Chronic liver disease is a major cause of morbidity and mortality worldwide. In the United States alone, nearly 40,000 patients progress to liver failure and death annually. Liver transplantation is currently the standard treatment for various forms of severe liver disease, including acute liver failure or end-stage liver disease due to any etiology, as well as selected metabolic and neoplastic conditions.
Chronic liver diseases are often silent until cirrhosis with clinical decompensation ensues. Events that define decompensation include ascites, hepatic encephalopathy, variceal bleeding, and nonobstructive jaundice, which are often triggered by precipitating factors such as bacterial infections, portal vein thrombosis, surgery, or hepatocellular carcinoma. Acute liver failure, on the other hand, is characterized by jaundice, coagulopathy, and the development of hepatic encephalopathy within 8 weeks of the onset of symptoms. Subacute liver failure has a similar presentation, except hepatic encephalopathy develops later, between 8 and 24 weeks of the onset of jaundice.
For both decompensated cirrhosis and acute/subacute liver failure, liver transplantation represents a lifesaving treatment. Unfortunately, however, liver allografts remain a scarce resource and, despite great improvements in the organ allocation system, approximately 2,500 to 3,000 patients are removed from the waiting list every year in the United States due to death or poor overall clinical status (Organ Procurement and Transplantation Network [OPTN] data, 2008 to 2015).
The organ transplant system in the United States is currently managed by a nonprofit organization—the United Network for Organ Sharing (UNOS). In 2002, the UNOS adopted the model of end-stage liver disease (MELD) score as the organ allocation system. The MELD score is an objective, laboratory-based measurement which has been validated in predicting 3-month waiting list mortality.1
For children 12 years of age and under, the pediatric end-stage liver disease (PELD) score replaces MELD. In addition, some conditions are listed for liver transplant as UNOS status 1A, which takes priority over the MELD allocation score: acute liver failure, primary nonfunction of transplanted livers, hepatic artery thrombosis, and Wilson disease. The benefit of transplantation should be weighed against posttransplant morbidity
and mortality, being reserved for patients with considerable decline in their quality of life and/or high mortality without transplant.
Figure 20.1 Etiology of liver disease in patients undergoing liver transplantation in the United States From UNOS database ; 7,127 transplanted patients.
According to the UNOS database, 7,127 liver transplants were performed in the United States in 2015 and hepatocellular carcinoma was the most common indication for liver transplantation (often in the setting of cirrhosis due to various etiologies), followed by hepatitis C-related cirrhosis, alcoholic liver disease, and nonalcoholic fatty liver disease (Fig. 20.1
). Acute liver failure accounts for less than 10% of all transplants performed in the United States annually, with acetaminophen overdose being responsible for almost half of these cases.
20.2 EVALUATION OF DONOR LIVER BIOPSY
Liver biopsy is an important adjunct tool in the evaluation of potential allografts. Its main role is to confirm the quality/viability of the donor organ and exclude features that would either contraindicate transplan-tation or increase the likelihood of various adverse short and long-term outcomes. As a result of organ shortage, increasing utilization of allografts from suboptimal donors has occurred in recent years, leading to a higher proportion of donor livers requiring pretransplantation frozen section evaluation. The so-called “extended donor criteria” includes several established risk factors for posttransplantation organ dysfunction and failure, such as old age (>60 years), hypernatremia (>155 mEq/L), macrovesicular steatosis >40%, cold ischemia time >12 hours, partial liver allografts, and donation after cardiac death (DCD) donors.2
Suboptimal livers also include allografts from donors with prolonged hemodynamic instability, use of vasopressors, chronic viral
hepatitis (B or C), as well as the presence of mass lesions or significant fibrosis. Although most risk factors are defined clinically, the main role of pathologists in the pretransplantation frozen section evaluation is to assess steatosis and exclude any miscellaneous findings (significant inflammation, necrosis, granulomatous processes, fibrosis, neoplasms, etc.) which may not have been detected on clinical evaluation.
In order to avoid diagnostic errors, the liver sample being submitted for frozen section by the transplant surgeon must be obtained and handled properly. Although a single biopsy is generally adequate (either needle core or wedge biopsy), additional samples should be obtained if the appearance of the liver is grossly heterogeneous. A large-core biopsy (i.e., away from the liver capsule) is preferred if there is concern for fibrosis, because subcapsular wedges can show nonspecific subcapsular fibrosis that is suboptimal for staging purposes.
