RISK OF ACUTE MYELOGENOUS LEUKEMIA AND MYELODYSPLASIA FOLLOWING CANCER-TREATMENT

Modern cancer treatment has substantially increased the duration of su rvival and curability of patients with various malignancies. Cure rate s have increased dramatically for a number of paediatric malignancies, Hodgkin's disease (I-ID) and testicular cancer. Less impressive, but nonetheless clear improvements in survival have been achieved for brea st cancer, ovarian cancer and non-Hodgkin's lymphoma (NHL). The succes sful treatment of these malignancies has involved the use of multi-age nt chemotherapy (CT) and high-dose radiation therapy. Now that a subst antial group of cancer patients has such a favourable prognosis, it ha s become increasingly important to evaluate the long-term complication s of treatment. Paradoxically, research conducted over the last two de cades has clearly demonstrated that some treatments used to control ca ncer have the potential to induce new (second) malignancies. Of all tr eatment-related second malignancies, leukaemia is considered to be one of the most serious (Tucker, 1993). The excess risk of leukaemia is h igh following treatment of various primary malignancies. Moreover, sec ondary leukaemia has a poor prognosis, since most of these leukaemias are resistant to therapy. Increased risk of second leukaemia has been observed after both radiotherapy (RT) and CT. The ability of radiation to induce leukaemia was first observed in patients irradiated for ank ylosing spondylitis (Court-Brown and Doll, 1957; Upton, 1975). Most kn owledge about the leukaemogenic effect of radiation in humans has come from epidemiological studies of atomic bomb survivors in Japan, occup ationally irradiated workers, and patients treated with radiation for benign and malignant diseases (Boice, 1988). Insight into the complex dose-response curve for radiation-induced leukaemia has been gained on ly recently (Boice et al, 1987; Preston et al, 1994). The carcinogenic potential of CT was recognized much later than that of ionizing radia tion. This has obviously to do with the fact that chemotherapeutic age nts were not introduced in cancer control until the late 1940s (Rieche , 1984), while modern multi-agent combination CT, which is now known t o have the strongest carcinogenic potential, was not used until the 19 60s. Until the introduction of combination CT, patients treated with a ntineoplastic agents did not live long enough for an increased risk of second malignancies due to treatment to become manifest. A review of the literature indicates that, generally, it takes 5-20 years from the introduction of a drug into clinical practice before a carcinogenic e ffect of the agent becomes evident (Sieber and Adamson, 1975; Stolley and Hibberd, 1982; Rieche, 1984). Evidence of the carcinogenicity of c hemotherapeutic agents has come not only from clinical observations of second malignancies in patients treated with these drugs, but also, t o a great extent, from in vivo and in vitro laboratory studies. Pionee ring work in this field (Haddow et al, 1948; Shimkin et al, 1966; Schm ahl et al, 1977) was conducted before clinical studies had shown incre ased risk of second malignancies following CT. The leukaemogenicity of CT in man was first discovered in patients treated for multiple myelo ma. The first report suggesting a role of alkylating agents was publis hed in 1970 (Kyle et al, 1970), and the association was confirmed in a number of subsequent studies (Rosner and Grunwald, 1974; Bergsagel et al, 1979). The increased incidence of leukaemia in myeloma patients f ollowed the introduction of melphalan and other alkylating agents in 1 962. MOPP combination CT for HD (consisting of mechlorethamine, vincri stine, procarbazine and prednisone) was introduced in 1967; the leukae mogenic potential of this regimen became evident in reports published in 1973, 1975 and 1977 (Bonadonna et al, 1973; Canellos et al, 1975; C oleman et al, 1977). The predominant type of leukaemia associated with alkylating agent CT has been acute myelogenous leukaemia (AML). After the early recognition of increased risk of AML in survivors of HD and multiple myeloma, strongly increased risks have now also been demonst rated following combination CT for a large number of other malignancie s (Levine and Bloomfield, 1992). CT appears to be far more potent than RT in inducing leukaemia. This review addresses the risk of secondary AML and myelodysplasia (MDS) following treatment of individual primar y malignancies. Rather than giving a comprehensive review, this discus sion will focus on large patient series that were published recently. Emphasis will be on the contribution of various treatment factors to t he risk of AML and MDS.