Liver Pathology

Sepsis/ICU Associated Cholestasis with Rapidly Progressive Fibrosis

Zachary Goodman
Armed Forces Institute of Pathology
Washington, DC

Clinical History:
The patient is a 45 year old man who was injured by an improvised explosive device (IED) in Iraq, suffering severe blast and shrapnel wounds to both lower extremities and the right hip with open comminuted fracture of the iliac wing. He was hypotensive on arrival at the combat support hospital in Baghdad, where he was transfused with 33 units of blood and packed red cells and other blood products before and during surgery to debride the wounds and evacuate a large retroperitoneal hematoma. He developed acute renal failure, but two days later when stable he was transferred to a medical center in Germany. He spent the next three weeks in intensive care on assisted ventilation and hemodialysis with gradually improving cardiopulmonary and renal function. The abdominal, hip and lower extremity wounds were left open for washout and debridement every 2 to 3 days. Complications included at least 2 episodes of gram negative sepsis and an episode of ventilator-associated pneumonia as well as infections in the hip and lower extremity wounds, treated with multiple antibiotics, primarily levofloxacin, meropenem and vancomycin. Total parenteral nutrition (TPN) was administered from the eighth to the 22nd day after injury. On the 16th day after injury an open cholecystectomy and liver biopsy were performed because of elevated bilirubin and alkaline phosphatase, fever, leukocytosis and thickened gallbladder (? ultrasound). The abdominal wound was gradually closed over the next 6 days, and he was weaned from the ventilator, dialysis and TPN. The patient continued to improve and was sent to a hospital in his home country on the 25th day after injury.


Slide 1
Click to view with ImageScope
Click to view with a Web-Based Viewer

Slide 2
Click to view with ImageScope
Click to view with a Web-Based Viewer


Biopsy Findings:
The biopsy shows considerable bile stasis, most prominent in acinar zone 3 ("centrilobular"). There is also zone 3 steatosis with some hepatocellular ballooning and occasional irregular cytoplasmic inclusions but not true Mallory-Denk bodies. The portal areas are edematous and contain numerous neutrophils, and there is acute cholangitis affecting many acinar bile ducts, along with a zone 1 (periportal) ductular reaction and associated acute inflammation. The Masson trichrome shows early zone 1 fibrosis accompanying the ductular proliferation and also zone 3 perisinusoidal/pericellular fibrosis in the areas with bile stasis and steatosis. There are also several acini with confluent fibrosis of zone 3, and a number of ductules are present in the same areas. Immunostains for alpha smooth muscle actin (SMA) show numerous activated myofibroblasts accompanying the reactive ductules in zone 1 and also in areas of pericellular fibrosis and confluent fibrosis in zone 3.

Comment
This case poses a number of questions:

1) Why was the patient jaundiced?
Bile flow is a complex, highly energy dependent phenomenon. Bilirubin, cholesterol and other lipids, bile acids and other organic anions are cleared from the sinusoidal blood by membrane bound ATP dependant transport proteins and receptors. In the hepatocytes they are conjugated or metabolized to increase their aqueous solubility and transported into the bile by a number of canalicular transport proteins against a concentration gradient. Water and electrolytes accompany bile salts and other solutes by osmotic forces to produce normal bile. Hydrostatic pressure along with pericanalicular contractile proteins help bile to flow through the small bile channels, and larger ducts have smooth muscle to produce peristaltic movement of the luminal bile. There are many redundancies in the system so that minor injuries can be absorbed without development of significant disease, and even some more significant injuries can be mitigated. However, when an injury is sufficiently severe, cholestasis (failure of bile flow) occurs, and as serum bilirubin rises, the patient becomes jaundiced.

Sepsis, and to a lesser extent bacterial infections that do not cause generalized sepsis, is one of the more common causes of non-obstructive intrahepatic cholestasis. The pathogenesis is complex and multifactorial, especially in a patient like the current example who has had severe trauma. Multiple transfusions result in hemolysis and increased bilirubin production, while hypotensive episodes and decreased hepatic perfusion cause delayed clearance of bilirubin and other substances. Poor hepatic perfusion also contributes by retarding metabolism and decreasing the production of ATP, thus inhibiting the energy dependent transport processes necessary for bile secretion. Endotoxin and proinflammatory cytokines inhibit multiple steps in bile acid uptake, transport and excretion by the various transport proteins. Drugs, especially antibiotics can cause idiosyncratic severe cholestatic reactions in susceptible individuals, but lesser degrees of interference with transport proteins may also occur and contribute to the cholestasis. Finally, TPN is also associated with intrahepatic cholestasis by poorly understood mechanisms, many of which probably overlap with sepsis associated cholestasis.

