Abstract: Magic bullets armed with nuclear war head? Monoclonal antibodies labeled with radioactive isotope, when taken orally, seek and attach to tumor cells and bring radiation close to a range of destruction.

Hybridoma/monoclonal antibody (mAb) technology as described in Kohler and Milstein’s work resurrected Ehrlich’s century old concept of “magic bullets.” This seminal publication described fusion of a plasmacytoma (tumor of activated B lymphocytes) with spleen cells and subsequent isolation of hybrids that secreted mAb with pre-defined specificity. Generation of mouse mAbs against tumor-associated antigens became a focus in the 1980’s. Pre-clinical studies provided proof-of-concept for the potential of mAbs for therapy although inherent limitations of these models demonstrated discordance in predictability of actual efficacy. Clinical investigations also illustrated deficiencies; foremost, an inevitable immune response, i.e., the production of human anti-murine immunoglobulin antibodies (HAMA). Other limitations included (1) inadequate tumor dose delivery; (2) insufficient activation of effector function(s); (3) slow blood clearance; (4) low affinity and avidity; (5) normal organ targeting; (6) tumor antigen heterogeneity; and (7) insufficient tumor penetration (Milenic et al., 2004; Knox et al., 2000). Most of these were handled with genetic engineering or chemical modification, but some obstacles remained. HAMA production has been addressed by chimerization (combining portions of mouse and human antibodies) or complete humanization (grafting only the key antigen-binding portions of mouse antibody to a human antibody framework) of the mAb.

With the elimination of many obstacles combined with a better understanding of inherent mAb limitations and support from industry, several radiolabeled mAbs (Figure 1) have been, or are currently being evaluated in Phase III clinical trials (Table 1).

In the past 2 years, the FDA has approved two radiolabeled mAbs (Zevalin and Bexxar) for treating non-Hodgkin’s lymphoma (NHL), setting the stage for additional targeted radiation therapies.

The Radioisotopes

Choosing the most appropriate radionuclide is crucial. No single radionuclide is likely to address every therapeutic requirement, as disease does not exclusively present in a single form. Isotope selection is often driven by economic rather than medical or scientific considerations, which may negatively impact both pre-clinical and clinical trials. Ultimately, defining the limitations of the targeting vectors and radionuclides will allow for more rational design of clinical trials (Milenic et al., 2004; O’Donoghue et al., 1995; Hassfjell et al., 2001).

Critical variables for successful targeted radiation include: emission type, energy/range of emission, and half-life. A sampling of isotopes (Table 2) points to several alternatives comprising (Figure 2) three emission types: β^--particles, α-particles, and

Auger electrons. Advantages of β^--emitters include the potential to bypass tumor antigen heterogeneity and differential mAb tumor penetration. Uniform targeting of the entire lesion becomes possible when the emission range exceeds the radius of the targeted lesion (O’Donoghue et al., 1995). Convenience, availability, and familiarity have supported radio-iodines, e.g., 131I. Other clinically relevant β^--emitters include 90Y, 67Cu, 186Re, and 177Lu. Emission energy and half-life requirements will likely be met with a small group of isotopes that are already available. The emission path lengths of β^--emitters are relatively long, yet sparse (mean range of 275 μm, 500-600 μm for 90Y) with low linear energy transfer (LET) (Figure 3). Energy deposition takes place away from the actual decay event. Therapeutic benefit results from crossfire, i.e., the cell targeted with the radionuclide on the surface, or internalized within the cell, is not necessarily the effective target, underscoring the constraints of β^--emitters. Single cell metastatic diseases, leukemias, and disseminated diseases therefore cannot be adequately treated with β^--emitters. Despite this, β^--emitters continue to dominate pre-clinical and clinical trials; Zevalin and Bexxar, both FDA approved, are armed with β^--emitters.

The list of qualified α-emitters is relatively short, largely owing to half-life constraints. Presently, only 212Bi, 213Bi, 211At, and 225Ac (Table 2) are actively studied (Hassfjell et al., 2001).

α-Emitters have high-energy particles (4-9 MeV), characterized by dense, short (40-100 μm) emission path lengths of high LET, ~400 times greater than β^-–emitters. Energy deposition initiates directly from the decay site (Figure 2). Most cytotoxicity is directed at the cell targeted and the immediate neighboring cells. In contrast to β^-–emitters, 1-3 nuclear traversals are required to kill a cell with an α-emitter possessing a dose rate of 1 cGy/hr. Their short physical half-life coupled with short emission path length, has generally limited their use to leukemias, highly vascularized tumors, and metastatic/disseminated disease where adequate access, targeting time, and appropriate disease size converge.

