Epstein–Barr virus (EBV) establishes latency in B cells following primary infection forming an episome or circular DNA rendering it resistant to most conventional antiviral agents which target linear viral DNA.1 Following allogeneic hematopoietic cell transplantation (HCT), donor-derived B cells that harbor EBV in a latent state are liable to transformation of the virus to a lytic state and its rapid proliferation in the face of posttransplant T-cell immunodeficiency.2 EBV reactivation is thus largely donor-derived following HCT, unlike solid-organ transplantation, except for nonablative conditioning, where residual host B cells might be the source of EBV proliferation.3,4 The major concern with uncontrolled or sustained EBV proliferation is the development of clonal B-cell lymphoproliferative disorders (LPDs), which are often fatal. Both T-cell depletion (TCD) and haploidentical HCT have been identified as major risk factors for both EBV reactivation and EBV-LPD.5 EBV-specific cytotoxic T cells have been shown to protect against EBV-LPD in the posttransplantation period.6 However, similar data on natural killer (NK) cell subsets are lacking, although certain NK subsets have been shown to be protective after primary infection in normal hosts.7-10 In addition, donor EBV seropositivity has been shown to increase the risk of chronic graft-versus-host disease (GVHD), but the pathophysiology is unclear.11
Our group had developed a novel approach to haploidentical HCT, combining posttransplant cyclophosphamide (PTCy) and T-cell costimulation blockade with cytotoxic T-lymphocyte-associated protein 4 Immunoglobulin (CTLA4Ig).12,13 In patients with severe aplastic anemia, introduction of CTLA4Ig, both pre and posttransplantation, markedly reduced the incidence of alloreactivity.12 The same phenomenon was observed in younger patients with malignant diseases,14 who otherwise experienced a very high incidence of both acute GVHD and posttransplantation hemophagocytic syndrome when transplanted on conventional PTCy protocol.15,16 Although naive T cells are susceptible to CTLA4Ig-mediated blockade of activation, NK cells are resistant to CTLA4Ig.17 In addition, CTLA4Ig was shown to promote anti-tumor effect of NK cells in preclinical models.18 Given the efficacy of CTLA4Ig in abrogating T-cell–mediated alloreactivity, we had explored the possibility of using CTLA4Ig-primed donor lymphocyte infusions (DLIs) in exploiting NK cell–mediated antileukemia effect without incurring T-cell–mediated alloreactivity. This was successfully demonstrated in 30 patients with advanced acute leukemia with a progression-free survival of 75%, the incidences of acute GVHD and nonrelapse deaths being <10%.13 In all these cohorts, despite a very low incidence of acute GVHD, about one-fifth of the patients developed chronic GVHD. We prospectively studied the impact of the protocols employing CTLA4Ig with PTCy on EBV reactivation and EBV-LPD, particularly in relation to various subsets of NK cells and its impact on GVHD.
PATIENTS AND METHODS
Patients between 2 and 65 years of age without a suitable matched family donor were enrolled if they possessed a haploidentical family donor. Approval was obtained from institutional review board of the institution (ECR 2372; protocol number: H0401G), and informed consents were obtained from all patients and donors in accordance with Declaration of Helsinki.
Conditioning regimens for malignant and nonmalignant diseases have been described in detail in our previous publications and illustrated in Figure 1. Graft source was mobilized peripheral blood stem cells. All patients received PTCy at 50 mg/kg on days +3 and +4, 24 hours apart.19 In the protocol MCFH0401G, patients received CTLA4Ig (abatacept) at 10 mg/kg on days −1, + 6, +20, and +35.12 In those with nonmalignant disorders, abatacept was administered every 4 weeks after day +35 until day +180 and sirolimus was given from day −8 as previously described to maintain trough level of 4–12 ng/mL.12 Those with malignant diseases received cyclosporine from D+5 at 1.5 mg/kg, 12 hours apart to maintain a trough level of 75–150 ng/mL along with. DLI at 1 × 106/kg was administered 12 hours following CTLA4Ig on day +7 and at 5 × 106/kg on days +21 and +35 in the absence of GVHD.13 The treatment schema has been illustrated in Figure 1.
