Hematopoietic stem cell donor vaccination with CMV Triplex augments frequencies of functional and durable CMV-specific T cells in the recipient: a novel strategy to limit antiviral prophylaxis

Abstract

To enhance protective cytomegalovirus (CMV)-specific T cells in immunosuppressed recipients of an allogeneic hematopoietic cell transplant (HCT), we evaluated post-HCT impact of vaccinating healthy HCT donors with Triplex. Triplex is a viral vectored recombinant vaccine expressing three immunodominant CMV antigens. The vector is modified vaccinia Ankara (MVA), an attenuated, non-replicating poxvirus derived from the vaccinia virus strain Ankara. It demonstrated tolerability and immunogenicity in healthy adults and HCT recipients, in whom it also reduced CMV reactivation. Here, we report feasibility, safety and immunological outcomes of a pilot phase 1 trial (NCT03560752 at ClinicalTrials.gov) including 17 CMV-seropositive recipients who received an HCT from a matched related donor (MRD) vaccinated with 5.1 ×10⁸ pfu/mL of Triplex before cell harvest (median 15, range 11–28 days). Donor and recipient pairs who committed to participation in the trial resulted in exceptional adherence to the protocol. Triplex was well tolerated with limited adverse events in donors and recipients, who all engrafted with full donor chimerism. On day 28 post-HCT, levels of functional vaccinia- and CMV-specific CD137⁺CD8⁺ T cells were significantly higher (p<0.0001 and p=0.0174 respectively) in recipients of Triplex vaccinated MRD than unvaccinated MRD (control cohort). Predominantly central and effector memory CMV-specific T cell responses continued to steadily expand through 1-year follow up. CMV viremia requiring antivirals developed in three recipients (18%). In summary, this novel approach represents a promising strategy applicable to different HCT settings for limiting the use of antiviral prophylaxis, which can impair and delay CMV-specific immunity, leading to CMV reactivation requiring treatment.

1. Introduction

Cytomegalovirus (CMV) reactivation after allogeneic hematopoietic stem cell transplantation (HCT), a curative treatment modality for a wide range of hematologic disorders, remains a leading clinical concern¹. Though prophylactic and/or preemptive antiviral therapy (PET) greatly reduced the risk of CMV disease and mortality post-HCT, they are associated with significant economic burden, toxicity, immune reconstitution impairment^2–5 and rates of failure⁶. Letermovir prophylaxis is effective with minimal myelosuppression and renal toxicity, however by 24 weeks post-HCT, clinically significant viremia rebounds, resistance occurs^7,8 and delayed onset of CMV infection is frequent^9,10. These findings are consistent with reports of CMV resistance developing in cell culture suggesting a low viral genetic barrier¹¹. Additionally, polyfunctional pp65 and IE1 T cell responses are markedly decreased in HCT patients receiving letermovir prophylaxis compared with recipients receiving PET¹². T cell immunity is central to lifelong control of this persistent virus and recovery of CMV-specific T lymphocytes ends the risk period¹³. Transfer of CMV memory T cells within the peripheral blood stem cell (PBSC) graft from a CMV-seropositive donor helps to reconstitute protective immunity¹⁴. Adoptively transferred CMV-specific T cells have been shown to establish protective immunity, hence development of CMV-cellular immunity at early time points post-HCT is essential for long-term viral control¹⁵. Nonetheless, success of immune-based therapeutic approaches can be critically hindered when immunosuppressive steroid therapy is required to control graft versus host disease (GVHD)¹⁶.

Here, we report the outcome of a pilot trial designed to provide early post-HCT robust and durable protection from CMV reactivation, by vaccinating HCT donors with Triplex before PBSC graft collection. The goal of this approach is to transfer CMV immunity elicited in the immunocompetent HCT donor to the vulnerable recipient with the graft infusion. Especially early post-HCT when GVHD treatment^17,18 most frequently occurs, the recipient’s ability to mount a vaccine response is compromised and viral reactivation cannot be controlled¹⁹. Triplex, a modified vaccinia Ankara (MVA) vectored CMV multi-antigen vaccine demonstrated tolerability and immunogenicity in healthy adults, autologous and allogenic and HCT recipients, in whom it reduced CMV viremia^16,20,21. MVA, a safe attenuated poxvirus that is replication-deficient in mammalian cells, has been frequently used as a viral vector, to prevent smallpox and more recently as a vaccine for monkeypox^22–25.

