Over the past decades, survival rates after pediatric liver transplantation have increased from about 76% to over 90% more than 10 years after transplantation (Tx). Follow-up care of liver-transplanted children includes both monitoring of graft functioning and monitoring for associated comorbidities. Monitoring the allograft is comprised of a number of features, ranging from the innate hepatic metabolic and synthetic functions to late surgical complications, such as biliary and vascular obstruction, and to transplant-specific, immune-mediated complications, such as graft fibrosis, graft hepatitis, and acute or chronic rejection.
Protocols for follow-up care are usually comprised of regular evaluation of routine biochemical liver function tests (e.g., transaminase levels, parameters of cholestasis, parameters of hepatic synthetic function), as well as ultrasound examinations to monitor hepatic vasculature and hepatic morphology.
Hepatic synthetic and metabolic function can easily be monitored by routine biochemical parameters. Biliary obstruction will usually manifest first by biochemical changes such as raised bilirubin and gamma-glutamyltransferase (GGT) levels and, in advanced stages, might cause clinical signs and symptoms such as jaundice, pale stools, and pruritus and possibly dilation of bile ducts detectable on ultrasound. Acute rejection will be suspected in cases of raised transaminase levels, but confirmation requires liver biopsy.
Chronic graft changes such as graft hepatitis, graft fibrosis, and chronic rejection might be present even in the presence of normal or near-normal biochemical liver function tests, rendering early stages almost unrecognizable for routine follow-up. Again, their detection requires liver biopsy. The use of protocol biopsies at regular intervals after pLTx has not yet been universally endorsed by pediatric liver transplantation (pLTx) centers. Published series of late protocol biopsies after pLTx indicate a prevalence of graft fibrosis of 70% more than 10 years after pLTx, and a prevalence of chronic inflammation (graft hepatitis) of about 60%.
The main aim of integrating serological biomarkers into the routine monitoring of the liver allograft would be to identify graft changes and dysfunctions that cannot be detected by routine biochemical parameters. Biomarkers that could potentially obviate the need for liver biopsy in (pediatric) liver transplantation have been evaluated for the diagnosis of acute cellular rejection and for the diagnosis of graft fibrosis.
Graft fibrosis can be detected in up to 30% of grafts 1 year after transplantation and increases in frequency over the years. Graft fibrosis is seen in 60% to 80% of grafts 10 years and more after pLTx. Fibrosis is accompanied by some degree of inflammation in more than 50% of cases.
Over the past decade, a number of approaches have been developed and tested to facilitate the noninvasive diagnosis of hepatic fibrosis without recurrence to liver biopsy. These approaches include imaging—namely, ultrasound and magnetic resonance imaging (MRI)-based techniques—and serological markers of fibrosis. One unifying feature for all these approaches is that they were primarily developed in adult hepatology, predominantly in the context of adults with hepatitis C, and in a pre-transplantation setting. Transfer of knowledge to the context of pLTx has been slow, thereby limiting the applicability of results for pediatric transplant hepatologists further (see Figure 30.1 ).
Serological biomarkers for the diagnosis of hepatic graft fibrosis can be divided into markers of hepatic fibrogenesis and extracellular matrix (ECM) deposition and markers that indicate the functional sequelae of hepatic fibrosis ( Table 30.1 ).
|Test||Adult LTx||Pediatric CLD||Pediatric LTx||Numbers of Patients in Biopsy-Controlled Studies after pLTx||Diagnostic Value in pLTx|
|Hyaluronic acid||+||+||+||n = 28 8||Rule in, rule out fibrosis|
|ELF||+||++||+||n = 23 14||No|
|FibroSure||++||+||+||n = 28 14 |
n = 5 (8 biopsies in 5 patients) 16
|APRI||+||++||++||n = 39 20 |
n = 36 21
n = 38 22
n = 30 23
|FibroScan||++++||++++||++||n = 36 14 |
n = 36 21
n = 16 30
|Serial follow-up, not as a single shot|
|ARFI||++++||+++||++||n = 23 35 |
n = 30 23
|Diagnose, or rule out significant fibrosis; not for lower degrees of fibrosis|
Markers that build on circulating levels of ECM components and factors of ECM degradation are the FibroSure (FS) test and enhanced liver fibrosis (ELF) score:
The FS test uses the ECM component alpha-2-macroglobulin with apolipoprotein A1, alanine aminotransferase (ALT), GGT, bilirubin, and haptoglobin.
The ELF score combines the ECM components of hyaluronic acid and the amino terminal peptide of type III procollagen with a factor of ECM degradation, tissue inhibitor of metalloproteinase-1.
In addition, the ECM component hyaluronic acid has been individually assessed as a marker for hepatic fibrosis.