Donor Selection and Mobilization Protocol
The methods followed for HLA typing, killer immunoglobulin-like receptor (KIR) genotyping, and defining NK cell alloreactivity and the mobilization protocol for hematopoietic stem cells for both malignant and nonmalignant diseases have been described in our earlier studies.12,14,19,20
All patients were treated in protective isolation rooms provided with high-efficiency particle air filters. Antimicrobial prophylaxis was instituted as per the departmental guidelines. Cytomegalovirus (CMV) prophylaxis was guided by pre-emptive monitoring of viral CMV load by quantitative polymerase chain reaction twice a week until day 100.
Acute GVHD was graded according to modified Glucksberg criteria, and chronic GVHD was scored based on National Institutes of Health global severity criteria.21,22 Posttransplantation hemophagocytic syndrome was defined as previously published.16
Detection and Quantification of EBV DNA
Whole blood was collected from the patients 1 week before transplantation and every week thereafter for a minimum of 100 days for detection of EBV DNA. DNA extraction was carried out from whole blood using QAMamp mini kit (QIAGEN GmbH, Hilden, Germany). EBV detection and quantitation were carried out using EBV R-gene kit (bioMerieux SA, Marcy I' Etoile, France) by real-time polymerase chain reaction on Step One PLUS (Applied Biosystems).23
Definitions and Treatment of EBV Reactivation
EBV reactivation was defined as detection of >0.5 × 103 virus copies/mL on 2 separate samples. Presence of EBV reactivation with biopsy-proven lymphoproliferative disease was defined as EBV-LPD. EBV reactivation >1 × 103 copies/mL on consecutive samples was considered for pre-emptive treatment with rituximab. Rituximab was administered at 375 mg/m2 weekly until EBV was <0.5 × 103 copies/mL.
Flow Cytometric Assessment of T- and NK Cell Subtypes
This was carried out on the peripheral blood samples of patients at days +30, +60, +90, +150, and every 3 months thereafter for 1 year following HCT.14,20 The NK cell and T-cell immunophenotypes were carried out by 8 color-flow Cytometry in Navios (Beckman Coulter Inc) using the following mouse anti-human mAbs from Beckman Coulter, Immunotech (Marseille, France): CD45 (J33), CD3 (UCHT1), CD4 (13B8.2), CD8 (B9.11), CD56 (N901), CD16 (3G8), CD80 (MAB104), CD86 (HA5.2B7), CD158a (EB6B), CD158b (GL183), CD158e (Z27.3.7), CD159a (Z199) and Miltenyi Biotec GmbH (Bergisch Gladbach, Germany): CD159c (REA2015). T regulatory cells were analyzed using mouse anti-human mAbs from BD Biosciences (San Jose, CA); CD4 (SK3), CD25 (2A3), CD127 (HIL-7R-M21).
The transplant outcomes studied were EBV reactivation, grade 2–4 acute GVHD, chronic GVHD, nonrelapse mortality (NRM), EBV-related mortality, and overall survival. The various immune parameters were compared between subgroups at multiple time-points. Binary variables were compared between the groups using chi-square test. The continuous variables were analyzed using nonparametric tests (Mann–Whitney U test) for medians across the groups and t test with Levene test for equality of variances for mean values between groups. Probabilities of survival were estimated using the Kaplan–Meier product-limit method. An outcome was determined to be significantly different if the observed P was <0.05. All analyses were performed using statistical software IBM SPSS Statistics Version 21 (Armonk, NY).
Incidence of EBV Reactivation
All patients and donors were seropositive for EBV. Eight out of 71 patients developed EBV reactivation with a cumulative incidence of 13.8% (95% confidence interval [CI], 9.1-18.5). The median time to onset of EBV reactivation was 96 days (range, 65–307). The median age of patients with EBV reactivation was 28 years (range, 2–53). All were transplanted for malignant diseases (acute myeloid leukemia-5, acute lymphoblastic leukemia-3). The characteristics of patients with and without reactivation have been detailed in Table 1.
