OBJECTIVE: Preliminary studies have demonstrated that the Ras family and related guanosine triphosphate-dependent proteins are overactivated in malignant gliomas and that inhibition of the activation of such proteins, by blockade of their post-translational processing, reduces tumor cell growth in vitro. The current study evaluates the utility of this therapeutic strategy in vivo, using preclinical glioma model systems. METHODS: We examined the efficacy against U-87 human malignant glioma cells, in both subcutaneous and intracranial nude mouse models, of selective peptidomimetic inhibitors of farnesyltransferase (FTI-276) and geranylgeranyltransferase (GGTI-297), which are involved in critical steps in the post-translational processing of Ras and related guanosine triphosphate-dependent proteins. For the subcutaneous model, 2 x 105 U-87 cells were implanted; after measurable tumors were detected on Day 7, animals were treated with either FTI-276, GGTI-297, or vehicle, administered by continuous infusion for 7 days. Differences in tumor volumes among the treatment groups were examined for significance using a Student's t test. For the intracranial model, 2 x 105 U-87 cells were implanted in the right frontal lobe and treatment was initiated on Day 7. In initial studies, animals received a 7-day course of either FTI-276, GGTI-297, or vehicle. In subsequent studies, a 28-day treatment period was used. Comparisons of survival times among treatment groups were performed using a rank-sum test. RESULTS: Although the two agents exhibited comparable antiproliferative activities in previous in vitro studies, an obvious difference in efficacy was apparent in this study. Whereas the geranylgeranyltransferase inhibitor failed to improve survival rates, compared with those observed for control animals, in either the subcutaneous or intracranial model, the farnesyltransferase inhibitor produced objective regression of tumor growth in the subcutaneous model and significant prolongation of survival times in the intracranial model, without apparent toxicity. In the subcutaneous model, tumor volumes for the control, GGTI-297- treated, and FTI-276-treated animals on Day 28 after implantation were 621 ± 420, 107 ± 104, and 18.5 ± 12.7 mm3, respectively (P < 0.05). In the 7- day-treated intracranial model, survival times for the control, GGTI-297- treated, and FTI-276-treated groups were 27.7 ± 2.9, 29.8 ± 2.1, and 43.6 ± 2.7 days, respectively (P < 0.001). In the 28-day-treated intracranial model, survival times for the control, GGTI-297-treated, and FTI-276-treated groups were 29.2 ± 3.7, 28.3 ± 3.9, and 58.7 ± 6.2 days, respectively, with five of six animals in the latter group surviving more than 55 days after tumor implantation (P < 0.001). CONCLUSION: These studies demonstrate that farnesyltransferase inhibition is effective in diminishing the growth of human glioma cells in vivo. Evaluation of this treatment approach in clinical trials is warranted.