In the current study, we aim to assess whether CMV-specific T cell immunity elicited by Triplex in the HCT donor could be transferred and expanded in the recipient. Safely infusing a graft with enhanced levels of functional and durable CMV-specific T cells can enable the recipient to control CMV reactivation, therefore reducing or eliminating the need for immunosuppressive antivirals.

2. Methods

2.1. Clinical study design and oversight

This prospective study was an investigator-initiated phase 1 trial conducted at City of Hope (COH; Duarte, CA) to evaluate feasibility, safety and immunogenicity of vaccinating HCT matched related donors (MRDs) with Triplex vaccine. The trial protocol was approved by COH institutional review board and was undertaken in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki. All study participants provided written informed consent. The primary objectives of this trial were to establish the feasibility and safety (for both donor and recipient) of priming CMV immunity in donors by Triplex vaccination before PBSC harvest; and to longitudinally assess the therapy induced changes in frequency and quality of CMV-specific T cells. The secondary objectives included assessing the impact of Triplex vaccination on CMV viremia and use of antivirals. Adult MRDs and their CMV-seropositive recipients of a planned allogeneic HCT were approached to be enrolled in this study, registered at ClinicalTrials.gov as NCT03560752. Enrolled donors received one injection of 5.1 ×10⁸ pfu/mL Triplex vaccine¹⁶, 10–60 days before stem cell mobilization therapy. Use of myeloablative, reduced intensity, and non-myeloablative conditioning regimens were permitted pre-transplant. Conditioning regimen selection was allowed as per standard institutional guidelines. Eligible patients could not receive T cell depleting agents (e.g. anti-thymocyte globulin) or an ex-vivo T cell-depleted graft and were not planned to be administered letermovir prophylaxis. Recipients were clinically and immunologically evaluated up to 1-year post-HCT. Adverse events (AEs) and graft versus host disease (GVHD) were monitored for all participants as necessary, and not less than bi-weekly from day 28 until day 100 post-HCT. Afterwards, GVHD was monitored as necessary or monthly until 6 months, and subsequently as per standard of care (SOC). Further inclusion and exclusion criteria details and study schema are outlined in the supplemental material, available with the online version of this article.

2.2. Study participant specimens

Peripheral blood mononuclear cells (PBMC) from enrolled HCT MRD and recipient pairs (N=17) were separated from blood, following standard protocols¹⁶ and stored in liquid nitrogen. Immunological data from PBMC collected on day 28 post-HCT from a cohort of CMV-seropositive recipients receiving stem cells from unvaccinated MRD¹⁶ were used as a retrospective control dataset (Table 1S). These control recipients, who had consented to use their left-over specimens and data for future studies, did not receive letermovir prophylaxis, and on day 28 post-HCT no study (NCT02506933) intervention or administration of CMV preventive antiviral therapy had been provided¹⁶. As detailed¹⁶, rigid exclusion, protocol specified criteria were applied for the retrospective MRD cohort, which differ from the current MRD study recipients, who could be eligible as long as they were adult patients scheduled for a planned HCT (Intention-To-Treat-population).