Kinetics of EBV DNA Load in Response to Treatment and Outcome
The initial viral load varied between 0.05 and 1.45 × 104 copies/mL (median, 1.1 × 104). The maximum viral load was 0.7–3.9 × 104/mL. The viral load peaked at a median of 2 weeks. Seven patients with EBV reactivation had CMV reactivation preceding the onset, and all had undetectable CMV viral load at the time of EBV reactivation. None were on ganciclovir or foscarnet at the onset of EBV reactivation. The median duration of reactivation was 2 weeks. Only 3 patients received pre-emptive rituximab for 2 doses. All had EBV DNA load below 0.05 × 104 copies/mL by 3 weeks, except 1 patient who had low level of reactivation for 6 weeks (<1.0 × 104 copies/mL). No recurrence of EBV reactivation was witnessed apart from 1 patient who developed the first reactivation at day +90 and had a recurrence at day +307, while on systemic corticosteroids for chronic GVHD. This was treated with 2 doses of weekly rituximab without further recurrence. None developed EBV-LPD at a median follow-up of 2 years (range, 1–4 y).
Risk Factors for EBV Reactivation
EBV reactivation tended to be more in those with malignant diseases (18.3% [95% CI, 11.6-25] versus 0% in nonmalignant diseases; P = 0.09). There was no impact of patient or donor age, donor NK-KIR haplotype, NK ligand mismatch, and graft composition on EBV reactivation. The time to engraftment was similar between the 2 groups.
Impact of EBV Reactivation on GVHD
The overall incidence of grade 2–4 acute GVHD was 10.2% (95% CI, 6.6-13.8). None with EBV reactivation developed acute GVHD, compared with 14.4% in those without (P = 0.5). The incidence of chronic GVHD was 19% (8/69; 95% CI, 12.9-25.1). Five out of 8 patients with EBV reactivation went on to develop chronic GVHD at a median of 145 days (range, 124–187) with an incidence of 62.5% (events, 5/8; 95% CI, 45.4-79.6), compared with only 8% (events, 3/61; 95% CI, 3.5-12.5; onset 100–132 d) in those without EBV reactivation (P = 0.001; Figure 2). There was no relationship between prior rituximab therapy and chronic GVHD. However, in patients who experienced chronic GVHD following EBV reactivation, the involvement was predominantly cutaneous with lichenoid histology without any major organ involvement and was protracted. Those who developed chronic GVHD without prior EBV reactivation had oral or ocular GVHD which was mild and self-limiting in nature and did not require any systemic therapy. No other risk factor was identified for chronic GVHD.
NRM, Progression-free, and Overall Survival
The overall incidence of NRM was 4.2% (95% CI, 1.8-6.6). Progression-free and overall survival were 75.6% (95% CI, 70.2-81.0) and 82.3% (95% CI, 77.6-87.0), respectively, at a median follow-up of 24 months (range, 14–60). There was no impact of EBV reactivation on any of these outcomes.
EBV Reactivation and NKG2Apos CD56dim Subset of NK Cells
There was no difference in NK or T-cell subsets between those with and without EBV reactivation. This is detailed in Table 2. There was no difference in the overall CD56bright or CD56dim NK cell populations either. However, the expression of NKG2A in CD56dim 16pos NK cells before, during, and after EBV reactivation showed significant alterations. The CD56dim NKG2Apos NK cells before EBV reactivation at day +60 was 21.9% ± 6.4%, which increased to 44.7% ± 15.2% at day +90 and 54.4% ± 17.3% at day +150 (P < 0.01). This was associated with a reciprocal decrease in NKG2Cpos CD56dim population. All 5 patients who developed chronic GVHD did so within 8 weeks.
NKG2Apos CD56dim NK Cells and Chronic GVHD
The kinetics of NKG2Apos CD56dim NK cells were compared with 10 patients, who were transplanted for acute leukemia but did not develop EBV reactivation, disease progression, and acute or chronic GVHD (control cohort). In contrast to those with EBV reactivation and chronic GVHD (EBV-cGVHD cohort, n = 5), the control cohort showed a trend to decrease in NKG2Apos CD56dim NK cells at mean values of 37.9% ± 18.7%, 26.2% ± 11.9%, and 23.6% ± 13.5% at days +60, +90, and +150, respectively (P = 0.04; Figure 3A). In the EBV-cGVHD cohort, the NKG2Apos CD56dim NK cells were significantly higher at day +90 (52.5% ± 13.04%) and day +150 (59.08% ± 16.3%) (Figure 4A), correlating with the onset of chronic GVHD, compared with the control cohort (P = 0.004).
