Christopher J. Sonnenday
Presentation
A 59-year-old man with cirrhosis secondary to chronic hepatitis C infection undergoes an annual screening liver ultrasound. A 2.5-cm solid, well-circumscribed mass in the posterior aspect of the right hepatic lobe is identified. The liver is noted to be nodular in appearance.
Differential Diagnosis
While the differential diagnosis of a liver mass is broad and includes infectious lesions as well as both benign and malignant lesions, a new mass detected in a patient with chronic liver disease should be presumed to be a primary hepatic malignancy until proven otherwise. Hepatocellular cancer (HCC) is the leading cause of death in clinically compensated cirrhotics, and individuals with cirrhosis secondary to viral hepatitis have a 10% to 20% 5-year cumulative risk of developing HCC. For this reason, HCC screening has been shown to have a profound impact on HCC-related mortality among cirrhotics. The American Association for the Study of Liver Disease currently recommends serial hepatic ultrasound and serum alpha-fetoprotein (AFP) every 6 to 12 months in at-risk populations (e.g., any patient with cirrhosis).
While the evaluation of a liver mass in a patient with cirrhosis is aimed at diagnosing HCC, other etiologies may be considered. Regenerative nodules may appear mass-like and may represent an early stage in the development of hepatocellular neoplasms. Similarly, adenomas may be diagnosed in patients with chronic liver disease. These lesions are at high risk for malignant transformation, especially in this population, and imaging characteristics alone may be insufficient to distinguish hepatic adenomas from well-differentiated HCC. Thus, surgical resection or ablative therapies should be considered in adenomas. Other benign lesions that typically do not require surgical resection include focal nodular hyperplasia (FNH) and hemangioma.
Patients with chronic liver disease and cirrhosis are also at increased risk for developing intrahepatic cholangiocarcinoma, the other major primary hepatic malignancy. While classically described in patients with cholestatic liver disease (e.g., primary sclerosing cholangitis), patients with cirrhosis are also more prone to developing cholangiocarcinoma and mixed tumors that include both HCC and cholangiocarcinoma cell types.
Workup
Any mass lesion suspected on screening ultrasound should be further investigated with contrast-enhanced cross-sectional imaging, either computed tomography (CT) or magnetic resonance imaging (MRI). MRI appears to be slightly more sensitive and specific than CT, and particular imaging characteristics—arterial phase enhancement with early washout of contrast on the delayed phases of the scan—are considered diagnostic for HCC (Figure 1). Contrast-enhanced MRI may provide definitive diagnosis of FNH and hemangioma as well. Image-guided percutaneous biopsy is reserved for cases in which the diagnosis is in doubt following adequate imaging or in cases where it is thought to change management based on clinical suspicion.
FIGURE 1 • MRI of hepatocellular carcinoma, demonstrating characteristic enhancement on arterial phase imaging (left panel) with washout of contrast on delayed-phase imaging (right panel).
Once the diagnosis of HCC is established by imaging or biopsy, the choice of appropriate therapy is made based upon tumor burden, severity of underlying liver disease, and patient performance status. Staging evaluation should include measurement of serum AFP and chest CT. Bone scan may be appropriate when clinically indicated by symptoms or suspicion on cros-ssectional imaging.
Evaluation of underlying liver function and synthetic reserve is critical to providing safe treatment for HCC. While multiple clinical classification systems exist to predict severity of liver disease, the Child-Turcotte-Pugh (CTP) classification may be the simplest and most helpful (Table 1). Consensus exists that CTP class C patients should not undergo hepatic resection due to excessive perioperative mortality, and CTP class B patients should only be considered for minor hepatic resections (resection of two or fewer Couinaud segments) when they have excellent performance status. The evaluation of CTP class A patients for hepatic resection is more difficult, as these patients can vary substantially in their risk of perioperative mortality and postoperative liver failure.
