Chemokines and graft-versus-host disease - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 410

would reduce GVHD and enhance (or at least maintain) GVT effects. Target-
ing either CCL3 or CCR1 resulted in decreased GVHD severity but preserva-
tion of GVT [33,66] , suggesting that certain chemokines or their receptors
may allow for specific intervention. However, both the complexity of the
chemokine receptor:ligand network orchestrating immune responses and
the redundancy of chemokine expression on tumor cells can make selec-
tive inhibition problematic [103] . For example, using a xenograft model of
human breast cancer, administration of antibodies to CCL2 was initially
found to result in significant tumor regression in vivo. While attempts to
demonstrate anti-tumor effects using anti-CXCL8 antibodies were unsuc-
cessful (many tumors also express CXCL8), the administration of antibodies
to CXCL8 and CCL2 in combination resulted in much greater anti-tumor
effects [104] . This suggests that neutralization of multiple chemokines may
be necessary for significant and sustained in vivo anti-tumor effects. It
would be of interest to use such an approach in tumor-bearing mice receiv-
ing an allogeneic HCT in order to develop strategies that will successfully
separate the toxicity of GVHD from the beneficial GVT effect.

Advances in the clinical application of chemokine
biology during GVHD

415

As highlighted throughout this chapter, laboratory insights have led to
the development of focused translational research studies evaluating the
expression of specific chemokines and their receptors from leukocytes iso-
lated from either peripheral blood, lesional tissue, or both, from patients
with acute or chronic GVHD. In a more general approach, Jaksch and col-
leagues evaluated the expression of CCR1, CCR2, CCR5, and CXCR3 by real-
time PCR from the blood of 50 patients after allogeneic HCT [105] . Thirty-six
of 50 patients developed GVHD and 10 developed recurrent disease for a
total of 46 episodes of GVHD that could be interrogated. The expression of
each chemokine receptor was increased in at least 70% of samples collected
during GVHD. The median increase in the four markers ranged from 3- to
12-fold, and enhanced expression was observed at, or just before, the onset
of GVHD in the majority of cases. Chemokine receptor expression was also
elevated at the time of bacterial, fungal, or viral infections so that an increase
in mRNA expression for this panel of receptors was not specific for GVHD.

Investigators have also used a number of discovery, proteomics-based
strategies to identify biomarkers predictive for GVHD. One of the more
common proteomics approaches utilizes high-performance liquid chro-
matography followed by mass spectroscopy for protein isolation after the
removal of common proteins such as albumin, IgG, and transferrin. Hori
and colleagues used this approach in mice to identify the precursor of CCL8
and found that its expression was markedly increased in mice with GVHD
compared to controls [106,107] . To confirm this clinically, they used a simi-
lar strategy on the serum of three HCT recipients, of which two developed
GVHD, and one healthy control. In the patients with GVHD, CCL8 increased
at days 9-10 post-HCT and continued to be increased through the first
month post-HCT, whereas CCL8 was minimally detected in the patient
without GVHD and the healthy control. To validate these findings, CCL8
was measured in serum by ELISA. Fourteen patients who underwent HCT