Shaping Graft Immunity

Shaping graft immunity:

Prevention of relapse, viral reactivations and graft-versus-host disease

after allogeneic stem cell transplantation

Suzanne van Dorp

Shaping graft immunity:

Prevention of relapse, viral reactivation and graft-versus-host diseaseafter allogeneic stem cell transplantation

Vormgeving van het transplantaat ter voorkoming van relapse, viralereactivaties en graft-versus-host ziekte na allogene stamceltransplantatie

(met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag

van de rector magnificus, prof. dr. G. J. van der Zwaan, ingevolge het besluit

van het college voor promoties in het openbaar te verdedigen op

dinsdag 4 juni 2013 des ochtends te 10.30 uur

door

Suzanne van Dorp

geboren op 8 juli 1982 te Haarlem

Promotor: Prof. dr. H.M. LokhorstCo-promotoren: Dr. J.H.E. Kuball

Dr. E. Meijer

iv

Voor Henk van Dorp

v

Contents

1 Introduction and scope of the thesis 1

I Therapeutic potential of an allogeneic stem

cell transplantation

2 Single-centre experience with nonmyeloablative allogeneic stem celltransplantation in patients with multiple myeloma: prolonged re-missions induced 11

II Graft-versus-Host Disease and mechanisms

of disease

3 Rituximab treatment before reduced intensity conditioning trans-plantation associates with a decreased incidence of extensive chronicGVHD 25

4 The immunological phenotype of rituximab-sensitive chronic graft-versus-host disease: a phase II study 41

III Towards new effectors with the potential to

control viral infections and leukemia

5 γ9 and δ2CDR3 domains regulate functional avidity of T-cells har-boring γ9δ2T-cell receptors 57

6 γδT-cells elicited by CMV-reactivation after allo-SCT cross-recognize CMV and leukemia 85

7 General discussion 119

Bibliography 129

vi

8 Nederlandse samenvatting 155

Curriculum Vitæ 165

List of publications 167

Dankwoord 169

vii

viii

Chapter 1

Introduction and scope of the thesis

1.1 Therapeutic potential of an allogeneic stem cell

transplantation

Allogeneic stem cell transplantation is used as curative treatment for a wide

variety of hematological malignancies, such as multiple myeloma [5, 13] and

acute myeloid leukemia [66, 185]. The treatment effect is partially due to the

conditioning regimen given before transplantation, but mainly to the graft-

versus-tumor effect that is seen after infusion and repopulation of the new im-

mune system. Graft-versus-tumor (GvT) effect is induced by competent donor

lymphocytes and is clearly illustrated by sustained remission after infusion of

donor lymphocytes after allo-SCT [120, 138, 228]. Donor T cells are key players

in the curative GvT effect. In an HLA-matched setting donor T-cells can re-

act against polymorphic peptides, minor histocompatibility antigens (mHags),

presented by HLA-molecules to donor T cells. mHags can be expressed on all

cells, but can also be specific to hematopoietic cells and tumor cells [99, 206]. T

cells directed against mHags, specifically presented on tumor cells, can induce

GvT effect.

Despite the GvT effect, relapse is not uncommon. 10 − 45% of patients will

ultimately develop relapse of disease [5, 31, 43, 171]. This is the main cause

of treatment failure. Furthermore non-relapse mortality (NRM), mainly caused

by graft-versus-host disease (GVHD) and infections, such as viral reactivations,

gravely impairs the success rate of allo-SCT. After myeloablative and nonmye-

loablative allo-SCT NRM of respectively 32% and 20% was reported [59, 204].

Thus, there is still a substantial need to improve allo-SCT. This can be achieved

1

either by modulating the transplantation in order to minimize the risk of GVHD.

However, this strategy is usually hampered by an increase in relapse [81, 91].

Alternatively, achieving a better control once GVHD occurs represents a valu-

able option. Finally, understanding the rather complex molecular mechanisms

for virus and tumor control after allo-SCT could allow the development of new

treatment strategies with improved efficacy in terms of tumor control and lim-

ited toxicity. Therefore, we will briefly introduce so far know mechanisms of

GVHD (Section 1.2) and introduce potentially valuable new effector cells in or-

der to control leukemia as well as viral infections after allo-SCT (Section 1.3).

1.2 Graft-versus-Host Disease as major challenge

after allo-SCT

The presence of mHags on normal cells can also induce a severe and possibly

lethal complication of allo-SCT, GVHD [57]. Acute GVHD is seen mostly in the

first three months after allo-SCT and is characterized by dermatitis, hepatitis

and enteritis. Acute GVHD is seen in up to 60% of patients and morbidity and

mortality are high [80]. T-cell depletion of the graft is associated with lower

incidence of acute GVHD, however also impairs the GvT effect and relapse of

disease is more often seen in patients treated with a T-cell depleted graft [91,

148]. This again emphasizes the importance of donor T cells in the pathogenesis

of acute GVHD. Acute GVHD is treated with immunosuppressants, such as

high dose prednisone, ciclosporin-A and mycophenolate mofetil (MMF) [80].

Promising results are also seen with mesenchymal stromal cells (MSC’s), which

immunomodulatory effect could taper the allo-reactivity of donor T cells [112,

129, 215].

Chronic GVHD is the major long-term complication of allo-SCT. Up to 70% of

patients surviving 100 days post-SCT will develop chronic GVHD [14]. This

disease is invalidating and decreases quality of life enormously. Moreover it is

the major cause of late mortality after allo-SCT [131, 151]. Chronic GVHD can

affect various organs and has a heterogenous phenotype [70]. In the majority of

patients the skin is involved, either locally or generally [71]. Chronic GVHD can

cause sclerosis of the skin, with similarities with systemic sclerosis (SSc). This

2

CHAPTER 1. INTRODUCTION AND SCOPE OF THE THESIS

can influence movability and chest wall compliance. Moreover chronic GVHD

causes ulcerations of the skin, which are, besides painful and dysmorphic, a

port d’ entrée for pathogens [201]. Other often affected sites are mucosal sur-

faces of eyes, mouth, vagina and gastro-intestinal tract. Furthermore chronic

GVHD can cause bronchiolitis obliterans syndrome, hepatitis and nephritis [71].

The general finding in affected organs is fibrosis. The pathogenesis of organ

damage and fibrosis in chronic GVHD is poorly understood. In both mouse

and human studies fibrosis was not only restricted to the affected sites, but was

generally seen in various organs, even in the spleen [van Dorp, unpublished

data] [100]. Activation of fibroblasts, causing fibrosis, can be mediated by sev-

eral cytokines and growth factors, such as IL-6 [15], TGFβ [104, 127] and PDGF

[1]. A role for T cells in this process is undoubted. Epithelial damage and T-

cell infiltrates at the affected sites are seen in both acute and chronic GVHD.

These infiltrates are, when present, however less fulminant in chronic GVHD

[35]. The addition of anti-thymocyte globuline (ATG) to the conditioning regi-

men, causing an in vivo depletion of both host and donor T cells, lowers the

incidence of chronic GVHD, however not in the same extent as seen in acute

GVHD [114, 122]. Furthermore the differences in phenotype between acute

and chronic GVHD suggest a different pathogenesis. The similarities in phen-

otype with auto-immune diseases like SSc, suggest a role for auto-antibodies

and therefore B cells in development of the fibrosis seen in chronic GVHD. The

findings of plasma cell infiltrates and depositions of immunoglobulins in tissues

affected by chronic GVHD also contribute to this hypothesis [36]. Most im-

portantly however, promising results are seen in patients with chronic GVHD,

refractory to steroids, when treated with B-cell depleting therapy [36, 51, 179].

However, so far limited data on efficacy and potential mechanisms of action of

B-cell depleting therapies have been reported in men. Therefore, we investig-

ated in more detail the potential of B-cell depleting therapies in preventing or

controlling chronic GVHD.

3

1.3 Towards new effectors with the potential to con-

trol viral infections and leukemia

Viral infections and relapse of the underlying disease are the other two chal-

lenges after allo-SCT. Control of leukemia has been reported to be mainly me-

diated by NK cells [174] and αβT cells [150]. NK cells are known to recognize

tumor cells due to loss of inhibitory signals, caused by downregulation of MHC

molecules on tumor cells. Furthermore several activating NK receptors, such as

NKG2D and NKp30, contribute to this recognition by binding ligands upregu-

lated on malignantly transformed cells, i.e. the MHC-like molecules MICA and

MICB and ULPs [141]. αβT cells can recognize tumor specific antigens presen-

ted by MHC molecules through their-T cell receptor (TCR) [19, 160]. In patients

with leukemia the control mechanisms are not sufficient enough to clear tumor

cells. To overcome this problem and to optimize the anti-tumor response in

vivo, it has been suggested to generate substantial amounts of tumor specific T

cells by adoptive transfer of engineered T cells bearing a tumor specific αβTCR

[106, 150, 218]. Although promising, this technique has several limitations.

First, a distinct αβTCR is restricted to an HLA type, therefore the amount of

patients that can be treated with one TCR is limited. Secondly, redirecting αβT

cells, with a high affinity αβTCR against tumor specific antigens, can induce

‘on-target’ auto-reactivity, since these specific antigens are often also expressed

on normal tissues, although to a lesser extent [107, 162]. However, even when

an αβTCR directed against antigens uniquely expressed on tumor cells is used,

autoreactivity can occur. Introduction of a specific or engineered αβTCR can

induce pairing with the endogenous αβTCR. This mispairing induces αβT cells

with unknown specificity. When these unwanted TCRs gain specificity towards

normal tissues, this can lead to ‘off-target’ auto-reactivity [20, 67]. The major

challenge in adoptive TCR transfer is to select TCRs that can be broadly used,

with high tumor specificity and affinity, while on- and off-target auto-reactivity

are prevented.

An attractive option to meet all these conditions at once is the use of a γδTCR.

γδT cells are a minor population in the peripheral blood and more widespread

in epithelial tissues, such as the skin, lung, gastro-intestinal and reproductive

4


tract [89]. γδT cells express a heterodimer composed of rearranged γ and δ

chains, comparable to the αβTCR. γδT cells rearrange and express clonally di-

verse TCRs, however the variability of the variable (V) and joining (J) elements

in the TCRγ and TCRδ loci is more limited in number than of the TCRα and

TCRβ. In humans there are at this moment 6 known Vγ genes and 5 Vδ genes

[12, 89, 95]. The diversity of the γδTCR is enhanced by a special feature of

the TCRδ locus. The TCRδ locus contains, like the TCRβ locus, diversity (D)

elements. TCRδ chains frequently use both of their D elements to create more

diverse complementarity determining regions (CDR3). The TCRδ locus is loc-

ated within the TCRα locus and the TCRγ locus is located in the proximity of

the TCRβ locus. Despite the proximity of their locations rearranging of genes of

the TCRα and TCRδ, or TCRβ and TCRγ, is very rare. Transfer of a γδTCR into

αβT cells could prevent mispairing between the endogenous and the introduced

TCR and thereby formation of TCRs with unwanted specificity. Using the αβT

cells as carrier of the γδTCR also solves the problem that function and prolif-

eration capacity of γδT cells can be impaired in cancer patients [234], whereas

αβT cells can, even in end stage cancer, induce potent immune responses [126].

γδTCRs, like the αβTCR, use the signalling domain of the CD3 molecule after

activation [161]. However they recognize antigens in a major histocompatibility

complex (MHC) independent matter [89, 110]. It is thought that not one, but

a various number of ligands causes activation of the γδTCR. They specifically

recognize both self- and non-self ligands, upregulated on stressed cells. Self-

ligands, upregulated on infected or malignantly transformed cells and activat-

ing γδT cells, are endogenous low molecular mass phosphoantigens, which are

metabolites of the mevalonate pathway, such as isoprenylpyrophosphate (IPP)

[34, 109, 214], and the mitochondrial F1ATPase [198]. Furthermore, γδT cells

recognize known ligands of NK-receptors, upregulated on infected or trans-

formed cells, such as the MHC-like molecules MIC-A and MIC-B [87, 238], and

ULBP4 [119]. This recognition is mediated by both the γδTCR as NKG2D, which

is also expressed by most γδT cells [110, 184]. Moreover CD1c, a MHC-like mo-

lecule, presenting self lipids, is upregulated on inflammation induced monocyte

differentiation into dendritic cells (DCs) is shown to activate γδT cells [190]. A

broad spectrum of bacterial species and parasites are described to induce prolif-

eration and cytokine production by γδT cells. They are recognized by non-self

5

ligands, mainly exogenous phospho-antigens, like HMBPP [17, 42]. These are

products of the equivilant of the mevalonate pathway, as used by different mi-

crobes [109].

In this context, we investigated in this thesis the role of γδT cells and defined

γδTCRs to control leukemia. However, as γδT cells are also activated by cyto-

megalovirus (CMV) and other herpesvirus antigens we investigated also the

role of γδT cells and defined γδTCRs after allo-SCT and CMV-reactivation and

focused on mainly Vδ1pos γδT cells as CMV infection induces proliferation of

mainly Vδ1pos γδT cells. Healthy individuals, who are CMV seropositive, have

higher percentages of circulating Vδ1pos γδT cells, with an effector memory

phenotype [176]. This suggests a role for Vδ1pos γδT cells in controlling CMV

virus after primary infection. Proliferation of Vδ1pos γδT cells has also been

described upon reactivation of CMV disease after kidney transplantation [55].

After allo-SCT reactivation of CMV, contributes to the non-relapse mortality

and morbidity, causing pneumonia, retinitis and most frequently colitis [28]. In

immunocompetent individuals there is a tight balance between the virally in-

fected cells and the host immune system, mainly virus specific T cells. This fine

balance is disturbed in patients with a compromised immune system as seen

after intensive chemotherapy and allo-SCT. The virus is no longer controlled by

the immune system and can reactivate from its latency [28]. Interestingly, there

is recent evidence that virus specific αβT cells in the graft can protect against

severe CMV reactivations [175]. It is hypothesized that, since the graft of a

seronegative donor does not contain virus specific αβT-cells, a competent im-

mune response against CMV will take even longer than in patients treated with

the graft of a seropositive donor. However in many cases immune responses

early after allo-SCT are not yet adequate enough to control the virus, even after

tapering of immunosuppressants. Therefore, we asked as to whether individual

γδT-cell lines, clones or even individual receptors might not only provide tools

to improve control of leukemia after allo-SCT but also to further prevent viral

reactivations.

In the context of above described major challenges during an allo-SCT we

defined the scope of this thesis: In Part I the potential of allo-SCT in multiple

myeloma patients has been addressed. In Part II (Chapters 3 and 4) the effect

6


of both host and donor B-cell depletion on the development of chronic GVHD

has been evaluated. In Part III (Chapters 5 and 6) the potentially therapeutic

efficacy of γδT cells for the control of leukemic and viral infections has been in-

vestigated. This includes the impact in the context of an allo-SCT as well as the

development of next generation immune therapies for the clearance of leukemia

and infections.

Part I. Therapeutic potential of an allogeneic stem cell transplantation

Chapter 2. Single-centre experience with nonmyeloablative allogeneic

stem cell transplantation in patients with multiple myeloma: pro-

longed remissions induced.

Part II. Graft-versus-Host Disease and mechanisms of disease

Chapter 3. Rituximab treatment before reduced-intensity conditioning

transplantation associates with a decreased incidence of extensive

chronic GVHD.

Chapter 4. The immunological phenotype of rituximab-sensitive chronic

graft-versus-host disease: a phase II study.

Part III. Towards new effectors with the potential to control viral infections and

leukemia

Chapter 5. γ9 and δ2CDR3 domains regulate functional avidity of T-cells

harboring γ9δ2T-cell receptors.

Chapter 6. γδT cells elicited by CMV-reactivation after allo-SCT cross-

recognize CMV and leukemia

7

8

Part I

Therapeutic potential of an

allogeneic stem cell

transplantation

9

Chapter 2

Single-centre experience withnonmyeloablative allogeneic stemcell transplantation in patients withmultiple myeloma: prolongedremissions induced

Suzanne van Dorp1, Ellen Meijer

1, Niels W.C.J. van de Donk1,

Adriaan W. Dekker1, Karel Nieuwenhuis

1, Monique C. Minnema1,

Eefke Petersen1, Roger Schutgens

1, Leo F. Verdonck1

and

Henk M. Lokhorst1

Netherlands Journal of Medicine, 2007

1Department of Haematology, University Medical Centre Utrecht, Utrecht, the Netherlands

11

Abstract

Background: The role of allogeneic stem cell transplantation in multiplemyeloma is not yet established.Methods: We retrospectively evaluated the outcome of nonmyeloablativeallogeneic stem cell transplantation (NMA) in patients with multiple myel-oma treated at the department of Haematology of the University MedicalCenter Utrecht. Thirty-six patients received NMA as part of the first-linetreatment; 23 patients as part of salvage therapy. Conditioning regimen waslow-dose total body irradiation (TBI, 2 Grays) only; fludarabine was addedin patients without previous autologous stem cell transplantation and pa-tients with matched unrelated donors received antithymocyte globulin inaddition to fludarabine and TBI.Results: Following NMA overall response increased from 84 to 90%, com-plete remission rate from 15 to 32%. As part of first-line treatment NMA in-duced complete remission in 50% of patients versus one patient (4%) treatedfor relapsed multiple myeloma. Median progression-free survival was 26months (13 months for the salvage group, 38 months for the ‘upfront’ pa-tients). Median overall survival has not been reached yet. The achievementof complete remission following NMA as part of first-line treatment wasassociated with prolonged progression-free and overall survival. Major tox-icities were acute and chronic graft-versus-host disease occurring in 64%(23% grade 3-4) and in 54% (49% extensive) patients, respectively. Sevenpatients (12%) died from nonrelapse mortality (NRM), five patients (9%)directly related to toxicity of NMA.Conclusion: NMA in multiple myeloma is feasible, is associated with ac-ceptable NRM and may induce prolonged complete remission. In pretreatedpatients the result of NMA is disappointing which urges new strategies.

2.1 Introduction

Allogeneic stem cell transplantation (ASCT) is probably the only treatment with

a curative potential for multiple myeloma. This is due to the graft-versus-

myeloma effect, mediated by immune competent donor lymphocytes, best il-

lustrated by the induction of sustained remissions following donor lymphocyte

infusions after ASCT [5, 13, 181]. However, the necessity of performing ASCT

in multiple myeloma is disputed as no survival advantage has been obtained

compared with autologous SCT, in particular when myeloablative conditioning

12

CHAPTER 2. NONMYELOABLATIVE ASCT IN MULTIPLE MYELOMA

for the ASCT is applied [24]. An important factor for this is the high nonrelapse

mortality associated with myeloablative conditioning [24, 139].

In an attempt to lower nonrelapse mortality and make ASCT available to more

patients, nonmyeloablative conditioning was introduced. Nonmyeloablative

ASCT (NMA) is associated with reduced acute toxicity, while antitumor activity

is probably maintained [6, 76, 147, 204]. In this retrospective single center study

we show that NMA in multiple MM is feasible, with acceptable NRM and that

prolonged remissions may be induced in patients who received NMA as part of

first line treatment and achieved a complete response following SCT.

2.2 Patients and methods

2.2.1 Selection of patients. Patients with MM, who received a NMA at the

University Medical Center Utrecht, The Netherlands, between September 2001

and September 2005 were included in this retrospective study. In this period

in all newly diagnosed patients younger than 66 years and their siblings the

human leucocyte antigens (HLA) class I (HLA-A, HLA-B, HLA-C) and class II

(HLA-DR, HLA-DP, HLA-DQ) were typed in the first three months after dia-

gnosis. If a HLA matched sibling donor was available (1 factor class I or class

II mismatch was allowed), patients could proceed to NMA between 2 and 6

months after high dose melphalan (HDM) 200 mg/m2 and autologous stem

cell rescue, that followed 3 courses of induction therapy with Vincristine, Ad-

riamycin, Dexamethasone (VAD) or Thalidomide, Adriamycin, Dexamethasone

(TAD) [77]. Also patients with a relapse after preceding treatment, but respons-

ive to salvage therapy, and having an HLA-matched related or unrelated donor

were also allowed for subsequent NMA.

2.2.2 Conditioning. The conditioning regimen before allogeneic stem infusion

for the patients with completely matched HLA identical siblings donors con-

sisted of 1 course of low dose Total Body Irradiation (TBI) (2 Gy) only, in case

they had received HDM 200 mg/m2 within the preceding 2 and 6 months (tan-

dem auto-NMA). Fludarabine 30 mg/m2 intravenously for 3 days was added in

case no preceding autologous SCT was performed. The conditioning regimen

13

before allogeneic stem infusion for the patients with an HLA mismatched or

unrelated donor consisted of anti-thymocyte globulin (ATG; 2 mg/kg/day for 4

days) followed by fludarabine 30 mg/m2 intravenously for 3 days and 1 course

of low dose TBI (2 Gy).

2.2.3 Immune suppression. In the post-transplantation period all patients were

treated with the immunosuppressive drugs cyclosporine A (CSP) and my-

cophenolate mofetil (MMF). Patients received 30 mg/kg/day MMF for 60− 90

days and 2 × 4.5 mg/kg/day CSP for 3 − 6 months according to the Seattle

scheme [147].

2.2.4 GVHD grading and treatment. For diagnosing and grading acute Graft

versus Host Disease (GVHD) the Gluckberg criteria [77] were used. Chronic

GVHD was graded according to the Seattle classification [202]. Time of onset of

acute and chronic GVHD and grade of GVHD were monitored. Acute GVHD >

grade I was treated with prednisone 1− 2 mg/kg/day and when necessary top-

ical prednisone treatment was applied. In these cases the doses of CSP and/or

MMF were increased or continued. In case of steroid refractory acute GVHD

other drugs were used, such as sirolimus, tacrolimus, rituximab or more exper-

imental drugs, such as alemtuzumab and dacluzimab. Chronic GVHD of the

skin was treated with topical prednisone. In severe cases of extensive chronic

GVHD prednisone 1 mg/kg/day was given.

2.2.5 Definitions. Response and progression were determined according to the

European Group for Blood and Marrow Transplantation (EBMT) criteria [25].

In short, a partial response (PR) was defined as ≥ 50 reduction of serum M-

proteine or ≥ 90% reduction in 24 hour excretion of Bence Jones proteinuria

in case of light chain disease (LCD). A complete response (CR) was defined as

complete disappearance of serum and urine M-proteine as determined by im-

mune fixation of serum and 10 times concentrated urine. In addition monclonal

myeloma cells as determined by immune phenotyping had to be absent in a

representative bone marrow aspirate or biopsy. Non relapse mortality (NRM)

was defined as any death not related to progressive or relapsed myeloma.

14


2.2.6 Statistical analysis. For the statistical analysis SPSS 12.0.1 for Windows

(SPSS Inc., IL, and USA) was used. Overall survival (OS) was measured in

months and defined as the time from the date of transplantation until date

of death or last follow-up. Progression-free (PFS) survival was measured in

months and defined as the time from the date of transplantation until the date

of progression or death from any cause or last follow-up. Time to acute or

chronic GVHD was calculated from the date of transplantation until occurrence

of acute or chronic GVHD. Probabilities of overall survival, progression-free

survival, and NRM were calculated using the Kaplan-Meier method. Kaplan-

Meier curves were generated to illustrate survival and the log-rank test was

used to compare survival curves between sub-groups. Univariate Cox regres-

sion analysis was used to determine the prognostic value of various variables

for overall survival and progression-free survival. The predictive value of acute

and chronic GVHD for OS and PFS was calculated using a time-dependent uni-

variate Cox regression analysis.

2.3 Results

2.3.1 Patient characteristics. Fifty-nine patients were included in this study.

The median age was 55 years (range 35 to 67). There were 42 males (71%) and

17 females (28%). The median follow-up duration of survivors was 25.2 months

(range 6.8 to 54.6) (Table 2.1). In 36 patients (61%), NMA was part of first-line

treatment and in 23 patients (39%) it was part of salvage treatment. At the time

of transplant, nine patients (15%) were in complete remission and 40 patients

(68%) were in partial remission.

Forty-four patients (74%) had a matched related donor, four patients (7%) had a

partially matched related donor and six patients (10%) had a matched unrelated

donor, and five patients (9%) had a partially matched unrelated donor. In 16

cases (27%) there was a female donor and a male recipient. Thirty-five patients

(59%) were conditioned with TBI only (2 Gy) and 24 (41%) with TBI and fluda-

rabine (30 mg/m2/day for 3 days) [6]. Fifteen patients (25%) received ATG as

in vivo T-cell depletion. At the time of diagnosis 21 out of 50 patients (42%) had

chromosome 13 abnormalities in FISH analysis and 20 out of 44 patients (46%)

15

had an elevated β2-microglobulin (≥ 3.0 mg/l).

No. of patients (%)

Sex

Male 42 (71.2)

Female 17 (28.8)

Age (years)

Median 55

Range 35− 67

Median follow up1 (months)

Median 25.2

Range 6.8− 54.6

Extent of prior therapy

First-line treatment 36 (61.0)

Relapse treatment 23 (39.0)

Donor

MRD 44 (74.6)

PMRD 4 (6.8)

MUD 6 (10.2)

PMUD 5 (8.5)

Conditioning regimen

TBI 35 (59.3)

TBI and fludarabine 24 (40.7)

Donor sex match

Female to male 16 (27.1)

Other 43 (72.9)

Deletion of chromosome 132

Presence of deletion of chromosome 13 21 (42.0)

Absence of deletion of chromosome 13 29 (58.0)

β2-microglobulin3

< 3 mg/L 24 (54.5)

> 3 mg/L 20 (45.5)

Status at the time of Allo-SCT

CR 9 (15.3)

No CR 50 (84.7)

Table 2.1: Patient characteristics (N = 59). Allo-SCT indicates allogeneic stem celltransplantation; CR, complete response; MRD, matched related donor; MUD matchedunrelated donor; PMRD, partial matched related donor; PMUD, partial matched unre-lated donor; TBI, total body irradiation. 1Follow-up duration of survivors. 2Determinedin 50 patients (84.7%). 3Determined in 44 patients (74.6%).

16


Total No. ofpatients (%)

First-line treatmentn (%)

Relapse treatmentn (%)

p

Remission state beforeallo-SCT

0.007

CR 9 (15.3) 9 (25) 0 (0)

PR 40 (67.8) 24 (66.7) 16 (69.6)

NR 10 (16.9) 3 (8.3) 7 (30.4)

Remission state afterallo-SCT

< 0.001

CR 19 (32.2) 18 (50) 1 (4.3)

PR 35 (59.3) 17 (47.2) 18 (78.3)

NR 5 (8.5) 1 (2.8) 4 (17.4)

Table 2.2: Response rates to non-myeloablative allo-SCT in first-line and relapse treat-ment. Allo-SCT, allogeneic stem cell transplantation; CR, complete response; PR, partialresponse; NR, no response; Differences in categorical variables were determined withthe Pearson χ2-test.

2.3.2 Response and survival. Total response rate following NMA increased

from 83% (n = 49) to 92% (n = 54); complete response rate increased from

15 to 32%. NMA as part of first-line treatment induced a complete remission

in 50% of patients, as compared with achievement of a complete remission in

one patient (4%) treated for relapsed multiple myeloma. An ongoing response,

defined as improvement of partial to complete response and from no response

to partial or complete response occurred in 24% of patients; 28% in patients who

received NMA as part of first-line treatment and 17% in patients who received

NMA as part of relapse treatment (Table 2.2).

Twenty-five patients (42%) relapsed or progressed after NMA, two from com-

plete remission and 23 from partial remission. At the time of analysis 48 patients

were alive. Eleven patients (19%) had died, four from progressive disease and

seven (12%) from nonrelapse mortality. The estimated overall survival of the

whole group of patients at two years was 84% (Figure 2.1). Median progression-

free survival was 23.5 months (range 1.0 to 38.0 months; Figure 2.2). In patients

who received NMA as part of first-line therapy, overall and progression-free sur-

vival at two years were 88.9 and 68.9%, respectively (Figures 2.3A and B). The

achievement of complete remission after NMA in this group of patients was

associated with superior overall survival and progression-free survival (Fig-

ures 2.4A and B). Also the presence of complete remission before NMA was

17

Figure 2.1: Overall survival follow-ing nonmyeloablative allogenic stem celltransplantation.

Figure 2.2: Progression-free survival fol-lowing nonmyeloablative allogenic stemcell transplantation.

associated with prolonged progression-free survival (p = 0.037), but not with

prolonged overall survival (p = 0.234). The occurrence of chronic GVHD was

associated with prolonged overall survival (p = 0.012) but not with progression-

free survival (p = 0.3). The 11 patients with acute GVHD grade III and IV had

inferior overall survival due to fatal outcome of this complication in five pa-

tients (p = 0.001). No fatal deaths were observed in the patients with acute

GVHD grade 0 to II. None of all other factors tested including age, gender of

recipient or donor, conditioning regimen, use of ATG, family or a matched un-

related donor, deletion of chromosome 13 (FISH), β2 microglobulin ≥ 3 mg/ml,

had an impact on overall or progression-free survival. In the patients who re-

ceived NMA as part of the treatment for relapsed myeloma overall survival and

progression-free survival at two years were 77.5 and 23.9%, respectively (Fig-

ure 2.5). None of the factors tested including age, gender as described above

had an impact on progression-free and overall survival. It should be mentioned,

however, that the statistical analysis must be interpreted with caution due to the

small number of patients.

2.3.3 Toxicity. Nonrelapse mortality at 12 months was 12% (Figure 2.6). Five

patients (9%) died from acute GVHD grade III to IV. One patient died from

18


Figure 2.3: (A) Overall survival in patients who received NMA as first-line treatment.(B) Progression-free survival in patients who received NMA as first-line treatment.

Figure 2.4: (A) Overall survival in patients who recieved NMA as first-line treatmentand did reach complete remission afterwards and patients who did not reach completeremission. (B) Progression-free survival in patients who recieved NMA as first-line treat-ment and did reach complete remission afterwards.

19

Figure 2.5: (A) Overall survival in patients who recieved NMA as treatment for relapsedmyeloma. (B) Progression-free survival in patients who recieved NMA as treatment forrelapsed myeloma.

complications occurring after heart catheterisation and one relapsed patient re-

fused further treatment, including a stem cell boost for secondary aplasia and

ultimately died from overwhelming septicaemia. Acute GVHD following NMA

occurred in 38 patients (64%): grade I in 12 (20%), grade II in 12 (20%), and

grade III or IV in 14 patients (24%). Chronic GVHD following NMA occurred

in 32 patients (54%), with three patients (5%) experiencing limited disease and

29 patients (49%) extensive disease. NMA as first-line treatment was associated

with a higher incidence of grades II to IV acute GVHD, when compared with

NMA as relapse treatment (56 versus 26%; p = 0.034). The use of ATG sig-

nificantly reduced the incidence of chronic GVHD (20 versus 66%; p = 0.003).

This may explain the lower incidence of chronic GVHD in patients with an un-

related or mismatched donor. All other factors tested were not associated with

occurrence of acute or chronic GVHD.

2.4 Discussion

Several conclusions can be drawn from this retrospective study. The first one is

that NMA is feasible in multiple myeloma, even in heavily pretreated patients.

Nonrelapse mortality after NMA compares very favourably with nonrelapse

mortality after myeloablative ASCT [24, 139]. What is remarkable is the absence

20


Figure 2.6: Nonrelapse mortality.

of nonrelapse mortality in the patients receiving a transplant from a matched

unrelated donor, probably due to the administration of ATG. The second ob-

servation is that NMA as part of first-line therapy results in a high percentage

of complete responses which seems to be predictive for prolonged progression-

free and overall survival, while all patients not achieving a complete response,

including the vast majority of the relapsed patients, have remissions of short

duration. Longer observation, however, is needed to determine the quality and

durability of these complete remissions. Late relapses from complete remissions

are not uncommon after ASCT for multiple myeloma [44]. The third conclusion

is that overall survival is remarkably good even in the pretreated patients. This

may be due to the efficacy of novel agents such as thalidomide, bortezomib and

DLI given to the patients who relapsed after NMA [220, 221]. Acute and chronic

GVHD were the most important toxicities and responsible for the fatal outcome

in five patients (9%). Nonrelapse mortality percentage may still increase due

to the considerable number of patients with chronic extensive GVHD. Chronic

GVHD, the most important negative factor for quality of life after NMA with

full stem cell grafts, is however a significant factor for prolonged progression-

free and overall survival.

Although our results and results from other studies are encouraging, the role of

NMA for myeloma is not yet established [147, 243]. In the recently published

prospective study by the French IFM, high-risk myeloma patients with an HLA-

21

identical family donor and treated with tandem autologous/NMA-ASCT had

comparable progression- free and overall survival to the patients with no donor

who were treated with double autologous SCT [74]. In this study in vivo T-cell

depletion was performed with high- dose ATG as part of the nonmyeloablative

conditioning regimen in all patients. The beneficial effect of in vivo T-cell deple-

tion is the low incidence of acute and chronic GVHD; the detrimental effect is

the elimination of the graft-versus-myeloma (GvM) effect [137]. The importance

of immune-competent donor T cells for GvM effect is illustrated by responses

to DLI and the occurrence of chronic GVHD [140]. European study groups,

including the Dutch Haemato-Oncology Association (HOVON), Spain’s Pro-

grama para el estudio y tratamiento de las hemopatias malignas (PETHEMA),

and the European Group for Blood and Marrow Transplantation (EBMT), are

performing comparable prospective donor versus no-donor studies. The res-

ults of these studies have to be awaited for more definite conclusions about the

value of NMA in multiple myeloma. In anticipation of the outcome of these

studies it is necessary to explore new strategies, which are aimed at stimulating

the cytotoxic efficacy of the donor T cells towards the residual myeloma cells

without enhancing GVHD. The suggestion that the novel antimyeloma agents

such as bortezomib, thalidomide, and lenalidomide may preferentially stimu-

late the graft-versus-tumor effect and not GVHD is fascinating in this respect

[121, 221].

In conclusion, NMA ASCT as part of first-line treatment of multiple myeloma

is feasible, is associated with acceptable transplant-related mortality and may

induce a high percentage of complete remissions of good quality and prolonged

duration. The outcome of prospective donor versus no-donor studies, however,

has to be awaited to better define the role of this treatment for multiple my-

eloma. In extensively pretreated patients response rate and progression-free

survival are disappointing and in this category of patients new strategies need

to be explored. These strategies should be aimed at enhancing the graft-versus-

tumor effect, probably by incorporating novel agents.

22

Part II

Graft-versus-Host Disease and

mechanisms of disease

23

Chapter 3

Rituximab treatment before reducedintensity conditioningtransplantation associates with adecreased incidence of extensivechronic GVHD

Suzanne van Dorp1,2, Floor Pietersma

2, Matthias Wölfl3,

Leo F Verdonck1, Eefke J Petersen

1, Henk M Lokhorst1, Edwin Martens

4,

Matthias Theobald1, Debbie van Baarle

2, Ellen Meijer1, and

Jürgen Kuball1,2

Biology of Blood and Marrow Transplantation, 2009

1Department of Hematology and Van Creveld Clinic, UMC Utrecht, The Netherlands2Department of Immunology, UMC Utrecht, The Netherlands3Children’s Hospital, University of Würzburg, Germany4Julius Center for Health Sciences and Primary Care, UMC Utrecht, The Netherlands

25

Abstract

Chronic graft-versus-host-disease (cGVHD) is the major cause of late mor-bidity and mortality after allogeneic stem cell transplantation. B cells havebeen reported to be involved in mediating cGVHD. To assess whether pre-emptive host B-cell depletion prevents extensive cGVHD after allogeneicreduced-intensity-conditioning-transplantation (RICT), 173 patients treatedwith RICT for various haematological diseases, who have or have not re-ceived Rituximab (Rtx) within 6 months prior to RICT, were analyzed ret-rospectively. Rtx-treatment within 6 months prior RICT reduced extensivecGVHD significantly from 45.8% to 20.1%. We hypothesize that most likelyhost B cells initiate cGVHD and, thus, a systematic B-cell depletion priorto RICT by Rtx might be a valuable strategy in order to reduce extensivecGVHD after RICT.

3.1 Introduction

Chronic graft-versus-host-disease (GVHD) is the major long term complication

of allogeneic stem-cell-transplantation (allo-SCT) as up to 70% of all survivors

of allo-SCT beyond day 100 develop chronic GVHD (cGVHD) [10, 130, 187].

Several lines of investigation indicate that B cells are involved in the develop-

ment of cGVHD [130, 210] and B-cell depletion using the monoclonal anti-CD20

antibody Rituximab (Rtx) has demonstrated benefit in the treatment of steroid-

refractory cGVHD with a success rate of up to 70% [36, 51, 179, 242]. Given

poor clinical responses in patients with steroid-refractory cGVHD [130] a po-

tential prophylactic value of Rtx in cGVHD warrants further pursuit. To date,

the answer to this question has been approached by others, but remains un-

answered. A study of patients after myeloablative allo-SCT suggested no differ-

ence in cGVHD in 35 patients treated with Rtx as part of the allo-SCT condition-

ing regimen, but indicated a possible decrease in overall acute GVHD (aGVHD)

incidence [113]. Therefore we studied, whether depletion of host B cells prior

to transplantation by Rtx can reduce the incidence of this quality-of-life and

life-threatening complication in patients who underwent allogeneic reduced-

intensity-conditioning-transplantation (RICT). Patients with and without Rtx-

treatment within 6 months prior to RICT were analyzed retrospectively and

were compared for the incidence of acute and cGVHD, graft-versus-leukemia-

effect (GVL) and overall survival (OS).

26

CHAPTER 3. DUAL ROLE FOR HOST B CELLS IN RICT

3.2 Design and methods

3.2.1 Power analysis and selection of patients. To calculate the minimum

amount of patients needed to see a significant difference in extensive cGVHD

between the Rtx pre-treated and non-Rtx-group an effect size, thus the expec-

ted difference between two groups was set at 25%. The Chi-square calculation

indicated that 18 patients per group were needed to achieve a power of 80%

(α = 0.05). 173 patients with various hematological diseases (Table 3.1), who

received a RICT at the University Medical Center Utrecht, from September 2001

until April 2007, were included in this retrospective study and allowed to fur-

ther increase the number of Rtx treated patients to 29 Rtx patients and 144

control patients, which increased the power from 80% to > 99%. One patient

in the Rtx-group has been treated with alemtuzumab 6 months prior to RICT.

RICT was given as curative or as rescue treatment to patients younger than 70

years with an available HLA-matched related or unrelated donor (1 HLA class

I or class II mismatch was allowed). Patients were treated according to clinical

protocols approved by the local ethics board and gave their informed consent.

Total population Rtx No Rtx p-value

N (%) 173 (100) 29 (17) 144 (83)

Median age (years) (range) 56 (20− 69) 58 (29− 67) 56 (20− 69) 0.662

Median follow up (months) (range) 28 (1− 40) 21.9 (1− 40) 28.8 (1− 40) 0.011

Sex % (female/male) 34/66 28/72 35/65 0.417

Disease, n < 0.001

ALL 6 0 6

AML 29 0 29

MDS 8 0 8

CLL 13 8 5

CML 6 0 6

MM 62 0 62

NHL 32 18 14

SAA 6 0 6

Myelofibrosis 2 0 2

Other 9 3 6

Donor, n (%) 0.925

MRD 111 (64) 19 (66) 92 (64)

PMRD 10 (6) 1 (3) 9 (6)

MUD 38 (22) 7 (24) 31 (22)

Continued on next page

27

Continued from previous page

Total population Rtx No Rtx p-value

PMUD 14 (8) 2 (7) 12 (8)

HLA-mismatch, n (%) 24 (14) 3 (10) 21 (15) 0.547

Patient male/donor female, n (%) 37 (21) 5 (17) 32 (22) 0.551

ATG as part of conditioning, n (%) 70 (41) 11 (38) 59 (41) 0.761

RT as part of pre-treatment, n (%) 25 (15) 5 (17) 20 (14) 0.639

Remission before RICT, n (%) 0.1

CR 55 (32) 4 (14) 51 (35)

PR 74 (43) 16 (55) 58 (40)

NR/progression 44 (25) 9 (31) 35 (24)

Median time of Rtx treatment before RICT(months) (range)

- 2.0 (1− 6) - -

Duration of IS, n (%) 0.235

Short 89 (51) 12 (41) 77 (54)

Long 84 (49) 17 (59) 67 (46)

Organs affected by limited cGVHD, n (% oflimited cGVHD)

Skin 16 (100) 5 (100) 11 (100) -

Liver 3 (19) 0 (0) 3 (27) 0.195

Organs affected by extensive cGVHD, n (%of extensive cGVHD)

Skin 50 (78) 4 (80) 46 (78) 0.916

Liver 17 (27) 2 (40) 15 (25) 0.479

Oropharynx 56 (88) 3 (60) 53 (95) 0.053

Eyes 30 (47) 2 (40) 28 (47) 0.748

Lungs 12 (19) 1 (20) 11 (19) 0.941

Gut 7 (11) 0 (0) 7 (12) 0.414

Kidneys 2 (3) 0 (0) 2 (3) 0.676

Tendons/Joints 2 (3) 1 (20) 1 (2) 0.024

Genitals 3 (5) 1 (20) 2 (3) 0.092

Pericard 1 (2) 0 (0) 1 (2) 0.769

Table 3.1: Patient characteristics. ∗p-values: Mann-Whitney U test for age and follow-up; χ2-test for other factors; ALL: acute lymfoblastic leukaemia; AML: acute myeloidleukaemia; ATG: antithymocyte globulin; CLL: chronic lymfocytic leukaemia; CML:chronic myeloid leukaemia; CR: complete remission; Disease: disease for which RICTwas given as treatment; cGVHD: chronic graft-versus-host disease; HLA: human leuk-ocyte antigen; IS: immunosuppression; MDS: myelodysplastic syndrome; MM: multiplemyeloma; MRD: matched related donor; NHL: non-Hodgkin lymphoma; No Rtx: Norituximab treatment within 6 months prior to RICT; NR: no remission; PMRD: par-tial matched related donor; PR: partial remission; RICT: reduced intensity conditioningtransplantation; RT: radiotherapy; Rtx: Rituximab treatment within 6 months prior toRICT; SAA: severe aplastic anemia.

