Cytomegalovirus-Specific Immunity Recovers More Slowly after Cord Blood Transplantation Compared with Matched Sibling Donor Allogeneic Transplantation

Abstract

BACKGROUND:

Rapid quantitative recovery of NK cells but slower recovery of T-cell subsets along with frequent viral infections are reported after umbilical cord blood (UCB) compared with matched sibling donor (MSD) hematopoietic cell transplantation (HCT). However, it remains unclear whether increased propensity for viral infections is also a result of slower recovery of virus-specific immunity after UCB as compared to MSD HCT.

OBJECTIVES:

We examined the differences in the function of virus-specific peripheral blood mononuclear cells (PBMC) after UCB (N=17) vs. MSD (N=9) using previously collected patient blood samples at various time points after HCT.

METHODS:

Interferon-gamma (IFN-γ) enzyme-linked immune absorbent spot (ELISpot) assay was used to quantify the PBMC frequencies that secrete IFN-γ in response to 11 immunopeptides from 5 common viruses. We included the patients who received the same reduced intensity conditioning regimen without ATG, no systemic glucocorticoids and had no relapse or acute/chronic graft-versus-host disease within 1 year after HCT.

RESULTS:

The CMV-reactive PBMC frequencies were higher in CMV seropositive vs. seronegative patients after HCT. Among CMV seropositive patients, the frequency of CMV-reactive PBMC was lower after UCB compared to MSD throughout one year of HCT. We observed no differences in virus-specific PBMC responses towards HHV6, EBV, BK, and adenovirus antigens between UCB and MSD.

CONCLUSION:

Our data demonstrate that the reconstitution of CMV-specific immunity is slower in CMV seropositive recipients of UCB vs. MSD HCT in contrast to other viruses which had similar recoveries. These study findings support implementation of more potent prophylactic strategies for preventing CMV reactivation in CMV seropositive patients receiving UCB HCT.

INTRODUCTION

Delayed immune reconstitution is a serious complication after hematopoietic cell transplantation (HCT) that can result in increased mortality from serious infections.^1-3 Cytomegalovirus (CMV) infection, in particular, remains one of the most fatal complications of HCT.⁴ We previously reported more rapid quantitative recovery of NK cells but slower recovery of T-cell (CD3+, CD8+, and CD4+) subsets along with higher frequency of viral infections after umbilical cord blood (UCB) compared with peripheral blood (PB) matched sibling donor (MSD) HCT.¹ Increased viral infections after UCB HCT as compared to MSD HCT have been attributed to lack of transfer of memory immune cells from donor to recipient and delayed immune cell recovery after UCB HCT.^2,5 However, prior studies focused on the quantitative recovery of immune cells after UCB compared to adult donor HCT.^1,3 Thus, it remains unclear whether the increased propensity for viral infections in UCB as compared to T cell replete PB MSD recipients is primarily explained by the slower recovery of immune cell numbers or also by the slower recovery of virus-specific immune cell function among PB mononuclear cells (PBMC). We therefore examined virus-specific immune cell recovery on previously collected blood samples at scheduled time points after UCB and PB MSD HCT.

METHODS

We analyzed 107 PB specimens from 26 adults that were prospectively collected for immune monitoring at days 60, 100, 180, and 360 after HCT. All study patients provided written informed consent prior to HCT and this study was approved by the University of Minnesota Institutional Review Board. In order to have a homogeneous patient population and to avoid factors affecting immune reconstitution we included only those patients who received the same reduced intensity conditioning (RIC) regimen without ATG, no systemic glucocorticoids and had >95% CD3 and CD33 PB donor chimerisms by day 60, no relapse or acute/chronic graft-versus-host disease (GVHD) within 1 year after HCT. Conditioning regimen consisted of fludarabine 150 mg/m², cyclophosphamide 50 mg/kg and 200cGy total body irradiation, and GVHD prophylaxis of cyclosporine and mycophenolate mofetil.⁶ Recipients of UCB and MSD HCT received the same supportive care, including acyclovir dose of 800 mg 5 times a day for CMV seropositive (CMV^pos) and 400 mg twice daily for CMV seronegative (CMV^neg) patients through day 100 of HCT, then all received acyclovir 400 mg twice daily for at least one year after HCT.^1,6,7

