Systematic Investigation of Elevated Cholestatic Enzymes During The Third Posttransplant Month

There are no commonly accepted grading systems for gauging cholestasis after liver transplantation (LT). Severe intrahepatic cholestasis has been defined as a serum total bilirubin level >100 μmol (>5.9 mg/dL), an increase in the alkaline phosphatase level 3 times the upper limit of normal, or a combination of the two.1 Although severe cholestasis is a rare event, it is commonly associated with morbidity that may lead to a requirement for retransplantation. In contrast, transient and minor elevations of cholestatic liver tests are reportedly of limited clinical significance.2 Nevertheless, cholestasis with either an alkaline phosphatase level >2 times the upper limit of normal or a combined elevation of both bilirubin and alkaline phosphatase levels is associated with increased mortality. According to this less stringent definition, approximately 1 in 4 patients in the LT program at the University of Alberta Hospital met the criteria for cholestasis in the third month after LT in a review of more than 900 adult LT recipients with available hepatic biochemistry data (Fig. 1). In this analysis, graft survival was worse for patients with cholestasis (median survival: 88 ± 15 versus 183 ± 10 months, P < 0.001), and patient survival also decreased (median survival: 115 ± 20 versus 186 ± 9 months, P < 0.001). The increased mortality rate for patients with cholestasis was mainly observed in the first 2 years after LT (Fig. 1). Multiple risk factors are associated with the development of early severe intrahepatic cholestasis (Table 1). Several of these factors may combine to predispose patients to chronic allograft damage, and a reassessment of perioperative events should guide the management of cholestatic disease in the third posttransplant month (Fig. 2). Graft nonfunction/dysfunction is associated with severe intrahepatic cholestasis and represents a major cause of graft failure along with hepatic artery thrombosis (HAT). This problem may occur as a result of a preservation/reperfusion injury in a critically ill patient who receives a suboptimal allograft with a prolonged cold ischemia time. Additional risk factors also include older donors and increased steatosis in the allograft.1-3 The effects of cold and warm ischemia can result in severe postoperative cholestasis in LT recipients. Biochemical evidence of preservation/reperfusion injury is usually observed directly after LT with elevated transaminases and cholestasis; although the former normalize rapidly, the correction of bilirubin can take several months.1-3 The process may become complicated by the subsequent development of sepsis. Although the hepatic parenchyma may be quite resilient to cold ischemia, preservation injury directly affects the microhepatic circulation and can lead to the detachment of biliary epithelium from the basement membrane. Previous studies have reported that hepatic cold ischemia extending from 10 to 12 hours leads to nonanastomotic biliary stricturing in 25% of LT recipients, and the frequency rises to greater than 50% when cold ischemia is prolonged for more than 13 hours.2 Histologically, the diagnosis of preservation/reperfusion injury can be made through the observation of steatosis, cholestasis, and ballooning degeneration of hepatocytes. The use of ABO-incompatible grafts is linked with an approximately 2-fold increase in mortality 3 months after LT.4 Both immune and vascular mechanisms have been implicated in the development of allograft dysfunction because ABO-incompatible recipients have an increased risk of developing HAT, biliary strictures, and chronic rejection. ABO-incompatible recipients have a higher incidence of antibody-mediated rejection, which usually occurs within the first month after LT.4 This may be because the implicated blood antigens have been localized to bile ducts and vascular endothelium within the allograft.1 The diagnosis of antibody-mediated rejection can be established by the detection of donor-specific antibodies with diffuse complement component 4d deposition involving endothelium or stroma in the majority of portal tracts or sinusoids.4 Accordingly, the use of ABO-incompatible allografts is restricted to emergency LT or living donor LT, and recipients usually require prophylactic antibody-depleting maneuvers. Infections frequently occur after LT (60%-80%), and sepsis is the commonest cause of severe intrahepatic cholestasis.1, 3 As a result, cephalosporin prophylaxis is commonly employed in the perioperative period. Most infectious episodes are observed within the first 2 months and are usually related to abdominal surgery or nosocomial infections. Infections often complicate other processes outlined in Table 1. Accordingly, cholestatic patients should be actively investigated for the presence and source of infections (Fig. 2), especially because clinical signs and symptoms