FALSE-POSITIVE IMMUNOREACTIVITY WITH MUSCLE-SPECIFIC ACTINS IN NON-HODGKINS-LYMPHOMAS

Antibodies to forms of muscle actin have been developed that are highl y specific for tissue of myogenic origin and for which false-positive reactions are seldom reported. We studied 23 cases of B-cell malignant lymphoma by immunohistochemistry using monoclonal antiactin antibodie s of two different specificities. Eight of the 23 cases were positive with an antibody against smooth-muscle actin that had been produced as ascitic fluid. Reactivity was abolished by the use of antibody from t he same clone (1A4) but produced as a tissue culture supernatant. Nine of 23 cases were positive with an antibody against muscle actin (clon e HHF35) produced as ascitic fluid. An antibody from the same clone (H HF35), but produced as a tissue culture supernatant, gave weak stainin g of both lymphoid cells and connective tissue background. Lymphoma st aining by both HHF35 antibodies (and background staining) was abolishe d by the addition of ethylenediaminetetraacetic acid (EDTA) to the ant ibody. (EDTA is reported to eliminate nonspecific protein binding.) Ea ch ascitic fluid-derived antibody stained the lymphomas in an apparent ly specific manner. There was intense granular deposition of reaction product localized to specific portions of the cell (plasma membrane or diffuse cytoplasmic staining and, in two signet ring cell lymphomas, globular staining of intracytoplasmic inclusions), and there was subst antial concordance of the staining of the two antibodies. This finding of false-positive reactivity by two different antiactin monoclonal an tibodies, both prepared as ascitic fluid, supplements recent reports o f a similar phenomenon with ascitic fluid preparations of the antimela noma monoclonal antibody HMB45. It appears likely that ascitic fluid p reparations of some monoclonal antibodies may contain altered antibodi es or nonspecific reacting substances that are a potential source of d iagnostic error.