What Is Known
• Recurrence of pruritus and cholestasis can occur in a minority of progressive familial intrahepatic cholestasis type 2 patients after liver transplantation.
• This has been shown to be the result of anti-bile salt export pump antibodies produced by the recipient against the neo-autoantigen.
• This condition is often refractory to standard anti-cellular rejection immunosuppressants.
What Is New
• There is complement activation in recurrent bile salt export pump disease with C4d deposition.
• Because most laboratories do not have the ability to detect anti-bile salt export pump antibodies, C4d may act as a surrogate marker for antibody-mediated complement activation and graft dysfunction.
• Patients with recurrent bile salt export pump respond to antibody-depletion therapy.
Progressive familial intrahepatic cholestasis type 2 (PFIC2) is the most severe form of bile salt export pump (BSEP) deficiency and usually manifests in infancy/early childhood with jaundice, pruritus, and progression to hepatic fibrosis with an increased risk for hepatocellular carcinoma (1,2). Liver indices typically show elevated serum bile acids and liver enzymes, with low or normal gamma-glutamyltranspeptidase (GGT). Histomorphologically, cholestatic giant cell hepatitis is typically seen in severe cases. Furthermore, there can be normal or reduced or absent BSEP expression on the hepatocyte canalicular membrane by immunohistochemistry (IHC) (3). Liver transplantation (LT) is offered to PFIC2 patients with refractory pruritus or end-stage liver disease (2,4).
Recurrent BSEP (rBSEP) disease has been reported in PFIC2 patients following LT, with development of de novo anti-BSEP antibodies as a proposed mechanism (5–10). There are many lines of evidence to support this. First, regimens used to treat humoral rejection (plasmapheresis, intravenous immunoglobulin, mycophenolate mofetil, rituximab, and azathioprine) are useful in rBSEP, whereas heightened immunosuppression aimed at ameliorating cellular rejection does not relieve the symptoms or normalize liver function indices. Second, ant-BSEP antibodies have been found in serum of patients with rBSEP. Third, immunoglobulin G (IgG) has been shown to be present at the canalicular surface in livers with rBSEP. These presumably anti-BSEP reactive antibodies have been postulated to block the function of BSEP through yet unknown mechanism (7,8).
We present a case with 3 episodes of rBSEP following LT, which demonstrates how anti-C4d staining—a common marker used in evaluation of antibody-mediated rejection in other organ transplants—may be applied to diagnosing rBSEP as well.
Majority of the clinical and pathology tests were performed as a part of standard medical care. Additional pathology tests were performed as research using departmental (pathology) funds. This report was drafted after appropriate permission from the institutional review board for the conduct of research on human subjects and also after parental permission to conduct and publish this study. Immunohistochemical staining was performed on formalin-fixed paraffin-embedded tissue using the following antibodies: BSEP (sc-74500; Santa Cruz Biotechnology, Dallas, TX), CD10 (PA0270; Leica Biosystems, Buffalo Grove, IL), C5b-9 (NC0292288; Fisher Scientific, Waltham, MA), C4d (PA0792; Cell Marque, Rocklin, CA), and IgG (A0423; Dako, Santa Clara, CA). Stains were performed on a Leica Bond III automated machine with Bond Polymer Refine Detection DAB detection kit (Leica Microsystems Inc, Buffalo Grove, IL) according to the manufacturer's instructions. Direct immunofluorescence (IF) for C4d (NC9158836; Fisher Healthcare, Waltham, MA) was performed on snap frozen tissue using an affinity purified secondary anti-mouse IgG antibody, prepared in horse serum (Vector Laboratories; Burlingame, CA). Direct IF for IgG on patient's liver sample (snap frozen) and indirect IF with patient's serum were performed as described previously (8) and published as patient number 3 in a previous report (10).
We report the case of a 16-month-old girl who presented to our institution with pruritus and evaluation for liver transplantation (LT). She had initially presented to another institution at 5 months of age with mild jaundice, low GGT cholestasis and elevated bile acids. Liver biopsy at that time showed cholestatic giant cell hepatitis. Furthermore, IHC performed on a research basis showed absent BSEP on canaliculi, strongly suggestive of PFIC2.
At presentation, the patient's pruritus was the most important complication, leading to excoriations of the skin. Medical treatment mainly with ursodeoxycholic acid and rifampin had been tried without success. Persistent ursodiol-associated diarrhea limited its use in her case even though she used it on a relatively routine basis. She had also undergone an ileal exclusion procedure (ileal-cecal anastomosis excluding distal 84 cm of ileum) at 13 months of age. The ileal exclusion did not ameliorate pruritus, but was associated with development of frequent loose stools. Given the persistent pruritus, her condition was expected to progress and she was listed for LT (Table 1). At 18 months, she underwent whole deceased donor liver transplantation along with ileal bypass correction.
