Evaluation of CMV T-cell Mediated Immunity in Children with CMV DNAemia After Allogeneic Hematopoietic Cell Transplantation

Abstract Background Cytomegalovirus (CMV) is among the most common infectious complications following allogeneic hematopoietic cell transplantation (allo-HCT). A commercially available CMV T-cell immune panel (CMV-TCIP, Viracor®), reported as percentages of CMV-specific interferon gamma releasing CD4 and CD8 cells (%CMV-CD4 or %CMV-CD8 IFN-γ), may be used to assess CMV cell-mediated immunity (CMV-CMI). However, the utility of this assay among pediatric allo-HCT recipients remains unknown. Methods A retrospective analysis among pediatric allo-HCT recipients (age ≤18 years) from 1/2020 to 11/2023 who developed CMV DNAemia and had CMV-TCIP testing was performed. Per institutional protocol, weekly quantitative plasma CMV PCR was performed until day +100. Clinically significant CMV infection (csCMVi) was defined as CMV viral load (VL) ≥500IU/ml and pre-emptive therapy (PET) was provided for csCMVi. The most frequent indications for CMV-TCIP testing were to evaluate CMV-CMI to inform the need of CMV virus-specific T-cell therapy (CMV-VST) in the setting of resistant/refractory CMV, to monitor CMV-CMI as a proxy of VST expansion post CMV-VST, or to evaluate the need of ongoing secondary antiviral prophylaxis. A %CMV-CD4 or %CMV-CD8 IFN-γ ≤0.2% was categorized as a poor response, whereas a result >0.2% was classified as an adequate response. Specimens with insufficient quality, high background, or discrepant results from %CMV-CD4 or %CMV-CD8 were classified as a discordant response. Demographics, underlying condition, HCT type, conditioning regimen, antiviral agents and use of VST, VL, and CMV-TCIP results were recorded. Descriptive and nonparametric statistics (median, interquartile range [IQR]) were used for analysis. Results Twenty-seven CMV-TCIP results from 10 unique allo-HCT recipients were analyzed. The median age at transplant was 8.5 years (IQR 2.75-14.3), predominantly for hematologic malignancy (70%). CMV donor/recipient (D/R) serostatus were D+/R+ (60%), D-/R+ (30%) and D+/R- (10%). CMV DNAemia developed post-HCT in 80% (8/10) with median onset of 4 days post-HCT (IQR 3-40) and 2 subjects had known DNAemia prior to HCT. Nine (90%) subjects were started on PET for csCMVi, whereas 1 (10%) received antiviral therapy when VL was <500IU/ml. The first CMV-TCIP was obtained at median of 112 days (IQR 42-130) post-HCT. Despite appropriate antiviral therapy, patients with a peak VL ≥500 IU/mL (obtained within 1 week before and after CMV-TCIP testing) demonstrated significantly lower absolute CD4 (median 12 vs. 198 cells/uL, p=0.02) and CD8 (median 79 vs. 443 cells/uL, p=0.04) counts and %CMV-CD4 IFN-γ (0.04% vs. 0.36%, p=0.02) compared to those with a peak VL <500 IU/mL. Conclusion Our results suggest that CMV-CMI, particularly %CMV-CD4 IFN-γ >0.2%, was observed concurrently with downward trending VL, and may serve as a proxy of VST expansion in patients receiving CMV-VSTs. Figure 1A. The distribution of peak VLs over time based on %CMV-CD4 and %CMV-CD8 IFN-γ are demonstrated in Figure 1B. CMV peak VLs were lower with TCIP testing results reported as adequate combined responses compared to those with poor combined responses. In five subjects (50%) who received CMV-VSTs as part of CMV therapy, increased %CMV-CD4 and %CMV-CD8 IFN-γ were observed post-infusion with an observed downward trend in VL.