`

`I 02 T. Toubai et al
`
`contributed to improved outcomes after allogeneic HCT. 1 Yet the major complication
`of allogeneic HCT, graft-versus-host disease (GVHD), remains lethal and limits its
`wider application.2 Traditionally, clinical GVHD occurring prior to I 00 days after
`HCT is called acute, and that occurring after I 00 days is called chronic GVHD. 3---S It
`is important to recognize that this traditional definition was based on the temporal
`rather than the clinical or pathophysiological nature of GVHD. This definition is not
`satisfactory, because acute and chronic GVHD have distinct clinical features that can
`sometimes present concomitantly and/or independently of the time after transplant.
`A recent National Institutes of Health (NIH) consensus classification of GVHD in(cid:173)
`cludes late-onset (after day I 00) acute GVHD and an overlap syndrome that has
`features of both acute and chronic GVHD.6
`The clinical manifestations of acute GVHD occur in the skin, gastrointestinal tact
`and liver.7 Patholofically, the sine qua non of acute GVHD is selective epithelial damage
`
`of target organs.8• It typically manifests within 20--40 days following full-intensity allo(cid:173)
`geneic HCT, and may be slightly more delayed following reduced-intensity allogeneic
`HCT.
`Chronic GVHD is a syndrome that typically presents more than I 00 days after
`transplant; it can occur either as an extension of acute GVHD (progressive), or after
`a disease-free interval (quiescent), or with no precedent (de novo) acute GVHD.4-6,IO
`It is important to note that strategies which have improved acute GVHD rates - such
`as cord-blood transplants - have not as clearly affected the incidence of chronic
`GVHD. 1 1 Additionally, strategies that have not significantly altered acute GVHD
`rates - such as peripheral blood stem cell transplants (PBSCTs) - appear to have in(cid:173)
`creased the incidence of chronic GVHD. 12 Chronic GVHD has myriad manifestations,
`and unlike acute GVHD it can affect multiple organ systems. In general, it can present
`as lichen-type features, dryness, strictures or sclerosis of the several organs including
`skin (sclerosis), mouth (xerostomia), eyes (xerophthalmia), vagina, esophagus, liver,
`lung (bronchiolitis obliterans), fasciitis, serositis Qncluding pericardia! or pleural effu(cid:173)
`sions), and rarely kidneys (nephrotic syndrome). ·13 Pathologically, depending on the
`severity, a plethora of features are observed. However, in contrast to acute GVHD,
`it is most often characterized by fibrosis of the affected organ. 14
`The discordance between acute GVHD rates and chronic GVHD, the time of onset,
`the organ system involvement, types of tissue damage, the differences in clinical fea(cid:173)
`tures, and the kinetics suggest that although both acute and chronic GVHD occur after
`allogeneic HCT, the underlying immunopathological mechanisms might be distinct. The
`pathophysiology of acute and chronic GVHD based on clinical data and experimental
`studies is discussed below.
`
`PATHOPHYSIOLOGY OF GVHD
`
`Rapid advances have allowed for refinements of the model that was proposed several
`decades ago by Billingham for the development of GVHD. 15 The availability of several
`experimental models that reflect the early GVH process has substantially enhanced
`our understanding of the biology of acute GVHD. 16 Target-tissue damage in acute
`GVHD is caused by cytopathic donor T cells that respond to the genetically disparate
`host polymorphic antigens, which are presented by host and/or donor antigen-pre(cid:173)
`senting cells (APCs). This damage is further amplified by the non-specific inflammatory
`mediators. This is now conceptualized to occur in sequential phases involving complex
`
`

