National Institutes of Health-Defined Chronic Graft-vs.-Host Disease in Pediatric Hematopoietic Stem Cell Transplantation Patients Correlates With Parameters of Long-Term Immune Reconstitution

Abstract

Recent data revealed the importance of immune reconstitution (IR) for the evaluation of possible biomarkers in National Institutes of Health (NIH)-defined chronic graft-vs.-host disease (cGVHD) and its clinical aspects. In this large pediatric study (n = 146), we have analyzed whether cellular and humoral parameters of IR in the long-term follow-up (FU) with a special emphasis on B-cell reconstitution correlate with NIH-defined cGVHD criteria. HYPOTHESIS: we were especially interested in whether meaningful cGVHD biomarkers could be defined in a large pediatric cohort. We here demonstrate for the first time in a highly homogenous pediatric patient cohort that both cGVHD (n = 38) and its activity were associated with the perturbation of the B-cell compartment, including low frequencies of CD19⁺CD27⁺ memory B-cells and increased frequencies of circulating CD19⁺CD21^low B-cells, a well-known hyperactivated B-cell subset frequently found elevated in chronic infection and autoimmunity. Notably, resolution of cGVHD correlated with expansion of CD19⁺CD27⁺ memory B-cells and normalization of CD19⁺CD21^low B-cell frequencies. Moreover, we found that the severity of cGVHD had an impact on parameters of IR and that severe cGVHD was associated with increased CD19⁺CD21^low B-cell frequencies. When comparing the clinical characteristics of the active and non-active cGVHD patients (in detail at time of analyses), we found a correlation between activity and a higher overall severity of cGVHD, which means that in the active cGVHD patient group were more patients with a higher disease burden of cGVHD-despite similar risk profiles for cGVHD. Our data also provide solid evidence that the time point of analysis regarding both hematopoietic stem cell transplantation (HSCT) FU and cGVHD disease activity may be of critical importance for the detailed investigation of pediatric cohorts. Finally, we have proven that the differences in risk factors and patterns of IR, with cGVHD as its main confounding factor, between malignant and non-malignant diseases, are important to be considered in future studies aiming at identification of novel biomarkers for cGVHD.

Keywords: B-cells; biomarker; chronic graft-vs.-host disease; immune reconstitution; pediatric hematopoietic stem cell transplantation.

Figure 1

Flow diagram. S1 General patient characteristics of study cohort (after exclusion of late acute GVHD); S2 age at time point of analyses and interval from HSCT to analyses; (1) patient characteristics regarding underlying diseases (comparison between malignant and non-malignant cohorts); (2) IR parameters in malignant diseases in no cGVHD cohort early (n = 47, a = 88) vs. late FU (n = 56, a = 197); (3) comparison between active (n = 29, a = 63) vs. no cGVHD (n = 67, a = 308) malignant only; (4) comparison between active (n = 29, a = 63) vs. resolved (n = 20, a = 65) cGVHD; malignant only; (5) comparison of resolved vs. no cGVHD, malignant only; (6) impact of NIH overall severity on IR parameters, malignant only; (7) clinical cGVHD characteristics of patients with malignant diseases. a, number of analyses. n, number of patients.

Figure 2

Box plots depict frequencies of CD19⁺CD21^low B-cells during long-term follow-up in malignant underlying diseases only, without the effect of cGVHD (a, number of analyses).

Figure 3

Box plots depict frequencies of CD19⁺CD21^low B-cells in correlation with activity of cGVHD at time of analyses (a, number of analyses).

Figure 4

Box plots depict frequencies of CD19⁺CD21^low B-cells in correlation with NIH-defined overall severity of cGVHD (a, number of analyses). In SPSS, small circle identified an outlier whereas ^* is an extreme value.