28


3.2.2 Determination of B-cell counts. Absolute B-cell counts were determined

in whole blood using TRUcount tubes R© (Becton Dickinson), according to the

manufacturers’ protocol. In brief, whole blood samples were incubated with

anti-human CD19 (Fluorescein Isothiocyanate (FITC)-labeled, Becton Dickin-

son) in TRUcount tubes R©. Erythrocytes were lysed with lysing buffer (Becton

Dickinson). The samples were acquired on LSR-II (Becton Dickinson) flow cyto-

meter. Results were analysed with FACS DIVA software (Becton Dickinson).

Absolute B-cell numbers were calculated according to manufacturers’ protocol.

3.2.3 Conditioning regimen before RICT. For patients with a matched related

donor the conditioning regimen consisted of fludarabin (30 mg/m2/day i.v.

for 3 days) and of one fraction low dose total body irradiation (TBI) (2 Grays).

The transplantation with GCS-F mobilized peripheral blood hematopoietic stem

cells was performed after TBI. In the case of a HLA-mismatched family donor

or any unrelated donor rabbit antithymocyte globulin (ATG; 2 mg/kg/day for

4 days) was added to the regimen and infused before fludarabin was given.

Patients with multiple myeloma, who were treated with RICT, within 3 months

after high dose melfalan (200 mg/m2) and autologous SCT, only received TBI.

3.2.4 Immunosuppression after allo-SCT. In the post-transplantation period all

patients were treated with the immunosuppressants cyclosporin A (CSA) and

mycophenolate mofetil (MMF). Patients received 2× 4.5 mg/kg/day CSA until

day +84 (short) or day +120 (long). Hereafter CSA was tapered if no GVHD

was present. CSA dose was lowered in the case of raised creatinin levels or

severe side effects. Patients received 15 mg/kg/day MMF (maximum of 3

g/day) until day +28 (short) or +84 (long), also followed by tapering in the

absence of GVHD.

3.2.5 Acute and chronic GVHD. aGVHD was diagnosed and graded accord-

ing to the Glucksberg criteria [77], cGVHD was graded according to the Seattle

classification [71, 132]. aGVHD > grade 1 was treated with prednisone 1− 2

mg/kg/day, in case of skin localization, topical prednisone treatment was ap-

plied. Additionally CSA and MMF doses were increased or continued. Steroid-

29

refractory aGVHD was treated with sirolimus, tacrolimus, Rtx or more experi-

mental drugs, such as alemtuzumab and dacluzimab. cGVHD of the skin was

treated with topical prednisone. In severe cases of extensive cGVHD prednisone

1 mg/kg/day was given. Time to acute and cGVHD was calculated from the

date of transplantation until occurrence of acute or cGVHD.

3.2.6 Statistical analysis. Progression free survival (PFS) was defined as the

probability of being alive with no indication of disease progression. Overall

survival (OS) was defined as the probability of survival without considering the

occurrence or non-occurrence of relapse. PFS and OS were measured in months

and calculated from the date of transplantation until the date of first signs of

progression or last date of follow-up. Incidence of acute and cGVHD, EBV and

CMV reactivations, 95% donor chimerism and probabilities of PFS and OS were

calculated using the 1-Kaplan Meier method. Kaplan Meier curves were gener-

ated to illustrate survival and the log-rank test was used to compare survival

curves between subgroups. Univariate Cox regression was used to determ-

ine the prognostic value of various variables for the development of acute and

cGVHD, EBV and CMV reactivation, 95% donor chimerism, PFS and OS. These

variables included sex, age, disease type, HLA-mismatch, sex-mismatch, anti-

thymocyte-globuline as part of conditioning regimen (ATG), radiotherapy as

part of the treatment regimen of the prior disease (RT), conditioning, remission

state prior to RICT, aGVHD grade II-IV or III-IV, limited or extensive cGVHD,

status of T cell and non-T cell chimerism and application of Rtx. For acute

and cGVHD a univariate time-dependent Cox regression analysis was used to

assess its predictive value. Variables that had a p-value ≤ 0.10 in univariate

analysis were included in a multivariate Cox regression analysis, because the

number of available potential predictors exceeded the maximum allowed num-

ber. Rtx treatment was always included in multivariate analysis. E.g., the effect

of disease stage before RICT was assessed first with univariate Cox-regression

analysis and when a p-value < 0.10 was reached, in multivariate Cox-analysis

together with Rtx. We thereby corrected for stage of disease. Not more than 7

variables were used at once in multivariate analysis. Thereby it was guaranteed

that the compared groups would not be too small. Variables were entered all at

once in the Cox model. We used a backward logistic regression model to calcu-

30


late the effect hierarchically. Statistical analyses were performed with SPSS 15.0

for Windows (SPSS Inc., IL, and USA).

The significance of the difference between the B-cell counts on time points pre-

and post-transplantation of patients who received Rtx and of patients who did

not, was assessed by using a Mann-Whitney U test. Analyses were performed

with GraphPad Prism 4.0 for Windows (GraphPad Software Inc., San Diego,

CA, USA). A probability level of 5% (p < 0.05) was considered significant in all

analyses.

3.3 Results

3.3.1 Influence of Rtx-treatment prior to RICT on B-cell counts post-RICT. 173

patients with hematological malignancies who underwent RICT were analyzed

retrospectively (Table 3.1). Patients were divided in a group that received (Rtx,

n = 29) or did not receive (no-Rtx, n = 144) Rtx within 6 months prior to

RICT as B-cell depletion has been reported to last for up to 6 months [7, 22, 37].

The median time of the last administration of Rtx before RICT was 2.0 months

(range 1.0− 6.0 months). Patients received 375 mg/m2 rituximab per applica-

tion and in 27 cases this was administered in combination with chemotherapy

(R-CHOP/R-PECC; 6− 8 cycles). One patient with EBV reactivation post-solid

organ transplantation received three single applications of Rtx, one patient re-

ceived twice Rtx for chronic active EBV infection. To investigate whether B-cell

depletion with Rtx within 6 months prior RICT effects B-cell counts pre- and

post-RICT, B-cell counts from patients who did (n = 5) and did not receive Rtx

(n = 6) within 6 months prior to RICT where investigated from the study cohort

when B-cell counts were available prior and after RICT at least until 9 months

post-RICT. B-cell counts from patients were compared to B-cell counts from 14

healthy controls (median: 212.5/µl; range 88− 418/µl). Median B-cells counts

pre-RICT (13.0/µl; (n = 4)) and 3 months post-RICT (23.5/µl; (n = 6)) were

significantly lower (p < 0.01) in patients who received Rtx, as compared to a

healthy control group (n = 14) (Figure 3.1). In contrast, median of B-cell counts

in patients who did not receive Rtx prior to RICT differed not significantly be-

fore (182/µl; (n = 3)) and 3 months after (148.5/µl; (n = 6)) RICT as compared

31

Figure 3.1: B-cell counts pre- and post-RICT in patients who received (white bars) anddid not receive (black bars) Rtx within 6 months prior to RICT as compared to a healthycontrol group (grey bar). Statistical analyses compare B-cell counts of patients of eachtime point to B cell counts of the healthy control group. Counts before RICT (within 1week), or months after RICT are indicated. Bars indicate median and range. Numbersindicate the number of samples measured per time point. Only significant changes(p < 0.05) are indicated. Statistical analyses were performed with a Mann-Whitney Utest.

to the healthy control group (n = 14). 6 and 9 months after RICT, no significant

difference in B-cell counts was observed for both, the Rtx- and the no-Rtx-group,

when compared to the healthy control group (Figure 3.1). Thus, Rtx-treatment

within 6 month prior RICT reduces total number of B cells post-RICT at least

until 3 months after RICT.

3.3.2 Impact of Rtx-treatment prior to RICT on aGVHD and cGVHD. The in-

cidence of aGVHD was analyzed by both univariate and multivariate analysis.

There was no significant difference in the incidence of grade II-IV and grade III-

IV between the Rtx- (48.2% grade II-IV (n = 14); 17.9% grade III-IV (n = 5)) and

the no-Rtx-group (45.7% grade II-IV (n = 64) and 15.6% grade III-IV (n = 22))

(Figure3.2A and B). However, the median onset of grade III-IV aGVHD was sig-

nificantly earlier in the Rtx- as compared to the no-Rtx-group (Rtx-group 0.60

months, no-Rtx-group 1.60 months (p = 0.013); Figure 3.2C). Only ATG preven-

ted grade III-IV aGVHD significantly (p = 0.041, HR 2.57 (95% CI: 1.04− 6.38))

in multivariate-analysis with Cox regression (Table 3.2).

32


In order to assess whether Rtx-treatment prior to RICT prevents the develop-

ment of limited and extensive cGVHD, first the overall frequency of limited and

extensive cGVHD was determined and not significantly different between the

Rtx- and the no-Rtx-group (42.5% (n = 10) versus 54.9% (n = 69), p = 0.468), in

uni- nor in multivariate analysis (Figure 3.3A; Table 3.2). However, in the Rtx-

group 20.1% of patients developed extensive cGVHD (n = 5), while in the no-

Rtx-group 45.8% of patients developed extensive cGVHD (n = 58) (p = 0.053)

(Figure 3.3B). The main sites affected were skin, liver, mucosae and lung with no

statistical difference between the Rtx- and the no-Rtx-group (Table 3.1). After

multivariate analysis significantly less extensive cGVHD was detected in pa-

tients treated with Rtx (p = 0.035, HR 2.67 (95% CI: 1.07− 6.68)) (Table 3.2).

Only the application of ATG as part of the conditioning regimen (p < 0.001,

HR 4.35 (95% CI: 2.26− 8.38)) but no other analyzed factors also decreased the

incidence of extensive cGVHD in a multivariate analysis. Vice versa limited

cGVHD was increased in the Rtx-group (25.6% (n = 5)) as compared to the

no-Rtx- group (12.9% (n = 11), p = 0.088) (Figure 3.3C). Again this difference

reached significance after multivariate analysis (p = 0.040, HR 0.33 (95% CI:

0.11− 0.95)) (Table 3.2). In summary, Rtx-treatment prior to RICT did not af-

fect the frequency of grade II-IV or III-IV aGVHD but attenuated substantially

extensive cGVHD and increased thereby presumably the frequency of limited

cGVHD.

3.3.3 Rituximab prior to RITC did not associate with a difference in either PFS

or OS. Frequently, manipulations which result in a reduction of GVHD are

also associated with a reduction of GVL and vice versa [10, 130]. Therefore it

was assessed whether host B-cell depletion by Rtx-treatment might not only im-

pair GVHD but also influence GVL. As surrogate marker for GVL, progression

free survival (PFS) was assessed separately in the Rtx- and no-Rtx group. PFS

was decreased by HLA mismatch (p = 0.038, HR 0.51 (95% CI: 0.27 − 0.96))

and radiotherapy as part of the treatment regimen prior to RICT (p = 0.009,

HR 0.43 (95% CI: 0.23− 0.81)). Although Rtx-treatment was associated with a

significantly reduced incidence of cGVHD, PFS was not influenced (p = 0.306)

(Figure 3.4A; Table 3.2) by any other analyzed factors including type of dis-

ease. Also OS at 40 months differed not significantly (p = 0.159) between the

33

Figure 3.2: 1-Kaplan Meier curves of acute GVHD grade II-IV (A) and grade III-IV (B)in patients who received Rtx prior to RICT and in patients who did not. (C) Onset ofacute GVHD grade III-IV in patients who received Rtx prior to RICT and in patients whodid not.

34


Acute GVHD grade II-IV Univariate p-value Multivariate p-value HR 95% CI

Disease 0.051 0.806 1.022 0.861− 1.212

No HLA mismatch 0.058 0.065 2.189 0.951− 5.034

No Rtx 0.61 0.621 0.864 0.485− 1.541

Acute GVHD grade III-IV Univariate p-value Multivariate p-value HR 95% CI

No Rtx 0.702 0.733 0.844 0.320− 2.230

No ATG 0.034 0.041 2.574 1.039− 6.379

Totalchronic GVHD Univariate p-value Multivariate p-value HR 95% CI

No HLA mismatch 0.093 0.895 1.066 0.409− 2.779

No sex mismatch 0.036 0.132 0.664 0.391− 1.130

No Rtx 0.468 0.528 1.241 0.634− 2.427

No ATG 0.001 0.002 2.216 1.339− 3.667

No RT 0.099 0.34 1.515 0.645− 3.560

Acute GVHD 0.014 0.026 1.658 1.063− 2.586

Extensive chronic GVHD Univariate p-value Multivariate p-value HR 95% CI

Disease 0.054 0.16 0.846 0.670− 1.068

No HLA mismatch 0.069 0.727 0.801 0.232− 2.773

No Rtx 0.053 0.035 2.673 1.070− 6.675

No ATG < 0.001 < 0.001 4.354 2.261− 8.384

No Remission 0.07 0.978 0.99 0.487− 2.013

No aGVHD 0.046 0.102 1.515 0.921− 2.493

Limited chronic GVHD Univariate p-value Multivariate p-value HR 95% CI

Male sex 0.075 0.177 2.423 0.671− 8.746

Disease 0.011 0.026 1.469 1.047− 2.060

No Rtx 0.088 0.04 0.328 0.113− 0.950

Progression free survival Univariate p-value Multivariate p-value HR 95% CI

No HLA mismatch 0.006 0.038 0.507 0.267− 0.963

No Rtx 0.306 0.252 1.652 0.700− 3.901

No RT 0.001 0.009 0.434 0.232− 0.814

Non relapse mortality Univariate p-value Multivariate p-value HR 95% CI

Male sex 0.097 0.094 2.312 0.866− 6.174

Age < 55 yrs 0.042 0.038 0.378 0.150− 0.950

No Rtx 0.055 0.074 0.449 0.187− 1.081

Table 3.2: Multivariate Cox regression analysis of outcome in terms of acute GVHDgrade II-IV, acute GVHD grade III-IV, total chronic GVHD, extensive chronic GVHD,limited chronic GVHD, progression free survival, non relapse mortality. p-values <

0.05 are considered significant. ATG: anti-thymocyte globuline; CI: Confidence Interval;GVHD: graft-versus-host disease; HLA: human leukocyte antigen; HR: Hazard Ratio;RT: radiotherapy; Rtx: rituximab within 6 months prior to RICT.

35

Figure 3.3: 1-Kaplan Meier curves of total chronic GVHD (A) extensive chronic GVHD(B), and limited chronic GVHD (C) in patients who received Rtx prior to RICT and inpatients who did not.

36


Figure 3.4: Kaplan Meier curves of progression free survival (A) and overall survival(B) in patients who received Rtx prior to RICT and in patients who did not.

Rtx-group (n = 18; 61.4%) and no-Rtx-group (n = 103; 67.4%) (Figure 3.4B).

This observation is surprising as PFS and OS [8], in contrast to aGVHD and

cGVHD [50, 51], might have been heavily influenced by the imbalance in the

type of disease in this study cohort. However, the low number for certain en-

tities in this study cohort might have hampered the analysis. In summary, our

data demonstrate a reduced cGVHD in Rtx-treated patients, while GVL is not

affected.

3.4 Discussion

Discussion B cells play an important but yet unclear role in the pathogenesis

of cGVHD [130, 210] and B- cell depletion of donor B cells after allo-SCT with

the monoclonal anti-CD20 antibody, Rtx, has been shown to improve steroid

refractory cGVHD [36, 51, 179, 242]. We asked whether B-cell depletion by Rtx

prior to RICT could decrease the incidence of cGVHD and pre-emptive B-cell

depletion could therefore be a strategy to prevent this major complication after

allo-SCT. Our data indicate indeed that Rtx-treatment prior to RICT decreases

the incidence of extensive cGVHD while relapse rate was not influenced. This

result is in contrast to a previous study which retrospectively compared acute

and cGVHD in 35 leukemia patients, who received Rtx, and 31 control pa-

37

tients [113]. The authors reported no influence of rituximab on the incidence of

chronic GVHD and a possible but not significant decrease of the incidence of

acute GVHD. The main explanation for this difference might be the myeloab-

lative conditioning regimen which depletes the majority of immune cells, thus

also B cells. Consequently, the reported overall incidence of cGVHD was in both

arms very low (∼11%).

A significantly shorter follow-up period in Rtx patients than patients who did

not receive Rtx might have influenced the incidence of cGVHD. However, the

latest time point of onset of extensive cGVHD was 18.2 months (median 5.0

months (range 1.5− 18.2 months)) and the median follow-up was in both groups

> 20 months. So even though the follow-up of the Rtx group was shorter than

follow-up of the no-Rtx group, it was still long enough for extensive cGVHD to

develop.

B cells have been suggested to be final effector-cells in cGVHD by secreting IL6,

a known fibroblast-growth-factor [237]. However, in patients of whom B-cell

counts were available at longer follow up (6 or 9 months post-RICT), no signi-

ficant differences could be observed in total B-cell counts at the onset of cGVHD

and in most patients a full donor chimerism was reached (data not shown). This

suggests that the early depletion of most likely host B-cells rather than donor

B-cells are important for the development of pathogenesis of cGVHD. Thus,

alternative mechanisms must be responsible for the reduced incidence of ex-

tensive cGVHD if patients are treated with Rtx prior to RICT. Host B cells have

been proposed to serve as professional-antigen-presenting-cells, present minor-

histo-compatibility antigens, and induce GVL and GVHD [197]. In this light our

data suggest that host antigen-presenting B cells are under certain conditions

[153] required for initiating priming of donor T cells which mediate maximum

GVHD but no substantially GVL. Thus, host B cells might rather influence shap-

ing of a self-reactive donor T cell repertoire than directly mediating cGVHD.

We can only speculate as to why Rtx-treated patients who developed grade

III-IV aGVHD developed it earlier and conclusions have to been drawn very

cautiously due to the very low number of patients in the Rtx-arm (n = 5). One

explanation for this more rapid course of aGVHD could be that B cells serve as

regulatory B cells, as reported in mice, and provide IL10 to attenuate the course

38


of disease [188]. However, we cannot exclude that also other modulators of the

immune system such as regulatory T cells [165, 244] or Th17 [241] cells were

indirectly effected by B-cell depletion.

We are aware of major limitations of this study, such as imbalance in disease,

which primarily might have influenced the analysis of PFS and OS but not the

incidence of acute and cGVHD. Thus, our data indicate that most likely host

B cells are important in the pathogenesis cGVHD. In order to reduce cGVHD,

pre-emptive B-cell depletion might be therefore beneficial prior to RICT.

39

40

Chapter 4

The immunological phenotype ofrituximab-sensitive chronicgraft-versus-host disease: a phase IIstudy

Suzanne van Dorp1,2,∗, Henrike Resemann

2,∗, Liane te Boome2,

Floor Pietersma1, Debbie van Baarle

1,3, Frits Gmelig-Meyling1, Roel de

Weger4, Eefke Petersen

2, Monique Minnema2, Henk Lokhorst

2,

Saskia Ebeling1, Marijke van Dijk

4, Ellen Meijer2

and Jürgen Kuball1,2

Haematologica, 2011

1Dept of Immunology, UMC Utrecht, Utrecht, the Netherlands2Dept of Hematology, UMC Utrecht, Utrecht, the Netherlands3Dept of Internal Medicine and Infectious Diseases, UMC Utrecht, Utrecht, the Netherlands4Dept of Pathology, UMC Utrecht, Utrecht, the Netherlands∗Both authors contributed equally to this work

41

Abstract

Chronic graft-versus-host disease is the major long-term complication afterallogeneic stem cell transplantation with a sub-optimal response rate to cur-rent treatments. Therefore, clinical efficacy and changes in lymphocyte sub-sets before and after rituximab treatment were evaluated in a prospectivephase II study in patients with steroid-refractory chronic graft-versus-hostdisease. Overall response rate was 61%. Only responding patients werefound to have increased B-cell numbers prior to treatment. B cells had anaïve-antigenpresenting phenotype and were mainly CD5 negative or hada low CD5 expression. Normal B-cell homeostasis was reestablished in re-sponding patients one year after ritxumab treatment and associated with asignificant decline in skin-infiltrating CD8+ T cells, suggesting that host Bcells play a role in maintaining pathological CD8+ T-cell responses. Imbal-ances in B-cell homeostasis could be used to identify patients a priori witha higher chance of response to rituximab treatment (Eudra-CT 2008-004125-42).

4.1 Introduction

Graft-versus-host-disease (GVHD) is the most common and life-threatening

complication after allogeneic stem cell transplantation (allo-SCT) [84, 200]. B-

cell depletion with rituximab (RTX) has been successful in steroid-refractory

chronic GVHD, showing response rates of 43− 80% [36, 51, 159, 179, 216, 242].

However, the nature of B-cell contribution, as well as to what extent B-cell de-

pletion can restore physiological conditions, has so far not been clarified. Hy-

potheses obtained from mouse models of chronic GVHD and retrospective ana-

lysis of patient materials have been conflicting. For example, a correlation was

found between high levels of B-cell activating factor (BAFF) in a retrospective

analysis of 45 patients with active chronic GVHD [192]. However, no significant

correlation could be found between the levels of BAFF and RTX-responsiveness

in the most recently published prospective phase II study of 37 patients [115].

In active chronic GVHD prior to response to immunosuppressants, expansion

of activated CD27+ B cells has been observed [192]. Another retrospective study

of 35 patients suffering from active chronic GVHD, in which treatment included

prednisone and calcineurin inhibitors, showed significantly lower numbers of

memory B cells (CD27+) [83]. To our knowledge, no prospective comparisons

42

CHAPTER 4. RITUXIMAB-SENSITIVE CHRONIC GVHD

have been made between immune subsets before and after B-cell depletion ther-

apy of steroid-refractory chronic GVHD, so details on reestablishment of nor-

mal B-cell pools after RTX treatment, as well as changes in immune-pathology

of the skin, still have to be clarified. Consequently, the aim of this study was

to demonstrate effectiveness of B-cell depletion therapy in steroid-refractory

chronic GVHD and to identify potentially involved cell subsets by a compre-

hensive immunological analysis in a prospective clinical trial.

4.2 Design and methods

4.2.1 Patients and patient material. In the course of a prospective study

(Eudra-CT 2008-004125-42), a cohort of 20 chronic GVHD patients who received

allo-SCT due to various hematologic malignancies was treated with 4 weekly

doses of 375 mg/m2 RTX (F. Hoffmann-La Roche Ltd., Basel, Switzerland). All

patients had chronic GVHD with at least skin symptoms. Inclusion criteria

were age over 18 years, life-expectancy of more than six months, and a World

Health Organization (WHO) performance status of 2 or under. All patients were

steroid-refractory or steroid-dependent. Refractory chronic GVHD was defined

as progression of disease after at least two weeks of prednisone treatment (ap-

proximately 1 mg/kg) or no response after four weeks of prednisone treatment.

Steroid-dependent chronic GVHD was defined as an inability to completely

taper immunosuppressive treatment. Patients received RTX treatment at the

University Medical Center Utrecht between March 2007 and February 2010 ac-

cording to clinical protocols approved by the local ethics board. In cases of

progression or recurrence of chronic GVHD, for which an alternative systemic

therapy was needed, patients were excluded from the study and follow up was

ended. Response criteria were set as the following: complete response (CR) and

partial response (PR) defined according to the recently published National Insti-

tutes of Health (NIH) criteria [173]. For laboratory analysis, blood samples were

taken from each patient before and after treatment, and consecutively every two

months until one year after treatment. Peripheral blood mononuclear cells (PB-

MCs) were immediately isolated and lymphocyte numbers measured by Tru-

Count (according to the manufacturer’s protocol, BD Biosciences). PBMCs were

then frozen and stored in liquid nitrogen until further analysis. Plasma and

43

serum were stored at −80◦C until further analysis. Furthermore PBMC, plasma

and serum samples from allo-SCT recipients without GVHD (No-GVHD) and

healthy donors were obtained. An informed consent was obtained for other

control samples to be used for analysis. Skin biopsies were obtained before

RTX therapy and five months after treatment from adjacent sites if possible,

otherwise from the most active site, and stored in 4% formalin and embedded

in paraffin. Eye involvement was measured before RTX, three and seven months

after treatment, using a Schirmer’s test. FACS staining, cytokine analysis [54], B-

cell clonality, chimerism [224], and auto-antibodies [103], histological stainings

[36] and the control groups used are described in the Supplementary Design

and Methods section and Supplementary Tables 4.3 and 4.4.

4.2.2 Statistical analysis. For prospective data analysis of the cohort of 20 pa-

tients SPSS (IBM, Chicago, USA) was used. Receiver operating characteristic

(ROC) curve was used to determine a cut-off point for B-cell numbers that was

predictive for responsiveness to treatment. Hazard ratios were calculated us-

ing a Cox’s regression analysis. Incidence of response was calculated using the

Kaplan-Meier method. A Kaplan-Meier curve was used to illustrate responsive-

ness and a log rank test was used to compare incidence between patients with

high and low B-cell numbers. Data from FACS staining, plasma and serum

analysis were analyzed using GraphPad Prism 5 for Windows (GraphPad Soft-

ware, La Jolla, USA). Differences in lymphocyte subsets were compared using

two-way ANOVA for normally distributed data. Gaussian-distributed groups

were compared using Student’s t-test. Groups of data, which were not nor-

mally distributed were compared using Mann-Whitney U tests. In either case,

a probability level of 5% (p < 0.05) was found to be significant.

4.3 Results and Discussion

In order to prospectively test clinical efficacy of B-cell depletion therapy in

steroid-refractory chronic GVHD, a cohort of 20 patients presenting with at

least skin involvement was treated with rituximab and followed until one year

after treatment or until relapse of chronic GVHD. Two patients had to be ex-

cluded from further study; one due to an allergic reaction to rituximab and

44


Res

pons

en

(%)

Med

ian

TT

RR

espo

nse

dura

tion

Mon

ths

(ran

ge)

Mon

ths

(ran

ge)

OR

CR

PRSD

Prog

ress

ion

Tota

l(n

=18

)11

(61)

011

(61)

3(1

7)4

(22)

3(1−

4)12

(1−

12)

Org

anin

volv

emen

tO

RC

RPR

SDPr

ogre

ssio

n

Skin

(n=

18)

Eryt

hern

a(n

=17

)13

(76)

4(2

4)9

(53)

2(1

2)2

(12)

3(1−

8)6

(3−

12)

Ulc

era

(n=

3)0

00

1(3

3)2

(67)

xx

Mov

able

scle

rosi

s(n

=5)

4(8

0)2

(40)

2(4

0)0

1(2

0)1

(1−

3)10

(3−

12)

Dee

psc

lero

sis

(n=

12)

9(7

5)4

(33)

5(4

2)2

(17)

1(8

)1.

5(1−

3)8

(3−

12)

Eyes

(n=

15)

6(4

0)4

(27)

2(1

3)7

(47)

2(1

3)2

(1−

6)6.

5(4−

12)

Ora

lmuc

osa

(n=

8)3

(38)

1(1

2)2

(25)

5(7

5)0

1(1−

6)n.

d.

GI

trac

t(n

=7)

3(4

3)2

(29)

1(1

4)4

(57)

01

(x)

n.d.

Dos

ere

duct

ion

n.d

Pred

niso

ne(n

=18

)9

(50)

4(2

2)5

(28)

9(5

0)x

3(1−

7)

Cyc

losp

orin

eA

(n=

5)2

(40)

2(4

0)0

3(6

0)x

3(2−

4)

MM

F(n

=4)

2(5

0)1

(25)

1(2

5)2

(50)

x3.

5(3−

4)

Tabl

e4.

1:To

tal

resp

onse

rate

s,re

spon

sera

tes

per

orga

nan

ddo

sere

duct

ion

ofim

mun

osup

pres

sant

saf

ter

trea

tmen

tw

ith

RTX

.CR

indi

cate

sco

mpl

ete

resp

onse

/red

ucti

on;

GI:

gast

roin

test

al;

MM

F:m

ycop

heno

late

mof

etil;

OR

:ove

rall

resp

onse

/red

ucti

on;

PR:p

arti

alre

spon

se/r

educ

tion

;SD

:sta

ble

dise

ase/

dose

;TTR

:tim

eto

resp

onse

/red

ucti

on.

45

Figure 4.1: Possible predictive markers for responsiveness to RTX treatment and im-munological changes after RTX treatment. (A) Absolute numbers of naïve (CD27−), andactivated (HLA-DRhigh) B cells in responding, non-responding patients, No-GVHD andhealthy donor controls. Arrows indicate the control group on which statistical analysiswas performed by a Mann-Whitney U test (∗p < 0.05). (B) Representative dot plots ofcell surface expression of CD5 and CD20 in responding patients at T = 0 and T = 12,non-responding patients at T = 0, No-GVHD and healthy-donor controls. Number ofskin-infiltrating (C) CD4+ and CD8+ T cells before and five months after RTX treatmentin responding patients as well as percentage of skin-infiltrating CD8+ T cells before andafter RTX treatment in responders and healthy controls. Statistical analysis was per-formed using an unpaired t test and a two-way ANOVA (∗p < 0.05, ∗∗p < 0.01).

46


Factor HR 95% CI p-value

Age 0.985 0.914− 1.061 0.689

Sex 0.796 0.210− 2.980 0.728

ATG 0.925 0.198− 4.307 0.920

Acute GVHD 1.585 0.418− 6.009 0.498

B cells > 0.40× 109/L 10.683 1.237− 92.235 0.031

Low IL-21 3.615 0.435− 30.057 0.234

Ulcerative skin 0.033 0.000− 16.323 0.281

Table 4.2: Predictive value of various factors for responsiveness of RTX treatment.ATG indicates anti-thymocyte globuline as part of conditioning regimen; CI: confidenceintenval; GVHD: graft-versus-host disease; HR: hazard ratio. HR and p-values: univari-ate Cox’s regression model.

one due to relapse of leukemia. Eighteen patients could, therefore, be included

for further analyses. Patients’ characteristics are shown in the Supplementary

Table 4.5. Overall response rate was 61% (n = 11). Only partial responses

were seen during the time of follow up. Median time to response was three

months (range 1− 4 months) and 55% of responders had an ongoing response

(n = 6). Median response duration, measured until last time of follow up, was

12 months (range 1− 12 months) (Table 4.1). Dosage of prednisone could be

reduced in 50% of patients (n = 9) and completely stopped in 4 patients (22%).

Median time to dose reduction of prednisone was three months (range 1− 7

months) (Table 4.1). To investigate whether the production of auto-antibodies

was associated with symptoms of chronic GVHD as reported [172, 196], serum

before and after rituximab treatment of patients and No-GVHD controls was

tested for a panel of antibodies correlated with Systemic Sclerosis (SSc) in terms

of quality (type) and quantity. Several auto-antibodies were found in serum of

both responders and non-responders, as well as in serum of No-GVHD con-

trols. However, no significant associations between presence of antibodies and

chronic GVHD could be found (data not shown) as also reported by others

[203]. Conflicting data have also been reported on the correlation between

BAFF-levels or BAFF-to-B-cell ratios in RTX-responding patients, as the recently

published prospective study of 37 patients [115] did not show a significant

correlation in contrast to a retrospective study of 20 patients [193]. Also, our

prospective study of 18 patients did not show any correlation between BAFF-

47

levels and RTX response. However, differences in these studies and our data

could also be partially a consequence of the fact that patients received different

doses of corticosteroids in different studies, and high doses of corticosteroids

as used in our study have been reported to partially inhibit BAFF [105]. In

our study, only IL-21 was significantly decreased in responding as compared

to non-responding patients (data not shown). Peripheral blood mononuclear

cells from different groups were analyzed by flow cytometry for lymphocyte

subsets, and there was no significant difference in total lymphocyte numbers

between patient groups, No-GVHD and healthy donor controls. No significant

differences were observed between responders and non-responders when com-

paring CD8+ and CD4+ T cells in the peripheral blood. Also, regulatory T cells

(Tregs, CD3+CD4+CD25−CD127+FoxP3+), naïve, effector memory and central

memory, distinguished on the basis of CD62L and CD45RO expression, did not

show any significant difference between all groups at any time point, as well as

T cells expressing early (CD69), intermediate (CD137) and late (HLA-DR) activ-

ation markers (data not shown). B-cell numbers in responding patients before

treatment (T = 0) were increased with significantly higher absolute numbers

of naïve B cells (CD19+CD20+CD27−) and CD86+ and HLA-DRhigh B cells

when compared to both non-responders, No-GVHD, and healthy donor con-

trols (against all controls all < 0.05, Figure 4.1A). A cut-off point for B cells of

0.40× 109/L was calculated using a ROC curve. The relative risk of respons-

iveness in patients with absolute B-cell numbers more than 0.40× 109/L was

estimated using a univariate Cox’s regression analysis. Patients with absolute

B-cell numbers of more than 0.40× 109/L had a 10.7 times higher chance of be-

ing responsive to rituximab treatment (HR 10.7 [95% CI 1.2− 92.3]; p = 0.031).

B-cell number was the only factor found to be of significant predictive value

(Table 4.2). Rituximab sufficiently depleted all B cells in the peripheral blood

of all patients as indicated by the lack of CD19+ cells at one month after the

first rituximab dose (T = 1; data not shown). B-cell reconstitution in peripheral

blood could only be observed 10− 12 months after start of rituximab. The MFI

of CD5+ B cells was significantly decreased in responding patients before treat-

ment (T = 0) and normalized until T = 12 (p < 0.05), and was, therefore, then

again comparable to all controls. These differential expression patterns of CD5

in patients and all controls are shown in representative dot plots in Figure 4.1B.

At T = 12, also CD5− B-cell numbers, naive B-cell numbers, and HLA-DRhigh B-

48


cell numbers in responding patients were again comparable to all controls, thus

with a significant difference when compared to T = 0 (p < 0.05). Knowledge

about CD5 and its signaling functions in lymphocytes is still very limited and

the biological role of CD5− and CD5+ B cells differ in mice and men [52, 61]. In

contrast to our study, an increase in CD5+ B cells has been observed in patients

with other autoimmune disease such as lupus erythematosus [61]. However,

regardless of its function, loss of CD5 expression on B cells might assist in

identifying patients who will benefit from B-cell depletion.

To investigate whether responding and non-responding patients also show dif-

ferences in infiltrating immune cells of the skin, skin sections of responders,

non-responders, and healthy donors were additionally stained for T- and B-cell

markers. As reported [35], only minimal immune cell infiltrates could be ob-

served in patients with chronic GVHD: total numbers of T cells were within

the range of T-cell infiltrates usually observed in healthy individuals (data not

shown). However, the percentage of skin-infiltrating CD8+ T cells was signi-

ficantly higher when compared to healthy controls (p < 0.05; Figure 4.1C). No

significant difference was observed in skin infiltrating T cells before rituximab

treatment between responding and non-responding patients (data not shown).

There was no difference in numbers of skin infiltrating B cells between respond-

ing and non-responding patients, and B cells disappeared after rituximab treat-

ment (data not shown). After rituximab treatment, only ongoing responding

patients were analyzed at five months after study entry as nonresponding pa-

tients had already undergone other therapies and were thus no longer eligible.

A significant decrease in CD8+ T cells was observed after five months in re-

sponding patients (p < 0.01), whereas there was no significant change in CD4+

T-cell numbers (Figure 4.1C and D). This resulted again in normalization of

the percentage of skin-infiltrating CD8+ T cells when compared to the healthy

control group (Figure 4.1C). This suggests that B-cell depletion by rituximab

not only reduces the number of potentially antigen-presenting B cells, but also

reduces the skin infiltrating CD8+ T-cell compartment. This, therefore, sup-

ports recent hypotheses of a T-cell to B-cell crosstalk in the setting of chronic

graft-versus-host disease in man [203].

In summary, to our knowledge this is the first prospective comprehensive study,

49

which describes the immunological phenotype of RTX-sensitive as compared to

RTX-unresponsive chronic graft-versus-host disease in the peripheral blood and

the skin. Elevation of B-cell numbers with a dominant naïve, antigen-presenting

phenotype, as well as skewing towards CD5− B cells and B cells with a low CD5

expression was selectively found in responding patients and was the only pre-

dictive factor of responsiveness to rituximab. Physiological B-cell homeostasis

was re-established in responding patients one year after treatment and associ-

ated with a reduced skin infiltration of CD8+ T cells in responding patients.

This suggests that an imbalanced B-cell repertoire can contribute to chronic

graft-versus-host disease by sustaining skin-infiltrating CD8+ T cells. These

findings could also be useful in identifying in advance those patients who will

benefit from rituximab treatment and provide a basis for larger confirmatory

prospective clinical trials.

4.4 Supplementary Design and Methods

4.4.1 FACS staining. For phenotypic analysis, PBMCs from patients, No

GVHD controls and HD controls were stained with antibodies against the

following markers, with fluorescent labels as indicated: CD3-PerCP, CD4-

PerCP, CD80-R-PE, IFN-g-FITC, CD69-FITC, CD137-PE, CD5-PE and CD62L-

PE-Cy5 (all from BD Pharmingen), CD8-PerCP, CD69-APC, CD19-APC, CD138-

PerCP, IL-17-PacBlue (all from BioLegend), CD3-eFlour-450, CD4-PE-Cy7, CD4-

Alexa Flour 750, CD8-APC, CD25-FITC, CD127-PE-Cy7, HLA-DR-Alexa Flour

750,CD38-PE-Cy7, CD86-PE-Cy5, CD27-eFour 780, FoxP3-APC, CD20-PacBlue,

IL-4-PE-Cy7, and IL-10-PE (all from eBioscience).

For FACS analysis test samples of 300, 000 cells were analyzed. For evaluation

of cytokine production capacities of lymphocytes, cells were stimulated with IL-

2 (20 u/ml, Novartis Pharmaceuticals) and PHA-L (30µg/ml, Sigma Aldrich).

After 4 hours of stimulation, cells were stained for extracellular and intracellu-

lar markers. FoxP3-staining was performed according to the manufacturer’s in-

structions for intracellular staining (eBioscience). Samples were processed with

a LSR-II flow cytometer (BD Biosciences). The acquired data were analyzed

using FACS Diva software (BD Biosciences).

50


4.4.2 Cytokine analysis. For cytokine analysis, plasma samples from patients,

No GVHD and HD controls were examined for their content of IL-2, IL-10,

IL-12p70, interferon-gamma (IFN-γ), tumor necrosis factor-alpha (TNF-α), IL-4,

IL-13, IL-6, IL-17, and IL-21 using multiplex immunoassays as described earlier

[54]. Transforming growth factor-beta (TGF-β), BAFF and platelet-derived

growth factor-AA (PDGF-AA) were measured with ELISA according to manu-

facturer’s instructions (BD Biosciences, Bender MedSystems (TGF-β and BAFF)

and Antigenic America (PDGF-AA). PDGF-AA was measured in plasma, while

TGF-β and BAFF were measured in serum samples.

4.4.3 B-cell clonality, chimerism and auto-antibodies. For clonality assessment

of B cells, genomic DNA was isolated from patient PBMC samples using a

nucleospin blood quick pure kit (Qiagen). B-cell receptor diversity was analyzed

using BIOMED multiplex PCR assays as described earlier [224].

Chimerism analysis of T and B cells was performed by PCRbased amplific-

ation of short tandem repeats sequences as described earlier [103]. For ana-

lysis of auto-antibodies, an immunoblot for SSc-specific auto-antibodies was

used according to the manufacturer’s instructions (Euroimmun Lübeck, Ger-

many). Sera of patients and No-GVHD controls were analyzed for IgG antibod-

ies against Scl-70, CENP A, CENP B, RP11, RP155, Fibrillarin, NOR90, Th/To,

PM-Scl100, PM-Scl75, Ku, PDGFR and Ro-52.

4.4.4 Histological stainings. Skin biopsies were stored in 4% formalin and em-

bedded in paraffin. Slides were stained with hematoxylin and eosin (H&E,

Klinipath), and monoclonal antibodies against CD3 (A0452), CD8 (M7103),

CD20 (M0755; all from Dako), CD4 (Monosan, monx10326), CD5 (Novocasta,

NCL-CD5-4C7) and FoxP3 (eBioscience, 14-4776). Slides were stained for all

markers, except FoxP3, using a BondmaX stainer (Leica). Slides were stained

for FoxP3 manually. Epidermal involvement and dermal sclerosis was scored

as described earlier [36]. Pathologists were clinically blinded during analysis.

Nine skin biopsies of the leg, arm and trunk, which were obtained from healthy

donors after receiving their informed consent served as controls.

51

No-GVHD controls (n = 5)

Median age (yrs; range) 50 (45− 63)

Sex M/F (%) 60/40

Disease (n)

• AML 3

• CML 1

• NHL 1

Related donor (n,%) 4 (80)

NMA conditioning (n,%) 5 (100)

ATG (n,%) 1 (20)

Acute GVHD (n) 0

Table 4.3: Characteristics of No-GVHD control group used for flow-cytometry analyses.AML: acute myeloid leukemia; ATG: antithymocyte globuline; CML: chronic myeloidleukemia; NHL: non-Hodgkin’s lymphoma; NMA: non-myeloablative.

Healthy controls (n = 5)

Median age (yrs; range) 30 (23− 40)

Sex M/F (%) 40/60

Table 4.4: Characteristics of healthy controls used for flow-cytometry analyses.

52


Tota

lpop

ulat

ion

Res

pond

ing

pati

ents

Non

-res

pond

ing

pati

ents

p-va

lue

N(%

)18

(100

)11

(67)

7(3

3)

Med

ian

age

(yea

rs;r

ange

)53

(39−

66)

53(3

9−

64)

55(4

4−

66)

0.75

1

Sex

mal

e/fe

mal

e(%

)78

/22

73/2

786

/14

0.48

5

Med

ian

follo

w-u

p(m

onth

s;ra

nge)

7(1−

13)

8(4−

13)

4(1−

4)0.

001

Num

ber

ofpr

e-tr

eatm

ents

11

11.

000

Med

ian

tim

eaf

ter

allo

-SC

T(m

onth

s;ra

nge)

34(1

0−

61)

38(9−

77)

34(1

1−

62)

0.31

9

Mon

ths

from

onse

tof

chro

nic

GV

HD

unti

lRTX

-tre

atm

ent

(med

ian;

rang

e)12

(2−

51)

15(3−

51)

8(2−

43)

0.44

1

Dis

ease

0.45

0

•A

MI/

MD

S2

02

•C

LL2

02

•C

ML

11

0

•M

M8

53

•M

yelo

fibro

sis

11

0

•N

HL

44

0

Rel

ated

dono

r(n

;%)

14(7

8)9

(82)

5(7

1)0.

