Patients treated with natural (leukocyte‐derived) interferon (IFN)‐α do not develop IFN antibodies
A. Österborg,K. Engman,M. Björeman,K. Carlson,E. Kimby,M. Ohrling,J. Samuelsson,E. Ösby,H. Mellstedt
DOI: https://doi.org/10.1111/j.1600-0609.1991.tb01561.x
1991-09-01
European Journal Of Haematology
Abstract:To the editor: In 1987 Dr von Wussow and colleagues reported the occurrence of neutralizing antibodies to recombinant interferon-alpha,, (rIFN-a,,) in 25 % of IFNtreated chronic myelogenous leukemia patients (1). The observation has later been confirmed by others (2,3). IFN antibodies may develop also upon therapy with rIFN-a,b (15%) (4) and lymphoblastoid IFN (4%) (5). Dianzani et al. (6) reported that among 354 patients treated with rIFN-a,, 17 % had neutralizing antibodies, as had 6% of those treated with rIFN-a,,. A higher number of patients had IFN-binding antibodies. Induction of IFN antibodies may render the patients refractory to further IFN therapy (1, 7). Thus, it should be of clinical importance to prevent the emergence of IFN antibodies. Here we report the incidence of IFN-binding antibodies in multiple myeloma (MM) patients treated with natural (leukocyte-derived) IFN-a as primary therapy . 44 patients with MM stages I1 or I11 were studied. The patients received melphalan/prednisone repeated every 6th week together with 7 x lo6 units (U)/m2 of natural IFN-a (Finnferon alpha, Red Cross Transfusion Service, Helsinki, Finland) S.C. for 5 consecutive days repeated every 3rd wk (8). At the time of clinical response (defined as a 50% reduction of the pretreatment serum M component concentration) the IFN dosage was changed to 3 x lo6 U daily 3 d a week continuously. Various numbers of treatment cycles were given during the induction period. At the time of test, 15 patients received induction therapy, 24 patients were on the continuous low-dose IFN schedule and in 5 patients IFN had been withdrawn due to side-effects or progression. The test was performed after a median IFN treatment period of 17 (3-5 1) months. The median cumulative dose of IFN was 846 (1902964) x lo6 U. Serum samples were collected between 72 h and 4 months after the last IFN administration. The presence of IFN-binding antibodies was analyzed by ELISA (BioNative, Ume$ Sweden). Briefly, serum samples were added to wells in micro-ELISA lates recoated with human leukocyte IFN immoklizedty goat polyclonal anti-IFN antibodies. After incubation for 1 h at room temperature (RT), the wells were washed and an alkaline phosphataseconjugated goat anti-human IgG was added for 2 h at RT. After washing x 4, substrate was added and absorbance was read at 405 nm after optimal color development. The absorbance values were corrected for unspecific binding by subtraction of the absorbance values in parallel control wells not coated with IFN. The accuracy of the ELISA test was confirmed by reference sera kindly provided by Drs P. von Wussow, Hannover, Germany, E. Lundgren, Ume$ Sweden and Hoffman-La Roche, Nutley, New Jersey, USA (data not shown). IFN-binding antibodies were not detected in any of the sera of the 44 MM patients, despite repeated S.C. injections and comparatively high cumulative doses of IFN. In 3 patients, IFN was withdrawn 1-4 months before test due to progression of the disease. The treatment failure on IFN in those patients was probably not due to the occurrence of IFN antibodies. MM patients have an impaired humoral immunity. It cannot be excluded that the absence of IFN antibodies in the present study might be due to the inability of the patients to mount an anti-IFN immune response. However, we do not believe that this is the case since IFN-binding antibodies have been detected in 5-7% of MM patients treated with recombinant IFN (9, 10). It is concluded that antibodies against natural IFN-a used for therapy might appear in a very low frequency in MM patients. Similar results have been seen in patients with neuroendocrine tumors treated with natural IFN-a (K. Oberg, pers. comm.). Although a proportion of patients with neutralizing antibodies to rIFN-a may respond to therapy with natural IFN-a, there are those who do not respond. Thus, considering the possibility of developing antibodies to therapeutically administered IFN and thereby a subsequent possible inhibition of the antitumor activity, natural IFN-a may be preferable from this point of view.