Natural CD4 CD25 FOXP3 regulatory T cells in graftversus-host disease - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 254

after reduced intensity, conditioning, suggesting that homeostatic Treg pro-
liferation in lymphopenic recipients may be a confounding factor. For BM
grafts, however, an influence of Treg numbers or frequencies on incidence or
severity of acute GVHD could not be confirmed in a limited number of BMT
recipients [129] . Overall, these few studies suggest that the number of donor
Treg in the graft may influence the development of acute GVHD after SCT,
but definitive conclusions cannot be drawn as only low patient numbers
were included in these trials and the large variability in the transplantation
procedures and immunosuppressive regimens may have a stronger influ-
ence than Treg numbers. Unfortunately, systematic and quality-controlled
phenotypic analyses of stem cell graft compositions are not performed
routinely in the clinical setting. Yet, even in animal models, in which uni-
form conditioning regimens are applied and no additional pharmacologic
immunosuppression is administered, only modest differences are observed
in GVHD severity and survival if donor Treg are deleted from the stem cell
graft. A clinically significant donor Treg effect was mainly observed if the fre-
quency of cotransplanted donor Treg was increased to unphysiologic levels.

Treg monitoring after SCT

257

Miura and colleagues were the first to describe an inverse correlation of
Treg numbers and the incidence and severity of acute GVHD after BMT
[130] . As antibodies against FOXP3 were not yet available, they used qRT-
PCR to determine Treg contents in peripheral blood at the time of GVHD
manifestation. Compared to time-matched transplant recipients without
GVHD, the GVHD cohort showed lower levels of FOXP3, and FOXP3 further
declined with increasing GVHD severity in both acute and chronic GVHD.
The parallel quantification of T cell receptor gene rearrangement excision
circles suggested that patients with GVHD had a diminished thymic output
of both Treg and Tconv. These findings were recently confirmed in a large
prospective single-center study using freshly isolated blood samples and
state of the art flow cytometry for the quantification of Treg cell numbers.
In this study, Treg frequencies were determined in 90 patients after autolo-
gous SCT, in 65 patients after allogeneic SCT without GVHD, and sequen-
tially in 60 patients before and after GVHD onset. The authors also found
diminished Treg cell numbers in peripheral blood in patients with acute
GVHD, declining Treg numbers with increasing disease severity, and signifi-
cant differences in nonrelapse mortality and overall survival if patients were
grouped according to the median Treg frequency in peripheral blood at the
time of disease onset [131] . Based on these findings, the authors concluded
that the determination of Treg frequencies may be a valuable prognostic
marker for acute GVHD. Yet, the gradual differences in Treg frequencies and
the arbitrary cut-off for high versus low Treg numbers (median) most likely
provide insufficient sensitivity and specificity for a reliable biomarker. A
series of smaller immune monitoring trials also described reduced periph-
eral blood Treg numbers in acute GVHD [129,132-134] , although some
investigators did not observe a difference [135-138] . In chronic GVHD the
situation is even more disturbing, as reduced [139,140] , normal [136,141] ,
and even increased [132,137,142,143] Treg cell numbers or frequencies have
been described. In a few studies Treg were measured not only in periph-
eral blood, but also in GVHD target tissues. Two groups found reduced Treg