Early D-Dimer Value Is Highly Predictive of Non-Relapse Mortality in Children after Haploidentical Hematopoietic Stem Cell Transplantation

Abstract

Background Alloreactivity has been implicated with endothelial damage, severe complications, and death after hematopoietic stem cell transplantation (HSCT). Allogenic reaction in HLA-haploidentical HSCT (HaploHSCT) is established early after graft infusion, and therefore, the severity of this reaction could be measured in the first days of graft infusion. Plasmatic D-Dimer (DD) has been used as a surrogate indicator of endothelial dysfunction and inflammation. We explored whether days +2 or +3 DD were predictive of one-year non-relapse mortality (1y-NRM) in post-transplantation cyclophosphamide-treated children after HaploHSCT. Methods In 2013, we started to measure DD in the first 7 days after graft infusion. We focus only on patients with peak DD taken on days +2 or +3. Data source was our pediatric HSCT registry, which is prospectively filled using information from the clinical charts, imaging, laboratory, and pathology reports. We excluded Fanconi anemia patients. The period of analysis was from Jan/2013 to June/2018. We used plasma DD measured by ELISA (ng/mL-FEU) and as reported by the clinical laboratory. DD was log-transformed for achieving normality. We divided DD in tertiles and we took the value between the 2nd and 3rd tertile as the final cutoff point. We used the cumulative hazards method to estimate 1y-NRM. Cox's multivariate regression models were used to adjust hazard ratios (HR) by covariates derived from the recipient, the graft, and the donor. Non-parametric ROC curve was constructed to estimate the AUC. Results We included thirty-two patients (0-18 years of age). Table 1 presents recipient, graft, and donor characteristics by DD. Nine patients died and had a DD geometric mean of 3048 ng/mL FEU, which is higher when compared to living patients (n=23; 1229 ng/mL FEU; P Conclusion We found 12-times more risk of 1y-NRM in children with DD ≥1700 ng/mL-FEU in day+2 or +3 after HaploHSCT. This relationship was independent of other covariates including aGvHD. This finding could be used to separate patients with high risk of NRM in an early HSCT phase, and therefore to introduce preemptive strategies to mitigate this risk. More research is needed to understand the mechanisms underlying this finding.