The fresh tissue should ideally be carefully wrapped in a paper towel or gauze soaked in preservation solution and transported to the pathology lab immediately. Prolonged exposure to normal saline can cause significant histologic artifact, characterized by cellular discohesiveness and pyknosis-like chances of hepatocytes, mimicking necrosis (Fig. 20.2
). Dry absorbent materials and sample compression can result in tissue dehydration and absorption of fat, potentially resulting in significant histologic changes and underestimation of steatosis. Tissue freezing itself frequently causes small cytoplasmic vacuoles in hepatocytes that are essentially indistinguishable from microvesicular steatosis on hematoxylin and eosin (H&E)-stained sections. Oil red O or other fat stains may aid in this distinction, but require experience for proper interpretation.
The importance of donor steatosis as a risk factor in the setting of liver transplantation is well established and consistently observed in several clinical and experimental studies.3
Macrovesicular steatosis, defined histologically as one or a few large fat droplets that displace the hepatocyte nucleus to the periphery of the cell, represents the main finding in this context and has been associated with various adverse outcomes during the early transplantation period, including primary graft dysfunction and nonfunction, as well as prolonged intensive care unit and hospital stay.7
This is thought to be related in part to the
displacement of fat droplets from the cytoplasm of hepatocytes into sinusoids (because of mechanical factors and/or ischemia), leading to sinusoidal obstruction, lipid peroxidation and formation of free radicals upon reperfusion, and endothelial injury.7
Examination of failed allografts owing to primary nonfunction and early posttransplant biopsies of steatotic allografts often show large, coalescent fat droplets within hepatic sinusoids—referred to as lipopeliosis (Fig. 20.3
)—which illustrates one of the primary pathogenic mechanisms in this scenario.
Figure 20.2 Saline preservation artifact. Hepatocytes appear discohesive, with glassy cytoplasm and piknotic-appearing nuclei, resembling early necrosis.
Figure 20.3 Lipopeliosis. Large coalescent fat globules extruded from hepatocytes are seen within congested sinusoids.
Some confusion exists with regard to the nomenclature for the types of steatosis, making interpretation of the literature and communication of pathological findings quite challenging in certain situations. For example, the term microvesicular steatosis
is often inappropriately used in the literature to refer to small-droplet steatosis, which virtually always coexist with typical large-droplet steatosis (as seen in nonalcoholic fatty liver disease, for example). In contrast, true microvesicular steatosis is characterized by very small lipid droplets which impart a “bubbly” appearance to the hepatocyte cytoplasm and that are often difficult to recognize on H&E stains (Fig. 20.4
). Both small- and large-droplet fat are currently considered part of the spectrum of macrovesicular steatosis. Microvesicular steatosis is typically seen in specific clinical contexts that includes Reye syndrome, acute fatty liver of pregnancy, and toxicity by certain medications, and is thought to be related to mitochondrial toxicity. Unfortunately, in the context of liver transplantation, we must keep in mind that the terms microvesicular steatosis
and small-droplet steatosis
have been used interchangeably by some authors. See Chapter 16
for further discussion.
Although the specific method of interpretation differs slightly among pathologists, macrovesicular steatosis should be evaluated quantitatively (as a percentage). Macrovesicular steatosis is estimated as the percentage area of the hepatic lobules occupied by large-droplet steatosis (Fig. 20.5
). Although no specific guidelines exist, livers with 30% large-droplet steatosis or less are considered adequate, without significant increased risk of adverse outcome, whereas allografts showing >60% are generally considered inadequate and, therefore, are not used for transplantation. Allografts with moderate (30% to 60%) large-droplet steatosis may or may not be utilized, depending on local
preferences and additional risk factors related to both donor and recipient. Finally, digital/computerized tools are available to assist in this interpretation but yield values that are significantly different (approximately half) compared to visual (i.e., “eyeball”) estimates by pathologists.10
Therefore, validation of values yielded by digital analysis and correlation with outcomes is necessary before this technique is applied during liver allograft procurement.
Figure 20.4 Microvesicular steatosis versus small-droplet steatosis. The left side of the panel shows true microvesicular steatosis, with very small fat droplets that impart a “bubbly” appearance to the cytoplasm. In contrast, small-droplet steatosis (seen on the right side of the panel in conjunction with large-droplet steatosis) is characterized by larger droplets than microvesicular steatosis, with greater variation in droplet size.
