The role of marrow grafts in irradiation accidents

Abstract

In our first paper describing attempted marrow grafts in human beings, my colleagues and I pointed out the potential application of marrow grafting for treatment of irradiation victims (10). Fortunately, in the intervening years irradiation accidents have been infrequent. The only patient clearly to have benefited from a marrow infusion was fortunate enough to have an identical twin (11). The use of marrow grafts after the accident in Yugoslavia and the Chernobyl disaster has attracted a great deal of publicity, but has not clearly benefited any of the recipients (1, 4, 6). Evaluation and supportive care of individuals exposed to potentially lethal doses of total body irradiation have been described in detail (8). In an actual disaster, management is confounded by associated problems of burns and trauma as described following the Chernobyl accident. For this reason and because of problems in assessing total dose and dose distribution, an appropriate application of marrow grafting to irradiation victims is extremely difficult. Studies in mice had suggested that an infusion of allogeneic marrow might actually be harmful if the recipient had received a barely lethal exposure, the so called “midlethal” effect ( 12). This question was studied in the canine model in which the LD-100 dose is approximately 400 cGy. Dogs were exposed to 360, 450, and 540 cGy and given an infusion of allogeneic marrow from unrelated dogs. The hematologic data and the time of death were approximately the same as for dogs given irradiation only. Dogs given autologous marrow infusions at these exposure levels all recovered. It was concluded that an allogeneic marrow infusion from unrelated animals was neither harmful nor beneficial (9). More recently, this same question has been studied in irradiated dogs given an infusion of DLA compatible marrow. Dogs given 450 cGy showed transient engraftment which permitted them to live long enough for their own marrow to recover. At 600 cGy, mixed marrow chimerism was often observed. At increasing irradiation doses of 600,700, and 800 cGy, fewer dogs survived with autologous marrow recovery and more showed sustained allogeneic engraftment (7). These studies demonstrated that an infusion of histocompatible marrow is not harmful and may result in temporary donor marrow function which may prevent lethality until host marrow can recover. At higher lethal exposures, host marrow may not recover and survival depends upon function of the allogeneic graft. It seems clear that to have any chance of success in protection against lethal irradiation, the patient must be histocompatible with the donor, which means that the donor must be syngeneic, autologous, or HLA-identical. Autologous marrow can be cryopreserved for long periods of time (2). The logistics and cost of marrow storage for large numbers of people at risk of irradiation accidents make this approach impractical. A major problem, as emphasized previously (8), concerns histocompatibility typing of the irradiation victim. Blood samples must be obtained promptly after the irradiation exposure before lymphocytes disappear from the circulating blood. When sufficient numbers of viable cells for standard HLA typing are lacking, it may be possible to complete an informative HLA family study by collecting from the patient nucleated cells by marrow aspiration or skin biopsy to provide DNA for restriction fragment-length polymorphism (RFLP) analysis using one or two restriction enzymes and hybridization with class I (HLA-A, B, C) and class II (HLADR, DQ and DP) probes. This test can be used to identify an HLA identical sibling, but will not be adequate for precise matching with unrelated donors. An RFLP test requires 5 to 7 days to complete. Within 1 to 2 years, sufficient numbers of oligonucleotide probes for DR. DQ, and DP genes should be available to allow for typing of DNA amplified by the polymerase chain reaction (PCR) (3). This method requires very small amounts of DNA, a single cell may be an adequate source, and the entire oligo-