The histologic differential diagnosis of the cholestatic injury in this case is between intrahepatic cholestasis and large duct mechanical biliary obstruction, and the histologic findings do not allow the distinction on morphologic grounds alone. Even without definitive evidence for absence of obstruction, recognition of the likelihood of sepsis associated cholestasis could justify conservative management of the patient. Nevertheless, once mechanical biliary obstruction is excluded, as it was in the present case at the time of surgery, then sepsis is the likely principal cause with many contributing factors, including hepatic ischemia, renal failure, multiple transfusions, hemolysis, TPN and other drugs.

2) Why is there a ductular reaction?
The prominent ductular reaction in this case consists of ductules with well-defined lumina. They are located for the most part within and at the edges of fibrotic portal tracts and area accompanied by neutrophils (so-called acute cholangiolitis). This type of ductular reaction was previously referred to as "typical" and felt to be the result of proliferation of pre-existing bile ducts in acute biliary obstruction, as opposed to "atypical" ductular reaction consisting of small anastomosing ductules with poorly defined lumina that are associated with scarring in many chronic liver diseases. This terminology (typical vs. atypical) is now discouraged, since the two types are often difficult to distinguish, and it is now believed that expansion of the hepatic progenitor cells located in the canals of Hering along with an unknown contribution from bone marrow derived progenitor cells may be the source of most of the reactive ductules. Most studies that have shown this have been in chronic fibrosing liver diseases, where the ductular reaction accompanies scarring, but hepatocellular regeneration from the progenitor cell compartment also occurs in massive hepatic necrosis, so it seems likely that an acute biliary obstruction would also trigger a progenitor cell expansion. In the present case, the patient did not have a mechanical obstruction, but the sepsis-related cholestasis appears to have produced a functional obstruction with acute cholangitis of the acinar bile ducts. The term "cholangitis lenta" historically has been used for non-obstructive cholangitis related to generalized sepsis. In its most severe form there is inspissated bile in dilated periportal ductules, a lesion that has been called "bile ductular cholestasis" which when seen in a liver biopsy is virtually pathognomonic of sepsis.

In addition to the periportal ductular reaction, there are prominent duct-like structures in some areas that have zone 3 ("centrilobular") fibrosis, presumably related to recent ischemic/hypoxic injury. How they got to this location is unclear, but it is possible that they were derived from intra-acinar progenitor cells of canals of Hering, which can extend into the parenchyma for about some distance from the portal tract to central vein. Alternatively, they may be derived from circulating bone marrow-derived progenitor cells. In either case, they probably play a role in fibrogenesis, as discussed next.

3) Why is there fibrosis after such a short time?
Hepatic fibrosis, defined as accumulation of excess extracellular matrix components, is considered to be the liver's wound healing response to chronic injury. Whether fibrosis occurs, where it occurs, how much takes place, and the rapidity of fibrosis progression are complicated, imperfectly understood phenomena. The cells that produce collagen and other matrix components are predominantly activated myofibroblasts that can be identified in tissue by the presence of alpha smooth muscle actin.

There are at least four possible sources of these myofibroblasts in injured liver tissue. The hepatic stellate cells (previously called perisinusoidal lipocytes or Ito cells), which reside in the perisinusoidal space of Disse where they store vitamin A and regulate sinusoidal blood flow, react to tissue injury by transformation into activated myofibroblasts. Portal fibroblasts and myofibroblasts also become activated, particularly in biliary tract disease. Biliary epithelial cells and possibly also hepatocytes can undergo epithelial-mesenchymal transition to become activated myofibroblasts. And finally, cells whose origin is in the bone marrow may travel through the circulation to sites of injury and differentiate into myofibroblasts. The hepatic stellate cells have been most intensely studied, since their lipid content allows them to be isolated and grown in tissue culture, but the relative roles of the various populations is uncertain.

The mesenchymal cells that give rise to the myofibroblasts involved in production of fibrosis accompany the ductular reaction that occurs in many types of liver injury. As the ductules derive from hepatic epithelial progenitor cells, they are supported by and interact with mesenchymal cells, including stellate cells, endothelial cells, and other stromal cells. Experimental models of fibrosis, such as the bile duct ligated rat, suggest that in biliary disease, portal myofibroblasts accompanying the ductular reaction are the major source of fibrosis. On the other hand, in diseases with intralobular fibrosis, such as steatohepatitis and vascular disease, the stellate cells may be the major source. Other diseases may have varying proportions of the two processes.