Figure 20.5 Semi-quantitative assessment of steatosis. Mild (5% to 33%), moderate (33% to 66%), and severe (>66%) steatosis are illustrated in the left, middle, and right parts of the panel, respectively.
20.6 SMALL-FOR-SIZE SYNDROME
The typical clinical scenario in portal hyperperfusion/small-for-size syndrome is unexplained cholestasis, coagulopathy, and ascites in the early posttransplant period, most commonly occurring in the setting of reduced-size or living donor allografts. The clinical features are not specific for this entity, mimicking several other posttransplant complications, but liver transplant surgeons are often aware of this possibility, being alerted when there is splanchnic congestion and portal hypertension upon revascularization of the allograft.
Etiology and pathogenesis
Portal hyperperfusion/small-for-size syndrome is thought to be related primarily to portal hyperperfusion, which leads to decreased levels of adenosine (a vasodilator) from the hepatic circulation (i.e., “adenosine washout”), which in turn causes hepatic artery vasoconstriction, with subsequent low arterial flow, thrombosis, and ischemic cholangitis.
The histologic changes in portal hyperperfusion/small-for-size syndrome are rather subtle and nonspecific and a high degree of suspicion is required for this diagnosis. In early phases (starting minutes after transplantation), there is denudation of the endothelium of portal veins and sinusoids as well as hemorrhage into portal connective tissue. Subsequently, there is regeneration of the endothelium, reactive endothelial cells, subendothelial edema, and, eventually, fibrointimal hyperplasia, luminal obliteration, and recanalization. Nodular regenerative hyperplasia is also seen, presumably owing to vascular flow abnormalities. The hepatic lobules show a constellation of nonspecific abnormalities, including centrilobular cholestasis, microvesicular steatosis, parenchymal atrophy, and necrosis. In severe cases, ischemic cholangitis may be seen.18
The prognosis of portal hyperperfusion/small-for-size syndrome will depend on the severity of the disease. Some cases are effectively treated by decreasing portal venous flow using octreotide, splenic artery ligation, or mesocaval shunts.
20.8 ACUTE CELLULAR REJECTION
In spite of significant improvements in immunosuppressive therapy and lower rates of acute cellular rejection after the introduction of current immunosuppressive agents, rejection remains an important cause of liver allograft dysfunction and graft loss.36
Compared to other solid organ allografts, however, the liver is significantly more resistant to the deleterious effects of cellular rejection, with most episodes being either subclinical (i.e., seen in protocol biopsies without laboratory evidence of liver dysfunction) or mild, without associated long-term organ damage following treatment.39
Clinical factors associated with an increased risk for acute cellular rejection include older allograft donor age (≥30 years), prolonged cold ischemic time (≥15 hours), “healthy” recipients (young age, low Child-Pugh score, normal creatinine, etc.), HLA-DR mismatch, and baseline autoimmune conditions (autoimmune hepatitis, primary biliary cirrhosis, primary sclerosing cholangitis, etc.). Lower rates of rejection have been associated with pretransplant alcoholic liver disease.41
Mild acute cellular rejection is often clinically asymptomatic, typically presenting with isolated mild elevations of the liver enzymes. In moderate to severe rejection, fever, enlargement and tenderness of the allograft, and decreased bile output may be noted, along with leukocytosis and eosinophilia in some cases. Although allograft rejection characteristically presents with a predominantly “cholestatic” pattern of liver enzyme elevations (i.e., predominant elevation of bilirubin, alkaline phosphatase, and γ-glutamyl transpeptidase, with relatively minor elevation of transaminases), nonselective enzyme elevation is also rather common. Clinical and laboratory findings overall lack both sensitivity and specificity for the diagnosis of acute cellular rejection and histopathologic examination of a biopsy sample is required to confirm the diagnosis.47
In practice, most cases of acute cellular rejection occur early in the posttranplantation period, most commonly in the first month. Late rejection (3-6 months posttransplantation or later) is less frequent and generally associated with suboptimal immunosuppression.
Etiology and pathogenesis
Although complex and still not entirely understood, the basic pathogenesis of acute cellular rejection is related to the recognition of alloantigens by antigen-presenting cells, with subsequent cytotoxic T-lymphocyte activation and cell-mediated injury to the allograft, primarily targeting interlobular bile ducts and hepatic microvasculature.
