CXCR5⁺PD-1⁻ circulating T follicular helper cells identify vaccine readiness in pediatric allogeneic hematopoietic cell transplantation

Abstract

Abstract Background:
Children undergoing allogeneic hematopoietic cell transplantation (allo-HCT) are highly susceptible to vaccine-preventable infections due to prolonged delays in immune reconstitution. Current revaccination guidelines vary widely but typically rely on a combination of time from transplant and clinically available, broad measures of immune reconstitution. Unfortunately, many individuals continue to have poor vaccine response. Individualized measures of vaccine readiness are needed to optimize the timing of and response to vaccination. CD4+ circulating T follicular helper (cTfh) cells, defined by CXCR5 expression, are a promising candidate to predict vaccine readiness. Tfh are critical for germinal center function and long-lived antibody responses, and productive T:B cell interactions are required to generate normal frequency of cTfh. However, the dynamics of cTfh reconstitution after allo-HCT is not known, and the potential role of these cells in predicting vaccine readiness is not well established. Methods:
We conducted a prospective longitudinal study of pediatric patients at Children's Hospital of Philadelphia receiving SARS-CoV-2 mRNA vaccination following αβ T cell–depleted allo-HCT (n=11), alongside healthy controls (n=9) and solid organ transplant (SOT) recipients as immunosuppressed controls (n=22). Serial blood samples were collected pre- and post-vaccination. Vaccine responses were assessed by quantifying IgG against the receptor-binding domain (RBD) of the SARS-CoV-2 spike protein. Peripheral blood mononuclear cells (PBMCs) were analyzed by high-parameter spectral flow cytometry to characterize T and B cell reconstitution, including cTfh subsets (CXCR5⁺PD-1⁺ and CXCR5⁺PD-1⁻), functional Tfh1/Tfh17 polarization (CXCR3, CCR6), and memory B cells (CD27, IgD). Antigen-specific cellular responses were measured using flow cytometric activation-induced marker (AIM) assays and SARS-CoV-2 B cell probes. Results:
Serologic vaccine responses, measured by RBD-specific IgG, were impaired in allo-HCT recipients compared to healthy controls (p<0.01), while SOT recipients has variable non-responsiveness. Longitudinal sampling allowed for assessment of the recovery kinetics for CD4 T cell subpopulations. While CXCR5⁺PD-1⁺ cTfh cells recovered early in both groups, CXCR5⁺PD-1⁻ cTfh cells remained significantly reduced in allo-HCT recipients beyond 1.5 years post-transplant. Frequencies of CXCR5⁺PD-1⁻ cTfh cells were markedly lower in vaccine non-responders compared to responders in the HCT group (p<0.05), but not in SOT. In parallel, memory B cell recovery was also impaired in allo-HCT recipients who failed to respond to vaccination, with persistent reductions in both class-switched and unswitched CD27⁺ memory B cells. Interestingly, Tfh1-skewed cTfh polarization (CXCR3⁺) was associated with vaccine responsiveness in SOT recipients (p<0.05). CXCR5⁺PD-1⁻ cTfh showed strong correlation with serologic response (Spearman ρ = 0.72, p = 0.012) and whereas CXCR5⁺PD-1+ cTfh (ρ = 0.28, p = 0.41) did not. Moreover, CXCR5⁺PD-1⁻ cTfh also correlated significantly with the recovery of memory B cells (ρ = 0.79, p = 0.0092). These correlations were specific to the HCT group and not observed in the SOT. SARS-CoV-2-specific B cells were quantified using fluorescent spike and RBD probes. The frequency of the RBD+ memory B cells was reduced in HCT and was the lowest in the non respneders. Additionally, the presence of spike⁺ MBC cells positively correlated with CXCR5⁺PD-1⁻ cTfh frequencies (ρ = 0.78, p = 0.0068). Lastly, antigen-specific CD4⁺ T cell responses were assessed using activation-induced marker (AIM) assays (CD69⁺OX40⁺) following stimulation with SARS-CoV-2 peptide pools. Among allo-HCT recipients, 7/11 (64%) showed detectable CD4⁺ T cell responses. Conclusions:
Our findings suggest that CXCR5⁺PD-1⁻ cTfh cells are a key correlate of vaccine readiness in pediatric allo-HCT recipients. Our data suggest that reliance on fixed time intervals post allo-HCT may be insufficient to guide re-vaccination vaccination timing. Instead, individualized immune profiling using cTfh and B cell reconstitution could enable more effective vaccine timing. This approach has broad implications for clinical decision-making in transplant immunology, particularly as novel antigens and vaccines are introduced into post-transplant care. Larger prospective studies are needed to ensure generalizability across transplant protocols and across vaccine classes.