In the present case, the periportal fibrosis can be attributed to portal myofibroblast activation accompanying the ductular reaction secondary to the cholangitis of sepsis. The diffuse zone 3 fibrosis can be attributed to stellate cell activation and collagen production due to the acute cholestatic injury with perhaps some contribution from the steatosis. The areas of confluent zone 3 fibrosis are most likely due to ischemic injury with progenitor cell proliferation producing the ductules and myofibroblast activation causing the scarring. The fact that so much fibrosis can occur in only sixteen days shows that this is a highly dynamic process and that under the right circumstances, fibrosis can occur very rapidly.

4) What are the long-term consequences of this injury?
At the time of discharge, the patient had shown remarkable improvement clinically, and it is quite possible that the fibrosis may entirely disappear as the source of the injury is removed. Liver extracellular matrix, like other tissues, contains a number of remodeling proteins, notably the matrix metalloproteinases. Part of the process of scar formation involves inhibition of the matrix metalloproteinases, while remodeling and resolution of fibrosis can take place by removal of activated myofibroblasts by apoptosis, decreased matrix production and increased fibrolysis. Experimental models of liver fibrosis and cirrhosis (such as rats with bile duct ligation or chronic carbon tetrachloride intoxication) spontaneously improve and may revert to normal after the cause of injury is removed. Improvement in fibrosis and even possible reversal of cirrhosis in humans has been observed following successful therapy of chronic viral hepatitis, hemochromatosis, autoimmune hepatitis and chronic biliary obstruction, so it is likely that this patient's liver fibrosis will improve, and it is possible that his liver could be essentially normal within a few months.

Bullet Points
1. Bile flow is a complex, energy dependant process, and many types of liver injury interfere with bile flow, producing cholestasis.

2. Ductular reaction is a response to many types of liver injury.

3. Alpha smooth muscle actin positive myofibroblasts accompany reactive ductules and produce hepatic fibrosis in many types of liver injury.

References:

Cholestasis in Sepsis
  1. Zimmerman HJ, Fang M, Utili R, Seeff LB, Hoofnagle JH: Jaundice due to bacterial infection. Gastroenterology 1979; 77:362-374.

  2. Lefkowitch JH: Histological assessment of cholestasis. Clin Liver Dis 2004; 8:27-40.

  3. Chand N, Sanyal AJ: Sepsis-induced cholestasis. Hepatology 2007; 45:230-241.
Ductular Reaction
  1. Roskams TA, Theise ND, et al: Nomenclature of the finer branches of the biliary tree: canals, ductules, and ductular reactions in human livers. Hepatology 2004; 39:1739-1745.

  2. Roskams T, Desmet V: Ductular reaction and its diagnostic significance. Semin Diagnostic Pathol 1998; 15:259-269.
Fibrosis
  1. Cassiman D, Roskams T: Beauty is in the eye of the beholder: emerging concepts and pitfalls in hepatic stellate cell research. J Hepatol 2002; 37:527-535.

  2. Desmouliere A, Darby IA, Gabbiani G: Normal and pathologic soft tissue remodling: Role of the myofibroblast, with special emphasis on liver and kidney fibrosis. Lab Invest 2003; 83:1689-1707

  3. Ramadori G, Saile B: Portal tract fibrogenesis in the liver. Lab Invest 2004; 84:153-159.

  4. Beaussier M, Wendum D, Schiffer E, Dumont S, Rey C, Lienhart A, Housset C: Prominent contribution of portal mesenchymal cells to liver fibrosis in ischemic and obstructive injuries. Lab Invest 2007; 87:292-303.

  5. Rygiel KA, Robertson H, Marshall HL, Pekalski M, Zhao L, Booth TA, Jones DEJ, Burt AD, Kirby JA: Epithelial-mesenchymal transition contributes to portal tract fibrogenesis during human chronic liver disease. Lab Invest 2008; 88:112-123.

  6. Friedman SL: Hepatic stellate cells: Protean, multifunctional, and enigmatic cells of the liver. Physiol Rev 2008; 88:125-172.

  7. Gieling RG, Burt AD, Mann DA: Fibrosis and cirrhosis reversibility – Molecular mechanisms. Clin Liver Dis 2008; 12:915-937.