To evaluate how well antibodies to one asparaginase preparation predict or correlate with antibodies to another preparation in acute lymphoblastic leukemia (ALL) and lymphoma patients who did and did not have hypersensitivity reactions during chemotherapy. In all, 24 children with newly diagnosed ALL or lymphoma, who received Escherichia coli asparaginase 10 000 IU/m(2) IM thrice weekly for nine doses as part of multiagent induction and reinduction chemotherapy, and seven monthly doses during the first 7 months of continuation treatment, were studied. Plasma samples were collected at postinduction and at postreinduction. Six of 24 patients had no overt clinical reactions (nonreacting) and received only the E. coli preparation. Of these, 18 patients who had allergic reactions were switched to Erwinia asparaginase. A total of 18 patients had an anaphylactoid reaction to Erwinia asparaginase and were switched to receive polyethylene glycol (PEG) asparaginase. Antibody levels were measured by enzyme-linked immunoadsorbent assay against all the three asparaginase preparations. At postinduction, antibodies against E. coli were higher in reacting patients (0.063+/-0.066) than in nonreacting patients (0.019+/-0.013) (P=0.03). At postreinduction, anti-Erwinia antibodies were significantly higher in reacting patients (0.431+/-0.727) than in nonreacting patients (0.018+/-0.009) (P=0.007). Anti-E. coli antibodies correlated with anti-PEG antibodies at postinduction (r=0.714, P<0.001) and at postreinduction (r=0.914, P<0.001), but did not correlate with anti-Erwinia antibodies at postinduction (r=0.119, P=0.580) and at postreinduction (r=0.078, P=0.716). The results indicate a crossreactivity between patient antibodies raised against natural E. coli and PEG asparaginase but not Erwinia asparaginase.