Auger electron emitters such as 67Ga, 195mPt, 123I, and 125I (Table 2) have received the least attention. The accepted premise is that their extreme cytotoxicity and efficacy is limited by a prerequisite for emissions to occur within the cell nucleus. Despite this paradigm, Auger emitters have demonstrated potential as therapeutics for microscopic residual disease (Milenic et al., 2004).

Linking Radionuclides to Proteins

Researchers have sought to balance the conditions required to achieve an adequately stable radiolabeled product within the constraints imposed by isotope chemistry and half-life (Milenic et al., 2004). As this field has matured, the choices of realistic isotopes and chelating agents have narrowed. Direct radio-iodination (131I, 125I, 123I) is well established and will not be addressed here. All metallic radionuclides require chelation chemistry for radiolabeling mAbs (Figure 4). These chelates must possess specific functional groups that permit both conjugation to proteins and stable complex formation with metallic radionuclides. Likewise, 211At also requires a linking agent, as direct radio-halogenation is inappropriate.

The suitable radio-metals have diverse properties and coordination chemistry, precluding a universal bifunctional chelating agent (BCA). Numerous criteria must be considered in the choice of BCA, e.g., its design and its actual use. A BCA may form and maintain an adequately stable metal complex, but kinetics may render it impractical. DOTA (1,4,7,10-tetra-azacylcododecane-N,N’,N”,N”’-tetraacetic acid) forms highly stable and kinetically inert complexes with radiobismuth isotopes 212Bi or 213Bi, however, complex formation kinetics require 15-45 mins. Their short half-lives make this choice of chelate impractical (Hassfjell et al. 2001).

The Protein

The contribution of the targeting vector to therapy should not be discounted (Milenic et al., 2004). Direct tumor cell killing may be achieved via antibody-dependent cell cytotoxicity. Cell killing may proceed through a cell-dependent (phagocytosis) or a cell-independent mechanism (lysis). Receptors are integral to cellular function; triggering or blocking them can induce either cell growth or death. Reduced therapeutic efficacy against breast cancer and lymphoma by Trastuzumab (Herceptin) and Rituximab (Rituxan), respectively, has been noted when Fc (antibody’s non-variable region, see Figure 4) receptor activation is absent. Rituximab appears to invoke signaling events including increased phosphorylation and induction of suicide of B lymphocytes. Trastuzumab binds HER2 (a member of the receptor tyrosine kinase family) and has efficacy in metastatic breast cancer patients. This mAb has been postulated to exercise a direct anti-proliferative signaling effect through blocking of receptor-ligand interactions and inducing receptor depletion. Daclizumab (Zenapax), which recognizes interleukin-2 (a key cytokine) receptor (CD25), achieves therapy through cell signaling interference and function by blocking interleukin-2 binding thereby inhibiting tumor proliferation.

Modification of cell signaling may result in synergistic effects when combined with chemo- or conventional radiotherapy. Studies in animal models have confirmed enhanced efficacies of anti-HER2 mAbs in concert with external beam radiation in model systems where radiation or mAb alone had minimal effects. Trastuzumab, combined with paclitaxel or doxorubicin, enhanced both rates and duration of response in patients. Clinical trials are investigating the effects of combined drug / targeted radiation therapy to integrate radioimmunotherapy into mainstream clinical cancer therapy.