Reduction in NKG2Cpos CD56dim NK Cells in the EBV-cGVHD Cohort
The pattern of NKG2C expression inversely correlated with that of NKG2A expression. In the EBV group, CMV reactivation occurred at a median of 30 days (range, 18–45) and resolved at a median of 60 days (range, 30–65). NKG2C expressions on day +30 were 3.5% ± 2.03%, which increased to 16.9 ± 5.8 at day +60 (P = 0.004). The NKG2A expressions in the same cohort decreased from 45.3% ± 21.7% at day +30 to 21.9% ± 6.4% at day +60 (P = 0.03). While the NKG2Cpos CD56dim NK cells showed a similar trend in the EBV-cGVHD cohort as well as the control cohort at day +60 (16.9% ± 5.8% and 20.1% ± 6.3%), this subset decreased markedly during the study period from 16.9% ± 5.8% to 4.1% ± 2.8% at day +150 (P = 0.005; Figure 4B), concomitant with the increase in NKG2Apos CD56dim 16pos NK cells in the EBV-cGVHD cohort from 20.0 ± 5.15 at day +60 to 59.08 ± 16.3 at day +150 (P = 0.0001; Figure 4A).
Restoration of NKG2A and NKG2C Subsets After Remission of Chronic GVHD
These subsets were reassessed 3 months after subsidence of chronic GVHD (median day +365 posttransplantation) when patients were off immunosuppression in the EBV-cGVHD cohort. Expressions of both NKG2Cpos (34.8% ± 10.9%) and NKG2Apos (16.5% ± 4.65%) subsets reversed to resemble the pattern seen in the control cohort (P < 0.01; Figure 4A and B).
In the 3 patients without EBV reactivation who developed mild and self-limiting chronic GVHD, the pattern of reconstitution was similar to the control cohort, with a progressive decline in NKG2AposCD56dim cells and increase in NKG2Cpos CD56dim populations (data not shown).
T-cell–depleted HCT is associated with several-fold higher risk of EBV reactivation and subsequent development of EBV-LPD. In a study on 972 patients, it was shown that the risk of EBV reactivation was 65% in TCD-HCT, compared with 31% in those not receiving a TCD graft.24 EBV-LPD in this study was witnessed only in recipients of TCD graft. Another registry-based study identified both haploidentical HCT and TCD as dominant risk factors for EBV-LPD.5 In this context, the impact of T-costimulation blockade with CTLA4Ig-based haplo-HCT as developed by our group assumes significance.
We had earlier shown that there is no increase in CMV reactivation following haploidentical HCT, combining CTLA4Ig and PTCy, and there was virtually no recurrent CMV infection.25 In addition, the incidence of adenovirus infections was significantly reduced with this protocol. CTLA4Ig has long been used in patients with rheumatoid arthritis. In patients with rheumatoid arthritis, EBV load is estimated to be several times higher than the normal population. However, CTLA4Ig did not increase EBV proliferation in this population even when used over a prolonged period.26 In our study, EBV reactivation was only 13.8% in patients undergoing haplo-HCT for very high–risk diseases. More importantly, the viral load was modest in quantity as well as the duration, with only 3 patients requiring 2 doses of rituximab. None of the patients had viral copy numbers exceeding 1 × 104/mL, and none developed EBV-LPD.
The median time to EBV reactivation in our study was 96 days, mostly within first 100 days, which is in keeping with most prospective studies of this nature.24 However, the most significant finding from our study was the temporal relationship between EBV and the onset of chronic GVHD within 6 weeks in the majority. There was no other predisposing factor for chronic GVHD observed in our study. In an European Society of Blood and Marrow Transplantation registry-based study on 11 364 patients with acute leukemia, donor EBV seropositivity was strongly associated with chronic GVHD.11 However, a pathophysiologic correlation has remained conjectural.
