Mediastinal bulk in Hodgkin disease. Method of measurement versus prognosis.
The presence of a large mediastinal mass (bulk disease) in patients with newly diagnosed Hodgkin disease is believed by many to predict a poorer prognosis and to warrant more aggressive treatment. These masses are formed by an aggregate of mediastinal lymph nodes. The determination of bulk disease is confusing, with at least 27 definitions having been proposed. This study seeks to determine the best definition, and determine the role of thoracic computed tomography (CT) versus chest radiographs in the evaluation of mediastinal bulk disease. One hundred seven consecutive newly diagnosed adult patients with Hodgkin disease were evaluated using 13 commonly used definitions of mediastinal bulk. Of the 76 patients with mediastinal disease, 73 had bulk disease as defined by at least one definition. Of the 16 patients who had recurrence of mediastinal disease, only the presence of bulk disease according to one definition (hilar adenopathy, greater than or equal to 2 cm) was statistically significant in its prediction (P = .05). No definition based on the size of the mediastinal nodal mass reliably predicted those patients with recurrence. No differences in our data were found for differing stages or disease cell types, the presence of extension, or with differing treatment regimens. This study highlights the confusion and controversy surrounding the use of bulk disease of the mediastinum as an adverse prognostic indicator. The numerous methods of measuring mediastinal bulk in patients with newly diagnosed Hodgkin disease are confusing, overlap, and are not statistically reliable in predicting recurrence. Efforts to create a standard or ideal definition were unsuccessful. Thoracic CT was useful in those patients whose bulk disease distorted only one side of the mediastinal silhouette on chest radiographs.
Hopper, KD; Diehl, LF; Lynch, JC; McCauslin, MA
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)