Pretransplantation evaluation included negative panel reactive antibodies by flow cytometry. The explanted liver showed bland hepatocanalicular cholestasis, scattered giant cell transformation (Fig. 1A), minimal portal fibrous expansion (Fig. 1B), and complete absence of canalicular BSEP protein expression by IHC (Fig. 1C). Subsequent genetic testing at age 4 (when testing became readily available) agreed with the diagnosis, identifying heterozygous mutations of c.1723C>T (p.R575X) (nonsense mutation expected to eliminate protein product), and c.2178+1G>T (splice site splice mutation expected to produce a functionally deficient or truncated protein) in the ABCB11 gene.
The patient's post-LT course has been characterized by 3 main issues. First, she presented 6.7 years after LT with vomiting, bloody diarrhea, increasing abdominal pain and weight loss. Both computational tomography examination and endoscopic biopsies of the terminal ileum and cecum showed features of Crohn disease. Liver indices were unremarkable at that time. She was started with 400 mg mesalamine and 40 mg prednisone for Crohn disease. Budesonide (3 mg, tid) and azathioprine (37.5 mg increased to 75 mg) were added to the regimen a few months later due to poor response. She was maintained on this regimen along with 1 mg daily tacrolimus for about a year (Table 2). All medications for Crohn disease were discontinued 8.5 years post-LT when she was symptom free and endoscopically normal.
Second, 2 episodes of biopsy-confirmed mild acute cellular rejections were encountered early in the post-transplant course. The first episode was at 1.8 years post-LT with abdominal pain and dark urine, and the second at 2.8 years post-LT with transaminitis following an upper respiratory infection (Table 1). Tacrolimus trough level was lower than the target level for the first episode, but within target level during the second episode. Canalicular expression of BSEP by IHC was preserved in both liver biopsies (data not shown), and she responded to steroid pulse, initiation of mycophenolate, and increased tacrolimus (Table 1).
Third, she continued to have pruritus. The first complaints were mild and self-resolving, in the first 6 months after transplant. The pruritus, however, re-appeared 6 years after transplant when inflammatory bowel disease was diagnosed. Initially, it was unaccompanied by changes in liver laboratory tests, and was treated successfully with metronidazole. At 8.9 years post-LT, she presented with pruritus and elevated liver enzymes. This was followed by 2 additional episodes of similar presentation at 10.3 and 11.0 years post-LT, respectively (Table 3). Tacrolimus trough levels were lower than target (5–7 ng/mL) for the first and third episodes of pruritus, and the third episode was preceded by recent low-grade fever and influenza in family members. Her serum bile acids were also elevated. Both the symptoms and biochemical profile were similar to her initial pretransplant presentation of PFIC2. Circulating IgG antibodies with affinity to canalicular surface of hepatocytes were identified by indirect IF using normal human liver control tissue, during all 3 episodes of pruritus.
All 3 biopsies showed moderate lobular disarray, patchy feathery degeneration/cholate injury, and rare multinucleated giant hepatocytes, recapitulating PFIC2, with canalicular cholestasis only seen in the second biopsy (Fig. 2A). By IHC, BSEP expression was lost in areas with feathery degeneration in all 3 biopsies (Fig. 2B), whereas it was retained in areas with relatively normal morphology. As a control, staining for other canalicular proteins (CD10 and MDR3) was maintained (data not shown). There was a clear canalicular staining pattern for human IgG antibodies co-localizing with MRP2 in the patient's liver by direct IF (10). IHC for IgG showed patchy canalicular staining in all 3 biopsies that was accentuated in areas of feathery degeneration (Fig. 2C). In addition, mild acute cellular rejection was seen in the first and third biopsies (Fig. 2D). IHC for C4d showed portal vein positivity in the first (Fig. 2E) and second biopsies (not shown), but was negative in the third. Direct IF for C4d was only performed on the last 2 biopsies, it was negative in the second biopsy, but diffusely positive along the sinusoids in the third (Fig. 2F). Complement membrane attack complex (C5b-9) was detected in hepatic artery endothelial cells and basement membrane by IHC in the second biopsy (Fig. 2E inset), and rarely in Kupffer cells in the third biopsy (data not shown).
She responded well to regimens lowering circulating antibodies (rituximab with or without intravenous immunoglobulins) resulting in resolution of pruritus and decreased serum bile acid levels (Table 3). Patient is currently stable on sirolimus with normal liver indices at her most recent follow-up, 17 months from the last episode of rBSEP. Viral studies by PCR (EBV, HHV6, HHV7, HSV 1/2, CMV, adenovirus, parvovirus) of liver tissue and serology (HBV, HCV, HAV) had been negative in all rejection episodes. No structural biliary tract abnormalities were identified during these episodes of pruritus using standard imaging modalities.
