Humoral and cellular immune responses to COVID-19 mRNA vaccines in immunosuppressed liver transplant recipients

Liver transplant recipients (LTRs) are especially vulnerable to severe COVID-19 due to chronic immunosuppression and comorbidities. Prior reports indicate that LTRs exhibit reduced immunogenicity to SARS-CoV-2 mRNA vaccines compared with healthy donors (HDs) and other immunosuppressed cohorts, but the extent and durability of humoral and cellular immune memory remain insufficiently characterized. This study aimed to define the magnitude and quality of vaccine-induced anti–receptor-binding domain (RBD) antibodies and spike-specific CD4+ and CD8+ T-cell responses in LTRs compared with HDs, to determine how immunosuppressive regimens influence these responses, and to assess whether a third vaccine dose augments humoral and cellular immunity, including cross-reactivity to Omicron sublineages.

Existing studies have shown lower serologic responses to SARS-CoV-2 mRNA vaccines in liver and other solid organ transplant recipients relative to healthy individuals, with improvements after a third dose in some cohorts. Anti-CD20 therapies dramatically blunt humoral immunity while partially preserving T-cell responses in other diseases. In healthy cohorts, mRNA vaccines induce durable humoral and cellular immunity, including cross-reactive memory, but CD8+ T-cell induction can be modest and affected by variant evolution. Bivalent Omicron-containing boosters increase breadth of neutralization against Omicron sublineages in animal models and humans. The role of calcineurin inhibitors and combination immunosuppression in shaping vaccine-elicited T-cell help and antibody production in LTRs has been highlighted but is not fully resolved.

Design and participants: Longitudinal cohort of 98 individuals from Osaka University, Japan: 44 healthy donors (HDs) and 54 liver transplant recipients (LTRs) receiving BNT162b2 or mRNA-1273. Time points included pre-vaccination and serial collections at approximately 1, 3, and 6 months after the second dose, pre-third dose (6 months), and 1 month after the third dose. Samples: Peripheral blood mononuclear cells (PBMCs) isolated by density gradient centrifugation with CPT tubes and cryopreserved; plasma collected for serology and neutralization. Serology: ELISA to quantify total IgG to SARS-CoV-2 spike RBD (recombinant RBD-coated plates; biotinylated anti-human IgG detection; HRP-streptavidin; OD 450 nm; endpoint titers determined via standard curves; Omicron BA.1 reference used for titer calibration). Neutralization: Pseudotyped virus neutralization titers (pVNT50) on HEK-293T cells expressing ACE2 and TMPRSS2; luciferase readout (Firefly); pVNT50 defined as plasma dilution with 50% inhibition. Cellular assays: Flow cytometry of PBMCs after stimulation with overlapping spike peptides; detection of spike-specific CD4+ T cells by CD154 and 4-1BB and of CD8+ T cells by activation/cytokine markers (e.g., CD69, 4-1BB) with memory subset markers (CD27, CD45RO, CD57) to define CM/EM/Effector phenotypes; intracellular staining for cytotoxic molecules (GZMA, GZMB, Perforin). Variant analyses: Antibody and T-cell reactivity assessed to Wuhan-1 and Omicron sublineages (e.g., BA.1, BA.2, BA.5, BQ.1.1, XBB); fold-changes calculated relative to Wuhan-1. Statistics: Group comparisons by Mann–Whitney U and Wilcoxon matched-pairs signed-rank tests; correlations by Spearman’s rank; multivariable logistic regression to identify predictors of weak/strong antibody responses; P-value significance thresholds denoted as * <0.05, ** <0.01, *** <0.001, **** <0.0001. Data collection and analysis were performed with GraphPad Prism, SPICE, and R.