2.3. HLA typing and chimerism

Both tests were done as a part of pre- and post-HCT routine clinical evaluation, as per SOC. Donor and recipient pairs HLA typing were performed using Next Generation Sequencing Kits (Scisco Genetics). Measurements of donor chimerism were performed following current guidelines on approximately days +30, +100, and +180 post-HCT, using quantitative PCR (GenDx) from peripheral blood, bone marrow or CD3 subset. CD3 was enriched using magnetic beads (Stemcell Technologies) prior to DNA isolation, according to manufacturer’s instructions

2.4. CMV-specific immune monitoring

CMV-specific cellular immunogenicity was monitored in PBMC of donors and recipients by measuring levels of T cells expressing the CD137 activation marker, and assessing the CD28, CD45RA memory phenotype profiles^20,26–29. Following 24 hours stimulation with either pp65, IE1 or IE2 peptide libraries and medium as control, PBMC from blood draws obtained before Triplex vaccination in the donors and at day 28, 42, 56, 70, 84, 100, 140, 180, 270, and 365 post-HCT were labeled and analyzed by fluorescence-activated cytometry (FC; Gallios^™, Beckman Coulter with Kaluza analysis software, Brea, CA)². Concentrations of pp65-, IE1- or IE2-specific CD3⁺CD4⁺CD137⁺ and CD3⁺CD8⁺CD137⁺ T cells were longitudinally measured using multiparameter (6 colors) FC. The lower limit of detection for CD3⁺ CD8⁺CD137⁺ or CD3⁺CD4⁺CD137⁺ T cells was 0.02% or 0.1 cells/μl. When either CMV-specific CD3⁺CD8⁺CD137⁺ T cell or CD3⁺CD4⁺CD137⁺ T cell populations were ≥0.2%, a further analysis for CD28 and CD45RA memory membrane markers was feasible and performed following previously reported gating strategies. CD45RA⁺CD28⁺ cells were classified as naïve, CD45RA⁻ CD28⁺ cells were classified as central memory (TCM), and CD28⁻ cells were classified as effector T cells. Within the effector T cell group, two subpopulations were identified: CD45RA⁻CD28⁻ cells (TEM) and CD45RA⁺CD28⁻ effector “revertant” T cells, re-expressing the RA isoform of the CD45 surface marker (TEMRA).

2.5. MVA vector-specific immune analysis and vector persistence.

To monitor Triplex viral vector-specific T cell responses from recipient blood collected on day (d)+28 post-HCT, PBMC were stimulated with HLA Class I Supertype A and B, epitopes of vaccinia virus (MVA is and attenuated vaccinia virus derived from the chorioallantois vaccinia virus strain Ankara²²), as described²⁰. MVA DNAemia persistence was monitored in the recipients for up to 1 year, as previously reported^16,20.

2.6. Statistical analysis

Descriptive statistics were used to analyze donor and patient characteristics. Groups were compared with the Mann–Whitney test using GraphPad Prism 9.4.1. Levels of memory T cell subsets were transformed using the arcsine square root formula to be compatible with the normal distribution. Linear mixed models were used to address the statistical significance of changes in memory T cell subsets, at specific days post-HCT. The Loess scattered plots were obtained using the ggplot2 package in R (https://cran.r-project.org/web/packages/ggplot2/index.html).

3. Results

3.1. Characteristics of the study population

Between August 2018 to July 2021, thirty-six matched related donors (MRD; 8/8 A, B, C, DRB1 high resolution) and recipient (R) pairs of a planned HCT for the treatment of hematologic malignancy were approached. Eighteen either declined study participation or were ineligible after screening, and another eighteen were enrolled in the trial (Table 1 and Figure 1S). Eligible donors were vaccinated prior to the start of stem cell mobilization and subsequent harvest for recipient infusion (median 15, range 11–28 days). No MRD or R discontinued the study for personal reasons or refused to donate blood during the multiple, per protocol planned blood draws. One recipient (UPN18) became ineligible for HCT before donor vaccination and the MRD/R pair was excluded from the study. Seventeen MRD and their CMV-seropositive recipients were included in the final analysis. UPN11 is the only trial recipient whose donor was CMV seronegative. Recipient UPN09 was lost to follow up on day +270 due to insurance issues. UPN16 (on day +40) and UPN17 (on day +287) relapsed and were censored, as per protocol. UPN15 expired 5 months post-HCT. Details of demographic, clinical characteristics and the summary of the trial outcomes are shown in Table 1.

Table 1.

