INTRODUCTION: Severe transplant-associated thrombotic microangiopathy (TMA) is a serious complication of allogeneic hematopoietic cell transplantation (HCT), which is frequently linked with acute GVHD. The association between serum biomarkers predictive of acute GVHD and TMA remains understudied in adult HCT patients.

METHODS: Established and novel biomarkers of both acute GVHD and TMA were selected including serum concentrations of ST2, ANG-2 (angiopoietin-2), EGF (epidermal growth factor), VEGFR2, sFLT-1 (soluble fms-like tyrosine kinase-1), PlGF (placental growth factor), cystatin-C, VDBP (vitamin D binding protein), sC5b-9, NETs (neutrophil extracellular traps), and F-actin measured pre-HCT in a centralized laboratory using enzyme-linked immunosorbent assay. ST2 and sC5b-9 were analyzed as log-transformed variables. TMA was adjudicated by a blinded three-member panel. Time to event multivariable analyses were performed for severe TMA, grade III-IV acute GVHD, NRM and OS using a predictive approach including pre-HCT variables significant (P<0.01) in the univariable analysis with backward elimination.

RESULTS: 368 HCT patients had prospectively collected biospecimens for serum biomarker evaluation from 3 participating centers of the MIDAS (Microangiopathy, endothelial Damage in Adults undergoing Stem cell transplantation) consortium. At time of HCT, median age was 61 years (range 19-78), ~60% were male, 88% were non-Hispanic White, 42% had HCT comorbidity index >/=3, myeloid malignancies were the most common (82%) indication, ~60% were in CR, 72% received RIC regimens, and peripheral blood was the predominant (96%) graft source. Most frequent donor type was 10/10 HLA matched unrelated donor (MUD, 56.5%), with <10/10 HLA-mismatched unrelated (MMUD, 16.6%), haploidentical (14.4%) and HLA-matched related donor (12.5%) less frequently used. GVHD prophylaxis was tacrolimus (Tac) and methotrexate (MTX) based regimen in 44% and post-HCT cyclophosphamide (PTCy) based regimen in 54% (35% with Tac and 19% with sirolimus). Median follow up of survivors was 9.6 months (range, 1.0-12.9).

The incidence of severe TMA by day + 100 was 18.5%, about half of which occurred before onset of acute GVHD and half occurred after acute GVHD onset. Acute GVHD grade III-IV was less frequent in PTCy-based prophylaxis compared to MTX-based prophylaxis (11.5% vs 20.5%), however severe TMA occurred in a similar proportion of patients with PTCy-based prophylaxis (22% vs. 19%). In addition, severe TMA occurred in 12% of patients with a maximum grade of 0-I and 18% of patients with grade II acute GVHD demonstrating that severe TMA occurs in ~10-20% of patients without severe grade III-IV acute GVHD and in those receiving PTCy-based prophylaxis.

In multivariable analysis of severe TMA, we discovered effect modification between pre-HCT creatinine and sex, therefore stratified models are presented. In males, severe TMA risk was increased 6-fold with each unit (1 mg/dL) increase of pre-HCT creatinine (HR=6.34, 95% CI 2.55-15.72; p<0.001). In females, severe TMA risk was increased 36-fold with each one-unit increase in pre-HCT creatinine (HR=36.34, 95% CI 13.91-94.91; p<0.0001) and was increased with detectable PlGF (HR=2.84, 95% CI 1.39-5.81, P=0.004).

In multivariable analysis, grade III-IV acute GVHD was significantly reduced in PTCy-sirolimus compared to MTX-based regimens (HR=0.23, 95% CI 0.07-0.73; p=0.01), significantly higher with increasing pre-HCT serum creatinine (HR=2.66, 95% CI 1.22-5.80; p=0.01) and soluble C5b-9 (HR=1.72, 95% CI 1.09-2.71; p=0.02). NRM increased with elevated pre-HCT serum creatinine (HR=4.17, 95% CI 2.13-8.16; p<0.001) and detectable PlGF (HR=1.94, 95% CI 1.05-3.61; p=0.035). In stratified models, overall mortality was significantly higher with increasing pre-HCT serum creatinine in males (HR=4.41, 95% CI 1.43-13.61; p=0.01), and females (HR=8.3, 95% CI 1.68-40.97; p=0.009).

CONCLUSIONS: In this prospective study, we show that distinct biomarkers measured pre-HCT are associated with severe TMA, acute GVHD, NRM and OS. Pre-HCT serum creatinine was the strongest factor associated with all 4 outcomes. Severe TMA incidence was similar between PTCy and MTX prophylaxis, despite a significant reduction in acute GVHD with PTCy. Additional analyses of biomarkers measured post-HCT are ongoing and our results are worthy of further validation.

This content is only available as a PDF.
2025
Sign in via your Institution