515

NM

Aco

ndit

ioni

ng(n

;%)

16(8

9)10

(91)

6(8

6)0.

641

ATG

(n;%

)4

(22)

2(1

8)2

(29)

0.51

5

Acu

teG

VH

D(n

;%)

14(7

8)8

(73)

6(8

6)0.

428

Tabl

e4.

5:A

llo-S

CT

indi

cate

sal

loge

neic

stem

cell

tran

spla

ntat

ion.

AM

L:ac

ute

mye

loid

leuk

emia

;ATG

:ant

ithy

moc

yte

glob

ulin

e;C

LL:

chro

nic

lym

phoc

ytic

leuk

emia

;C

ML:

chro

nic

mye

loid

leuk

emia

;D

isea

se,d

isea

sefo

rw

hich

allo

-SC

Tw

asgi

ven;

GV

HD

:gra

ft-v

ersu

s-ho

stdi

seas

e;M

DS:

mye

lody

spla

stic

synd

rom

e;M

M:m

ulti

ple

mye

lom

a;N

HL:

non-

Hod

gkin

’sly

mph

oma;

NM

A:n

on-m

yelo

abla

tive

.∗ p

-val

ues:

Man

n-W

hitn

eyU

test

for

age,

follo

wup

,mon

ths

afte

ral

lo-S

CT

and

chro

nic

GV

HD

;Fis

hers

exac

tte

stfo

rot

her

fact

ors.

53

54

Part III

Towards new effectors with

the potential to control viral

infections and leukemia

55

Chapter 5

γ9 and δ2CDR3 domains regulatefunctional avidity of T-cellsharboring γ9δ2T-cell receptors

Cordula Gründer1, Suzanne van Dorp

1, Samantha Hol1, Esther Drent

1,

Kirsten Scholten1, Sabine Heijhuurs

1, Wouter Scheper1, Zsolt

Sebestyen1, Roland Strong

2and Jürgen Kuball

1

Blood, 2012

1Department of Hematology and Immunology, University Medical Center Utrecht, Utrecht, TheNetherlands

2Fred Hutchinson Cancer Research Center, Seattle, United States

57

Abstract

Immunotherapy with innate immune cells has recently evoked broad in-terest as a novel treatment option for cancer patients. γ9δ2T-cells in particu-lar are emerging as an innate cell population with high frequency and stronganti-tumor reactivity, which makes γ9δ2T-cells and their receptors prom-ising candidates for immune-interventions. However, clinical trials have sofar reported only limited tumor control by adoptively transferred γ9δ2T-cells. As a potential explanation for this lack of efficacy, we found unexpec-tedly high variability in tumor recognition within the physiological humanγ9δ2T-cell repertoire, which is substantially regulated by the CDR3 domainsof individual γ9δ2T-cell receptors (TCR). We demonstrate that here repor-ted molecular needs of CDR3 domains to interact with a targetcell shapethe physiological γ9δ2T-cell repertoire and most likely limit protective andtherapeutic efficacy of γ9δ2T-cells, the first defense against tumors. Basedon these findings, we propose combinatorial-γδTCR-chain-exchange (CTE)as an efficient method for designing high-affinity γ9δ2TCRs that mediateimproved anti-tumor responses, both in vitro and in vivo in a humanizedmouse model.

5.1 Introduction

Immunotherapy with innate immune cells has become widely used because this

approach obviates the need to match a cellular product to a defined human-

leukocyte-antigen (HLA) haplotype, allowing adoptive immunotherapies to be

used in virtually any cancer patient without extensive in vitro selection or ma-

nipulation of the cellular product [82, 189, 207]. γ9δ2T-cells are promising as

an innate cell population for this purpose. They are usually observed at high

frequencies in the human peripheral blood and provide a very strong antitu-

mor reactivity against various solid and hematological cancers [38]. However,

within γ9δ2T-cell populations, individual clones display great diversity in the

repertoire due to the activating or inhibitory receptors expressed [164]. Select-

ing innate cell products for certain cell types, like those with a low level of in-

hibitory receptors, therefore seems plausible, especially considering the limited

efficacy of adoptively transferred innate immune cells in clinical trials [75, 167].

An alternative proposal is to engineer cells to express defined activating innate

receptors that mediate strong anti-tumor-reactivity, such as a defined γ9δ2T-

58

CHAPTER 5. γ9 AND δ2CDR3 DOMAINS

cell receptor (TCR) 8, which could pave the way for readily available and more

effective cellular products. However, the molecular details of how a γ9δ2TCR

interacts with its target are not fully understood, making it challenging to select

defined γ9δ2T-cells or to engineer T-cells with defined γ9δ2TCRs.

In ‘classical’ immunoreceptors, such as αβTCRs or immunoglobulins (Igs), the

complementary determining regions (CDR) determine affinity and specificity

for a specific peptide epitope. V(D)J recombination allows the creation of a

highly variable CDR repertoire ensuring recognition of an immense collection

of antigens. γ9δ2T cells also possess a rearranged TCR that mediates recog-

nition. Phosphoantigen isopentenyl pyrophosphate (IPP) has been suggested

to be a key player in γ9δ2TCR mediated activation [38, 78, 111], but no dir-

ect interaction between a γ9δ2TCR and IPP or any other phosphoantigen has

ever been demonstrated. It was previously suggested that positively charged

residues within the γ9δ2TCR are crucial for the response to negatively charged

phosphoantigens [4, 163] and a potential IPP-binding groove has been proposed

[4]. Interestingly, it appeared that responsiveness to phosphoantigens depends

in particular on germline-encoded residues within all CDRs apart from δCDR3

[232], extending the footprint of recognition to a much larger region than ini-

tially predicted.

Sequence alignment studies suggested that no defined δCDR3 motif is re-

quired for recognition beyond a hydrophobic residue at position δ109 (Kabat-

numbering δ97 [134]), which suggests a less dominant role for δCDR3 [33,

157, 232]. Consequently, it is still unclear why variations in the γ9δ2CDR3

regions—which are particularly abundant in δCDR3—have evolved in humans

and whether this variability is important in regulating activation of a γ9δ2T

cell. Understanding the reason for this variation would help to explain either

the specificity or the regulation of functional avidity of a γ9δ2T cell. This would

also provide insight into the role of a γ9δ2TCR during the selection process of

a γ9δ2T cell and allow engineering of therapeutic cells with higher anti-tumor

activity. In our study, we therefore asked the following research questions: 1)

what is the clonal diversity in terms of tumorspecificity and functional avidity

within γ9δ2T cells once they express an identical set of activating and inhibit-

ory receptors, 2) what is the specific role of individual γ9δ2TCRs, and 3) can we

59

engineer a γ9δ2TCR with improved and broader anti-tumor-reactivity?

5.2 Material and Methods

5.2.1 Cells and cell lines. Daudi, MDAMB231, SW480, K562, MCF-7, BT-

549, Phoenix-Ampho and Jurkat cells were obtained from ATCC, MZ1851RC

from Barbara Seliger (University Halle, Germany). Saos2 was kindly provided

by Arnold Levine (Princeton University, USA), FaDu and SCC9 by Niels

Bovenschen (UMC Utrecht, the Netherlands) and Mogens Claesson (Univer-

sity Copenhagen, Denmark), Psf5 and MRC5 by Bodo Plachter (Mainz, Ger-

many). Phoenix-ampho cells were cultured in DMEM+1%Pen/Strep (Invitro-

gen) +10%FCS (Bodinco), all other cell lines in RPMI+1%Pen/Strep+10%FCS.

PBMCs were isolated from buffy coats obtained from the Sanquin Blood Bank

(Amsterdam, The Netherlands). PBMC samples from AML patients were a kind

gift from Matthias Theobald (Mainz, Germany) and were collected according to

GCP and Helsinki regulations.

5.2.2 TCR Mutagenesis, Cloning and Sequencing. All γ9δ2TCR modifications

are based on codon-optimized genes of γ9- or δ2-TCR chain G115 flanked

by NcoI and BamHI restriction sites (synthesized by GeneArt, Regensburg,

Germany). To generate alanine-mutations, site-directed mutagenesis was per-

formed by overlap extension PCR [97] or whole plasmid mutagenesis [158]

as previously described, using a proofreading polymerase (Phusion, Bioké).

Mutated NcoI-BamHI digested γ9- or δ2-TCR chains were ligated into the ret-

roviral vector pBullet and sequenced by BaseClear c©

5.2.3 Retroviral transduction of T cells. γ9δ2TCRs were transduced into αβT-

cells as previously described [149]. In brief, packaging cells (phoenix-ampho)

were transfected with gag-pol (pHIT60), env (pCOLT-GALV) [209] and two ret-

roviral constructs (pBullet) containing either γ9-chain-IRES-neomycine or δ2-

chain-IRESpuromycine, using Fugene6 reagent (Takara, Gennevilliers, France).

Human PBMC activated with αCD3 (30 ng/ml) (Orthoclone OKT R©3, Janssen-

Cilag, Tilburg, The Netherlands) and IL-2 (50 IU/ml) (Proleukin R©, Novartis,

60


Arnhem, The Netherlands) were transduced twice with viral supernatant within

48 hours in the presence of 50 IU/ml IL2 and 4µg/ml polybrene (Sigma-

Aldrich, Zwijndrecht, The Netherlands). Transduced T-cells were expanded

by stimulation with αCD3/CD28 DynabeadsTM

(0.5× 106 beads/106 cells) (In-

vitrogen) and IL-2 (50 IU/ml) and selected with 800µg/ml geneticin (Gibco,

Karlsruhe, Germany) and 5µg/ml puromycin (Sigma-Aldrich, Zwijndrecht, The

Netherlands). Following selection, TCR-transduced T-cells were expanded in

vitro based on a previously described REP protocol [183].

5.2.4 Flow cytometry. γ9δ2TCR expression was analyzed by flow cytometry

using the following antibodies: Vγ9-PE (clone B3, BD), Vδ2-FITC (clone B6, BD),

pan-γδTCR-PE (clone IMMU510, Beckman Coulter), and γδTCR-APC (clone B1,

BD). Fold change was calculated based on MFI values of γ9- G115wt/δ2-G115wt

transduced T cells set to 1 and mock transduced T cells to 0. The expression

pattern of individual γ9δ2T-cell clones was analyzed using the following an-

tibodies: NKG2D-PE (clone 1D11, Biolegend), NKAT2-PE (clone DX27, BD),

CD158a-FITC (clone HD-3E4, BD) and NKB1-FITC (clone DX9, BD). Cell-cell

conjugate assays were performed with CFSE (Invitrogen)-labelled T cells and

Lavacell (Active Motif Europ)-labeled Daudi cells at room-temperature for 2

hours and analyzed by flow cytometry.

5.2.5 Functional T-cell assays. 51Chromium-release assay for cell-mediated

cytotoxicity was previously described [125, 226]. In brief, target cells were

labeled overnight with 100µCu 51Cr (150µCu for primary cells) and incubated

for 4− 6h with transduced T cells in five effector-to-target ratios (E:T) between

30:1 and 0.3:1. Fold change was calculated when compared to reactivity of en-

gineered T cells expressing unmutated γ9δ2TCR.

IFNγ ELISpot was performed using anti-hu IFNγ mAb1-D1K (I) and mAb7-B6-

1 (II) (Mabtech-Hamburg, Germany) following the manufacturer’s recommen-

ded procedure [23]. Target and effector cells (E:T 3:1) were incubated for 24h in

the presence of pamidronate (Calbiochem, Germany) where indicated.

IFNγ ELISA was performed using ELISA-ready-go! Kit (eBioscience) following

manufacturer’s instructions. Effector and target cells (E:T 1:1) were incubated

61

for 24h in the presence of pamidronate as indicated. Where specified, fold

change was calculated when compared to reactivity of engineered T-cells ex-

pressing unmutated γ9δ2TCR.

5.2.6 Animal models. The RAG-2-/-/γc-/--BALB/C mice, originally obtained

from AMCAS b.v. (Amsterdam, The Netherlands), were bred and housed in

the specific pathogen-free (SPF) breeding unit of the Central Animal Facility

of Utrecht University. Experiments were conducted according to Institutional

Guidelines after acquiring permission from the local Ethical Committee and in

accordance with current Dutch laws on Animal Experimentation. To induce tu-

mor xenografts, sublethal total body irradiated (2 Gy) RAG-2-/-/γc-/--BALB/C

mice (11− 17 weeks) were injected i.v. with 0.5× 106 Daudi-Luc cells (a kind

gift from Genmab Utrecht, The Netherlands) [26, 149]. Outgrowing tumors

were visualized in vivo by Biospace bioluminescent imaging. Mice were anes-

thetized by isoflurane before they received an intraperitoneal injection (100µl)

of 25mg/ml Beetle Luciferin (Promega). Bioluminescence images were acquired

and analyzed with M3Vision software (Photon Imager; Biospace Laboratory).

107 γ9δ2TCR positive transduced T-cells were intravenously injected together

with Daudi-Luc cells. Mice received 0.6× 106 IU of IL2 (Proleukin R©, Novartis)

in IFA s.c. on day 1 and every 21 days till the end of the experiment. Pamidro-

nate (10mg/kg body weight) was applied in the indicated groups at day 1 i.v.

and every 21 days i.p.

5.3 Results

5.3.1 Anti-tumor reactivity of individual γ9δ2T-cell clones. To investigate

whether individual γ9δ2T-cell clones mediate differential activity against tu-

mor cells compared to the parental γ9δ2T-cell population, γ9δ2T cells from a

healthy donor were cloned by limiting dilution and tested against a broad panel

of tumor cells in an IFN-γ ELISpot (Table 5.1). High variability in tumor re-

cognition was observed between individual γ9δ2T-cell clones (cl) in specificity

and functional avidity; compared to the original bulk population, cl5 and cl13

produced twice as many IFN-γ spots in response to the target Daudi and select-

ively generated significant amounts IFN-γ when challenged with K562, BT549

62


!" #9$2 #9$2 #9$2 #9$2 #9$2 #9$2 #9$2 #9$2T-cells bulk cl3 cl4 cl5 cl7 cl8 cl13 cl15

PBMCs 3 2 10 2 7 0 3 12 13

K562 0 14 9 9 62 181 7 56 12

Daudi 0 206 211 55 458 318 268 500 244

MZ1851RC 0 205 4 94 114 216 77 93 19

BT549 0 2 1 0 44 1 2 41 8

MCF-7 0 2 6 1 64 1 1 58 10

SW480 0 0 11 2 3 1 1 4 11

MDA MB 231 0 4 2 1 12 1 1 14 10

< 40 40 - 100 > 100IFN-! spots/15,000 cells:

Table 5.1: Anti-tumor reactivity of individual γ9δ2T-cell clones.

and MCF-7. In contrast, cl3 and cl15 recognized solely Daudi cells. A vari-

able expression of NK and KIR receptors has been reported in innate immune

cells 25-27 and might have contributed to the observed differential activity of

selected clones. Therefore, surface expression of γ9δ2TCR, NKG2D, CD158a,

NKAT-2 and NKB-1 was examined (Figure 5.6). However, no correlation was

found between receptor expression patterns and anti-tumor reactivity of tested

γ9δ2T-cell clones. We hypothesized that the diversity within the γ9δ2TCR con-

tributes to the differential activity of examined γ9δ2T-cell clones. Therefore,

γ9δ2TCRs of the highly tumor-reactive cl5 and the weakly tumor-reactive cl3

were chosen for detailed analysis and compared to γ9δ2TCR G115 [4, 149].

5.3.2 Anti-tumor-reactivity mediated by individual γ9δ2TCRs. To elucidate

differences among γ9δ2TCRs of tumor-reactive clones, sequences of wild type

(wt) γ9- and δ2-TCR chains of cl3 (γ9-cl3wt/δ2-cl3wt) and cl5 (γ9-cl5wt /δ2-

cl5wt) were determined and aligned with γ9δ2TCR G115. All three γ9δ2TCRs

only differed in their CDR3 domains: 1-3 amino acids between γ109 and γ111

63

in γCDR3 and 4-8 amino acids between δ108 and δ112 in δCDR3 (Table S1,

numbering according to IMGT [134]). To determine whether distinct γ9δ2TCRs

mediate differential anti-tumor reactivity, individual γ9δ2TCR chains were

cloned into the retroviral vector pBullet and linked to a selection marker as

described [231]. The wt-combinations γ9-cl3wt/δ2-cl3wt, γ9-cl5wt/δ2-cl5wt and

γ9-G115wt/δ2-G115wt were transduced into peripheral blood αβT-cells, selected

by antibiotics and further expanded. γ9δ2TCR G115 (γ9-G115wt/δ2-G115wt)

[4, 149] served as control, as did cells transduced with an empty vector cassette

(mock). First, the transductants were tested for expression of the introduced

γ9δ2TCR by flow cytometry using a pan-γδTCR antibody and comparable re-

ceptor expression was detected (data not shown). Next, γ9δ2TCR-transduced T-

cells were functionally tested against the tumor target Daudi in a 51Cr-release as-

say (Figure 5.1A). T cells expressing γ9-cl3wt/δ2-cl3wt had a 50 percent reduced

ability to lyse tumor cells (p < 0.01), whereas T cells with γ9-cl5wt/δ2-cl5wt

were nearly twice as potent (p < 0.01) as the control γ9-G115wt/δ2-G115wt.

To determine whether the phenotypes of γ9δ2TCR-transduced cells with de-

creased or increased functional avidity are also present on the cytokine level,

IFN-γ production was measured by ELISA. Daudi cells were then co-incubated

with pamidronate, leading to an increased production of the phosphoantigen

isopentenyl pyrophosphate (IPP) by the tumor cells. High IPP levels allowed

enhanced cytokine secretion of the transductants, which was measured in the

supernatant after 24h (Figure 5.1A). In line with changes observed for lytic ca-

pacity, IFN-γ secretion was 50 percent lower for T cells transduced with γ9-

cl3wt/δ2-cl3wt, and nearly twice as high for T-cells expressing γ9-cl5wt/2-cl5wt,

relative to control γ9-G115wt/δ2-G115wt. To assess whether or not this outcome

depends on the differential amount substrate of the mevalonate pathway (pre-

sumably IPP), the transductants were further tested in a pamidronate-titration

assay and the half maximal effective concentration (EC50) was calculated. Des-

pite different plateaus in IFN-γ secretion, all selected mutants and the wt con-

trol had a comparable pamidronate-EC50 (∼30pg/ml) (Figure 5.1B). This indic-

ates that distinct γ9δ2TCR clones mediate different functional avidity and the

high variability among parental γ9δ2T-cell clones in tumor recognition seems

to be substantially regulated by the CDR3 domains of individual γ9δ2T-cell re-

ceptors. Based on these results, the correlation between CDR3 domains and

functional avidity was investigated.

64


Figure 5.1: Anti-tumor reactivity mediated by γ9δ2TCRs. Peripheral blood T-cells werevirally transduced with indicated wt γ9δ2TCRs or CTE-engineered γ9δ2TCRs and (A,C) tested against Daudi in the absence (51Cr-release assay, E:T 3:1) or presence (IFN-γELISA, E:T 1:1) of 300µM pamidronate. Fold change was calculated when compared

65

to reactivity of γ9-G115wt/δ2-G115wt engineered T-cells. IFN-γ secretion of indicatedγ9δ2TCR engineered T-cells against Daudi was measured by ELISA (B, D) in the pres-ence of indicated amounts of pamidronate or (E) different E:T ratios. (F) Percentagesof cell-cell conjugates of Daudi and T-cells engineered with indicated γ9δ2TCR weredetermined by flow cytometry. Data represent the mean±SD. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001 by 1-way ANOVA.

5.3.3 Combinatorial-γδTCR-chain-exchange (CTE) as rapid method to mod-

ulate functional avidity of engineered T-cells. To make the above determina-

tion, we devised a strategy named combinatorial- γδTCR-chain-exchange (CTE),

which results in the expression of newly combined γ9- and δ2- TCR chains

on engineered T cells. During this process, γ9-G115wt was combined with

δ2-cl3wt or δ2-cl5wt and δ2-G115wt with γ9-cl3wt or γ9-cl5wt. These combin-

ations were retrovirally transduced into αβT-cells and equivalent expression

was detected by FACS using a pan-γδTCR antibody (data not shown). En-

gineered T cells were functionally tested against the tumor target Daudi in a51Cr-release assay and an IFN-γ ELISA (Figure 5.1C). The exchange of γ9- or

δ2-chains indeed caused notable differences. Compared to the original TCR

γ9- G115wt/δ2-G115wt, the combination of γ9-G115wt/δ2-cl3wt, γ9-G115wt/δ2-

cl5wt or γ9-cl5wt/δ2-G115wt mediated 50 to 70 percent increased specific lysis

of tumor cells (all p < 0.05) and a 70 to 100 percent increased IFNγ-secretion

(all p < 0.001). In contrast, combining γ9-cl3wt and δ2-G115wt resulted in a sig-

nificant reduction in both IFNγ-secretion and lytic activity (both p < 0.01). Ac-

cordingly, the same magnitude of recognition was observed when IFNγ produc-

tion of transductants was tested in a pamidronate titration assay (Figure 5.1D).

Moreover, only the combination γ9-cl3wt/δ2-G115wt led to decreased IFNγ pro-

duction of transduced cells at all pamidronate concentrations (max. 100pg/ml),

while all other CTE-γ9δ2TCRs mediated an increased IFNγ-secretion (max.

≥ 1000pg/ml) as compared to wt TCR γ9-G115wt/δ2-G115wt (max. 800pg/ml).

Equal pamidronate-EC50s of ∼ 30pg/ml were calculated for all responsive

γ9δ2TCR-transduced cells.

To determine whether cell-cell interaction influences the response-kinetics dif-

ferently than pamidronate stimulation, CTE-γ9δ2TCR γ9-G115wt/δ2-cl5wt and

control TCR γ9-G115wt/δ2-G115wt were tested in an effecter-to-target ratio (E:T)

titration assay (Figure 5.1E), and an E:T-50 was calculated. Interestingly, T cells

66


with γ9-G115wt/δ2-cl5wt responded differently with an E:T-50 of 1:1, compared

to an E:T-50 of 10:1 calculated for control cells expressing γ9-G115wt/δ2-G115wt.

To test whether the interaction between different TCRs and ligands—thus the

affinity—is indeed increased, cell-cell conjugates between Daudi and T-cells

expressing either potentially high (γ9-G115wt/δ2-cl5wt) or low (γ9-cl3wt/δ2-

G115wt) affinity TCRs were measured by flow cytometry. Significantly more

cell-cell interactions were observed when γ9-G115wt/δ2-cl5wt was expressed

as compared to γ9-cl3wt/δ2-G115wt and mock transduced T-cells (Figure 5.1F).

This effect did not depend on the presence of pamidronate (data not shown).

G115wt/δ2-cl5wt is therefore a high affinity γ9δ2TCR. Hence, CTE appears to

be an efficient method to rapidly engineer γ9δ2TCRs with increased affinity,

mediating improved functional avidity in transduced T-cells.

5.3.4 Residues in δCDR3 and Jδ1 are involved in γ9δ2TCR stability and in

mediating functional avidity of engineered αβT-cells. To elucidate the mo-

lecular requirements of δCDR3 to mediate optimal functional avidity, alanine-

mutagenesis of a model δCDR3 (clone G115) was performed including the

whole Jδ1 segment, as important residues have also been reported within Jγ1

[157]. During an initial screening, five sequence areas were found to either in-

fluence TCR expression or functional avidity of γ9δ2TCR transduced T cells

(data not shown). To clarify the degree to which single residues are responsible

for impaired γ9δ2TCR expression and lower TCR mediated functional avidity,

single alanine mutations were generated. The mutated and wt δ2-G115 chains

were expressed in combination with γ9-G115wt in αβT-cells and tested for

γ9δ2TCR expression using an antibody specific for the δ2-chain (Figure 5.2A).

Three single alanine mutations caused a 70 percent lower TCR expression when

compared to the unmutated δ2-G115wt, namely δ2-G115L116A, δ2-G115F118A and

δ2-G115V124A (Table 5.3). Comparable results were observed using antibodies

directed against the γ9-chain or the constant domain of the γδTCR (data not

shown), indicating the importance of δ2-G115L116, δ2-G115F118 and δ2-G115V124

for stable TCR expression. The crystal structure of γ9δ2TCR G115 supports our

findings: δ2-G115L116, δ2-G115F118 and δ2-G115V124 are located in hydrophobic

cores (Figure 5.2B) and could thus be crucial for the structural stability of the

γ9δ2TCR G115.

67

Figure 5.2: γ9δ2TCR expression and functional avidity of transduced T cells expressingsingle alanine mutated δ2 chain of clone G115. Peripheral blood T cells were virallytransduced with indicated γ9 and δ2 TCR chains and (A) analyzed by flow cytometryusing a δ2-specific antibody. Shown is the fold change in mean fluorescent intensity(MFI) in comparison to wt control expressing δ2-G115wt. (C) Lytic activity of trans-ductants was tested in a 51Cr-release assay against the tumor target Daudi (E:T 10:1).Specific lysis is indicated as fold change 51Cr-release measured in the supernatant after5h. Fold change was calculated as compared to reactivity of unmutated wt (δ2-G115wt).∗∗p < 0.01, ∗∗∗p < 0.001 by 1-way ANOVA. Arrows indicate mutations in δ2-G115 thatimpaired receptor expression (dotted arrows) or functional avidity (continuous arrows).(B, D) Crystal structure of γ9δ2TCR G115 indicating relevant amino acids (red arrows),δ chain (in blue), δCDR3 (in green), γ chain (in brown).

68


To address the impact of single alanine mutations on functional avidity, a 51Cr-

release assay was performed (Figure 5.2C). As expected, transductants with low

TCR expression (δ2-G115L116A, δ2-G115F118A and δ2-G115V124A) could not lyse

tumor cells effectively, as they demonstrated an 80 percent lower lytic capacity

when compared to cells transduced with δ2-G115wt. However, T cells with

mutation δ2-G115L109A and δ2-G115I117A (Table 5.3) properly expressed the TCR

but showed a 70 percent reduced lytic activity when compared to δ2- G115wt

expressing cells. Similar results were obtained when alanine substitutions δ2-

G115L109A and δ2-G115I117A were introduced into the δ2-chain of γδTCR clone

3 (data not shown). These results indicate that not only residue δL109 [33, 157,

232] but also δI117 in δCDR3 is generally important for γ9δ2TCRs to mediate

functional avidity (Figure 5.2D). However, sequence alignments between δ2-

chains of clones 3, 5 and G115 indicated that δL109 and δI117 are conserved

(Table 5.2), making it unlikely that these residues mediate different functional

avidities of the γ9δ2TCR transduced cells studied here.

5.3.5 Influence of CDR3 length on functional avidity of γ9δ2TCR transduced

T-cells. Surprisingly, alanine substitutions during alanine-scanning mutagen-

esis of γ9δ2TCR G115 could replace large parts of the δCDR3 domain without

functional consequences. That raises the possibility that the crucial factor for

the differing functional avidities of distinct γ9δ2TCR combinations is not a

defined amino acid, but the relative length between the functionally import-

ant residues δ2-G115L109 and the structurally important residue δ2-G115L116.

Therefore, different δ2-G115 length mutants were generated. Since the triple δ2-

G115T113−K115 is also important for stable surface expression (data not shown),

nine length mutants (δ2-G115LM) with 0 to 12 alanine between δ2-G115L109

and δ2-G115T113 were generated and equally expressed in αβT-cells, again in

combination with γ9-G115wt (Figure 5.3A). To test the functional avidity of δ2-

G115LM transduced T cells, an IFNγ ELISA in response to Daudi was per-

formed in the presence of pamidronate (Figure 5.3B). Interestingly, engineered

T cells expressing δ2-G115LM0 and δ2-G115LM1 were unable to produce IFNγ

and T-cells expressing δ-G115LM4 or δ-G115LM12 secreted only about half the

amount of IFNγ compared to δ2-G115wt transduced cells. All other mutants

(δ2-G115LM2,3,5,6,9) induced comparable amounts of IFNγ in engineered T cells

69

relative to transductants expressing δ2-G115wt. Mutants with functional impair-

ment (δ2-G115LM0,1,4,12, Table 5.3) were further tested against various amounts

of pamidronate and an EC50 was calculated. Despite different plateaus in max-

imal IFNγ secretion, all selected δ2-G115LM transduced cells and the wt control

had a comparable pamidronate-EC50 (∼30pg/ml) (Figure 5.3C). Length muta-

tions were also studied in γCDR3 of γ9δ2TCR G115 by engineering stretches of

1-6 alanines between γ9-G115E108 and γ9-G115E111.1 (γ9-G115LM1−6). However,

this did not affect functional avidity (Figure 5.8A).

This indicates that considerable alanine stretches within γ9 and δ2CDR3 do-

mains can be tolerated, likely because CDR3 regions are relatively exposed parts

of the TCR (Figure 5.3F). However, too short and very long alanine stretches

between δ2-G115L109 and δ2-G115T113 in particular, as well as stretches with

four alanines, are associated with reduced or absent function of a γ9δ2TCR

(Figure 5.3B and C). Loss of binding in mutants with short alanine-stretches is

most likely because the middle segment of δCDR3 is crucial for binding to the

ligand. That suggests the existence of an optimal δCDR3 length for γ9δ2TCRs.

Therefore, the CDR3 length within the γ9δ2TCR repertoire was studied.

5.3.6 Consequences for the physiological γ9δ2T-cell repertoire. The ImMun-

oGeneTics (IMGT) database [134] was searched for reported stretches between

γ9-G115E109 and γ9-G115E111.1 as well as δ2-G115L109 and δ2-G115T113. A prefer-

ential length for reported γ9-chains was found for CDR3 regions corresponding

to γ9-G115LM2 and γ9-G115LM3, but shorter stretches were also reported (Fig-

ure 5.8B). In contrast, δ2-chains with short δCDR3 domains, such as δ2-G115LM1

or δ2-G115LM0, were not reported (Figure 5.3D), in line with our observation

that such chains are not functional. The majority of listed γ9δ2TCRs contain

δCDR3 lengths which correspond to δ2-G115LM5,6,7. These findings support

the hypothesis that positive selection favors γ9δ2TCRs with an optimal δCDR3

length of 5-7 residues between δ2-G115L109 and δ2-G115T113. However, the indi-

vidual sequence might still play a role in γ9δ2TCR mediated functional avidity.

70


Figure 5.3: γ9δ2TCR expression and functional avidity of transduced T-cells expressingγ9δ2TCR G115 with δ2-CDR3 length mutations. (A) γ9δ2TCR expression of indicatedtransductants was analyzed by flow cytometry using a γδTCR-pan antibody. Shown isthe fold change in mean fluorescent intensity (MFI) in comparison to wt control express-ing δ2-G115wt. (B) IFN-γ secretion of δ2-G115LM transduced T cells against the tumor

71

target Daudi (E:T 1:1) was measured by ELISA after 24h incubation in the presence of100µM pamidronate. Shown is the fold change in IFN-γ production when comparedto reactivity of transductants expressing wt δ2-G115wt. (C) Transductants expressing δ2-G115LM0,1,4,12 were tested in a titration assay against the tumor target Daudi with increas-ing amounts of pamidronate as indicated. IFN-γ production was measured after 24h byELISA. (D) Generated δ2-G115LMs were matched in a BLAST search with γ9δ2TCRs de-scribed in the IMGT database. Shown is the number of citations compared to δ2-G115LM

of similar δCDR3 length. (E) Transductants with δ2-G115LM2,4,6 were compared side-by-side to transductants expressing individual γ9δ2TCRs of the same δCDR3 length. IFN-γsecretion of transduced T-cells against the tumor target Daudi (E:T 1:1) was measuredby ELISA after 24h in the presence of 100µM pamidronate. Shown is the fold changein IFN-γ production compared to reactivity of transductants expressing wt δ2-G115wt.Data represent the mean±SD. ∗∗p < 0.01, ∗∗∗p < 0.001 by 1-way ANOVA. (F) Crystalstructure of γ9δ2TCR G115; the region that was used for alanine stretches within δCDR3is shown in white, residual δCDR3 in green, δ chain in blue, γ chain in brown.

5.3.7 Influence of the CDR3 sequence on γ9δ2TCR mediated functional avid-

ity. To test the hypothesis that both the length and sequence of δCDR3 can

be important to mediate optimal functional avidity, γ9δ2TCR length mutants

δ2-G115LM2, δ2-G115LM4, and δ2-G115LM6 were transduced into T-cells in com-

bination with γ9-G115wt. IFNγ-secretion of transductants in response to Daudi

was compared to cells transduced with wt sequences from δ2-cl3wt (corresponds

in length to δ2-G115LM2), δ2-cl5wt (corresponds in length to δ2-G115LM4), and

δ2-G115wt (corresponds in length to δ2-G115LM6) (Table 5.4). Although T-cells

transduced with δ2-G115LM6 and δ2-G115wt did not differ in the amount of

cytokine secretion, all other combinations of wt chains showed a more than

two-fold increase in IFNγ when compared to the length mutant that select-

ively contained alanines (Figure 5.3E). These results were confirmed when the

lytic capacity of transduced cells was tested (data not shown). The sequence

in δCDR3 is therefore also a significant factor for the optimal functioning of a

γ9δ2TCR.

Accordingly, the sequential importance of γCDR3 was studied. Thereby, γ9-

G115LM1−3 were transduced into T-cells in combination with δ2-G115wt. IFNγ-

secretion of transductants in response to Daudi was compared to cells trans-

duced with γ9-cl3wt (corresponds in length to γ9-G115LM1), γ9-cl5wt (corres-

ponds in length to γ9-G115LM2) and γ9-G115wt (corresponding to γ9-G115LM3)

72


Figure 5.4: Functional avidity of transduced T-cells expressing γ9δ2TCR G115 withγ9-CDR3 length mutations. (A) Peripheral blood T-cells were virally transduced withindicated γ9 and δ2 TCR chains. Lytic activity of transductants was compared side-by-side to T-cells expressing individual γ9δ2TCRs of the same γ9CDR3 length. Specificlysis is indicated as fold change 51Cr-release measured in the supernatant after 5h. Datarepresent the mean±SD. ∗∗p < 0.01 by 1-way ANOVA. (B) Crystal structure of γ9δ2TCRG115 indicating γ9CDR3 in gray including amino acids γ9-G115A109, γ9-G115Q110 andγ9-G115Q111 (red arrows), δCDR3 is shown in green; δ chain in blue; γ chain in brown.

(Table 5.4). T cells expressing γ9-cl3wt/δ2-G115wt selectively produced lower

amounts of IFN-γ when compared to their wt equivalent γ9-G115LM1 (Fig-

ure 5.4A). Previously, the same γ9δ2TCR combination was also found to medi-

ate reduced functional avidity (Figure 5.1C and D). Interestingly, loss of activity

could be restored to normal levels (referred to γ9δ2TCR G115wt) by mutating

γCDR3E109 in γ9-cl3wt to γCDR3A109, which demonstrates that a single change

in the variable sequence of γ9CDR3 is sufficient to regulate functional avidity

of γ9δ2TCR transduced T cells tested here.

In summary, the length and sequence of the δ2CDR3 domain between L109

and T113 play a crucial role in γ9δ2TCR mediated functional avidity. In ad-

dition, the individual sequence between γ9CDR3E108 and γ9CDR3E111.1 can

hamper the activity of a γ9δ2TCR, and in G115 γCDR3A109 is most likely cru-

cial for ligand interaction (Figure 5.4B). This provides not only the rationale

for CTE-engineered γ9δ2TCRs but also for random mutagenesis within both γ9

and δ2CDR3 regions.

73

Figure 5.5: Anti-tumor-reactivity of T cells transduced with CTE-engineered γ9δ2TCRs.(A) Peripheral blood T cells were virally transduced with indicated γ9δ2TCRs and testedagainst indicated tumor cell lines and healthy control tissue. Transductants were in-cubated with target cells (E:T 1:1) in the presence of 10µM pamidronate. IFN-γ pro-duction was measured after 24h by ELISA. (B) Peripheral blood T cells were virallytransduced with clinical grade vector MP71 containing indicated γ9δ2TCR and tested

74


against primary AML blasts and healthy progenitor cells in an IFN-γ ELISpot assay (E:T3:1) in the presence of 10µM pamidronate. Data represent the mean±SD. ∗p < 0.05,∗∗p < 0.01, ∗∗∗p < 0.001 by 1-way ANOVA. (C, D) The functional avidity of T-cells ex-pressing CTE-γ9δ2TCR γ9-G115wt/δ2-cl5wt or control γ9δ2TCR (γ9-G115wt/δ2-G115wt)was studied in Rag2-/-γc-/- double knockout mice. After total body irradiation (2Gy)on day 0, mice were intravenously injected with 0.5× 106 Daudi-luciferase cells and 107

CTE-γ9δ2TCR transduced T-cells at day 1. Additionally, 6 × 105 IU IL-2 in IFA andpamidronate (10mg/kg body weight) were injected at day 1 and every 3 weeks untilthe end of the experiment. (C) Tumor outgrowth was assessed by in vivo biolumines-cence imaging (BLI) by measuring the entire area of mice on both sides. Data representthe mean of 4 animals measured. ∗∗p < 0.01 by 1-way ANOVA (day 42). (D) Overallsurvival of treated mice was monitored for 72 days.

5.3.8 From bench to bedside. CTE-engineered γ9δ2TCRs with increased activ-

ity against tumor cells are interesting candidates for TCR-gene therapeutic

strategies. This leads to the question of whether changes in functional avid-

ity mediated by CTE-γ9δ2TCRs constitute a unique phenomenon of a defined

γ9δ2TCR pair in response to the tumor cell line Daudi, or if this is a general

response to most tumor targets. Therefore, CTE-γ9δ2TCRs that mediated in-

creased (γ9- G115wt/δ2-cl5wt) or reduced (γ9-cl3wt/δ2-G115wt) activity were

tested against various tumors in an IFNγ ELISA in the presence of pharma-

cological concentrations of pamidronate (10µM) (Figure 5.5A) [149]. In line

with the observed response to Daudi, tumor reactivity was significantly in-

creased for K562 and SW480 when taking advantage of γ9-G115wt/δ2-cl5wt,

as compared to γ9-G115wt/δ2-G115wt, but was significantly reduced or even

absent for all other targets using γ9-cl3wt/δ2-G115wt. Furthermore, transduct-

ants expressing γ9-G115wt/δ2-cl5wt also recognized the tumor cell lines Saos-2,

MZ1851RC, SCC9, MDA-MB231, Fadu and MCF-7, which were not recognized

by control T-cells expressing γ9-G115wt/δ2-G115wt. Moreover, CTE-engineered

T-cells with increased activity against tumor cells still did not show any re-

activity towards healthy tissue such as PBMCs and fibroblasts. Therefore, CTE-

engineered γ9δ2TCRs can provide higher tumor reactivity against a broad panel

of tumor cells while not affecting normal tissue, and thus have the potential to

increase efficacy of TCR-engineered T cells.

To assess the potential clinical impact of CTE-engineered γ9δ2TCRs, we tested

whether an increased efficacy of CTE-γ9δ2TCRs is also present when vectors

75

for clinical applications are used and primary blasts of patient cells are chosen

targets. Therefore, γ9- G115wt was linked via an E2F element to either δ2-

G115wt or to δ2-cl5wt (CTE-γ9δ2TCR with highest efficacy) and cloned into

the clinical grade vector MP71. Transduction efficacy was comparable (∼30%),

and CTE-γ9δ2TCR transduced T cells were tested against primary AML blasts

and healthy CD34 progenitor cells in an IFNγ-ELISpot (Figure 5.5B). Transduct-

ants expressing γ9-G115wt/δ2-cl5wt recognized AML blasts of patient 1 (AML1)

equally, and were superior in recognizing AML blasts of patient 2 (AML2) when

compared to cells with γ9-G115wt/δ2-G115wt. Furthermore, CD34 progenitor

cells were not recognized by T-cells expressing either γ9-G115wt/δ2-cl5wt or

γ9-G115wt/δ2-G115wt. In light of these findings, CTE-engineered TCR γ9-

G115wt/δ2-cl5wt appears to be a promising candidate for clinical application.

Finally, to demonstrate that CTE-γ9δ2TCRs are safe and function with increased

efficacy when compared to the original constructs in vivo, adoptive transfer of

either CTE-(γ9-G115wt/δ2-cl5wt) or wt-engineered T cells was studied in a hu-

manized mouse model: protection against outgrowth of Daudi-luc in Rag2-/-

γc-/- double knockout mice, where the frequency of T-cell infusion was re-

duced relative to our previously reported model [149]. This resulted in loss

of protection with wt-engineered T-cells when tumor growth was measured

by bioluminescence imaging (BLI) or overall survival was assessed (Figure 5.5C

and D). However, CTE-engineered T cells expressing γ9-G115wt/δ2-cl5wt clearly

reduced tumor outgrowth in vivo (20, 000 counts/min, day 42) as compared to

wt-engineered T cells (180, 000 counts/min, day 42). Furthermore, only mice

treated with CTE-engineered T cells had an increased overall survival of ∼2

months relative to mice treated with T cells expressing γ9-G115wt/δ2-G115wt.

These results indicate that CTE-engineered γ9δ2TCRs efficiently mediate anti-

tumor reactivity in vivo, which points to CTE as a potential tool to optimize

γ9δ2TCRs for clinical application.

5.4 Discussion

γ9δ2T cells are innate lymphocytes that provide strong anti-tumor reactivity

against solid and hematological cancers [46, 152]. However, despite posit-

ive preclinical data, adoptive transfer of γ9δ2T cells in clinical studies has

76


provided only limited tumor control [167]. In our study, we determined one

factor preventing the successful translation of this strategy into humans: the

strong tumor-reactive potential of γ9δ2T-cells is not a universal feature among

all γ9δ2T-cells. Individual γ9δ2T-cell clones differ in their anti-tumor reactiv-

ity in specificity and functional avidity. The latter is substantially regulated by

CDR3 domains of individual γ9δ2TCRs. Most likely, single amino acid sub-

stitutions in CDR3 impact the affinity of a γ9δ2TCR to its ligand. Finally, we

have provided a means to engineer immune cells that harbor γ9δ2TCRs with in-

creased anti-tumor reactivity in vitro and in vivo which are promising candidates

for clinical applications.

