Flu redux and RIT: friendly fire in the treatment of low-grade lymphoma

Comment on McLaughlin et al, page 4573, and Bennett et al, page 4576

Two reports describe the problem of secondary malignancies in patients with LGL. Bennett and colleagues find that I¹³¹ tositumomab (Bexxar) following alkylating agent therapy does not increase the risk of tMDS or tAML, while McLaughlin and colleagues find that this risk is increased following fludarabine-based therapy.

The problem that Boice¹  has called “friendly fire,” or the development of secondary malignancies after treatment, has been studied extensively over the past 25 years. As patients live longer, we will be faced with more questions about late effects. In low-grade lymphoma (LGL), the demonstration of the CD20 target and the subsequent development of monoclonal antibodies (rituximab) as therapeutic ligands has improved response rates and allowed more therapeutic options. Based on the observation that external beam radiation is an effective modality that is limited by the confines of the beam, investigators have developed novel radioimmunoconjugates that use low-energy I¹³¹ tositumomab²  or higher-energy Y⁹⁰ ibritumomab tiuxetan (Zevalin)³  attached to a murine anti-CD20 antibody. Data have suggested that response rates are high and durable and that there is no cross-resistance with either rituximab or prior chemotherapy. The fear, however, was that the radiation would lead to a higher incidence of second malignancies. This concern was not unreasonable, as the radionuclide P³² was associated with a higher-than-expected incidence of leukemia in patients with polycythemia vera,⁴  but large studies with I¹³¹ in thyroid disease have found no increase in treatment-related acute myeloid leukemia (tAML) or treatment-related myelodysplastic syndrome (tMDS).⁵  In this issue of Blood, Bennett and colleagues observed that the risk of tAML or tMDS was not higher than would be expected (about 1.1% per year in confirmed cases) from alkylating agents alone when radioimmunotherapy (RIT) was used after alkylating agent exposure in 995 heavily treated patients and was 0% in 76 previously untreated patients. This should allow further careful trials of RIT used earlier in the course of the disease. The observation that the incidence of tAML or tMDS was highest in patients with thrombocytopenia and in those who had been previously exposed to the purine analogue fludarabine deserves attention. In earlier studies with Zevalin, prior exposure to fludarabine was associated with prolonged thrombocytopenia.

In this issue of Blood, McLaughlin and colleagues report on 202 patients with LGL who were treated with either fludarabine, mitoxantrone, and dexamethasone (FND) with or without rituximab, followed by interferon alpha (IFNα), or with alternating triple therapy (ATT) with or without rituximab, followed by IFNα (ATT has no fludarabine). With a median follow-up of 42 months, 8 patients (4%) developed tMDS or tAML. Of these, 6 received FND with or without rituximab and IFNα, but 2 also received alkylating agents. Two received ATT with or without rituximab and IFNα. The cytogenetic profile of these cases included frequent chromosome 5 (5q-) and 7 (monosomy 7) abnormalities, a pattern that has been described before, especially after alkylating agent exposure. No patient had an 11q23 abnormality. These observations, combined with the finding by Bennett et al that prior fludarabine exposure was associated with a higher risk of tMDS or tAML in patients receiving RIT and with a report suggesting that in chronic lymphocytic leukemia combinations of alkylating agents and fludarabine were also associated with a greater risk of secondary tMDS or tAML,⁶  should give investigators and clinicians pause. We must rethink our combination regimens (alkylating agents and purine analogues) and consider testing RIT before fludarabine in future clinical trials in LGL.

Pedersen-Bjergaard et al⁷  have provided direction by describing a genetic pathway that characterizes therapy-related tMDS or tAML. In this model (see figure), 8 pathways based on specific cytogenetic abnormalities are described, and within these pathways, 3 etiologies emerge: (1) exposure to alkylating agents, (2) exposure to topoisomerase II inhibitors, and (3) spontaneous endogenous recombinations unrelated to chemotherapy. As we improve our therapeutic armamentarium by using purine analogues and RIT and offer either sequential or combined therapy to patients, we may find new patterns and increased risk. Since many new agents, including fludarabine and RIT, are now available off the shelf, we are obligated to follow the lead of the investigators in the 2 reports in this issue of Blood and test new agents and combinations in the context of carefully monitored clinical trials, in order to study the biologic and clinical implications of our actions. ▪