Radioimmunotherapy of Non-Hodgkin’s Lymphoma Revisited

TO THE EDITOR:

We appreciate the views communicated by Britton (1) in his comments on the papers by Koral et al. (2) and Wiseman et al. (3) pertaining to therapy of non-Hodgkin’s lymphoma (NHL) with ¹³¹I-tositumomab and ⁹⁰Y-ibritumomab tiuxetan, respectively, and desire to offer some other thoughts. Because the carriers of radioactivity for therapy of differentiated thyroid cancer and neuroendocrine tumors are not active against these neoplasms absent the radioactivity delivered, it is not appropriate to use these as examples of a similar paradigm in NHL radioimmunotherapy, because the antibody carriers have in fact been shown to be active on their own, especially rituximab. Predosing with rituximab and adding high amounts when giving the murine radiolabeled antibody certainly differentiate this from any other non-antibody-based therapy. Indeed, experimental evidence has been presented that some biologic effects exerted by the naked antibody can enhance the effects of radiation (4).

We would be gratified if there were indeed a direct relationship between targeting and estimated tumor dose delivered with therapeutic response in radioimmunotherapy. However, evidence is accumulating that tumors can have significant responses despite receiving lower estimated doses than others that receive a higher dose. Even tumors that are not visualized (i.e., tumor dosimetry cannot be determined) can undergo major responses (5,6).

Although Koral et al. (2) defined a trend in previously untreated patients whereby those receiving higher radiation doses to the tumor were more likely to have a complete response, even they acknowledged other factors that can contribute to a response and did not advocate, as Britton does, the use of tumor targeting and dosimetry as a means of selecting which NHL patients should receive radiolabeled antibody treatment. Although taking this course might enhance the likelihood of a response, there is as yet no body of data that proves the contrary; namely, that patients do not respond if their tumors are not targeted with the radioactive antibody. Indeed, Koral et al. did emphasize the limitations of dosimetry methods for tumors and the effects of unlabeled antibody and also found that some tumors responded at lower radiation doses and others failed to respond at higher doses.

Most clinical trials have a regulatory requirement that targeting be confirmed before therapeutic doses are given, but definitive tumor targeting is not required for the approved agents. Trying to select patients on the basis of dosimetry is not unlike the selection of patients for cytotoxic chemotherapy, for which in vitro assays of tumor sensitivity have not been predictive. But when a specific target molecule is recognized before therapy, such as the expression of cluster designation (CD) 20 on lymphoma cells, then use of a radiolabeled anti-CD20 monoclonal antibody is more justified than when the cells have no demonstrable CD20. However, how much CD20 is enough? Is its presence in 10% of cells, 20% of cells, or more cells sufficient, and is weak staining adequate? In vivo imaging of a patient with extensive disease can be problematic and misleading if proper doses and procedures are not followed, including dose titration. Therefore, having both in vitro and in vivo data of antigen expression and antibody accretion is certainly the best situation, but when the naked antibody can have immunotherapeutic activity and when there is still an incomplete understanding of factors that influence a response, targeting the isotope to the tumor may not, by itself, be sufficient for selecting a candidate patient. For these reasons, we do not agree with the implication of Britton’s view that even patients who have failed prior therapies should not be given radioimmunotherapy if pretreatment targeting and dosimetry are not convincing of a probable response, especially when therapeutic doses of naked monoclonal antibodies are being given as a part of the therapy. Britton states, “there is clearly a lower limit at which insufficient therapy has no benefit to the patient with a tumor,” but there are no data to support this conclusion. In many early phase I trials, responses are observed at the starting therapeutic dose level, and responses have been seen even after the pretherapy imaging dose, but of course the therapeutic dose is always escalated because of the assumption that higher doses will be more beneficial. Certainly the higher complete response rate at myeloablative doses supports the “more is better” dictum, but clearly we must be more cautious in advocating an expanded use of doses requiring hematopoietic support and instead direct trials toward showing efficacy for radioimmunotherapy with less toxicity in a frontline setting. Indeed, use of ¹³¹I-tositumomab as a first therapy in indolent, follicular NHL showed lower toxicity but also a higher response rate than when ¹³¹I-tositumomab was given at the same dose after multiple prior drug courses (7), but it is not as yet clear whether there is a difference in tumor targeting and accretion between these 2 patient groups.

Also deserving of comment is Britton’s example that the dose administered to a patient on the basis of body weight seemingly could have been increased were it adjusted to the dose limit allowed by the estimated bone marrow dose. Although the calculation of red marrow absorbed dose aids in the prediction of hematologic toxicity, there are other factors to consider, particularly in NHL (8–11), and therefore it would be irresponsible to suggest that another method be used in place of those that have carefully been evaluated for the 2 approved radioimmunotherapy agents. The simple observation that the majority of patients who receive radioimmunotherapy experience myelotoxicity, sometimes even grade 4, suggests that the current dosing methods (i.e., whole-body clearance and body weight plus blood count) are still imperfect. However, this observation also means that most patients are receiving the highest possible dose without having to rely on more drastic measures to control severe myelosuppression. Nevertheless, between at least 8% and 15% do require hematopoietic growth factors or blood cell transfusions (Bexxar [Corixa Corp.] and Zevalin [Biogen IDEC] product labels). Under the currently approved indication, patients will most likely have undergone several cycles of chemotherapy that can have a profound effect on the marrow reserve. From a practical perspective, reducing the dose of radiolabeled antibody according to bone marrow involvement (with >25% involvement excluding treatment) and baseline platelet counts has been shown to mitigate myelosuppression. Use of red marrow dosimetry in combination with a biologic marker, such as FLT3-L, improved the prediction of myelotoxicity in patients with solid tumors (12), but use of this or other markers of hematopoietic status needs further examination to determine utility in lymphoma.

Radioimmunotherapy remains complex because of the combination of immunology and radiation medicine, each having its own set of problems and prospects. This complexity was confirmed by the long journey that this modality traveled before the first products reached (only recently) clinical practice. Yet, despite many differences between the first 2 radioimmunotherapy agents approved for use, they are interestingly similar in their ability to achieve higher response rates than either prior chemotherapy or use of the naked antibody by itself (13), and they are indeed also gaining interest as potential frontline therapies for NHL (7). In this setting, we prefer to advocate that all appropriate patients whose tumors express CD20 should be candidates for CD20 radioimmunotherapy, that those who express CD22 should likewise be treated with radiolabeled CD22 mAbs, and so forth, at least until more reliable in vivo methods of selecting the best responders are confirmed. But standards for determining adequate expression of such markers are not established, and whether these determinations should be made before or after patients fail chemotherapy remains a subject of future investigation and could include study arms with and without patient selection based on prior external imaging.