• Anti-RBD IgG titers were significantly lower in LTRs than in HDs after the second vaccination; combination immunosuppressive therapy (CNI plus additional drugs) was associated with the weakest antibody responses compared with CNI monotherapy.
• Multivariable analysis indicated the number of immunosuppressive drugs had the largest negative impact on antibody titers.
• A third vaccine dose increased anti-RBD IgG titers in LTRs and HDs, though LTR titers generally remained below those of HDs; positive induction rates after the third dose were lower in CNI-treated groups than in HDs.
• Neutralizing activity (pVNT50) improved modestly after a third dose, with greater gains against BA.5 than against BQ.1.1 or XBB, where neutralization remained poor.
• Substantial fold-reductions in anti-RBD binding titers against Omicron sublineages were observed before the third dose, particularly for BQ.1 and XBB (e.g., HDs: ~8.43-fold and ~11.9-fold reductions; CNI: ~5.23-fold and ~6.35-fold; CNI + other drugs: ~4.41-fold and ~4.41-fold, respectively). Similar patterns persisted post-third dose for several sublineages.
• Spike-specific CD4+ T-cell responses were reduced in LTRs receiving multiple immunosuppressants vs HDs after the second dose (e.g., at 1, 3, and 6 months: p = 0.0117, 0.0208, 0.0047), but increased after the third dose, narrowing differences between groups.
• CD4+ T-cell responses (CD154+ total and Tfh) correlated positively with anti-RBD IgG titers at 1 month after the second dose and at later time points (HDs: r = 0.429, p = 0.049; LTRs: r = 0.488, p = 0.001), linking T-cell help with humoral immunity.
• Spike-specific CD8+ T cells were detectable in most participants (100% of HDs and 93% of LTRs at 1 month after the second dose) but were quantitatively lower in LTRs; memory phenotype shifted from central memory to effector memory over time in both cohorts.
• No qualitative differences in spike-specific CD8+ T cells were observed between groups by cytotoxic molecule expression (GZMA, GZMB, Perforin); most spike-specific CD8+ T cells expressed GZMA before and after the third dose.
• A third vaccination did not boost spike-specific memory CD8+ T-cell frequencies in either HDs or LTRs.
• CD4+ and CD8+ T-cell responses showed cross-reactivity to Omicron sublineages (e.g., BA.5, BQ.1.1, XBB), with preserved T-cell recognition despite reduced antibody binding/neutralization.

This study addresses how immunosuppression in liver transplant recipients modulates mRNA vaccine-induced humoral and cellular immunity. The findings demonstrate that combination immunosuppressive regimens markedly impair antibody induction and attenuate CD4+ T-cell responses, implicating reduced T-cell help as a contributor to poor serologic responses. A third vaccine dose improved anti-RBD IgG and CD4+ T-cell responses in LTRs, supporting booster vaccination to enhance protection, although LTR titers often remained below those of HDs. In contrast, memory CD8+ T-cell responses did not increase after a third dose in either cohort, suggesting that current Wuhan-1–based mRNA boosters inadequately expand spike-specific CD8+ T-cell memory. Importantly, despite strong antigenic drift limiting antibody recognition—especially against BQ.1.1 and XBB—both CD4+ and CD8+ T cells retained cross-reactivity to Omicron sublineages, indicating that cellular immunity may contribute to protection from severe disease even when neutralization is compromised. Clinically, these observations support additional or variant-adapted vaccination strategies in LTRs and highlight the need to tailor immunosuppression and vaccine approaches to optimize both humoral and T-cell immunity.

mRNA vaccination elicits measurable humoral and spike-specific CD4+ T-cell responses in liver transplant recipients, but responses are blunted by multi-drug immunosuppression and remain inferior to healthy donors. A third vaccine dose augments antibodies and CD4+ T cells yet fails to boost memory CD8+ T-cell frequencies. While T-cell cross-reactivity to Omicron sublineages is largely preserved, antibody recognition and neutralization of certain sublineages (e.g., BQ.1.1, XBB) remain reduced. Future work should evaluate variant-updated (e.g., bivalent) or next-generation vaccines, strategies to enhance CD8+ T-cell memory, and the impact of tailoring immunosuppressive regimens on vaccine efficacy in LTRs.

Single-center cohort with modest sample size; heterogeneous immunosuppressive regimens may confound group comparisons; assays were generally performed once per figure/time point; vaccine platform was Wuhan-1–based mRNA, limiting inference to variant-updated vaccines; limited follow-up on clinical outcomes (e.g., breakthrough infection severity); no recent anti-CD20–treated LTRs were included, limiting generalizability to that subgroup.