Abstract

Background:RHOE is a uniquely farnesylated antagonist of RHOA that regulates cytoskeleton dynamics, cell mobilization and metalloproteinase (MMP)‐mediated tissue invasion. Recent data suggest that CXCL12 can be cleaved by MMPs generating a toxic fragment CXCL12(5–67) that no longer binds to its receptor CXCR4 but that can interact with CXCR3‐B and induce cell death.Aims:To characterize the mechanism of action of tipifarnib in AML, herein, we describe the identification of RHOE as a potential target of tipifarnib that could mediate CXCR3 activation by CXCL12 and response to farnesyl transferase inhibition in AML patients.Methods:CTEP20 was a phase 2, multicenter, open‐label study that investigated the efficacy and safety of the farnesyl transferase inhibitor tipifarnib in 158 older adults with previously untreated, poor‐risk AML. INT11 was a multicenter phase 2 study that evaluated tipifarnib in 252 pts with relapsed/refractory AML. Global gene expression data was generated using the Affymetrix U133A gene chip from 92 pretreatment BM samples from CTEP20 and INT11 pts (GSE8970, GSE5122) and analyzed retrospectively with respect to study outcomes. The Kaplan‐Meier method was employed to estimate survival and progression free survival (PFS) and the analysis of prognostic variables. Clinical trial information: NCT00027872, NCT00101296.Results:We have previously shown that CXCL12 expression determines bone marrow homing and sensitivity to tipifarnib in AML (Gualberto, ASH 2017). We recently identified RHOE in a screening of AML databases for genes of uniquely farnesylated proteins that are co‐expressed with CXCL12 (TCGA AML, provisional N = 163, r = 0.723, p < 0.0001). RHOE and CXCL12 gene expression were also strongly correlated with VCAM1 expression (p < 0.0001), suggesting stromal origin. We then investigated the activity of tipifarnib in AML patients with bone marrow expression of RHOE, CXCL12 and CXCR3. Analysis of bone marrow samples from AML patients enrolled in CTEP20 showed that pre‐treatment expression of CXCR3, CXCL12, and RHOE were highly significantly (p < 0.0001) predictive of complete response (CR) on tipifarnib treatment with ROC AUC values of 0.862, 0.782, and 0.736, respectively. The most discriminating index was the product of the expression of CXCL12 (ligand) and CXCR3 (receptor). The 80th percentile of the product of the expression of CXCL12 and CXCR3 had a 100% PPV and 94% NPV to identify a CR outcome, assuming a true 10% CR rate in the unselected population. PFS in the high CXCL12 x CXCR3 subset was 433 days versus 64 days in the low expressing subset (HR = 0.24, p = 0.005, N = 34). CXCR3 expression also predicted long term survival in study INT11. Median survival times in relapsed/refractory AML patients with CXCR3 expression tertiales 1st‐3rd were respectively 54, 75 and 182 days (p = 0.009, N = 58). Ex vivo treatment of bone marrow stromal cell cultures with tipifarnib showed loss of CXCL12 immunoreactivity and current work investigates its potential tipifarnib‐induced metabolism.Summary/Conclusion:These findings provide a potential RHOE‐mediated mechanism of action of tipifarnib in AML patients with bone marrow CXCL12 and CXCR3 expression.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call