`

`GVHD pathophysiology: acute versus chronic
`
`I 03
`
`interactions between a variety of cytokines, chemokines, ad~tive and innate immune
`19
`subsets (summarized in several excellent recent reviews).2· 1
`-
`Similar advances in the understanding of the biology of chronic GVHD have not
`been made, in part due to the absence of appropriate experimental models that mimic
`all the features of chronic GVHD. Some murine models - depending on the strain
`combinations (the type of genetic disparity), conditioning regimen, and type and
`amount of donor cells - produce certain features of chronic GVHD, such as the fibro(cid:173)
`sis of skin, lung or lupus nephritis or liver damage. 12 But no single model captures all of
`the features and kinetics of chronic GVHD. This lack of appropriate experimental
`models might be because of the differences between humans and experimental spe(cid:173)
`cies. For example, in contrast to murine models, the kinetics of clinical chronic
`GVHD in humans is slower and is observed only after prophylaxis and/or treatment
`for acute GVHD. Even when clinical chronic GVHD arises de novo and in the absence
`of active immunosuppression, it is not possible to definitively rule out the impact of
`either GVH prophylaxis and/or subclinical acute GVHD on the subsequent develop(cid:173)
`ment of chronic GVHD. It is therefore important to consider these caveats when
`attempting to understand the biology of chronic GVHD from experimental models.
`The genetic and immunological basis of acute and chronic GVHD biology is discussed
`below.
`
`GENETIC BASIS OF GVHD
`
`Human leukocyte antigens (HLA) matching
`
`The most important immunogenic proteins that contribute to the intensity of GVH re(cid:173)
`action are the human leukocyte antigens.20 HLA proteins are highly polymorphic and
`are encoded by the major histocompatibility complex (MHC). Class I (HLA-A, -B,
`and -C) antigens are expressed on almost all nucleated cells. Class II antigens (DR,
`DQ, and DP) are primarily expressed on hematopoietic cells, but their expression
`can be induced in many other cell types following inflammation or injury.20 Several lines
`of evidence demonstrate that regardless of the type of graft or the intensity of the pre(cid:173)
`parative regimen, acute GVHD is directly related to the degree of HLA mismatch
`(approximately 40% in recipients of HLA-identical grafts but increases to 60-80% in re(cid:173)
`cipients of unrelated or one-antigen HLA-mismatched grafts).21
`25 These clinical obser(cid:173)
`-
`vations have been validated in experimental models wherein acute GVHD leading to
`rapid mortality is observed across many MHC disparate strain combinations. 16
`Although the role of H LA disparity has been analyzed in the overall outcomes after
`allogeneic HCT, relatively fewer studies have attempted to correlate it with the inci(cid:173)
`dence and severity of chronic GVHD. Some clinical studies have demonstrated a direct
`28 In con(cid:173)
`association between HLA-A, -B, -C (class I) disparity and chronic GVHD.2
`6--
`trast to acute GVHD, only a few MHC-mismatched animal models recapitulate some
`features of chronic GVHD. A recent study demonstrated that acute GVHD in an
`MHC-mismatched model resulted in the emergence of donor-reactive donor T cells
`in the host, which caused severe 'autoimmune' colitis upon adoptive transfer back
`into donor but not host type mice.29 Collectively, these observations nonetheless
`make a strong case for the critical role of HLA disparity in both acute and chronic
`GVHD. However, it remains to be determined whether specific mismatches are
`more critical for chronic than acute GVHD.
`
`

