INVITED COMMENTARY ␣-Particle Emitters in Radioimmunotherapy: New and Welcome Challenges to Medical Internal Dosimetry sulting in the release of the daughter as the cellular S value methodology to Over the past decade, there has a free element. include time-dependent partial contri- been progressively stronger interest in Two very different approaches can butions of the various daughter emis- the use of ␣-particle emitters for radio- be applied to the dosimetry of ␣-parti- sions in the serial decay chains of immunotherapy (1–6). With proper lo- cle emitters. One is microdosimetry, in 225Ac, 221At, 213Bi, and 223Ra. In their calization of the labeled antibody, the which the probabilistic nature of approach, a cutoff time (␶0) is selected high linear energy transfer of ␣-parti- ␣-particle emission and its trajectory before which free elemental daughter cles provides a correspondingly high through the cell and cell nucleus are radionuclides are considered to remain probability of mitotic cell kill when explicitly considered (1,8–10). In a in the same source configuration as compared with an equivalent number microdosimetric analysis, probability that assumed for the parent. At short of cellular traversals by lower linear density functions of specific energy are cutoff times, the daughter radionu- energy transfer ␤-particles. Conse- obtained (stochastic expressions of en- clides, which are released as free ele- quently, much developmental work ergy imparted per unit mass to small ments after the ␣-decay of the parent, has been initiated in the production, targets), as well as frequencies of zero- diffuse or migrate far from the site of chemistry, and preclinical trials of can- dose contribution. Input data for such the parent decay and thus the cellular an analysis, however, require detailed target dose results only from the decay didates for radioimmunotherapy such knowledge of geometric features such of the parent. The authors note that for as 211At, 212Bi, 212Pb, 225Ac, 213Bi, and as the spatial distribution and size of parent decays in blood circulation, 223Ra (7). In general, ␣-emitters with the source and target regions (e.g., cel- short values of ␶ are applicable. For half-lives that are either relatively 0 lular and nuclear sizes and subcel- tumor interstitium, intermediate values short or relatively long compared with lular distribution of the radionuclide). of ␶ would be appropriate; thus, the transient times in blood as well as dif- 0 Meaningful correlations to biologic re- total cellular dose is contributed by the fusion and binding times in disease sponse further require data on the tim- parent and a ␶ -dependent fraction of tissues may be considered. Those 0 ing of the decays within the phases of the cumulative decays of the daughters -emitters with relatively short half- ␣ the cell cycle and the variations of of the serial decay chain. It is clear that 213 lives, such as Bi, will most likely be cellular radiosensitivity during these this approach to cellular dosimetry restricted in their application to small, phases. In many cases, such data are lends itself nicely to broader consider- readily accessible tumors. For treat- not available in the clinical setting. ations of the biokinetics and dosimetry ment of larger solid tumors, longer- A simpler approach is to extend the of radionuclides with multiple unstable 225 lived ␣-emitters such as Ac and MIRD schema to the cellular level and daughters as proposed under a matrix 223 Ra can also be considered. How- estimate mean absorbed dose to the formalism developed by these same ever, longer-lived radionuclides re- cells or cell nuclei through the appli- authors. quire more extensive normal organ do- cation of cellular S values. In its 1997 Several issues and challenges of simetry and biokinetics of their monograph, the MIRD committee pub- ␣-particle dosimetry are highlighted multiple unstable daughters in evaluat- lished extensive tables of cellular S through this approach. First, what ing clinical efficacy. With high proba- values for a wide range of ␣- and value of ␶0 is appropriate and under bility, the recoil energy of the ␣-emis- ␤-emitting radionuclides (11). These what conditions of the cellular micro- sion will result in destruction of their tabulations include S values for the cell environment? What is the spatial mo- chemical bonds with the antibody, re- and cell nucleus as target regions and bility of these daughter radionuclides for the cell, the cell surface, the nu- within tissues and cellular microenvi- cleus, and the cytoplasm as potential ronment that would permit quantitative Received Mar. 30, 2001; revision accepted Apr. 9, 2001. source regions. selections of ␶0? To correctly perform For correspondence or reprints contact: Wesley In this issue of The Journal of Nu- this analysis, detailed knowledge of E. Bolch, PhD, Department of Nuclear and Radio- logical Engineering, University of Florida, 202 clear Medicine, Hamacher et al. (12) the chemical diffusion coefficients for NSC, Gainesville, FL 32611-8300. have provided an elegant extension of each elemental species within various 1222 