TABLE 1. The CTP Classification of Liver Disease Severity
CTP class A patients with overt evidence of portal hypertension are generally considered suboptimal candidates for hepatic resection. Clinical signs of portal hypertension, such as a history of variceal hemorrhage, esophageal or gastric varices on upper endoscopy, visible upper abdominal varices on cross-sectional imaging, or grossly apparent ascites are all contraindications to resection. Thrombocytopenia is another critical clinical indicator of surgical risk with hepatic resection, reflecting the hypersplenism of advanced cirrhosis and portal hypertension. A platelet count under 100,000 is considered a contraindication to major hepatectomy.
A number of quantitative liver function tests have been investigated to assess hepatic reserve prior to hepatic resection, including indocyanine green (ICG) clearance, galactose elimination capacity, and technetium-99m galactosyl human serum albumin scan, among others. The ICG clearance study is the most commonly used internationally, though it is not commonly used or available in the United States. Many surgeons use volumetric assessment as a proxy for hepatic reserve. This technique relies on the use of manual or automated serial measurement of cross-sectional liver volumes produced from a thin-section helical CT scan. The volume of the liver segments to be preserved following resection is then divided by the total estimated liver volume, which produces a percentage of future liver remnant (FLR) volume. In patients with normal liver parenchyma and function, an FLR of 25% to 30% is considered adequate, if two contiguous Couinaud segments are preserved. In patients with cirrhosis, an FLR of 40% to 50% is desired.
Presentation Continued
In the present case, contrast-enhanced MRI reveals arterial phase enhancement with delayed phase contrast washout of a 2.5-cm solitary lesion in segment 6 of the right hepatic lobe. Chest CT shows no evidence of metastasis. The patient is active and independent in his activities of daily living and works full-time. He has no history of ascites or hepatic encephalopathy. Upper endoscopy shows no esophageal varices. Laboratory evaluation reveals an albumin of 4.1 g/dL, total bilirubin of 1.1 mg/dL, and a platelet count of 148,000. CT volumetry suggests that resection of the posterior sector (segments 6 and 7) would leave an FLR of 65%.
Diagnosis and Treatment
Treatment options for HCC are diverse and require careful consideration of both tumor stage and liver disease severity. Resection offers the best likelihood of survival in select patients with resectable disease, superior to nonsurgical therapies with 30% to 70% 5-year survival. In patients with CTP class B or C liver disease, liver transplantation may be the more appropriate choice of therapy. Posttransplant survival has been shown to be excellent (65% to 80% 5-year survival) when patients are selected for transplant according to strict selection criteria, known as the Milan criteria (solitary tumor under 5 cm, or 3 or fewer tumors each under 3 cm). Obviously donor organ availability and the significant medical risks and costs associated with liver transplantation limit its expansion to all patients with HCC.
Among patients with more extensive tumor burden, and/or decompensated liver disease, ablative therapies may be considered. Radiofrequency ablation may offer prolonged survival and local tumor control, particularly in small solitary lesions (<3 to 4 cm). Transarterial chemoembolization has been shown to extend survival in patients with unresectable disease, though the procedure may also precipitate hepatic decompensation and is best applied to CTP class A or select CTP class B patients. In patients with metastatic or recurrent disease, systemic chemotherapy with the multikinase inhibitor sorafenib may provide additional months of survival in patients not eligible for other therapies.
In cirrhotic patients without an adequate predicted FLR who are otherwise good candidates for surgical therapy, portal vein embolization (PVE) may be considered as a way to augment the size of the remnant liver. PVE takes advantage of the contralateral hypertrophy and ipsilateral atrophy that takes place in response to selective portal vein occlusion. Repeat imaging is typically performed 3 to 6 weeks following embolization, with repeat liver volume estimates. Failure to respond to PVE portends a poor outcome following resection and should be considered a contraindication to proceeding with surgical therapy. Most centers will aim to operate on patients with an appropriate response at 21 to 30 days following PVE, capitalizing on the peak hypertrophic response at this time period.
Surgical Approach
While hepatic resection can be accomplished safely by both open and minimally invasive techniques, cirrhotic patients present particular challenges in terms of transection of the fibrotic hepatic parenchyma and risk of blood loss. Therefore, only experienced laparoscopic liver surgeons should take on minimally invasive hepatic resections in cirrhotic patients. Laparoscopic liver surgery is discussed elsewhere in this text; this section describes open hepatic resection (Table 2).
