Liver Transplantation for the Treatment of Hepatocellular Carcinoma
Liver Transplantation for the Treatment of Hepatocellular Carcinoma
ABSTRACT: The rise of infection with hepatitis C virus worldwide and the lack of effective treatment for this infection has led to a rise in the incidence of hepatocellular carcinoma (HCC). Moreover, it is now accepted that the cirrhotic liver, regardless of etiology, is a nidus for tumor formation. A landmark study of 60 patients by Mazzaferro et al, published in 1996, forever changed the way in which surgeons approach this disease. The Milan criteria (one lesion < 5 cm or three lesions < 3 cm each) were adopted by the United Network for Organ Sharing in 2002 as the criteria by which patients would be given exception points for liver transplantation for HCC. Resection for HCC in the cirrhotic patient has poor outcomes, and many patients' livers cannot be resected due to underlying portal hypertension and thrombocytopenia. Results of transplantation are excellent, with an 80% 5-year survival. Finally, it has been demonstrated that the use of sirolimus for immunosuppression management in these patients may improve long-term disease-free survival; however, no consensus has been reached across the transplant community.
Hepatocellular carcinoma (HCC) is the fifth most common cancer in the world and the third most common cause of cancer-related death. In the majority of cases, HCC develops in the setting of cirrhosis. Treatment with curative intent is possible in only 20% to 25% of cases and consists of resection or liver transplantation.
Epidemiology of HCC
The incidence of HCC continues to rise worldwide, due in large part to the rising number of persons who are positive for infection with hepatitis C virus (HCV). HCV is an RNA virus that infects around 300 million persons worldwide, with approximately 4 million new cases per year. At this time, there is no effective therapy that can reliably cure the infection, nor is there a preventive vaccine. Current therapy with interferon and ribavirin is often difficult to tolerate because of the side effect profiles of both agents. Moreover, the majority of patients are infected with HCV genotype 1, which exhibits the lowest response to therapy—namely, a 50% sustained viral response (SVR) after 48 weeks of therapy, compared with the 80% SVR after 24 weeks of therapy that is seen with genotypes 2 and 3. In the setting of HCV infection, the odds ratio of HCC developing in a patient with cirrhosis is 6.1 compared with a non-cirrhotic, with a yearly incidence of 2% to 8%. These data come from a clinic-based study, which also showed that the patients who were HCV-positive were at a 20-fold increased risk of developing HCC compared with the HCV-negative cohort.[3-5] In the HALT-C (Hepatitis C Antiviral Long-term Treatment against Cirrhosis) study, the 5-year risk of non-cirrhotic HCV-positive patients developing HCC was 4.8%.
Infection with hepatitis B virus (HBV) also portends a risk of HCC, but not to the same degree as HCV infection. In one study by Beasley et al, the annual incidence in HBV-positive patients was 0.5%; in the setting of known cirrhosis, the incidence was 2.5%, with a relative risk (RR) of 100 compared with non-infected matched controls.[7,8] Interestingly, Asian carriers of the virus continue to have a risk of HCC even in the setting of normal liver morphology. This trend is not seen in Caucasians who have cleared the virus and have normal morphology. Therefore, surveillance for HCC is not warranted in non-cirrhotic HBV-positive Caucasian patients. In contrast, surveillance should be continued from age 40 onward in Asian men of similar status, in whom there is an annual incidence of HCC of 0.2%. Of course, in patients co-infected with either HCV or HIV, the incidence of HCC is exponentially greater, necessitating surveillance in these patients.[6-8]
The most rapidly rising cause of cirrhosis, especially in the developed world, is non-alcoholic fatty liver disease.[8,9] The incidence of HCC in the non-viral cirrhotic is not accurately known. In older studies, the presence of alcoholic liver disease and cirrhosis accounted for 32% of HCC cases.[10,11] Whether or not these cohorts were infected with HCV as well is not known. What can be said, based on a review of all relevant studies, is that any form of cirrhosis increases the risk of HCC developing in the native liver. Genetic hemochromatosis is also known to pose a high risk of HCC, with a RR of 20 and a yearly incidence of 3% to 4%.[12-14] Lastly, autoimmune hepatitis is associated with a 1.1% incidence of HCC per year, and end-stage primary biliary cirrhosis poses the same risk as HCV infection. In these latter groups, yearly surveillance is recommended.