Histopathologic examination represents the gold standard for the diagnosis of acute cellular rejection and is based on a triad of features which include: (1) portal inflammation, (2) lymphocytic cholangitis/bile duct injury, and (3) endotheliitis (Table 20.2
). At least two of the components of this triad must be present for the diagnosis of acute cellular rejection according to the widely adopted Banff criteria for liver allograft rejection (Table 20.3
Although the composition and intensity of portal inflammation varies significantly from case to case, depending on various factors (mainly posttransplantation period, as discussed below), typical “rejection-type” infiltrates are characterized by a mixture of lymphocytes that often display an “activated” or “blast-like” morphology—with slightly larger nuclei and with more cytoplasm. Scattered eosinophils (which tend to be more prominent in patients treated with steroid-sparing or lymphocyte-depleting immunosuppressive regimens), neutrophils, macrophages, and plasma cells can also be seen (Fig. 20.9
). Interface activity is absent in mild cases but is seen in moderate and severe acute cellular rejection (Fig. 20.10
Recognition of interlobular bile duct inflammation and injury is critical in the diagnosis of acute cellular rejection. In its early/mild form, there is lymphocytic infiltration through the epithelial basement membrane and lymphocytes are seen among bile duct epithelial cells (Fig. 20.11
). Morphologic evidence of bile duct
epithelial injury subsequently appears in the form of nuclear enlargement, nuclear overlapping, occasional mitotic figures, and cytoplasmic vacuolization (Fig. 20.12
). In severe cases, luminal obliteration occurs and the small bile ducts may be difficult to recognize within the inflammatory infiltrate (in which case a keratin immunostain may help). With persistent or recurring bile duct injury, nuclear pleomorphism, hyperchromasia, dyspolarity, cytoplasmic eosinophilia, and atrophic appearance of ducts (collectively referred to as bile duct senescence
) can be seen and are generally thought to represent features of early chronic rejection.47
In cases with significant bile duct injury and destruction, hepatocanalicular cholestasis can be present (see “Differential diagnosis” below for further discussion).
Table 20.2 Banff Grading Criteria for Acute Cellular Rejection (Global Assessment)
Portal inflammatory infiltrate that fails to meet the criteria for the diagnosis of acute rejection
Rejection infiltrate in a minority of the triads, which is generally mild and confined to the portal spaces
Rejection infiltrate expanding most or all portal tracts
As above, with spillover of inflammatory cells into the periportal areas, with moderate to severe perivenular inflammation that extends into the hepatic parenchyma and is associated with perivenular liver cell necrosis
Adapted from Demetris AJ, Adeyi O, Bellamy COC, et al. Liver biopsy interpretation for causes of late liver allograft dysfunction. Hepatology. 2006;44(2):489-501, with permission from American Association for the Study of Liver Diseases. doi:10.1002/hep.21280.
Table 20.3 Banff Grading Criteria for Acute Cellular Rejection (Rejection Activity Index, RAI)
Mostly lymphocytic inflammation involving, but not noticeably expanding, a minority of the triads
Expansion of most or all of the portal tracts by a mixed inflammatory infiltrate containing lymphocytes with occasional blasts, neutrophils, and eosinophils
Marked expansion of most or all of the triads by a mixed infiltrate containing numerous blasts and eosinophils, with inflammatory spillover into the periportal parenchyma
Bile duct inflammation/damage
A minority of the ducts are cuffed and infiltrated by inflammatory cells and show only mild reactive changes such as increased nuclear/cytoplasmic ratio of the epithelial cells
Most or all of the ducts infiltrated by inflammatory cells. More than an occasional duct shows degenerative changes such as nuclear pleomorphism, disordered polarity and cytoplasmic vacuolization of the epithelium
As above for 2, with most or all of the ducts showing degenerative changes or focal lumenal disruption
Venous endothelial inflammation (endotheliitis)
Subendothelial lymphocytic infiltration involving some, but not a majority of the portal and/or hepatic venules
Subendotheilal infiltration involving most or all of the portal and/or hepatic venules
As above for 2, with moderate or severe perivenular inflammation that extends into the perivenular parenchyma and is associated with perivenular hepatocyte necrosis
Adapted from Demetris AJ, Adeyi O, Bellamy COC, et al. Liver biopsy interpretation for causes of late liver allograft dysfunction. Hepatology. 2006;44(2):489-501, with permission from American Association for the Study of Liver Diseases. doi:10.1002/hep.21280.