Limited evidence has been provided for an important role of the variable do-

mains of a γ9δ2TCRs in mediating function. The requirement for germline

encoded residues has been reported only within γCDR3 and a hydrophobic

residue at position δ109 within δCDR3 [33, 232]. Here, Jδ1 residue δI117

was found to be important for tumor recognition, which shows that another

germline encoded residue is mandatory in the δ2TCR chain. Mutation of

residue δ109 or δI117 partially abrogates γ9δ2TCR-mediated activation, indic-

ating that such residues may be crucial for a first binding of a γ9δ2TCR to its

target. However, we have shown that single amino acids in the highly variable

part of CDR3, such as γCDR3A109, can clearly impact functional avidity. In the

γ9δ2TCR G115, γCDR3A109 points back towards the γ-chain, while γCDR3Q110

and γCDR3Q111 point towards the back of the TCR, but without contacting

any δ-chain residues. In combination with our functional data, this modeling

suggests that γCDR3A109 in G115 is important for directly mediating ligand in-

teractions. However, the A/E and Q/E substitutions are fairly non-conservative

and could also indirectly alter the structure of the γCDR3 loop and the TCR

binding site through global conformational changes. Additionally, our data in-

dicate that diverse amino acid compositions in δ2CDR3 can impact functional

avidity; the combination of a defined γ9 and δ2 chain is particularly important.

We hypothesize that the highly variable parts of γ9 and δ2CDR3 complement

each other to form a structure or allow a conformational change that is favorable

or unfavorable for target recognition. Therefore, in contrast to previous reports

[239], δCDR3 alone does not correctly reflect the full interaction of a γ9δ2TCR

with its target. Receptor flexibility is apparently necessary, for example to ad-

77

just to a variable cell surface of an antigen that might be presented in different

ways [30], or to respond with different affinities, as recently demonstrated for

T22 reactive γδT cells with variable δCDR3-domains in mice [191]. This is also

in line with a two-step binding model reported for αβTCRs, which requires a

preliminary interaction and then adjustment [236]. However, these hypotheses

are limited by the fact that no direct interaction of a γ9δ2TCR with a ligand has

been reported, which prevents further visualization of the proposed receptor-

ligand-interactions.

Our data suggest that certain limitations in the γ9δ2TCR repertoire are me-

diated by the δCDR3 length. If δCDR3 becomes too short, γ9δ2TCRs are not

functional, and such receptors have not been reported within the human δ2TCR

repertoire. Although the here used IMGT database for human γ9δ2TCR rep-

ertoire is certainly not complete, it is plausible that alterations in δCDR3 in

particular can limit the positive selection of a γ9δ2T cell. The functionally tol-

erated variability in CDR3 sequences and lengths reported in our study sup-

port the observation that γ9δ2T-cell responses to phosphoantigen stimulation

do not further select for defined CDR3 sequences [65, 219]. In line, we ob-

served identical dose-response kinetics in pamidronate titration experiments,

although the magnitude of response differed significantly. This indicates that

γ9CDR3 and δ2CDR3 binding does not involve substrates directly regulated by

pamidronate, which supports the hypothesis that a secondary signal is present

[182] irrespective of the mevalonate pathway.

Interestingly, T cells engineered to express defined γ9δ2TCRs did reflect the

functional avidity of the original γ9δ2T-cell clone tested here. Therefore, differ-

ences in CDR3 domains of γ9δ2TCRs can be responsible for differential func-

tional avidities observed between individual γ9δ2T-cell clones. However, it is

still likely that the functionality of distinct γ9δ2T cells is orchestrated by differ-

ent stimulatory molecules, e.g. NKG2D [53], and co-stimulatory signals derived

from molecules that are additionally expressed on γ9δ2T cells [182]. The pleth-

ora of specificities and functional avidities of distinct γ9δ2T-cell clones must

therefore be taken into account when such cells are ex vivo expanded and ad-

optively transferred [82].

To generate tumor-reactive T-cells that mimic the reactivity of a γ9δ2T cell, we

78


proposed to engineer αβT-cells to express a defined γ9δ2TCR [149]. This al-

lows rapid engineering of tumor-reactive T cells that are not limited by HLA-

restrictions and are readily available for nearly any patient with any cancer.

Moreover, it allows a γδT-cell repertoire to be repleted. The functionality of this

repertoire is usually heavily impaired in cancer patients [234]. To further im-

prove functional avidity mediated by a TCR, laborious display strategies have

been used for αβTCRs [101]. Based on our observation that mainly the γ9

and δ2CDR3 domains are involved in mediating functional avidity, we have

proposed the concept of combinatorial-γδTCR-exchange (CTE) as an efficient

method to design γ9δ2TCRs that mediate broad and strong anti-tumor re-

sponse. The CTE-γ9δ2TCR transduced T-cells tested remained tumor specific

and did not respond to healthy tissue. This indicates that pairing distinct γ9

chains or δ2 chains only strengthens the response towards malignant cells in-

stead of altering specificity, which reduces the likelihood of unwanted specificit-

ies. Hence, CTE-engineering provides an elegant strategy to redirect T cells

more effectively against a broad range of tumor cells.

5.5 Acknowledgments

We appreciate the technical support by Anton Martens and helpful discussions

with Charles Frink, Jeanette Leusen, Bart Nagelkerken, and Tuna Mutis.

79

!9-chain

104 105 106 107 108 109 110 111 111.1 112.1 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

!9-cl3wt C A L W E E · · E L G K K I K V F G P G T K L I I T

!9-cl5wt C A L W E I · Q E L G K K I K V F G P G T K L I I T

!9-G115wt C A L W E A Q Q E L G K K I K V F G P G T K L I I T

"2-chain

104 105 106 107 108 109 110 111 111.1 112.2 112.1 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128

"2-cl3wt C A C D L L G · · · · Y T D K L I F G K G T R V T V E P

"2-cl5wt C A C D A L K R · · T D T D K L I F G K G T R V T V E P

"2-G115wt C A C D T L G M G G E Y T D K L I F G K G T R V T V E P

!CDR3 J!P

"CDR3 J"1

Table 5.2: CDR3 sequence alignment of γ9δ2TCR G115, clone 3 and clone 5.

104 105 106 107 108 109 110 111 111.1 112.2 112.1 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128

!2-G115wt C A C D T L G M G G E Y T D K L I F G K G T R V T V E P impairment

!2-G115L116A C A C D T L G M G G E Y T D K A I F G K G T R V T V E P stability

!2-G115F118A C A C D T L G M G G E Y T D K L I A G K G T R V T V E P stability

!2-G115V124A C A C D T L G M G G E Y T D K L I F G K G T R A T V E P stability

!2- G115 C A C D T L G M G G E Y · · · L I F G K G T R V T V E P stabilitydeletionT113-K115A

!2-G115L109A C A C D T A G M G G E Y T D K L I F G K G T R V T V E P avidity

!2-G115I117A C A C D T L G M G G E Y T D K L A F G K G T R V T V E P avidity

!2-G115LM0 C A C D T L 0 x A T D K L I F G K G T R V T V E P avidity




!CDR3 J!1

single alanine mutants and deletions

alanine length mutants

Table 5.3: Alanine mutations in δCDR3 and the Jδ1 segment with impact on receptorstability or functional avidity.

80


Figure 5.6: Expression pattern of individual γ9δ2T-cell clones. Surface expression ofA) γδTCR B) NKG2D C) CD158a D) NKAT-2 and E) NKB-1 on individual γ9δ2 T-cellclones was analyzed by flow cytometry using the following antibodies: pan-γδTCR-PE (clone IMMU510), NKG2D-PE (clone 1D11), NKAT2-PE (clone DX27), CD158a-FITC(clone HD-3E4) and NKB1-FITC (clone HP-3E4).

81

!9-chain J!P

111 112104 105 106 107 108 109 110 111 .1 .2 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

!9-cl3wt C A L W E E · · E L G K K I K V F G P G T K L I I T!9-G115LM1 C A L W E A · · E L G K K I K V F G P G T K L I I T!9-cl5wt C A L W E I · Q E L G K K I K V F G P G T K L I I T!9-G115LM2 C A L W E A · A E L G K K I K V F G P G T K L I I T!9-G115wt C A L W E A Q Q E L G K K I K V F G P G T K L I I T!9-G115LM3 C A L W E A A A E L G K K I K V F G P G T K L I I T

"2-chain J"1

111 112 113104 105 106 107 108 109 110 111 .1 .2 .3 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128

"2-cl3wt C A C D L L G · · · · Y T D K L I F G K G T R V T V E P"2-G115LM2 C A C D T L A · · · · A T D K L I F G K G T R V T V E P"2-cl5wt C A C D A L K R · · T D T D K L I F G K G T R V T V E P"2-G115LM4 C A C D T L A A · · A A T D K L I F G K G T R V T V E P"2-G115wt C A C D T L G M G G E Y T D K L I F G K G T R V T V E P"2-G115LM6 C A C D T L A A A A A A T D K L I F G K G T R V T V E P

!CDR3

"CDR3

Table 5.4: CDR3 length mutants of γ9δ2TCR G115 and length matched CDR3 domainsof clone 3 and clone 5.

Figure 5.7: Functional avidity of γ9δ2TCR transduced T cells with different γCDR3length. γ9δ2TCRs with γCDR3 length mutations (γ9-G115LM) or unmutated wt (γ9-G115wt) were virally transduced in peripheral blood T cells. (A) Lytic activity of trans-ductants was tested in a 51Cr-release assay against the tumor target Daudi (E:T 10:1).Specific lysis is indicated as fold change 51Cr-release measured in the supernatant after5h. Fold change was calculated as compared to reactivity of unmutated wt (δ2-G115wt).(B) Generated γ9-G115LM were matched in a BLAST search with γ9δ2TCRs describedin the IMGT database. Shown is the number of citations compared to γ9-G115LM of thesame length.

82


8

8

A)

B)

LCL 6

LCL 8

LCL 1

2

LCL 2

0

LCL 2

2

LCL 3

6

LCL 3

7

LCL 4

8

LCL 6

1

LCL 6

9

LCL 7

0

LCL 8

3

LCL 9

3

LCL 9

4

LCL 9

6

0

50

100

150

200

250

mock

9-G115wt/!2-G115wt 9-cl3wt/!2-G115wt

******

*** ***

*** *****

***

***

***

***

*

***

*

**

***

********* ***

******

*** ******

**** ***

***

sp

ots

/15000 c

ells

Figure S3: !TCR downmodulation and reduced alloreactivity of CTE-engineered T-cells. A) 9!2TCRs

were virally transduced in peripheral blood T-cells and analyzed for "#TCR and !TCR expression, respectively,

using an "#TCR-pan and a !2-chain specific antibody. The mean fluorescence intensity (MFI) of "#TCR and

!TCR are indicated. B) Allo-reactivity of CTE-engineered T-cells were tested aginst EBV-LCLs in an IFN-

ELISpot assay. Only EBV-LCLs which elicited allo-reactivity against mock are shown. Data represent the

mean±SD as compared to mock transduced cells. *p<0.05, **p<0.01, ***p<0.001 by 1-way ANOVA.

Figure 5.8: αβTCR downmodulation and reduced alloreactivity of CTE-engineeredT cells. (A) γ9δ2TCRs were virally transduced in peripheral blood T cells and analyzedfor αβTCR and γδTCR expression, respectively, using an αβTCR-pan and a δ2-chainspecific antibody. The mean fluorescence intensity (MFI) of αβTCR and γδTCR are in-dicated. (B) Allo-reactivity of CTE-engineered T cells were tested aginst EBV-LCLs inan IFN-γ ELISpot assay. Only EBV-LCLs which elicited allo-reactivity against mock areshown. Data represent the mean±SD as compared to mock transduced cells. ∗p < 0.05,∗∗p < 0.01, ∗∗∗p < 0.001 by 1-way ANOVA.

83

84

Chapter 6

γδT-cells elicited byCMV-reactivation after allo-SCTcross-recognize CMV and leukemia

Wouter Scheper1,∗, Suzanne van Dorp

1,∗, Sabina Kersting1,

Floor Pietersma1, Caroline Lindemans

2, Samantha Hol1,

Sabine Heijhuurs1, Zsolt Sebestyen

1, Cordula Gründer1,

Victoria Marcu-Malina1, Arnaud Marchant

3,

Catherine Donner4, Bodo Plachter

5, David Vermijlen3,6,

Debbie van Baarle1,7

and Jürgen Kuball1

Leukemia, 2013

1Department of Hematology and Immunology, University Medical Center Utrecht, the Nether-lands

2Pediatric Blood and Marrow Transplantation Program, University Medical Center Utrecht, theNetherlands

3Institute for Medical Immunology , Université Libre de Bruxelles, Belgium4Department of Obstetrics and Gynecology, Hôpital Erasme, Brussels, Belgium5Institute for Virology, University Medical Center of the Johannes Gutenberg-University, Mainz,Germany

6Institute of Pharmacy, Université Libre de Bruxelles, Belgium7Department of Internal Medicine and Infectious Diseases, University Medical Center Utrecht, theNetherlands∗Both authors contributed equally to this work

85

Abstract

Human cytomegalovirus (CMV) infections and relapse of disease remainmajor problems after allogeneic stem cell transplantation (allo-SCT), inparticular in combination with CMV-negative donors or cordblood trans-plantations. Recent data suggest a paradoxical association between CMV-reactivation after allo-SCT and reduced leukemic relapse. Given the po-tential of Vδ2-negative γδT-cells to recognize CMV-infected cells and tu-mor cells, the molecular biology of distinct γδT-cell-subsets expanding dur-ing CMV-reactivation after allo-SCT was investigated. Vδ2neg γδT cellexpansions after CMV-reactivation were observed not only with conven-tional but also cordblood donors. Expanded γδT-cells were capable ofrecognizing both CMV-infected cells as well as primary leukemic blasts.CMV- and leukemia-reactivity were restricted to the same clonal population,whereas other Vδ2neg T-cells interact with dendritic cells (DCs). Cloned Vδ1-TCRs mediated leukemia-reactivity and DC-interactions, but surprisinglynot CMV-reactivity. Interestingly, CD8αα expression appeared to be a sig-nature of γδT-cells after CMV exposure. However, functionally CD8αα wasprimarily important in combination with selected leukemia-reactive Vδ1-TCRs, demonstrating for the first time a co-stimulatory role of CD8αα fordistinct γδTCRs. Based on these observations, we advocate the explora-tion of adoptive transfer of unmodified Vδ2neg γδT cells after allo-SCT totackle CMV-reactivation and residual leukemic blasts, as well as applicationof leukemia-reactive Vδ1-TCR-engineered T cells as alternative therapeutictools.

6.1 Introduction

Human cytomegalovirus (CMV) is a widely prevalent herpesvirus that, after

primary infection, persists life-long in the human host. Although infections are

asymptomatic in most immunocompetent individuals, reactivation of the virus

in immunocompromised patients after allogeneic stem cell transplantation (allo-

SCT) can lead to life-threatening complications including colitis and pneumonia

[27]. Moreover, CMV infection is associated with increased risk of acute graft-

versus-host disease (GVHD) [63, 213] and predisposes to secondary infections

due to CMV-induced immunesuppression [166]. Paradoxically, recent evidence

shows that CMV reactivation after allo-SCT reduces the risk of leukemic relapse

[18, 64], suggesting an unexpected favorable association between CMV infection

86

CHAPTER 6. CMV- AND LEUKEMIA-REACTIVE γδT CELLS AFTER ALLO-SCT

and clearance of tumor.

Multiple cell populations have been reported to be involved in clearance of

CMV infection. A vast body of information has been gathered for CMV-specific

αβT cells and NK cells [49]. For NK cells it has been hypothesized that they

may cross-recognize CMV-infected cells and cancer cells by responding to CMV-

infected residual AML blasts [64], which may contain considerable CMV copy

numbers. An alternative population that might also contribute to a better con-

trol of leukemia after CMV-reactivation is represented by γδT cells. In recent

years numerous studies have established the importance of γδT cells, a minor

T-cell population in peripheral blood but prominently present at sites of CMV

replication such as epithelial tissues, in both anti-viral immunity and tumor-

surveillance [29]. Contrary to αβT cells, activation of γδT cells does not rely on

antigen-presentation by MHC but is instead mediated by pathogen-derived an-

tigens or self molecules that are upregulated on infected, transformed or other-

wise stressed cells. In adult peripheral blood the major γδT-cell subset expresses

T-cell receptors (TCRs) composed of Vδ2 and Vγ9 gene segments (therefore also

referred to as Vδ2pos γδT cells) and is activated by small, nonpeptidic phospho-

antigens of pathogen or self origin [42, 78]. In contrast, γδT cells that reside in

epithelial tissues express TCRs composed of mainly Vδ1 or Vδ3 chains paired

with diverse Vγ chains, and a proportion of these γδT cells (collectively called

Vδ2neg γδT cells) expresses CD8αα [32, 58].

The involvement of γδT cells in the immune response against CMV has been

established by studies in transplant patients as well as healthy individu-

als, showing that CMV infection associates with marked in vivo expansions

of specifically Vδ2neg γδT cells that are reactive against CMV-infected cells

[56, 118, 176, 178, 229]. Furthermore, expansion of Vδ2neg T cells upon CMV

infection was shown to correlate with clearance of the virus [128]. In addition to

the anti-CMV response, numerous studies have implicated γδT cells in tumor

host defense: γδT cells have been found to infiltrate tumors of diverse origin

in vivo [41, 68, 144] and both Vδ2neg and Vδ2pos subsets have been abundantly

shown to be cytotoxic to cancer cells in vitro [45, 78, 86, 198].

Taken together these reports have strongly established the importance of Vδ2neg

γδT cells in the immune response against CMV and in tumor-surveillance. In

87

the present study we therefore evaluated the potential anti-leukemia capacity

of γδT cells that expand upon CMV-reactivation in a population of patients

with hematological malignancies receiving allo-SCT from either conventional

or cordblood donors. We show that in this cohort CMV-reactivation after allo-

SCT associates with in vivo expansions of CMV-reactive Vδ2neg γδT cells. These

CMV-induced γδT cells are capable of cross-recognizing hematological cancers,

and thus may explain the favorable effect of CMV-reactivation on risk of leuk-

emic relapse. In addition, such cells can serve as tools either from third party

donors to tackle CMV infection and leukemia or by taking advantage of here-

identified receptors to redirect T cells against leukemia.

6.2 Methods

6.2.1 Cell lines and antibodies. (see Supplementary Methods)

6.2.2 Patients, allo-SCT and blood sampling. A cohort of 26 patients with vari-

ous hematological diseases (Supplementary Table 6.1), who received an allo-

SCT at the UMC Utrecht, from December 2005 until August 2008, was analyzed.

Allo-SCT was given as curative or as rescue treatment to patients younger than

70 years with available HLA-matched related or unrelated donors, or with cord-

blood grafts. Patients were treated according to clinical protocols approved by

the local ethics board and gave their informed consent. Outcome of allo-SCT

of these patients was retrospectively analyzed in terms of hematopoietic re-

covery, viral reactivations, acute and chronic GVHD and progression free and

overall survival [225, 226]. After allo-SCT, patients were weekly monitored for

3 months for CMV-reactivation by real-time automated CMV-DNA PCR using

a TaqMan R© probe. For patients with conventional stem cell donors, PBMCs

of these time points were isolated and stored in liquid nitrogen until phen-

otypic analysis or expansion. Blood samples of cordblood patients were col-

lected 50− 100 days after transplantation. Absolute counts of CD3+, γδTCR+

and Vδ2+ T-cells were determined using TRUcount tubes R© (BD), according to

the manufacturer’s protocol. PBMCs were stained for γδTCR, Vδ2, CD3, CD4,

CD8α, CD16, CD25, CD27, CD45RO, CD56, CD80, and HLA-DR. The cohort of

newborns with CMV infection has been described recently [229].

88


6.2.3 Expansion and isolation of γδT-cell lines. γδT cells were isolated and

expanded using a previously described REP-protocol [183] (see Supplementary

Methods).

6.2.4 Functional T cell and DC maturation assays. IFNγ-ELISPOT,

51Chromium-release and DC maturation assays were performed as previ-

ously described [125, 149] (see Supplementary Methods).

6.2.5 Cloning of γδTCRs and retroviral transduction of T cells. γδTCRs were

isolated and sequenced as described in Supplementary Methods. Clone TCRs,

Vγ9Vδ2-TCR clone G115 [4] and a HLA-A*0201-restricted WT1126-134-specific

αβTCR [123] were transduced into αβT cells as described [149, 183, 209] (see

Supplementary Methods).

6.2.6 Statistical analyses. Differences were analyzed using indicated statistical

tests in GraphPad Prism 5.0 for Windows (GraphPad Software Inc.).

6.3 Results

6.3.1 CMV-reactivation after allo-SCT associates with expansion of Vδ2neg

γδT cells in both CMV-positive and CMV-negative stem cell donors. In or-

der to test whether an increase in Vδ2neg γδT cells is observed during CMV-

reactivation after allo-SCT, blood samples of 26 patients with umbilical cord-

blood (n = 10) and adult stem cell donors (n = 16, Supplementary Table 6.1)

were collected after allo-SCT and analyzed by flow cytometry. Nine patients

(56%) with adult stem cell donors developed a CMV-reactivation within 3

months after allo-SCT. In agreement with previous work in the context of allo-

SCT [118] but also other transplantation settings [56, 178], CMV-reactivation

but not EBV-reactivation associated with a significant increase in absolute num-

bers of circulating donor γδT cells in patients with conventional adult stem

cell donors (Fig. 6.1A, left panel; Supplementary Table 6.2). Expression ana-

lysis of CD45RO and CD27 indicated significantly lower percentages of naïve

(CD45ROnegCD27pos) γδT cells in CMV-reactivating patients (Supplementary

89

Fig. 6.7A), suggesting expansion of effector cells in these patients. Also all

patients with cordblood grafts, thus CMV-naïve donors, that developed a CMV-

reactivation (n = 6) had significantly increased numbers of circulating γδT

cells when compared to time-matched non-CMV-reactivating patients (n = 4;

Fig. 6.1A, right panel), although due to logistic challenges no time-course evalu-

ation was possible in this cohort. In both patient populations the increase in γδT

cells was due to an increase in the Vδ2neg subset (Fig. 6.1B), while Vδ2pos γδT

cells did not significantly differ between CMV-reactivating and non-reactivating

patients (adult grafts: median 2.17 versus 2.39 cells/µl, P = 0.37; cordblood

grafts: median 4.38 versus 0.81, P = 0.38). No differences were observed in

total CD3+ T-cell numbers between CMV-reactivating and non-reactivating pa-

tients (adult grafts: median 437 versus 355 cells/µl, P = 0.80; cordblood grafts:

median 58 versus 97 cells/µl, P = 0.38). In order to assess whether the increase

in Vδ2neg γδT cells during CMV-reactivation was mainly driven by γδT cells

expressing a public Vγ8Vδ1-TCR, which has been reported to play a substantial

role in congenitally infected newborns [229], clonality of increased cell fractions

was analyzed by spectratyping. However, when analyzing clonality of Vδ1, Vδ2

and Vδ3 γδT cells no such enrichment was observed in selected patients (Sup-

plementary Fig. 6.8). Interestingly, the increase in γδT cells in CMV-reactivating

patients preceded the increase of αβT cells, as a significant difference in αβT

cells between CMV-reactivating and non-reactivating patients was not observed

until 3 months after allo-SCT (Supplementary Fig.6.7B).

Together, these data confirm in a rather small but apparently representative co-

hort that CMV-reactivation after transplantation of CMV-seropositive stem cell

grafts associates with a significant increase of donor Vδ2neg γδT cells. Notably,

Vδ2neg γδT cells were also elicited during CMV-reactivation when CMV-naïve

cordblood grafts were used.

6.3.2 Patient-derived Vδ2neg γδT cells specifically recognize CMV-infected and

transformed cells ex vivo. To functionally evaluate whether Vδ2neg γδT cells

that expanded in vivo upon CMV-reactivation could indeed contribute to an

anti-CMV immune response, γδT cells were isolated from CMV-reactivating

patients and analyzed ex vivo. Patient-derived bulk, Vδ2neg, and Vδ2pos γδT-cell

subsets were coincubated with CMV-infected fibroblasts and γδT-cell activation

90


Figure 6.1: Selective expansion of Vδ2neg γδT cells in patients with CMV-reactivationafter allo-SCT. (A) Blood samples of patients with conventional adult stem cell donorswere collected weekly after allo-SCT (left panel) or during CMV-reactivation in patientswith cordblood-derived grafts (right panel), and presence of γδT cells was analyzedby flow cytometry. In the left panel, median values of all patients are presented. Forpatients with conventional donors, most CMV-reactivations were observed in the secondand third month after transplantation. (B) Presence of Vδ2neg γδT cells was analyzed inpatients with conventional and cordblood donors by flow cytometry. In patients withconventional stem cell donors (left panel) Vδ2neg γδT cells were measured in the secondand third month after allo-SCT, in patients with cordblood grafts at the same timepointas in (A). In box plots, the line at the middle is the median, the box extends from the25th to 75th percentile, and the whiskers extend down to the lowest value and up to thehighest. Mann Whitney U test was used to assess differences between CMV-positive andCMV-negative patients, and significant differences are indicated (∗p < 0.05).

91

was measured by IFNγ-ELISPOT. γδT cells isolated from patients with conven-

tional stem cell grafts secreted significantly higher levels of IFNγ upon con-

tact with CMV-infected cells when compared to uninfected controls (Fig. 6.2A).

In line with previous studies [56, 118], CMV-reactivity of patient-derived γδT

cells was mediated exclusively by Vδ2neg γδT cells, but not Vδ2pos γδT cells

(Fig. 6.2A). Importantly, γδT cells isolated from cordblood patients produced

IFNγ in response to and were able to specifically lyse CMV-infected cells

(Fig. 6.2B).

Previously it has been reported that γδT cells that expand upon CMV-

reactivation are able to cross-recognize solid cancer cells [47, 90], however cross-

reactivity with leukemic cells has not been reported. Therefore, patient-derived

Vδ2neg and Vδ2pos γδT cells were coincubated with a variety of hematological

cancer cell lines and primary acute myeloid leukemia (AML) blasts. Indeed,

CMV-reactive Vδ2neg γδT cells were able to specifically recognize lymphoma

(Daudi), leukemia (BV173, K562 and KCL22), and myeloma (U266) cell lines,

and most importantly primary AML blasts (Fig. 6.2C). In contrast, Vδ2pos γδT

cells from selected patients responded to hematological cell lines but not to

primary AML samples, although reactivity could be induced after treating AML

cells with the bisphosphonate pamidronate, a compound that induces the ac-

cumulation of Vδ2pos γδT-cell-activating phosphoantigens in treated cells [149]

(data not shown).

In summary, CMV-reactivation after allo-SCT associates with an increase in mul-

tipotent Vδ2neg γδT-cell populations from both CMV-positive and naïve stem

cell donors that are able to recognize both CMV-infected cells and hematolo-

gical tumor cells.

6.3.3 CMV-reactive Vδ2neg γδT-cell clones cross-recognize leukemic cells, in-

cluding primary AML blasts. To investigate whether cross-reactivity of Vδ2neg

γδT cells to leukemic blasts and CMV-infected fibroblasts is restricted to dif-

ferent clonal populations, Vδ2neg γδT cells were cloned by limiting dilution.

All generated clones carried Vδ2pos γδTCRs and expressed the natural killer

receptor NKG2D (Fig. 6.3A and Supplementary Fig. 6.9). Two clones (B11 and

E1) heterogeneously expressed CD8αα. CMV-reactivity of generated clones was

92


subsequently analyzed by coincubation with either CMV-infected or uninfec-

ted fibroblasts. Two Vδ1pos γδT-cell clones (B11 and E1) responded to CMV-

infected fibroblasts by increased IFNγ production, while two other clones (D3

and E2) did not (Fig. 6.3B). None of the clones produced TNFα in response to

CMV-infected cells (data not shown). Next, CMV-reactive Vδ1pos γδT-cell clones

B11 and E1 were coincubated with the hematological tumor cell lines U266, T2

(T- and B-lymphoblastoid cell line), EBV-LCL (Epstein-Barr virus-transformed

lymphoblastoid cell line) or primary AML blasts. Both CMV-reactive clones dis-

played a potent IFNγ-response against all (clone B11) or most (clone E1) tested

tumor cell lines as well as primary AML samples (Fig. 6.3C) but not healthy

fibroblasts. However, leukemia-reactivity was not a feature of all isolated clones,

as clones D3 and E2 did not produce IFNγ or TNFα in response to leukemic cell

lines or blasts (data not shown). Together, these data suggest that here-isolated

CMV-reactive clones are able to cross-recognize hematological tumor cells.

6.3.4 The interaction of Vδ2neg γδT cells with DCs is clonally segregated from

CMV- and leukemia-reactivity and is mediated by individual γδTCRs. Because

isolated clones D3 and E2 did not show a cytokine response against CMV-

infected fibroblasts nor leukemic cells (data not shown), we hypothesized that

such clones elicited after CMV infection are involved in maturation of dendritic

cells (DCs) [136], and thereby may aid in mounting adaptive immune responses.

Therefore, monocyte-derived immature DCs were cultured alone or in the pres-

ence of Vδ1pos γδT-cell clones and expression of the maturation markers CD80

and CD86 on DCs was measured after 48 hours. Selectively in the presence of

Vδ1 γδT-cell clones D3 and E2, but not CMV- and leukemia-crossreactive clones

B11 and E1, a substantial and significant increase in CD80/CD86+ DCs was ob-

served compared to immature DCs alone (Fig. 6.3D), resembling the phenotype

of DCs matured by the classical maturation cocktail (prostaglandin E2, IL-1β,

IL-6 and TNFα) [108]. However, no detectable production of IL12p70 was in-

duced by Vδ1 γδT-cell clones (data not shown). Importantly, induction of this

characteristic mature phenotype of DCs by clones D3 and E2 was observed in

the absence of CMV-infected cells or CMV virions in cocultures, indicating that

maturation of DCs by Vδ2neg γδT-cell clones is independent of CMV-antigen.

93

Figure 6.2: Specific recognition of CMV-infected and leukemic cells by patient Vδ2neg

γδT cells. (A) γδT cells isolated from patients with conventional adult stem cell donorswere expanded and cultured ex vivo before MACS-sorting and use in functional analysis.Sorted Vδ2pos or Vδ2neg γδT cells were subsequently cocultured for 18 hours with CMV-infected or -uninfected human foreskin fibroblasts and γδT-cell activation was measuredby IFNγ-ELISPOT. Results from two representative patients are shown. (B) Left panel:γδT cells from patients with cordblood transplantations were tested for CMV-reactivity

94


as in (A). Unsorted γδT-cells isolated from these patients predominantly (up to 90%) con-sisted of Vδ2neg T-cells, as determined by flow cytometry. Right panel: Killing capacity ofγδT-cells from cordblood patients against CMV-infected fibroblasts was determined bycoincubating γδT-cells and CMV-infected fibroblasts in a 4− 6hr 51Chromium-releaseassay. Uninfected fibroblasts served as control. Data from three different patients areshown. (C) MACS-sorted Vδ2pos and Vδ2neg γδT-cells from the same patients as in(A) were used to test anti-tumor recognition. Vδ2pos or Vδ2neg γδT-cells were cocul-tured with indicated hematological cancer cell lines or primary leukemic blasts (at 3:1target:effector ratio) in IFNγ-ELISPOT. For both γδT-cell populations healthy unsortedT-cells served as negative control target. Error bars represent SEM. Student t test (A,B) or one-way ANOVA (C) was used to assess differences between γδT-cell responses(∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001).

To test whether induction of maturation markers on DCs by Vδ1pos T-cell clones

D3 and E2 is mediated by their γδTCRs, Vγ- (Vγ4, Vγ8 and Vγ9) and Vδ1-

chains of here-generated clones were sequenced (Supplementary Table 6.3) and

retrovirally transduced into αβT cells. In agreement with our previous data

on Vγ9Vδ2 TCRs [149], all clone-derived δ1-TCRs were efficiently expressed in

both CD4+ and CD8+ αβT cells and down-regulated endogenous αβTCRs (Sup-

plementary Fig. 6.10). The involvement of individual δ1-TCRs in the induction

of the mature-like phenotype of DCs was subsequently analyzed by incubat-

ing transduced T cells with immature monocyte-derived DCs. Selectively δ1

TCRs that were isolated from clones E2 and D3 but neither δ1-TCRs E1 and

B11 nor mock-transduced cells induced a marked (∼3.5 to 9-fold) upregulation

of CD80/CD86 on DCs (Fig. 6.3E) and increased TNFα secretion in culture su-

pernatants (Supplementary Fig. 6.11A). In addition, a higher mean expression

of CD40, CD83 and HLA-DR was detected on DCs after coincubation with E2-

TCR- and D3-TCR-transduced T cells, but not T cells expressing δ1 TCRs E1

and B11. This phenotype depended on both CD1c, a lipid-presenting molecule

previously reported to be involved in Vδ2neg γδT-cell-mediated DC maturation

[136], and TNFα (Supplementary Fig. 6.11B). As was observed in experiments

with original clones, DCs did not produce detectable levels of IL12p70 (data not

shown). Taken together, these data show that distinct clonal populations within

the expanded Vδ2neg γδT-cell subset are responsible for CMV- and leukemia-

reactivity and for interacting with DCs, and that the Vδ2neg γδT-cell-DC inter-

action involves defined δ1TCRs.

95

Figure 6.3: CMV-reactive Vδ2neg γδT-cell clones cross-recognize cancer cells, but donot interact with DCs. (A) Vδ2neg γδT-cell clones were generated by limiting dilutionand phenotyped by flow cytometry. (B) CMV-reactivity of generated clones was de-termined by incubating clones alone or in combination with CMV-infected or uninfec-

96


ted fibroblasts (at 3:1 target:effector ratio) for 18 hours in an IFNγ-ELISPOT assay. (C)CMV-reactive Vδ2neg γδT-cell clones E1 and B11 were cultured alone, with hematolo-gical cancer cell lines or with primary leukemic blasts for 18 hours and IFNγ-productionwas determined by ELISPOT. Healthy human fibroblasts served as negative control. (D)Vδ2neg γδT-cell clones were incubated with immature DCs for 48 hours and the percent-age CD80/CD86 double-positive DCs was measured by flow cytometry. (E) TCR γ- andδ-chains of original Vδ2neg T-cell clones were sequenced and retrovirally transduced intoαβT-cells. Transfer of DC-interacting capacities was tested by culturing mock-transducedT-cells or T-cells expressing clone-derived δ1-TCRs with immature DCs for 48 hours andmeasuring expression of maturation markers by flow cytometry. Error bars representSEM. Student t test (B) or one-way ANOVA (C, D) were applied to assess differencesbetween γδT-cell responses (∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001).

6.3.5 Cancer-reactivity, but not CMV-reactivity, is mediated by distinct

γδTCRs. To formally confirm that individual γδTCRs of Vδ1 clones B11 and

E1 mediate CMV-reactivity, as reported after in primo utero CMV-infection [229],

αβT cells transduced with clone-derived δ1TCRs and the previously reported

CMV-reactive Vγ8Vδ1TCR [229] were incubated with CMV-infected or uninfec-

ted fibroblasts. Surprisingly, only T cells expressing the public CMV-reactive

Vγ8Vδ1-TCR but neither here-cloned CMV-reactive nor non-reactive δ1-TCRs

produced IFNγ after contact with CMV-infected cells (Fig. 6.4A and data not

shown), suggesting that CMV-recognition by original clones B11 and E1 must

rely on alternative surface receptors.

In order to test the mechanism involved in tumor recognition, clones B11 and E1

were tested for expression of NKp30, a receptor recently reported to be involved

in anti-tumor reactivity by Vδ1pos γδT cells [45]. However, here-isolated clones

did not express NKp30 (Supplementary Fig. 6.9). Thus, an alternative mechan-

ism must mediate tumor-reactivity and could include the individual γδTCRs.

Therefore, δ1TCR- and mock-transduced αβT cells were cocultured with hem-

atological cancer cell lines or primary AML blasts and T-cell activation was

determined by IFNγ-ELISPOT. Selectively T cells transduced with δ1TCRs of

CMV- and cancer-reactive clones B11 and E1 but not mock-transduced T-cells

were able to recognize both hematological cancer cell lines and primary AML

cells, while healthy T cells were not recognized (Fig. 6.4B). Importantly, cancer-

reactivity of both δ1TCRs could be extended to solid cancers, since pharyn-

geal (Fadu) and breast cancer (MDA-MB231) cell lines also activated γδTCR-

97

Figure 6.4: Isolated δ1-TCRs transfer cancer-reactivity, but not CMV-reactivity, to αβT-cells. (A) αβT-cells transduced with empty vector, with a public CMV-reactive δ1-TCRor with clone-derived δ1-TCRs were incubated for 18 hours with CMV-infected or unin-fected foreskin fibroblasts and IFNγ-secretion was measured by ELISPOT. (B) αβT-cellstransduced with empty vector or with either the B11 or E1 δ1-TCR were cultured withprimary AML blasts and hematological and solid cancer cell lines in an IFNγ-ELISPOT.Healthy T-cells were used as negative control. Error bars represent SEM. Student ttest (A) or one-way ANOVA (B) were used and significant differences are indicated(∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001).

transduced T cells (Fig. 6.4B). T cells transduced with δ1TCRs D3 or E2 pro-

duced neither IFNγ nor TNFα against tested cancer cells (data not shown).

Thus, cancer-reactivity of selected Vδ1pos γδT-cell clones is mediated by their

respective δ1-TCRs and this reactivity can be transferred to previously non-

reactive αβT-cells.

6.3.6 CD8αα functions as critical coreceptor for selected tumor-reactive

δ1TCRs. δ1TCRs isolated from clones B11 and E1 can be suitable tools to redir-

ect αβT-cells against leukemias as reported for Vγ9Vδ2-TCR clone G115, which

reprograms both CD4+ helper and CD8+ cytotoxic αβT cells against a broad

98


panel of tumor cells [149]. Thus, we questioned whether also here-isolated

tumor-reactive Vδ1pos γδTCRs are able to redirect both subsets of αβT-cells,

CD4+ and CD8+, against cancer cells. CD4+ and CD8+ T-cells expressing

tumor-reactive δ1TCRs B11 or E1 were therefore separated and incubated with

T2 or Daudi target cell lines. CD4+ and CD8+ αβT cells transduced with δ1TCR

E1 produced similar levels of IFNγ in response to tumor target cells (Fig. 6.5A).

In sharp contrast, δ1TCR B11 was able to reprogram CD8+ but not CD4+ T

cells, even though the introduced B11 TCR was expressed at slightly higher

levels in CD4+ than in CD8+ T cells (Supplementary Fig. 6.10B), as reported

previously [124, 149]. This suggested that for full T-cell activation this γδTCR

requires a molecule present on CD8+ but not CD4+ αβT-cells, such as NKG2D

or CD8. To address this, CD8+ T-cells transduced with the B11 δ1TCR were

preincubated with blocking antibodies against CD8α, CD8β, or NKG2D and

subsequently coincubated with Daudi or T2 target cells. Blocking of NKG2D,

which is expressed on most CD8+ but not CD4+ αβT cells and can amplify

αβ- and γδT-cell responses [53, 85], had only an effect on target cell recogni-

tion when T-cells were transduced with a γ9δ2TCR as reported [149] (data not

shown). Strikingly, blocking CD8α but not CDβ resulted in a marked decrease

in IFNγ-secretion when compared to T-cells pretreated with control antibody

(Fig. 6.5B). Blocking capacity of CD8β antibody was confirmed by inhibiting

MHC class I-restricted αβT cells. CD8α-blocking on CD4+ T-cells expressing

the B11 or E1 δ1TCRs served as additional negative controls and did not influ-

ence T-cell responses. Thus, the CD8α but not CD8β domain is important in

the ligand interaction of the B11 δ1-TCR. These data indicate that depending

on the particular γδTCR, tumor-reactivity is mediated by CD8α-dependent and

-independent mechanisms, suggesting e.g. different affinities of here-cloned

TCRs to their ligands.

The original clone B11 expressed the CD8αα homodimer but not the CD8αβ het-

erodimer (Supplementary Fig. 6.9). To test whether CD8αα was also involved

in activation of the original B11 γδT-cell clone, clone B11 (CD8+) and clone E1

(CD8low) were cocultured with T2 target cells in the presence of CD8α- or CD8β-

blocking antibodies. Similar to the effect on T-cells transduced with the B11 δ1-

TCR, blocking of CD8α significantly inhibited IFNγ-production by the original

clone (Fig. 6.5C). However, the effect of CD8α-blocking was less pronounced

99

on the original clone compared to B11-transduced αβT cells, most likely due to

lower expression of CD8αα on the parental clone when compared to CD8αβ ex-

pression on transduced T-cells (data not shown). Again, blocking of CD8β did

not affect IFNγ-secretion, as expected based on the CD8αα-positive phenotype

of clone B11. As was observed in E1-transduced T-cells, CD8αβ-blocking did

not affect activation of clone E1 (Fig. 6.5C). To corroborate these observations,

additional CD8αα-positive Vδ1 T-cell clones were generated from a different

donor and the effect of CD8α-blocking on activation of clones was analyzed. Of

nine CD8αα+ clones tested, blocking CD8α but not CD8β inhibited activation of

one clone that reacted to the colorectal cancer cell line SW480 (clone FE11), as

measured by reduced IFNγ secretion (Fig. 6.5D). CD8α-blocking had no effect

on activation of the parental polyclonal Vδ2negCD8+ γδT-cell line of this donor

nor of two other donors (data not shown), suggesting that CD8-dependence of

defined γδT-cell clones is not a general phenomenon yet observed in a substan-

tial fraction (2 out of 10) of isolated clones.

CD8αα was in the majority of isolated clones not functionally involved in

tumor-reactivity, questioning whether CD8αα rather plays a general role in

CMV-reactivity. In order to assess whether an increase in CD8αα expression

on γδT cells might be linked to CMV-infection in vivo, the cohort of conven-

tional stem cell donors was analyzed for CD8 expression by flow cytometry.