`

`I 04 T. Toubai et al
`
`Minor histocompatibility antigen (miHAs}
`
`In the context of MHC-matched HCT, as is the case with most clinical allo-HCT, donor
`T cells recognize MHC-bound peptides derived from the protein products of polymor(cid:173)
`phic genes (minor histocompatibility antigens, miHAs) that are present in the host but
`not in the donor.3
`36 The critical role of miHA disparity for the development of acute
`0---
`GVHD has been demonstrated by the development of acute GVHD in substantial
`numbers of patients ~40%) receiving HLA-identical grafts and optimal post-grafting im(cid:173)
`mune suppression.33
`7 The pattern of expression of miHAs might account for the
`•
`unique target organ involvement in acute GVHD.31
`•37 This notion is supported by ex(cid:173)
`perimental murine studies, which show that distinct miHAs dictate the phenotype and
`target organ involvement of acute GVHD. 38 Experimental data have demonstrated that
`not all of the miHAs are equal in their ability to induce lethal GVHD, and that they
`40 Disparity in a single immunodominant
`show hierarchical immunodominance.39
`•
`miHA is not sufficient to cause acute GVHD, although T cells targeting single miHA
`42 However, to date no spe(cid:173)
`can induce tissue damage in a human skin explant model.41
`•
`cific miHAs that are common and have equivalent immunogenicity in most patients
`have been identified. Nonetheless, the polymorphic miHAs are the critical antigenic
`targets for inducing acute GVHD in MHC-matched HCT.
`By contrast, on the basis of its clinical features, chronic GVHD has been considered
`to be an 'autoimmune' disease. Some experimental studies have shown that T cells
`from animals with chronic GVHD are specific for a common (shared between host
`44
`and donor) determinant of MHC class II molecules43
`, and are therefore considered
`•
`to be 'autoreactive'. These autoreactive cells of chronic GVHD are associated with
`a damaged thymus and negative selection.7.4~B This would then indicate that the
`non-polymorphic antigens expressed in donor and recipient rather than the disparate
`polymorphic miHA antigens are the likely targets in chronic GVHD. Despite the ex(cid:173)
`perimental evidence and clinical similarity to autoimmune diseases, there are no clear
`clinical data on the isolation of donor-derived T cell clones that recognize non(cid:173)
`polymorphic antigens from both donor and recipient.
`By contrast, emerging clinical data show a strong correlation between the presence
`of immune responses against ubiquitously expressed miHAs and chronic GVHD.
`Furthermore, because chronic GVHD occurs (a) only after allogeneic HCT, (b) acute
`GVHD is its main risk factor 12,and (c) is distinct from 'syngeneic' GVHD caused by
`improper thymic selection48
`, it is possible that chronic GVHD is caused by T cells
`that have undergone chronic antigen stimulation due the presence of ubiquitous
`miHA antigens. The similarity to clinical features of 'autoimmune' diseases might there(cid:173)
`fore be the result of chronic stimulation-induced target organ damage, which perhaps
`happen to be miHAs for chronic GVHD and non-polymorphic 'auto-antigens' for au(cid:173)
`toimmune diseases. This concept is supported by the recent observations in female(cid:173)
`to-male HCT demonstrating a strong correlation with the presence of antibodies to
`Y-chromosome-encoded histocompatibility antigens and chronic GVHD.49,so In any
`event, even if miHA are the targets, it remains unknown whether these are the
`same as the ones targeted in acute GVHD. A recent elegant murine study in fact sug(cid:173)
`gested that the type and selection of immunodominant miHAs determines the target
`and character of GVHD damage.38
`Lastly, because of 'epitope spreading' and the failure of appropriate regulatory
`mechanisms - either as a consequence of acute GVHD or its treatment and/or
`prophylaxis -
`it is also conceivable that donor T cells which recognize both
`
`