Characteristics and outcomes of the MRD/R pairs

Characteristic at baseline*	Donors (n=17)	Recipients1 (n=17)
Median age (IQR), y	53 (40–60)	56 (44–60)
Female sex	10 (59)	9 (53)
Primary diagnosis
Acute lymphoblastic leukemia//lymphoblastic lymphoma		8 (47)
Acute myelogenous leukemia		7 (41)
Diffuse large B cell lymphoma		1 (6)
Precursor cell lymphoblastic lymphoma		1 (6)
T cell large granular lymphocytic leukemia		1 (6)
Karnofsky performance score (at conditioning for HCT)
100		3 (18)
90		10 (59)
80		4 (23)
Conditioning regimen
Reduced intensity		11 (65)
Myeloablative		6 (35)
CMV donor serostatus
Negative	1 (6)
Positive	16 (94)

Study Outcomes*
Primary outcomes
Feasibility
D/R pairs approached	36
D/R pairs enrolled	18 (50)
Safety
Grade 2–3 AE2 in donors	3 (18)
Non-relapse mortality at 100 days after HCT	0 (0)
Delayed engraftment	0 (0)
Severe (grade 3)3 acute GVHD	2 (12)
Grade 3–4 AE in recipients	0 (0)
Secondary outcomes
Clinically significant CMV viremia4	3 (18)
Relapse	2 (12)
All-cause mortality	1 (6)
Cellular immunity (see Figure 1)
Letermovir prophylaxis5	3 (18)

MRD/R = matched related/recipient; HCT = hematopoietic stem cell transplant; IQR = interquartile range; CMV = cytomegalovirus; D/R = donor/recipient; AE = adverse event; GVHD = graft-versus-host disease;

^*Values are numbers of patients (percentages) unless otherwise indicated.

¹All recipients were CMV seropositive.

²On the basis of Common Terminology Criteria for Adverse Events (CTCAE), version 5.0, at least probably or definitely related (see Table S2).

³No grade 4 was observed.

⁴CMV reactivation prompting antiviral therapy.

⁵Per physician decision.

3.2. Safety and CMV related outcomes

All recipients engrafted (Table 1) and achieved full donor chimerism (>95%)³⁰ early post-HCT (median day +33, data not shown). Adverse events (AEs, Table 2) after Triplex vaccination in donors were minimal, consistent with previous outcomes⁸. No serious AEs, grade 3 or 4 AEs probably or definitely related to donor vaccination, and no severe (grade 3 or 4) acute GVHD³¹ was observed in the recipients¹⁶. At all tested time points, MVA viral vector DNA was undetectable (data not shown)^16,20. CMV-viremia required PET in three recipients (18%: UPN11, 15 and 17). Per physician decision, letermovir prophylaxis was administered to UPN2 and 4 (3-month course⁹) and to UPN3 (10 days).

Table 2.

All attributable grade 2–4 (in donors) and 3–4 (in recipients) adverse events

From Triplex injection to cell harvest1
MedDRA2	Maximum Grade	Attribution	HCT donors (N = 17)
Administration site: Pain	3	Possible	1 (6%)
Fatigue	2	Probable	2 12%)
Fever	2	Probable	1 (6%)
Malaise	2	Probable	1 (6%)
Vertigo	3	Probable	1 (6%)
Headache	2	Probable	2 12%)
Headache	3	Probable	1 (6%)

Days 28 and 42 after stem cell infusion1
MedDRA2	Maximum Grade	Attribution	HCT recipients (N = 17)
None	3–4	Possible/Probable	0

¹Based on Common Terminology Criteria for Adverse Events (CTCAE, version 5.0). Attributions were possibly and probably related to vaccination in the donor from day of Triplex injection to cell harvest, and in the recipients on Days 28 and 42 after stem cell infusion. No worse attributions were observed.

²MedDRA denotes Medical Dictionary for Regulatory Activities.

