Treatment of Acute and Chronic Rejection

Immune Mechanisms and Pathology

AR is mediated predominantly by effector (killer) T cells that develop in response to donor antigens from the transplanted organ. This inflammation initially affects the biliary epithelium and subsequently the hepatocytes. Histological examination of liver tissue reveals a mixed cellular infiltrate composed predominantly of lymphocytes localized to the portal tracts and the central vein areas. Early histological changes, however, are often nonspecific and may be confused with recurrent hepatitis C infection. With time or severity, damage to bile ducts and to the central vein endothelium occurs.

Clinical Findings and Diagnosis

Most cases of AR are manifested as asymptomatic biochemical changes. These biochemical changes almost always occur before clinical signs and symptoms and are a manifestation of the cholestatic inflammation first affecting the biliary epithelium and subsequently the hepatocytes. Accordingly, subtle increases can first be seen in bilirubin, alkaline phosphatase, and γ-glutamyltransferase levels. Shortly thereafter, a rise in serum transaminase levels is seen. Although no absolute biochemical parameters can currently be used to define an episode of rejection, there are some experimental methods currently being evaluated. MicroRNAs have emerged as important regulators of immune function. Although currently not a proven method to detect acute rejection, early results show that detection of certain microRNAs in the serum can be linked to episodes of acute rejection, specifically, miR-122, miR-101, and let-7b, which are preferentially expressed by liver. These circulating microRNAs have been shown to be extremely stable and protected from RNase-mediated degradation. As examination of tissue biopsy specimens is costly, risky, and invasive, these microRNAs might prove to be beneficial in the future. Currently clinicians base their suspicions on a relative rise in liver function tests, the amount of time elapsed since transplantation, and the underlying disease.

Clinical findings tend to appear later in the course of rejection and reflect more significant injury to the graft. These signs and symptoms can include fatigue, fever, malaise, abdominal pain, decreased bile output, increasing ascites, and hepatomegaly ( Table 93-1 ). If bile is available for inspection through a T tube, it may be lighter in color and less viscous than usual.

The gold standard for the diagnosis of AR is percutaneous liver biopsy. Although invasive, this diagnostic procedure affords the greatest sensitivity and specificity in the diagnosis of AR. However, percutaneous liver biopsy is not without risk. Although bleeding is the most common complication, serious hemorrhagic complications can occur in 0.06% to 0.35% of cases. Other complications include pneumothorax, hemobilia, peritonitis, perforated viscus, and the development of an arteriovenous fistula. The risk for sustaining any of these complications is about 2%, and the risk for death hovers between 0.009% and 0.1%. Because the most common complication is bleeding, clinicians should attempt to correct any coagulopathy before performing this procedure.

On occasion, despite percutaneous biopsy, the diagnosis of rejection can remain in doubt, particularly because other causes of acute inflammation in the liver can be confused with AR. Most frequently, cytomegalovirus (CMV) infection and recurrence or persistence of hepatitis C make the diagnosis of AR quite difficult. Treating presumptive AR with increased immunosuppression can be devastating if, in fact, the actual diagnosis is hepatitis C. In these situations, repeat biopsy before treatment may be of help.

Protocol biopsies in the postoperative period are advocated in some centers as a way to treat rejection more effectively. By increasing sensitivity and specificity, protocol biopsies allow one to diagnose AR while it is still at the subclinical level. The underlying assumption in performing protocol biopsies is that early diagnosis and treatment of AR may be of benefit. Although no multicenter randomized controlled trial has ever been conducted, some centers treat based on histological criteria without biochemical evidence of graft dysfunction. These preemptive efforts have not been conclusively shown to be beneficial in liver transplant patients. Recent data suggest that this approach may be overly cautious and may lead to overtreatment.

From 15 studies of patients undergoing protocol biopsies, 302 patients were identified who had histological, but no biochemical evidence of rejection. Of these patients, only 36 subsequently experienced clinical rejection, 7 of which were steroid resistant and 9 of which were CR.

Despite the sensitivity and specificity of percutaneous liver biopsy, the procedure remains an invasive one with all the attendant risk for morbidity and mortality. Numerous investigators have searched for safer and less invasive diagnostic methods, as well as ones that would potentially diagnose AR at an earlier time point. Some centers have tried fine-needle aspiration. However, no large randomized controlled studies have examined this issue in liver transplant patients. Although the smaller bore of the fine needle may result in less trauma, this risk may be amplified by the need to perform multiple passes because insufficient quantity of tissue may be obtained with a single pass. In addition, the subtle changes of early AR may be missed because of sampling error, with a subsequent delay in diagnosis and treatment. “Overcalling” AR can also occur and result in a needless increase in immunosuppression with its attendant morbidity. Moreover, in cases in which CR is also being considered, fine-needle aspiration has no diagnostic role because a core needle biopsy is required to more accurately assess hepatic histological characteristics. A number of noninvasive methods have been examined in the laboratory to detect acute rejection early. The newest of these is microRNA profiling. MicroRNAs are a class of small nonencoding RNAs that are important regulators of gene expression.

