Background Poor Graft Function (PGF) characterized by multilineage cytopenias in the setting of complete donor chimerism is a recognized complication of allogeneic stem cell transplantation (alloSCT) resulting in bone marrow failure. Our prior study involving a historic cohort of 819 patients analyzed from 2000-2016 identified non-matched sibling donor, underlying MPN, ICU admission/positive blood cultures in 1st 30 days, acute GVHD, CMV and non CMV viremia as associated with the development of PGF (1) suggesting an immune pathophysiology to PGF. Here we report the results of three inter-related studies aimed at confirming clinical risk factors associated with PGF and the underlying immunobiology of PGF.

Methods To evaluate and validate the clinical risk factors of PGF, a retrospective analysis of alloSCT performed in our service between 2016-2020 were used as a validation cohort. To evaluate the immunobiology of PGF, digital spatial profiling (DSP), utilizing a NanoString™ immuno-oncology panel was applied to historic bone marrow core/trephine biopsies of patients with either PGF, good graft function (GGF) or healthy donors (HD). Patients with GGF were defined as having normal blood counts and complete donor chimerism at time of sample. Peripheral blood flow cytometry was performed on stored peripheral blood mononuclear cells (PMBCs).

ResultsRetrospective Analysis 308 patients had an alloSCT between 2016 and 2020 of which 49 patients fulfilled criteria for PGF. The median age of the cohort was 54 (16-74), demographics of the PGF versus non-PGF cohort are presented in Table 1. The cumulative incidence of PGF in the 2016-2020 cohort was 16% compared to 5% in the 2000-2016 cohort (p=0.001). On multivariate analysis early ICU admission (p=0.002), non-CMV viral infection (p=0.032) and GVHD (p<0.001) remained associated with the development of PGF in the 2016-2020 cohort. PGF patients had a worse 2-year survival than non- PGF (75% versus 60%), this was mainly driven by PGF patients without marrow recovery at last follow up, who had median survival of 8 months. The association of PGF with alloreactivity, infection and inflammation in this independent cohort supports an immunologic basis to PGF.

DSP and Flow cytometry. The bone marrow trephines of patients with PGF (n=20), GGF (n=20) or 20 Normal Controls (n=20) were evaluated by DSP. PBMC flow cytometry was also performed on patients with PGF (n=18), GGF (n=13) and HD (n=11)

Overall, results supported immune dysregulation as being associated with PGF. An increase in monocyte/macrophage activity demonstrated by CD14/CD163 and increased markers of immune activation such as CD44, TIM3 and HLA-DR combined with reduction in immunosuppressive molecules such as ARG1 and immunomodulatory VISTA in PGF by was detected by DSP. PB flow analysis showed similarities between PGF and GGF including increased PDL1 expression in monocytes and dendritic cells, suggesting that immune activation and exhaustion occurs in all patients following alloSCT. In GGF samples increased CD11c (tolerogenic dendritic cells) and reduced CD44 expression (reduced T-cell homing) were demonstrated. Together, these findings suggest that T cell activation following infection or GVHD in the setting of failed tolerogenic regulation leads to PGF

Conclusion The results of these studies confirm that dysregulated BM immunity is the basis of PGF, the incidence of which is increasing.

References

  1. Prabahran A, Koldej R, Chee L, Wong E, Ritchie D. Evaluation of risk factors for and subsequent mortality from poor graft function (PGF) post allogeneic stem cell transplantation. Leuk Lymphoma. 2021;62(6):1482-9.

Koldej:CRISPR Therapeutics: Research Funding. Chee:Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; Otsuka: Membership on an entity's Board of Directors or advisory committees. Ritchie:MSD: Honoraria; Amgen: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Takeda: Honoraria, Research Funding; BMS: Honoraria, Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution