The farnesyltransferase inhibitor tipifarnib (Zarnestra™) was originally developed to target malignancies harbouring RAS mutations. In the first clinical studies with tipifarnib, in adults with leukemia, it was found that patients who responded did not harbour any RAS mutations, suggesting a different mechanism of response. In a previous study we showed that 18% of 150 untreated pediatric AML patients harbour mutations in RAS, of which 30% were CBF-AML. We now studied 44 untreated and 13 relapsed pediatric AML, as well as 22 untreated ALL samples for mutations in RAS, using D-HPLC and direct sequencing. In vitro tipifarnib resistance was determined by a 4-day MTT assay (concentration 0.016-51μM, kindly provided by Janssen Research). The LC50 value, the concentration at which 50% of cells are killed by tipifarnib, was used as a measure of resistance. Patient characteristics were; for untreated AML: 64% boys; median age 9.3 years; median WBC 74.8x109/L; FAB 2xM0, 2xM1, 8xM2, 3xM3, 16xM4, 8xM5, 5x unclassified; for relapsed AML: 77% boys; median age 4.0 years; median WBC 41.6x109/L; FAB 2xM0, 2xM2, 3xM4, 2xM5, 2xM7, 2x unclassified; for untreated ALL: 73%boys; median age 6.0 years; median WBC 10.2x109/L; 15 B-cell precursor (BCP) ALL and 7 T-ALL.

We found RAS mutations in 14 (32%) untreated AML samples (N-RAS : 8 samples exon 1, 1 sample exon 2; K-RAS: 5 samples exon 1 mutations). In relapsed AML 2 samples showed an N-RAS exon 1 mutation (15.4%). In ALL 18.2% had a RAS mutation: an N-RAS exon 1 mutation was found in 2 patients (9.1%) and a K-RAS exon 1 mutation in another 2 patients (9.1%). The distribution of tipifarnib sensitivity was similar in RAS mutated- and non-mutated untreated AML patients [median LC50 RAS mutated 7.1μM (P25-P75: 6.0-9.6μM) vs. non-mutated 4.9μM (P25-P75 2.3-8.2μM); p=0.199]. When we compared N-RAS mutated samples with K-RAS mutated samples there was no statistically significant difference in sensitivity to tipifarnib (median LC50 [p25-p75] 3.2μM [2.9-3.9μM] and 4.9μM [3.7-23.1μM], p=0.20), and comparing them separately with non-mutated AML did not show differences in sensitivity to tipifarnib (p=0.172 and p=0.463 respectively). One out of 9 (11%) N-RAS mutated and 3 out of 5 (60%) K-RAS mutated samples had an LC50 value above the 75th percentile for non-mutated AML and were considered resistant. Within relapsed AML the 2 RAS mutated samples had LC50 values of 0.83 and 6.3μM, versus a median value of 6.9μM for non-mutated relapsed AML. In ALL, we found similar results [median LC50 RAS mutated 7.8μM (P25-75: 4.1-12.8μM) vs. non-mutated 17.4μM (P25-75: 4.5-22.9μM), p=0.3], but the groups were very small.

In conclusion, primary pediatric AML and ALL samples withRAS mutations show similar distributions of tipifarnib sensitivity as samples withoutRAS mutations. Hence, some RAS mutated samples may be relatively in-vitro resistant to tipifarnib, and some non-mutated samples may be relatively sensitive. Therefore, clinical studies with these compounds should not be restricted to RAS-mutated leukemia. Further studies are necessary to determine the molecular targets of farnesyltransferase inhibitors.

Author notes

Corresponding author

Sign in via your Institution