3.3. Immunogenicity

Figure 1 summarizes the immune-monitoring outcome of functionally activated CMV-specific CD4 and CD8 T cell in the patients receiving a graft infusion from a MRD immunized with the Triplex vaccine. Figure 1A shows that on day +28 post-HCT, levels of CMV-specific CD137⁺CD8⁺ T cells were significantly higher (p=0.0174, Mann-Whitney) in recipients of Triplex vaccinated MRD (N=17) than in unvaccinated control MRD¹⁶ (N=41). As visualized by the gray band detailed in Figure 1B legend, elevated frequencies of functionally activated pp65-, IE1-, IE2-specific T cells persisted through day 100 post-HCT and continued to vigorously expand during immune-reconstitution through the 1-year follow up. The marginal geometric means of CMV-specific T cells/μl estimated from the Loess regression plot (Figure 1B) were 10.21 (95% CI, 7.56 to 13.78) at day +28; 14.86 (95% CI, 11.29 to 19.55) at day +100, and 43.16 (95% CI, 21.79 to 85.49) at day +365, respectively, which indicates a sustained increasing trend in CMV-specific T cells in the recipient population. As previously found in healthy adults and in HCT recipients vaccinated with Triplex^16,20, pp65-specific T cell expansions were the most consistent and the highest in magnitude among the recipient population (Figure 2A), though robust levels of both IE1- and IE2-specfic T cells were also observed in some patients (Figure 2B). As for CMV viremic UPN11, response to Triplex vaccination in the CMV seronegative donor and transfer of CMV-specific T cells to the recipient were minimal (Figure 2S). For UPN15 and UPN17, CMV-specific T cell levels dropped before viremia onset, when recipients became lymphopenic during corticosteroid treatment for acute GVHD. In the three viremic patients, CMV specific cellular immunity increased after PET discontinuation.

Figure 1. CMV-specific T cell monitoring.

(A) Comparison of functionally activated CMV specific CD137⁺CD3⁺CD8⁺ (left plot) and CD137⁺CD3⁺CD4⁺ T cell levels at day +28 post-HCT in patients receiving a stem cell graft from unvaccinated (control recipient cohort, N=41) versus Triplex vaccinated matched related donors (study recipients, N=17). The difference between group was calculated using the Mann-Whitney test and the p values are reported in each plot (B) Longitudinal levels of combined CD137⁺CD3⁺ CD4⁺ and CD137⁺CD3⁺ CD8⁺ T cells specific for pp65, IE1 and IE2 antigens. The band shown was computed using the loess scatterplot smoother providing the marginal geometric mean concentrations through time. A 95% confidence band is shown in gray, and individual measurement trajectories are shown for each participant up to 7 days before CMV reactivation, requiring antivirals (UPN 11, 15 and 17). Logarithmic spacing is used to aid visualization. (C) Box plots showing percentages on the arcsine scale of pp65-specific, CD137⁺CD8⁺ naïve, T central and effector memory phenotype (on each plot top, the memory subset is indicated). The box spans the interquartile range, the central bar shows the median, and the whiskers extend to 1.5 times the interquartile range. Levels of memory subsets were compared using linear mixed models. The calculated p values at days post-HCT, reported on the corresponding boxes with a symbol (*p<0.05; †p<0.01; ‡p<0.001; §p<0.0001), refer to comparisons of naive vs. central memory, naïve vs. effectors or central memory vs. effectors. Significantly lower levels of naive vs. central memory are indicated with green symbols on the corresponding day post-HCT central memory box; significantly lower levels of naive vs. effector memory with green symbols on the corresponding day post-HCT effector memory box. Likewise, the significantly lower levels of central memory vs. effectors starting on day 100 through 365 post-HCT are reported with blue symbols on the corresponding day post-HCT effector memory box; and the significantly lower levels of effectors vs. central memory on day 28 post-HCT is reported with the red symbol on the corresponding central memory box. MRD = matched related; HCT = hematopoietic stem cell transplant; SEM = standard error of the mean.

Figure 2. MVA and CMV-specific T cell responses.

