Prospective assessment of risk biomarkers of sinusoidal obstruction syndrome after hematopoietic cell transplantation

BACKGROUND Currently, no laboratory tests exist to stratify for the risk of developing sinusoidal obstruction syndrome (SOS), an early endothelial complication after hematopoietic cell transplantation (HCT). Risk biomarkers of SOS have not been verified in a prospective cohort accounting for differences between practices across institutions. Herein, we aimed to define risk groups for SOS occurrence using 3 proteins: L-ficolin, hyaluronic acid (HA), and stimulation 2 (ST2). METHODS Between 2017 and 2021, we prospectively accrued 80 pediatric patients across 4 US centers. Biomarkers were tested by ELISA blind to patient groupings and associated with SOS incidence on day 35 after HCT, and overall survival (OS) on day 100 after HCT. Cutpoints were identified using retrospective cohorts and applied to the prospective cohort. RESULTS Combination of the 3 biomarkers measured on day 3 after HCT in the prospective cohort provided 80% (95% CI 55%–100%) sensitivity and 73% (95% CI 62%–83%) specificity for risk of SOS occurrence. Patients with low L-ficolin were 9 times (95% CI 3–32) more likely to develop SOS, while patients with high HA and ST2 were 6.5 (95% CI 1.9–22.0) and 5.5 (95% CI 2.3–13.1) times more likely to develop SOS. These 3 markers also predicted worse day 100 OS — L-ficolin: HR, 10.0 (95% CI 2.2–45.1), P = 0.0002; HA: HR, 4.1 (95% CI 1.0–16.4), P = 0.031; and ST2: HR, 3.9 (95% CI 0.9–16.4), P = 0.04. CONCLUSION L-ficolin, HA, and ST2 levels measured as early as 3 days after HCT improved risk stratification for SOS occurrence and OS and may guide risk-adapted preemptive therapy. TRIAL REGISTRATION ClinicalTrials.gov NCT03132337. FUNDING NIH.

. Sensitivity, specificity, PPV, and NPV using optimal cutpoints in retrospective cohort Table S6. Sensitivity, specificity, PPV, and NPV using optimal cutpoints in prospective cohort Table S7. Correlation between biomarker levels and maximum total serum bilirubin Table S8. Correlation between biomarker levels and total bilirubin at day 3 Table S9. Correlation between biomarker levels and alkaline phosphatase at day 3 Table S10. Correlation between biomarker levels and AST at day 3 Table S11. Correlation between biomarker levels and ALT at day 3 Table S12. Demographics for other post-HCT complications Table S13. Correlation between biomarker levels and TNFR1 at day 3 Table S14. Association between biomarker levels and thrombotic microangiopathy (TMA) Table S15. Association between biomarker levels and idiopathic pneumonia syndrome (IPS) Table S16. Association between biomarker levels and overall graft-versus-host disease (GVHD) severity Table S17. Association between IL-6 and TNFR1 levels and overall graft-versus-host disease (GVHD) severity Table S18. Association between biomarker levels and infections Table S19. Descriptive statistics of biomarkers in subcategorized age in patients without SOS.

Study Patients.
Detailed eligibility criteria for the study are described below

Inclusion Criteria
Age ≤ 25 years undergoing HCT for any reason who fulfill any ONE of the following criteria: 1. History of hepatic disease as defined by: a. Viral hepatitis (i.e., hepatitis C virus [HCV]) b. Liver tumor before HCT c. Hepatic fibrosis or cirrhosis before HCT as proven by liver biopsy d. High aspartate aminotransferase (AST) (> 2x ULN) before HCT (pretransplant evaluation) e. High alanine transaminase (ALT) (> 2x ULN) before HCT f. High bilirubin (> 1.2x ULN) before HCT 2. HCT high-risk features including: a. Conditioning with high-risk modalities including: i. Busulfan (BU)-containing regimen particularly with oral BU + cyclophosphamide ii. TBI-containing regimen, particularly cyclophosphamide + total-body irradiation (TBI) b. ≥ 2 HCT c. Allo-HCT for leukemia > or = second relapse d. Unrelated donor (URD) HCT e. Human leukocyte antigen (HLA) mismatch HCT (less than 10 of 10 for bone marrow/peripheral blood stem cell [BM/PBSC] or anything less than 6 of 6 for UCB) f. Use of sirolimus + tacrolimus prophylaxis for GVHD 3. High-risk disease states including: a. Juvenile myelo-monocytic chronic leukemia (JMML) b. Primary hemophagocytic lymphohistiocytosis (HLH) c. Adrenoleukodystrophy d. Osteopetrosis 4. Other high-risk features including: a. Prior treatment with gemtuzumab ozogamicin b. Use of hepatotoxic drugs 1 month before HCT and during HCT c. Iron overload (i.e., thalassemia/sickle cell) with serum ferritin > 1000ng/ml d. Deficit of ATIII, T-PA (i.e., < 30% normal values), and resistance to activated protein C if clinical indication (these values do not have to be specifically checked if no clinical history) e. Young age < 2 years but more than 1 month

Exclusion Criteria
Patients who are transplanted but do not fulfill any of the above-mentioned criteria.

ELISA.
Antibody pairs were purchased as described in Supplemental Table 2. Proteins were measured in samples using commercially available enzyme-linked immunosorbent assays (ELISA) and following the manufacturer' recommendations and using a sequential ELISA approach previous described. All samples and standards were tested in duplicate. All washes were performed using the Aquamax 2000 plate washer (Molecular Devices, Sunnyvale, CA). Absorbance was measured immediately after termination of the substrate reaction using a SpectraMax ABS Plus plate reader and results were calculated using SoftMax Pro Version 7.1 (Molecular Devices, Sunnyvale, CA).