`

`GVHD pathophysiology: acute versus chronic
`
`I 05
`
`non-polymorphic and mi HA epitopes might cause and perpetuate chronic GVHD. Thus,
`in contrast to the antigenic targets for acute GVHD, the nature of the relevant immuno(cid:173)
`genic targets for chronic GVH D remains, as yet, largely speculative.
`
`KIR and cytokine polymorphisms
`
`The role of human killer-cell immunoglobulin-like receptors (KIRs) has been a focus of
`intense study in recent years.20·51 ·52 Polymorphisms in the transmembrane and cyto(cid:173)
`plasmic domains of KIR receptors govern whether the receptor has inhibitory
`potential or activating potential, and the balance between them regulates NK cell
`activation.53 Recent experimental data and some clinical observations suggest that
`KIR mismatch in GVH direction is associated with GVL without exacerbation of acute
`GVHD, while some others have not shown any clear association.20·51·52·54·55 There is
`a paucity of both experimental and clinical data on the role of KIR mismatches and out(cid:173)
`comes of chronic GVHD.
`Several cytokines, such as interleukins IL- I, IL-6, IL-8, tumor necrosis factor
`rJ. (TNFr1.), released during the •~okine storm' phase of acute GVHD, play a critical
`role in amplifying acute GVHD. 1 Gene polymorphism studies with TNF, IL-I 0, inter(cid:173)
`feron le (IFNy) variants have correlated with acute GVHD in some but not all stud(cid:173)
`ies.56--- 8 Genetic polymorphisms of proteins involved in innate immunity, such as
`NOD2/CARD I 5, in both the donors and recipients were recently shown to have
`a strong association with acute gastrointestinal GVHD.59 Despite the obvious inflam(cid:173)
`matory state of chronic GVHD, there are no clear experimental data on the role of
`proinflammatory cytokines in chronic GVHD, although increased levels of TNFr1.
`and IFNy transcription might predict the onset of extensive chronic GVHD.60
`Despite the lack of experimental data, some clinical studies have evaluated the role
`of certain donor and recipient cytokine polymorphisms. These studies showed that IL(cid:173)
`I 0, IL- I rJ., IL- I Ra and IL-6 polymorphisms have shown variable association with chronic
`GVHD.58 A common problem with most of these studies is the heterogeneity and the
`small numbers of patients. Nonetheless these data, when taken together, suggest that
`both acute and chronic GVHD are likely to be modulated by non-HLA polymorphisms
`in addition to mismatched HLA and miHAs.
`Thus genetic variation in both patient and donor can significantly affect both acute
`and chronic GVHD by causing disparity of transplant antigens and modulating the in(cid:173)
`tensity of the immune responses by alteration of the activation and function of several
`immune cells, as discussed below.
`
`IMMUNE-CELL SUBSETS IN GVHD
`
`APCs in GVHD
`
`The earliest phase of acute GVHD is set in motion by the damage caused by the
`underlying disease and the conditioning regimens that cause multiple changes and
`6
`enhance the secretion of proinflammatory cytokines such as TNFr1. and IL-1.61
`---6
`This results in activation of host APCs which is further amplified by the systemic
`translocation of immunostimulatory microbial products such as lipopolysaccaride
`(LPS).62·67---69 This scenario is concordant with the clinical observation that the risk
`of GVHD increases with conditioning regimens that cause extensive injury to epithelial
`and endothelial surfaces62.66, and has been supported by elegant murine studies.63
`
`