At least 20 patients with rBSEP disease due to anti-BSEP antibodies have been reported since its first recognition in 2009 (5–10). The estimated prevalence of rBSEP disease is approximately 8% (6). Similar to the prototypical disease PFIC2, the common presenting symptoms include pruritus, jaundice and occasionally diarrhea. Patients tend to have normal to low levels of GGT and increased transaminases. Biliary bile acid concentration is reduced with a reciprocal increase in serum bile acids. The interval between liver transplantation and the first episode of cholestasis is variable, ranging from 3 months to 17 years. Our patient's episodes of rBSEP disease (8.9, 10.3, and 11 years post-LT) fell into this time frame.
With detection of circulating anti-BSEP antibodies and demonstration of IgG deposition at the canalicular surface of hepatocytes in rBSEP disease (8–10), the pathophysiology has been proposed to resemble a type II hypersensitivity/autoimmune reaction. In this form of autoimmunity, IgG or immunoglobulin M binds with autoantigens located on the cell surfaces, leading to functional changes by either receptor activation or inactivation. The classic example of this form of autoimmunity is Grave's disease with anti-TSH receptor antibodies. In our patient, there was definite hepatocyte injury with paucity of inflammatory cells, and absent detectable complement deposition (C4d and MAC) along the canaliculi. These combined features support a non-cytotoxic form of hypersensitivity reaction. The absence of BSEP immunoreactivity in abnormal hepatocytes in the allograft may be secondary to endogenous anti-BSEP antibodies masking the epitopes on neo-autoantigen. The variable results with respect to BSEP detection by IHC in 6 other reported rBSEP patients (5,7,9) may also result from variable epitopes being masked and/or variable degrees of structural similarity between reagent anti-BSEP antibody and endogenous antibody within the antigen-binding region (Fab). The neo-autoantigen BSEP may become exposed to the immune system with ongoing typical hepatocyte turnover. Most patients with rBSEP disease have detrimental ABCB11 gene mutations resulting in early mRNA degradation, absent protein expression, and consequently lack of auto-tolerance to BSEP antigen (11). Alternatively, some ABCB11 mutations may significantly alter the tertiary structure of BSEP protein, but allow its expression on the canalicular membrane. This second category of PFIC2 patients can also develop rBSEP disease somewhat analogous to the hemolytic transfusion reaction in D variant recipients harboring anti-D autoantibodies (12).
Documented triggering factors for rBSEP disease are change in the therapeutic immunosuppression regimen, decreased immunosuppression, viral infections (EBV and HBV), and high-dose antibiotics (7,9). In our patient, a preceding triggering event was not identified in the first 2 episodes, while a possible influenza infection was noted before the last episode. It is possible that a preceding immune disturbance may stimulate proliferation of a particular clone of memory B cells involved in the alloantibody response, and an episode of acute hepatocyte damage from toxins or viruses may result in increased exposure/availability of the neo-antigen. Both of these scenarios could lead to increased binding of the BSEP by anti-BSEP antibodies that overwhelm the phagocytic capacity of the reticuloendothelial system, resulting in rBSEP disease. Because patients are on continuous immunosuppressants, de novo inflammatory bowel disease after LT is rare, with most reported cases being ulcerative colitis (13), it is uncertain if our patient's immune system predisposed her to rBSEP disease.
There are no previous reports demonstrating sinusoidal C4d immunoreactivity associated with rBSEP episodes. C4d is a stable by-product during the activation of the classic and lectin complement pathways, with currently no known biological function. It binds covalently with the endothelium near the site of complement activation, allowing detection by IHC and serving as a marker of complement activation. Complement activation is initiated by antibodies in the immunoglobulin M and/or IgG classes, and it is widely used in assessing humoral rejection in heart, kidney and pancreas allografts (14). Diffuse linear sinusoidal C4d immunoreactivity by direct IF during the last pruritic episode in our patient is consistent with complement activation and may suggest antibody-mediated or humoral rejection. Non-human leukicyte antigen (HLA) antibodies against epitopes on endothelial or epithelial cells or C4d deposition in the absence of circulating donor-specific HLA antibodies have been increasingly recognized in antibody-mediated graft dysfunction (15,16). The findings in the present case are congruent with the existing literature (15), implicating non-HLA antibodies, such as anti-BSEP being involved in complement fixation via mechanisms typically described for humoral rejection associated with anti-HLA antibodies. Further work is necessary to confirm these observations, with respect to rBSEP disease.