TABLE 2. Key Technical Steps and Potential Pitfalls in Open Hepatic Resection
Perioperative management of the hepatic resection patient can have a profound influence on outcomes, with maintenance of a low central venous pressure as a central tenet in intraoperative management. Establishment of large-bore venous access is essential, and arterial line placement for systemic blood pressure monitoring is also advised. Central venous access, which has been advocated both for large-volume resuscitation and for CVP monitoring, can be helpful in the most complex cases but is probably not necessary in limited resection cases. Use of large-bore peripheral intravenous catheters (14 or 16 gauge) and conservative volume resuscitation can achieve similar goals.
Incision choice is critical to exposure and efficiency. Left-sided hepatic lesions can be approached from an upper midline incision, but typically a right subcostal incision is preferred. A right subcostal incision with midline extension is a very versatile incision, allowing exposure to the suprahepatic vena cava, making a bilateral subcostal or Chevron incision necessary only for cases with difficult exposure due to body habitus or large tumors. The abdomen should be carefully explored to assess for metastatic disease. Limited mobilization of the liver is performed to allow access for intraoperative ultrasound (IOUS).
Careful IOUS should follow a consistent three-step sequence, beginning with definition of the Couinaud segments based on the portal and hepatic venous anatomy. Attention is paid to important anatomic variants such as early or late division of the right portal pedicle or large accessory hepatic veins. The second phase of the IOUS exam should be a methodical scan through the entire liver parenchyma, with identification and measurement of all lesions. The final step in an IOUS exam is planning of intended resection with note taken of the important critical vasculature to be included or avoided in a segmental resection.
The dissection phase of the operation begins with isolation of the portal structures by opening the pars flaccida (gastrohepatic ligament) and passing a finger through the foramen of Winslow such that the porta can be encircled with a tape. This maneuver facilitates quick access to the porta when necessary, and can facilitate episodic inflow occlusion (Pringle maneuver). The amount of portal dissection necessary is then determined by the extent of hepatectomy planned. A peripheral nonanatomic or segmental resection will not require much additional portal dissection. A formal or extended lobectomy may be approached with a formal portal dissection and pedicle control prior to parenchymal transection.
Parenchymal transection may be performed by multiple techniques, with use of surgical energy devices, clips, or ligatures to control small vessels. Stapling devices may be used to control larger pedicles, including the hepatic veins. The role of inflow occlusion (Pringle maneuver) during parenchymal transection may be utilized to limit excessive blood loss when appropriate. While some surgeons use this technique routinely, others try to limit any potential ischemic injury to the liver remnant, preferring selective pedicle isolation and ligation. When inflow occlusion is utilized, it appears that intermittent periods of clamping followed by periods of reperfusion can limit the ischemic injury to the liver remnant.
Once the parenchymal transection is completed, achieving final hemostasis is critical. Significant bleeding vessels should be controlled with nonabsorbable fine sutures or clips. The argon beam coagulator is effective for small vessels and raw surfaces. Application of additional hemostatic agents such as fibrin- and thrombin-based agents can also be helpful. Care should be taken to identify occult biliary leaks from the parenchymal surface, which should be controlled with fine sutures when identified. Leaving a surgical drain after hepatectomy is a controversial practice and appears to be only partially successful in allowing diagnosis and adequate drainage of a postoperative biliary leak. Selective use of drains for cases with concomitant biliary reconstruction and/or difficult perihilar dissection is probably a reasonable strategy.
Special Intraoperative Considerations
IOUS should be used throughout the hepatectomy operation, not just at the stage of planning the resection. Surgeons should use IOUS to reassess margin status throughout the parenchymal transection. Furthermore, IOUS can be used to identify large vascular pedicles that will need to be controlled in the parenchymal division. Finally, a completion IOUS with Doppler exam can assess the preserved inflow and outflow to the liver remnant, especially in larger segmental resections.