To date, hyaluronic acid (HA) has been assessed in one small cohort of 40 children after liver transplantation, the majority of whom underwent protocol biopsies. Fibrosis was present in only 12.5% of the studied transplantation cohort, thereby limiting the possibility to calculate meaningful cutoff values. Healthy children universally displayed HA levels below 30 ng/mL. A HA level of 50 ng/mL appears to be a sensible cutoff value to rule out significant fibrosis (Ishak staging system, ≥ F3). A HA level of 200 ng/mL lacks sensitivity but is highly specific for the diagnosis of significant fibrosis.
Enhanced Liver Fibrosis Score
The ELF score is one of the first serological fibrosis biomarkers that became commercially available. It was originally developed in the context of adult nonalcoholic fatty liver disease (NAFLD) or nonalcoholic steatohepatitis (NASH). The ELF score has been extensively evaluated in nontransplanted children and adolescents with NASH/NAFLD, showing good correlation with fibrosis stage and good diagnostic performance for the diagnosis of significant fibrosis. Published cutoff values for the diagnosis of fibrosis in children with NASH/NAFLD differ from those published for the commercial application in adults. In transplanted children, ELF scores did not correlate with the degree of fibrosis and could not discriminate among different stages of hepatic fibrosis. Cutoffs from nontransplanted children with NAFLD therefore cannot be automatically applied. Liver-transplanted children with normal grafts on liver biopsy have higher ELF scores than healthy children. Increased ELF scores have also been described in liver-transplanted children with elevated versus normal transaminase levels, in children with a history of venous outflow obstruction, and in children receiving a mammalian target of rapamycin inhibitor as a primary immunosuppressant. Potential alterations in ECM turnover after liver transplantation as an explanation for these observations still remain a topic of speculation.
In adults, ELF post-transplantation has been evaluated in the context of patients with hepatitis C virus (HCV) infection. At a threshold of 10.5, the ELF score could predict disease recurrence with an accuracy of 74%.
The FibroSure (FS) test was developed for the noninvasive diagnosis of hepatic fibrosis in adults with HCV infection. In this setting, it has been shown to have excellent sensitivity and specificity before and after liver transplantation (HCV). Comparison of the FS was usually made with histological grading of fibrosis according to the METAVIR score, which again was specifically developed for patients with HCV infection.
In children after liver transplantation, FS has been evaluated in children undergoing indication biopsies, but not in conjunction with protocol biopsies.
Correlation with the degree of fibrosis in indication biopsies is poor. In general, FS tends to underestimate fibrosis. Comparison with standard values obtained in healthy children reveals FS to be within the normal range in 62% of patients with an underestimation of fibrosis. Serial evaluation suggests that an increase in FS over time might be associated with the occurrence of clinical events, such as cholestasis or ductopenic rejection. In liver-transplanted adults without hepatitis C, sensitivity for diagnosing advanced fibrosis with a cutoff of 0.5 is poor (33.3%), although specificity is high (90.9%).
Serological biomarkers that indicate functional sequelae of hepatic fibrosis include the APRI and fibrosis score 4 (FIB-4). Both markers use routine biochemical or hematological blood tests such as AST, ALT, and platelet count, which in turn reflect hepatocellular damage or portal hypertension.
Fibrosis Score 4
FIB-4 combines the patient’s age, AST, ALT, and platelet count into an algorithm that was primarily developed for adults with HCV infection. An FIB-4 score of 3.3 and more at 1 year after liver transplantation was associated with an increased risk of death and graft loss in liver-transplanted adults. Similarly, an increase in the FIB-4 score of more than 1.4 from 12 to 24 months after transplantation carried an increased risk of death and graft loss. To date, an evaluation of FIB-4 in children after liver transplantation has not been performed.
Aspartate Aminotransferase Platelet Ratio Index
The APRI is calculated using the formula AST [U/L]/(AST upper limit of normal) × 100/platelet count [10 9 /L]. A comparison of APRI performance with histological assessment of graft fibrosis has been made using the Ishak score, the liver allograft fibrosis (LAF) score, and the Batts-Ludwig score. Results are equivocal. Cutoff values for the diagnosis of significant fibrosis (Ishak ≥ F3, Batts-Ludwig ≥ F2) range from 0.4 to 0.45. In contrast, some studies describe no or only very little diagnostic value for the APRI in the context of pLTx.
In summary, currently available serological markers for liver allograft fibrosis in children should be used with caution because universally applicable cutoffs for diagnosing or ruling out fibrosis are still lacking.
Imaging techniques to diagnose and quantify hepatic fibrosis include transient elastography using the FibroScan, acoustic radiation force imaging (ARFI), and MRI elastography. All these techniques have been extensively investigated in pediatric liver disease in a pre-transplantation setting, but only limited experience is available for children after liver transplantation.
