Pharmacodynamic Monitoring Predicts Outcomes of CCR5 Blockade as Graft-versus-Host Disease Prophylaxis

Abstract

Blocking lymphocyte trafficking after allogeneic hematopoietic stem cell transplantation is a promising strategy to prevent graft-versus-host disease (GVHD) while preserving the graft-versus-tumor response. Maraviroc, a CCR5 antagonist, has shown promise in clinical trials, presumably by disrupting the migration of effector cells to GVHD target organs. We describe a phosphoflow assay to quantify CCR5 blockade during treatment with maraviroc and used it to evaluate 28 patients in a phase II study. We found that insufficient blockade of CCR5 was associated with significantly worse overall survival (HR, 10.6; 95% CI, 2.2 to 52.0; P = .004) and higher rates of nonrelapse mortality (HR, 146; 95% CI, 1.0 to 20,600; P = .04) and severe acute GVHD (HR, 12; 95% CI, 1.9 to 76.6; P = .009). In addition, we found that pretransplant high surface expression of CCR5 on recipient T cells predicted higher nonrelapse mortality and worse GVHD- and relapse-free survival. Our results demonstrate that pharmacodynamic monitoring of CCR5 blockade unravels interpatient variability in the response to therapy and may serve as a clinically informative biomarker.

Keywords: Chemokine receptor blockade; GVHD; Maraviroc.

Fig. 1. CCR5 blockade sufficiency can be determined by phosphoflow

(A) Representative flow plot from patient 04712-015 on day −6 pre-transplant, prior to initiation of maraviroc (MVC). CD8 T-cell pCCR5 mean fluorescence intensity (MFI) increases in response to CCL4 stimulation and decreases in response to exogenous MVC. (B) Average fold-change in pCCR5 MFI on day −6 (patient not yet on MVC) and day 0 (patient on MVC) CD8 T-cells. Stimulation has an effect on day −6 but not day 0; exogenous MVC has an effect on day −6 that is ameliorated by day 0. (C) Representative flow plots and MFIs of sufficiently blocked patient 04712-018 (top) and insufficiently blocked patient 04712-005 (bottom). (D) Heterogeneity in %change in pCCR5 MFI on day 0 in response to exogenous MVC. Those patients whose MFI dropped by more than 20% compared to the unstimulated control were considered to have insufficient blockade, as additional MVC had an additional effect. *** = p<0.001.

Fig. 2. Blockade insufficiency is predictive of worse outcomes

(A–D) Insufficient blockade on day 0 is associated with increased risk of death (A. HR=10.6, p=0.004), GRFS (B. HR=3.4, p=0.04), aGVHD3–4 incidence (C. HR=12, p=0.009), and non-relapse mortality (D. HR=146, p=0.04).

Fig. 3. Peri-transplant increase in T-cell CCR5 surface expression

CCR5 expression on CD4 (A) and CD8 (B) T-cells. On both CD4 and CD8 T-cells, recipient pre-transplant CCR5 expression was elevated compared to healthy donor control, increased by day of transplant in response to maraviroc and chemotherapy, and dropped to an intermediate level. (C) Representative CCR5 expression flow plots of pre-transplant “CCR5lo” and “CCR5hi” patients’ CD8 T-cells. * = p<0.05, ** = p<0.01, *** = p<0.001.

Fig. 4. Recipient pre-transplant CD8 T-cell CCR5 expression is predictive of outcomes

(A–C) High CD8 T-cell CCR5 surface expression is associated with increased risk of death (A. HR=3.9, p=0.06), GRFS (B. HR=3.2, p=0.03), and non-relapse mortality (C. HR=6.2e08, p<0.001).