Patients with myelodysplastic syndrome (MDS) suffer from, and often succumb to, serious infections or bleeding because of severe pancytopenia, with or without progression to acute leukemia. Current therapy for MDS does not yield a survival advantage, and no form of therapy has proven advantageous enough to warrant being considered standard. Mutated RAS genes are found in about 20 percent-30 percent of patients with MDS. Furthermore, in patients without RAS mutations, Ras may still be activated indirectly because of the effects of other genetic aberrations. It is now known that in order for even mutated Ras to be active, it must move from the cytoplasm to the plasma membrane, a process in which the addition of a farnesyl group by the farnesyltransferase enzyme (FTase) plays an important role. Recently, novel compounds which act as FTase inhibitors and hence interfere with Ras activation have been developed. The goal of this project will be to perform the first clinical study of a novel FTase inhibitor (R115777) in MDS patients and to ascertain the optimum dose for biologic and clinical response. (R115777 is available from the Cancer Therapy and Evaluation Program of the National Cancer Institute.) Our hypothesis is that the patients most likely to respond are those who have baseline Ras activation which can be downregulated after treatment with an FTase inhibitor. We will start with a Phase I dose-finding study. The objective of the initial trial will be to ascertain patient tolerance to R115777, and to estimate optimum biologic dose. These trials will then be expanded to determine subsets of patients who are most likely to show biologic and clinical response. The optimum dose and subsets of patients with biologic response to R115777 will be ascertained by measuring the relationship between administered dose and plasma concentration of R115777 and effects on biologic endpoints: FTase activity, Ras and lamin B farnesylation, and downstream effectors of Ras (MAP kinase and IL-1 (the latter based on our recent results demonstrating that Ras activation leads to autocrine IL-1 production in leukemia)). In summary, this work will determine a dose of R115777 which is both tolerable and maximizes the impact of this molecule on the FTase/Ras system. This dose will then be used to determine if R115777 is effective in the treatment of MDS and if certain molecular/biologic markers are predictive of response. These studies will also enhance our understanding of the relationship between inhibition of farnesylation, effects on downstream effectors, and antitumor activity in MDS, and should be useful as a paradigm for the clinical application of other FTase inhibitors.