`

`I 06 T. Toubai et al
`
`Host type APCs are critical for the induction, while donor type APCs exacerbate
`acute GVHD. In certain experimental models, donor type APC chimeras also induce
`76
`acute GVHD.68
`7
`•
`0--
`Amongst the cells with antigen-presenting capability, dendritic cells (DCsj are the
`most potent cells that play an important role in the induction of acute GVHD 7, a find(cid:173)
`ing that correlates well with the clinical observation that presence of host DCs after
`day 100 was associated with the severity of acute GVHD.74 Experimental data also sug(cid:173)
`gest that acute GVHD can be regulated by qualitatively or quantitatively modulating
`84 Absence of RelB signaling in host type DCs or enriching
`distinct DC subsets.7
`B--
`for host cos+ 'lymphoid' DC subsets early after HSCT substantially reduced acute
`GVHD. 17
`82 Other cells with antigen-presenting capabilities - such as monocytes/mac(cid:173)
`•
`rophages or semi-professional APCs - might also play a role in this phase. In accor(cid:173)
`dance with this concept, recent data suggest that host type B cells reduce acute
`GVHD in certain contexts.85 Host- or donor type non-hematopoietic stem cells,
`such as mesenchymal stem cells or stromal cells when acting as APCs, also reduce
`acute GVHD, although the mechanisms remain unclear.2· 17
`In contrast to the earliest phase of acute GVHD, little is known about the factors
`that initiate chronic GVHD. This is in large part due to the absence of animal models
`in which chronic GVHD occurs following induction, prophylaxis and/or treatment of
`acute GVHD. Given that acute GVHD is the biggest risk for the incidence of chronic
`GVHD, it is likely that damage and inflammation caused by acute GVHD and its
`treatment might play an important role in the onset of chronic GVHD. Chronic
`GVHD occurs later, perhaps after complete donor reconstitution, and therefore it
`stands to reason that indirect antigen presentation by donor APCs might be impor(cid:173)
`tant in its pathogenesis. This notion is supported by the observation that: (a) there is
`a clinical correlation between the presence of high numbers of donor mononuclear
`cells in PBSCT and a greater risk of chronic GVHD62
`86
`87
`; (b) experimental models
`•
`•
`demonstrating the 'autoimmune' nature of chronic GVHD also suggest a role for
`donor-derived APCs48
`; and (c) recent experimental data from an MHC-matched
`miHA-disparate model of CD4+-mediated chronic GVHD demonstrate that either
`donor or host APCs can initiate skin chronic GVHD, while only the donor APCs
`played a dominant role in intestinal chronic GVHD.88 Little is known about the spe(cid:173)
`cific subsets of APCs that induce or enhance chronic GVHD. Clinical data on the
`therapeutic benefit of rituximab89
`90 indicate that B cells might be pathogenic in
`•
`chronic GVHD. These observations, however, do not clarify whether B cells play
`a role in priming donor T cells (as APCs) or function as effectors due to dysregu(cid:173)
`lation of donor T-helper cells.
`Collectively, most of the experimental and clinical correlative data suggest that
`functional hematopoiesis-derived APCs are critical for the induction of both acute
`and chronic GVHD. A recent study, however, demonstrated that hematopoietic
`APCs are not obligatory for acute GVHD, at least for thymic epithelial cell (TEC)
`damage following allo-HCT.91
`
`NK, ya Tand NKT cells
`
`NK cells have recently been recognized to have modulatory effects on acute GVHD. In
`a parent into FI model of allo-HCT donor NK cells recognized the absence of donor
`class I on host APCs and eliminated them, resulting in reduction of GVHD.51 These
`observations are supported by some but not all clinical observations.92 To our
`
`

`

`GVHD pathophysiology: acute versus chronic
`
`I 07
`
`knowledge the role of NK cells has not been systematically studied in experimental
`models of chronic GVHD.
`Infusion of donor type yo T cells increased acute GVHD, while absence of host
`type yo T cells was associated with reduced APC activation and acute GVHD in an
`80 By contrast, in the absence of host yo T cells, GVHD
`MHC-mismatched model. 17
`•
`severicy was not altered in an MHC-matched, miHA-disparate model of chronic
`GVHD.93 One clinical study suggested that acute GVHD was greater in patients
`who received greater numbers of donor yo T cells.94 It is not known whether these
`distinct effects of yo T cells on acute and chronic GVHD reflect the differences in the
`immunobiology between the two or is merely a consequence of variation in the strain
`combinations/experimental models.
`Emerging data suggest that NKT cells - depending on their origin, donor or host -
`have seemingly opposite effects on the induction and severity of acute GVHD. 17
`95
`•
`These observations for acute GVHD were validated in a recent human studl which,
`interestingly, did not show any differences in the incidence of chronic GVHD. 6 There
`are nonetheless few data directly evaluating the role of donor or host NKT cells in the
`experimental models of chronic GVHD.
`
`T cells
`
`Donor T cells are an absolute requirement for the development of both acute and
`chronic GVHD.2
`15 Thus, not surprisingly, pharmacologic inhibition of T cells forms
`'
`the cornerstone of therapy for both acute and chronic GVHD. Several lines of evi(cid:173)
`dence demonstrate that not all subsets of T cells are cytopathic in GVHD, and that
`certain subsets might have salutary effects on the induction and severity of both acute
`and chronic GVHD (discussed below).2 This raises questions about the current pan-T(cid:173)
`cell suppression modalities of prophylaxis and treatment for GVHD, and makes a com(cid:173)
`pelling case for targeting T cell subsets to preserve an appropriate balance between
`cytopathic and regulatory T cell subsets.
`
`T cell co-stimulation
`
`The T cell receptor (TCR) of the donor T cells recognize HLA/miHA alloantigens or
`the non-polymorphic antigens on either host APCs (direct presentation) or donor
`98 T cell activation and differentiation requires both
`APCs (indirect presentation).97
`•
`TCR ligation and co-stimulation via a 'second' signal through interaction between
`the T cell co-stimulatory molecules and their ligands on APCs. 99 Accordingly, some
`experimental data have demonstrated that interruption of the second signal by block(cid:173)
`ade of various positive co-stimulatory molecules (CD28, ICOS, CD40, CD30, 4-1 BB
`and OX40) reduces GVHD, while anta~onism of the inhibitory signals (PD- I and
`100-- os Whether these can be directly translated
`CTLA-4) exacerbates acute GVHD.2
`•
`into the clinical context remains unclear, as does their applicability to chronic GVHD.
`For example, although antibody to OX40 ligand decreased experimental acute GVHD,
`OX40 expressing CD4+ cells were shown to be associated with onset of chronic
`107 A recent study demonstrated
`GVHD but not with acute GVHD in humans. 106
`•
`that CD80/86-dependent co-stimulation is critical for the induction of chronic
`GVHD.88 It also demonstrated that CD40 co-stimulation, in addition to CD80/86, is
`critical for causing intestinal chronic GVHD, but only CD80/86 is required for skin
`chronic GVHD.88 These data suggest that there are distinct T cell co-stimulatory re(cid:173)
`quirements in acute and chronic GVHD for various types of target organ damage.
`
`

