Fatores prognósticos em crianças e adolescentes portadores de Linfoma Não-Hodgkin.

Abstract

This prospective/retrospective study comprises 98 children with non-Hodgkins lymphoma diagnosed between 1981 and 2006 (26 years) at the Hospital das Clínicas, Federal University of Minas Gerais. Prognostic factors for surviving the disease were analyzed. The immunohistochemistry contribution to increase the accuracy of the diagnosis was also evaluated. The majority of patients were males (male/female ratio 2.3). The median age at diagnosis was 70 months (range 8.4 to 189.5 months). Malnutrition was defined as a Z score -2 (2 standard deviation below the median for the reference WHO population). Taking into consideration the weight-for-age Z score, 20.2% of children were considered malnourished; for the height-for-age Z score the frequency was 8.2%. An abdominal mass was the commonest form of presentation (49%). WHO histologic classification and the staging devised by St. Jude Childrens Research Hospital was used throughout. Burkitts lymphoma was the most frequent histologic type (52%) and the majority of patients presented as advanced stage at diagnosis (70.2%). Follow up varied from 0.1 to 26.3 years (median 6.1y) for those who did not die. Complete clinical remission rate was 83.7%. The estimated probabilities of overall survival and event free survival at 5 years were 73% (SE 4.4%) and 72% (SE 4.7%), respectively. Death during induction (n = 14) were due to infections and severe metabolic disturbances. Relapses occurred in 5 patients, all localized to central nervous system (4 isolated and one combined with marrow relapse). Up to 1987 modified Sloan-Kettering LSA2L2 protocol was adopted for treatment (n = 20). From 1987 onwards BFM-based protocols were employed. In univariate analysis significant prognostic factors for dying were increased levels of serum uric acid and urea at diagnosis (P < 0.001 and P = 0.03, respectively), advanced stage (P = 0.014) and the way histologic specimen for diagnosis was obtained (partial versus total ressection; P = 0.03). Gender, age, nutritional status, histologic type, serum levels of LDH, type of treatment protocol, or response to treatment were not statistically significant prognostic factors. Improving the accuracy of the morphologic diagnosis with immunohistochemistry was not either associated with a higher probability of survival.