Interferon-gamma (IFN-γ) enzyme-linked immune absorbent spot assay (ELISpot) was performed using IFN-γ ELISpot kits (Bio-Techne Inc) to quantitate the frequency of PBMC that secrete IFN-γ in response to 11-mer overlapping immunopeptides: IE1 and pp65 for CMV, U90 for human herpesvirus 6 (HHV6), EBNA1, LMP2 and BZLF1 for Epstein-Barr virus (EBV), VP1 and large T for BK virus (BKV), and hexon and penton for adenovirus (HAdV3) (JPT Peptide Technologies, Gmbh). PBMC were thawed and kept overnight in L-Glutamine containing RPMI-1640 medium supplemented with 10% fetal bovine serum (Invitrogen) and 10 ng/mL human recombinant interleukin 2 (Bio-techne). The following day the PBMC were seeded at the 10⁸ cells/L concentration into duplicate wells of 96-well plate (200 microL volume), pulsed with peptides (concentration adjusted per JPT manufacturer recommendations), and incubated for 24 hours. Stimulation with the recombinant CD3/CD28 antibodies (100 ng/mL, Bio-Techne), lipopolysaccharides (10 ng/mL, Sigma-Aldrich) served as positive controls, while the dimethyl sulfoxide (0.01%) (Sigma-Aldrich) served as a negative control. Plates were developed according to the manufacturer’s instruction (Bio-Techne), scanned on Cellular Technology Limited (CTL) instrument, and the images were analyzed using SmartCount™ software (ImmunoSpot).

Immune reconstitution parameters after UCB and MSD HCT were summarized by median (95% CI, range) at each time point after HCT. Wilcoxon Rank Sum test and Kruskal-Wallis test for nonparametric data were used for the comparison between the groups. The cut-off significance level for all p-values was 0.05. All statistical analyses were implemented using Statistical Analysis System software version 9.4 (SAS Institute Inc., Cary, NC).

RESULTS

We examined the differences in quantitative and virus-specific immunity after UCB (N=17) versus PB MSD (N=9) after HCT for hematological malignancies. The recovery of blood NK cells was more rapid and particularly pronounced after UCB compared to MSD at day 100 and day 180 after HCT (Table). In contrast, when comparing T-cells after UCB versus MSD at day 60 after HCT, we observed lower median absolute counts (x10⁶/L) of blood CD3+ (208.8 vs. 408.3, p=0.03), CD8+ (39.1 vs. 112.1, p=0.02) and CD4+ (136.1 vs. 264.1; p=0.11). In subgroup analyses, the recovery of absolute T-cell counts was not significantly different in CMV^pos versus CMV^neg recipients when UCB and MSD cohorts were examined separately (data not shown).

Table.

Absolute immune cell count recovery after UCB vs. MSD HCT

Immune cell type	UCB N=17	MSD N=9	p-value
	Median number x106/L (range)
Day 60 after HCT
NK cells	238.7 (99.6-1002.0)	196.5 (74.4-306.9)	0.08
CD3+ T cells	208.8 (0.6-930.0)	408.3 (263.2-3279.6)	0.03
CD8+ T cells	39.1 (0.0-144.3)	112.1 (37.9-350.9)	0.02
CD4+ T cells	136.1 (0.1-791.4)	264.1 (182.4-2794.2)	0.11
Day 100 after HCT
NK cells	306.0 (167.2-1221.7)	114.9 (18.1-277.2)	<0.01
CD3+ T cells	266.8 (17.3-729.0)	452.9 (50.7-2324.4)	0.14
CD8+ T cells	57.6 (3.2-437.4)	223.7 (20.5-358.3)	0.10
CD4+ T cells	230.4 (3.6-370.0)	278.4 (18.0-1994.3)	0.22
Day 180 after HCT
NK cells	300.6 (135.3-1400.0)	111.6 (19.8-218.0)	<0.01
CD3+ T cells	447.2 (181.9-3792.9)	546.3 (118.4-1375.6)	0.86
CD8+ T cells	166.3 (14.9-2693.0)	140.1 (46.3-521.4)	0.79
CD4+ T cells	285.0 (131.4-834.4)	293.8 (67.8-740.1)	0.86
Day 365 after HCT
NK cells	252.0 (133.3-1014.0)	188.3 (114.4-548.1)	0.44
CD3+ T cells	624.0 (321.9-1951.7)	760.1 (191.8-804.3)	0.44
CD8+ T cells	242.0 (34.8-1362.3)	342.8 (63.3-423.1)	0.95
CD4+ T cells	326.9 (166.3-926.3)	328.2 (115.3-335.2)	0.51

UCB, umbilical cord blood. MSD, matched sibling donor. HCT, hematopoietic cell transplantation.