Radioimmunoconjugates Tested in Clinical Trials

The majority of cytotoxic agents have a low therapeutic index (benefit vs. risk). Selective localization of a cytotoxic agent via mAb may enhance the therapeutic index. Radioimmunoconjugates have been evaluated in clinical trials across the full spectrum of malignancies (Knox et al., 2000). Dissimilarities in experimental design and execution have complicated comparisons of the results. Variations in dosing (protein and radioactivity), administration methodology, radionuclide, targeting vehicle, and selection of both endpoint and the manner of its determination are further complications. The majority of radioimmunotherapy clinical trials have employed intact murine IgG mAbs. Chimeric and humanized mAbs have recently undergone clinical evaluation. Substantial optimization continues with goals of rapid targeting, maximized retention through increased target antigen expression or re-expression, and minimized exposure of normal tissue. Limited successes have been achieved in treating solid tumors whereas the greatest accomplishments have been with lymphohematopoietic malignancies despite myelosuppression (reduced bone marrow activity and therefore reduced platelets, white and red blood cell counts) toxicity issues. Tumor accessibility, ability to characterize tumor phenotypes, coupled with the ability to determine differentiation, and intrinsic radiosensitivities suggest that these cancers are ideal for radioimmunotherapy. Antigenic heterogeneity and lack of tumor penetration are not major impediments to radioimmunotherapy of lymphohematopoietic disease, whereas these variables coupled with myelosuppression and other toxicities, e.g., hepatic, renal, gastro-intestinal, have limited therapy of solid tumors.

Zevalin (ibritumomab tiuxetan) is the first therapeutic radiolabeled mAb approved by the FDA (Feb. 2002) (Hernandez et al., 2003). Zevalin is a 90Y-labeled murine anti-CD20 mAb using the BCA Tiuxetan. This agent is approved for treatment of relapsed or refractory low grade, follicular, or transformed B-cell NHL. Patients with Rituxan-refractory follicular NHL are eligible as are those receiving Rituxan as part of a therapeutic regimen. Phase I/II studies demonstrated efficacy and safety treating recurrent B-cell lymphoma, and established “pre-treating” patients with Rituxan as part of the protocol. Complete and partial responses have resulted with progression-free survival of 72% 3-29+ months and a progression-free survival of 78% with minimal hematological toxicity. Pre-treatment of patients with Rituxan improved tumor uptake, decreased accumulation in the spleen and decreased urinary excretion of Zevalin. Additional trials have established patient doses and comparisons to Rituxan have indicated an improved overall response rate over that of Rituxan monotherapy. Furthermore, 74% of the patients who failed Rituxan therapy were responsive to Zevalin.

A second anti-CD20 mAb Bexxar (tositumomab and 131I-tositumomab) labeled with 131I, has also received FDA approval for treatment of CD20 positive, follicular, NHL, with and without transformation, disease refractory to Rituxan, and relapsed after chemotherapy (Knox et al., 2000). A multi-center trial confirmed efficacy and safety in patients with relapsed or transformed low-grade NHL. Improved survival of chemo-refractory NHL patients following treatment with Bexxar has also been reported. The median response duration and complete responses for those receiving targeted radiation therapy were greatly improved versus those receiving only chemotherapy only. A comparison of tositumomab alone and in concert with 131I-tositumomab (namely, Bexxar) was conducted to determine any additional benefit resulting from 131I. To quote the investigators, “All therapeutic outcome measures were significantly enhanced by the conjugation of 131I to tositumomab” (Davis et al., 2003). No increased incidence of treatment-related myelodysplastic disease (reduced white cells and platelets; MDS) or acute myeloid leukemia resulted when Bexxar was the initial therapy.

The complete response rate, remission duration and the partial response from single or fractionated doses treating B-cell malignancies have been more impressive when high myeloablative (destruction of bone marrow) doses are combined with autologous bone marrow transplantation or stem cell transplantation (Pagel et al., 2002). Complete remissions lasting up to 53 months have been noted in patients that underwent bone marrow transplantation after receiving myelosuppressive doses of 131I-labeled anti-CD20 mAbs. Autologous stem cell transplantation following bone marrow eliminative Bexxar therapy also resulted in increased partial and complete responses. These patients experienced an improved progression-free survival and overall survival with a median follow-up of 2 years. High dose targeted radioimmunotherapy (HD-RIT) has also been compared to conventional high dose radiotherapy; higher progression-free and overall survival were reported for the HD-RIT group. The majority of the HD-RIT patients eventually relapsed, however, leading to clinical trials that included chemotherapy in the radioimmunotherapy regimen. A Phase II trial combined CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) chemotherapy followed by tositumomab/131I-tositumomab for untreated follicular NHL. A low dose of 131I-tositumomab was included in the regimen for imaging and dosimetric calculations. This scheme resulted in 67% complete and 23% partial responses with a 2 year progression-free survival rate of 81% and overall survival rate of 97% with a median follow-up of 2.3 years. CHOP therapy is also being evaluated in conjunction with Rituximab. A strategy of combining chemotherapy, total body irradiation, and radioimmunotherapy has also shown potential for treatment of advanced acute lymphocytic leukemia and MDS. Each of these patient groups has experienced an improved median survival or disease-free survival using the aforementioned treatment. Issues pertaining to either Bexxar or Zevalin eliciting HAMA have been minimal. HAMA occurs <2% of the time with Zevalin while there has been ~5-10% occurrence with Bexxar; re-treatment has not been limited by this issue.