In the absence of any quantitative correlation between reconstitution of T- and NK cell subsets and EBV reactivation or development of GVHD, we had analyzed the NKG2A and NKG2C expressions at various time-points before and after EBV reactivation. NKG2A and NKG2C belong to the C-type lectin family and exist in a dimeric state with CD94. NKG2A is an inhibitory receptor with HLA-E as its cognate ligand. NKG2A is expressed at a very early stage of differentiation, mostly in CD56 bright immature NK cells, probably as an evolutionary mechanism in the prevention of autoreactivity, until the education and licensing of NK cells are ensured.27 With stepwise maturation of NK cells, the expression of NKG2A reduces with a reciprocal increase in the expression of NKG2C, which is an activating receptor and a subset of NKG2C cells expresses a memory phenotype in response to CMV infections.28
The introduction of CTLA4Ig and CTLA4Ig-primed DLI was associated with rapid rise in CD56dim population with higher KIR and lower NKG2A expressions as described previously.13 However, in all patients with EBV reactivation, there was a surge in CD56dim NKG2Apos population during and after EBV reactivation, leading on to the onset of chronic GVHD, which was not seen in those without EBV reactivation or chronic GVHD. This population of CD56dim NKG2Apos cells has been shown to attenuate the severity of primary EBV infection in children.7 More importantly, this population was shown to persist at high levels for several months after EBV infection, which is not witnessed in the natural course of NK cell maturation. In another study, NKG2A+ NK cells were demonstrated to be the predominant population responsible for eliminating EBV-infected autologous lymphoblastoid cell line.29 The CD56dim NKG2Apos NK cells are one of the subsets recognized at the early stage of differentiation, which are soon replaced by CD56dim NKG2Aneg and subsequently NKG2Cpos populations, a pattern observed in the cohort of patients who did not have EBV reactivation or chronic GVHD. This suggests that this alteration might have played a pathogenetic role in chronic GVHD.
The CD56dim NKG2Apos population has not been studied widely, and its role if any in the genesis of GVHD is not succinctly clear. However, another study which had also studied the kinetics of both NKG2Apos and NKG2Cpos populations following allogeneic HCT found a strong inverse correlation between these populations and occurrence of chronic GVHD, more so in recipients of HLA-mismatched grafts.30 They observed a significantly higher NKG2A/NKG2C ratio in patients with chronic extensive GVHD. These findings were akin to our observations, but the pathophysiology behind this alteration was not studied. EBV infection has been shown to promote NKG2A expressing NK cells, both CD56bright and CD56dim, without upregulating NKG2C expression, unlike CMV infection.31 Because the patients were monitored prospectively, it is clear from our data that alterations in NK cell subsets preceded the onset of chronic GVHD, thus raising the possibility that low-grade EBV reactivations might predispose to the development of more extensive GVHD.
Despite only a limited number of patients developing chronic GVHD on this protocol, the prospective correlation between EBV and the reciprocal alterations of NKG2A and NKG2C expressions with the onset and remission of chronic GVHD is compelling enough to suggest a causative association. Because the use of rituximab for EBV reactivation had no impact on the occurrence, severity, or pattern of chronic GVHD, the triggering process might be an adverse effect of CD56dimNKG2Apos subset on the CD4+ T cells.32 NK cells might aggravate or dampen T-cell alloreactivity through its effect on dendritic cells or via direct engagement with T cells.33 It is possible that the CD56dimNKG2Apos subset might be lytic for EBV-infected B cells but has a promoting effect on T cells which might have primarily helped contain EBV proliferation.7 However, persistence of this population which promotes T-cell activation, coupled with inverse reduction of NKG2Cpos population which is lytic for dendritic cells as well as activated CD4+T cells, might explain the pathogenesis of chronic GVHD in the context of EBV reactivation.34 Finally, the recovery and interaction of immune subsets are products of a specific transplant protocol and the pattern of NK cell alterations described here might be unique to our protocol employing CTLA4Ig in haploidentical HCT linking EBV reactivation with chronic GVHD. However, the pathophysiology of chronic GVHD is likely to be multifaceted and this might be one of the several pathways to be considered.
In conclusion, our study demonstrates a significant moderation in EBV reactivation and complete absence of EBV-LPD following haploidentical HCT employing sequential CTLA4Ig and PTCy. However, our study also suggests a probable pathogenetic role of EBV in initiation of chronic GVHD, possibly mediated by altered NK cell subsets. Further studies delving into this possible pathway linking EBV reactivation with chronic GVHD deserves further attention.
