Prediction of splenic involvement in children with Hodgkin's disease. Significance of clinical and intraoperative findings. A retrospective statistical analysis of 154 patients in the German therapy study DAL-HD-78

Abstract
In 154 splenectomized children and adolescents with histologically proven Hodgkin's disease in the therapy study DAL-HD-78, the incidence of splenic involvement was 39%. In single-parameter analyses 6 of 16 examined pre- and intraoperative findings showed significant correlation to splenic involvement: B-symptoms, palpable splenic enlargement, mediastinal/lung hilus involvement, nodular changes of splenic surface, enlarged lymph nodes at splenic hilus/pancreatic tail, or enlargement of other upper-abdominal lymph nodes. The results of multivariant analyses (Cox regression model) of these six parameters showed that the two most significant intraoperative parameters—changes of splenic surface and enlargement of lymph nodes at splenic hilus/pancreatic tail-gave almost all of the information which can be obtained about splenic involvement. With these two parameters, an intraoperative decisional strategy for selective splenectomy has been developed which allows the omission of splenectomy in about two thirds of children with Hodgkin's disease while still obtaining detailed information about infradiaphragmatic spread of disease. Since minor splenic involvement remains undetected in about 10% of the nonsplenectomized patients (i.e., 6% of all patients), this method should be used only in combination with chemotherapy.