Variable expression of microRNAs has been linked to pathological conditions in humans including malignant, metabolic, cardiovascular, and autoimmune diseases. Specifically, hepatocyte-derived microRNAs (HDmiRs) have been studied in the laboratory as potential markers for hepatocyte injury. It has been shown that one can detect a rise in serum HDmiR levels before a measurable rise in transaminase levels is observed. In the setting of acute rejection, the serum HDmiRs are elevated by up to 20-fold. These data suggest that this may be an emerging modality to detect liver injury and acute rejection early, even before a transaminitis occurs. Another new method involves donor circulating cell-free DNA in the serum of the recipient. Initially used in cardiac transplantation, it has recently been studied in liver recipients. In one study it was shown that donor cell-free DNA levels are 90% immediately after transplant and fall to less than 15% by day 10. In patients that had biopsy-proven rejection, these levels rose up to 50%. Biliary aspirates from indwelling T tubes, cytokine profiling, histological analysis of cellular aspirates, cellular Fas ligand analysis, and isotope scanning, along with other noninvasive technologies, have also been investigated in the laboratory and the clinic. All these methods, however, have been hampered by suboptimal sensitivity or specificity because of similarities between AR and other forms of acute inflammation. Recently quantification of granzyme expression in fine-needle aspirates has been investigated with moderate success. Although this technique has been used with a high level of accuracy for the diagnosis of AR in renal allografts, the presence of CMV infection appears to decrease the accuracy of this method in liver transplants.

Thus, despite numerous attempts to use noninvasive techniques over the past 20 years, none of these diagnostic methods have been validated or meet the sensitivity and specificity of percutaneous liver biopsy. Core needle liver biopsy remains the procedure of choice for the diagnosis of AR.

Before initiating treatment of AR, the clinician should consider other diagnoses ( Table 93-2 ). Potentially confounding differential diagnoses are often time dependent. Within the first month after transplantation, other possible diagnoses include preservation injury, early hepatic artery thrombosis, portal vein thrombosis, biliary leak, CMV infection, and intra-abdominal sepsis. After the early postoperative period until approximately a year after transplantation, recurrence or persistence of hepatitis C, drug toxicity, biliary strictures, and opportunistic infections may be confused with AR. After the first year, late mimics of AR can include late hepatic artery thrombosis, recurrent autoimmune hepatitis, recurrent primary sclerosing cholangitis, CR, and posttransplantation lymphoproliferative disorder (PTLD). Once treatment has started, every effort should be made to also eliminate or decrease potentially aggravating factors ( Table 93-3 ). Conditions such as diarrhea and drug interactions can lead to suboptimal levels of immunosuppression and hamper effective therapy and resolution of AR. As many as 50% of patients suffering from AR may have one or more of these aggravating factors.

Treatment

After the diagnosis of AR is established, therapy first consists of optimization of maintenance immunosuppression ( Fig. 93-1 ). The dosage of the patient’s mainstay calcineurin inhibitor can be increased, and other immunosuppressive agents such as mycophenolate mofetil, azathioprine, and rapamycin, if not currently being taken, can also be added. If the AR does not appear to respond to these initial measures, intravenous and high-dose steroid therapy can be implemented. Antilymphocytic antibodies and newer agents such as interleukin-2 receptor antagonists (IL-2RAs) are reserved for rarer cases of steroid-resistant AR or for the occasional severe case of untreated AR.

Such treatment is associated with a constellation of potential toxicities, including a risk for excessive immunosuppression and infection. Of concern as well is the risk for PTLD. These risks and benefits must be considered in each individual case. Although treatment of the AR episode will ameliorate or reduce symptoms, liver grafts generally do not suffer from long-term sequelae as a result of early episodes of AR. Unlike renal allografts, there is no direct correlation between the number of AR episodes and long-term graft or patient survival. (There is some controversy about these statements.) Generally, however, once AR is diagnosed from histological and clinical or biochemical evidence, the benefits of treatment outweigh the risks.

After AR is adequately treated, maintenance immunosuppression may be changed or increased for several months to reduce the probability of relapse. For patients being treated with cyclosporine-based immunosuppressive regimens, changing maintenance therapy to tacrolimus may also help treat and decrease the incidence of acute and chronic rejection. Other immunosuppressive agents may be added, and patients previously receiving dual therapy may temporarily be maintained on triple or even quadruple therapy.

Pediatric patients in whom AR is diagnosed have very specific concerns in their treatment, and these issues are reviewed later.