The Prognostic Value of YKL-40 in Allogeneic Hematopoietic Cell Transplantation

Although allogeneic hematopoietic cell transplantation (HCT) is a treatment option for a variety of malignant hematologic diseases, complications such as graft versus host disease (GVHD), infections and relapse are still major causes of morbidity and mortality. Prognostic scores such as the disease risk index and the HCT comorbidity index have proven useful in predicting relapse and treatment related mortality (TRM). However, to appropriately balance the likelihood of disease control against the risk of debilitating complications, still more accurate pretransplant markers are necessary to predict outcome.

YKL-40 (chitinase-3-like-1 (CHI3L1)) is mainly secreted by cancer cells, macrophage and neutrophils. YKL-40 regulates vascular endothelial growth factor and has a role in angiogenesis and inflammation, cell proliferation and differentiation, and remodeling of the extracellular matrix. High plasma YKL-40 levels have been associated with poor prognosis in patients with different types of cancers and inflammatory diseases

A previous study in HCT found that YKL-40 >95% age-adjusted level in recipients was associated with relapse and lower disease-free and overall survival. High YKL-40 levels in donors were associated with more grade II-IV acute GVHD.

We investigated the prognostic value of pre-transplant plasma YKL-40 in a validation cohort of 784 recipients and donors undergoing allogeneic HCT for acute myeloid leukemia (AML) (n=626) or myelodysplastic syndrome (n=158) (MDS) with 8/8 matched unrelated donors after myeloablative (n=566) or reduced intensity (n=218) conditioning. Transplantations were facilitated through the National Marrow Donor Program (NMDP) and data collection and analysis were performed under the auspices of the Center for International Blood and Marrow Transplant Research (CIBMTR). Samples were provided by the NMDP/CIBMTR and obtained pre-transplant from recipients and prior to mobilization from donors. P<0.05 was considered significant for validation of previous observations.

Mean recipient pre-transplant YKL-40 plasma concentrations were elevated in patients with HCT-comorbidity index (CI) scores ≥5 (HCT-CI 0: n=266, 106 (range 20-1446) ng/ml; HCT-CI 1-2: n= 244, 105 (range 20-1176) ng/ml; HCT-CI 3-4: n= 206, 106 (range 20-636) ng/ml; HCT-CI ≥5: n=68, 171 (range 20-4644) ng/ml; p<0.04). There was no difference (p=0.91) in YKL-40 concentrations between recipients with AML (111 (range 20-4644) ng/ml) and MDS (113 (range 20-714) ng/ml). There was no association between recipient YKL-40 concentrations and cytogenetics or AML disease status (first complete remission versus beyond first remission). In addition, MDS YKL-40 concentrations did not significantly increase with successively worse cytogenetics (p=0.67), IPSS (p=0.72) or Karnofsky score (p=0.86).

Recipient YKL-40 levels above the age-adjusted 95% percentile were not associated with any outcomes (p>0.198). Donor YKL-40 levels >95% was associated with acute GVHD II-IV (HR 1.39, 95% confidence interval 1.00-1.92, p=0.0466) as observed in the previous study. There was no association between donor YKL-40 and other outcomes. When age-adjusted YKL-40 percentile was used as a continuous variable, recipient and donor pre-transplant elevated YKL-40 were not associated with any outcomes (p>0.088). Analyses were also conducted to investigate the impact of extremely low and high YKL-40 level expression and no associations were observed.

In contrast to the prior study no association between recipient pre-transplant YKL-40 concentrations and clinical outcome were observed. However, in agreement with the previous study, donor YKL-40 concentrations were associated with increased acute GVHD, suggesting that donors with a higher degree of inflammation prior to mobilization could increase risk of post transplant complications. However, this did not translate into increased risk of TRM. Although it is not known whether the associations between YKL-40 and outcome are causal or just bystander effects, our observations suggest that an inflammatory biomarker, such as YKL-40, that potentially defines a higher-risk donor population could be a valuable tool complementing clinical risk scores in HCT. In conclusion, age adjusted YKL-40 levels in the donor, but not recipient, were a prognostic indicator for acute GVHD risk in this population.