In summary, 13 monoclonal antibodies are currently approved by the FDA; 2 are radiolabeled, Bexxar, a mAb with 131I, and Zevalin, with 90Y, both for the treatment of CD20+ NHL (Milenic et al. 2004). A 67% overall response rate has been achieved in patients receiving Zevalin. Experience with a single dose of Bexxar in 5 clinical trials has resulted in responses lasting from 3 months to 5.5 years. Complete responses occurred in 28% of the patients with the median response lasting 4.8 years. These results demonstrate the enhanced efficacy from a targeted radionuclide. The results also illustrate the rational approach of targeting a cell surface antigen that is not shed or modulated which can also directly signal cell death through apoptosis. This inherent activity of the target provides a true molecularly-targeted combination therapy, a key factor in the success of both of these agents. Considerable efforts continue towards fully understanding the mechanisms of action as well as in combination with various drugs (e.g., Taxol, doxorubicin) and radiosensitizers (Gemzar). Some evidence exists, particularly with the anti-CD20 mAbs, Herceptin, and Erbitux, that additive, synergistic or complementary modes of action may render cells more sensitive to radiation thereby enhancing their therapeutic index.

Conclusion

Despite the wealth of knowledge and genetic engineering technology, the first 2 radiolabeled mAbs approved for therapy are murine in nature and subject to all of the limitations therein. Clinical application of radiolabeled mAbs remains in its infancy in many aspects, particularly so in regards to therapies beyond lymphohematological diseases. The functional integration of targeted radiation therapy into established chemotherapies and external beam therapies still requires rational construction of therapeutic regimens that include fractionated dosing and drug combination cocktail therapies. Clear evidence exists that enhanced results are achieved by execution of these strategies.

Dominance of mAb therapies for the lymphohematopoietic malignancies and their success therein reflects disease accessibility and radiosensitivity. Consensus appears to support that mAb-based therapies of solid tumors may remain restricted to the treatment of minimal, residual, or micrometastatic disease, and as a component of a multi-modality treatment regimen. However, these limitations may be a reflection of inappropriate targeting agents, sub-optimal chemistry, incorrect radionuclide choice, and less than rational experimental design. There remains a continuing effort to refine and optimize all of the components to improve efficacy and minimize toxicity. The next decade should prove exciting as rational exploration and investigation applying the cumulative knowledge to date makes targeted radiation therapy a reality and mainstream component for the treatment and management of cancer.

References and Further Readings

Milenic DE, Brady ED, Brechbiel MW. Antibody-Targeted Radiation Cancer Therapy. Nature Reviews Drug Discovery 3:488-498, 2004.

Knox, S. J., Meredith, R. F. Clinical radioimmunotherapy. Seminars in Radiation Oncology 10:73-93, 2000.

O’Donoghue JA, Bardies M, Wheldon TE. Relationships between tumor size and curability for uniformly targeted therapy with beta-emitting radionuclides. Journal of Nuclear Medicine 36:1902-1909 (1995).

Hassfjell S, Brechbiel MW. The Development of the α-particle emitting radionuclides 212Bi and 213Bi for therapeutic applications. Chemical Reviews 101:2019-2036, 2001.

Hernandez MC, Knox SJ. Radiobiology of radioimmunotherapy with 90Y ibritumomab tiuxetan (Zevalin). Seminars in Oncology 30:6-10, 2003.

Davis T, et al. Long-term observation of a randomized trial comparing tositumomab and iodine-131 tositumomab (BEXXAR) with tositumomab alone in patinets with relapsed or refractory low grade (LG) or transformed low grade (T-LG) non-Hodgkin’s lymphoma (NHL). Blood 102:405a-406a, 2003.

Pagel JM, Matthews DC, Applebaum FR, Berstein ID, Press OW. The use of radioimmunoconjugates in stem cell transplantation. Bone Marrow Transplantation 29:807-816, 2002.

[Discovery Medicine, 4(22):213-219, 2004]