The authors thank all the patients and family members who participated in this study. The authors also thank each and every member of our department for their help in patient care and execution of the study.
1. Cohen JI. Epstein-Barr virus infection. N Engl J Med. 2000; 343:481–492
2. Rasche L, Kapp M, Einsele H, et al. EBV-induced post transplant lymphoproliferative disorders: a persisting challenge in allogeneic hematopoetic SCT. Bone Marrow Transplant. 2014; 49:163–167
3. Buyck HC, Ball S, Junagade P, et al. Prior immunosuppressive therapy with antithymocyte globulin increases the risk of EBV-related lymphoproliferative disorder following allo-SCT for acquired aplastic anaemia. Bone Marrow Transplant. 2009; 43:813–816
4. Cohen JM, Cooper N, Chakrabarti S, et al. EBV-related disease following haematopoietic stem cell transplantation with reduced intensity conditioning. Leuk Lymphoma. 2007; 48:256–269
5. Landgren O, Gilbert ES, Rizzo JD, et al. Risk factors for lymphoproliferative disorders after allogeneic hematopoietic cell transplantation. Blood. 2009; 113:4992–5001
6. Chakrabarti S, Milligan DW, Pillay D, et al. Reconstitution of the Epstein-Barr virus-specific cytotoxic T-lymphocyte response following T-cell-depleted myeloablative and nonmyeloablative allogeneic stem cell transplantation. Blood. 2003; 102:839–842
7. Azzi T, Lünemann A, Murer A, et al. Role for early-differentiated natural killer cells in infectious mononucleosis. Blood. 2014; 124:2533–2543
8. Williams H, McAulay K, Macsween KF, et al. The immune response to primary EBV infection: a role for natural killer cells. Br J Haematol. 2005; 129:266–274
9. Chijioke O, Azzi T, Nadal D, et al. Innate immune responses against Epstein Barr virus infection. J Leukoc Biol. 2013; 94:1185–1190
10. Chijioke O, Müller A, Feederle R, et al. Human natural killer cells prevent infectious mononucleosis features by targeting lytic Epstein-Barr virus infection. Cell Rep. 2013; 5:1489–1498
11. Styczynski J, Tridello G, Gil L, et al. Impact of donor Epstein-Barr virus serostatus on the incidence of graft-versus-host disease in patients with acute leukemia after hematopoietic stem-cell transplantation: a study from the acute leukemia and infectious diseases working parties of the European Society for Blood and Marrow Transplantation. J Clin Oncol. 2016; 34:2212–2220
12. Jaiswal SR, Bhakuni P, Zaman S, et al. T cell costimulation blockade promotes transplantation tolerance in combination with sirolimus and post-transplantation cyclophosphamide for haploidentical transplantation in children with severe aplastic anemia. Transpl Immunol. 2017; 43–44:54–59
13. Jaiswal SR, Bhakuni P, Joy A, et al. CTLA4IG primed donor lymphocyte infusion: a novel approach to immunotherapy after haploidentical transplantation for advanced leukemia. Biol Blood Marrow Transplant. 2019; 25:673–682
14. Jaiswal SR, Bhakuni P, Joy A, et al. Higher CD45RA+ regulatory T cells in the graft improves outcome in younger patients undergoing T cell-replete haploidentical transplantation: where donor age matters. Biol Blood Marrow Transplant. 2018; 24:2025–2033
15. Jaiswal SR, Chakrabarti A, Chatterjee S, et al. Haploidentical transplantation in children with unmanipulated peripheral blood stem cell graft: the need to look beyond post-transplantation cyclophosphamide in younger children. Pediatr Transplant. 2016; 20:675–682
16. Jaiswal SR, Chakrabarti A, Chatterjee S, et al. Hemophagocytic syndrome following haploidentical peripheral blood stem cell transplantation with post-transplant cyclophosphamide. Int J Hematol. 2016; 103:234–242
17. Kean LS, Hamby K, Koehn B, et al. NK cells mediate costimulation blockade-resistant rejection of allogeneic stem cells during nonmyeloablative transplantation. Am J Transplant. 2006; 6:292–304