Transient Elastography by FibroScan
Transient elastography (TE) by FibroScan uses the propagation velocity of a pressure-induced shear wave within the liver tissue as a surrogate marker for liver stiffness and hence liver fibrosis. Different sizes of measuring probes are available for small and older children, and care needs to be taken to pick the appropriate sized probe because results may vary according to which probe is used. Normal liver stiffness values for healthy children range from 3 to 6 (median 4.5) kPa, with an upper limit of normal of 6.47 kPa. Application is limited in very young children because excessive movement during the examination and small intercostal spaces may result in invalid or falsely increased liver stiffness values. Liver stiffness values as measured by TE correlate well with histological degrees of liver fibrosis in a variety of pediatric liver diseases. However, TE has yielded only imprecise results in mild to moderate degrees of fibrosis, with considerable overlap of values among fibrosis classes. Its main application, therefore, appears to be to rule out fibrosis or rule in advanced fibrosis and cirrhosis. Cutoff values for advanced fibrosis have been reported as 8.6 kPa (METAVIR ≥ F3). The presence of inflammation can increase liver stiffness, leading to an overestimation of hepatic fibrosis. In children with no or mild fibrosis (METAVIR, F0–F2), the proportion of patients with liver stiffness values above the cutoff of 8.6 kPa increased up to 60%, depending on ALT levels as a result of inflammation.
In the context of pLTx, several technical issues need to be considered. The prescribed approach of obtaining TE measurements in the anterior axillary line frequently cannot be applied in patients with reduced size or left split grafts. The technical success rate is generally lower in left split recipients compared with full-size graft recipients owing to smaller grafts, limited intercostal access, and midline position of the grafts.
Comparison of TE and histological staging of fibrosis in liver-transplanted children has been made both for the METAVIR score and the Ishak score. For METAVIR, a cutoff to detect significant fibrosis (≥ F2) has been described at 5.6 kPa, with a sensitivity of 75% and a specificity of 95.8%, and at 6.9 kPA with a sensitivity of 100% and specificity of 80%. For severe fibrosis (≥ F3), a cutoff value has been described at 10.5 kPa, with 60% sensitivity and 75% specificity.
For the Ishak score, cutoff values to diagnose the presence of any fibrosis (≥ F1), moderate fibrosis (≥ F2), or advanced fibrosis (≥ F3) are given at values of 7.2, 9.3, and 12.7 kPa, respectively.
Liver stiffness values in transplanted children without fibrosis tend to be higher than reference values obtained in healthy children. Care should therefore be taken not to apply reference values and upper limits of normal without scrutiny. Also, at identical degrees of fibrosis, liver-transplanted children show higher liver stiffness values than children with chronic liver disease of varying causes. Values might be higher in split graft recipients compared with full-size graft recipients owing to altered anatomy, position, and scar tissue. The presence of rejection and of biliary or venous outflow obstruction is an additional confounder that might lead to an overestimation of fibrosis. Given the wealth of confounding factors, the use of TE to assess liver fibrosis in liver-transplanted children needs to take into account the clinical situation, anatomy, and medical history. Determination of an individual baseline for each patient with follow-up over time might be more useful than a one-off measurement.
Acoustic Radiation Force Imaging
ARFI is an ultrasound-based elastography technique that uses a short-duration acoustic pulse to produce a mechanical stimulus, thereby inducing a shear wave within the liver tissue, the propagation velocity of which is determined by ultrasound. Its main advantage over TE is the possibility to determine the region of interest for elastography assessment freely.
Normal values for shear wave velocity obtained in healthy children are 1.07 to 1.12 m/s ± 0.10 m/s. In healthy children, shear wave velocity (SWV) values are higher in the left lobe compared with the right lobe of the liver and decrease with increasing depth of the measurement.
Comparison with liver histology has shown good correlation of ARFI SWV values with the degree of hepatic fibrosis in children before and after liver transplantation. In children with biliary atresia, ARFI SWV values greater than 2 m/s, if measured repeatedly, are considered as predictors for the need for liver transplantation.
In children after liver transplantation, higher degrees of fibrosis (METAVIR of F3 or F4) are clearly associated with increased SWV values. Distinction of the absence of fibrosis from low-grade fibrosis (F0 vs. F1) is more difficult because some transplanted patients without fibrosis display SWV values in a range that is normally considered pathological. A SWV cutoff of 1.34 m/s is given for the diagnosis of any fibrosis (> F0) according to the METAVIR scoring system, whereas a cutoff of 2 m/s predicts fibrosis over F2 with a sensitivity of 100%. Cutoff values vary according to the scoring system used to assess fibrosis ( Table 30.2 ). Similar to TE, the presence of hepatic inflammation and obstructive cholestasis can lead to an overestimation of liver stiffness. There is as yet no certainty about the influence of different types of graft (left split, right split, whole liver) on ARFI results.