Radiobiology of radioimmunotherapy: Targeting CD20 B-cell antigen in non-Hodgkin's lymphoma

Abstract

The radiobiology of radioimmunotherapy is an important determinant of both the toxicity and the efficacy associated with the treatment of B-cell non-Hodgkin's lymphoma with radiolabeled anti-CD20 monoclonal antibodies. The properties of the target, CD20, and the mechanisms of action of both the monoclonal antibodies and the associated exponentially decreasing low-dose-rate radiotherapy are described. The radiation dose and dose-rate effects are discussed and related to both the tumor responses and normal organ toxicity. Finally, the use of either unlabeled or radiolabeled anti-CD20 monoclonal antibodies as a component of combined modality therapy (including the sequential or concurrent use of sensitizers) and future directions of the field are discussed.

Introduction

Lymphomas are very radiosensitive tumors. Radiotherapy (RT), alone or combined with chemotherapy, has been used to treat these diseases. Despite the high initial response rate of most lymphomas to conventional therapy, this therapy fails in many patients, who develop recurrent/chemotherapy-refractory disease. Radioimmunotherapy (RIT) is a very promising new therapy for the treatment of recurrent and refractory B-cell non-Hodgkin's lymphoma (NHL). RIT combines the principles of systemic treatment with the local delivery of radiation. This is accomplished using radioimmunoconjugates, which are composed of antibodies to tumor-associated antigens that are chemically linked to radionuclides. Monoclonal antibodies (MABs) are produced by a single clone of antibody-producing B cells and are highly specific for a particular antigenic target. The selection of a radionuclide for use in RIT is based on its emission profile, particle energy, mean path length in tissue, and physical half-life. Ibritumomab tiuxetan (Zevalin, IDEC Pharmaceuticals, San Diego, CA) is a radioimmunoconjugate composed of a murine monoclonal anti-CD20 IgG1 κ-antibody labeled with the high-energy, pure β-emitter ⁹⁰Y. Bexxar (Corixa and GlaxoSmithKline) is another radioimmunoconjugate used in the treatment of NHL. It is composed of a murine anti-CD20 IgG 2a antibody named tositumomab, which is labeled with ¹³¹I, a γ- and β-emitter. The anti-CD20 antibodies used in Zevalin and Bexxar (ibritumomab and tositumomab, respectively) are both murine antibodies, because of the desirable circulating half-life of these antibodies (compared with chimeric/humanized antibodies, which have significantly longer half-lives). This results in maximal tumor uptake 48–72 h after administration with relatively little radiation to normal organs.

A number of clinical trials using murine anti-CD20 MABs conjugated to ¹³¹I (¹³¹I-tositumomab or Bexxar) or ⁹⁰Y (Y⁹⁰-ibritumomab tiuxetan or Zevalin) have demonstrated their efficacy in patients with relapsed or refractory indolent B-cell NHL 1, 2, 3, 4, even after the use of rituximab 5, 6. The Bexxar (tositumomab and ¹³¹I-tositumomab) regimen consists of the sequential administration of a dose of the “cold” or unlabeled MAB, tositumomab, to improve tumor localization 7, 8, with the subsequent administration of ¹³¹tositumomab (Bexxar). The regimen includes a dosimetric study to enable the calculation of a patient-specific therapeutic dose to deliver an absorbed whole body dose of 75 cGy in patients with a platelet count of ≥150,000. The use of this regimen has resulted in significant durable clinical responses in some patients, supporting its recent approval for marketing. Approval was given on the basis of the findings from a study of 40 patients with relapsed/refractory disease after rituximab therapy and supported by the demonstration of durable responses in four other studies enrolling 190 patients with relapsed/refractory disease after chemotherapy. Tumor responses were documented in approximately 60% of patients, with complete responses (CRs) in approximately 30%. The median duration of response has been >12 months, with occasional durable CRs 1, 2, 3, 9.

The Y⁹⁰-ibritumomab tiuxetan (Zevalin) regimen uses the chimeric anti-CD20 antibody Rituxan for predosing. In the pivotal study supporting Food and Drug Administration approval of Zevalin, patients were randomized to receive either Rituxan alone (four weekly doses of 375 mg/m²) or RIT with 250 mg/m² Rituxan on Day 0 followed by 5 mCi ¹¹¹In-labeled ibritumomab tiuxetan as a tracer for imaging/dosimetry. On Day 7, they received 250 mg/m² Rituxan followed by 0.4 mCi/kg ⁹⁰Y-labeled ibritumomab tiuxetan. A total of 143 patients were randomized in this study. On the basis of the International Workshop NHL Response Criteria, the overall response rate was 80% for Zevalin vs. 56% in the Rituxan arm (p = 0.002). CRs were achieved in 30% with Zevalin vs. 16% with Rituxan (p = 0.04) (4). Of those patients who have since had progression, the time to the next therapy cycle was 11.5 months after Zevalin and 7.8 months after Rituxan.