18. Peng Y, Luo G, Zhou J, et al. CD86 is an activation receptor for NK cell cytotoxicity against tumor cells. Plos One. 2013; 8:e83913
19. Jaiswal SR, Zaman S, Chakrabarti A, et al. Improved outcome of refractory/relapsed acute myeloid leukemia after post-transplantation cyclophosphamide-based haploidentical transplantation with myeloablative conditioning and early prophylactic granulocyte colony-stimulating factor-mobilized donor lymphocyte infusions. Biol Blood Marrow Transplant. 2016; 22:1867–1873
20. Jaiswal SR, Zaman S, Nedunchezhian M, et al. CD56-enriched donor cell infusion after post-transplantation cyclophosphamide for haploidentical transplantation of advanced myeloid malignancies is associated with prompt reconstitution of mature natural killer cells and regulatory T cells with reduced incidence of acute graft versus host disease: a pilot study. Cytotherapy. 2017; 19:531–542
21. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995; 15:825–828
22. Jagasia MH, Greinix HT, Arora M, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. The 2014 diagnosis and staging working group report. Biol Blood Marrow Transplant. 2015; 21:389–401.e1
23. Jeulin H, Salmon A, Bordigoni P, et al. Comparison of in-house real-time quantitative PCR to the adenovirus R-gene kit for determination of adenovirus load in clinical samples. J Clin Microbiol. 2010; 48:3132–3137
24. van Esser JW, van der Holt B, Meijer E, et al. Epstein-Barr virus (EBV) reactivation is a frequent event after allogeneic stem cell transplantation (SCT) and quantitatively predicts EBV-lymphoproliferative disease following T-cell–depleted SCT. Blood. 2001; 98:972–978
25. Jaiswal SR, Bhakuni P, Joy A, et al. Early expansion of CD56dimNKG2Alow with late surge and persistence of CD56dimNKG2AnegNKG2Cbright NK cells attenuate cytomegalovirus (CMV) replication and recurrence as well as leukemia relapse following haploidentical HSCT with T cell co-stimulation blockade and PTCy. Biol Blood Marrow Transplant. 2019; 25:S328
26. Balandraud N, Texier G, Massy E, et al. Long term treatment with abatacept or tocilizumab does not increase Epstein-Barr virus load in patients with rheumatoid arthritis - a three years retrospective study. Plos One. 2017; 12:e0171623
27. Béziat V, Descours B, Parizot C, et al. NK cell terminal differentiation: correlated stepwise decrease of NKG2A and acquisition of KIRs. Plos One. 2010; 5:e11966
28. Della Chiesa M, Sivori S, Carlomagno S, et al. Activating KIRs and NKG2C in viral infections: toward NK cell memory? Front Immunol. 2015; 6:573
29. Hatton O, Strauss-Albee DM, Zhao NQ, et al. NKG2A-expressing natural killer cells dominate the response to autologous lymphoblastoid cells infected with Epstein-Barr virus. Front Immunol. 2016; 7:607
30. Kordelas L, Steckel NK, Horn PA, et al. The activating NKG2C receptor is significantly reduced in NK cells after allogeneic stem cell transplantation in patients with severe graft-versus-host disease. Int J Mol Sci. 2016; 17:E1797
31. Hendricks DW, Balfour HH Jr, Dunmire SK, et al. Cutting edge: NKG2C(hi)CD57+ NK cells respond specifically to acute infection with cytomegalovirus and not Epstein-Barr virus. J Immunol. 2014; 192:4492–4496
32. Cook KD, Kline HC, Whitmire JK. NK cells inhibit humoral immunity by reducing the abundance of CD4+ T follicular helper cells during a chronic virus infection. J Leukoc Biol. 2015; 98:153–162
33. Pallmer K, Oxenius A. Recognition and regulation of T cells by NK cells. Front Immunol. 2016; 7:251
34. Laffont S, Seillet C, Ortaldo J, et al. Natural killer cells recruited into lymph nodes inhibit alloreactive T-cell activation through perforin-mediated killing of donor allogeneic dendritic cells. Blood. 2008; 112:661–671