`

`I 08 T. Toubai et al
`
`T cell subsets
`
`MHC class I (HLA-A, -B, -C) differences stimulate CDS+ T cells, and MHC class II
`I The individual roles
`(HLA-DR -DP, -DQ) differences stimulate CD4+ T cells. IOB-I I
`of CD4+ and cos+ T cells in both acute and chronic GVHD have been exhaustively
`studied in mouse models which collectively demonstrated that, depending on the type
`of mismatch between the strain combinations, both CD4+ and cos+ T cells can in(cid:173)
`duce GVHD.2· 17 On the other hand, clinical trials of CD4+ or cos+ depletion have
`.
`I
`.
`112
`een mconc us1ve.
`b
`Naive CD4+ T cells, upon encountering the antigen on APCs, differentiate into
`either Th I /Th2/Th 17 cells/or induced regulatory (Tr/Th3) cells, depending on the
`cytokine milieu. Th I cytokines (IFNy, IL-2 and TNFrx) have been implicated in the
`pathophysiology of acute GVHD. 113
`114 IL-2 production by donor T cells remains
`•
`the main target of many current clinical therapeutic and prophylactic approaches to
`both acute and chronic GVHD. Emerginj data indicate an important role for IL-2 in
`the generation and maintenance of CD4 CD2S+Foxp3+ Tregs (more below), suggest(cid:173)
`ing that prolonged interference with IL-2 may have an unintended consequence in the
`118
`prevention of the development of long-term tolerance after allogeneic HCT. 11
`5---
`Furthermore, experimental data show that early increase in Th I polarization of donor
`T cells can attenuate acute GVHD. This suggests that physiological and adequate
`amounts of Th I cytokines are critical for GVHD induction, while inadequate produc(cid:173)
`119
`tion (extremely low or high) could regulate acute GVHD. 114
`122 Likewise, several
`•
`-
`approaches that polarize donor T cells to Th2 reduced acute GVHD in some models,
`while some studies failed to show a beneficial effect of Th2 polarization on acute
`GVHD. 12
`131 By contrast, non-irradiated parent into Fl models of chcchcronic
`3-
`GVHD demonstrated that Th I polarization reduced while Th2 phenotype enhanced
`chronic GVHD. 13
`133 Thus the Thl/Th2 paradigm of donor T cells in the immunopa(cid:173)
`2,
`thogenesis of acute and chronic GVHD has evolved over the past years, and is com(cid:173)
`plex and incompletely understood. Moreover, all of these studies were performed
`prior to the identification of a critical role for Th 17 cells in the pathogenesis of several
`diseases that were previously considered to be Th I -dependent. To our knowledge, the
`role of Th 17 cells in acute and chronic GVHD remains as yet unexplored.
`Induced regulatory cells secrete either IL- IO (Tr cells} or TGF-f3 (Th3 cells). Their
`role in regulating experimental acute GVHD is unclear.1 4 By contrast, recent clinical
`data demonstrate a strong association between IL- IO polymorphisms and the severity
`of GVHD.57 TGF-f3, another suppressive cvtokine, was shown to suppress acute
`GVHD but to exacerbate chronic GVHD. 13s'
`Several studies have demonstrated a suppressive role for natural donor
`co4+co2s+ Foxp3+ regulatory T CTreg} cells - obtained from na'ive animals or gen(cid:173)
`erated ex vivo - in acute and chronic GVHD. 136 Donor co4+co25+ T cells sup(cid:173)
`pressed early expansion of alloreactive donor T cells and their capacity to induce
`acute GVHD without abrogating GVL. 137
`138 Donor CD4+co25+ T cells induced/
`•
`generated by immature or regulatory host or donor type APCs were also able to sup(cid:173)
`press acute GVHD.79 However, a clinical study in humans after matched sibling donor
`grafts found that, in contrast to the murine studies, donor grafts containinJ larger
`numbers of co4+co25+ T cells developed more severe acute GVHD. 13 These
`data suggest that co-expression of CD4+ and CD25+ is an insufficient marker for eval(cid:173)
`uating the numbers/impact of donor natural Tregs on GVHD outcomes. Other studies
`found that Foxp3 mRNA expression was significantly decreased in peripheral blood
`mononuclear cells from patients with acute GVHD. 140
`141 But Foxp3 expression in
`•
`
`

