Dynamics of Immune Reconstitution in Adults with Sickle Cell Disease Undergoing Nonmyeloablative Matched Sibling Donor Stem Cell Transplantation: A Conventional and Computational Analysis

Background

Allogeneic hematopoietic stem cell transplantation (HSCT) with a matched sibling donor and nonmyeloablative conditioning with alemtuzumab/3Gy total body irradiation (TBI) is a curative therapy for adults with sickle cell disease (SCD). Despite mostly favorable outcomes, mixed chimerism, graft failure and infectious events still pose significant challenges, which might be related to perturbations in immune reconstitution. A comprehensive analysis of adaptive immune cells before and after allogeneic HSCT in SCD patients is lacking.

Aims

To characterize the immune phenotype of peripheral blood mononuclear cells (PBMC) before and after allogeneic HSCT in adults with SCD, compared to age- and ethnicity-matched healthy controls (HCs). To evaluate PBMC subsets in relation to infections and donor chimerism after transplantation.

Methods

We collected PBMCs of adults with SCD undergoing nonmyeloablative HSCT with alemtuzumab/TBI conditioning at baseline and 3, 6, 12, and 24 months post-transplant. Samples were analyzed using a 5x14-color flow cytometry panel to detect subsets of B, T, T-follicular helper (Tfh), NK cells, monocytes, and dendritic cells (DCs). Measurements were standardized using rainbow calibration beads. Cell populations were identified with unbiased clustering (FlowSOM) and with manual gating (FlowJo) of predefined subsets. Statistics included correction for multiple testing (Benjamini-Hochberg).

Results

We analyzed 85 longitudinal samples from 23 transplant recipients (median age 26 years, range 18 – 49, 48% females, 78% HbSS). Immunosuppression was stopped between 12 - 18 months post-transplant. Median donor T-cell chimerism 12 months post-transplant was 66% (IQR 54 - 78). Twelve HCs (30 years, 22 – 42, 43% females) were included as controls.

Within pre-gated B cells (CD45+CD3-CD14-CD19+), profound differences between SCD patients and HCs were identified at baseline: SCD patients had a higher proportion of cells in metaclusters containing naive transitional (IgD+IgM+CD27-CD38+CD24+IgG-) B cells (70% vs 54%, p <0.001) and a lower proportion in metaclusters containing memory (IgD-CD27+) and double negative (IgD-CD27-) B cell subsets (0.8% vs 3.4%, p <0.001 and 4.0% vs 8.8%, p= 0.008, respectively). After 24 months, B cells were reconstituted to proportions comparable with baseline. However, compared to HCs, significant differences in B cell profile persisted and were not ameliorated by HSCT. Manually gated analysis matched the unsupervised analysis.

Gated and unsupervised analysis of T cells (CD45+CD19-CD14-CD3+) showed CD4+ cells reconstituted slower than CD8+ cells. Naive (CCR7+45RA+) T cells were nearly absent at 3 months but recovered by 12 months to levels comparable with baseline and HCs. Compared to baseline, the proportions of effector (CCR7-45RA+, p= 0.019) and effector memory (CCR7-45RA-, p <0.001) T cells were larger at 3 months and decreased to baseline values by 12 months post-transplant. Central memory (CCR7+45RA-) T cells reconstituted slowly: In FlowSOM, one metacluster containing CD4+ central memory T cells remained diminished after 24 months (7.8%), compared to HCs (15%, p= 0.012). Regulatory T cells (CD4+CD8-CD25+CD127-FoxP3+) were enriched after 3 months (p= 0.027), decreasing to normal values by 12 months. Double negative (DN, CD3+CD4-CD8-) T cells were enriched at baseline, absent at 3 months, but at 12 months, comparable to proportions seen in HCs.

Monocyte, DC, and NK cell metaclusters reconstituted to proportions comparable with HCs. Proportions of conventional DC1 cells were significantly lower in patients at baseline and 12 months, but reconstituted to normal levels at 24 months post-HSCT.

Conclusion

SCD is characterized by altered B and T cell homeostasis, which is only partly restored after allogeneic HSCT, even with nonmyeloablative conditioning resulting in mixed lymphoid chimerism. Transplant recipients had a persistent increase of naive and reduced memory B cell subsets, probably due to irreversible SCD-related hyposplenia. Central memory T cells were less abundant at 24 months post-transplant. Our findings highlight the importance of re-vaccination, patient education, and pneumococcal prophylaxis even after successful cure.

Correlation analyses between PBMC subsets at the various time points and chimerism levels and infections are currently underway.