While obtaining a clear surgical margin has obvious oncologic applications, preservation of hepatic parenchyma is of particular concern in patients with cirrhosis. Thus, resection should be planned by IOUS to include an adequate but not excessive margin, with preservation of inflow and outflow to adjacent segments. In the case of hepatic surgery, it appears that the size of the margin is not oncologically relevant, as long as it proves to be histologically negative.
Improved outcomes with hepatic resection have allowed the cautious expansion of advanced hepatobiliary techniques to allow resection of more locally advanced tumors, including those with vascular involvement. Total vascular isolation of the liver may allow resection and reconstruction of tumors involving the vena cava, hepatic veins, or portal structures. However, given the increased risk of postoperative hepatic dysfunction in patients with cirrhosis, these techniques are not advisable in these patients.
Postoperative Management
Perioperative monitoring should include observation for evidence of hemorrhage, particularly in thrombocytopenic or coagulopathic patients. While surgical site infections in hepatectomy patients are relatively rare when compared to patients undergoing gastrointestinal operations, biliary leaks from the cut surface of the liver may predispose to biloma and abscess formation. Typically low-volume leaks can be controlled with adequate percutaneous drainage and observation, but larger, more central leaks may benefit from endoscopic sphincterotomy and endobiliary stent placement to divert bile flow away from the site of leak.
Obviously in the cirrhotic patient, monitoring for signs of hepatic dysfunction is the most important postoperative strategy. Mild hepatic dysfunction can present with progressive jaundice or ascites. If the patient remains otherwise clinically well and free of infection, these clinical problems will often resolve over the first 1 to 2 postoperative weeks. More ominous signs of liver failure include progressive coagulopathy, lactic acidosis, renal dysfunction, vasodilatation, and encephalopathy. Evaluation of a patient with progressive liver dysfunction should include liver duplex to establish patent hepatic vascular inflow and outflow. Renal replacement therapy should be considered in more severe cases to address volume overload and more hepatic injury from congestion. Supportive care is the only intervention in most cases, although salvage liver transplantation may be considered in patients who are otherwise appropriate candidates who have surgical pathology revealing tumors within Milan criteria and without vascular invasion.
Case Conclusion
The patient undergoes an open segment 6/7 resection with IOUS. The liver is stiff and cirrhotic, making division of the hepatic parenchyma challenging. Estimated blood loss is 800 mL. Postoperatively, the patient develops mild ascites and lower extremity edema, and his bilirubin rises to 2.1 mg/dL by postoperative day 5.
Over the ensuing week, the ascites resolves with gentle diuresis, and the bilirubin is corrected. Final pathology from the resection specimen reveals a 2.3-cm moderately differentiated HCC with negative margins and no vascular invasion.
At 18 months postoperatively, the patient remains free of disease with stable liver function.
TAKE HOME POINTS
· A liver mass identified in a patient with cirrhosis should be considered malignant until proven otherwise.
· Contrast-enhanced MRI is the preferred confirmatory diagnostic study for investigation of a liver mass in a cirrhotic patient.
· Arterial phase enhancement with delayed phase washout is diagnostic for HCC.
· Evaluation of a cirrhotic patient for hepatic resection includes careful assessment of both tumor burden and severity of underlying liver disease.
· CTP class B and C patients, and patient with obvious portal hypertension, are not appropriate candidates for hepatic resection.
SUGGESTED READINGS
Azoulay D, Castaing D, Smail A, et al. Resection of nonresectable liver metastases from colorectal cancer after percutaneous portal vein embolization. Ann Surg. 2000; 231:480–486.
Bruix J, Sherman M. Management of hepatocellular carcinoma. Hepatology. 2005;42:1208–1236.
Fattovich G, Stroffolini T, Zagni I, et al. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology. 2004;127:S35–S50.
Marrero JA, Hussain HK, Nghiem HV, et al. Improving the prediction of hepatocellular carcinoma in cirrhotic patients with an arterially-enhancing liver mass. Liver Transpl. 2005;11:281–289.