Among CMV^pos patients, there were 4 cases of CMV reactivation out of 10 UCB and none out of 4 MSD. The frequencies of CMV-reactive PBMC were higher in CMV^pos as compared to CMV^neg patients (Figure). However, among CMV^pos patients, particularly in those with CMV reactivation, the frequency of CMV-reactive PBMC was lower after UCB compared with MSD throughout one year of HCT.

Figure. Virus-specific immune reconstitution after UCB vs. MSD HCT by CMV serostatus.

The boxes show the interquartile range of spot-forming cell count (SFC) per 100,000 PBMC specific to each viral antigen by MSD vs. UCB donor types and by CMV serostatus. The bold horizontal lines inside the boxes and the numbers at the bottom of each panel indicate the median SFC per 100,000 PBMC; p-values on top compare the donor type/CMV serostatus groups. The whiskers represent 1.53 the height of the box (or minimum/maximum values if there is no value in that range); and the circles indicate the outliers. The individual figures show the SFC specific to (A) CMV_IE1 peptide by donor type and CMV serostatus; (B) CMV_IE1 peptide in CMV seropositive patients by CMV reactivation status; (C) CMVA_PP65 peptide by donor type and CMV serostatus; (D) CMVA_PP65 peptide in CMV seropositive patients by CMV reactivation status; (E) HHV6_U90 by donor type and CMV serostatus; (F) EBV_EBNA1 peptide by donor type and CMV serostatus, the results against EBV_LMP2 and EBV_BZLF1 peptides were similar to EBV_EBNA1 peptide (not shown); (G) BKV_LTA peptide by donor type and CMV serostatus, the results against BKV_VP1 peptide was similar to BKV_LTA (data not shown); and (H) HAdV3 peptide by donor type and CMV serostatus, the results against HAdV5 peptide was similar to HAdV3 (data not shown).

Other viral events included 6 HHV6 and 1 adenovirus reactivation in UCB recipients, and 1 EBV reactivation in MSD recipients. We detected no differences in virus-specific PBMC responses towards HHV6, EBV, BK and adenovirus antigens between UCB and MSD recipients throughout one year of HCT.

DISCUSSION

Our data suggest that, despite more rapid quantitative recovery of NK cells after UCB HCT, reconstitution of CMV-specific adaptive immunity is slower within first year of HCT in CMV^pos patients transplanted with UCB versus PB MSD. Thus, higher frequency of viral reactivations/infections after UCB HCT are explained not only by the slower recovery of T-cell numbers but also by the slower recovery of CMV-specific T-cell immunity. Relative protection from excessive viral infections in UCB recipients is likely explained by robust recovery of NK cells after UCB HCT.^1-3 NK cells are among the earliest cell types to respond to latent viral reactivations via secretion of IFN-γ, among other cytokines.⁸ However, without additional signals from T cells (e.g. IL-2 and IL-12) IFN-γ production by NK cells remains suppressed.⁹ We observed longitudinal increase in IFN-γ spot-forming units toward CMV antigens over time, which is likely related to an improvement in communication between innate and adaptive immune cells as previously reported.¹⁰

Promising results against CMV infection have been reported with use of novel antiviral prophylactic agents^11-13 and immunotherapy strategies.^14-16 Emerging prevention strategies also include CMV peptide vaccines and virus-specific T-cell adoptive transfer.^14-18 Poxvirus vectored triplex CMV vaccine administrated to CMV seropositive patients after HCT was recently reported to be effective in reducing the CMV viremia events.¹⁶ Similarly, adoptively transferred donor-derived multi-antigen- or CMV-specific cytotoxic T lymphocytes after HCT resulted in lower rates of CMV replication and tissue infection.^15,19

Our study findings support the implementation of more effective antiviral prophylactic strategies in order to reduce the risk of viral infections and thereby improve survival in CMV^pos patients receiving UCB HCT.

Highlights.

1. The CMV-reactive PBMC frequencies are higher in CMV seropositive vs. CMV seronegative patients after HCT.

2. Among CMV seropositive patients, the frequency of CMV-reactive PBMC is lower after UCB compared to MSD throughout one year of HCT.

3. We observed no differences in virus-specific PBMC responses towards HHV6, EBV, BK, and adenovirus antigens between UCB and MSD.