(A) T cell levels of MVA (Triplex vaccinia virus vector)-specific CD137⁺CD3⁺CD8⁺ T cells in HCT control recipients (* and grey symbols; N=28) who received a stem cell graft from unvaccinated MRDs (17 of 28 MRDs were born before 1972 and may have received smallpox vaccination) compared to those measured in the study HCT recipients (red symbols; N=17) who received a stem cell graft from Triplex vaccinated MRDs (10 of 17 MRDs were born before 1972 and may have received smallpox vaccination). For the study HCT recipients, pp65, IE1 and IE2 CD137⁺CD3⁺CD8⁺T cell levels are also reported (black symbols). The p value refers to Mann-Whitney test used to compare the groups. (B) Longitudinal immune profiles of representative recipients UPN03 and UPN09 (lost to follow up at day +270) showing robust and sustained levels of post-HCT T cells specific for pp65, IE1 and IE2. The syringe symbol indicates the pre-transplant day of Triplex injection in the MRD.

The memory phenotype analysis (Figure 1C) focused on pp65-specific T cell responses and showed that functionally activated pp65-specific CD137⁺CD8⁺ T cells had a predominant central memory profile early post-HCT, with markedly reduced naïve T cell subsets. During post-HCT immune reconstitution, effectors gradually became the most prevalent subset. The memory phenotype of pp65-specific CD137⁺CD4⁺ T cells was largely composed of central memory T cell subsets (CD45RA⁻ CD28⁺, TCM; all median >90%)^27,32.

3.4. MVA vector immunity

To further confirm transfer of donor Triplex-driven immunity, we assessed whether vaccinia specific-T cells, elicited by the MVA poxvirus vector could be as well measured in recipients²⁰. As shown in Figure 2A, levels of MVA-specific CD137⁺CD3⁺CD8⁺ T cells were significantly higher (p<0.0001, Mann-Whitney) in the study participants (N=17) who received an infusion graft from a Triplex immunized MRD compared to available specimens (N=28) from a cohort of patients receiving an HCT from unvaccinated MRDs¹⁶. Within each cohort, levels of MVA-specific CD137⁺CD3⁺CD8⁺ T cells in recipients whose MRD was born before 1972, and presumably received smallpox vaccination (this vaccination was compulsory in the US), did not significantly differ from those measured in recipients whose MRD was born after 1972 (data not shown), as previously reported²⁰.

4. Discussion

There is an unmet need for an alternative safer approach than antivirals for controlling CMV reactivation post-HCT^33,34. We describe a novel strategy of immunizing donors pre-HCT with Triplex vaccine to prevent CMV reactivation, requiring toxic treatment and causing significant complication in CMV-seropositive recipients^35–37. This immunotherapeutic approach was designed to overcome the post-HCT immune impairment early post-HCT, when the risk of CMV reactivation is highest³⁸, and to improve and accelerate immune reconstitution. The outcome of this phase 1 pilot trial is the first reported evidence of a feasible, safe and successful attempt to increase protective CMV-specific T cell immunity pre-HCT through donor vaccination, in CMV-seropositive recipients. The trial met its primary endpoints: the HCT donor vaccination treatment resulted in augmented frequency of functional and durable CMV specific T cells, starting early post-HCT and displaying a persistent phenotype of experienced, central memory T cells.

Vaccination is the main preventative approach to limit morbidity and mortality in immunocompromised hematopoietic stem cell transplant recipients at high risk of post-HCT infections³⁹. Emphasis is usually placed on post-transplant vaccination, but there are data on the efficacy of pretransplant vaccinations, and trials have shown promising results if both recipients and donors receive vaccinations before transplantation^40,41. Recent evidence supports that immunization of donors and recipients before HCT can significantly enhance immunogenic T and B cell responses and increase response to vaccination post-HCT^29,42. However, there is limited availability of long-term safety data on using vaccination in donors and recommendation for donor vaccination still faces logistical and ethical challenges^43–45.

