Extended survival of 9- and 10-gene-edited pig heart xenografts with ischemia minimization and CD154 costimulation blockade-based immunosuppression

Abstract

Background: Xenotransplantation has made significant advances recently using pigs genetically engineered to remove carbohydrate antigens, either alone or with addition of various human complement, coagulation, and anti-inflammatory ''transgenes''. Here we evaluated results associated with gene-edited (GE) pig hearts transplanted in baboons using an established costimulation-based immunosuppressive regimen and a cold-perfused graft preservation technique.

Methods: Eight baboons received heterotopic abdominal heart transplants from 3-GE (GalKO.β4GalNT2KO.hCD55, n = 3), 9-GE (GalKO.β4GalNT2KO.GHRKO.hCD46.hCD55. TBM.EPCR.hCD47. HO-1, n = 3) or 10-G (9-GE+CMAHKO, n = 2) pigs using Steen's cold continuous perfusion for ischemia minimization. Immunosuppression (IS) included induction with anti-thymocyte globulin and αCD20, ongoing αCD154, MMF, and tapered corticosteroid.

Results: All three 3-GE grafts functioned well initially, but failed within 5 days. One 9-GE graft was lost intraoperatively due to a technical issue and another was lost at POD 13 due to antibody mediated rejection (AMR) in a baboon with a strongly positive pre-operative cross-match. One 10-GE heart failed at POD113 with combined cellular and antibody mediated rejection. One 9-GE and one 10-GE hearts had preserved graft function with normal myocardium on protocol biopsies, but exhibited slowly progressive graft hypertrophy until elective necropsy at POD393 and 243 respectively. Elevated levels of IL-6, MCP-1, C-reactive protein, and human thrombomodulin were variably associated with conditioning, the transplant procedure, and clinically significant postoperative events.

Conclusion: Relative to reference genetics without thrombo-regulatory and anti-inflammatory gene expression, 9- or 10-GE pig hearts exhibit promising performance in the context of a clinically applicable regimen including ischemia minimization and αCD154-based IS, justifying further evaluation in an orthotopic model.

Keywords: costimulation pathway blockade; genetic engineering; heart transplantation; swine; xenotransplantation.

Fig 1.. Illustration of ischemia minimization, heterotopic cardiac transplantation model and immunosuppressive treatment regimen

(A) A pig heart is suspended over a reservoir and. cold sanguineous Steen perfusate is infused in the brachio cephalic artery and after cross-clamping the aortic arch. An incision in the left ventricular apex assures drainage to avoid distention. (B) The same heart after being abdominally implanted and shortly after reperfusion, with the apical LV venting site secured by a Rummel snare and pledgets. (C) Cartoon illustration of the intra-abdominal heterotopic heart transplant model. (D) The immunosuppressive treatment regimen: Antithymocyte globulin (ATG) and αCD20 were given over the week before the surgery. TNF inhibitor, Interleukin 6 receptor inhibitor, thromboxane synthase inhibitor, antihistamines and corticosteroid were given prior to reperfusion of the cardiac xenograft. Maintenance therapy included co-stimulation blockade with αCD154 (TNX-1500 or 5C8H1), mycophenolate mofetil, tapered corticosteroid and low dose acetylsalicylic acid. Intravenous heparin was given in the first 10–14 days and was followed in some cases by low molecular weight heparin.

Fig 2.. Xenograft survival, function, and hypertrophy +animal weight + aCD154

(A) The contractility was subjectively classified as good [3], moderate [2], weak [1], or absent [0], using ultrasound and palpation. The red cross on the graph indicates graft loss. (B) Interventricular septum (IVS) thickness in ultrasound. Initial growth in diameter was seen in all animals in the first 4 months. A progressive hypertrophy in B7221-10GE anticipated graft rejection, while both B2622-10GE & B2721-9GE reached a stable level after 16–20 weeks.(C) Tonix-1500, the primary αCD154 reagent in this study, reached the targeted level of >700 μg/ml by the third week, and was generally maintained between 1000 to 1500 μg/ml thereafter during ongoing treatment (solid line). B2721-9GE and B22622-10GE were converted to primatized 5C8H1 (red arrows indicate the timing), due to an interruption in TNX-1500 supply. 5C8H1 level of around 1000 μg/ml was maintained up to week 28 (the (dashed lines). Reducing the dose to 10mg/kg/week on POD180, was associated with drop in level to around 500 μg/ml (grey background). (D) Initial drop in weight and weak appetite were seen in all animals after transplant, followed by rapid recovery, except for B2622-10GE, which failed to gain weight, developed anemia and lost his appetite, leading to euthanasia at POD 243 (E). Moderate perioperative anemia not attributable to perioperative blood loss resolved spontaneously in all animals by week 2 (F). WBC counts dropped following induction treatment, but returned to normal within 4 weeks in all surviving animal (G). Thrombocytopenia was observed in association with thrombotic microangiopathy and consumptive coagulopathy (TM/CC) in two of the 3GE pig heart recipients, in association with AMR in B6921-10GE, and transiently within the first month in the 3 long-surviving 9- and 10GE (H). No significant creatinine elevation, and indicator of kidney injury, was seen in this study

Fig 3.. Xenografts at necropsy

At necropsy, all 3GE grafts were severely damaged. B3022-3GE graft was distended with ruptured left ventricular wall (the arrows in A); B2922-3GE graft chambers were filled with occluding thrombus and the left atrium had ruptured (B). On POD13, B6921-9GE graft was distended with intracavitary thrombi, myocardial edema, and epicardial and interstitial hemorrhage (C). B2721-9GE graft appeared normal at necropsy after 393 days, with moderate hypertrophy, limited intramyocardial fatty changes (arrows in D), and prominent pericardial fat. Apart from moderate hypertrophy, B2622-10GE graft appeared normal at POD243 (E). The very large size of the graft in comparison to the native heart at necropsy resulted from large pig donor heart size at transplant (donor weight 19.2 kg; recipient weight 9.1 kg) coupled with post-transplant hypertrophy (Figure 4b). Massive hypertrophy, scarring, and fatty infiltration were evident in the B7221-10GE xenograft at necropsy (F), with LV obliteration (left of image) and RV hypertrophy (right of image).