Strikingly, CMV-reactivating patients had significantly more circulating CD8+

γδT cells compared to non-reactivating patients (Fig. 6.6A). This observation

was confirmed in a complementary cohort of congenitally CMV-infected new-

borns (Fig. 6.6B). In this cohort, CD8α expression on γδT cells associated with a

differentiated effector (CD27neg/lowCD28neg) [229] phenotype (Supplementary

Fig. 6.12). Microarray gene expression profiling revealed highly upregulated

expression of CD8α but not CD8β upon CMV-infection (Fig. 6.6C), and flow

cytometry on blood samples of infected individuals indeed showed that CMV-

associated expression of CD8 on γδT cells is preferentially of the αα homodimer

(Fig. 6.6D). Of note, CD8+ αβT cells sorted from the same CMV-infected new-

borns did not show increased expression of CD8α (Fig. 6.6D). To test whether

CD8αα plays a functional role in CMV-reactivity by Vδ2neg γδT cells, clones

B11 and E1 were coincubated with CMV-infected or uninfected fibroblasts in

the presence of CD8α-blocking antibody. However, blocking CD8αα inhibited

100


not only the specific recognition of CMV-infected cells but also the occasionally

observed background reactivity of clone B11 but not clone E1 against freshly

plated fibroblasts (data not shown), suggesting that CD8αα may interact rather

with a general stress-antigen than an antigen specific for CMV infection.

In summary, these data show that CD8αα expressed on human Vδ2neg γδT-

cells associates with CMV-infection in vivo and is able to function as a critical

costimulator on selected clones as well as on γδTCR-reprogrammed αβT cells

when coincubated with tumor cells.

Figure 6.5: CD8αα acts as a coreceptor for selected δ1TCRs. (A) CD4+ and CD8+ αβT-cells transduced with B11 or E1 δ1-TCRs were sorted and subsequently cocultured withT2 or Daudi cell lines in an IFNγ-ELISPOT. (B) CD4+ and CD8+ transduced αβT-cellswere coincubated with T2 target cells as in (A), but now in the presence of a control anti-body or blocking antibodies against CD8α or CD8β. αβT cells expressing a WT1126-134-specific αβTCR [123] that were coincubated with T2 cells pulsed with 10− 6 M WT1126-134 peptide served as positive control for CD8α- and CD8β-blocking. (C) Original clonesB11 and E1 were incubated with T2 target cells as in (B). (D) Clone FE11 was generatedby limiting dilution, phenotyped by flow cytometry (left panel), and coincubated withSW480 target cells as in (B). Error bars represent SEM. Student t test (A) or one-way AN-OVA (B, C, D) were used and significant differences are indicated (∗p < 0.05; ∗∗p < 0.01;∗∗∗p < 0.001).

101

Figure 6.6: CMV reactivation after allo-SCT and congenital CMV infection associatewith increased expression of CD8 on γδT cells. (A) The percentage of CD8+ γδT cells

102


of patients with conventional stem cell donors was measured in the second and thirdmonth after allo-SCT by flow cytometry. (B) Cord blood from fetuses congenitally infec-ted (n = 11) or not infected (n = 16) was collected at term delivery and the percentage ofγδT cells expressing CD8 was analyzed by flow cytometry. (C) Gene expression analysisof γδT cells derived from three CMV-infected newborns versus γδT cells derived fromthree CMV-uninfected newborns. MA plot of differentially expressed genes in γδT cellsupon CMV infection. M (log2 of fold change) reflects the differential expression of agene. Positive and negative values indicate genes which are up- and down-regulated,respectively, upon CMV infection. A (mean expression) reflects the overall expressionlevel of a gene. Note that a similar figure, with indication of other genes, has been pub-lished before [229]. The highly up-regulated expression of CD8α RNA is indicated. Thedown-regulation of CD28 RNA is indicated as well for comparison. (D) The majority ofCD8 on γδT cells of congenitally infected newborns is composed of the CD8αα homod-imer. Percentages of CD8+ γδT cells and CD8+ αβT cells expressing the CD8α+CD8β+

phenotype were determined by flow cytometry in cordblood samples from eight congen-itally infected newborns (left panel). Representative flow cytometry plots (right panel)illustrate the staining patterns of CD8α and CD8β on αβT cells and γδT cells. MannWhitney U test (A, B) and Student t test (D) were used and significant differences areindicated (∗p < 0.05; ∗∗∗p < 0.001).

6.4 Discussion

The contribution of Vδ2neg γδT cells to controlling CMV-infection has received

considerable attention in recent years, and it is now well-established that these

unconventional T-cells play important roles in the immune response to CMV-

infection [56, 90, 118, 176, 229]. Combined with their widely reported reactivity

towards a variety of (mainly solid) tumors [41, 86, 144], this has made Vδ2neg

γδT cells a promising cell population for immunotherapeutic application. In the

present study we demonstrate that Vδ2neg γδT cells that expand upon CMV-

reactivation after allo-SCT are capable of responding to both CMV-infected and

leukemic cells. Additionally, by demonstrating that tumor-reactivity of Vδ2neg

γδT cells can be transferred by γδTCR gene-transfer, and by identifying a novel

role for CD8αα in the antigen-restriction of γδTCRs, we provide a solid basis

for the therapeutic exploration of Vδ2neg γδT cells and their γδTCRs.

Our observation that the occurrence of a single event (i.e. CMV-infection) is

able to induce expansion of γδT-cell subsets with anti-CMV- and anti-leukemia-

reactivity, including reactivity against primary leukemic blasts, provides an al-

103

ternative explanation for recent unexpected findings of a reduced relapse rate

in patients with CMV-reactivation after allo-SCT [18, 64]. Furthermore, it is in

line with a report in kidney transplant patients demonstrating that expansion

of Vδ2neg γδT cells following CMV-infection associated with a reduced risk of

developing solid cancer post-transplantation [47]. γδT cells isolated from these

patients reacted against both CMV-infected cells and epithelial tumor cells in

vitro. Thus, although CMV-reactivation after allo-SCT is still associated with

substantial non-relapse-related mortality (e.g. GVHD, colitis, and secondary in-

fections), reactivation of the virus reduces the risk of mortality due to relapse

of leukemia, and we show that one possible link is a CMV-induced expan-

sion of leukemia-reactive γδT cells. This hypothesis is further substantiated by

clinical data demonstrating that increased numbers of γδT cells after allo-SCT

are associated with improved disease-free survival, without higher incidence

of GVHD [79].

Mechanistically, little is known about the requirements for γδT-cell activation,

and the identity of the molecules on CMV-infected and leukemic cells that

are recognized by here-generated γδT-cell clones so far remain elusive. Dual-

reactivity of Vδ2neg γδT-cell clones to CMV and solid cancer cells has been

reported and has led to the hypothesis that γδTCRs of dual-reactive cells recog-

nize shared antigens on CMV-infected and transformed cells [90, 235]. How-

ever, our γδTCR-gene transfer experiments show that cancer-reactivity, but not

CMV-reactivity is mediated by δ1TCRs isolated in this study, indicating that

alternative immune receptors may be responsible for CMV-reactivity of the

original clones or that the γδTCR is involved but depends on additional mo-

lecules not expressed on αβT cells. In line with this, it was recently reported

that the γδTCR isolated from a CMV-reactive Vγ4Vδ5 clone requires costim-

ulation by CD11a-CD18 (LFA-1) [235]. However, here-isolated CMV-reactive

clones E1 and B11 as well as αβT cells transduced with their respective γδTCRs

expressed high levels of CD11a (see Supplementary Fig. 6.13), suggesting that

other mechanisms must be involved. Alternatively, it was recently shown that

Vδ2neg γδT cells could be stimulated by IgG-opsonized CMV virions via the

IgG receptor CD16 (FcγRIIIa), independent of γδTCR-engagement [48]. How-

ever, here-isolated CMV-reactive clones did not express CD16 (see Supplement-

ary Fig. 6.9).

104


We report here for the first time that in human γδT cells CD8αα functions as

restriction element for target recognition by distinct δ1TCRs. Although long de-

scribed to be expressed on γδT cells [32, 58], the function of CD8αα on these cells

has so far remained unknown. In our experiments, blocking CD8α resulted in a

marked and significant inhibition of tumor recognition by different clones and

distinct δ1TCRs, which was not observed when tested on the bulk population.

These data put CD8αα into the field of coreceptors for δ1TCRs for a defined sub-

set of tumor-reactive γδT cells. Moreover, we report that CMV-infection associ-

ates with an increase in CD8αα-expressing γδT cells in both allo-SCT patients

as well as congenitally infected newborns, suggesting a link between CD8αα

and the immune response against CMV in vivo. However, the functional in-

volvement of CD8αα in γδT-cell-mediated CMV-reactivity remains to be further

defined. Within the αβT-cell compartment, CD8αα-positive T-cells are enriched

in mucosal tissues such as intestine and these cells are described to display a

characteristic innate-like phenotype [39]. However, on these cells CD8αα does

not function as a classical MHC class I-binding αβTCR coreceptor as CD8αβ

does, but more likely serves as suppressor of αβTCR-mediated T-cell activa-

tion [40]. A subset of NK cells also expresses CD8αα, and these cells posses

greater killing capacity than CD8αα-negative NK cells [208]. This effect was

attributed to enhanced resistance to apoptosis that was specifically mediated

through CD8αα-signaling [2]. Superior cytotoxicity of CD8αα-expressing NK

cells has been associated with clinical remission of leukemia patients [142, 143],

indicating that CD8αα on innate immune cells may be relevant to clinical out-

come after allo-SCT. Finally, CD8αα on murine innate-like intestinal αβT cells

was shown to enhance αβTCR-mediated T-cell activation by binding the non-

classical MHC-I molecule thymus leukemia (TL) [135]. Thus, the expression of

CD8αα on innate(-like) immune cells may indicate a universal role for CD8αα

as regulatory receptor in innate immune responses.

To tackle CMV-infections in immuno-compromised patients, several clinical

trials have focused on the adoptive transfer of CMV-reactive αβT cells [69,

133]. However, major obstacles are presented by the MHC-restricted antigen-

recognition of αβT cells and the challenge to generate sufficient numbers of

CMV-reactive αβT cells within the time constraints of severe infection [98]. Our

data suggest that Vδ2neg γδT cells are an interesting alternative source of CMV-

105

reactive T cells for such patients as we observe that in vivo generated Vδ2neg

γδT cells react against not only CMV-infected cells but also leukemic cells in

vitro. Moreover, we demonstrate that CMV-reactive γδT cells can also be ob-

tained from the naïve umbilical cordblood repertoire, underscoring the value of

this third-party stem cell source for application in allo-SCT, in particular also

for patients with CMV-negative donors. In summary, we advocate the explor-

ation of adoptive transfer of unmodified Vδ2neg γδT cells in CMV- and tumor-

immunotherapies and the application of leukemia-reactive Vδ1-TCR-engineered

T cells. Clinical trials will need to be pursued in order to test efficacy and safety

of the application of such strategies.

6.5 Acknowledgements

We thank the members of the stem cell facility at the UMC Utrecht for technical

assistance. We also thank Margreet Brouwer for her expert technical assistance.

6.6 Supplementary Methods

6.6.1 Antibodies and flow cytometry. Antibodies used for flow cytometry

included: γδTCR-APC (clone B1, BD), γδTCR-PE (clone IMMU510, Beck-

man Coulter), γδTCR-FITC (clone 11F2, BD), Vδ2-PE and -FITC (clone B6,

BD), Vδ1-FITC (clone R9.12, Beckman Coulter), αβTCR-PE-Cy5 (IP26A, Beck-

man Coulter), CD3-eFluor450 (clone OKT3, eBioscience), CD3-pacific blue

(clone SP34-2, BD), CD4-PE-Cy7 (clone RPA-T4, BD), CD8α-APC (clone RPA-

T8, BD), CD8α-PE-Cy7 (clone SFCI21Thy2D3, Beckman Coulter), CD8β-PE

(clone 2ST8.5H7, BD), CD16-PE (clone CB16, eBioscience), CD27-APC-eFluor780

(clone 0323, eBioscience), CD27-APC (clone L128, BD), CD28-ECD (clone

CD28.2; Beckman Coulter), CD40-APC (clone HB14, Biolegend), CD45RO-

PE-Cy7 (clone UCHL1, BD), CD56-PE (clone B159, BD), CD80-PE (clone

L307.4, BD), CD83-FITC (clone HB15e, BD), CD86-PE-Cy5 (clone IT2.2, eBios-

cience), NKp30-APC (clone P30-15, Biolegend), NKG2D-APC (clone 1D11, BD),

CD158a(NKAT1)-FITC (clone HP-3E4, BD), CD158b(NKAT2)-PE (clone DX27,

BD), NKB1(NKAT3)-FITC (clone DX9, BD), HLADR-APC-Cy7 (clone L243, Bio-

legend). All allo-SCT samples were processed with FACSCanto-II or LSR-II flow

106


cytometers (BD) and analyzed with FACSDiva software (BD). Whole cord blood

samples derived from infected and uninfected newborns were run on the CyAn

flow cytometer and data were analyzed using Summit 4.3 (Dako).

6.6.2 Cell lines and primary acute myeloid leukemia cells. Daudi, K562,

KCL22, T2, BV173, SW480, MDA-MB231, U266, foreskin fibroblasts and

Phoenix-Ampho cell lines were obtained from ATCC. EBV-LCL was kindly

provided by Phil Greenberg (Seattle, WA). Fadu was kindly provided by Niels

Bovenschen (UMC Utrecht, The Netherlands). Fibroblasts and Phoenix-Ampho

cells were cultured in DMEM supplemented with 1% Pen/Strep (Invitrogen)

and 10% FCS (Bodinco), all other cell lines in RPMI with 1% Pen/Strep and

10% FCS. Fresh PBMCs were isolated by Ficoll-Paque (GE Healthcare) from

buffy coats supplied by Sanquin Blood Bank (Amsterdam, The Netherlands).

Where indicated, foreskin fibroblasts were infected with culture supernatants

of fibroblasts previously infected with human CMV strain AD169 at a multipli-

city of infection (MOI) of 2. After 24 hours, infected and uninfected fibroblasts

were washed before being used in functional assays. Frozen primary acute my-

eloid leukemia (AML) samples were a kind gift from Matthias Theobald (Mainz,

Germany) and were collected in compliance with GCP and Helsinki regulations.

6.6.3 Expansion and isolation of γδT-cell lines. PBMCs were stimulated for

14 days with 1µg/ml PHA-L (Sigma-Aldrich), 50U/ml IL-2 (Novartis Pharma),

5ng/ml IL-15 (R&D Systems), and irradiated allogeneic PBMCs, Daudi and

EBV-LCLs. Fresh IL-2 was added twice a week. After first expansion, poly-

clonal γδT-cell lines were obtained by MACS-isolation (TCRγδ+ T-cell isolation

kit, Miltenyi Biotec) with a purity of > 90% and were further expanded using

again the REP-protocol. Vδ2pos and Vδ2neg γδT-cell fractions were obtained by

MACS-depleting Vδ2pos γδT cells from bulk cultures using Vδ2TCR-PE anti-

body and anti-mouse IgG microbeads (Miltenyi Biotec). γδT cells isolated from

patients receiving cordblood grafts typically contained up to 90% Vδ2neg γδT

cells and were therefore not further MACS-sorted. Vδ2neg γδT-cell clones were

generated from a CMV-seropositive healthy donor by limiting dilution. All γδT-

cell cultures were stimulated biweekly using the REP-protocol.

107

6.6.4 Spectratyping and microarray experiments. Spectratyping analysis and

microarray experiments were performed as previously described [229]. Mi-

croarray data and procedures were deposited at Array Express (www.ebi.ac.

uk/arrayexpress) under accession no. E-MEXP-2055.

6.6.5 Dendritic cell maturation assay. Monocytes were isolated from PBMCs

by plate adhesion and differentiated into immature dendritic cells (iDCs) by cul-

turing for 4 days in AIM-V medium in the presence of 500U/ml IL-4 (Peprotech)

and 800U/ml GM-CSF (Peprotech). Next, iDCs were cocultured with T-cells at

a ratio of 1:1 for 48 hours and expression of CD40, CD80, CD83, CD86 and HLA-

DR was measured by flow cytometry. Where indicated, CD1c-blocking antibody

(clone L161, Biolegend), TNFα-blocking antibody (clone MAb1, eBioscience), or

control antibody was added to cultures at a concentration of 20µg/ml. Secretion

of TNFα and IL12p70 was measured by ELISA (eBioscience).

6.6.6 Functional T-cell assays. IFNγ-ELISPOT was performed by coculturing

15, 000 T-cells and 50, 000 target cells (ratio 0.3:1) for 24 hours in nitrocellulose-

bottomed 96-well plates (Millipore) precoated with anti-IFNγ antibody 1-D1K

(Mabtech). Plates were washed and incubated with biotinylated antibody 7-B6-

1 (Mabtech) followed by streptavidin-HRP (Mabtech). IFNγ spots were sub-

sequently visualized with TMB substrate (Sanquin) and spots were quantified

using ELISPOT Analysis Software (Aelvis). With regard to γδT-cell clones, re-

activity to CMV-infected cells and cancer cells was generally determined in

the same experiment. Where indicated, blocking of CD8α was performed us-

ing 10µg/ml anti-CD8α antibody clone OKT8 (eBioscience), blocking of CD8β

with 10µg/ml anti-CD8β clone 2ST8.5H7 (Abcam), and NKG2D-blocking with

10µg/ml anti-NKG2D clone 149810 (R&D Systems).

51Chromium-release assays was performed as described [125, 149]. Target cells

were labeled overnight with 150µCu 51Cr and subsequently incubated with γδT-

cells in four effector-to-target ratios (E:T) between 30:1 and 1:1. 51Cr-release in

supernatant was measured 4− 6hr later.

108

www.ebi.ac.uk/arrayexpress

www.ebi.ac.uk/arrayexpress


6.6.7 Cloning of γδTCR genes and retroviral transduction of T-cells.

mRNA of γδT-cell clones was isolated using the Nucleospin RNA-II kit

(Macherey-Nagel) and reverse-transcribed using SuperScript-II reverse tran-

scriptase (Invitrogen). TCRγ- and TCRδ-chains were amplified by PCR us-

ing Vδ1 (5’-gatcaagtgtggcccagaag-3’), Vγ2-5 (5’-ctgccagtcagaaatcttcc-

3’), Vγ8 (5’-gctgttggctctagctctg-3’) and Vγ9 (5’-tccttggggctctgtgtgt-

3’) sense primers, and Cδ (5’-ttcaccagacaagcgaca-3’) and Cγ (5’-

ggggaaacatctgcatca-3’) antisense primers. PCR products were sequenced

by Baseclear c© (Leiden, the Netherlands). Codon-optimized sequences of clone

TCRs were subsequently synthesized by Geneart R© (Regensburg, Germany) and

subcloned into pBullet.

Packaging cells (Phoenix-Ampho) were transfected with gag-pol (pHIT60),

env (pCOLT-GALV) [209] and pBullet constructs containing TCRγ-chain-IRES-

neomycine or TCRδ-chain-IRES-puromycin, using Fugene6 (Promega). PB-

MCs preactivated with αCD3 (30ng/ml) (clone OKT3, Janssen-Cilag) and IL-2

(50U/ml) were transduced twice with viral supernatant within 48 hours in the

presence of 50U/ml IL-2 and 4µg/ml polybrene (Sigma-Aldrich). Transduced

T-cells were expanded by stimulation with αCD3/CD28 Dynabeads (0.5× 106

beads/106 cells) (Invitrogen) and IL-2 (50U/ml) and selected with 800µg/ml

geneticin (Gibco) and 5µg/ml puromycin (Sigma) for one week. Where indic-

ated, CD4+ and CD8+ TCR-transduced T cells were separated by MACS-sorting

using CD4- and CD8-microbeads (Miltenyi Biotec). Following selection, TCR-

transduced T-cells were stimulated biweekly using the REP-protocol [183].

109

Patient groups

CMV-reactivation No CMV-reactivation

Conventional graft cohort

N 9 7

Median age (range) 56 (33-62) 49 (35-68)

Sex M/F (%) 89/11 57/43

Donor/recipient relation

RD 5 (56) 4 (57)

MUD 4 (44) 3 (43)

Diagnosis

AML 1 (11) 4 (57)

CLL 1 (11) 1 (14)

CML 1 (11) 0 (0)

MM 5 (56) 2 (29)

NHL 1 (11) 0 (0)

Conditioning

NMA 9 (100) 9 (100)

MA 0 (0) 0 (0)

ATG 5 (56) 3 (43)

GVHD 8 (89) 3 (43)

CMV+ Patient 8 (89) 4 (57)

CMV+ Donor 5 (56) 1 (14)

OS at 2 years 5 (56) 5 (71)

Cordblood graft cohort

N 6 4

Median age (range) 2 (1− 10) 2 (1− 15)

Sex M/F (%) 67/33 0/100

Diagnosis

AML 2 (33) 0 (0)

ALL 2 (33) 3 (75)

JMML 0 (0) 1 (25)

NMID 1 (17) 0 (0)

NMMD 1 (17) 0 (0)

Conditioning

NMA 0 (0) 0 (0)

MA 6 (100) 4 (100)

ATG 6 (100) 4 (100)

GVHD 2 (33) 1 (25)

CMV+ Patient 6 (100) 4 (100)

CMV+ Donor 0 (0) 0 (0)

OS at 2 years 5 (83) 3 (75)

Table 6.1: Patient characteristics. ALL, acute lymphocytic leukemia; AML, acute my-

110


eloid leukemia; ATG, antithymocyte globuline; CLL, chronic lymfocytic leukemia; CML,chronic myeloid leukemia; CMV, cytomegalovirus; F, female; GVHD, graft-versus-hostdisease; JMML, juvenile myelomonocytic leukemia; M, male; MA, myeloablative; MM,multiple myeloma; NHL, non-Hodgkins lymphoma; NMA, non-myeloablative; NMID,non-malignant immunodeficiency; NMMD, non-malignant metabolic disease; OS, over-all survival; RD, related donor; MUD, matched unrelated donor.

Patient groups

EBV-reactivation No EBV-reactivation p-value

Conventional graft cohort

N 6 10

% αβT-cells / lymphocytes 30.4 50 0.66

% γδT-cells / lymphocytes 1.2 1.2 0.81

% CD56posCD16pos cells / CD3neg lymphocytes 34.1 74.6 0.01

Table 6.2: Comparison of γδT cells, αβT cells and NK cells between patients with andwithout EBV-reactivation. EBV, Eppstein-Barr virus. p-values: Mann-Whitney U test.

Figure 6.7: Naïve γδT cells and total αβT cells after allo-SCT. (A) PBMCs of patientswith conventional adult stem cell donors were collected weekly after allo-SCT, and thepercentage of naïve CD27posCD45ROneg γδT cells was analyzed by flow cytometry. (B)Absolute counts of αβT cells after allo-SCT with conventional donors was measured byflow cytometry. A Mann Whitney U test was performed at all time points and significantdifferences are indicated (∗p < 0.05).

111

Figure 6.8: γδTCR clonality analysis of γδT cells from CMV-reactivating patients. Rep-resentative spectratype analyses of Vδ1, Vδ2 and Vδ3 γδTCR clonality in blood samplesof CMV-reactivating patients that received stem cells from conventional adult donors(A) or cordblood donors (B). All patients were analyzed during CMV-reactivation. TheCDR3δ1 size of 11 amino acids, corresponding with the CDR3δ1 size of the publicVγ8Vδ1 TCR [229], is indicated with arrows.

112


Figure 6.9: Phenotyping of Vδ2neg γδT-cell clones. Vδ2neg T-cell clones were generatedby limiting dilution and surface expression of indicated receptors was measured by flowcytometry. Gating was established based on appropriate isotype controls.

113

Figure 6.10: Efficient retroviral expression of δ1TCRs in CD4+ and CD8+ αβT cells. (A)Isolated δ1TCRs were retrovirally transduced into αβT cells and surface expression ofendogenous αβTCR and introduced γδTCR was determined by flow cytometry. Indic-ated in plots are percentages of quadrants and MFIs of γδTCR and αβTCR stainings. (B)Transduced αβT cells were costained for CD4 and expression levels (MFI) of γδTCRs onCD4− (i.e. CD8+) and CD4+ αβT cells is indicated in plots.

114


Figure 6.11: Upregulation of DC maturation markers by γδTCR-transduced T cellsinvolves TNFα and CD1c. (A) Immature DCs (iDCs) were cultured alone, with mock-transduced αβT cells, or with αβT cells expressing clone-derived γδTCRs for 48 hoursand TNFα levels in culture supernatants were measured by ELISA (one-way ANOVA:∗p < 0.05, ∗∗p < 0.01). (B) iDCs were cultured as in (A) but now in the presence ofcontrol antibody or blocking antibodies against CD1c or TNFα. After 48 hours CD83expression on DCs was measured as a representative marker of DC maturation.

115

Figure 6.12: CD8α expression is associated with a differentiated effector phenotype(CD27neg/lowCD28neg) of γδT cells in CMV-infected newborns. Association of CD8α

expression with CD27neg/low γδT cells (left panel) and CD28neg γδT cells (right panel).Stainings are representative for 11 CMV-infected newborns. Plots represent lymphocytesgated on CD3+γδTCR+ phenotype.

Figure 6.13: Expression of CD11a on original clones, γδTCR-transduced αβT cells andJurkat cells. Expression of CD11a is shown as a fold increase of MFI of the specificstaining over MFI of the staining with control antibody.

116


TC

Rγ

chai

ns

Clo

neV

nam

e3’

V-R

EGIO

NN

5’J-

REG

ION

Jna

me

B11

TRG

V4*

02TGTGCCACCTGGGATGG.

CCAGGAAGG

...TTATTATAAGAAACTCTTT

TR

GJ1

*01

D3

TRG

V8*

01TGTGCCACCTGG......

TCCAGGGGGG

...CCACTGGTTGGTTCAAGATATTT

TR

GJP

1*01

E1TR

GV

9*01

TGTGCCTTGTGGGAG...ACTTCCTACCTC

....TATTATAAGAAACTCTTT

TR

GJ1

*01

E2TR

GV

9*01

TGTGCCTTGTGGGAG...

CCC

.AATTATTATAAGAAACTCTTT

TR

GJ2

*01

TC

Rδ

chai

ns

Clo

neV

nam

e3’

V-R

EGIO

NN

1P

D1-

REG

ION

N2

D2-

REG

ION

N3

P5’

J-R

EGIO

NJ

nam

eD

1na

me

D2

nam

e

B11

TRD

V1*

01TGTGCTCTTGGGGAAC.

AGGTCG

..TTCCTA.TTGATCT....GGGGAT...

TCC

GT

ACACCGATAAACTCATCTTT

TR

DJ1

*01

TR

DD

2*01

TR

DD

3*01

D3

TRD

V1*

01TGTGCTCTTGGG.....

AAAAGTGGCA

...GGGGGAT...

CACCA

......ATAAACTCATCTTT

TR

DJ1

*01

TR

DD

3*01

E1TR

DV

1*01

TGTGCTCTTGGGGAACTCGGACGGGGAGGGATACTGGGGGA....

AATT

..ACCGATAAACTCATCTTT

TR

DJ1

*01

TR

DD

3*01

E2TR

DV

1*01

TGTGCTCTTGGGGAACT

.....CTAC

ACA

..TGGGGGATAC.AGCCTT

CTTTGACAGCACAACTCTTCTTT

TR

DJ2

*01

TR

DD

2*01

TR

DD

3*01

Tabl

e6.

3:C

DR

3se

quen

ces

ofV

δ2ne

gT-

cell

clon

esIm

Mun

oGen

eTic

s(I

MG

TR ©

)Ju

ncti

onA

naly

sis

outp

ut(www.imgt.org

)

117

www.imgt.org

118

Chapter 7

General discussion

Allogeneic stem cell transplantation is used as a curative treatment for haemat-

ological malignancies, as shown in Part I for multiple myeloma. Treatment

effect is due to both the conditioning regimen before transplantation and the

graft-versus-tumor effect after transplantation. However, relapses often occur

and also non-relapse mortality (NRM) and morbidity is very high in both myel-

oablative (1-year NRM 32%) and nonmyeloablative (1-year NRM 20%) allo-SCT

[59, 204].

The main goals of future allo-SCT strategies remain therefore to improve GVL

and prevent or develop options to effectively treat GVHD. In Part II and III

of this thesis different approaches to reach these goals were presented. First,

the role of B cells and B-cell depletion in treatment and prevention of chronic

GVHD, as described in Part II, will be discussed. Secondly, γδT cells and op-

tions to use them as an additional tool for immunotherapy for the effective

treatment and therapy of leukemia and infectious diseases, as investigated in

Part III, will be considered.

7.1 The role of B cells in GVHD

Several, mainly retrospective, studies have shown clinical efficacy of B-cell de-

pletion with the monoclonal antibody rituximab, as a treatment for steroid-

refractory chronic GVHD [36, 51, 172, 179, 242], thereby suggesting a significant

role of B cells in the development of chronic GVHD. However the exact mech-

anism behind these promising results has yet to be elucidated. In theory both

donor and host B cells could mediate allo-reactivity in chronic GVHD [197].

119

In both scenarios B cells could induce allo-reactivity directly, via production of

auto-antibodies. Increase in several auto-antibodies is seen in various studies

[36, 73, 212], however there is no consensus on which antibodies are important

in the pathogenesis of chronic GVHD. Also several B cell subsets and activation

markers are proposed to be of importance, though mostly conflicting results

are published on the exact phenotype of B cells involved in the pathogenesis

of chronic GVHD [83, 192]. Alternatively or additionally it is suggested that B

cells can act not only as direct mediators [192], but also indirectly, by present-

ing allo-antigens and therefore inducing allo-reactivity in the T-cell population

[197]. Also this could play a role for host (direct presentation) and donor B cells

(indirect presentation).

Pre-emptive B-cell depletion before allo-SCT, thus depletion of host B cells has

been reported to affect mainly the frequency of acute GVHD [60, 113, 180].

In contrast, our data (Chapter 3) suggest that pre-emptive B-cell depletion can

also substantially affect the incidence of chronic GVHD. These results suggest

that not only donor B-cells can contribute to chronic GVHD, also host B cells

can prime the new developing immunity towards allo-reactivity, and therefore

B-cell depletion as part of the conditioning regimen could be considered to

prevent chronic GVHD. However, in Chapter 3 we also revealed a potential

threat by depletion of B cells prior to allo-SCT. Severe acute GVHD occurred

faster in patients treated with rituximab prior to allo-SCT. In mice, host B cells

were shown to produce IL-10 and function as regulatory cells, attenuating the

course of acute GVHD [188]. However two retrospective studies showed an

even reduced incidence of acute GVHD in patients treated with rituximab prior

to allo-SCT [113, 180], although they did not comment on the time of onset of

acute GVHD. The patient numbers used to show the fast onset of severe acute

GVHD in Chapter 3 were very limited. Since in a larger cohort a beneficial

effect was seen on the incidence of acute GVHD and also on NRM, this threat

might not be of great influence, although prospective studies addressing this

issue remain necessary to answer this question.

Prophylactic depletion of donor B cells after allo-SCT could also be an altern-

ative preventive strategy. The first clinical trial using rituximab as prophylactic

treatment for chronic GVHD, administered 2 months post allo-SCT, showed an

120

CHAPTER 7. GENERAL DISCUSSION

incidence of total chronic GVHD of 20% [11]. This is a very promising result,

since generally up to 70% of patients develop chronic GVHD [10, 130, 170, 225].

Although It has been documented that B-cell depletion in the first 6 months after

(T-cell depleted) allo-SCT caused prolonged life-threatening cytopenias [155],

Arai and colleagues showed a non relapse mortality of only 3% [11], suggest-

ing this problem could be overcome. Up to now no results for a randomized

controlled trial have been published, therefore this chapter cannot be closed yet.

Despite here-presented evidence for the prophylactic B-cell depletion in order

to prevent GVHD, B-cell depletion is so far mainly used once chronic GVHD oc-

curred. In Chapter 4 we demonstrate that patients with a sclerodermatous skin

phenotype were shown to be more responsive to B-cell depletion than patients

with an ulcerative skin phenotype. In patients with an ongoing response, the

response could be correlated with a change in B-cell homeostasis. B-cell imbal-

ances were found in patients with active chronic GVHD of the sclerodermatous

skin phenotype. These imbalances were corrected or even reset by treatment

with rituximab. After depletion the new B-cell compartment was comparable

to patients without chronic GVHD. This suggests that chronic GVHD with a

sclerodermatous skin phenotype is B-cell driven and patients with this pheno-

type will benefit from anti-B cell therapy.

To a large extent B-cell homeostasis is controlled by B cell activating factor be-

longing to the TNF family (BAFF) [195, 217]. In patients with chronic GVHD

disturbances in B-cell homeostasis have been reflected in disturbances in BAFF

levels [3, 192]. In patients with active chronic GVHD a high BAFF/B-cell ratio

and low numbers of naïve B cells suggest a shift towards activated memory B

cells, responsible for the allo-reactivity that leads to chronic GVHD [116]. In

our prospectively studied cohort and in a prospective study by Kim et al, no

significant differences in BAFF levels were found [115]. Moreover, we found

high numbers of naïve B cells with an antigen-presenting phenotype. These

controversial results could be due to the fact that patients in our cohort received

higher amounts of corticosteroids then in other cohorts, which have been de-

scribed to inhibit BAFF and therefore could influence B-cell activation [105]. In

steroid-refractory chronic GVHD patients, that are responsive to B-cell deple-

tion, B cells seem to escape the inhibitory effect of steroids and an imbalanced

121

B-cell homeostasis can be maintained.

The exact definition of B-cell imbalances needs further exploration, since dif-

ferent authors show different and even contradictory results [3, 83, 192, 226].

Regardless of the underlying mechanism one could argue that patients with

a sclerodermatous skin phenotype and proven B-cell imbalances should be

treated early with B-cell depletion, thus even before clinical signs of chronic

GVHD.

Although B cell depletion appears as promising strategy in order treat patients

with chronic GVHD, none of the patients developed a complete resolution of

GVHD and other did not respond at all. In Chapter 4 a more ulcerative skin

phenotype of chronic GVHD was described as less responsive. In these patients

lesions of the skin were more comparable to those seen in patients with acute

GVHD. B-cell phenotype was not altered and B-cell depletion did not result

in improvement of symptoms of chronic GVHD. In our cohort 17% of patients

suffered from this ulcerative type of chronic GVHD, not responding to B-cell

depletion. This is a substantial population; therefore one could argue that in

times of personalized medicine, including B-cell depletion in the conditioning

regimen for all patients seems rather blunt. When these patients can be identi-

fied a priori, they can also be treated differently. A suggestion worth researching

is treating these patients with tools now used in the treatment of acute GVHD,

such as extracorporeal photopheresis [9, 93] and mesenchymal stromal cells

(MSC). The immune modulatory effect of MSCs is successfully used in the treat-

ment of steroid-refractory acute GVHD [186, 230], as well as first line treatment

[112]. Also in chronic GVHD preliminary data show remission of both sclero-

derma and ulcera after administration of MSC’s in steroid-refractory patients

[245]. Finally also tyrosine-kinase inhibitors, such as imatinib and nilotinib,

could be an alternative option to treat unresponsive patients or to reach com-

plete resolution of symptoms in partially responding patients. Imatinib has

been shown to decrease the amount of fibrosis seen in pulmonary fibrosis [1].

Also in systemic sclerosis and chronic GVHD there has been rationale to treat

patients with tyrosine-kinase inhibitors [156, 169]. In these patients fibroblasts

show an activated myofibroblast phenotype as the result of stimulation by not

only higher levels of pro-fibrotic cytokines such as transforming growth factor

122


(TGF)-β [94, 154] and PDGF [211], but also due to production of stimulatory

anti-PDGFR antibodies [16, 212]. By inhibiting the platelet-derived-growth-

factor (PDGF)- receptor, tyrosine-kinase inhibitors can attenuate the activation

of fibroblasts by PDGF and TGF-β as well as their stimulation by anti-PDGFR

antibodies. First clinical trials treating patients with steroid-refractory chronic

GVHD with imatinib show overall response rates up to 79% [145, 146, 168].

To improve the treatment response of patients with steroid-refractory chronic

GVHD treated with B-cell depletion, we recently extended the clinical trial, de-

scribed in Chapter 4. Patients with steroid-refractory chronic GVHD with a skin

involvement will be treated with rituximab and thereafter nilotinib will be given

for 6 months (http://www.hovon.nl/studies/studies-perziektebeeld/sct.

html?action=showstudie&studie_id=90&categorie_id=11).

In conclusion, we propose that chronic GVHD should be treated as a hetero-

geneous disease. This is clinically suggested by the fact that various sites are

affected with a different type of pathology [92, 201] and supported by the obser-

vation that immunological phenotype of chronic GVHD can differ [3, 116, 226].

Patients, who will benefit from B-cell depletion, might be identified by their

aberrant B-cell profile early and they should receive an early treatment with

rituximab. Patients, who show no imbalances in B-cell phenotype, will most

likely not respond to B-cell depletion and should be treated differently. How-

ever, larger randomized trials are needed for this conclusion backed up by phase

I/II trials to explore new drugs for the treatment of chronic GVHD.

7.2 γδT cells as valuable tool for cellular immuno-

therapy

αβT cells directed against minor-histocompatibility antigens restricted to

haematological stem cells or even leukemic cells have been shown to induce a

powerful antitumor response, while GVHD is prevented [205, 233]. In addition,

also αβT cells directed against defined tumor-associated antigens have been

reported to control leukemia after allo-SCT [117, 222]. Thus, αβT cells are un-

doubtedly important tools in immunotherapy of both solid and haematological

malignancies. However, major limitations arise through the fact that αβTCRs

123

http://www.hovon.nl/studies/studies-perziektebeeld/sct.html?action=showstudie&studie_id=90&categorie_id=11

http://www.hovon.nl/studies/studies-perziektebeeld/sct.html?action=showstudie&studie_id=90&categorie_id=11

recognize their target in an HLA dependent manner and usually only one or

a limited amount of peptides [117]. Consequently, each individual patient re-

quires and optimal set of TCR pMHC interactions in order to control leukemia

and viral infections. In addition, immune responses against tumor-associated

antigens frequently underlie mechanisms of tolerance, thus an optimal set of

TCR-pMHC interactions can frequently not identified for patients [124]. Finally,

frequencies of in particular minor-histocompatibility-antigen reactive T-cells is

usually low. Thus an extensive amplification [183] or engineering [124] of im-

mune cells is required for a potential clinical application. The here-discussed

multiple obstacles hamper so far the broad application of an individualized im-

mune therapy with αβT cells in nearly all clinical scenarios including allo-SCT

and urged us to search for additional immunological tools in order to control

leukemia and infections.

We propose in this thesis that one cell population, namely γδT cells and their

receptors might overcome multiple major limitations of αβT cells. Firstly, we

demonstrate in Chapter 5 that the frequency of tumor-reactive γδT cells in

healthy individuals is usually high. Secondly, we demonstrate as reported by

others that γδT cells and in particular the γδTCR recognizes its target HLA-

independently [72], thus γδT cells or their receptors could be used for as broader

patient population.

We investigated in detail the benefit of γδTCRs as tool for adoptive immune

therapies and suggest many advantages. Firstly, engineering of high affinity

γ9δ2TCRs is feasible and allows the selection of a highly tumor-reactive re-

ceptor without substantial side effects [88]. Secondly, in contrast to αβTCRs

[123], γδTCRs do not pair with endogenous αβTCR chains [149, 223], thus this

strategy prevents generation of unwanted (potentially autoreactive) specificities.

Thirdly, these genetically modified cells are not expected to be immunogenic as

they do not harbor fusion or non-self-proteins as reported for multiple genetic-

ally modified T cells [21]. Forthly, γδT cells and their receptors do not display

substantial GVHD properties [111]. Fifthly, expression of γ9δ2TCR in αβT cells

helps to overcome defects in proliferation seen in γ9δ2Tcells. Finally, in contrast

to most αβTCRs, expression of high-affinity αβTCRs leads to efficient repro-

gramming of carrier cells (CD4+ and CD8+ αβT-cells) and to enhanced immune

124


responses [88, 149]. Collectively, these features plus the MHC-independent fash-

ion of antigen recognition make γδTCR an ideal tool for the generation of novel

T-cell based cellular immune cells [88, 111, 149, 194].

Although many reasons have been suggested as to why γδT cells and their

receptors could be valuable tools for immune therapies, up till now clinical

studies exploiting the antitumor reactivity of γδT cells in adoptive T-cell trans-

fer are rather disappointing [167]. In Chapter 5 we gave a possible explanation

and solution for this problem. We showed that the strong anti-tumor reactivity

of γ9δ2T cells was not a feature of the total population of these cells. Individual

γ9δ2T-cell clones showed differences in their anti-tumor response in specificity

and functional avidity. We showed that these features are regulated by the

CDR3 regions of both γ9 and δ2 chain. Single amino acid changes in the highly

variable part of the CDR3 region of the γ9 chain, diverse amino acid compos-

itions in the CDR3 region of the δ2 chain and length of total CDR3 region of

both γ9 and δ2 chain can all influence functional avidity. We showed that a spe-

cific combination of a γ9 and a δ2 chain is particularly important and therefore

we hypothesize that highly variable parts of the CDR3 regions of both γ9 and

δ2 chain complement each other to form a structure that defines target recog-

nition. We used a new strategy, combinatorial-γδTCR-chain-exchange (CTE),

which results in the expression of newly combined γ9- and δ2-TCR chains on

engineered T-cells. CTE allows us to design γ9δ2TCRs that mediate broad and

strong anti-tumor response. These TCRs can be used for rapid engineering of

tumor-reactive T-cells that are not limited by HLA restrictions and will not sur-

prise us with unwanted specificity. CTE-engineered γ9δ2TCRs provided higher

tumor reactivity against a broad panel of tumor cells, derived from both solid as

well as haematological malignancies. Therefore one could argue that in theory

these cells could be used for any patient with any type of cancer, making them

a precious additional tool for future cellular immunotherapies.