Figure 20.9 Acute cellular rejection, inflammation. Mixed portal inflammation with relatively large, “activated” lymphocytes and eosinophils.
Figure 20.10 Acute cellular rejection, interface activity. A case of moderate to severe rejection showing interface activity.
Finally, endotheliitis—the third component of the acute cellular rejection triad—is characterized
by lymphocytic inflammation targeting the vascular endothelium of portal veins, central veins or, less commonly, hepatic arteries. Endotheliitis is histologically characterized by lymphocyte attachment to endothelial cells, generally associated with endothelial injury/reactive changes, or by the presence of subendothelial lymphocytes (subendotheliitis) causing detachment, or “lifting,” of endothelial cells from the vascular basement membrane (Fig. 20.13
). Portal vein branches are the most commonly affected structures, whereas central vein endotheliitis (often accompanied by centrilobular inflammation and injury—i.e., central perivenulitis) is more common in moderate to severe rejection or in late-occurring rejection (Fig. 20.14
). Inflammatory or necrotizing arteritis is identified in rare instances of severe rejection, typically involving larger arteries near the hepatic hilum (therefore, not usually present in biopsy samples).47
Figure 20.11 Acute cellular rejection, bile duct injury. Mild bile duct injury with subtle epithelial changes and focal lymphocytic infiltration.
Figure 20.12 Acute cellular rejection, bile duct injury. More significant bile duct injury is seen in this case, which shows prominent epithelial injury and partial bile duct destruction.
Figure 20.13 Acute cellular rejection, portal vein endotheliitis.
Figure 20.14 Acute cellular rejection, central vein endotheliitis.
In the most common form of acute cellular rejection, occurring in the initial 1 to 3 months posttransplantation, the histologic findings tend to be “typical,” as described above, with lymphocytic portal inflammatory infiltrates, portal vein endotheliitis, and readily recognizable bile duct injury, with interface activity and central vein endothelialitis/central perivenulitis developing as rejection progresses toward the moderate to severe end of the spectrum. The diagnosis of acute cellular rejection during this period is usually obvious. In the late posttransplantation period (3 to 6 months or later), however, variant histologic patterns
of rejection become more prevalent and may cause diagnostic difficulties52
(described in detail below).
Finally, moderate to excellent interobserver agreement has been reported51
for the various histologic criteria of acute cellular rejection: portal inflammation (κ= 0.86 to 0.88), subendothelial inflammation (κ = 0.39 to 0.63), and bile duct damage (κ = 0.42 to 0.49). The interobserver reproducibility for the overall diagnosis of acute cellular rejection in this study was good (κ = 0.50 to 0.62 in 2 separate interpretations), whereas the intraobserver reproducibility was good to excellent among five participating pathologists (κ = 0.53 to 0.89).
Several variant histologic patterns are recognized in liver allografts and are thought to represent forms of alloimmune attack to the liver (i.e., rejection), all occurring predominantly in the late posttransplantation period.
Plasma cell-rich hepatitis
This term refers to a chronic hepatitis with markedly increased plasma cells, typically with prominent interface and lobular activity, resembling autoimmune hepatitis in native livers (Fig. 20.15
). When occurring in patients with underlying autoimmune liver diseases (autoimmune hepatitis, primary biliary cirrhosis, primary sclerosing cholangitis, etc.) in their native livers, this is often considered to represent either recurrent or de novo autoimmune/alloimmune hepatitis. One study also identified an IgG4-rich subpopulation within cases of plasma cell hepatitis, which was associated with more aggressive histology and high rates of response to increased immunosuppression.53
Figure 20.15 Acute cellular rejection, plasma cell-rich variant. Clusters of plasma cells are seen within this portal tract. Also notice the significant amount of bile duct injury and occasional eosinophils.
