Future analyses of total body irradiation

Early analyses of the effects of “accidental” Total Body Irradiation (TBI) led to the discovery of the possibilities of bone marrow transplantation (4-6). The logical consequence of these new insights was the design of treatment protocols in which patients with lethal bone marrow disorders were given deliberate TBI followed by an infusion of unaffected bone marrow cells. Some of such treatment protocols were found to be curative and constitute an important improvement in the management of previously lethal hematologic disorders (1,8, 10). TBI causes bone marrow aplasia. Bone marrow transplant teams are usually headed by medical specialists who have great experience with management of bone marrow aplasia, that is, medical oncologists or hematologists. Routinely, these specialists use chemotherapy for cancer patients which will cause temporary bone marrow failure that is reversible with time or with infusion of bone marrow from a donor source. However, these specialists are not familiar with radiation in small fields or large fields, such as TBI. Bone marrow transplantation is a toxic procedure with at least a 10% procedural lethality rate in experienced centers ( 1, 10). Most of the morbidity and mortality of bone marrow transplantation is usually blamed on the use of TBI, as TBI constitutes to the average bone marrow transplant physician the least familiar component in the preparation (conditioning) of patients. There are known advantages to the use of TBI over chemotherapy in the conditioning of patients, such as TBI’s current status as the most effective immunosuppressive conditioning agent ( 15). Moreover, all cells in the body will be treated (irrespective of vascular supply or sanctuary site); cells outside S-phase are sensitive to radiation and parts of the body can receive more or less radiation at will by changes in the TBI technique (e.g., field within a field, or partial shielding and in the future radiolabeled antibodies) ( 14, 15) and prior treatment will not have induced resistance to radiation in tumor cells. Gale and coworkers discuss their hypotheses about the role of TBI in conditioning patients prior to BMT (3). The important issues in any analysis of TBI are dose, fraction size, and endpoint selection. These issues are not adequately addressed in Gale’s editorial and will be dealt with briefly here. In contrast to their review that new “exotic” mechanisms must be invoked to explain the presumed benefits and risks of TBI, we are convinced that regular physics, radiation biology, and radiation oncology concepts apply. Most of the TBI “problems” can be analyzed in experimental animals. Some (e.g., tumor recurrence after TBI) can only be explored in human patients. In a recent AAPM report, the dosimetric problems of TBI are well illustrated (11). The contour of the human body, tissue inhomogeneities in the human body, and beam off-axis factors are the major contributors to wide differences in dose within a single patient, between patients and most importantly, between institutions. Unfortunately, dosimetric studies are not performed or reported in sufficient detail by the great majority of transplant centers to allow for a) a prospective or retrospective dose determination in the most important subsections of the body and b) comparisons with other centers. Relatively small dose differences are biologically significant, as it is well known from experimental animal as well as clinical studies of TBI that the dose effect curves for most endpoints are steep. Fraction sizes from 1.1 to 9.0 Gy have been used, with total doses up to approximately 9.0 Gy for single fraction TBI and 1214 Gy for fractionated TBI. Small fractions (~2.0 Gy) are suspected to cause fewer late effects. Their introduction in TBI procedures is probably not warranted, as late effects (with the exception of lung) are not dose limiting, immunosuppressive and tumoricidal effects of TBI decrease with fraction size, and the logistics of small, repeated TBI fractions are unduly complicated (11). It is clear that different TBI procedures using a different total dose and fraction size cannot be compared, without extensive radiobiological “normalization.” Normalization formulas will be different for different endpoints. The major reason why the analysis of TBI is complex is the multitude of endpoints: a) Acute side effects. b) Immunosuppressive effects. c) Tumoricidal effects and d) Late side effects. Elsewhere we have argued