A variety of radionuclides can be used in RIT, including α- and β-emitters, and those that work through electron capture or internal conversion 10, 11, 12. Each radionuclide has a unique emission profile that is associated with advantages and disadvantages, depending on the particular clinical situation (e.g., bulky disease vs. microscopic residual disease). ⁹⁰Y has an average energy of 0.93 MeV, a half-life of 2.7 days, and a mean path length in tissue of 2.5 mm. ¹³¹I has an average β-particle energy of 0.19 MeV, a half-life of approximately 8 days, and a mean path length of 0.3 mm in tissue (12). These properties are ideal for the treatment of macroscopic disease because of the significant crossfire/bystander effect achieved primarily by the β-particles, and because of the relatively good match between the physical half-life of the radionuclides and the biologic half-life of the murine anti-CD20 antibodies.

Immunotherapy became an approved treatment modality for B-cell NHL in 1997, with the Food and Drug Administration approval of rituximab (Rituxan, Genentech, South San Francisco, CA and IDEC, San Diego, CA). Unlabeled MABs can kill tumor cells through multiple mechanisms, including apoptotic signal transduction pathways, antibody-dependent cell mediated cytotoxicity (ADCC), and complement fixation (depending on the particular antibody and model studied). RIT is advantageous compared with the use of unlabeled antibody alone, because of the additive effect of radiation-induced cytotoxicity and the ability of the associated radioactivity to kill tumor cells some distance from the bound radiolabeled antibody. This means that with RIT it is not necessary to have antibody bind to every cell in the tumor to kill all the tumor cells. Two randomized clinical trials comparing unlabeled antibody with radiolabeled antibody have demonstrated that RIT is significantly more efficacious than unlabeled anti-CD20 MAB therapy alone 4, 13, 14. However, as expected, RIT was associated with more toxicity than treatment with unlabeled antibody alone. Estimated tumor doses ranged from 580 to 6710 cGy (mean 1700 cGy for Zevalin 15, 16 and 895 cGy for Bexxar [17]).

Bone marrow suppression has been the dose-limiting toxicity from RIT in NHL. For example, in the pivotal study of Bexxar RIT in a heavily pretreated patient population (median of four prior chemotherapy regimens), 18%, 22%, and 0% of patients had Grade 4 neutropenia, thrombocytopenia, and anemia, respectively (2). An integrated safety analysis of data from five clinical trials with Zevalin at 0.4 mCi/kg revealed a 30% incidence of Grade 4 neutropenia, with 10% and 3% Grade 4 thrombocytopenia and anemia, respectively (18). The hematologic toxicity was generally reversible, and nonhematologic toxicity was usually Grade 1 or 2 and not dose limiting.

It is conceivable that low-dose, low-dose-rate (LDR) RT may have an enhanced toxic effect on the bone marrow 19, 20, 21, 22. However, given the estimated dose to bone marrow from these RIT treatments (e.g., median of 0.69 Gy to the red marrow from 0.4 mCi/kg ⁹⁰Y in Zevalin-treated patients) (23), and the fact that these patients had compromised bone marrow reserves as a result of previous cytotoxic therapies, with bone marrow involvement in a significant subset of patients, the extent of myelosuppression induced by this treatment is not surprising. Therefore, in terms of acute bone marrow toxicity, the data to date do not suggest that there has been clinically significant enhancement of the effect of the low-dose, LDR RT on the bone marrow. However, given the heterogeneity of the patient populations studied and the presence of numerous confounding variables, such an effect may exist, but not be detectable within the limitations of the study designs used. It has also been hypothesized that low-dose, LDR RT may result in increased leukemogenic transformation 19, 20. The incidence of myelodysplastic syndrome and secondary leukemia is presently within the expected background rate for heavily pretreated (e.g., with alkylating chemotherapy with/without prior RT) patients 24, 25, but data from ongoing long-term follow-up studies with both Bexxar and Zevalin will help to clarify this in the future.

Radioimmmunotherapy has potential advantages over external beam RT (EBRT) in that it can simultaneously treat multiple tumor sites throughout the body. It is a targeted therapy that spares most normal tissues with the delivery of relatively low (well below organ tolerance) doses to normal organs and adjacent structures 15, 16, 26, 27. It is often difficult or impossible to encompass diffuse disease (e.g., Stage III or IV NHL) within EBRT fields that would be associated with acceptable toxicity. Therefore, because of dose-limiting normal tissue toxicity, it is often not possible to treat all known or symptomatic disease sites in NHL patients with conventional EBRT. Also, although EBRT uses high-dose-rate (HDR) RT, the radiation from RIT is very LDR, approximately 2.5 orders of magnitude lower in intensity than HDR RT. The LDR RT associated with RIT is continuous and exponentially decreasing with time. Data have suggested that some lymphomas may be more sensitive to LDR than HDR RT (28), and normal tissues are better able to tolerate LDR than HDR RT (29), presumably because of increased ability to repair radiation damage during LDR RT. Dose and dose-rate effects are discussed further in “Biologic effects of RT.” The use of either unlabeled or radiolabeled anti-CD20 MABs in combination with other therapies is discussed in “Combined Modality Therapy.” The final section discusses future directions of RIT for NHL.