The nomenclature (autoimmune-like hepatitis54
and posttransplant plasma cell hepatitis55
) and the precise etiology in patients with concurrent recurrent hepatitis C has been a subject of debate. The prevailing evidence is that this pattern represents an alloimmune/rejection phenomenon, rather than a variant pattern of recurrent hepatitis C, as evidenced by an increased frequency of antinuclear antibodies and other autoantibodies, association with subtherapeutic levels of immunosuppression, and a trend toward a more favorable prognosis in response to increased immunosuppression. This pattern has been associated with progressive disease and poor prognosis.54
At Mayo clinic, our general approach in these posttransplant plasma cell-rich cases is to use the term plasma cell hepatitis, followed by grading and staging, favoring this pattern to represent an alloimmune phenomenon/rejection variant, noting that this pattern can sometimes be associated with an aggressive course.
Isolated central perivenulitis
is a term recognized by the Banff Working Party, which refers to centrilobular changes that include central vein endotheliitis, perivenular inflammation, and perivenular hepatocyte injury/dropout (Fig. 20.16
). In allografts, central venulitis is most often associated with portal-based acute cellular rejection and commonly seen in the context of moderate to severe rejection. However, central perivenulitis occasionally can be seen as an isolated finding, with portal tracts that are either normal or
show only mild nonspecific inflammation. This entity is called isolated central perivenulitis and is generally thought to be a low-grade smoldering rejection, one that can often be treated with optimization of immunosuppressants.
Figure 20.16 Acute cellular rejection, central perivenulitis. Prominent inflammation and hepatocyte dropout are seen adjacent to a central vein.
One study identified this pattern of injury in 28% of liver transplant patients with long-term follow-up.56
It was seen primarily in the late posttransplantation period, with many patients having histories of previous episodes of central perivenulitis in the early posttransplant period (usually with portal-based features of acute cellular rejection). Clinically, this pattern is most often associated with mild liver enzyme abnormalities. When untreated, a small minority of patients develop complications included zone 3 fibrosis, ductopenia, and de novo autoimmune/alloimmune hepatitis.
Chronic hepatitis pattern
When presenting in the late posttransplantation period, acute cellular rejection-related features are often more difficult to recognize, as bile duct injury, “activated” lymphocytes, and endotheliitis, can be sparse and some features can be absent. Thus the biopsy has to be carefully searched for features of acute cellular rejection and other causes of chronic hepatitis (such as recurrent disease, chronic hepatitis E, drug effect, etc.) have to be carefully excluded. When no cause is identified and no definite features or rejection are seen, the pattern of injury is called idiopathic posttransplantation hepatitis
). A significant proportion of allografts evolving to cirrhosis without a clear etiology persistently show this histologic pattern.57
Chronic hepatitis E virus in particular should be carefully excluded, because chronic hepatitis E is a treatable cause of allograft dysfunction and failure63
(see detailed discussion in the “Viral hepatitis” in Chapter 8
Figure 20.17 Idiopathic posttransplant hepatitis. Chronic portal inflammation without diagnostic features of acute cellular rejection.
Grading of acute cellular rejection
The grading system proposed by the Banff group47
—an international consensus document devised by a panel of transplant pathologists, clinicians, and surgeons in 1997—remains the preferred method of grading rejection in transplant centers around the world. An important concept to keep in mind is that this system was designed exclusively as a grading tool. Therefore, the scoring system associated with this method cannot be used to assess a sample for the diagnosis of rejection per se—only to grade it once the diagnosis has independently been established.
The Banff system endorses two forms of rejection assessment: the “global assessment,” whereby acute cellular rejection is classified as mild, moderate, or severe (or grade I-III), according to a specified set of criteria, and the “rejection activity index” (RAI), in which all three components of acute cellular rejection (portal inflammation, bile duct injury, and endothelialitis) are evaluated separately and ascribed a score from 0 to 3, with the overall rejection activity index being the sum of all three individual scores. This method has been validated by a number of subsequent studies, which have shown the Banff schema to be an accurate, reproducible, and clinically useful method for grading acute cellular rejection40
One of the difficulties in grading rejection relates to the presence of histologic variants. Isolated central perivenulitis, for example, cannot be adequately graded by standard acute cellular rejection Banff criteria. According to the Banff group recommendations for interpretation of late liver allograft dysfuction,67
these cases should be interpreted as minimal/indeterminate for rejection when perivenular inflammation involves a minority of central veins with patchy perivenular hepatocyte loss, mild rejection when the changes involve the majority of central veins, moderate rejection when there is at least focal confluent perivenular hepatocyte dropout, but without bridging necrosis, and severe rejection when confluent central-to-central necrosis is present (Table 20.4