Lung Transplantation: DSA to AMR Trajectory

Abstract

Donor specific antibodies(DSA) are a major risk factor for antibody mediated rejection (AMR) and allograft failure in lung transplant recipients (LTs). The trajectory from DSA to AMR and its association with allograft injury remains poorly defined. Thus, we detailed DSA and their trajectory to AMR. 417 LTs from three prospective cohort studies were included. LTs had pre-transplant DSA testing, post transplantation surveillance and clinically indicated DSA testing. An AMR Working Group reviewed clinical data and used ISHLT criteria to adjudicate AMR. Allograft injury was assessed by %ddcfDNA in serially collected plasma samples via shotgun sequencing. Positive DSA: preformed or de novo if DSA was first detected before or after 14 days post-transplant. The interval from DSA detection to AMR or no AMR was computed. Preformed and de novo DSA was subcategorized in +AMR and -AMR subgroups based on whether or not DSA was associated with clinical AMR. %ddcfDNA levels were compared for these subgroups. Treated LTs prior to allograft dysfunction or AMR were removed. 185 subjects had positive HLA DSA (54.73%) detected at a median of 33 days post-transplantation. (IQR 9.0-236.8) DSA was predominantly de novo (n=119, 64.3%) than preformed (n=66, 35.7%), class II than Class I (76.44% vs. 20.24%, p<0.01). Of 185 DSA+ patients, 92 developed clinical AMR, 27 no AMR, and 66 had insufficient data to adjudicate clinical AMR. AMR developed at a medium of 130 days from DSA detection. (IQR 58- 504) The prevalence of clinical AMR was 23.65%(N=92): definite (N=6, 6.5%), probable (N=16, 17.4%) and possible (N=70, 76%).De novo DSA showed a higher rate of progression to AMR than preformed DSA (62.5% vs. 32.5%, p<0.01). %ddcfDNA levels correlated with AMR risk: median %ddcfDNA for de novo DSA that lead to AMR was 2X higher than de novo DSA that did not lead to AMR (1.55 vs. 0.88 p=0.016). Preformed DSA that lead to AMR had a median %ddcfDNA of 1.47 versus 0.87 in the preformed DSA that did not lead to AMR (p=0.014). DSA that did not lead to AMR showed higher %ddcfDNA levels than DSA negative, non-rejection timepoints. (0.87 vs. 0.33, p<0.01) CONCLUSION: De novo and preformed DSA showed greater allograft injury than non-rejection DSA- timepoints. De novo DSA progresses to AMR at a higher rate than preformed DSA. Preformed or de novo DSA that progresses to AMR causes more lung injury than DSA with no AMR. %ddcfDNA could be used to identify and treat high-risk DSA.