Tolerance to liver allograft after allogeneic hematopoietic cell transplantation for severe aplastic anemia from the same HLA-matched sibling donor

Development of severe aplastic anemia (SAA) has been well described following non-A and non-B hepatitis or liver transplantation (LT), most often in the pediatric population. Its pathophysiology is unclear, but thought to be due to immune-mediated injury to liver and hematopoietic cells by unidentified viral organisms. In young and fit SAA patients, allogeneic hematopoietic cell transplantation (HCT) from HLA-matched sibling donor is the treatment of choice. Successful cases of HCT for SAA following LT have been described. However, all required long-term immunosuppressive therapy (IST) to prevent liver rejection or GVHD. Requirement for long-term immunosuppression predisposes transplant recipients to significant risks from opportunistic infections and secondary malignancies. While allograft tolerance has been reported in combined HCT and renal or lung transplantation from the same donor,1, 2, 3 outcomes of IST withdrawal following sequential LT and HCT for patients with SAA are unknown. We report a case of a 30-year-old man who presented with fulminant hepatic failure with jaundice and encephalopathy (aspartate transaminase (AST) 2014 U/L (5–34 U/L), ALT 2725 U/L (7–40 U/L), bilirubin 521 μmol/L (<22 μmol/L), international normalized ratio (INR) 9.8 (0.80–1.20)). A detailed history and physical examination did not identify an etiology of his liver failure. Viral hepatitis markers, autoimmune and metabolic workup were all negative. A liver biopsy revealed severe hepatitis of nonspecific etiology, with parenchymal collapse and submassive necrosis, mild-to-moderate portal and parenchymal lymphoid infiltrate. The patient underwent a living-related LT from his sister within 2 weeks, and received post-LT immunosuppression with mycophenolate mofetil, tacrolimus and prednisone. Two months later, he developed severe transfusion-dependent pancytopenia (neutrophil count 0.0 × 109/L, platelet count 2 × 109/L) and reticulocytopenia (reticulocyte count 6 × 109/L), unresponsive to G-CSF and withdrawal of mycophenolate mofetil. He was referred to hematology 4 weeks after onset of pancytopenia. BM biopsy revealed hypoplastic marrow with <10% cellularity, no evidence of dysplasia and normal cytogenetics, all consistent with SAA. Paroxysmal nocturnal hemoglobinuria (PNH) screen was negative. There was no family history or stigmata of inherited marrow failure disorders, and the Fanconi screen was negative. Subsequently, he received HCT from the same matched sibling donor of his LT. The conditioning regimen consisted of alemtuzumab 20 mg s.c. on day −7 and 30 mg s.c. on day −6, high-dose CY 50 mg/kg i.v. once daily from days −5 to −2, followed by infusion of 4.2 × 108/kg nucleated cells. Tacrolimus was used as an additional GVHD and liver rejection prophylaxis, and prednisone was tapered off. Hematological recovery was achieved on day 20 after HCT. Donor chimerism on the whole blood as measured by STR analysis was 100% on day 30 and remained stable between 90–99% on day 60, 120, 240, 270 and day 330 after HCT. Tacrolimus taper was started slowly 6 months after HCT with close monitoring of liver rejection. He was successfully weaned off all IST by 9 months after HCT. Unfortunately, his sibling donor died from metastatic stomach adenocarcinoma 10 months after stem cell donation. The recipient's malignancy screen via endoscopy and imaging was negative. At 45 months after HCT, he continues to be in excellent health, with no evidence of GVHD or liver rejection off all immunosuppression. Although the association of SAA following non-typeable hepatitis or LT has been well described, the diagnosis of SAA is often delayed. In our case, the etiology of pancytopenia was initially attributed to myelosuppression from mycophenolate mofetil, reinforcing the importance of early consideration of SAA as a potential cause of new onset pancytopenia following hepatitis or LT. Ten other cases of allogeneic HCT in patients who developed SAA after LT for fulminant hepatitis have been reported4, 5, 6, 7, 8, 9, 10, 11, 12 (Table 1). All except one occurred in the pediatric population. Eight patients underwent cadaveric orthotopic LT and two received living-related LT followed by HCT from different donors, all required long-term IST to prevent rejection. One patient developed acute GVHD and mild liver rejection while on CYA and prednisone,4 another developed rejection during IST tapering.8 Our case is unique in its use of the same donor for LT and HCT for SAA, which permitted successful IST withdrawal and donor-specific tolerance through the induction of hematopoietic chimerism. In the non-SAA setting, there have been two case reports of allograft tolerance following living-related LT and HCT from the same donor: a 4-month-old girl underwent LT followed by haploidentical HCT from her mother for familial hemophagocytic lymphohistiocytosis,13 and a 41-year-old man received an HCT for ALL followed by a LT from the same matched sibling donor for cholangiocarcinoma.14 Both patients maintained stable liver graft function off all IST for up to 22 months, with no signs of GVHD or rejection. Cases of fulminant GVHD following HCT and solid organ transplantation (SOT) from the same donors have been described, raising the possibility of increased risk for severe GVHD from host tolerance to donor allo Ags presented by previous allografts. Alemtuzumab-based conditioning regimens have been shown to be associated with very low incidence of acute and chronic GVHD following HCT for SAA in a previous pilot study from our centre.15 In addition, use of alemtuzumab obviates the need for MTX for GVHD prophylaxis and reduces the risk of hepatotoxicity in patients with compromised liver function. Our case shows that high-dose CY and alemtuzumab can be safely delivered as a part of the conditioning regimen for SAA in this setting. Cautious interpretation is warranted for case reports of successful allograft tolerance induced by combined SOT and HCT because of a positive reporting bias, short follow up in some reports, and the difference between operational tolerance and true tolerance. Liver allografts appear to be less susceptible to rejection than other SOT, operational tolerance has been reported in patients with non-adherence to IST and in up to 20% of selected LT after withdrawal of IST. Despite clinical evidence of stable graft function off IST, low-grade rejection may take place leading to slow progressive allograft dysfunction. To the best of our knowledge, this is the first report of successful induction of long-term liver allograft tolerance by subsequent HCT from the same matched sibling donor for SAA following LT. Selected SOT recipients with hematological malignancies may benefit from combined SOT and HCT to intentionally achieve donor-specific allotolerance and obviate the need for chronic IST. The authors declare no conflict of interest.