Value of barium studies for predicting primary versus secondary non-Hodgkin's gastrointestinal lymphoma

Background: To determine the predictive value of a single lesion versus mul tifocal disease for differentiating primary and secondary gastrointestinal (GI) lymphoma on barium studies. Methods: Our study group consisted of 90 cases of non-Hodgkin's GI lymphoma from the radiologic archives of the Armed Forces Institute of Pathology in which barium studies had been performed. Each of those studies was reviewe d jointly by two of the authors to determine in a blinded fashion whether t he patients had a single lesion or multifocal disease (defined as two or mo re discrete lesions). Medical and pathologic records were then reviewed to determine whether these patients had primary (defined as lymphoma arising i n the GI tract) or secondary (defined as disseminated lymphoma with associa ted GI tract involvement) GI lymphoma. The final clinical diagnosis was the n correlated with the radiographic findings to test the hypothesis that a s ingle lesion is more likely to be associated with primary GI lymphoma and t hat multifocal disease is more likely to be associated with secondary GI ly mphoma. Results: Sixty-eight patients had a single lesion in the GI tract on barium studies; 52 (76%) of these patients had primary GI lymphoma and 16 (24%) h ad secondary GI lymphoma. Another 22 patients had multifocal disease on bar ium studies; 20 (91%) of these patients had secondary GI lymphoma and two ( 9%) had primary GI lymphoma. Conclusion: The predictive value of a single lesion for primary GI lymphoma on barium studies was 76%, and the predictive value of multifocal disease for secondary GI lymphoma was 91%. Thus, our data suggest that it is often possible to differentiate these two forms of GI lymphoma on the basis of th e radiographic findings.