8 parameters in 85 advanced stage classical HL patients, in order to determine whether tissue-based variables could add prognostic value to standard clinical parameters, thus contributing to better risk stratification at presentation. RESULTS: Univariate analysis confirmed 5 indicators of shorter overall survival (OS): Bcl-2 overexpression; increased CD68+ tumor-associated macrophages (TAM); international prognostic score (IPS) > 2; bulky disease; and total lymph node involvement (TLNI) with regard to neoplastic and inflammatory cells. Apart from TLNI, these parameters influenced lower event-free survival (EFS). Multivariate analysis identified 5 independent factors for OS: Bcl-2 overexpression; increased CD68+ TAM; TLNI; IPS > 2; and bulky disease. Increased CD68+ TAM, IPS > 2, and bulky disease affected the EFS. Utilizing the cumulative score of unfavorable prognostic factors for OS, we designed a prognostic model stratifying patients into 4 risk groups (with 0-1, 2, 3, or 4-5 factors), each with progressively reduced OS (p < 0.001). CONCLUSION: Our findings support the combination of tissue-based variables with clinical parameters at diagnosis, identifying patients who are at higher risk of poor outcome.