However, there still remain unsolved issues. One for example is the limited life

span of engineered effector cells. Expansion of tumor specific, adoptively trans-

ferred T cells has only been observed in some patients and only after lymphode-

pletion [62, 162]. In attempts to improve the survival of these cells in a patient

several techniques have been tried, for example stimulating transferred T cells

125

in vivo with IL-2 and pamidronate [149], transfer of central memory phenotype

T cells after in vitro treatment with cytokine cocktails [240], using dual reactive

TCRs against both persistent viruses and leukemia [227]. These studies all re-

ported improvement of survival of these cells in vitro or in murine models. More

data in clinical studies is required to optimize the transfer of transduced T cells.

The use of a dual reactive TCR could efficiently prolong survival and there-

fore anti-tumor reactivity of transferred tumor specific T cells. This could be

due to more frequent encounter of both viral and tumor antigens by the TCR

and therefore a more sustained impulse to proliferate is available [227]. This

method furthermore offers a possibility to tackle two problems at once. CMV

reactivations after allo-SCT cause life-threatening complications, such as CMV

pneumonia or colitis [28]. A T cell showing both anti-tumor and CMV-reactivity

would be a very powerful tool in creating a therapy to treat leukemia and at the

same time prevent CMV-reactivation after transplantation. Again γδT cells seem

interesting candidates to achieve this goal, since they recognize self-antigens

upregulated on transformed and virally infected cells. Their role in both anti-

tumor and anti-viral surveillance has well been established [29]. CMV infection

induces in both healthy individuals [176] as well as immunocompromised pa-

tients after allo-SCT [118] as well as in utero [229] in vivo expansion of specifically

Vδ2neg γδT-cells. In addition Vδ2neg γδT-cells have been shown cytotoxicity to

cancer cells [41, 45]. Interestingly, CMV reactivation after allo-SCT reduces the

risk of leukemic relapse [18, 166]. This suggests an association between the im-

mune response against CMV infection and clearance of tumor. In Chapter 6

we demonstrate that CMV-reactive Vδ2neg γδT cells from both conventional

and cordblood donors expanded in vivo after allo-SCT. These γδT cells were

capable of cross-recognizing haematological cancers. This might explain the re-

duced risk of relapse of cancer in patients with CMV-reactivation. Vδ2neg γδT

cells could therefore very well be used as a tool to attack both CMV infection

and leukemia. Since CMV-reactive γδT cells could also be obtained from the

naïve umbilical cordblood repertoire, our results emphasize the important role

of these cells as a third-party stem cell source for allo-SCT.

Recognition of CMV infected cells and cancer cells by Vδ2neg γδT cells is demon-

strated to be TCR dependent [90, 110, 235]. Therefore we hypothesized that

126


here-cloned γδTCRs could recognize antigens shared by CMV-infected and

transformed cells. However, transduction of the isolated γδTCRs into an αβT

cell showed that only cancer-reactivity, but not CMV-reactivity was transferred.

Thus, either an alternative immune receptor is responsible for CMV-reactivity

of the original clones, or the γδTCR depends on additional molecules that are

not expressed on αβT cells.

The exact mechanism of interaction of γδT cells with their target remains, des-

pite thorough and extensive research, somewhat of a mystery. Involvement of

the TCR itself in both tumor recognition [41, 86, 110] and infectious targets

[17, 90, 96, 199] is undeniable. However, a role for natural killer (NK) recept-

ors, such as NKG2D and NKp30, has also been described in target recognitions

by γδT cells [53, 87, 102, 177, 184]. Our group demonstrated that transfer of

a tumor-specific γ9δ2TCR to NKG2Dneg and NKG2Dpos αβT cells resulted in

the same anti-tumor reactivity. However blocking NKG2D on γ9δ2TCR trans-

duced NKG2Dpos αβT cells decreased reactivity to tumor cells [149]. These

results taken together with the diverse package of suggested target molecules,

such as phosphoantigens, F1ATPase, MHC-like molecules such as MICA and

ULBP’s, tempt us to hypothesize that recognition by γδT cells is not solely done

by the TCR or solely done by NK-receptors, but is the result of an interaction

between those receptors. Since many αβT cells also express NKG2D this feature

can also be transferred to αβT cells and be of use in clinical applications of ad-

optive transfer of tumor specific γδTCRs. Within the here-presented thesis, we

were able to also elucidate a novel role of CD8αα as an alternative costimulat-

ory molecule necessary for Vδ2neg γδT cells to interact with tumor cells [194].

Blocking of CD8α on CD8αα- expressing distinct γδT-cell clones as well as trans-

duced CD8+ αβT-cells resulted in a significant inhibition of tumor recognition.

However, the precise role of CD8αα for tumor recognition in combination with

Vδ2neg γδTCRs needs further investigation.

7.3 Conclusion

Our presented papers contribute to the improvement of allo-SCT by offering

several possibilities to shape the graft into a more refined tool to treat haemato-

logical malignancies. We demonstrated that altering the composition of a graft

127

by early or late B-cell depletion provides interesting options in order to reduce

GVHD [225, 226]. In addition we introduced γδT cells and their receptors as in-

teresting additional tools for (engineered) adaptive immune therapies [88, 194],

which might one day replace the conventional allo-SCT.

128

Bibliography

[1] A Abdollahi, M Li, G Ping, C Plathow, S Domhan, F Kiessling, L B Lee, G McMa-hon, H J Gröne, K E Lipson, and P E Huber. Inhibition of platelet-derived growthfactor signaling attenuates pulmonary fibrosis. J Exp Med, 201(6):925–935, Mar2005.

[2] E G Addison, J North, I Bakhsh, C Marden, S Haq, S Al-Sarraj, R Malayeri, R GWickremasinghe, J K Davies, and M W Lowdell. Ligation of CD8alpha on humannatural killer cells prevents activation-induced apoptosis and enhances cytolyticactivity. Immunology, 116(3):354–361, Nov 2005.

[3] J L Allen, M S Fore, J Wooten, P A Roehrs, N S Bhuiya, T Hoffert, A Sharf, A MDeal, P Armistead, J Coghill, D A Gabriel, R Irons, A Essenmacher, T C Shea,K Richards, C Cutler, J Ritz, J Serody, A S Baldwin, and S Sarantopoulos. B cellsfrom patients with chronic GVHD are activated and primed for survival via BAFF-mediated pathways. Blood, 120(12):2529–2536, Sep 2012.

[4] T J Allison, C C Winter, J J Fournié, M Bonneville, and D N Garboczi. Structure ofa human gammadelta T-cell antigen receptor. Nature, 411(6839):820–824, Jun 2001.

[5] E Alyea, E Weller, R Schlossman, C Canning, P Mauch, A Ng, D Fisher, J Gribben,A Freeman, B Parikh, P Richardson, R Soiffer, J Ritz, and K C Anderson. Out-come after autologous and allogeneic stem cell transplantation for patients withmultiple myeloma: impact of graft-versus-myeloma effect. Bone Marrow Transplant,32(12):1145–1151, Dec 2003.

[6] E P Alyea, H T Kim, V Ho, C Cutler, J Gribben, D J DeAngelo, S J Lee, S Windawi,J Ritz, R M Stone, J H Antin, and R J Soiffer. Comparative outcome of nonmyeloab-lative and myeloablative allogeneic hematopoietic cell transplantation for patientsolder than 50 years of age. Blood, 105(4):1810–1814, Feb 2005.

[7] J H Anolik, J W Friedberg, B Zheng, J Barnard, T Owen, E Cushing, J Kelly, E CMilner, R I Fisher, and I Sanz. B cell reconstitution after rituximab treatment oflymphoma recapitulates B cell ontogeny. Clin Immunol, 122(2):139–145, Feb 2007.

[8] J. H. Antin. Reduced-Intensity Stem Cell Transplantation: “. . . whereof a littleMore than a little is by much too much.” King Henry IV, part 1, I, 2. HematologyAm Soc Hematol Educ Program, pages 47–54, 2007.

[9] N Apisarnthanarax, M Donato, M Körbling, D Couriel, J Gajewski, S Giralt,I Khouri, C Hosing, R Champlin, M Duvic, and P Anderlini. Extracorporeal pho-topheresis therapy in the management of steroid-refractory or steroid-dependent

129

cutaneous chronic graft-versus-host disease after allogeneic stem cell transplanta-tion: feasibility and results. Bone Marrow Transplant, 31(6):459–465, Mar 2003.

[10] F R Appelbaum. Haematopoietic cell transplantation as immunotherapy. Nature,411(6835):385–389, May 2001.

[11] S Arai, B Sahaf, B Narasimhan, G L Chen, C D Jones, R Lowsky, J A Shizuru, L JJohnston, G G Laport, W K Weng, J E Benjamin, J Schaenman, J Brown, J Ramirez,J L Zehnder, R S Negrin, and D B Miklos. Prophylactic rituximab after allogeneictransplantation decreases B-cell alloimmunity with low chronic GVHD incidence.Blood, 119(25):6145–6154, Jun 2012.

[12] B Arden, S P Clark, D Kabelitz, and T W Mak. Human T-cell receptor variablegene segment families. Immunogenetics, 42(6):455–500, 1995.

[13] M Arora, P B McGlave, L J Burns, J S Miller, J N Barke, T E Defor, and D J Weis-dorf. Results of autologous and allogeneic hematopoietic cell transplant therapyfor multiple myeloma. Bone Marrow Transplant, 35(12):1133–1140, Jun 2005.

[14] K Atkinson, M M Horowitz, R P Gale, D W van Bekkum, E Gluckman, R AGood, N Jacobsen, H J Kolb, A A Rimm, and O Ringdén. Risk factors for chronicgraft-versus-host disease after HLA-identical sibling bone marrow transplantation.Blood, 75(12):2459–2464, Jun 1990.

[15] T C Barnes, M E Anderson, and R J Moots. The many faces of interleukin-6: therole of IL-6 in inflammation, vasculopathy, and fibrosis in systemic sclerosis. Int JRheumatol, 2011:721608–721608, 2011.

[16] S S Baroni, M Santillo, F Bevilacqua, M Luchetti, T Spadoni, M Mancini, P Frati-celli, P Sambo, A Funaro, A Kazlauskas, E V Avvedimento, and A Gabrielli. Stim-ulatory autoantibodies to the PDGF receptor in systemic sclerosis. N Engl J Med,354(25):2667–2676, Jun 2006.

[17] C Behr, R Poupot, M A Peyrat, Y Poquet, P Constant, P Dubois, M Bonneville,and J J Fournie. Plasmodium falciparum stimuli for human gammadelta T cellsare related to phosphorylated antigens of mycobacteria. Infect Immun, 64(8):2892–2896, Aug 1996.

[18] C E Behrendt, J Rosenthal, E Bolotin, R Nakamura, J Zaia, and S J Forman. Donorand recipient CMV serostatus and outcome of pediatric allogeneic HSCT for acuteleukemia in the era of CMV-preemptive therapy. Biol Blood Marrow Transplant,15(1):54–60, Jan 2009.

[19] I Bellantuono, L Gao, S Parry, S Marley, F Dazzi, J Apperley, J M Goldman, and H JStauss. Two distinct HLA-A0201-presented epitopes of the Wilms tumor antigen1 can function as targets for leukemia-reactive CTL. Blood, 100(10):3835–3837, Nov2002.

[20] G M Bendle, C Linnemann, A I Hooijkaas, L Bies, M A de Witte, A Jorritsma, A DKaiser, N Pouw, R Debets, E Kieback, W Uckert, J Y Song, J B Haanen, and T N

130

BIBLIOGRAPHY

Schumacher. Lethal graft-versus-host disease in mouse models of T cell receptorgene therapy. Nat Med, 16(5):565–570, May 2010.

[21] C Berger, M E Flowers, E H Warren, and S R Riddell. Analysis of transgene-specific immune responses that limit the in vivo persistence of adoptively trans-ferred HSV-TK-modified donor T cells after allogeneic hematopoietic cell trans-plantation. Blood, 107(6):2294–2302, Mar 2006.

[22] N L Berinstein, A J Grillo-López, C A White, I Bence-Bruckler, D Maloney,M Czuczman, D Green, J Rosenberg, P McLaughlin, and D Shen. Association ofserum Rituximab (IDEC-C2B8) concentration and anti-tumor response in the treat-ment of recurrent low-grade or follicular non-Hodgkin’s lymphoma. Ann Oncol,9(9):995–1001, Sep 1998.

[23] K Besold, N Frankenberg, S Pepperl-Klindworth, J Kuball, M Theobald, G Hahn,and B Plachter. Processing and MHC class I presentation of human cytomeg-alovirus pp65-derived peptides persist despite gpUS2-11-mediated immune eva-sion. J Gen Virol, 88(Pt 5):1429–1439, May 2007.

[24] B B Björkstrand, P Ljungman, H Svensson, J Hermans, A Alegre, J Apperley,J Bladé, K Carlson, M Cavo, A Ferrant, A H Goldstone, A de Laurenzi, I Ma-jolino, R Marcus, H G Prentice, K Remes, D Samson, A Sureda, L F Verdonck,L Volin, and G Gahrton. Allogeneic bone marrow transplantation versus autolog-ous stem cell transplantation in multiple myeloma: a retrospective case-matchedstudy from the European Group for Blood and Marrow Transplantation. Blood,88(12):4711–4718, Dec 1996.

[25] J Bladé, D Samson, D Reece, J Apperley, B Björkstrand, G Gahrton, M Gertz,S Giralt, S Jagannath, and D Vesole. Criteria for evaluating disease responseand progression in patients with multiple myeloma treated by high-dose therapyand haemopoietic stem cell transplantation. Myeloma Subcommittee of the EBMT.European Group for Blood and Marrow Transplant. Br J Haematol, 102(5):1115–1123, Sep 1998.

[26] W K Bleeker, M E Munk, W J Mackus, J H van den Brakel, M Pluyter, M J Glennie,J G van de Winkel, and P W Parren. Estimation of dose requirements for sus-tained in vivo activity of a therapeutic human anti-CD20 antibody. Br J Haematol,140(3):303–312, Feb 2008.

[27] M Boeckh and A P Geballe. Cytomegalovirus: pathogen, paradigm, and puzzle. JClin Invest, 121(5):1673–1680, May 2011.

[28] M Boeckh, W G Nichols, G Papanicolaou, R Rubin, J R Wingard, and J Zaia.Cytomegalovirus in hematopoietic stem cell transplant recipients: Current status,known challenges, and future strategies. Biol Blood Marrow Transplant, 9(9):543–558,Sep 2003.

[29] M Bonneville, R L O’Brien, and W K Born. Gammadelta T cell effector functions: ablend of innate programming and acquired plasticity. Nat Rev Immunol, 10(7):467–478, Jul 2010.

131

[30] W K Born, L Zhang, M Nakayama, N Jin, J L Chain, Y Huang, M K Aydintug,and R L O’Brien. Peptide antigens for gamma/delta T cells. Cell Mol Life Sci,68(14):2335–2343, Jul 2011.

[31] J R Brown, H T Kim, P Armand, C Cutler, D C Fisher, V Ho, J Koreth, J Ritz,C Wu, J H Antin, R J Soiffer, J G Gribben, and E P Alyea. Long-term follow-upof reduced-intensity allogeneic stem cell transplantation for chronic lymphocyticleukemia: prognostic model to predict outcome. Leukemia, 27(2):362–369, Feb 2013.

[32] R P Bucy, C L Chen, and M D Cooper. Tissue localization and CD8 accessorymolecule expression of T gamma delta cells in humans. J Immunol, 142(9):3045–3049, May 1989.

[33] J F Bukowski, C T Morita, H Band, and M B Brenner. Crucial role of TCR gammachain junctional region in prenyl pyrophosphate antigen recognition by gammadelta T cells. J Immunol, 161(1):286–293, Jul 1998.

[34] J F Bukowski, C T Morita, Y Tanaka, B R Bloom, M B Brenner, and H Band. Vgamma 2V delta 2 TCR-dependent recognition of non-peptide antigens and Daudicells analyzed by TCR gene transfer. J Immunol, 154(3):998–1006, Feb 1995.

[35] M R Canninga-van Dijk, C J Sanders, L F Verdonck, R Fijnheer, and J G van denTweel. Differential diagnosis of skin lesions after allogeneic haematopoietic stemcell transplantation. Histopathology, 42(4):313–330, Apr 2003.

[36] M R Canninga-van Dijk, H M van der Straaten, R Fijnheer, C J Sanders, J G van denTweel, and L F Verdonck. Anti-CD20 monoclonal antibody treatment in 6 patientswith therapy-refractory chronic graft-versus-host disease. Blood, 104(8):2603–2606,Oct 2004.

[37] G Cartron, H Blasco, G Paintaud, H Watier, and C Le Guellec. Pharmacokineticsof rituximab and its clinical use: thought for the best use? Crit Rev Oncol Hematol,62(1):43–52, Apr 2007.

[38] B Castella, C Vitale, M Coscia, and M Massaia. Vγ9Vδ2 T cell-based immuno-therapy in hematological malignancies: from bench to bedside. Cell Mol Life Sci,68(14):2419–2432, Jul 2011.

[39] H Cheroutre. Starting at the beginning: new perspectives on the biology of mu-cosal T cells. Annu Rev Immunol, 22:217–246, 2004.

[40] H Cheroutre and F Lambolez. Doubting the TCR coreceptor function ofCD8alphaalpha. Immunity, 28(2):149–159, Feb 2008.

[41] A Choudhary, F Davodeau, A Moreau, M A Peyrat, M Bonneville, and F Joter-eau. Selective lysis of autologous tumor cells by recurrent gamma delta tumor-infiltrating lymphocytes from renal carcinoma. J Immunol, 154(8):3932–3940, Apr1995.

132

BIBLIOGRAPHY

[42] P Constant, F Davodeau, M A Peyrat, Y Poquet, G Puzo, M Bonneville, and J JFournié. Stimulation of human gamma delta T cells by nonpeptidic mycobacterialligands. Science, 264(5156):267–270, Apr 1994.

[43] J J Cornelissen, W E Fibbe, A V Schattenberg, E J Petersen, R Willemze, T J de Witte,B Löwenberg, L F Verdonck, A vd Biezen, and R Brand. A retrospective Dutchstudy comparing T cell-depleted allogeneic blood stem cell transplantation vs Tcell-depleted allogeneic bone marrow transplantation. Bone Marrow Transplant, 21Suppl 3:66–70, Jun 1998.

[44] P Corradini, M Cavo, H Lokhorst, G Martinelli, C Terragna, I Majolino,P Valagussa, M Boccadoro, D Samson, A Bacigalupo, N Russell, V Montefusco,C Voena, G Gahrton, and Chronic Leukemia Working Party of the European Groupfor Blood and Marrow Transplantation (EBMT). Molecular remission after myelo-ablative allogeneic stem cell transplantation predicts a better relapse-free survivalin patients with multiple myeloma. Blood, 102(5):1927–1929, Sep 2003.

[45] D V Correia, M Fogli, K Hudspeth, M G da Silva, D Mavilio, and B Silva-Santos.Differentiation of human peripheral blood Vδ1+ T cells expressing the naturalcytotoxicity receptor NKp30 for recognition of lymphoid leukemia cells. Blood,118(4):992–1001, Jul 2011.

[46] M Corvaisier, A Moreau-Aubry, E Diez, J Bennouna, J F Mosnier, E Scotet, M Bon-neville, and F Jotereau. V gamma 9V delta 2 T cell response to colon carcinomacells. J Immunol, 175(8):5481–5488, Oct 2005.

[47] L Couzi, Y Levaillant, A Jamai, V Pitard, R Lassalle, K Martin, I Garrigue, O Haw-char, F Siberchicot, N Moore, J F Moreau, J Dechanet-Merville, and P Merville.Cytomegalovirus-induced gammadelta T cells associate with reduced cancer riskafter kidney transplantation. J Am Soc Nephrol, 21(1):181–188, Jan 2010.

[48] L Couzi, V Pitard, X Sicard, I Garrigue, O Hawchar, P Merville, J F Moreau, andJ Déchanet-Merville. Antibody-dependent anti-cytomegalovirus activity of humanγδT cells expressing CD16 (FcγRIIIa). Blood, 119(6):1418–1427, Feb 2012.

[49] T Crough and R Khanna. Immunobiology of human cytomegalovirus: from benchto bedside. Clin Microbiol Rev, 22(1):76–98, Jan 2009.

[50] C Cutler, S Giri, S Jeyapalan, D Paniagua, A Viswanathan, and J H Antin. Acuteand chronic graft-versus-host disease after allogeneic peripheral-blood stem-celland bone marrow transplantation: a meta-analysis. J Clin Oncol, 19(16):3685–3691,Aug 2001.

[51] C Cutler, D Miklos, H T Kim, N Treister, S B Woo, D Bienfang, L B Klickstein,J Levin, K Miller, C Reynolds, R Macdonell, M Pasek, S J Lee, V Ho, R Soiffer, J HAntin, J Ritz, and E Alyea. Rituximab for steroid-refractory chronic graft-versus-host disease. Blood, 108(2):756–762, Jul 2006.

[52] A Dalloul. CD5: a safeguard against autoimmunity and a shield for cancer cells.Autoimmun Rev, 8(4):349–353, Feb 2009.

133

[53] H Das, V Groh, C Kuijl, M Sugita, C T Morita, T Spies, and J F Bukowski. MICAengagement by human Vgamma2Vdelta2 T cells enhances their antigen-dependenteffector function. Immunity, 15(1):83–93, Jul 2001.

[54] W de Jager, K Bourcier, G T Rijkers, B J Prakken, and V Seyfert-Margolis. Pre-requisites for cytokine measurements in clinical trials with multiplex immunoas-says. BMC Immunol, 10:52–52, 2009.

[55] J Déchanet, P Merville, F Bergé, G Bone-Mane, J L Taupin, P Michel, P Joly, M Bon-neville, L Potaux, and J F Moreau. Major expansion of gammadelta T lymphocytesfollowing cytomegalovirus infection in kidney allograft recipients. J Infect Dis,179(1):1–8, Jan 1999.

[56] J Déchanet, P Merville, A Lim, C Retière, V Pitard, X Lafarge, S Michelson,C Méric, M M Hallet, P Kourilsky, L Potaux, M Bonneville, and J F Moreau. Implic-ation of gammadelta T cells in the human immune response to cytomegalovirus. JClin Invest, 103(10):1437–1449, May 1999.

[57] J M den Haan, N E Sherman, E Blokland, E Huczko, F Koning, J W Drijfhout,J Skipper, J Shabanowitz, D F Hunt, and V H Engelhard. Identification of a graftversus host disease-associated human minor histocompatibility antigen. Science,268(5216):1476–1480, Jun 1995.

[58] K Deusch, F Lüling, K Reich, M Classen, H Wagner, and K Pfeffer. A major fractionof human intraepithelial lymphocytes simultaneously expresses the gamma/deltaT cell receptor, the CD8 accessory molecule and preferentially uses the V delta 1gene segment. Eur J Immunol, 21(4):1053–1059, Apr 1991.

[59] R Diaconescu, C R Flowers, B Storer, M L Sorror, M B Maris, D G Maloney, B MSandmaier, and R Storb. Morbidity and mortality with nonmyeloablative com-pared with myeloablative conditioning before hematopoietic cell transplantationfrom HLA-matched related donors. Blood, 104(5):1550–1558, Sep 2004.

[60] A Dominietto, E Tedone, M Soracco, B Bruno, A M Raiola, M T Van Lint, S Geroldi,T Lamparelli, B Galano, F Gualandi, F Frassoni, and A Bacigalupo. In vivo B-celldepletion with rituximab for alternative donor hemopoietic SCT. Bone MarrowTransplant, 47(1):101–106, Jan 2012.

[61] T Dörner, A M Jacobi, and P E Lipsky. B cells in autoimmunity. Arthritis Res Ther,11(5):247–247, 2009.

[62] M E Dudley, J R Wunderlich, T E Shelton, J Even, and S A Rosenberg. Generationof tumor-infiltrating lymphocyte cultures for use in adoptive transfer therapy formelanoma patients. J Immunother, 26(4):332–342, Jul-Aug 2003.

[63] H Einsele, G Ehninger, H Hebart, P Weber, S Dette, H Link, H P Horny, V Meuter,S Wagner, and H D Waller. Incidence of local CMV infection and acute intest-inal GVHD in marrow transplant recipients with severe diarrhoea. Bone MarrowTransplant, 14(6):955–963, Dec 1994.

134

BIBLIOGRAPHY

[64] A H Elmaagacli, N K Steckel, M Koldehoff, Y Hegerfeldt, R Trenschel, M Ditsch-kowski, S Christoph, T Gromke, L Kordelas, H D Ottinger, R S Ross, P A Horn,S Schnittger, and D W Beelen. Early human cytomegalovirus replication aftertransplantation is associated with a decreased relapse risk: evidence for a pu-tative virus-versus-leukemia effect in acute myeloid leukemia patients. Blood,118(5):1402–1412, Aug 2011.

[65] P S Evans, P J Enders, C Yin, T J Ruckwardt, M Malkovsky, and C D Pauza.In vitro stimulation with a non-peptidic alkylphosphate expands cells express-ing Vgamma2-Jgamma1.2/Vdelta2 T-cell receptors. Immunology, 104(1):19–27, Sep2001.

[66] F Fagioli, A Bacigalupo, F Frassoni, M T Van Lint, D Occhini, F Gualandi, T Lam-parelli, M Clavio, V Vitale, and G Sogno. Allogeneic bone marrow transplantationfor acute myeloid leukemia in first complete remission: the effect of FAB classific-ation and GVHD prophylaxis. Bone Marrow Transplant, 13(3):247–252, Mar 1994.

[67] J Ferrara, P Reddy, and S Paczesny. Immunotherapy through T-cell receptor genetransfer induces severe graft-versus-host disease. Immunotherapy, 2(6):791–794,Nov 2010.

[68] M Ferrarini, S Heltai, S M Pupa, S Mernard, and R Zocchi. Killing of lamininreceptor-positive human lung cancers by tumor infiltrating lymphocytes bearinggammadelta(+) t-cell receptors. J Natl Cancer Inst, 88(7):436–441, Apr 1996.

[69] T Feuchtinger, K Opherk, W A Bethge, M S Topp, F R Schuster, E M Weissinger,M Mohty, R Or, M Maschan, M Schumm, K Hamprecht, R Handgretinger, P Lang,and H Einsele. Adoptive transfer of pp65-specific T cells for the treatment ofchemorefractory cytomegalovirus disease or reactivation after haploidentical andmatched unrelated stem cell transplantation. Blood, 116(20):4360–4367, Nov 2010.

[70] A H Filipovich. Diagnosis and manifestations of chronic graft-versus-host disease.Best Pract Res Clin Haematol, 21(2):251–257, Jun 2008.

[71] A H Filipovich, D Weisdorf, S Pavletic, G Socie, J R Wingard, S J Lee, P Martin,J Chien, D Przepiorka, D Couriel, E W Cowen, P Dinndorf, A Farrell, R Hartzman,J Henslee-Downey, D Jacobsohn, G McDonald, B Mittleman, J D Rizzo, M Robin-son, M Schubert, K Schultz, H Shulman, M Turner, G Vogelsang, and M E Flowers.National Institutes of Health consensus development project on criteria for clinicaltrials in chronic graft-versus-host disease: I. Diagnosis and staging working groupreport. Biol Blood Marrow Transplant, 11(12):945–956, Dec 2005.

[72] P Fisch, M Malkovsky, E Braakman, E Sturm, R L Bolhuis, A Prieve, J A Sosman,V A Lam, and P M Sondel. Gamma/delta T cell clones and natural killer cell clonesmediate distinct patterns of non-major histocompatibility complex-restricted cyto-lysis. J Exp Med, 171(5):1567–1579, May 1990.

[73] H Fujii, G Cuvelier, K She, S Aslanian, H Shimizu, A Kariminia, M Krailo, Z Chen,R McMaster, A Bergman, F Goldman, S A Grupp, D A Wall, A L Gilman, and K R

135

Schultz. Biomarkers in newly diagnosed pediatric-extensive chronic graft-versus-host disease: a report from the Children’s Oncology Group. Blood, 111(6):3276–3285, Mar 2008.

[74] F Garban, M Attal, M Michallet, C Hulin, J H Bourhis, I Yakoub-Agha, T Lamy,G Marit, F Maloisel, C Berthou, M Dib, D Caillot, B Deprijck, N Ketterer, J L Har-ousseau, J J Sotto, and P Moreau. Prospective comparison of autologous stem celltransplantation followed by dose-reduced allograft (IFM99-03 trial) with tandemautologous stem cell transplantation (IFM99-04 trial) in high-risk de novo multiplemyeloma. Blood, 107(9):3474–3480, May 2006.

[75] M A Geller, S Cooley, P L Judson, R Ghebre, L F Carson, P A Argenta, A LJonson, A Panoskaltsis-Mortari, J Curtsinger, D McKenna, K Dusenbery, R Bliss,L S Downs, and J S Miller. A phase II study of allogeneic natural killer cell therapyto treat patients with recurrent ovarian and breast cancer. Cytotherapy, 13(1):98–107, Jan 2011.

[76] S Gerull, M Goerner, A Benner, U Hegenbart, U Klein, H Schaefer, H Goldschmidt,and A D Ho. Long-term outcome of nonmyeloablative allogeneic transplantationin patients with high-risk multiple myeloma. Bone Marrow Transplant, 36(11):963–969, Dec 2005.

[77] H Glucksberg, R Storb, A Fefer, C D Buckner, P E Neiman, R A Clift, K G Lerner,and E D Thomas. Clinical manifestations of graft-versus-host disease in human re-cipients of marrow from HL-A-matched sibling donors. Transplantation, 18(4):295–304, Oct 1974.

[78] H J Gober, M Kistowska, L Angman, P Jenö, L Mori, and G De Libero. HumanT cell receptor gammadelta cells recognize endogenous mevalonate metabolites intumor cells. J Exp Med, 197(2):163–168, Jan 2003.

[79] K T Godder, P J Henslee-Downey, J Mehta, B S Park, K Y Chiang, S Abhyankar, andL S Lamb. Long term disease-free survival in acute leukemia patients recoveringwith increased gammadelta T cells after partially mismatched related donor bonemarrow transplantation. Bone Marrow Transplant, 39(12):751–757, Jun 2007.

[80] H Goker, I C Haznedaroglu, and N J Chao. Acute graft-vs-host disease: pathobio-logy and management. Exp Hematol, 29(3):259–277, Mar 2001.

[81] J M Goldman, R P Gale, M M Horowitz, J C Biggs, R E Champlin, E Gluckman,R G Hoffmann, S J Jacobsen, A M Marmont, and P B McGlave. Bone marrowtransplantation for chronic myelogenous leukemia in chronic phase. Increased riskfor relapse associated with T-cell depletion. Ann Intern Med, 108(6):806–814, Jun1988.

[82] A Q Gomes, D S Martins, and B Silva-Santos. Targeting γδ T lymphocytes for can-cer immunotherapy: from novel mechanistic insight to clinical application. CancerRes, 70(24):10024–10027, Dec 2010.

136

BIBLIOGRAPHY

[83] H T Greinix, D Pohlreich, M Kouba, U Körmöczi, I Lohmann, K Feldmann, C Ziel-inski, and W F Pickl. Elevated numbers of immature/transitional CD21- B lympho-cytes and deficiency of memory CD27+ B cells identify patients with active chronicgraft-versus-host disease. Biol Blood Marrow Transplant, 14(2):208–219, Feb 2008.

[84] L M Griffith, S Z Pavletic, S J Lee, P J Martin, K R Schultz, and G B Vogels-ang. Chronic Graft-versus-Host Disease–implementation of the National Insti-tutes of Health Consensus Criteria for Clinical Trials. Biol Blood Marrow Transplant,14(4):379–384, Apr 2008.

[85] V Groh, R Rhinehart, J Randolph-Habecker, M S Topp, S R Riddell, and T Spies.Costimulation of CD8alphabeta T cells by NKG2D via engagement by MIC in-duced on virus-infected cells. Nat Immunol, 2(3):255–260, Mar 2001.

[86] V Groh, R Rhinehart, H Secrist, S Bauer, K H Grabstein, and T Spies. Broad tumor-associated expression and recognition by tumor-derived gamma delta T cells ofMICA and MICB. Proc Natl Acad Sci U S A, 96(12):6879–6884, Jun 1999.

[87] V Groh, A Steinle, S Bauer, and T Spies. Recognition of stress-induced MHCmolecules by intestinal epithelial gammadelta T cells. Science, 279(5357):1737–1740,Mar 1998.

[88] C Gründer, S van Dorp, S Hol, E Drent, T Straetemans, S Heijhuurs, K Schol-ten, W Scheper, Z Sebestyen, A Martens, R Strong, and J Kuball. γ9 andδ2CDR3 domains regulate functional avidity of T cells harboring γ9δ2TCRs. Blood,120(26):5153–5162, Dec 2012.

[89] W Haas, P Pereira, and S Tonegawa. Gamma/delta cells. Annu Rev Immunol,11:637–685, 1993.

[90] F Halary, V Pitard, D Dlubek, R Krzysiek, H de la Salle, P Merville, C Dromer,D Emilie, J F Moreau, and J Déchanet-Merville. Shared reactivity of Vdelta2(neg)gammadelta T cells against cytomegalovirus-infected cells and tumor intestinalepithelial cells. J Exp Med, 201(10):1567–1578, May 2005.

[91] G Hale and H Waldmann. Control of graft-versus-host disease and graft rejectionby T cell depletion of donor and recipient with Campath-1 antibodies. Resultsof matched sibling transplants for malignant diseases. Bone Marrow Transplant,13(5):597–611, May 1994.

[92] P Häusermann, R B Walter, J Halter, B C Biedermann, A Tichelli, P Itin, and A Grat-wohl. Cutaneous graft-versus-host disease: a guide for the dermatologist. Derma-tology, 216(4):287–304, 2008.

[93] A H Hautmann, D Wolff, J Hahn, M Edinger, N Schirmer, J Ammer, B Holler,K Landfried, M G Hautmann, N Ahrens, P Ugocsai, R Andreesen, and E Holler.Extracorporeal photopheresis in 62 patients with acute and chronic GVHD: resultsof treatment with the COBE Spectra System. Bone Marrow Transplant, 48(3):439–445,Mar 2013.

137

[94] M Hinchcliff, C C Huang, W Ishida, F Fang, J Lee, N Jafari, M Wilkes, S Bhat-tacharyya, E Leof, and J Varga. Imatinib mesylate causes genome-wide transcrip-tional changes in systemic sclerosis fibroblasts in vitro. Clin Exp Rheumatol, 30(2Suppl 71):86–96, Mar-Apr 2012.

[95] T Hinz, D Wesch, F Halary, S Marx, A Choudhary, B Arden, O Janssen, M Bon-neville, and D Kabelitz. Identification of the complete expressed human TCR Vgamma repertoire by flow cytometry. Int Immunol, 9(8):1065–1072, Aug 1997.

[96] M Ho, P Tongtawe, J Kriangkum, T Wimonwattrawatee, K Pattanapanyasat, L Bry-ant, J Shafiq, P Suntharsamai, S Looareesuwan, H K Webster, and J F Elliott.Polyclonal expansion of peripheral gamma delta T cells in human Plasmodiumfalciparum malaria. Infect Immun, 62(3):855–862, Mar 1994.

[97] S N Ho, H D Hunt, R M Horton, J K Pullen, and L R Pease. Site-directed mutagen-esis by overlap extension using the polymerase chain reaction. Gene, 77(1):51–59,Apr 1989.

[98] W Y Ho, H N Nguyen, M Wolfl, J Kuball, and P D Greenberg. In vitro methodsfor generating CD8+ T-cell clones for immunotherapy from the naïve repertoire. JImmunol Methods, 310(1-2):40–52, Mar 2006.

[99] W Hobo, K Broen, W J van der Velden, A Greupink-Draaisma, N Adisty,Y Wouters, M Kester, H Fredrix, J H Jansen, B van der Reijden, J H Falkenburg,T de Witte, F Preijers, T Schattenberg, T Feuth, N M Blijlevens, N Schaap, andH Dolstra. Association of disparities in known minor histocompatibility antigenswith relapse-free survival and graft-versus-host disease after allogeneic stem celltransplantation. Biol Blood Marrow Transplant, 19(2):274–282, Feb 2013.

[100] M C Hogenes, S van Dorp, J van Kuik, F R Monteiro, N ter Hoeve, M R van Dijk,A C Martens, and R A de Weger. Histological assessment of the sclerotic graft-versus-host response in the humanized RAG2-/-γc-/- mouse model. Biol BloodMarrow Transplant, 18(7):1023–1035, Jul 2012.

[101] P D Holler, P O Holman, E V Shusta, S O’Herrin, K D Wittrup, and D M Kranz. Invitro evolution of a T cell receptor with high affinity for peptide/MHC. Proc NatlAcad Sci U S A, 97(10):5387–5392, May 2000.

[102] K Hudspeth, M Fogli, D V Correia, J Mikulak, A Roberto, S Della Bella, B Silva-Santos, and D Mavilio. Engagement of NKp30 on VÎt’1 T cells induces the pro-duction of CCL3, CCL4, and CCL5 and suppresses HIV-1 replication. Blood,119(17):4013–4016, Apr 2012.

[103] C Huisman, R A de Weger, L de Vries, M G Tilanus, and L F Verdonck. Chimerismanalysis within 6 months of allogeneic stem cell transplantation predicts relapsein acute myeloid leukemia. Bone Marrow Transplant, 39(5):285–291, Mar 2007.

[104] H Ihn, K Yamane, M Kubo, and K Tamaki. Blockade of endogenous transforminggrowth factor beta signaling prevents up-regulated collagen synthesis in sclero-derma fibroblasts: association with increased expression of transforming growthfactor beta receptors. Arthritis Rheum, 44(2):474–480, Feb 2001.

138

BIBLIOGRAPHY

[105] C A Jacobson and J Ritz. B-cell-directed therapy for chronic graft-versus-host dis-ease. Haematologica, 95(11):1811–1813, Nov 2010.

[106] B Jena, G Dotti, and L J Cooper. Redirecting T-cell specificity by introducing atumor-specific chimeric antigen receptor. Blood, 116(7):1035–1044, Aug 2010.

[107] L A Johnson, R A Morgan, M E Dudley, L Cassard, J C Yang, M S Hughes, U SKammula, R E Royal, R M Sherry, J R Wunderlich, C C Lee, N P Restifo, S LSchwarz, A P Cogdill, R J Bishop, H Kim, C C Brewer, S F Rudy, C VanWaes, J LDavis, A Mathur, R T Ripley, D A Nathan, C M Laurencot, and S A Rosenberg.Gene therapy with human and mouse T-cell receptors mediates cancer regressionand targets normal tissues expressing cognate antigen. Blood, 114(3):535–546, Jul2009.

[108] H Jonuleit, U Kühn, G Müller, K Steinbrink, L Paragnik, E Schmitt, J Knop, andA H Enk. Pro-inflammatory cytokines and prostaglandins induce maturation ofpotent immunostimulatory dendritic cells under fetal calf serum-free conditions.Eur J Immunol, 27(12):3135–3142, Dec 1997.

[109] D Kabelitz. Small molecules for the activation of human gammadelta T cell re-sponses against infection. Recent Pat Antiinfect Drug Discov, 3(1):1–9, Jan 2008.

[110] D Kabelitz, A Glatzel, and D Wesch. Antigen recognition by human gammadeltaT lymphocytes. Int Arch Allergy Immunol, 122(1):1–7, May 2000.

[111] D Kabelitz, D Wesch, and W He. Perspectives of gammadelta T cells in tumorimmunology. Cancer Res, 67(1):5–8, Jan 2007.

[112] P Kebriaei, L Isola, E Bahceci, K Holland, S Rowley, J McGuirk, M Devetten,J Jansen, R Herzig, M Schuster, R Monroy, and J Uberti. Adult human mesen-chymal stem cells added to corticosteroid therapy for the treatment of acute graft-versus-host disease. Biol Blood Marrow Transplant, 15(7):804–811, Jul 2009.

[113] P Kebriaei, R M Saliba, C Ma, C Ippoliti, D R Couriel, M de Lima, S Giralt, M HQazilbash, J L Gajewski, C S Ha, R E Champlin, and I F Khouri. Allogeneichematopoietic stem cell transplantation after rituximab-containing myeloablativepreparative regimen for acute lymphoblastic leukemia. Bone Marrow Transplant,38(3):203–209, Aug 2006.

[114] H J Kim, W S Min, B S Cho, K S Eom, Y J Kim, C K Min, S Lee, S G Cho, J Y Jin, J WLee, and C C Kim. Successful prevention of acute graft-versus-host disease usinglow-dose antithymocyte globulin after mismatched, unrelated, hematopoietic stemcell transplantation for acute myelogenous leukemia. Biol Blood Marrow Transplant,15(6):704–717, Jun 2009.

[115] S J Kim, J W Lee, C W Jung, C K Min, B Cho, H J Shin, J S Chung, H Kim, W S Lee,Y D Joo, D H Yang, H Kook, H J Kang, H S Ahn, S S Yoon, S K Sohn, Y H Min, W SMin, H S Park, and J H Won. Weekly rituximab followed by monthly rituximabtreatment for steroid-refractory chronic graft-versus-host disease: results from aprospective, multicenter, phase II study. Haematologica, 95(11):1935–1942, Nov 2010.

139

[116] S J Kim and J H Won. B cell homeostasis and the development of chronic graft-versus-host disease: implications for B cell-depleting therapy. Leuk Lymphoma,53(1):19–25, Jan 2012.

[117] F M Kloosterboer, S A van Luxemburg-Heijs, R A van Soest, H M van Egmond,A M Barbui, M P Strijbosch, R Willemze, and J H Falkenburg. Minor histocompat-ibility antigen-specific T cells with multiple distinct specificities can be isolated bydirect cloning of IFNgamma-secreting T cells from patients with relapsed leukemiaresponding to donor lymphocyte infusion. Leukemia, 19(1):83–90, Jan 2005.

[118] A Knight, A J Madrigal, S Grace, J Sivakumaran, P Kottaridis, S Mackinnon, P JTravers, and M W Lowdell. The role of Vδ2-negative γδ T cells during cytomeg-alovirus reactivation in recipients of allogeneic stem cell transplantation. Blood,116(12):2164–2172, Sep 2010.

