Use of the NIH consensus criteria in cellular and soluble biomarker research in chronic graft-versus-host disease: A systematic review

Objectives: Chronic graft-versus-host disease (cGvHD) is the most frequent cause of late non-relapse mortality after allogeneic haematopoietic stem cell transplantation (alloHCT). Nevertheless, established biomarkers of cGvHD are still missing. The National Institutes of Health (NIH) Consensus Development Project on Criteria for Clinical Trials in cGvHD provided recommendations for biomarker research. We evaluated to which extent studies on cellular and soluble biomarkers in cGvHD published in the last 10 years complied with these recommendations. Also, we highlight the most promising biomarker candidates, verified in independent cohorts and/or repeatedly identified by separate studies. Methods: We searched Medline and EMBASE for "cGvHD", "biomarkers", "soluble" and "cells" as MeSH terms or emtree subject headings, and their variations on July 28th, 2021, limited to human subjects, English language and last ten years. Reviews, case reports, conference abstracts and single nucleotide polymorphism studies were excluded. Criteria based on the set of recommendations from the NIH group for biomarker research in cGvHD were used for scoring and ranking the references. Results: A total of 91 references encompassing 15,089 participants were included, 54 prospective, 17 retrospective, 18 cross-sectional, and 2 studies included both prospective and retrospective cohorts. Thirty-five papers included time-matched controls without cGvHD and 20 studies did not have any control subjects. Only 9 studies were randomized, and 8 were multicentric. Test and verification cohorts were included in 11 studies. Predominantly, diagnostic biomarkers were explored (n=54). Assigned scores ranged from 5-34. None of the studies fulfilled all 24 criteria (48 points). Nevertheless, the scores improved during the last years. Three cell subsets (CXCR3(+)CD56(br)(ig)(ht) NK cells, CD19(+)CD21(low) and BAFF/CD19(+) B cells) and several soluble factors (BAFF, IL-15, CD163, DKK3, CXCL10 and the panel of ST2, CXCL9, MMP3 and OPN) had the highest potential as diagnostic and/or prognostic biomarkers in cGvHD. Conclusion: Despite several limitations of this review (limited applicability for paediatric population, definition of verification, missing data on comorbidities), we identified promising candidate biomarkers for further evaluation in multicentre collaborative studies. This review confirms the importance of the NIH consensus group criteria for improving the quality and reproducibility of cGvHD biomarker research.