There was significant difference in the sinusoidal C4d deposition by IHC and direct IF, in the third biopsy. It has been shown that C4d immunoreactivity by direct IF correlates more closely with lymphocyte cross-match results than by IHC (17). Even though our patient's clinical presentation and response to treatment during the second pruritus episode indicate a bona fide antibody-mediated rejection, direct IF for C4d was negative in the biopsy. It is uncertain if this was due to the difference in the IgG class of anti-BSEP antibody between the second and third episodes, as IgG3 is the most efficient activator of complement, followed by IgG1 and IgG2. Also, various inflammatory conditions can activate complement pathways other than humoral rejection. The possibility that the influenza infection before the last episode of pruritus may have triggered complement activation cannot be excluded with certainty (18).
Our patient's graft dysfunction was not a conventional humoral rejection secondary to donor-specific anti-HLA class II antibodies detectable by lymphocyte crossmatch. She, however, responded favorably to anti-humoral rejection regimen, similar to other reported cases of rBSEP disease. Reported examples of patients with rBSEP disease requiring a second transplant are those who were offered immunosuppression regimens targeting acute cellular rejection (5,9). Even though early diagnosis and appropriate treatment of rBSEP disease appear to resolve post-transplant pruritus and cholestasis successfully, long-term graft survival is yet to be documented.
In conclusion, rBSEP disease should be considered when evaluating rejection in PFIC2 patients who have undergone LT, especially when severe pruritus is present. It is uncertain if this phenomenon secondary to non-HLA antibody production also applies to other liver specific protein-deficient disorders, for example urea cycle defects. A linear sinusoidal pattern of C4d deposition, by direct IF, may serve as a surrogate marker for circulating donor-specific antibodies and appropriate tissue triaging to save a snap frozen portion of liver core in OCT medium may be required when clinically indicated.
The authors thank Ms Karen Prince, graphic designer for her help with the figures.
1. Thompson R, Strautnieks S. BSEP: function and role in progressive familial intrahepatic cholestasis. Semin Liver Dis
2. Davit-Spraul A, Gonzales E, Baussan C, et al. Progressive familial intrahepatic cholestasis. Orphanet J Rare Dis
3. Evason K, Bove KE, Finegold MJ, et al. Morphologic findings in progressive familial intrahepatic cholestasis 2 (PFIC2): correlation with genetic and immunohistochemical studies. Am J Surg Pathol
4. Jacquemin E. Progressive familial intrahepatic cholestasis. Clin Res Hepatol Gastroenterol
2012; 36 (suppl 1):S26–S35.
5. Lin HC, Alvarez L, Laroche G, et al. Rituximab as therapy for the recurrence of bile salt export pump deficiency after liver transplantation. Liver Transpl
6. Siebold L, Dick AA, Thompson R, et al. Recurrent low gamma-glutamyl transpeptidase cholestasis following liver transplantation for bile salt export pump (BSEP) disease (posttransplant recurrent BSEP disease). Liver Transpl
7. Jara P, Hierro L, Martinez-Fernandez P, et al. Recurrence of bile salt export pump deficiency after liver transplantation. N Engl J Med
8. Keitel V, Burdelski M, Vojnisek Z, et al. De novo bile salt transporter antibodies as a possible cause of recurrent graft failure after liver transplantation: a novel mechanism of cholestasis. Hepatology
9. Maggiore G, Gonzales E, Sciveres M, et al. Relapsing features of bile salt export pump deficiency after liver transplantation in two patients with progressive familial intrahepatic cholestasis type 2. J Hepatol
10. Stindt J, Kluge S, Droge C, et al. Bile salt export pump-reactive antibodies form a polyclonal, multi-inhibitory response in antibody-induced bile salt export pump deficiency. Hepatology
11. Kubitz R, Droge C, Kluge S, et al. Autoimmune BSEP disease: disease recurrence after liver transplantation for progressive familial intrahepatic cholestasis. Clin Rev Allergy Immunol
12. Daniels G. Variants of RhD—current testing and clinical consequences. Br J Haematol
13. Naito T, Shiga H, Endo K, et al. De novo Crohn's disease following orthotopic liver transplantation: a case report and literature review. Intern Med
14. Murata K, Baldwin WM 3rd. Mechanisms of complement activation, C4d deposition, and their contribution to the pathogenesis of antibody-mediated rejection. Transpl Rev
15. Angaswamy N, Tiriveedhi V, Sarma NJ, et al. Interplay between immune responses to HLA and non-HLA self-antigens in allograft rejection. Hum Immunol
16. Tait BD, Susal C, Gebel HM, et al. Consensus guidelines on the testing and clinical management issues associated with HLA and non-HLA antibodies in transplantation. Transplantation
17. Kozlowski T, Andreoni K, Schmitz J, et al. Sinusoidal C4d deposits in liver allografts indicate an antibody-mediated response: diagnostic considerations in the evaluation of liver allografts. Liver Transpl
18. Monsalvo AC, Batalle JP, Lopez MF, et al. Severe pandemic 2009 H1N1 influenza disease due to pathogenic immune complexes. Nat Med