In our study population consisting of HCT donors, who were matched siblings or related to the recipients, adherence to the protocol was high, which was essential to ensure the overall success and feasibility of the trial. Both donors and recipients showed commitment, supportive and rigorous participation, which resulted in great follow-up compliance to the required multiple study visits, for assessing treatment safety and performing blood draws for immune monitoring research. In one single case, an out of state recipient (UPN09) was unable to complete follow-up and missed the last protocol visit (day 365) due to unresolved health insurance coverage issues.

As expected, and based on safety outcomes from previous trials of Triplex in healthy adults, allogeneic and autologous HCT, safety profiles in both donors and recipients were favorable, with no Triplex vaccine-associated safety concerns (Tables 1 and 2)^16,20,21.

The main finding (Figures 1 and 2) in the study recipients with Triplex vaccinated MRDs is the higher frequency of functionally activated CMV-specific T cells observed early post-HCT, compared to recipients with unvaccinated MRD. Moreover, the enhanced T cell levels were durable and continued to steadily expand during immune reconstitution. This is a highly favorable immunogenicity outcome, since T cell mediated cellular immunity is the most important factor in controlling CMV replication^46,47. Nonetheless a randomized, placebo-controlled trial is needed to confirm these encouraging, but preliminary comparisons. Several CMV proteins are targeted by the T cell response, including pp65, IE-1, IE-2^48–50. Future studies will assess additional T cell epitopes from immunogenic CMV encoded antigens⁴⁹, not expressed by Triplex to evaluate occurrence of subclinical CMV reactivation.

Early CMV reactivation remains associated with increased transplant-related mortality³⁷, and lack of CMV-specific CD4 and CD8 T cell immunity predicts early CMV reactivation and disease⁵¹. In the three recipients in whom PET was used to control early CMV viremia (Figure 2S), levels of CMV specific T cells were low before early viremia onset, due to corticosteroid treatment for acute GVHD. Among the study recipients, a single death (due to sepsis, on day +152 post-HCT), occurred in a viremic patient (UPN15) around two months after PET discontinuation (Figure 2S). In contrast, reconstitution of CMV-specific CD8 and CD4 T cell responses post-HCT correlates with long-lasting protection from uncontrolled CMV infection and improved outcome of CMV disease⁴⁶. High-dose post-transplant cyclophosphamide (PTCy)⁵² is increasingly used to prevent acute GVHD, and recent work indicate that PTCy neither induces pan-T cell depletion nor eliminates alloreactive T cells⁵³. Broad adoption of this new non-steroid GVHD prophylaxis could improve usage of CMV vaccination strategies for CMV seropositive recipients who remain at enhanced risk of CMV complications post-HCT^35–37.

Donor CMV serostatus is a strong predictor of the timing of T cell reconstitution post-HCT, highlighting the critical importance of donor-derived CMV-specific cellular immunity^12,14. CMV seropositive recipients (R⁺) transplanted using stem cells from a CMV seronegative donor (D⁻) are at heightened risk for increased duration of CMV reactivation and incidence of CMV disease⁵⁴. Triplex vaccination of CMV seronegative healthy volunteers and autologous HCT recipients was shown to induce robust and durable primary T cell responses^20,21. Therefore, assessing the impact of Triplex vaccinated HCT D⁻ in augmenting CMV specific protective T cells in vulnerable R⁺ is highly relevant. Unfortunately, due to the local (Los Angeles County²⁰) very high incidence of CMV seropositivity, we were able only to enroll one CMV seronegative donor in the current single center (COH) study. A future multicenter trial is planned, which we designed to include sufficient HCT D- vaccinated with Triplex to determine whether CMV serostatus of the donor influences the CMV outcome in the recipient.

