In general, red marrow is the dose-limiting tissue in non-myeloablative and lung for myeloablative radioimmunotherapy (RIT). Administration protocols are applied based on absorbed-dose values of the dose-limiting tissue and on an activity per body weight basis. Typical examples
Figure 3. Correlation between the whole body dose estimate based on 123I-MIBG pretherapy scans and the dose derived from 131I-MIBG posttherapy scans in patients treated for neuroendocrine tumors. The triangles represent the data of the first therapies, the crosses the data of retreatments. The straight line is the result of a linear regression to all data (R2 = 0.73).
Of these protocols are the 131I-labeled anti-CD20 antibody, tositumomab (Bexxar; Glaxo-SmithKline) (76) and the 90Y-labeled anti-CD20 ibritumomab tiuxetan (Zevalin; Biogen Idec) (77), respectively. These radiolabeled antibodies are used for treatment of non-Hodgkin’s lymphoma. The choice for an activity-based protocol for the 90Y-labeled antibody is based on the lack of correlation between absorbed dose and toxicity in the early studies. The explanation for the absence of a dose-response relationship can be found in different sources. In contrast to i31i, 90y is a pure b-emitter, and 90Y kinetics have to be derived from surrogate 111In imaging. Another point is that prior treatment of these patients and the bone marrow reserve have a strong effect on the bone marrow toxicity in this case. As patients undergoing RIT have been treated previously by chemotherapy, the impact of such prior therapy on the hematopoietic response to the RIT is important.
Although the necessity of patient-specific dosimetry is questionable in some applications of RIT where the dose-response observations for toxicity are poor, there is a general agreement that complete radiation dosimetry is necessary for each new application of a radiolabeled antibody in phase I and most probably also in phase II studies especially for safety reasons (78). An important argument for absorbed dose driven protocols in clinical phase I trials is that many patients are treated below the biologically active level due to the interpatient variability in activity based administration protocols. This implies data difficult to interpret in antitumor response and toxicity.
In view of the central role of red marrow toxicity in RIT methodologies, bone marrow dosimetry got already a lot of attention in the literature (79-85). In general, methods based on imaging as described in the section on 131I-MIBG therapy are used. Also, approaches to calculate the bone marrow dose based on blood activity measurements have been described, but these methods yield only reliable results when the activity does not bind specifically to blood or marrow components including tumor metastases in the marrow (79). By assuming rapid equilibrium of radiolabeled antibodies in the plasma and the extracellular fluid of the red marrow, a red marrow/blood concentration ratio of 0.3-0.4 can be derived. All red marrow dosimetry performed up to now uses a highly stylized representation of the red marrow over the body. More detailed representations are being generated especially for Monte Carlo calculations enhancing accuracy and reliability of the bone marrow doses (86).
Several studies have investigated the relation between the tumor dose and response especially in RIT of non-Hodgkin’s lymphoma (87-89) but the results are negative. Possible explanations are the therapeutic effect of the antibody, different confounding biological factors and the accuracy of tumor dosimetry. Here, standardization of data acquisition as presented in MIRD pamphlet No. 16 (13) may help in dose-response investigations. As discussed earlier in the section on 131I-MIBG therapy, a full patient-specific 3D dosimetric approach with imaging data from the combined SPECT-CT systems will improve substantially the accuracy of the tumor dosimetry results.