`

`GVHD pathophysiology: acute versus chronic
`
`I 09
`
`humans, unlike mice, may not be specific for T cells with a regulatory phenotype. 142
`Experimental data also showed that expression of CD62L and CD30 on donor T regs
`145
`might be required for optimal suppression of acute GVHD. 143
`-
`The role of natural T regs has been evaluated in chronic GVHD, where both host(cid:173)
`and donor type co4+co25+ T cells reduced the severity of disease. Association
`between reduced numbers of T-regulatory cells and deficient FOXP3 expression in
`141
`patients with chronic GVHD has been demonstrated both in peripheral blood 140
`•
`and in mucosal biopsies. 146 By contrast, another clinical study found that levels of
`Foxp3 mRNA in the CD25+ T cell compartment do not allow for predicting the de(cid:173)
`velopment of chronic GVHD, demonstrating that more than their mere presence or
`absence would need to be evaluated to understand the impact of Tregs on both acute
`and chronic GVHD. 147 Nonetheless, an intriguing possibility, given the negative impact
`of calcineurin inhibitors on T regs• is that chronic GVHD results - or is exacerbated - as
`a consequence of suppression of both the alloreactive donor cytopathic and the reg(cid:173)
`ulatory T cells.
`Emerging experimental data from many groups have found that the alloreactive
`na·ive (CD62L +Ji T cells, but not the memory (CD62L -) T cells, are associated with
`acute GVHD. 1 s- 15 1 Furthermore, expression of naive T cell marker CD62L was
`also found to be critical for regulation of GVHD by donor natural regulatory T
`144 It remains as yet unknown whether the reduced GVHD potential of mem(cid:173)
`cells. 143
`•
`ory-type T cells from a naive murine donor, in contrast to their ability to cause greater
`
`solid-organ allo-rejection 152, is due to a consequence of the intense conditioning reg(cid:173)
`imen and/or altered trafficking, or from a restricted repertoire and/or from a T cell
`intrinsic defect. Emerging data show that, in contrast to naive T cells that cause acute
`GVHD, alloantigen-primed memory T cells induce chronic colitis and liver injury that is
`more consistent with chronic GVHD. 153 Furthermore, a subset of post-mitotic
`CD441°CD62LhiCDS+ alloreactive memory T cells that develop during acute GVHD
`were shown to self-renew and induce GVHD with features resembling chronic
`GVHD upon transfer into secondary recipients. 154 However, these features were
`not well described. Moreover, the same group suggested that host thymic-derived do(cid:173)
`nor CD4+ T cells also caused GVHD with clinical features that are consistent with
`chronic GVHD. A few clinical observations have evaluated the correlation between ex(cid:173)
`pression of CD62L and CD44hiCCR7+ on donor CDS and CD4+ memory cells in
`patients with chronic GVHD, suggesting a role for memory T cells in acute and chronic
`GVHD. 155 Prospective clinical studies with these various cellular subsets are under
`way, and results might alter the future forms of allogeneic HCT.
`
`Host thymus-derived T cells
`
`Elegant experimental data have shown that central (thymic) deletion by establishment
`of a stable mixed hematopoietic chimeric state is an effective way to eliminate con(cid:173)
`tinued thymic !roduction of alloreactive CD4+ and CDS+ T cells and thus reduce
`acute GVHD. 1 6 The role of thymic deletion in the context of stable mixed chime(cid:173)
`rism for the development of chronic GVHD remains unexplored. However, as
`alluded to earlier, it has been postulated that the thymic damage by conditioning,
`acute GVHD, and age-related atrophy disrupt thymic education of T cells and cause
`chronic GVHD by the emergence of thymic-dependent autoreactive T cells. Consis(cid:173)
`tent with this notion, it was recently demonstrated that thymic-dependent donor
`bone-marrow-derived T cells that escape from negative selection have the ability
`to cause lethal chronic GVHD that is similar to human chronic GVHD, which was
`
`