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 42 • No. 8 • August 2001 tissue compartments (e.g., nucleo- organs as required in the evaluation of therapy utilizes dosimetry as a predic- plasm/cytoplasm, extracellular fluid, the clinical efficacy of these ␣-emitting tive tool for more near-term determin- cellular membranes, bone marrow) radionuclides for radioimmunotherapy istic effects. With this in mind, the would be needed. In most cases, such (15). A compartmental analysis of ac- ICRP recommendations for target tis- details for high-Z elements are not tivity in normal organs might include sue definitions cannot always be used. available. It is for this reason that the separate determinations of the cumula- For example, the ICRP methodology International Commission on Radio- tive decays within the parenchymal tis- for skeletal dosimetry focuses on en- logical Protection (ICRP), in its publi- sues of the organ (incorporated activ- dosteum and marrow stem cells as the cation 30 (13), makes the simplifying ity), and the cumulative decays within relevant targets in radiation protection. assumption that “daughter radionu- the vascular content of the organ (ac- In radionuclide therapy, however, clides produced from their parent tivity in transit through the organ). For these tissues might not be the only within the body stay with and behave photons and even high-energy ␤-parti- relevant targets within the skeleton metabolically like their parent.” Only cles, emissions within the larger to in- when predicting near-term marrow in the case of incorporated radium are termediate blood vessels of an organ toxicity. the longer-lived radon daughter radio- are considered to contribute to the Where does this leave us? Is the isotopes considered to have an inde- overall mean organ dose. For short- increased interest in ␣-emitters in ra- pendent biodistribution within the range ␣-particles, however, many of dionuclide therapy providing technical ICRP 30 framework. Clearly, research the blood source decays would yield challenges to medical dosimetry that in the area of intratissue mobility of energy deposition events restricted to are intractable? They might be if med- high-Z elements would be of great util- the vessel lumen (blood and blood el- ical dosimetry continues to rely solely ity both to radionuclide therapy and to ements) and thus make no contribution on the physics of energy deposition internal dosimetry for radiation protec- to the parenchymal tissue dose. This and geometric formulations of source tion considerations. fact motivates one to reconsider tradi- and target regions, even if treated sto- Second, the tabulations of cellular S tional models of normal organs as are chastically through microdosimetry. values implicitly consider only the used in the MIRD and ICRP method- The use of ␣-emitters provides the per- mean self-dose from activity originally ologies. Potential improvements in or- fect stimulus to the medical dosimetry bound to the target cell. As an unstable gan dosimetry would then require or- community to fully embrace new ad- daughter is released from its parent gan models in which larger vessels are vances in molecular biology and in decay site, it will diffuse further and explicitly delineated. This can be chal- vivo microimaging and to redefine and further away from the original target lenging in the context of geometric, expand its role and function as it seeks cell. The dose contribution to the target stylized models of organs. With newer improved methods for predicting bio- cell for each daughter emission would developments in tomographic compu- logic response. then abruptly transition to zero, as as- tational models, perhaps such intraor- sumed here, and would decrease con- gan tissue and vasculature differentia- tinuously with increasing distance tion may be feasible (16,17). Wesley E. Bolch University of Florida from the cell. For a higher energy ␣, Finally, ␣-particle dosimetry di- Gainesville, Florida the dose to the cell nucleus might ini- rectly challenges the MIRD schema in tially increase as the Bragg peak of the that, historically, the final quantity of REFERENCES ␣–ionization track is brought inside the interest has always been absorbed 1. Kvinnsland Y, Stokke T, Aurlien E. Radioimmu- cell nucleus. Also, as the unstable dose. Differences in biologic response notherapy with alpha-particle emitters: microdo- daughters migrate away from the target after equivalent energy deposition by simetry of cells with a heterogeneous antigen ex- cell, they will increase their dose con- photons/␤-particles and ␣-particles ob- pression and with various diameters of cells and nuclei. Radiat Res. 2001;155:288–296. tributions to neighboring cells. In fact, viously require scaling of the absorbed 2. McDevitt MR, Barendswaard E, Ma D, et al. An a further extension of the cellular S dose to arrive at a biologically equiv- alpha-particle emitting antibody ([213Bi]J591) for value methodology should consider alent dose quantity. Here, one may ini- radioimmunotherapy of prostate cancer. Cancer Res. 2000;60:6095–6100. multicellular clusters of cells (14).