In the study population, all recipients reached full donor chimerism at engraftment (median day +33). Hence, the robust levels of CMV-specific T cells observed at day +28 during the engraftment period were likely expanded in the donor by Triplex vaccination and transferred with the stem cell graft in the recipients. The concomitant finding of significantly enhanced levels of MVA viral vector-specific T cells at day +28 in Triplex vaccinated MRD recipients, compared to unvaccinated MRD recipients further supports that T cells, induced in donors by Triplex vaccination could be transferred into the recipients at infusion^29,42. Studies on durability of immunity after smallpox vaccination showed that vaccinia-specific antibodies may persist up to 75 years post-vaccination, whereas T cell responses have a half-life of 8–15 years⁵⁵. Data from the current trial showed similar trends. Participants who may have received previous smallpox vaccination who were born before 1972 had minimal pre-vaccination MVA-specific T cells (Figure 2A). Currently, evaluation of MVA vectored vaccines⁵⁶ for bivalent prophylaxis including monkeypox prevention is underway, in adult and pediatric transplant settings^57,58.

It is notable that in the study recipient population, immediately after stem cell transplant, pp65-specific CD8 T cells showed a memory phenotype pattern (Figure 1C) mainly consisted of antigen-experienced T lymphocytes, which subsequently acquired enhanced effector functions during immune reconstitution. Strong predominance of long-lived and functional effector T cells has been reported during primary CMV infection, since these cells are critical for viremia control^35,36,28,59. Our results also support the notion that after engraftment of allogeneic stem cells when complete donor chimerism is reached, circulating T cells in the recipient are mainly of donor origin, during the first year post-HCT. Donor derived T cells can promptly generate a functional anamnestic response upon recognition of antigens exposed to the donor preceding graft collection, including through vaccination before transplantation^40,41. Therefore, donor vaccination pre-graft can be a beneficial opportunity for the recipient to receive pools of mature and functional antigen specific T cells that can accelerate and augment a durable immune reconstitution, leading to control of post-HCT infections^40,41.

The current immunotherapeutic approach can be a desirable alternative to adoptive cellular therapy. Several studies have shown the clinical efficacy of adoptive transfer of pp65-specific T cells with effector cytotoxic functions to treat post-HCT CMV infection and disease^60–62. Nonetheless, its application has been limited by the labor-intensive nature of the in vitro culture methods to select and expand specific effector CMV-specific T cells from the HCT donor^46,63. Since both humoral and cellular immune responses have been implicated in protection against CMV viremia and disease in transplant patients³⁸, our group has developed a highly immunogenic MVA based vaccine, which simultaneously expresses pp65, the envelope pentamer complex (PC) and glycoprotein B (gB)⁶⁴. PC and gB are major targets of the B cell response and neutralizing antibodies^65–67. Future studies may reveal that immunizing HCT donors with vaccines stimulating both branches of the immune response may afford enhanced protection for the recipients. Nonetheless, Myers and collaborators in 1991 published⁶⁸ that minimal to no role of humoral immunity in resolution of CMV viremia post-HCT. Previous investigations have shown that both donor B and T cell immunity from pathogen exposure or vaccination pre-graft collection can be transferred to the recipient, expand during immune reconstitution, and can improve vaccination responses and/or control post-HCT infection^29,42.

This pilot phase 1 study was not powered to assess efficacy, and a phase 2 trial including a placebo arm will follow to confirm the current findings. Yet, the observed proportion of recipients requiring PET (18%, study secondary endpoint) suggests that the Triplex donor vaccination treatment induced a reduction in CMV reactivation rate when compared to MRD recipients prophylactically treated with letermovir⁹. This promising outcome suggests that recipient infusions containing abundant and durable donor derived central memory CMV-specific T cells, leading to enhanced supply of effectors^69,70 could have been the key correlate to reduced CMV reactivation^16,71–73. The upcoming phase 2 trial will establish the impact of enhanced specific CMV-specific T cell immunity on time to reactivation, duration of reactivation and the possible reduction of cost of care, compared to letermovir prophylaxis and/or PET when required. Based on feasibility, safety and marked immunogenicity in the current study, the Triplex HCT donor vaccination strategy could be also evaluated for higher risk unrelated donor and haploidentical HCT.

In summary, this novel approach has the potential to reduce the use of letermovir prophylaxis, overcoming breakthrough resistant viremia^8,74, delays and deficits in T cell reconstitution¹² and limit morbidity associated with high rates of CMV viremia rebound, requiring toxic antiviral treatment^9,75.