may be masked by immunosuppression. Bacterial and fungal infections most often involve other systems and do not mandate liver biopsy. When liver biopsy is performed, however, the histological picture of sepsis usually appears complex with neutrophil infiltration of the biliary tract, bile duct proliferation, and cholate stasis. Once biliary complications have been excluded by imaging, a viral infection of the allograft may require a histological evaluation to establish a diagnosis and rule out rejection. Cytomegalovirus (CMV) disease in the liver has been reported to occur in 1 of 5 patients with severe cholestasis, whereas cholestasis seldom occurs with CMV disease outside the liver.3 Historically, CMV infections have been found in up to 30% to 50% of LT patients in the first 1 to 3 months after LT and especially in seronegative recipients with seropositive donors. However, the common practice of CMV DNA monitoring with preemptive therapy and valganciclovir prophylaxis for seronegative recipients has considerably curtailed the incidence of CMV disease. Approximately a quarter of all cases of drug-induced liver injury occur in months 1 to 3 after LT.5 The common culprits include antibiotics (42%), immunosuppressive agents (14%), antihyperlipidemic agents (7%), antivirals (7%), antifungals (3%), and miscellaneous agents (21%), with trimethoprim-sulfamethoxazole as the commonest cause.5 Tacrolimus, cyclosporine, and azathioprine have all been implicated in cholestasis, but the effects are rarely severe and are easily reversed by the reduction or discontinuation of treatment.1, 5 Malignancy is covered extensively in this syllabus elsewhere. Approximately 50% of posttransplant lymphoproliferative disease cases present within the first year after LT. The disease is mainly observed in children and young adults, and cholestasis has been reported in up to 16% to 57% of patients.1 Management consists of imaging, the evaluation of Epstein-Barr virus DNA levels, the reduction of immunosuppression, and the consideration of rituximab.1 Biliary complications in LT recipients span the entire spectrum and range from bile leaks, strictures, and biliary cast formation to sphincter of Oddi dysfunction (Table 2).6 These may occur as a result of surgical complications, prior biliary disease, and perioperative factors (Table 1). Biliary reconstruction is commonly performed with a duct-to-duct anastomosis (choledochocholedochostomy), whereas a Roux-en-Y anastomosis (choledochojejunostomy) is reserved for patients with biliary disease and those with a size mismatch for the donor and recipient bile ducts. There has been a transition away from the use of T-tubes for biliary anastomoses because their use has been associated with decreased patient survival secondary to an increased risk of bile leaks and cholangitis.6, 7 Because of the impact of biliary complications on patient and graft survival, a duct-to-duct anastomosis is the preferred procedure: it reduces the operative time, minimizes the colonization of the biliary tract by bacteria, and also results in fewer biliary complications.6 Biliary complications are relatively common, and the incidence has increased with the emergence of living donor LT. For example, deceased donor LT is associated with a frequency of biliary strictures varying from 5% to 15%, whereas the incidence ranges from 28% to 32% with living donor LT.7 Bile leaks within the first weeks mainly arise at the anastomotic site and occur as a result of technical issues or ischemia. The presentation varies from an incidental finding on abdominal imaging with a fluid collection in an asymptomatic patient to a symptomatic patient presenting with abdominal pain and rising bilirubin levels with or without fever. For patients with a suspected bile leak arising from a duct-to-duct anastomosis, endoscopic retrograde cholangiopancreatography (ERCP) can be used for diagnosis and treatment with sphincterotomy and stent placement, whereas patients with a Roux-en-Y anastomosis may require external/internal biliary stent placement by interventional radiology or therapeutic ERCP with a double-balloon enteroscope; both may eventually require surgical correction.6, 7 Patients receiving a split allograft may also develop bile leaks from the cut surface. Clinically evident anastomotic biliary strictures tend to present in the second posttransplant month onward and usually occur as a result of the surgical procedure.6 Suspicion should be raised if a patient develops raised alkaline phosphatase levels even in the absence of biliary dilatation on ultrasound (US). Balloon dilatation followed by stent placement by ERCP is the treatment of choice, with repeat stenting every 3 months for approximately 1 year.7 Successful endoscopic treatment may be expected in 80% to 90% of deceased donor recipients versus 60% to 75% of living donor recipients.7 In contrast, nonanastomotic biliary strictures are less amenable to endoscopic