[119] Y Kong, W Cao, X Xi, C Ma, L Cui, and W He. The NKG2D ligand ULBP4binds to TCRgamma9/delta2 and induces cytotoxicity to tumor cells through bothTCRgammadelta and NKG2D. Blood, 114(2):310–317, Jul 2009.

[120] P Krishnamurthy, V T Potter, L D Barber, A G Kulasekararaj, Z Y Lim, R M Pearce,H de Lavallade, M Kenyon, R M Ireland, J C Marsh, S Devereux, A Pagliuca, andG J Mufti. Outcome of Donor Lymphocyte Infusion after T Cell-depleted Allogen-eic Hematopoietic Stem Cell Transplantation for Acute Myelogenous Leukemiaand Myelodysplastic Syndromes. Biol Blood Marrow Transplant, 19(4):562–568, Apr2013.

[121] N Kröger, A Shimoni, M Zagrivnaja, F Ayuk, M Lioznov, H Schieder, H Renges,B Fehse, T Zabelina, A Nagler, and A R Zander. Low-dose thalidomide and donorlymphocyte infusion as adoptive immunotherapy after allogeneic stem cell trans-plantation in patients with multiple myeloma. Blood, 104(10):3361–3363, Nov 2004.

[122] N Kröger, T Zabelina, W Krüger, H Renges, N Stute, J Rischewski, S Sonnenberg,F Ayuk, F Tögel, U Schade, H Fiegel, R Erttmann, C Löliger, and A R Zander.In vivo T cell depletion with pretransplant anti-thymocyte globulin reduces graft-versus-host disease without increasing relapse in good risk myeloid leukemia pa-tients after stem cell transplantation from matched related donors. Bone MarrowTransplant, 29(8):683–689, Apr 2002.

[123] J Kuball, M L Dossett, M Wolfl, W Y Ho, R H Voss, C Fowler, and P D Greenberg.Facilitating matched pairing and expression of TCR chains introduced into humanT cells. Blood, 109(6):2331–2338, Mar 2007.

[124] J Kuball, F W Schmitz, R H Voss, E A Ferreira, R Engel, P Guillaume, S Strand,P Romero, C Huber, L A Sherman, and M Theobald. Cooperation of humantumor-reactive CD4+ and CD8+ T cells after redirection of their specificity by ahigh-affinity p53A2.1-specific TCR. Immunity, 22(1):117–129, Jan 2005.

[125] J Kuball, M Theobald, E A Ferreira, G Hess, J Burg, G Maccagno, A P Barreiros,S Lüth, C C Schimanski, M Schuchmann, A Schwarting, M Neurath, G Otto, P RGalle, and A W Lohse. Control of organ transplant-associated graft-versus-host

140

BIBLIOGRAPHY

disease by activated host lymphocyte infusions. Transplantation, 78(12):1774–1779,Dec 2004.

[126] J Kuball, S F Wen, J Leissner, D Atkins, P Meinhardt, E Quijano, H Engler,B Hutchins, D C Maneval, M J Grace, M A Fritz, S Störkel, J W Thüroff, C Huber,and M Schuler. Successful adenovirus-mediated wild-type p53 gene transfer in pa-tients with bladder cancer by intravesical vector instillation. J Clin Oncol, 20(4):957–965, Feb 2002.

[127] M Kubo, H Ihn, K Yamane, and K Tamaki. Upregulated expression of transforminggrowth factor-beta receptors in dermal fibroblasts of skin sections from patientswith systemic sclerosis. J Rheumatol, 29(12):2558–2564, Dec 2002.

[128] X Lafarge, P Merville, M C Cazin, F Bergé, L Potaux, J F Moreau, and J Déchanet-Merville. Cytomegalovirus infection in transplant recipients resolves when circu-lating gammadelta T lymphocytes expand, suggesting a protective antiviral role. JInfect Dis, 184(5):533–541, Sep 2001.

[129] K Le Blanc, I Rasmusson, B Sundberg, C Götherström, M Hassan, M Uzunel, andO Ringdén. Treatment of severe acute graft-versus-host disease with third partyhaploidentical mesenchymal stem cells. Lancet, 363(9419):1439–1441, May 2004.

[130] S J Lee. New approaches for preventing and treating chronic graft-versus-hostdisease. Blood, 105(11):4200–4206, Jun 2005.

[131] S J Lee, J P Klein, A J Barrett, O Ringden, J H Antin, J Y Cahn, M H Carabasi,R P Gale, S Giralt, G A Hale, O Ilhan, P L McCarthy, G Socie, L F Verdonck,D J Weisdorf, and M M Horowitz. Severity of chronic graft-versus-host disease:association with treatment-related mortality and relapse. Blood, 100(2):406–414, Jul2002.

[132] S J Lee, G Vogelsang, and M E Flowers. Chronic graft-versus-host disease. BiolBlood Marrow Transplant, 9(4):215–233, Apr 2003.

[133] A M Leen, G D Myers, U Sili, M H Huls, H Weiss, K S Leung, G Carrum, R AKrance, C C Chang, J J Molldrem, A P Gee, M K Brenner, H E Heslop, C M Rooney,and C M Bollard. Monoculture-derived T lymphocytes specific for multiple virusesexpand and produce clinically relevant effects in immunocompromised individu-als. Nat Med, 12(10):1160–1166, Oct 2006.

[134] M. P. Lefranc. IMGT, the international ImMunoGeneTics database. Nucleic AcidsRes., 31(1):307–310, Jan 2003.

[135] A J Leishman, O V Naidenko, A Attinger, F Koning, C J Lena, Y Xiong, H C Chang,E Reinherz, M Kronenberg, and H Cheroutre. T cell responses modulated throughinteraction between CD8alphaalpha and the nonclassical MHC class I molecule,TL. Science, 294(5548):1936–1939, Nov 2001.

[136] D S Leslie, M S Vincent, F M Spada, H Das, M Sugita, C T Morita, and M BBrenner. CD1-mediated gamma/delta T cell maturation of dendritic cells. J ExpMed, 196(12):1575–1584, Dec 2002.

141

[137] H M Lokhorst. No RIC in high-risk myeloma? Blood, (107):3420–3421, 2006.

[138] H M Lokhorst, A Schattenberg, J J Cornelissen, L L Thomas, and L F Verdonck.Donor leukocyte infusions are effective in relapsed multiple myeloma after allo-geneic bone marrow transplantation. Blood, 90(10):4206–4211, Nov 1997.

[139] H M Lokhorst, C M Segeren, L F Verdonck, B van der Holt, R Raymakers, M Hvan Oers, R M Barge, H C Schouten, P H Westveer, M M Steijaert, J J Cornelissen,P Sonneveld, and Dutch-Belgian Hemato-Oncology Cooperative Group. PartiallyT-cell-depleted allogeneic stem-cell transplantation for first-line treatment of mul-tiple myeloma: a prospective evaluation of patients treated in the phase III studyHOVON 24 MM. J Clin Oncol, 21(9):1728–1733, May 2003.

[140] H M Lokhorst, K Wu, L F Verdonck, L L Laterveer, N W van de Donk, M H vanOers, J J Cornelissen, and A V Schattenberg. The occurrence of graft-versus-hostdisease is the major predictive factor for response to donor lymphocyte infusionsin multiple myeloma. Blood, 103(11):4362–4364, Jun 2004.

[141] E O Long. Tumor cell recognition by natural killer cells. Semin Cancer Biol, 12(1):57–61, Feb 2002.

[142] M W Lowdell, R Craston, D Samuel, M E Wood, E O’Neill, V Saha, and H GPrentice. Evidence that continued remission in patients treated for acute leukaemiais dependent upon autologous natural killer cells. Br J Haematol, 117(4):821–827,Jun 2002.

[143] M W Lowdell, N Ray, R Craston, T Corbett, M Deane, and H G Prentice. The invitro detection of anti-leukaemia-specific cytotoxicity after autologous bone mar-row transplantation for acute leukaemia. Bone Marrow Transplant, 19(9):891–897,May 1997.

[144] M J Maeurer, D Martin, W Walter, K Liu, L Zitvogel, K Halusczcak, H Rabinowich,R Duquesnoy, W Storkus, and M T Lotze. Human intestinal Vdelta1+ lymphocytesrecognize tumor cells of epithelial origin. J Exp Med, 183(4):1681–1696, Apr 1996.

[145] L Magro, B Catteau, V Coiteux, B Bruno, J P Jouet, and I Yakoub-Agha. Efficacy ofimatinib mesylate in the treatment of refractory sclerodermatous chronic GVHD.Bone Marrow Transplant, 42(11):757–760, Dec 2008.

[146] L Magro, M Mohty, B Catteau, V Coiteux, P Chevallier, L Terriou, J P Jouet,and I Yakoub-Agha. Imatinib mesylate as salvage therapy for refractory scleroticchronic graft-versus-host disease. Blood, 114(3):719–722, Jul 2009.

[147] D G Maloney, A J Molina, F Sahebi, K E Stockerl-Goldstein, B M Sandmaier,W Bensinger, B Storer, U Hegenbart, G Somlo, T Chauncey, B Bruno, F R Ap-pelbaum, K G Blume, S J Forman, P McSweeney, and R Storb. Allografting withnonmyeloablative conditioning following cytoreductive autografts for the treat-ment of patients with multiple myeloma. Blood, 102(9):3447–3454, Nov 2003.

142

BIBLIOGRAPHY

[148] D Maraninchi, E Gluckman, D Blaise, D Guyotat, B Rio, J L Pico, V Leblond,M Michallet, F Dreyfus, and N Ifrah. Impact of T-cell depletion on outcomeof allogeneic bone-marrow transplantation for standard-risk leukaemias. Lancet,2(8552):175–178, Jul 1987.

[149] V Marcu-Malina, S Heijhuurs, M van Buuren, L Hartkamp, S Strand, Z Sebestyen,K Scholten, A Martens, and J Kuball. Redirecting αβ T cells against cancer cells bytransfer of a broadly tumor-reactive γδT-cell receptor. Blood, 118(1):50–59, Jul 2011.

[150] V Marcu-Malina, S van Dorp, and J Kuball. Re-targeting T-cells against cancer bygene-transfer of tumor-reactive receptors. Expert Opin Biol Ther, 9(5):579–591, May2009.

[151] K J Martires, K Baird, S M Steinberg, L Grkovic, G O Joe, K M Williams, S AMitchell, M Datiles, F T Hakim, S Z Pavletic, and E W Cowen. Sclerotic-typechronic GVHD of the skin: clinical risk factors, laboratory markers, and burden ofdisease. Blood, 118(15):4250–4257, Oct 2011.

[152] S R Mattarollo, T Kenna, M Nieda, and A J Nicol. Chemotherapy and zoledronatesensitize solid tumour cells to Vgamma9Vdelta2 T cell cytotoxicity. Cancer ImmunolImmunother, 56(8):1285–1297, Aug 2007.

[153] C C Matte, J Liu, J Cormier, B E Anderson, I Athanasiadis, D Jain, J McNiff, andW D Shlomchik. Donor APCs are required for maximal GVHD but not for GVL.Nat Med, 10(9):987–992, Sep 2004.

[154] L L McCormick, Y Zhang, E Tootell, and A C Gilliam. Anti-TGF-beta treatmentprevents skin and lung fibrosis in murine sclerodermatous graft-versus-host dis-ease: a model for human scleroderma. J Immunol, 163(10):5693–5699, Nov 1999.

[155] Z McIver, N Stephens, A Grim, and A J Barrett. Rituximab administration within6 months of T cell-depleted allogeneic SCT is associated with prolonged life-threatening cytopenias. Biol Blood Marrow Transplant, 16(11):1549–1556, Nov 2010.

[156] F A Mendoza and S A Jiménez. Tyrosine kinase inhibitor therapy for systemicsclerosis: Quo Vadis? Arthritis Rheum, 63(11):3199–3203, Nov 2011.

[157] F Miyagawa, Y Tanaka, S Yamashita, B Mikami, K Danno, M Uehara, andN Minato. Essential contribution of germline-encoded lysine residues inJgamma1.2 segment to the recognition of nonpeptide antigens by human gam-madelta T cells. J Immunol, 167(12):6773–6779, Dec 2001.

[158] K Miyazaki. MEGAWHOP cloning: a method of creating random mutagenesislibraries via megaprimer PCR of whole plasmids. Methods Enzymol, 498:399–406,2011.

[159] M Mohty, O Avinens, C Faucher, P Viens, D Blaise, and J F Eliaou. Predictivefactors and impact of full donor T-cell chimerism after reduced intensity condi-tioning allogeneic stem cell transplantation. Haematologica, 92(7):1004–1006, Jul2007.

143

[160] J Molldrem, S Dermime, K Parker, Y Z Jiang, D Mavroudis, N Hensel,P Fukushima, and A J Barrett. Targeted T-cell therapy for human leukemia: cyto-toxic T lymphocytes specific for a peptide derived from proteinase 3 preferentiallylyse human myeloid leukemia cells. Blood, 88(7):2450–2457, Oct 1996.

[161] A Moretta, E Ciccone, G Pantaleo, G Tambussi, C Bottino, G Melioli, M C Mingari,and L Moretta. Surface molecules involved in the activation and regulation of T ornatural killer lymphocytes in humans. Immunol Rev, 111:145–175, Oct 1989.

[162] R A Morgan, M E Dudley, J R Wunderlich, M S Hughes, J C Yang, R M Sherry,R E Royal, S L Topalian, U S Kammula, N P Restifo, Z Zheng, A Nahvi, C Rde Vries, L J Rogers-Freezer, S A Mavroukakis, and S A Rosenberg. Cancer re-gression in patients after transfer of genetically engineered lymphocytes. Science,314(5796):126–129, Oct 2006.

[163] C T Morita, H K Lee, H Wang, H Li, R A Mariuzza, and Y Tanaka. Structuralfeatures of nonpeptide prenyl pyrophosphates that determine their antigenicityfor human gamma delta T cells. J Immunol, 167(1):36–41, Jul 2001.

[164] S Nedellec, M Bonneville, and E Scotet. Human Vgamma9Vdelta2 T cells: fromsignals to functions. Semin Immunol, 22(4):199–206, Aug 2010.

[165] V H Nguyen, S Shashidhar, D S Chang, L Ho, N Kambham, M Bachmann, J MBrown, and R S Negrin. The impact of regulatory T cells on T-cell immunityfollowing hematopoietic cell transplantation. Blood, 111(2):945–953, Jan 2008.

[166] W G Nichols, L Corey, T Gooley, C Davis, and M Boeckh. High risk of deathdue to bacterial and fungal infection among cytomegalovirus (CMV)-seronegativerecipients of stem cell transplants from seropositive donors: evidence for indirecteffects of primary CMV infection. J Infect Dis, 185(3):273–282, Feb 2002.

[167] A J Nicol, H Tokuyama, S R Mattarollo, T Hagi, K Suzuki, K Yokokawa, andM Nieda. Clinical evaluation of autologous gamma delta T cell-based immuno-therapy for metastatic solid tumours. Br J Cancer, 105(6):778–786, Sep 2011.

[168] A Olivieri, F Locatelli, M Zecca, A Sanna, M Cimminiello, R Raimondi, G Gini,N Mordini, A Balduzzi, P Leoni, A Gabrielli, and A Bacigalupo. Imatinib for re-fractory chronic graft-versus-host disease with fibrotic features. Blood, 114(3):709–718, Jul 2009.

[169] J Olivieri, S Coluzzi, I Attolico, and A Olivieri. Tirosin kinase inhibitors in chronicgraft versus host disease: from bench to bedside. ScientificWorldJournal, 11:1908–1931, 2011.

[170] S Ozawa, C Nakaseko, M Nishimura, A Maruta, R Cho, C Ohwada, H Saka-maki, H Sao, S Mori, S Okamoto, K Miyamura, S Kato, T Kawase, Y Morishima,Y Kodera, and Japan Marrow Donor Program. Chronic graft-versus-host diseaseafter allogeneic bone marrow transplantation from an unrelated donor: incidence,risk factors and association with relapse. A report from the Japan Marrow DonorProgram. Br J Haematol, 137(2):142–151, Apr 2007.

144

BIBLIOGRAPHY

[171] S Park, M Labopin, I Yakoub-Agha, J Delaunay, N Dhedin, E Deconinck, M Michal-let, M Robin, T De Revel, M Bernard, N Vey, B Lioure, S Lapusan, R Tabrizi, J HBourhis, A Huynh, Y Beguin, G Socié, F Dreyfus, P Fenaux, and M Mohty. Al-logeneic stem cell transplantation for chronic myelomonocytic leukemia: a reportfrom the Societe Francaise de Greffe de Moelle et de Therapie Cellulaire. Eur JHaematol, Jan 2013.

[172] F Patriarca, C Skert, A Sperotto, F Zaja, E Falleti, R Mestroni, F Kikic, E Calistri,C Filì, A Geromin, M Cerno, and R Fanin. The development of autoantibodies afterallogeneic stem cell transplantation is related with chronic graft-vs-host diseaseand immune recovery. Exp Hematol, 34(3):389–396, Mar 2006.

[173] S Z Pavletic, P Martin, S J Lee, S Mitchell, D Jacobsohn, E W Cowen, M LTurner, G Akpek, A Gilman, G McDonald, M Schubert, A Berger, P Bross, J WChien, D Couriel, J P Dunn, J Fall-Dickson, A Farrell, M E Flowers, H Greinix,S Hirschfeld, L Gerber, S Kim, R Knobler, P A Lachenbruch, F W Miller, B Mittle-man, E Papadopoulos, S K Parsons, D Przepiorka, M Robinson, M Ward, B Reeve,L G Rider, H Shulman, K R Schultz, D Weisdorf, G B Vogelsang, and ResponseCriteria Working Group. Measuring therapeutic response in chronic graft-versus-host disease: National Institutes of Health Consensus Development Project on Cri-teria for Clinical Trials in Chronic Graft-versus-Host Disease: IV. Response CriteriaWorking Group report. Biol Blood Marrow Transplant, 12(3):252–266, Mar 2006.

[174] D Pende, G M Spaggiari, S Marcenaro, S Martini, P Rivera, A Capobianco, M Falco,E Lanino, I Pierri, R Zambello, A Bacigalupo, M C Mingari, A Moretta, and L Mor-etta. Analysis of the receptor-ligand interactions in the natural killer-mediated lysisof freshly isolated myeloid or lymphoblastic leukemias: evidence for the involve-ment of the Poliovirus receptor (CD155) and Nectin-2 (CD112). Blood, 105(5):2066–2073, Mar 2005.

[175] F L Pietersma, S van Dorp, M C Minnema, J Kuball, E Meijer, R Schuurman, andD van Baarle. Influence of donor cytomegalovirus (CMV) status on severity ofviral reactivation after allogeneic stem cell transplantation in CMV-seropositiverecipients. Clin Infect Dis, 52(7):144–148, Apr 2011.

[176] V Pitard, D Roumanes, X Lafarge, L Couzi, I Garrigue, M E Lafon, P Merville,J F Moreau, and J Déchanet-Merville. Long-term expansion of effector/memoryVdelta2-gammadelta T cells is a specific blood signature of CMV infection. Blood,112(4):1317–1324, Aug 2008.

[177] A Poggi, C Venturino, S Catellani, M Clavio, M Miglino, M Gobbi, A Steinle,P Ghia, S Stella, F Caligaris-Cappio, and M R Zocchi. Vdelta1 T lymphocytes fromB-CLL patients recognize ULBP3 expressed on leukemic B cells and up-regulatedby trans-retinoic acid. Cancer Res, 64(24):9172–9179, Dec 2004.

[178] I Puig-Pey, F Bohne, C Benítez, M López, M Martínez-Llordella, F Oppenheimer,J J Lozano, J González-Abraldes, G Tisone, A Rimola, and A Sánchez-Fueyo. Char-acterization of γδ T cell subsets in organ transplantation. Transpl Int, 23(10):1045–1055, Oct 2010.

145

[179] V Ratanatharathorn, L Ayash, C Reynolds, S Silver, P Reddy, M Becker, J L Ferrara,and J P Uberti. Treatment of chronic graft-versus-host disease with anti-CD20chimeric monoclonal antibody. Biol Blood Marrow Transplant, 9(8):505–511, Aug2003.

[180] V Ratanatharathorn, B Logan, D Wang, M Horowitz, J P Uberti, O Ringden, R PGale, H Khoury, M Arora, S Spellman, C Cutler, J Antin, M Bornhaüser, G Hale,L Verdonck, M Cairo, V Gupta, S Pavletic, and Center for International Blood andMarrow Transplant Research, Milwaukee, WI, USA. Prior rituximab correlateswith less acute graft-versus-host disease and better survival in B-cell lymphomapatients who received allogeneic peripheral blood stem cell transplantation. Br JHaematol, 145(6):816–824, Jun 2009.

[181] C Reynolds, V Ratanatharathorn, P Adams, T Braun, S Silver, L Ayash, E Carson,A Eisbruch, L A Dawson, K McDonagh, J Ferrara, and J Uberti. Allogeneic stemcell transplantation reduces disease progression compared to autologous trans-plantation in patients with multiple myeloma. Bone Marrow Transplant, 27(8):801–807, Apr 2001.

[182] J C Ribot, A debarros, and B Silva-Santos. Searching for "signal 2": costimulationrequirements of γδ T cells. Cell Mol Life Sci, 68(14):2345–2355, Jul 2011.

[183] S R Riddell and P D Greenberg. The use of anti-CD3 and anti-CD28 monoclonalantibodies to clone and expand human antigen-specific T cells. J Immunol Methods,128(2):189–201, Apr 1990.

[184] B Rincon-Orozco, V Kunzmann, P Wrobel, D Kabelitz, A Steinle, and T Herrmann.Activation of V gamma 9V delta 2 T cells by NKG2D. J Immunol, 175(4):2144–2151,Aug 2005.

[185] O Ringdén, M Labopin, N C Gorin, N Schmitz, U W Schaefer, H G Prentice,L Bergmann, J P Jouet, F Mandelli, D Blaise, L Fouillard, F Frassoni, and AcuteLeukaemia Working Party of the European Group for Blood and Marrow Trans-plantation. Is there a graft-versus-leukaemia effect in the absence of graft-versus-host disease in patients undergoing bone marrow transplantation for acute leuk-aemia? Br J Haematol, 111(4):1130–1137, Dec 2000.

[186] O Ringdén, M Uzunel, I Rasmusson, M Remberger, B Sundberg, H Lönnies, H UMarschall, A Dlugosz, A Szakos, Z Hassan, B Omazic, J Aschan, L Barkholt,and K Le Blanc. Mesenchymal stem cells for treatment of therapy-resistant graft-versus-host disease. Transplantation, 81(10):1390–1397, May 2006.

[187] A M Rouquette-Gally, D Boyeldieu, A C Prost, and E Gluckman. Autoimmunityafter allogeneic bone marrow transplantation. A study of 53 long-term-survivingpatients. Transplantation, 46(2):238–240, Aug 1988.

[188] V Rowe, T Banovic, K P MacDonald, R Kuns, A L Don, E S Morris, A C Burman,H M Bofinger, A D Clouston, and G R Hill. Host B cells produce IL-10 follow-ing TBI and attenuate acute GVHD after allogeneic bone marrow transplantation.Blood, 108(7):2485–2492, Oct 2006.

146

BIBLIOGRAPHY

[189] L Ruggeri, M Capanni, E Urbani, K Perruccio, W D Shlomchik, A Tosti, S Posati,D Rogaia, F Frassoni, F Aversa, M F Martelli, and A Velardi. Effectiveness of donornatural killer cell alloreactivity in mismatched hematopoietic transplants. Science,295(5562):2097–2100, Mar 2002.

[190] A M Russano, G Bassotti, E Agea, O Bistoni, A Mazzocchi, A Morelli, S A Porcelli,and F Spinozzi. CD1-restricted recognition of exogenous and self-lipid antigensby duodenal gammadelta+ T lymphocytes. J Immunol, 178(6):3620–3626, Mar 2007.

[191] A Sandstrom, L Scharf, G McRae, A J Hawk, S C Meredith, and E J Adams. ÎsÎt’ Tcell receptors recognize the non-classical major histocompatibility complex (MHC)molecule T22 via conserved anchor residues in a MHC peptide-like fashion. J BiolChem, 287(8):6035–6043, Feb 2012.

[192] S Sarantopoulos, K E Stevenson, H T Kim, C S Cutler, N S Bhuiya, M Schowalter,V T Ho, E P Alyea, J Koreth, B R Blazar, R J Soiffer, J H Antin, and J Ritz. AlteredB-cell homeostasis and excess BAFF in human chronic graft-versus-host disease.Blood, 113(16):3865–3874, Apr 2009.

[193] S Sarantopoulos, K E Stevenson, H T Kim, W S Washel, N S Bhuiya, C S Cutler, E PAlyea, V T Ho, R J Soiffer, J H Antin, and J Ritz. Recovery of B-cell homeostasisafter rituximab in chronic graft-versus-host disease. Blood, 117(7):2275–2283, Feb2011.

[194] W Scheper, S van Dorp, S Kersting, F Pietersma, C Lindemans, S Hol, S Heijhuurs,Z Sebestyen, C Gründer, V Marcu-Malina, A Marchant, C Donner, B Plachter,D Vermijlen, D van Baarle, and J Kuball. γδT cells elicited by CMV reactivationafter allo-SCT cross-recognize CMV and leukemia. Leukemia, Jan 2013.

[195] P Schneider, F MacKay, V Steiner, K Hofmann, J L Bodmer, N Holler, C Ambrose,P Lawton, S Bixler, H Acha-Orbea, D Valmori, P Romero, C Werner-Favre, R HZubler, J L Browning, and J Tschopp. BAFF, a novel ligand of the tumor necrosisfactor family, stimulates B cell growth. J Exp Med, 189(11):1747–1756, Jun 1999.

[196] K R Schultz, D B Miklos, D Fowler, K Cooke, J Shizuru, E Zorn, E Holler, J Ferrara,H Shulman, S J Lee, P Martin, A H Filipovich, M E Flowers, D Weisdorf, D Cour-iel, P A Lachenbruch, B Mittleman, G B Vogelsang, and S Z Pavletic. Towardbiomarkers for chronic graft-versus-host disease: National Institutes of Healthconsensus development project on criteria for clinical trials in chronic graft-versus-host disease: III. Biomarker Working Group Report. Biol Blood Marrow Transplant,12(2):126–137, Feb 2006.

[197] K R Schultz, J Paquet, S Bader, and K T HayGlass. Requirement for B cells in T cellpriming to minor histocompatibility antigens and development of graft-versus-host disease. Bone Marrow Transplant, 16(2):289–295, Aug 1995.

[198] E Scotet, L O Martinez, E Grant, R Barbaras, P Jenö, M Guiraud, B Monsarrat,X Saulquin, S Maillet, J P Estève, F Lopez, B Perret, X Collet, M Bonneville, andE Champagne. Tumor recognition following Vgamma9Vdelta2 T cell receptor

147

interactions with a surface F1-ATPase-related structure and apolipoprotein A-I.Immunity, 22(1):71–80, Jan 2005.

[199] Y Shen, D Zhou, L Qiu, X Lai, M Simon, L Shen, Z Kou, Q Wang, L Jiang, J Estep,R Hunt, M Clagett, P K Sehgal, Y Li, X Zeng, C T Morita, M B Brenner, N L Letvin,and Z W Chen. Adaptive immune response of Vgamma2Vdelta2+ T cells duringmycobacterial infections. Science, 295(5563):2255–2258, Mar 2002.

[200] A Shimabukuro-Vornhagen, M J Hallek, R F Storb, and M S von Bergwelt-Baildon. The role of B cells in the pathogenesis of graft-versus-host disease. Blood,114(24):4919–4927, Dec 2009.

[201] H M Shulman, G E Sale, K G Lerner, E A Barker, P L Weiden, K Sullivan, B Gal-lucci, E D Thomas, and R Storb. Chronic cutaneous graft-versus-host disease inman. Am J Pathol, 91(3):545–570, Jun 1978.

[202] H M Shulman, K M Sullivan, P L Weiden, G B McDonald, G E Striker, G E Sale,R Hackman, M S Tsoi, R Storb, and E D Thomas. Chronic graft-versus-host syn-drome in man. A long-term clinicopathologic study of 20 Seattle patients. Am JMed, 69(2):204–217, Aug 1980.

[203] G Socié. Chronic GVHD: B cells come of age. Blood, 117(7):2086–2087, Feb 2011.

[204] M L Sorror, M B Maris, B Storer, B M Sandmaier, R Diaconescu, C Flowers, D GMaloney, and R Storb. Comparing morbidity and mortality of HLA-matchedunrelated donor hematopoietic cell transplantation after nonmyeloablative andmyeloablative conditioning: influence of pretransplantation comorbidities. Blood,104(4):961–968, Aug 2004.

[205] R Spaapen and T Mutis. Targeting haematopoietic-specific minor histocompat-ibility antigens to distinguish graft-versus-tumour effects from graft-versus-hostdisease. Best Pract Res Clin Haematol, 21(3):543–557, Sep 2008.

[206] R Spaapen, K van den Oudenalder, R Ivanov, A Bloem, H Lokhorst, and T Mutis.Rebuilding human leukocyte antigen class II-restricted minor histocompatibilityantigen specificity in recall antigen-specific T cells by adoptive T cell receptor trans-fer: implications for adoptive immunotherapy. Clin Cancer Res, 13(13):4009–4015,Jul 2007.

[207] S Srivastava, A Lundqvist, and R W Childs. Natural killer cell immunotherapy forcancer: a new hope. Cytotherapy, 10(8):775–783, 2008.

[208] E F Srour, T Leemhuis, L Jenski, R Redmond, and J Jansen. Cytolytic activity of hu-man natural killer cell subpopulations isolated by four-color immunofluorescenceflow cytometric cell sorting. Cytometry, 11(3):442–446, 1990.

[209] T Stanislawski, R H Voss, C Lotz, E Sadovnikova, R A Willemsen, J Kuball, T Rup-pert, R L Bolhuis, C J Melief, C Huber, H J Stauss, and M Theobald. Circumventingtolerance to a human MDM2-derived tumor antigen by TCR gene transfer. Nat Im-munol, 2(10):962–970, Oct 2001.

148

BIBLIOGRAPHY

[210] Y Sun, I Tawara, T Toubai, and P Reddy. Pathophysiology of acute graft-versus-host disease: recent advances. Transl Res, 150(4):197–214, Oct 2007.

[211] S Svegliati, R Cancello, P Sambo, M Luchetti, P Paroncini, G Orlandini, G Disce-poli, R Paterno, M Santillo, C Cuozzo, S Cassano, E V Avvedimento, and A Gabri-elli. Platelet-derived growth factor and reactive oxygen species (ROS) regulate Rasprotein levels in primary human fibroblasts via ERK1/2. Amplification of ROS andRas in systemic sclerosis fibroblasts. J Biol Chem, 280(43):36474–36482, Oct 2005.

[212] S Svegliati, A Olivieri, N Campelli, M Luchetti, A Poloni, S Trappolini, G Mor-oncini, A Bacigalupo, P Leoni, E V Avvedimento, and A Gabrielli. Stimulatoryautoantibodies to PDGF receptor in patients with extensive chronic graft-versus-host disease. Blood, 110(1):237–241, Jul 2007.

[213] Y Takemoto, H Takatsuka, H Wada, A Mori, K Saheki, M Okada, S Tamura,Y Fujimori, T Okamoto, E Kakishita, and A Kanamaru. Evaluation ofCMV/human herpes virus-6 positivity in bronchoalveolar lavage fluids as earlydetection of acute GVHD following BMT: evidence of a significant relationship.Bone Marrow Transplant, 26(1):77–81, Jul 2000.

[214] Y Tanaka, S Sano, E Nieves, G De Libero, D Rosa, R L Modlin, M B Brenner, B RBloom, and C T Morita. Nonpeptide ligands for human gamma delta T cells. ProcNatl Acad Sci U S A, 91(17):8175–8179, Aug 1994.

[215] Liane CJ te Boome, Cristina M Mansilla, Caroline A Lindemans, Lotte E Van derWagen, Marloes Cuijpers, Ineke Slaper-Cortenbach, Henk Rozemuller, Eefke JPetersen, Eric Spierings, Marc Bierings, Jaap J Boelens, Nico Wulffraat, and Jur-gen H Kuball. Treatment of Steroid Resistant Grade II to IV Acute Gvhd by In-fusion of Mesenchymal Stroma Cells Expanded with Human Plasma and PlateletLysate - a Phase I/II Study. ASH Annual Meeting Abstracts, 120(21):736, 2012.

[216] T Teshima, K Nagafuji, H Henzan, K Miyamura, K Takase, M Hidaka, T Miyamoto,K Takenaka, K Akashi, and M Harada. Rituximab for the treatment ofcorticosteroid-refractory chronic graft-versus-host disease. Int J Hematol, 90(2):253–260, Sep 2009.

[217] J S Thompson, P Schneider, S L Kalled, L Wang, E A Lefevre, T G Cachero,F MacKay, S A Bixler, M Zafari, Z Y Liu, S A Woodcock, F Qian, M Batten, C Madry,Y Richard, C D Benjamin, J L Browning, A Tsapis, J Tschopp, and C Ambrose.BAFF binds to the tumor necrosis factor receptor-like molecule B cell maturationantigen and is important for maintaining the peripheral B cell population. J ExpMed, 192(1):129–135, Jul 2000.

[218] W J Urba and D L Longo. Redirecting T cells. N Engl J Med, 365(8):754–757, Aug2011.

[219] E M Urban, A I Chapoval, and C D Pauza. Repertoire development and the controlof cytotoxic/effector function in human gammadelta T cells. Clin Dev Immunol,2010:732893–732893, 2010.

149

[220] N W van de Donk, N Kröger, U Hegenbart, P Corradini, J F San Miguel, H Gold-schmidt, J A Perez-Simon, M Zijlmans, R A Raymakers, V Montefusco, F A Ayuk,M H van Oers, A Nagler, L F Verdonck, and H M Lokhorst. Prognostic factorsfor donor lymphocyte infusions following non-myeloablative allogeneic stem celltransplantation in multiple myeloma. Bone Marrow Transplant, 37(12):1135–1141,Jun 2006.

[221] N W van de Donk, N Kröger, U Hegenbart, P Corradini, J F San Miguel, H Gold-schmidt, J A Perez-Simon, M Zijlmans, R A Raymakers, V Montefusco, F A Ayuk,M H van Oers, A Nagler, L F Verdonck, and H M Lokhorst. Remarkable activityof novel agents bortezomib and thalidomide in patients not responding to donorlymphocyte infusions following nonmyeloablative allogeneic stem cell transplant-ation in multiple myeloma. Blood, 107(8):3415–3416, Apr 2006.

[222] D van der Harst, E Goulmy, J H Falkenburg, Y M Kooij-Winkelaar, S A vanLuxemburg-Heijs, H M Goselink, and A Brand. Recognition of minor histocom-patibility antigens on lymphocytic and myeloid leukemic cells by cytotoxic T-cellclones. Blood, 83(4):1060–1066, Feb 1994.

[223] L T van der Veken, R S Hagedoorn, M M van Loenen, R Willemze, J H Falkenburg,and M H Heemskerk. Alphabeta T-cell receptor engineered gammadelta T cellsmediate effective antileukemic reactivity. Cancer Res, 66(6):3331–3337, Mar 2006.

[224] J J van Dongen, A W Langerak, M Brüggemann, P A Evans, M Hummel, F LLavender, E Delabesse, F Davi, E Schuuring, R García-Sanz, J H van Krieken,J Droese, D González, C Bastard, H E White, M Spaargaren, M González, A Par-reira, J L Smith, G J Morgan, M Kneba, and E A Macintyre. Design and standard-ization of PCR primers and protocols for detection of clonal immunoglobulin andT-cell receptor gene recombinations in suspect lymphoproliferations: report of theBIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia, 17(12):2257–2317, Dec2003.

[225] S van Dorp, F Pietersma, M Wölfl, L F Verdonck, E J Petersen, H M Lokhorst,E Martens, M Theobald, D van Baarle, E Meijer, and J Kuball. Rituximab treatmentbefore reduced-intensity conditioning transplantation associates with a decreasedincidence of extensive chronic GVHD. Biol Blood Marrow Transplant, 15(6):671–678,Jun 2009.

[226] S van Dorp, H Resemann, L te Boome, F Pietersma, D van Baarle, F Gmelig-Meyling, R de Weger, E Petersen, M Minnema, H Lokhorst, S Ebeling, S J Beijn,E F Knol, M van Dijk, E Meijer, and J Kuball. The immunological phenotype ofrituximab-sensitive chronic graft-versus-host disease: a phase II study. Haematolo-gica, 96(9):1380–1384, Sep 2011.

[227] M M van Loenen, R S Hagedoorn, M G Kester, M Hoogeboom, R Willemze, J HFalkenburg, and M H Heemskerk. Kinetic preservation of dual specificity of cop-rogrammed minor histocompatibility antigen-reactive virus-specific T cells. CancerRes, 69(5):2034–2041, Mar 2009.

150

BIBLIOGRAPHY

[228] L F Verdonck, E J Petersen, H M Lokhorst, H K Nieuwenhuis, A W Dekker, M GTilanus, and R A de Weger. Donor leukocyte infusions for recurrent hematologicmalignancies after allogeneic bone marrow transplantation: impact of infused andresidual donor T cells. Bone Marrow Transplant, 22(11):1057–1063, Dec 1998.

[229] D Vermijlen, M Brouwer, C Donner, C Liesnard, M Tackoen, M Van Rysselberge,N Twité, M Goldman, A Marchant, and F Willems. Human cytomegalovirus elicitsfetal gammadelta T cell responses in utero. J Exp Med, 207(4):807–821, Apr 2010.

[230] M von Bonin, F Stölzel, A Goedecke, K Richter, N Wuschek, K Hölig,U Platzbecker, T Illmer, M Schaich, J Schetelig, A Kiani, R Ordemann, G Ehninger,M Schmitz, and M Bornhäuser. Treatment of refractory acute GVHD with third-party MSC expanded in platelet lysate-containing medium. Bone Marrow Trans-plant, 43(3):245–251, Feb 2009.

[231] R H Voss, J Kuball, and M Theobald. Designing TCR for cancer immunotherapy.Methods Mol Med, 109:229–256, 2005.

[232] H Wang, Z Fang, and C T Morita. Vgamma2Vdelta2 T Cell Receptor recognitionof prenyl pyrophosphates is dependent on all CDRs. J Immunol, 184(11):6209–6222,Jun 2010.

[233] E H Warren, N Fujii, Y Akatsuka, C N Chaney, J K Mito, K R Loeb, T A Gooley,M L Brown, K K Koo, K V Rosinski, S Ogawa, A Matsubara, F R Appelbaum,and S R Riddell. Therapy of relapsed leukemia after allogeneic hematopoietic celltransplantation with T cells specific for minor histocompatibility antigens. Blood,115(19):3869–3878, May 2010.

[234] M Wilhelm, V Kunzmann, S Eckstein, P Reimer, F Weissinger, T Ruediger, andH P Tony. Gammadelta T cells for immune therapy of patients with lymphoidmalignancies. Blood, 102(1):200–206, Jul 2003.

[235] C R Willcox, V Pitard, S Netzer, L Couzi, M Salim, T Silberzahn, J F Moreau,A C Hayday, B E Willcox, and J Déchanet-Merville. Cytomegalovirus and tumorstress surveillance by binding of a human γδ T cell antigen receptor to endothelialprotein C receptor. Nat Immunol, 13(9):872–879, Sep 2012.

[236] L C Wu, D S Tuot, D S Lyons, K C Garcia, and M M Davis. Two-step bindingmechanism for T-cell receptor recognition of peptide MHC. Nature, 418(6897):552–556, Aug 2002.

[237] T A Wynn. Common and unique mechanisms regulate fibrosis in various fibro-proliferative diseases. J Clin Invest, 117(3):524–529, Mar 2007.

[238] B Xu, J C Pizarro, M A Holmes, C McBeth, V Groh, T Spies, and R K Strong.Crystal structure of a gammadelta T-cell receptor specific for the human MHCclass I homolog MICA. Proc Natl Acad Sci U S A, 108(6):2414–2419, Feb 2011.

[239] C Xu, H Zhang, H Hu, H He, Z Wang, Y Xu, H Chen, W Cao, S Zhang, L Cui,D Ba, and W He. Gammadelta T cells recognize tumor cells via CDR3delta region.Mol Immunol, 44(4):302–310, Jan 2007.

151

[240] S Yang, Y Ji, L Gattinoni, L Zhang, Z Yu, N P Restifo, S A Rosenberg, and R AMorgan. Modulating the differentiation status of ex vivo-cultured anti-tumor Tcells using cytokine cocktails. Cancer Immunol Immunother, 62(4):727–736, Apr 2013.

[241] T Yi, D Zhao, C L Lin, C Zhang, Y Chen, I Todorov, T LeBon, F Kandeel, S Forman,and D Zeng. Absence of donor Th17 leads to augmented Th1 differentiation andexacerbated acute graft-versus-host disease. Blood, 112(5):2101–2110, Sep 2008.

[242] F Zaja, A Bacigalupo, F Patriarca, M Stanzani, M T Van Lint, C Filì, R Scimè, G Mi-lone, M Falda, C Vener, D Laszlo, P E Alessandrino, F Narni, S Sica, A Olivieri,A Sperotto, A Bosi, F Bonifazi, R Fanin, and GITMO (Gruppo Italiano TrapiantoMidollo Osseo). Treatment of refractory chronic GVHD with rituximab: a GITMOstudy. Bone Marrow Transplant, 40(3):273–277, Aug 2007.

[243] R Zeiser and J Finke. Allogeneic haematopoietic cell transplantation for multiplemyeloma: reducing transplant-related mortality while harnessing the graft-versus-myeloma effect. Eur J Cancer, 42(11):1601–1611, Jul 2006.

[244] D Zhao, C Zhang, T Yi, C L Lin, I Todorov, F Kandeel, S Forman, and D Zeng. Invivo-activated CD103+CD4+ regulatory T cells ameliorate ongoing chronic graft-versus-host disease. Blood, 112(5):2129–2138, Sep 2008.