Fig 4.. Measurements of interventricular septum wall thickness and example of ventricular hypertrophy

Interventricular septum wall thickness (IVS) was used as indicator for hypertrophy and measured weekly by ultrasound through the entire study in a standardized fashion: long (A) and short axes (B) at the base of the LV outflow tract. An example of ventricular hypertrophy from B7221-10GE (C & D) with a septum measuring 24 mm at POD98.

Fig 5.. Inflammation markers, Troponin-I and human thrombomodulin levels

(A) Interleukin 6 and (B) C-reactive Protein elevations were seen in all animals following the initial induction treatment, primary surgical procedure, and subsequent interventions (e.g., central line removal around week 4). The rise in IL-6 was higher in two of the 3GE animals that lost their grafts. B2721-9GE demonstrated an IL-6 rise at week 2 associated with transiently diminished graft contractility, and a subsequent unexplained spike at week 36. B7221-10GE exhibited elevated CRP in association with graft failure, while a marked CRP elevation coincided with the pneumonia in case of B2721-9GE, and preceded euthanasia for clinical deterioration associated with subacute abdominal compartment syndrome and anemia in B2622-10GE. (C) Monocyte chemoattractant protein 1 was elevated for two weeks in all 9- and 10GE xenograft recipients, persisting at a reduced level beyond week 4 in B7221-10GE in association with progressive xenograft hypertrophy and immune injury. B2622-10GE exhibited generally rising levels of MCP-1 from week 8 through EOS (566.2 pg/mL) in association with progressive anemia, decreased appetite, and weight loss, clinically suggestive of chronic inflammation. In B2721-9GE MCP-1 levels returned to baseline by week 4 in association with stable graft function, and through successful management of atypical pneumonia around week 52. (D) Transient postoperative increase in troponin-I in all recipients, diagnostic of cardiomyocyte injury, returned to baseline levels in surviving animals by POD7. A second troponin-I peak was seen in association with graft immune injury and rejection. (E) Transient elevation in human thrombomodulin (hTBM) was detected in most 9- and 10-GE heart xenograft recipients within 12–24 hours after transplant. Subsequent hTBM elevation was observed at EOS in association with presumed antibody-mediated rejection in B6921-9GE; progressive xenograft hypertrophy and immune injury in B7221-10GE; and presumed abdominal compartment syndrome attributed to gradual xenograft hypertrophy in B2622-10GE. Minimal transient hTBM elevation was observed in B2721-9GE coincident with a pneumonia at POD364, with resolution after empiric antimicrobial treatment. As expected, 3-GE animals did not exhibit any significant elevation in hTBM associated with early graft failure.

Fig 6.. Anti-donor antibody levels

Level of IgM and IgG antibodies in recipient baboons against the donor pigs was measured retrospectively by incubating the baboons serum with donor pig CD2+CD20− T cells, which express the histocompatibility complex MHC class I, and CD-CD20+ B cell, which express both MHC class I and class II. Results are expressed as fold increase in mean fluorescent intensity in ELISA relative to a negative control (pig serum). B6921-9GE had a high level of IgM prior to transplant which increased further following removal of graft at POD13, most likely due to loss of absorption and stimulation of a sensitized, secondary immune response. B8522-3GE had a high anti-donor level of both IgM and IgG pretransplant that may explain the particularly rapid graft failure on POD0; demise of B2922-3GE and B3022-3GE grafts within 5 days was not associated with detectable. Initial high level of anti-class II IgM (only) was seen in B2622-10GE, which declined over the first month but remained detectable through euthanasia at 28 weeks. * Donor porcine aortic endothelial cells (PAECs), which express MHC I but not MHC II, were used instead of PBMCs for these recipients due to unavailability of donor pig leukocytes.

Fig 7.. Histologic findings

Necropsy of B3022-3GE graft showed extensive areas of myocyte ischemia and necrosis (A), which stained positive for C4d (B), consistent with acute humoral xenograft rejection or ischemia/reperfusion injury. B2721-9GE protocol graft biopsy at POD275 showed no evidence of rejection or thrombotic microangiopathy (C) and only focal C4d deposition along capillaries (D). Necropsy of B7221-10GE at POD113 showed extensive hemorrhage, thromboses, and myocyte necrosis (E) consistent with severe antibody-mediated rejection with intravascular thrombi consistent with thrombotic microangiopathy and CAV as well as foci diagnostic for ACR3R (not illustrated). The C4d was positive along capillaries and in necrotic myocytes (F). B2622-10GE showed at POD120 no evidence of rejection or thrombotic microangiopathy (G). At EOS necropsy, B2622-10GE (H) and B2721-9GE (I) xenografts exhibited macro- and microscopic areas of fat infiltrating or replacing normal myocardium, without evidence of rejection or TMA; pericardial fat several millimeters in thickness was observed in multiple long-surviving 9- and 10GE xenografts at explant.