Differentiation between lung allograft rejection and infection using donor-derived cell-free DNA and pathogen detection by metagenomic next-generation sequencing

BACKGROUND: In lung transplant recipients (LTRs), the primary causes of mortality are rejection and infection, which often present similar symptoms, making differentiation challenging. This study aimed to explore the diagnostic efficacy of plasma donor-derived cell-free DNA (dd-cfDNA) in conjunction with metagenomic next-generation sequencing (mNGS) for pathogen detection in differentiation between lung allograft rejection and infection in LTRs experiencing new-onset pulmonary complications. METHODS: We conducted a retrospective study on 188 LTRs who underwent lung or heart-lung transplantation at our institution from 2015 to 2021. The LTRs were categorized into three groups: stable, rejection, and infection. We measured plasma dd-cfDNA levels and utilized both mNGS and culture methods to identify pathogens in the bronchoalveolar lavage fluid (BALF). RESULTS: The rejection group exhibited the highest levels of plasma dd-cfDNA (median 1.34 %, interquartile range [IQR] 1.06–2.19 %) compared to the infection group (median 0.72 %, IQR 0.62–1.07 %) and the stable group (median 0.69 %, IQR 0.58–0.78 %) (both p < 0.001). Within the infection group, a significantly higher level of dd-cfDNA was observed in the cytomegalovirus infection subgroup (p < 0.001), but not in the fungal (p > 0.05) or bacterial infection subgroups (p > 0.05), when compared to the stable group. Elevated dd-cfDNA levels, in combination with negative mNGS results, strongly indicated lung allograft rejection, with a positive predictive value and negative predictive value of 88.7 % and 99.2 %, respectively. CONCLUSIONS: Plasma dd-cfDNA in combination with BALF pathogen detection by mNGS shows satisfactory accuracy in differentiating lung allograft rejection from infectious complications.