Andy Tully

Xenograft rejection is a complex process that advances quickly thus requiring a rapid and precise analysis of graft function. Traditional methods including biopsies and echocardiography do not have the capacity to detect subtle changes. There is therefore a desperate need to develop molecular methods with a rapid turnaround to institute treatments at the earliest incidence of impending rejection. In their recent publication, Loupy et al1 present an analysis using multimodal immunophenotyping of xenograft specimens from 2 decedent recipients of α-Gal knockout porcine renal xenotransplants performed at New York University (NYU), followed for 54 h posttransplant. Using wedge xenograft biopsy specimens pretransplant and at the study endpoint at 54 h, the investigators systematically applied an array of novel techniques. Tissue regions of interest were identified by blinded, experienced pathologists on standard hematoxylin and eosin (H&E) staining and then analyzed sequentially with immunostaining, bulk gene expression profiling, and spatial transcriptomics. Running xenograft samples through this systematic analysis in comparison with nontransplanted porcine, autotransplanted porcine, and nontransplanted ischemia–reperfusion porcine kidney controls, the authors formulated unified conclusions on the immunologic health of the grafts. Despite the favorable macroscopic appearance and graft function, the authors suggest that their techniques reveal early signs of antibody-mediated rejection (AMR). The described rejection pattern was subtle and perhaps represented an early signature of incipient AMR, demonstrated by microvascular inflammation with immunoglobulin deposition in glomerular and peritubular capillaries. Notably, C4d deposition was not observed indicating a lack of observed complement activation. Also noteworthy was the absence of microvascular thrombosis, a common feature of hyperacute rejection (HAR) and AMR in xenografts. This latter result is particularly noteworthy given that the xenografts did not express transgenes encoding for complement regulatory proteins. The investigators expand on this observation by identifying infiltrating innate immune cells via immunohistochemistry. Specifically, they identify CD69, CD15, and natural killer (NK) cells, particularly in the glomerular and peritubular capillaries of the grafts. Perhaps most intriguing, however, is the application of transcriptomics techniques to immunophenotyping. Using transcriptome profiling and sophisticated statistical analysis, the authors identified upregulated genes involved in antibody-mediated injury, endothelial cell activation (POSTN and LDB2), complement activation (C1QB), innate immune cell activity (CST3, PTPRB, and TXNIP), humoral response (HLA-B), interferon-γ response (IFI30, B2M, and C1QC), monocyte and macrophage activation (CD74), NK cell burden (FCGR3A and HLA-E), and nonimmune-related processes including metabolism. Notably, there was concordance between antibody deposition and innate immune infiltration with changes in gene expression. The discrepancy of upregulated complement genes in the absence of observed C4d staining will surely be the subject of further investigation. A significant achievement of this study is the successful prevention of HAR using a minimum of gene edits in the donor pig. As the field of xenotransplantation has advanced, xenografts have incorporated additional gene edits to prevent complement activation while mitigating hypercoagulability. Therefore, avoiding HAR with an α-Gal knockout xenograft represents a significant feat. This achievement is particularly significant with the absence of C4d deposition. How the pattern of early AMR detailed in this article would vary across xenografts with different genetics remains an open question. We hypothesize that the overexpression of human transgenes in the xenograft may protect porcine endothelial cells from complement-induced lysis and potential xenoimmune-related injuries, resulting in an altered transcriptome. There may be 1 limitation of the present study: the use of a pig donor with only a single gene edit (a substrate now in rare use) may limit the generalizability of the otherwise powerful analytical approach. Of note, in 2022 Moazami et al,2 also at NYU, transplanted 10 gene-edited porcine hearts overexpressing several human genes (TBM, EPCR, CD59, HO-1, CD46, and DAF) into 2 decedent human recipients and reported no evidence of cellular or AMR assessed by conventional histology, flow cytometry, and a cytotoxic crossmatch assay. Expanding this analysis to include the methods used in the work by Loupy et al1 would add significantly to the interpretation of their results and may yet reveal occult AMR in the cardiac grafts as well. Systematic deep phenotyping analysis, spatially specific to regions of interest identified on H&E, provides a new standard for early rejection detection. As demonstrated here, this technique is sufficiently powerful to tease out differences between underlying donor pathologies while being sufficiently detailed to demonstrate heterogeneity between recipients, which seems to be unexpectedly substantial, especially when using cell-type spatial profiling. This sort of analysis will be crucial to ascertain what parts of the biology seen in any given xenotransplant result from the donor, the transplant process, or the recipient so that treatment can be targeted appropriately. Overall, the work by Loupy et al1 demonstrates significant advances that emerge from collaborative efforts in xenotransplantation. Many excellent teams are currently working to translate xenotransplantation from nonhuman primate models into a viable solution responding to the current human organ shortage. Many innovative companies with unique gene-edited pigs and target-specific antibodies have fostered much of the current progress in this space. However, as demonstrated in the first instance of solid organ xenotransplantation into a living human, clinical translation and surveillance will be quite complex and will require the collaboration and expertise of all groups involved. A broad and detailed multicomponent analysis of the first clinical heart transplant recipient published by us in the Lancet,3 concluded that the failure of the cardiac xenograft was multifactorial in etiology. The complexities of clinical translation for xenotransplantation remain substantial but readily achievable through collaboration and powerful analytic techniques such as those presented by Loupy et al1. In conclusion, the multimodal gene profiling techniques described here ultimately need to be correlated with standard validated methods evaluating graft rejection. Currently, it appears challenging to determine the robustness of those data without this comparison. Nevertheless, specific changes in gene regulation could be used as early indicators of rejection or inflammatory responses to initiate efficient treatments. In addition, it is expected that the immune response will vary from pig to pig based on gene modifications and from one recipient to another based on preexisting comorbidities. With transcriptome profiling data in xenotransplantation so far limited to short-term experiments in decedent models in which factors in addition to rejection may have contributed, further research analyzing samples from live human xenotransplant cases are expected to enhance our understanding of this promising technology.