Roferon (rIFN-α2a) Is More Immunogenic Than Intron A (rIFN-α2b) in Patients with Chronic Myelogenous Leukemia

number of genetically engineered cytokines, including recombinant interferons (rIFNs), have become a option for the treatment of various malignant and infectious diseases. As well as natural IFN-a preparations, both rIFN-a2a and rIFN-a2b have proved to induce hematological remissions in most patients with chronic myelogenous leukemia (CML), and also to suppress to a significant extent the Philadelphia chromosome-positive cell clone in a minority of patients,'1'2) a result so far otherwise achieved only by bone marrow transplantation. Recent results from the Italian Co-operative Study Group on CML'3> even suggest that continued administration of an IFN-a preparation prolongs the duration of the chronic phase of CML, and so increases the life span of patients with this disease. It has been found that when cancer patients are treated with rIFN-a2, high-titered IFN neutralizing antibodies against this cytokine develop in a significant minority. Importantly, such high-titered antibodies are associated with reduced biological activity of the rIFN-a2 and loss of clinical response.'4) IFNs are not unique in this respect: recombinant preparations of human insulin, growth hormone, interleukin-2 (IL-2), and granulocytemacrophage colony-stimulating factor (GM-CSF) have also been reported to induce an immune response in a proportion of patients treated with these agents.'56) The reasons why these homologous recombinant proteins are immunogenic in man is not yet known, although there are some clues.'6) To compare the immunogenic potential of the two clinically available recombinant preparations of IFN-a2, 164 newly diagnosed CML patients enrolled in a single large multicenter trial were treated under identical conditions with one or the other, and their sera were screened for the emergence of antibodies against IFN-a2. In this prospective, randomized, three-armed multicenter trial, patients entered into the IFN arm were treated with either Roferon (rIFN-a2a, Hoffman LaRoche, Basel) or Intron A (rIFN-a2b, Essex Pharma, Munich) at a dose of 5 Mu/m2 daily. Before participating in this study, the principal investigator in each center chose which of the two rIFN-a2 preparations would be used to treat all the patients enrolled into the study at his/her institution, and only this one was used. In the whole study, the two groups of IFN-a2-treated patients did not differ statistically in terms of age, sex, initial leukocyte and platelet counts, Karnofsky index, and the proportion of those who were Philadelphia chromosome positive. Five of the original 164 patients were excluded from further analysis, 4 patients because of wrong diagnosis, and 1 due to deficient documentation. The remaining 159 patients were evaluable: 106 of these were treated with rIFN-a2a, and 53 with rlFNa2b. Twenty-three of these received rIFN-a2 treatment for less than 3 months, 7 patients because of intolerance of the side effects, 14 patients gained no clinical benefit, and 2 patients withdrew from the study at their own wish. Serum from 105 patients were screened for the presence of both binding antibodies and neutralizing antibodies against IFNa2. None of the patients had detectable antibodies before treatment was started, but these were found in 25 of the 105 patients (24%) during treatment. Thirty-four patients were treated with rIFN-a2b, and binding antibodies were found in 3 (9%); these were also found in 22 out of 71 patients (31%) treated with IFNa2a, a statistically significant difference (p = 0.024). These antibodies were first detected in the serum between 60 and 750 days from the start of the rIFN-a2 treatment (Fig. 1). Antibodies neutralizing rIFN-a2 were detected in only 1 of the 34 patients treated with rIFN-a2b (3%), but in 17ofthe71 treated with rlFNa2a (24%), again a statistically significant difference (p < 0.02). The association between the development of binding antibodies against a rIFN-a2 preparation and secondary resistance to the treatment was investigated. Within 12 months from the start of therapy, 18 patients (3 treated with rIFN-a2b, 15 with rlFN-a^) developed a secondary resistance after initially responding. Nine of these patients (50%) had high-titered rIFN-a2 antibodies (titer above 900 Interferon Binding Units (IBU)/ml), and relapsed despite continued IFN therapy: all had received rlFN-a^. Thus, there was a stringent association between such high-titered antibodies against rIFN-a2a and secondary resistance. The observed differences in the immunogenic potential of the two recombinant IFN-a2 preparations were surprisingly clearcut. rIFN-a2a induced the formation of anti-IFN-a2 antibodies significantly more often than rIFN-a2b, and the titers of the induced antibodies were higher. Since all the patients were treated with the two recombinant IFN-a2 under identical conditions, the