therapy because they tend to be longer, multiple, and located within the intrahepatic and extrahepatic bile ducts. The development of nonanastomotic biliary strictures is more complex and includes a confluence of vascular and immune factors that cause diffuse disease (Table 2). Approximately 50% of LT recipients with nonanastomotic biliary strictures require liver retransplantation or die waiting.7 Notably, the use of donors after cardiac death is associated with a high incidence of biliary complications (60%), and the likelihood of uncorrectable hepatic ischemia before organ retrieval may limit their utility despite the donor shortage.6 Gall stones, sludge, clots, and biliary casts occur with a frequency ranging from 3% to 12% and may be precipitated in part by biliary strictures, bile stasis, and bacterial infections.6 The majority of patients can be managed with sphincterotomy, but subjects with casts may require further manipulation with percutaneous transhepatic cholangiography (PTC); under these circumstances, the possibility of biliary ischemia should be entertained.6 HAT is discussed in detail elsewhere in this supplement. By definition, the effects of late HAT are observed more than 30 days after LT and are usually less severe than the effects of early HAT.8 The frequency of late-onset HAT is approximately 1% at experienced transplant centers that perform LT more than 30 times per year.1, 2 Depending on the extent of hepatic ischemia, the presentation may vary from asymptomatic disease with increased transaminases to biliary necrosis presenting with fever, abdominal pain, and cholangitis, which may progress to a bile duct leak or a hepatic abscess. The diagnosis is usually established by US/Doppler imaging, and a radiological intervention with angiography may be required to re-establish arterial flow (Fig. 2). The surgical risk factors include a prolonged cold ischemia time, technical problems with the arterial anastomosis (including a complex anatomy), the amount of blood products required, and the duration of surgery1,2; hypercoagulable states, cigarette smoking, and CMV infection have been identified as nonsurgical risk factors for HAT.1, 2 Historically, acute cellular rejection has been reported to occur in up to two-thirds of LT recipients and is usually seen within the first month with a median onset at 8 days.9 Depending on the severity, patients may present with elevated cholestatic liver tests, but hepatic biochemistry does not necessarily correlate well with the development of histological disease. The diagnosis is established histologically by the detection of endotheliitis, eosinophils, bile duct damage, and a mixed portal infiltrate. The frequency of rejection is heavily influenced by induction therapy and the immunosuppression regimen. It has long been recognized that acute cellular rejection does not negatively affect patient or allograft survival.9 Poor control of acute cellular rejection and antibody-mediated rejection have both been implicated in the development of ductopenic rejection.4 Many other factors, including an LT recipient with a prior diagnosis of primary biliary cirrhosis (PBC) or primary sclerosing cholangitis (PSC), a non-European recipient, a male donor for a female recipient, a human leukocyte antigen mismatch, a CMV infection, and a lack of azathioprine in the cyclosporine era, have been linked.2, 9 However, it should be recognized that some of these associations are somewhat tenuous because studies with small sample sizes have not been adequately reproduced. Specifically, it has proven difficult to discriminate the histological difference between recurrent PBC or PSC with ductopenic rejection. Histologically, a diagnosis of ductopenic rejection can be made when more than 50% of the portal tracts lack bile ducts; more rarely, foam cell arteriopathy or obliterative arteritis can be observed to help confirm the diagnosis.2, 9 Analogously to antibody-mediated rejection, both vascular and immune factors affect the development of ductopenic rejection; patients have histological evidence of obliterative arteriopathy, which is thought to contribute to biliary ischemia.2, 9 Clinically, ductopenic rejection manifests as a relentless increase in serum bilirubin and alkaline phosphatase levels. In contrast, patients with recurrent biliary disease usually have fluctuating cholestatic biochemistry tests and a slower progression of vanishing bile ducts. Patients with ductopenic rejection on cyclosporine maintenance therapy may respond to a change in the calcineurin inhibitor to tacrolimus, but the prognosis is usually poor, and retransplantation may be required. Notably, the prevalence has fallen considerably from an average of 8% in the cyclosporine era to a range of 2% to 5% in the tacrolimus era.2, 9 Hepatitis C virus (HCV) infection after LT is almost universal, its management remains challenging, approximately 1 in 5 patients will develop cirrhosis