Hurdles for Hematopoiesis: Reconstitution and Clonality after Hematopoietic Cell Transplantation

Hematopoietic stem cell transplantation (HCT) is a curative therapy for various hematologic and immunologic diseases. Despite its therapeutic potential, HCT presents unique biological challenges, both during early hematopoietic reconstitution and in sustaining long-term hematopoietic function. This thesis investigates the integrity of post-transplant hematopoiesis, applying complementary methodologies—ranging from blood count analysis to clonal tracking and single-cell multiomics—to identify mechanisms of hematopoietic dysfunction after HCT.
First, we assessed the reconstitution of peripheral blood counts in the first months post-HCT. We found that definitions of poor graft function (PGF) are highly heterogeneous across the literature, impeding comparability and clinical applicability. Using a large multicenter dataset and retrospective analysis of over 35,000 blood samples, we demonstrated that differences in PGF definitions significantly affect reported incidence, associated risk factors, and clinical outcomes. We advocate for standardized, lineage-specific thresholds and the incorporation of longitudinal blood count trajectories into predictive models to enhance early detection and risk stratification.
Second, we examined long-term hematopoietic integrity by studying clonal hematopoiesis (CH) in pediatric HCT survivors. Using sensitive targeted sequencing, we found a surprisingly high prevalence of CH, even in young recipients transplanted with young donors. All detected clones carried mutations in DNMT3A or TET2—canonical age-related CH drivers—and originated from donor-derived hematopoietic stem cells (HSCs). CH incidence was associated with donor age and inflammatory markers around the time of transplant, suggesting that the transplantation procedure selectively favors expansion of pre-existing mutant clones. These findings raise concerns about the long-term clinical consequences of donor-derived CH in young recipients.
To enable future studies into post-transplant hematopoiesis in children, we created a single-cell multimodal atlas of healthy pediatric bone marrow, integrating mRNA and surface protein profiles from over 90,000 cells. We identified age-dependent shifts in lineage bias, most notably a transition from B-cell–biased output in early childhood to more balanced or myeloid-skewed hematopoiesis in adolescence. This shift was linked to both intrinsic changes in lymphoid progenitor composition and extrinsic alterations in stromal signaling, including age-related changes in IL-7 and TGF-β1 signaling. By demonstrating important differences between pediatric, adolescent and adult bone marrow, this work highlights the need for a single cell reference map of pediatric bone marrow – provided in this thesis – when studying pediatric hematologic disease.
Together, the results presented in this thesis contribute to a better understanding of hematopoietic regeneration in the transplanted setting. Our findings highlight the need for harmonized clinical definitions, improved integration of blood count data in post-transplant monitoring, and attention to donor- and recipient-derived factors that shape the clonal and functional landscape of the hematopoietic system. These insights lay the groundwork for more precise, data-driven strategies to predict, monitor, and optimize hematopoietic recovery following stem cell transplantation.