`

`I IO T. Toubai et al
`
`prevented by thymectomy of the recipients prior to BMT.48 By contrast, data from
`another, sublethally irradiated, experimental model demonstrated that host th{mus
`158
`was not required for the development of 'autoimmune-like' chronic GVHD. 1 7
`•
`Taken together, existing experimental data suggest that mature donor T cells in
`the allograft are necessary and sufficient for acute GVHD, while both mature donor
`T cells and host thymic-dependent generation of donor T cells from the infused
`hematopoietic stem cells might play a role in the induction of chronic GVHD.
`
`B cells
`
`Recent data have brought into focus the role of B cells and humoral immunity in
`GVHD. Host type B cells might attenuate acute GVHD in an IL- I 0-dependent manner
`in certain contexts.85 By contrast, a potential role for B cells in chronic GVHD has
`long been speculated upon since the earliest reports of human chronic GVHD. 159 In
`addition, recent clinical data have provided a strong rationale for a pathogenic role
`of donor B cells in chronic GVHD. Specifically, the identification of a strong correlation
`between chronic GVHD and (a) the presence of antibodies to Y-chromosome-en(cid:173)
`coded histocom~atibility antigens50
`; (b) increased numbers of B cells with altered
`TLR9 responses 60
`; (c) the levels of B-cell-activating factor (BAFF), which promotes
`survival and differentiation ofactivated B cells 161
`; and (d) in some animal models, levels
`158 Moreover, emerging data showing clinical responses from de(cid:173)
`of autoantibodies.48
`•
`pletion of B cells with rituximab further strengthens the notion of a pathogenic role
`for B cells in chronic GVHD.90 Nonetheless, whether B cells are the effectors and/
`or inducers or amplifiers of chronic GVHD is not known.
`
`SUMMARY
`
`Donor T cells are critical for the development of both acute and chronic GVHD.
`There might however be differences in the prerequisites, target antigens, cytokine
`and cellular effectors that cause acute and chronic GVHD. Development of appropri(cid:173)
`ate animal models complemented by well-designed clinical correlative studies are ur(cid:173)
`gently required to better understand the apparently overlapping yet distinct
`pathophysiology of acute and chronic GVHD. On the therapeutic front, given the cur(cid:173)
`rent need, and based on understanding of the biology, albeit incomplete, it might be
`reasonable to meld strategies that are both complementary and distinct for prevention
`and treatment of acute and chronic GVHD.
`
`REFERENCES
`
`I. Appelbaum FR. Haematopoietic ce