Outcomes of ISHLT Lung Transplant AMR

Purpose ISHLT published diagnostic criteria for AMR in 2016 delineating the diagnostic certainty of AMR with definite, probable, and possible grading that considers PFTs, presence of DSA, histopathology, C4d, and ruling out other causes for graft dysfunction. The natural course of these AMR categories remains unknown. This study examines outcomes related to ISHLT AMR categories. Methods The study included 435 lung transplant recipients enrolled in two multicenter prospective cohort studies. An AMR Working Group reviewed clinical data and used ISHLT criteria to adjudicate for AMR and CLAD. DSA+ subjects were categorized as DSA+/no AMR and DSA+/AMR. The AMR group was further stratified into definite, probable, and possible. CLAD-free survival was defined as being alive and CLAD-free. Allograft injury was assessed as decline in FEV1 and via a validated biomarker (donor-derived cell-free DNA - %ddcfDNA). %ddcfDNA assayed by shotgun sequencing in serially collected post-transplant plasma samples. Allograft injury, DSA features and CLAD-free survival were compared between DSA+/AMR and DSA+/no AMR, as well as between AMR groups. Definite and probable AMR were combined into one group due to smaller numbers for each definite AMR. Results There were 185 DSA+ patients; 92 of which developed AMR. DSA+/AMR patients showed distinct features from DSA+/No AMR group. DSA+/AMR showed elevated %ddcfDNA levels (1.87 vs. 0.87), 3-fold higher hazard of NOT clearing DSA (95%CI 1.7-6.5, Figure A), and 4-fold higher hazard of dying or developing CLAD (95%CI 2.5-7.1, Figure B). On the other hand, definite/probable and possible AMR showed similar DSA mean florescence intensity (7241 vs 6023), %ddcfDNA (3.81 vs 3.11), %PFT decline (36% vs 28%) at the time of AMR diagnosis, and CLAD-free survival compared to possible AMR group. Conclusion DSA+/AMR subjects show increased risk of CLAD than DSA+ patients without AMR. However, all ISHLT AMR categories showed similar clinical and prognostic profiles, suggesting that the ISHLT diagnostic categories may not represent disease severity.