within 5 years, and the outcomes remain inferior in comparison with those for LT patients without infectious disease.10 The recurrence of a hepatitis B virus infection is seldom a problem because patients usually have adequate viral control before LT, and effective antiviral treatment is available after LT. Historically, approximately 5% of LT patients with hepatitis B virus developed severe intrahepatic cholestasis [called fibrosing cholestatic hepatitis (FCH)] with nearly universal fatal outcomes before the development of antiviral therapy. Similarly, FCH has affected LT recipients with HCV infections with a frequency reported in the range of 2% to 8% in different series.11 In order to better understand and study FCH in LT patients with HCV, a consensus definition was drafted in 2003: disease occurring more than 1 month after LT; a serum bilirubin level > 6 mg/dL (>102 μmol); a serum alkaline phosphatase level > 5 times the upper limit of normal; very high HCV RNA levels; and the characteristic histological ballooning of hepatocytes, cholangiolar proliferation, paucity of inflammation, and fibrosis in the absence of biliary complications or HAT.10 The study of this cholestatic phenotype of HCV recurrence highlights a major problem with recurrent disease management. The overuse of immunosuppression is associated with accelerated HCV progression, whereas too little permits rejection and a subsequent need for more immunosuppression. Indeed, a recent series on HCV-related FCH found that a strong independent predictor of FCH was a previous Banff score > 5; the implication was that patients likely received increased immunosuppression, a known trigger for FCH.11 This underscores the difficulty with liver biopsy interpretation when one is trying to differentiate alloimmunity from recurrent HCV because both processes may be ongoing at the same time without clear directives within the Banff scoring system of how to differentiate the two. Accordingly, it is recommended that an HCV RNA level be obtained to prevent the overuse of immunosuppression in patients with a high viral load. Another important observation was that total bilirubin happened to be the strongest laboratory predictor of cholestatic HCV.11 Indeed, a subanalysis of patients with HCV infections at our center revealed that graft survival and patient survival were worse for patients with cholestasis in the third post-LT month versus patients without cholestasis (Fig. 3). Accordingly, the presence of cholestasis in LT patients with HCV should signal a review of immunosuppression as well as options for antiviral treatment. The prevalence of recurrent PBC increases with time but does not appear to adversely affect survival in comparison with the survival of PBC LT recipients without histological disease. For our unit, we reported that approximately 1 in 4 PBC patients developed recurrent disease, and the earliest case was diagnosed within 2 months of LT.12 Indeed, it has been shown that the disease-specific phenotype of aberrant mitochondrial antigen expression in bile ducts can be observed in most PBC LT recipients soon after transplantation, and a similar proportion of patients (∼70%) have detectable anti-mitochondrial antibodies. Notably, cyclosporine use has consistently been reported to be protective against the development of recurrent PBC.12 The detection of cholestasis in the third LT month had no significant impact on our cohort of 122 PBC LT recipients with respect to graft or patient survival or the development of recurrent disease. Thus, it appears that there is a spectrum of recurrent PBC that responds favorably to ursodiol therapy without adversely affecting outcomes.12 Although the frequency of recurrent PSC is comparable to the frequency of PBC, LT recipients with recurrent PSC tend to fair worse with diminished graft and patient survival.13 In comparison with PBC recipients, there is an increased risk of retransplantation, and recurrent PSC occurs more frequently and with a more rapid onset in patients receiving a second graft. The possibility of an infectious (transmissible) etiology of PSC is also supported by the observation that patients with an intact colon at the time of LT are at higher risk for recurrent PSC. Other risk factors linked with recurrent disease include increased episodes of cellular rejection, immunosuppression, and the presence of inflammatory bowel disease, but data are inconsistent from study to study. Notably, in our patients with PSC (n = 92), the presence of cholestasis in the third posttransplant month did not affect graft survival or patient survival. However, the probability of recurrent PSC was higher for patients with PSC and cholestasis at 3 months versus PSC patients without cholestasis (Fig. 4); patients with PSC and cholestasis at 3 months had a hazard ratio of 2.62 for PSC recurrence (95% confidence interval = 1.11-6.15, P = 0.03).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.