Impact of the conditioning regimen - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 165

mucositis and more severe GVHD in patients receiving higher doses [20] .
On the other hand, clinical application of keratinocyte growth factors had
no major impact on GVHD although affecting mucositis [21] , and substitu-
tion of MMF for methotrexate in the setting of GVHD prophylaxis affected
mucositis but had no clear impact on GVHD.
Endothelial damage and toxicity
Further targets with high sensitivity to conditioning related damage are
endothelial cells (ECs). They recently gained major interest in the patho-
physiology of GVHD. In vitro studies revealed apoptotic damage of ECs by
irradiation doses as low as 4 Gy, and adding LPS enhanced both apoptosis
but also expression of adhesion molelcules such as ICAM 1 [22,23] . Clinically
circulating endothelial cells are detected more rapidly after TBI [24] , and
sera from patients obtained in the course of conditioning contain factors
inducing endothelial activation [25] . Endothelial damage is followed by early
neovascularization via donor cells which contribute to severity of GVHD as
recently shown by Penack [26,27] . In line with this concept, chimerism of
endothelial cells is more easily detected after full intensity conditioning [28] .
167
Role of inflammation and damage to immunological
organs
Conditioning-related inflammation and cytokine release
EXPERIMENTAL EVIDENCE
Besides translocation of bacterial ligands, direct induction of inflamma-
tory cytokines by irradiation and cytotoxic treatment has been postulated:
In vitro induction of NFκB and subsequent inflammatory cytokines has
been shown in blood cells [29,30] and several target tissues such as the lung
or brain. In parallel, adhesion molecules are induced which facilitate inva-
sion of inflammatory cells [31] . Xun and colleagues showed strong induc-
tion of TNF-α but also IL-1 and IL-6 by TBI and proved its involvement in
pathogenesis of GVHD by either postponing transplantation or prophylac-
tic use of TNF-neutralizing agents [1] . In parallel, we were able to show that
a single injection of a TNF-neutralizing antibody prior to TBI and trans-
plantation was able to protect from GVHD [32] .
CLINICAL EVIDENCE
In our clinical analysis of systemic TNF-α release in allogeneic transplanta-
tion, one of the most striking findings was that TNF-α release during pre-
transplant conditioning predicted a high incidence of treatment-related
mortality due to endothelial complications and GVHD [33] . A pathophysio-
logical role of early cytokine release was suggested by results of a phase II trial
using a TNF-neutralizing antibody during conditioning [34] which reduced
early GVHD compared to historical controls. Unfortunately, phase III trials
testing this approach could not be completed. A recent study employing
etanercept from pretransplant until day 56 failed to show beneficial effects
on GVHD, but again demonstrated that TBI is a stronger inducer of TNF and
soluble TNF-R release, which predisposes to more severe GVHD [35] .