Comment
Kidney xenotransplantation—bridge to allotransplantation or destination treatment?
Alexandre Loupy, Aurélie Sannier, Gabriel C Oniscu
Lancet 2026; 408: 976-978
https://doi.org/10.1016/S0140-6736(26)01479-0
The shortage of human kidneys remains a central problem in transplantation, and many candidates die or are removed from the waiting list before an organ becomes available.1Genetically engineered porcine kidneys have long been proposed to expand the donor pool,2 and over the past 3 years the field has rapidly moved from preclinical models to decedent models to living recipients.3, 4, 5 In their Article, Leonardo Riella and colleagues report a porcine kidney xenograft that supported a living human recipient for 271 days—the longest dialysis-free xenograft function described to date—followed by explantation and a subsequent deceased-donor human kidney transplantation.6 The authors position this sequence as a bridge to human transplantation, opening debate about where xenotransplantation fits in the treatment armamentarium. This milestone allowed the field to test long-standing questions on xenotransplantation in a living recipient.
The Article describes kidney xenotransplantation in a non-Hispanic White man aged 66 years, blood group O, with end-stage kidney disease from type 2 diabetes, who has been on haemodialysis for 2 years, and had a predicted 9% chance of transplantation in 5 years. The donor kidney (EGEN-2784) came from a Yucatan miniature pig with three xenoantigen knockouts, seven human transgenes, and CRISPR-inactivated porcine endogenous retroviruses. Induction immunosuppression combined rabbit antithymocyte globulin, corticosteroids, rituximab, and perioperative complement inhibition with pegcetacoplan; maintenance relied on CD40–CD154 costimulation blockade with tegoprubart, together with tacrolimus, mycophenolic acid, and prednisone, adjusted according to infection risk and immunological monitoring, with complement inhibition intermittently withdrawn.
The xenograft functioned immediately but the post-transplantation course included an early Banff IIA T-cell-mediated rejection at day 14, which resolved with corticosteroids and antithymocyte globulin, and a later phase of microvascular injury that progressed to thrombotic microangiopathy and prompted explantation at day 271. Donor-specific flow crossmatch remained negative throughout and anti-HLA antibodies were unchanged. 82 days after explantation, despite the earlier poor prediction of allograft access, the recipient received a deceased-donor kidney that was, fortuitously, a zero-HLA mismatch. The allograft functioned immediately, with creatinine near 1·0 mg/dL at 126 days of follow-up.
The most reassuring result in this Article is the absence of zoonosis. Intensive longitudinal surveillance, including metagenomic plasma sequencing, detected no porcine-derived pathogen (including porcine cytomegalovirus) at any point during or after the xenograft period.7 Cross-species infection has been a principal barrier to clinical xenotransplantation; these data support the view that the risk can be contained through pathogen-free donors, retrovirus inactivation, and surveillance.
Beyond these results, the case defines where science needs to advance now. The molecular and multiplex analyses, although extensive, could not be applied uniformly across the longitudinal course because of limited tissue availability. A fully reconstructed trajectory of graft injury from the day-14 T-cell-mediated rejection to the microvascular lesions was therefore not possible. Several findings, including early C4d deposition, mesangial IgM and IgA deposition associated with mesangial expansion, and microvascular inflammation with macrophage and natural killer cell predominance, could represent successive manifestations of a common injury process rather than unrelated processes. Whether this common injury process is driven by antibodies that remain undetectable with current assays, or by other mechanisms of endothelial activation, remains unresolved. Addressing this question will require longitudinal, standardised multimodal precision-pathology profiling.8
The Article also highlights where diagnostic tools need to mature. Assays for tracking the humoral response need strengthening: dedicated tests for antibodies against porcine antigens were not reported, and a donor-specific crossmatch alone cannot exclude humoral mechanisms while validated xeno-specific assays are still being established. A negative crossmatch supports—but does not yet prove—an antibody-independent process. The inability to exclude humoral mechanisms despite a negative crossmatch underscores the need for tissue-based deep phenotyping, since circulating antibody assays cannot capture antibodies already bound within the xenograft. The 126-day allograft follow-up, although reassuring, is too short to exclude late or memory humoral responses. Whether a previous xenotransplant affects allograft outcome remains an essential question if xenotransplantation is to serve as a bridge therapy.
As xenograft longevity is still limited to months, focus should shift from feasibility to how risks and benefits are balanced and how patients are selected in the context of available treatment options. With a maximum dialysis-free survival of now 271 days, the question is which patients should be offered a xenograft (as a bridge or destination therapy) rather than dialysis, who should be placed on the waiting list, and who should be considered for acceptable-mismatch allocation. Unlike heart or liver failure, dialysis is an effective alternative in end-stage renal disease, raising the threshold for justifying xenotransplantation. Perhaps the strongest case is for highly sensitised candidates with a calculated panel-reactive antibody approaching 99·9% and documented broad HLA sensitisation, rather than those who retain a realistic chance of a conventional offer. Explicit thresholds, full sensitisation characterisation, transparency about treatment risks and benefits, and evaluation within prospective clinical trials (rather than expanded-access pathways) should be the priority.9, 10
These choices are ultimately about patients, not only grafts.11 Exposing a recipient to months of intensive immunosuppression, repeated biopsies, and a sensitisation risk that could compromise a future allograft is defensible only when the alternative is genuinely worse, and that judgement must be made transparently and with the patient's informed participation.12 Equitable allocation of an allograft after a xenotransplant bridge must be protected so that access does not distort an already constrained system, and the economic burden—the engineered animal, immunosuppression, and monitoring, possibly followed by a conventional transplant—needs honest appraisal against dialysis.
Riella and colleagues have advanced the field with a carefully conducted case. Realising its promise will depend on the standardised deep phenotyping, humoral-response tools, and ethically grounded patient selection we have outlined.8