[245] H Zhou, M Guo, C Bian, Z Sun, Z Yang, Y Zeng, H Ai, and R C Zhao. Efficacy ofbone marrow-derived mesenchymal stem cells in the treatment of sclerodermatouschronic graft-versus-host disease: clinical report. Biol Blood Marrow Transplant,16(3):403–412, Mar 2010.

152

153

Hoofdstuk 8

Nederlandse samenvatting

Introductie

Allogene stamcel transplantatie (allo-SCT) wordt gebruikt in de behandeling

van verschillende hematologische maligniteiten zoals het multipel myeloom

(MM) en acute myeloide leukemie (AML). Het effect van de behandeling wordt

voor een deel verzorgd door de conditionering die wordt gegeven voor de allo-

SCT. Deze bestaat uit een hoge dosis chemotherapie en bestraling, waardoor de

kankercellen, maar ook het immuunsysteem van de patiënt vernietigd worden.

Hierna volgt de transplantatie van donor stamcellen, die uit zullen groeien tot

een nieuw immuunsysteem. De grootste bijdrage aan het slagen van de be-

handeling wordt geleverd door het zogeheten graft-versus-tumor effect. Dit

houdt in dat het getransplanteerde immuunsysteem de mogelijk overgebleven

kankercellen aanvalt en zorgt hierdoor dat de patiënt ook na de behandeling

kankervrij blijft. Het graft-versus-tumor effect ontstaat doordat de patiënt en

de donor verschillen, ondanks dat hun HLA-type gematched wordt. Hierdoor

zal het immuunsysteem van de donor de kankercellen van de patiënt als “niet

eigen” herkennen en deze aanvallen.

Er zijn drie belangrijke oorzaken van het falen van de behandeling met allo-SCT.

1. Ondanks het graft-versus-tumor effect krijgen veel patiënten (10-45%) relapse

van de ziekte.

2. Er is een belangrijke sterfte door de behandeling zelf (20-32%). Deze wordt

met name veroorzaakt door een bijwerking van het graft-versus-tumor effect,

graft-versus-host ziekte (GVHD). Hierbij worden niet alleen kankercellen,

155

maar ook de normale cellen van het lichaam van de patiënt als “niet eigen”

herkent en aangevallen door het nieuwe immuunsysteem. Er ontstaat een

afstotingsreactie van het lichaam van de patiënt. Dit kan kort na de allo-SCT

ontstaan (acute GVHD), maar ook op langere termijn zorgen voor invaliditeit

en sterfte (chronische GVHD).

3. Er ontstaan na de allo-SCT, door de immunosuppressiva die gebruikt moeten

worden om GVHD te voorkomen, veel infectieuze complicaties, met name

reactivaties van herpesvirussen zoals het Epstein Barr virus (EBV) en cyto-

megalovirus (CMV).

Verbetering van de allo-SCT is nodig om deze complicaties te verminderen.

Deze verbetering kan bereikt worden door de complicaties beter te behandelen.

Daarnaast kan men door betere kennis van de precieze mechanismen die ten

grondslag liggen aan het graft-versus-tumor effect, aan GVHD en het ontstaan

van infecties, nieuwe strategieën ontwikkelen om de effectiviteit van de allo-

SCT te vergroten en de complicaties te voorkomen. Dit proefschrift bestaat

uit drie delen. In deel 1 laten we het effect van allo-SCT zien in patiënten

met een multipel myeloom. In deel 2 gaan we in op de rol van B cellen in de

ontstaanswijze, behandeling en het voorkomen van chronische GVHD. In deel 3

gebruiken we γδT cellen als nieuwe strategie om leukemie en virale reactivaties

te controleren en GVHD te voorkomen.

(I) De therapeutische waarde van allo-SCT

Bij patiënten met een multipel myeloom zijn hun plasmacellen maligne ontaard.

Deze cellen houden zich met name op in het beenmerg. Er zijn verschillende

(chemotherapeutische) behandelingen, die deze ziekte kunnen onderdrukken

en een patiënt zelfs langdurig ziekte vrij kunnen laten zijn. De enige behan-

deling die de ziekte echt kan genezen, lijkt een allo-SCT te zijn. In deel 1 van

dit proefschrift werden de uitkomsten van de behandeling met een allo-SCT

van 63 patiënten met een multipel myeloom in het UMC Utrecht retrospectief

geanalyseerd. We hebben gekeken naar de respons op de behandeling (afname

van de ziekte activiteit), de overleving na allo-SCT, de ziekte vrije overleving

(als maat voor de hoeveelheid patiënten met een relapse van de ziekte), naar de

156

HOOFDSTUK 8. NEDERLANDSE SAMENVATTING

sterfte door de behandeling zelf en naar de patiëntengroep voor wie allo-SCT

nu het meest effectief is. Behandeling met allo-SCT zorgde voor een afname

van de ziekte activiteit bij 90% van de patiënten. Na 2 jaar was 84% van de

patiënten nog in leven. Helaas kreeg 42% van de patiënten in de 2 jaar follow-

up een relapse van de ziekte. Sterfte door de behandeling zelf kwam bij 12%

van de patiënten voor. 64% van de patiënten kreeg acute GVHD en 54% van

de patiënten kreeg chronische GVHD. Patiënten die snel na het ontstaan van

het multipel myeloom behandeld werden met een allo-SCT en die de allo-SCT

kregen op het moment dat er door de voorafgaande chemotherapie geen ziekte

activiteit meer was, reageerden het beste op de behandeling met allo-SCT. Deze

patiënten bleven de gehele follow-up duur ziekte vrij. De resultaten doen ons

concluderen, dat allo-SCT efficiënt is bij patiënten met een multipel myeloom.

De sterfte door de behandeling zelf in deze patiëntengroep was niet hoger dan

in andere groepen, maar blijft een probleem. Daarnaast waren er veel patiënten

met een acute en chronische GVHD. Deze complicaties beïnvloeden de kwaliteit

van leven na allo-SCT enorm.

(II) Het mechanisme van chronische GVHD

GVHD is de belangrijkste complicatie na allo-SCT. De acute vorm treedt in de

eerste maanden na allo-SCT op en kan een ernstig verloop hebben. We weten

dat bij acute GVHD specifieke afweercellen, zogeheten T cellen (gerijpt in de

Thymus), een belangrijke rol spelen in het ontstaan van de ziekte. Deze T cellen

zijn gericht tegen de normale cellen van de patiënt en maken deze kapot. Dit

resulteert in ernstige huiduitslag, met blaren en zweren, diarree en het falen

van de lever.

Chronische GVHD is de meest voorkomende complicatie, die op langere termijn

na allo-SCT kan ontstaan en kan tot jaren na de allo-SCT voor veel problemen

zorgen tot sterfte aan toe. Het is een ziekte, die de kwaliteit van leven vrese-

lijk in de weg staat. Chronische GVHD kan alle organen aantasten en kan veel

verschillende uitingsvormen hebben. Bij het grootste deel van de patiënten is

de huid aangedaan. Meestal zien we verlittekening van de huid. Deze verlitte-

kening tast niet alleen de huid aan, maar ook andere organen, zoals de longen,

de nieren, de (slok)darm en de lever. Door de verlittekening kunnen organen

157

minder goed functioneren. De huid kan zelfs dusdanig strak worden, dat het

als een harnas om het lijf zit en dat ademhalen bemoeilijkt wordt. Bij andere

patiënten zien we bij chronische GVHD juist meer zweren en wonden van de

huid.

Het verschil tussen acute en chronisch GVHD zit niet alleen in het tijdstip van

ontstaan, maar ook in de uitingsvorm. Met name de vorm van chronische

GVHD, waarbij veel verlittekening gezien wordt, lijkt niet op acute GVHD.

Dit verschil doet vermoeden dat ook de ontstaanswijze anders moet zijn. De

verlittekening bij chronische GVHD heeft veel weg van de verlittekening die

gezien wordt bij auto-immuunziekten, zoals systemische sclerose. Bij auto-

immuunziekten zijn niet alleen T cellen belangrijk, maar ook een ander soort

immuuncel, namelijk de B cel (Beenmerg gerijpt). De vergelijking met auto-

immuunziekten heeft ervoor gezorgd, dat er studies gedaan zijn, waarbij kleine

patiënten aantallen behandeld werden met rituximab, een geneesmiddel dat

specifiek B cellen verwijdert uit het lichaam (B-cel depletie). Rituximab wordt

veel gebruikt in de behandeling van lymfomen en chronische lymfatische leuke-

mie (CLL), aangezien bij deze ziektebeelden de B cellen ontaarden in kankercel-

len. B-cel depletie lijkt ook bij een aantal patiënten met chronische GVHD heel

goed te werken als behandeling. In deel II van dit proefschrift onderzochten we

als eerste de mogelijkheid om chronische GVHD te voorkomen.

In een grote groep patiënten, die een allo-SCT kregen voor uiteenlopende he-

matologische maligniteiten, hebben we onderscheid gemaakt tussen patiënten

die in hun behandeling voorafgaande aan de allo-SCT wel rituximab kregen

en patiënten die dit niet kregen. Bij patiënten, die in de 6 maanden voor hun

allo-SCT wel rituximab hadden gekregen en daarmee dus geen B-cellen meer

hadden, zagen we dat er minder en minder ernstige chronische GVHD ont-

stond. De B-cel depletie zorgde ervoor dat de B cellen van de patiënt verwij-

derd werden, maar had mogelijk ook effect op het uitgroeien van de B cellen

van de donor. Hier konden we geen onderscheid in maken in deze studie. Wel

kunnen we concluderen, dat rituximab een waardevolle aanvulling zou kunnen

zijn in de behandeling voorafgaande aan de allo-SCT om chronische GVHD te

voorkomen.

Daarna hebben we geprobeerd de ontstaanswijze van chronische GVHD beter

158


te begrijpen om zo beter gebruik te kunnen maken van de juiste behandeling

voor de juiste patiënten. Om dit te bereiken hebben we een studie uitgevoerd,

waarbij we patiënten met chronische GVHD, die niet goed reageerden op de

standaard therapie met prednison, behandelden met rituximab. Van deze pati-

ënten werden van te voren alle kenmerken en klachten in kaart gebracht en er

werd bloed afgenomen. Na de behandeling werden ze maandelijks teruggezien

en werd gekeken naar de response op de behandeling met rituximab. Wij za-

gen dat een groep patiënten wel en een groep patiënten nauwelijks reageerde

op de behandeling. We observeerden dat degenen die wel reageerden voorna-

melijk chronische GVHD hadden, die zich uitte in verlittekening van de huid

en leek op de auto-immuun ziekte systemische sclerose. Bij de patiënten, die

goed reageerden op de behandeling met rituximab, vonden we tevoren hogere

aantallen B cellen, die ook meer actief leken te zijn. Daarnaast hadden deze

B cellen andere kenmerken dan bij patiënten, die niet goed reageerden. De B

cellen in patiënten die niet goed reageerden, waren niet verhoogd aanwezig

en hadden dezelfde kenmerken als bij controle patiënten, die geen chronische

GVHD ontwikkelden. Na de behandeling met rituximab duurde het wel een

jaar voordat er weer B cellen gevonden werden in de patiënten. Bij degenen,

die goed reageerden op rituximab, hadden de B cellen, die na een jaar na be-

handeling teruggekomen waren, opvallend genoeg dezelfde kenmerken als bij

controle patiënten zonder GVHD. Het leek alsof de behandeling met rituxi-

mab de B cellen bij deze patiënten als het ware had gereset. Uit deze studie

konden we een aantal belangrijke dingen opmaken. Ten eerste is de populatie

patiënten met chronische GVHD geen eenheidsworst. Chronische GVHD heeft

veel verschillende verschijningsvormen en niet alle patiënten met de ziekte heb-

ben dezelfde kenmerken. De uiterlijke kenmerken van de ziekte verschillen,

maar blijkbaar verschillen ook de onderliggende mechanismen. Bij patiënten

die goed reageerden op de behandeling vonden we van te voren (dus tijdens de

chronische GVHD) hogere aantal B cellen met specifieke kenmerken, die na de

behandeling genormaliseerd waren. Dit geeft aan, dat deze geactiveerde B cel-

len heel goed de verlittekening kenmerkend voor de chronische GVHD kunnen

bewerkstelligen. Daarnaast zou je deze uiterlijke kenmerken van zowel patiën-

ten als van de B-cel populatie van de patiënten kunnen gebruiken om te bezien

bij welke patiënten het wel zin heeft om B-cel depletie te gebruiken als behan-

deling en bij welke patiënten dat geen zin heeft. Deze kennis brengt ons dichter

159

bij het precieze ontstaansmechanisme of liever gezegd mechanismen van chro-

nische GVHD en verklaart waarom bij sommige patiënten rituximab wel werkt

als behandeling en bij andere niet.

(III) Nieuwe strategieën om relapse en complicaties te

voorkomen

Niet alleen GVHD zorgt voor problemen na allo-SCT. Bij een groot aantal pa-

tiënten komt de onderliggende ziekte weer terug (relapse). Daarnaast zijn ook

infecties, zoals door herpesvirussen, een belangrijke oorzaak voor invaliditeit

en sterfte na de allo-SCT.

Normaalgesproken kan het immuunsysteem tumorcellen en cellen, die door een

virus geïnfecteerd zijn, herkennen en doden, zodat bijvoorbeeld leukemie en vi-

rusinfecties onder controle blijven. Belangrijke spelers in dit controlemechanis-

me zijn αβT cellen, die een specifiek eiwit aan de buitenkant dragen, namelijk

de αβT-cel receptor. Deze receptor bestaat uit twee delen, een α-keten en een β-

keten, en past precies op een molecuul aan de oppervlakte van normale cellen,

het Òhuman leukocyte antigenÓ (HLA) molecuul. Patiënt en donor worden ge-

matched op hun HLA type, zodat het nieuwe immuunsysteem kan functioneren

in het lichaam van de patiënt. Het HLA-molecuul bevat kleine eiwitdeeltjes uit

de cel en laat deze zien aan de αβT cellen. Wanneer deze eiwitdeeltjes afkomstig

zijn van een virus of wanneer ze afkomstig zijn uit een tumorcel, kan een αβT

cel, die deze eiwitdeeltjes specifiek herkent, de geïnfecteerde dan wel tumorcel-

len doden. Bij patiënten met leukemie heeft het controle mechanisme gefaald.

Een oplossing voor dit probleem zou kunnen zijn om grote aantallen αβT cel-

len, die specifiek de tumorcellen herkennen, als behandeling aan de patiënt te

geven. Dit kan door heel veel van deze cellen uit een donor te halen of door

de αβT-cel receptor, met de kwaliteit om specifiek tumorcellen te herkennen, op

andere, eerder niet specifieke, αβT cellen over te brengen. Er zijn een aantal na-

delen aan het gebruik van αβT cellen. Omdat ze precies passen op een bepaald

HLA-type, moeten er voor ieder HLA-type aparte αβT cellen gebruikt worden.

Daarnaast kan het overbrengen van een αβT-cel receptor op een andere αβT cel

zorgen voor zogeheten “on target” en “off target” auto-immuniteit. “On tar-

160


get” auto-immuniteit houdt in, dat de specifieke αβT-cel receptor stukjes eiwit

herkent, die met name gepresenteerd worden op tumorcellen, maar in mindere

mate ook op normale cellen voorkomen. Hierdoor zal deze specifieke αβT-cel

ook normale cellen doden. “Off target” auto-immuniteit houdt in, dat wanneer

de specifieke αβT-cel receptor overgebracht wordt op een andere αβT cel, de α-

en β-keten van de specifieke receptor combinaties maken met de α- en β-keten

van de αβT-cel receptor, die bij de niet specifieke αβT cel hoort. Er ontstaan dan

2 nieuwe soorten αβT-cel receptoren, waarvan we in het geheel niet weten wat

ze herkennen. Dit kan zorgen voor nieuwe receptoren, die ook weer normale

cellen herkennen en doden.

Een oplossing voor al deze nadelen is het gebruik van een ander soort T-cel

receptor, namelijk de γδT-cel receptor, die een γ- en een δ-keten heeft. T cellen

met deze receptor, γδT cellen, maken 5% uit van alle T cellen, die in het bloed

aanwezig zijn en bevinden zich in grotere hoeveelheden in de huid, long en

darm. Zij herkennen ook tumorcellen en geïnfecteerde cellen, maar doen dit

op een andere wijze dan αβT cellen. Zij passen niet op HLA-moleculen, maar

herkennen andere eiwitten, die alleen op tumorcellen en geïnfecteerde cellen

aanwezig zijn en zullen daarom normale cellen met rust laten, waardoor “on

target” auto-immuniteit voorkomen wordt. Daarnaast kan één γδT-cel receptor

voor patiënten gebruikt worden met diversen HLA-typen. Wanneer een γδT-cel

receptor overgebracht wordt op een αβT cel, zal deze niet combineren met de α-

en β-keten, die aanwezig is in de αβT cel en een αβT cel bevat geen γ- en geen

δ-keten. Hierdoor kunnen er geen T-cel receptoren met onbekende specificiteit

ontstaan en wordt “off target” auto-immuniteit voorkomen.

In deel III onderzochten we hoe en welke γδT-cel receptoren het beste gebruikt

kunnen worden om leukemie en virale infecties te controleren. We observeer-

den dat wanneer γδT cellen werden ingezet in de behandeling van leukemie

of andere soorten kanker, dit vaak teleurstellende resultaten had. Wij vonden

een mogelijke verklaring en een oplossing voor dit probleem. Een sterke reactie

tegen tumorcellen is niet alle γδT cellen gegeven. Sommige γδT-cel receptoren

zijn beter in het herkennen van tumorcellen dan andere. Wij hebben uit een

grote groep γδT cellen klonen gemaakt, met verschillende γδT-cel receptoren.

Hierdoor konden we de γδT-cel receptoren vergelijken, die wel goed waren en

161

niet goed waren in het herkennen van tumorcellen. Daarnaast konden we de

γδT-cel receptoren, die heel goed waren per aminozuur veranderen, waardoor

het duidelijk werd, welk deel van de γδT-cel receptor nu precies het verschil

bepaalde. We ontdekten dat de lengte van de δ-keten van belang was, maar ook

de combinatie van bepaalde γ- en δ-ketens. Veranderingen in zowel de γ- als

de δ-keten kon het reactievermogen van de γδT cellen op tumorcellen verande-

ren. Als je de juiste γ- en de juiste δ-keten gebruikt, en deze overbrengt op een

αβT cel, onafhankelijk van het type HLA, kun je een grote hoeveelheid T cel-

len genereren, die tumoren goed herkennen en doden en geen auto-immuniteit

geven. Hierdoor zouden leukemie en andere tumoren beter gecontroleerd kun-

nen worden. Dit principe hebben we “combinatorial γδT-cel receptor exchange”

(CTE) gedoopt.

Een andere observatie was dat tijdens CMV infecties na allo-SCT vaak in eerste

instantie γδT cellen zich vermenigvuldigden. Dit werd eerder ook al gezien bij

gezonde vrijwilligers met CMV en bij patiënten die een niertransplantatie had-

den ondergaan. Dit zou betekenen dat γδT cellen belangrijk zijn in de controle

van CMV infecties. Er bestaat een correlatie tussen CMV infectie na allo-SCT

en een langere leukemie vrije overleving. Dit deed ons vermoeden dat de γδT

cellen, die zich vermenigvuldigen tijdens CMV infectie, ook leukemie cellen

kunnen herkennen en doden en daarmee leukemie na allo-SCT kunnen contro-

leren. We vonden dat γδT cellen, die CMV herkenden, ook leukemie en ander

tumorcellen herkenden en doodden. Echter wanneer we de γδT-cel receptor

van deze cellen overbrachten op een αβT cel, herkende deze alleen tumorcel-

len en was de herkenning van CMV geïnfecteerde cellen verloren gegaan. Dit

betekent ten eerste dat we γδT cellen kunnen inzetten tegen leukemie en CMV

infecties na allo-SCT. Ten tweede hebben we opnieuw laten zien dat de γδT-cel

receptor zeer goed te gebruiken is om grote groepen T cellen te maken, die

leukemie cellen kunnen herkennen en doden. Maar dit betekent ook dat voor

de herkenning van CMV geïnfecteerde cellen moleculen nodig zijn, die op γδT

cellen aanwezig zijn en niet op αβT cellen en dat het overbrengen van alleen

de γδT-cel receptor wel genoeg is voor de herkenning van tumoren, maar niet

genoeg is voor de herkenning van CMV. Welke moleculen er precies nodig zijn

om ook CMV te herkennen, hebben we nog niet kunnen achterhalen.

162


Conclusie

In dit proefschrift worden verschillende manieren beschreven om het nieuwe

immuunsysteem dat gegeven wordt aan patiënten met leukemie, vorm te ge-

ven. Hierdoor kan allo-SCT een verfijndere behandeling worden. Door B cellen

te depleteren kan GVHD voorkomen en behandeld worden. Door gebruik te

maken van de unieke kenmerken van γδT cellen en de γδT-cel receptor kunnen

relapse van de ziekte en complicaties zoals GVHD en infecties vermeden wor-

den. Deze resultaten bieden mogelijkheden om in de toekomst de conventionele

manier van transplanteren te vervangen door een gerichte en geïndividualiseer-

de behandeling van leukemie.

163

164

Curriculum Vitæ

Suzanne van Dorp werd geboren op 8 juli 1982 te Haarlem. In 2000 behaalde zij

het gymnasiumdiploma aan het Stedelijk Gymnasium te Haarlem. Tijdens haar

studie Geneeskunde aan de Universiteit Utrecht deed zij onder andere coschap-

pen in Nicaragua en Malawi. Tijdens het co-schap interne geneeskunde in het

St. Antoniusziekenhuis ontstond haar interesse voor de hematologie. Haar

wetenschappelijke stage in het UMC Utrecht mondde uit in een promotieon-

derzoek onder leiding van Dr. J.H.E. Kuball en Dr. E. Meijer bij de afdelingen

Hematologie en Immunologie. De resultaten van dit onderzoek zijn beschreven

in dit proefschrift. Suzanne is in opleiding tot internist in het St. Antoniusziek-

enhuis te Nieuwegein.

165

166

List of publications

2 van Dorp S, Meijer E, van de Donk NW, Dekker AW, Nieuwenhuis K, Min-

nema MC, Petersen E, Schutgens R, Verdonck LF, Lokhorst HM. Single-centre

experience with nonmyeloablative allogeneic stem cell transplantation in pa-

tients with multiple myeloma: prolonged remissions induced. Neth J Med.

2007 May;65(5):178-84.

2 Kersting S, van Dorp S, Theobald M, Verdonck LF. Acute renal failure after

nonmyeloablative stem cell transplantation in adults. Biol Blood Marrow Trans-

plant. 2008 Jan;14(1):125-31.

2 Marcu-Malina V, van Dorp S, Kuball J. Re-targeting T-cells against cancer

by gene-transfer of tumor-reactive receptors. Expert Opin Biol Ther. 2009

May;9(5):579-91.

2 van Dorp S, Pietersma F, Wölfl M, Verdonck LF, Petersen EJ, Lokhorst HM,

Martens E, Theobald M, van Baarle D, Meijer E, Kuball J. Rituximab treat-

ment before reduced-intensity conditioning transplantation associates with a

decreased incidence of extensive chronic GVHD. Biol Blood Marrow Transplant.

2009 Jun;15(6):671-8.

2 Pietersma FL, van Dorp S, Jacobi R, Ran L, Nanlohy NM, Schuurman R,

Minnema MC, Meijer E, van Baarle D. High level of perforin expression in T

cells: An early prognostic marker of the severity of herpesvirus reactivation

after allogeneic stem cell transplantation in adults. Clin Infect Dis. 2010 Mar

1;50(5):717-25.

2 Westeneng AC, Hettinga Y, Lokhorst H, Verdonck L, van Dorp S, Rothova

A. Ocular graft-versus-host disease after allogeneic stem cell transplantation.

Cornea. 2010 Jul;29(7):758-63.

2 Minnema MC, van Dorp S, van de Donk N, Lokhorst H. Prognostic factors

and outcome in relapsed multiple myeloma after non-myeloablative allogen-

167

eic stem cell transplantation; a single center experience. Bone Marrow Trans-

plant. 2011 Feb;46(2):244-9.

2 Kuball J, de Boer K, Wagner E, Wattad M, Antunes E, Weeratna RD, Vicari

AP, Lotz C, van Dorp S, Hol S, Greenberg PD, Heit W, Davis HL, Theobald

M. Pitfalls of vaccinations with WT1-, Proteinase3- and MUC1-derived pep-

tides in combination with MontanideISA51 and CpG7909. Cancer Immunol

Immunother. 2011 Feb;60(2):161-71.

2 Pietersma FL, van Dorp S, Minnema MC, Kuball J, Meijer E, Schuurman R,

van Baarle D. Influence of donor cytomegalovirus (CMV) status on sever-

ity of viral reactivation after allogeneic stem cell transplantation in CMV-

seropositive recipients. Clin Infect Dis. 2011 Apr 1;52(7):e144-8.

2 van Dorp S∗, Resemann H∗, te Boome L, Pietersma F, van Baarle D, Gmelig-

Meyling F, de Weger R, Petersen E, Minnema M, Lokhorst H, Ebeling S,

Beijn SJ, Knol EF, van Dijk M, Meijer E, Kuball J. (∗Both authors contributed

equally) The immunological phenotype of rituximab-sensitive chronic graft-

versus-host disease: a phase II study. Haematologica. 2011 Sep;96(9):1380-4.

2 Hogenes MC, van Dorp S, van Kuik J, Monteiro FR, ter Hoeve N, van Dijk

MR, Martens AC, de Weger RA. Histological assessment of the sclerotic graft-

versus-host response in the humanized RAG2-/-γc-/- mouse model. Biol

Blood Marrow Transplant. 2012 Jul;18(7):1023-35

2 Gründer C, van Dorp S, Hol S, Drent E, Straetemans T, Heijhuurs S, Scholten

K, Scheper W, Sebestyen Z, Martens A, Strong R, Kuball J. γ9 and δ2CDR3

domains regulate functional avidity of T-cells harboring γ9δ2T-cell receptors.

Blood. 2012 Dec 20;120(26):5153-62.

2 Scheper W∗, van Dorp S∗, Kersting S, Pietersma F, Hol S, Heijhuurs S,

Lindemans C, Sebestyen Z, Gründer C, Marcu-Malina V, Marchant A, Don-

ner C, Becke S, Plachter B, Vermijlen D, van Baarle D, Kuball J. (∗Both au-

thors contributed equally) γδT-cells elicited by CMV-reactivation after allo-

SCT cross recognize CMV and leukemia. Leukemia. 2013 Jan 1. doi:

10.1038/leu.2012.374. [Epub ahead of print]

168

Dankwoord

Lieve pap, dit boek is opgedragen aan jou. Het was mooier geweest, als ik het je

had kunnen geven. Je hebt een belangrijke rol gespeeld in de keuze om te gaan

promoveren. Vroeger zei je vaak, dat je eigenlijk wel gewoon kon promoveren

op je afstudeerscriptie over Fockenbroch. Ik twijfelde daar toen al wat aan, en

na 6 jaar promoveren denk ik dat die twijfel wel terecht was. Maar promoveren

leek je wel wat, en je had gelijk. Het was zeker wat. Soms frustrerend, soms

stressvol, maar meestal inspirerend en een voorrecht om te kunnen doen.

Veel mensen hebben de afgelopen jaren voor die inspiratie gezorgd.

Als eerste wil graag ik mijn co-promotor Dr. J.H.E. Kuball bedanken. Danke

schön, Herr Doktor Kuball! Lieber Jürgen, het is ongelooflijk hoe jij in vijf

jaar van kortdurend mijn kamergenoot op AIO kamer 3 opgeklommen bent tot

waar je nu staat. Maar het is ook weer niet geheel verbazingwekkend. Jouw

intelligentie, je vindingrijkheid, je werklust en je absolute optimisme en door-

zettingsvermogen zijn een inspiratie en de grootste kracht achter deze promotie.

Ik ben heel erg blij, dat je mij geadopteerd hebt als AIO en het is een eer om

jouw eerste promovendus te zijn.

Daarnaast wil ik graag mijn co-promotor Dr. E. Meijer bedanken. Beste Ellen,

na mijn wetenschappelijke stage onder jouw hoede ben ik begonnen aan dit

promotie-traject. Helaas verliet je me na een jaar om je geluk elders te zoeken.

Ik heb een van de belangrijkste schatten in het onderzoek van jou geleerd, het

maken van een databank. Daarnaast heb ik tijdens onze hereniging in het St.

Antoniusziekenhuis, gezien hoe je een goede hematoloog moet zijn!

Dr. S.B. Ebeling, beste Saskia, dank je wel voor het begeleiden van mijn eerste

ervaringen met lab-onderzoek, pipeteren, muisexperimenten, wetenschappelij-

ke artikelen en überhaupt een volwassen baan hebben!

Ik wil graag mijn promotor Prof. dr. H. M. Lokhorst bedanken. Beste Henk. Jij

169

bracht mij in contact met Ellen en jij bracht mij mijn eerste paper! Je staat daar-

mee aan de basis van mijn wetenschappelijke carrière. Dank je wel daarvoor. Jij

bent voor mij hét voorbeeld voor hoe je én een geweldige klinische dokter kan

zijn én tegelijkertijd heel erg belangrijk in de onderzoekswereld.

De leden van de leescommissie: Prof. dr. E. Wiertz, Prof. dr. C.E. Hack,

Prof. dr. A. van de Loosdrecht, Prof. dr. J. Cornelissen, veel dank voor het

beoordelen van mijn manuscript. Prof. dr. P. Fisch, herzlichen Dank für die

Beurteilung meiner Dissertation.

Dr. H. Dolstra en Dr. T. Mutis, dank voor het opponeren. Tuna, ook erg bedankt

voor de spraakmakende discussies, de gezellige borrels en het salsa’en!

Dr. M. Minnema, Dr. E. Petersen, Dr. R. Schutgens, Dr. L. te Boome,

Dr. N. van de Donk, Dr. E. van der Spek , dank jullie wel voor de pretti-

ge samenwerking op de afdeling Hematologie. Monique en Eefke, speciale

dank voor het includeren van patiënten. Liane, dank je wel voor het oppakken

en voortzetten van de chronische GVHD studie en voor alle samenwerking in

group Kuball!

Prof Dr. M. van Dijk. Beste Marijke, dank je wel voor je hulp bij het analyseren

van alle coupes.

Monique Knies. Dank je wel voor je altijd supersnelle reacties, voor de fijne

interactie en voor het rennen naar de postkamer!

Group Kuball: Vica, Zsolt, Cordula, Wouter, Sam, Sabine, Sabina, Liane,

Caroline, Linda, Hakim, Julia, Carina, Edite: Thank you for all the “labbe-

sprechungen”, the discussions, the experiments together, the nice dinners and

celebrations. It was a pleasure and remember: “Gamma delta T cells rock!”

Lieve Wouter, mijn Muze! Je maakte het laatste jaar op het lab tot een waar

feest! Dank je wel voor al je hulp, je gezelligheid, je rust, je kunde en je heerlijke

humor. Mijn ideaalbeeld voor de toekomst zou zijn om met jou een onder-

zoeksgroep te hebben. Ik weet niet of het uit gaat komen, want je hebt wellicht

heel andere plannen, maar ik heb heel veel respect voor jou als onderzoeker en

collega en niet te vergeten als vriend! Zullen we in ieder geval nog vaak samen

“bangen” en mag ik dan alsjeblieft meezingen met lekkere hitjes?

170

DANKWOORD

Dear Vica, you have taught me so much! I admire how you take on projects

from scratch and make them into successes. I wish the Dutch could have won

you over, but I think that there is no place where you will be as happy as in

Israel with your family!

Dear Cordula, thank you for sharing your pure look on life and science. Your

perfectionism led to a Blood paper! Do not forget to put stuff in perspective

sometimes, since that can make things in life ánd science al lot easier! I’m sure

that you will find the way to do so and still be content with everything you do.

Thank you for organising all those nice group events!

Lieve Zsolt, toen jij kwam solliciteren, viel ik zowat van mijn stoel zo mooi

vond ik je. Maar niet alleen daarom bleek het heel gelukkig dat je group Kuball

kwam versterken. Je bracht een zeer hoog onderzoeksniveau, stijl, gezelligheid

en humor met je mee en een heel fijn Amsterdams juppengehalte samen met

Jan!

Lieve Sabina, Dr. Kersting. Wat was het fijn om met jou onderzoek te doen en

met Floor tussen ons in op de kamer te zitten! Je hebt je een ongeluk geFACSt

en ik ben heel blij dat daar uiteindelijk zo’n mooi paper uitgekomen is! We

hebben een onvergetelijk leuk congres in New Orleans gehad samen en ik hoop

dat daar nog een aantal van volgen! Ik bewonder hoe je je gezinsleven met zo’n

uitdagende baan combineert!

Lieve Sabine, Sabje! Dank je wel voor alle experimenten en al het isoleren en

kweken dat we samen hebben gedaan! Het was een waar genoegen om je te

zien groeien in de groep tot de onmisbare kracht die je bent!

Uiteraard wil ik graag de studenten bedanken, waarmee ik heb samengewerkt.

Henrike: Thank you for your persistence, your efficiency and your patience. It

must have been hard to work with a chaotic person like me, but the result is a

very nice paper in Hematologica! Nicole, ik vond het ontzettend leuk om een

deel van je begeleiding te kunnen doen! Minke, Agent Rab! Ja, dat was wel

fantastisch. MSC’s kweken, de FACS, AMC, en heel veel lachen: een perfecte

combinatie! Gelukkig werken we nu iedere dag samen in het Antonius!

Op de afdeling Immunologie in het WKZ was het goed toeven! Dat kwam

171

door de gedreven, maar ontspannen sfeer en de vanzelfsprekendheid en het

plezier waarmee er samengewerkt werd. Ik denk dat dat erg bijzonder is en wil

iedereen daar erg graag voor bedanken!

De staf van de afdeling Immunologie, zoals ik hem heb gekend: Dr. K. Tesse-

laar, Prof. dr. L. Meijaard, Dr. J. Leusen, Dr. D. van Baarle, Dr. J. Borghans,

Dr. A. Martens, Dr. K. Denzer, Dr. L. Bont, Prof. dr. P. Coffer, Prof. dr. F. Mie-

dema. Zeer veel dank voor de waardevolle bijdragen tijdens discussies en pre-

sentaties. Frank, ik wil jou speciaal nog bedanken voor je humorvolle en tege-

lijkertijd uiterst scherpe benadering van gepresenteerde data. Het was genieten

met jou bij een bespreking!

Petra, dank je wel voor het dynamic duo dat we waren in mijn eerste jaar. Ik

vind je heel bijzonder en kundig.

Yvonne en Saskia, zeer veel dank voor jullie oneindige flexibiliteit en hulpvaar-

digheid!

AIO kamer 3: Rogier, Lydia, Ellen, Floor, Kees, Kristof (Zotteke), Peter (de

Hongaarse muur), Jantine, Paul, Wouter, Cordula, Annelieke, Bart, Thijs, Hil-

de, Vica, Sabina, Charlotte, Kirsten, Hakim. Het waren mooie tijden! Het gek-

kenuurtje om half 5, de onuitputtelijke bron grappen en grollen, het doorbreken

van de bubbel, de dikke negermuziek, het tafeltennis, de vissen EBV en GVHD,

het klaterende gelach tussen de serieuze inspanningen door, maakten het de

leukste AIO kamer van alle AIO kamers ter wereld! Dank daarvoor.

Verder veel dank aan Jeffrey, Dan (de Man), Jorg, Veerle, Marieke, Henk Jan,

Richard, Henk, Vera, Liset, Jolanda, Ingrid, Ana, Justin, Guru, Marloes, Tessa,

Robbert, Robert, Stefin, Florijn, Frans, Joke, Walter, Nening, Ronald, Sanne,

Gerrit, Koos, Sigrid, Maaike, Marco, Marc en Wilco voor alle gezelligheid, de

fijne samenwerking en de geweldige borrels!!

Mijn paranimfen: Samantha, Sammie!! Jouw ongelooflijke efficiëntie en snel-

heid in het proeven doen, maakten dat er veel tijd was om koffie te drinken en

eindeloos te kletsen. Ik heb het zo leuk met je gehad en het was zo’n luxe en

voorrecht om met jou te mogen werken. Dank je wel. Heel fijn dat je aan mijn

zijde wil staan!

172

DANKWOORD

Floor, lieve Flora! Vanaf de eerste dag dat ik kwam werken in het WKZ was

je daar. Liefde op het eerste gezicht! Een onafscheidelijk duo was het gevolg

en daarmee heel veel hilariteit. Dansen op het lab, samen FACSen, SPSSen,

“datamassage”, samen uit eten met sammelebuca en birella’s en ook serieuzere

gesprekken, die geleid hebben tot een heel aantal publicaties (Pietersma, van

Dorp en van Dorp, Pietersma et al.), waar alleen Blood nog niet zo van ge-

charmeerd was. Dank je wel dat je het werken op een lab zo sprankelend hebt

gemaakt!

Gelukkig hebben ook buiten het lab mensen direct en indirect bijgedragen aan

dit boekje of aan mijn levensgeluk, zodat dit boekje makkelijker kon ontstaan.

Lieve Kees, jij hebt wel een heel uitzonderlijke positie ingenomen in het ont-

staan van dit proefschrift. Je maakte het al mooie leven in en buiten het lab nog

mooier! Dank je wel voor alles wat je me gegeven hebt en hebt laten zien. Ook

Cees (II), Nathalie, Marieke en Maarten heel erg bedankt voor de heerlijke

etentjes, de discussies, het altijd warme welkom en het varen!

Lieve Amsterdammers, liefste Wel, An, Ras, Soof, Maart, Daan, Shaun en Wes!

Jullie zijn bevrijdend! Heel erg bedankt voor jullie. Wellie, jij begrijpt mij als

geen ander, ik wil je speciaal bedanken, omdat je er onvoorwaardelijk voor me

bent.

Maart! Ik wil jou bedanken voor het feit, dat je altijd achter me gestaan hebt.

Je kent me door en door en dat maakt dat je heel belangrijk voor me bent.

Penny, dank je wel dat je me altijd een warme plek hebt gegeven, ook al heb

ik er de laatste jaren wat weinig gebruik van gemaakt. Daan, door jou wilde ik

professor worden (Grieks weliswaar). Dit is het begin!

Lieve Niels en Jeroen! Dank voor jullie geweldigheid! Ik ben blij dat heel

Nederland dat tegenwoordig inziet.

Tillemakinderen en juf! Heel erg bedankt voor de vertrouwde etentjes.

Antonianen, bedankt voor de bijzonder goede sfeer waarin we samenwerken!

Ik besef me heel goed, dat zoveel leuke collega’s eigenlijk niet kan! Dr. Geers,

dank voor de gelegenheid om dingen af te maken en de fijne, goed passende

supervisie! Dr. de Weerdt, lieve Okke! Door jou wil ik hematoloog worden.

173

Mark, dank je wel voor je absolute optimisme, je barricades voor bijvoorbeeld

de tonijn, je danskwaliteiten en Ticket to Ride! Azzie, lieverd! Ment to be vind

ik ons. Ik ben echt heel erg blij dat jij me op stond te wachten op de eerste dag!

Joost , mijn grote vriend! Ik kan werkelijk alles met je bespreken. Ik ben blij

met je keuze voor het mooiste aandachtsgebied dat er is!

Lieve Girlsch, Suzy Q, Hes en Lies. Ik ben heel blij met jullie en dat we samen

groot geworden zijn! Dank jullie wel voor altijd!! Daan en Vic, jullie zijn precies

goed voor de meisjes, niet veranderen!

Lieve Miekie! Dank je wel voor je trouwe vriendschap, je attentheid, je orga-

nisatorisch talent en alle leuke dingen die we samen gedaan hebben (niet te

vergeten jouw promotie!!). Erwin, jij ook dank je wel voor al die leuke avonden,

etentjes, feestjes en voor je interesse, fijnheid en Mario Cart!

Lieve Daph! Samen studeren, samen wonen, samen hardlopen, samen een Duit-

se baas hebben. Dank je wel dat we dat allemaal konden delen!

Lieve Willem. Dank je wel voor je beste vriendschap! Ik weet zeker dat mijn

kinderen in de zandbak toch van jouw duifjes van kinderen gaan winnen! En

ook weet ik zeker dat we er samen naar zullen gaan kijken!

Lieve Debster, Inator, mijn liefste huisgenoot! Ik ben zelden iemand tegengeko-

men bij wie het zo gemakkelijk was allemaal en bij wie ik zoveel overeenkom-

sten vond. Duizend maal dank!! Ik heb Paolo en John nogmaals benaderd voor

de woongroep. Ook dank voor het “delen” van Dit en Draak op de Koningslaan.

Dit, ik ben heel blij dat je er bent!!

Lieve Marie Louise, Hans, Kasper, Marlou en Jibbe. Jullie zijn geweldig! Dank

voor alle vrolijkheid, openheid en liefdevolheid, waarmee jullie me meteen om-

ringd hebben.

Lieve mamma en Harry, dank jullie wel voor alle liefde, gezelligheid, heerlijke

etentjes, adviezen, ondersteuning en het geweldige voorbeeld dat jullie zijn. Ik

hoop ook zo van het leven te gaan genieten als jullie dat doen en ik ben heel

blij dat jullie dat samen doen. Lieve Omi! Jij ook bedankt voor de weekenden

ontspannen en onze telefoongesprekken, waar van alles de revue passeert. Een

hele dikke iiiiiii!

174

DANKWOORD

Chris, mijn lieve broertje. Ik ben van alle mensen op de wereld het meest trots

op jou. Je hebt een heleboel voor je kiezen gehad en je bent er zo ontzettend

leuk en knap uitgekomen. Speciale dank voor het verzorgen van de lay-out en

de kaft van dit boekje! Het is echt prachtig geworden!

Arjan, lieve AJ! Ik heb héél erg veel zin in héél erg veel meer van jou.

Zoveel inspiratie! Dat moet een mooi feestje worden!

175

Documents

Shaping Graft Immunity