Diagnosis of HCC
The tests used to diagnose HCC include radiologic imaging and alpha-fetoprotein (AFP) serology. We will limit our discussion to those patients with cirrhosis because the incidence of HCC in the non-cirrhotic is quite low. Given the current sensitivity/specificity of MRI, a biopsy is rarely, if ever, needed.
There are no blood tests that are currently validated for use in the diagnosis of HCC. AFP levels have long been known to lack the sensitivity and/or specificity needed to diagnose HCC. AFP levels can be elevated in a number of conditions, including intra-hepatic cholangiocarcinoma and HCV infection. Because of this, AFP serology is no longer recommended as a diagnostic test for HCC; thus, the diagnosis is dependent on the radiologic appearance.[16,17]
Currently, most screening for HCC is done by ultrasound. Conventional ultrasound lacks the sensitivity and specificity needed to accurately diagnose small HCCs, and in fact, the American Association for the Study of Liver Diseases (AASLD) recommends that two imaging modalities be used if ultrasound is one of the tests.[18,19] Only contrast-enhanced ultrasound (CE-US) has the diagnostic sensitivity and specificity to be used as a screening tool. Unfortunately, ultrasound is so user-dependent that it has not been adopted globally as the preferred screening test. In a consensus conference hosted by the American Hepato-Pancreato-Biliary Association (AHPBA) in 2009, CE-US was left off as a possible surveillance modality because of this limitation. To prove the unreliability of US as a screening modality, a 1997 European study looked at 43 cirrhotic patients with multiple liver lesions (50 regenerative nodules, 6 benign lesions, 13 HCCs). All 43 patients received US, CT, and MRI imaging within 2 weeks; the lesions detected in the US group were as follows: 1/50 regenerative nodules, 2/6 benign lesions, and 2/16 HCCs. As is apparent from these numbers, US does a poor job of screening for HCC in the cirrhotic liver. This is especially true in the case of small lesions, where the sensitivity of US in one reported study was as low as 14%.
Multi-detector computed tomography (MDCT) and MRI are now the modalities of choice when screening for HCC in the cirrhotic liver (Figure 1). MDCT uses radiation and iodinated contrast to opacify the liver during the arterial phase, the portal venous phase, and the hepatic vein phase. The downside of MDCT is the nephrotoxicity caused by the contrast agent and the radiation exposure to the patient. The radiation exposure from just two CT scans is about 40 mSV, which is equivalent to the radiation exposure 2.4 km from Hiroshima's ground zero. Most large series demonstrate a sensitivity of about 70% (65% to 78%) for CT scanning in the detection of HCC in the cirrhotic liver. The benefit of CT is its more universal acceptance; also, image acquisition is not as onerous as with MRI.
Unlike MDCT, MRI is limited only by the code and acquisition sequences that the radiologist can envision. MRI is a non-invasive and non-radioactive imaging modality that views the human body at the tissue level. In multiple large series, the sensitivity of MRI in detecting small HCCs approaches 90%, and in lesions > 2 cm, it has a sensitivity of 100% and a specificity of 99%, with an overall accuracy of 99.1%. These data were prospectively collected and pathologically correlated. The benefit of MRI is truly realized when we compare its ability to detect small HCCs to that of CT. The sensitivity of CT in patients with small HCCs is around 50%, and in some series no better than 36%. Moreover, recent work by Martin et al shows that liver lesions > 2 cm on MRI can be diagnosed with greater than 99% accuracy without a liver biopsy.[25-37] Given these levels of sensitivity, specificity, and diagnostic accuracy, the usefulness of any additional imaging modality is questionable, and the pursuit of biopsy for pathologic correlation should be discouraged.
Partial Hepatectomy for HCC
The initial treatment decision in patients with HCC and cirrhosis is still based on the basic principles of hepatic surgery: inflow, outflow, and parenchymal reserve. The most important issue in the cirrhotic patient is the functional liver remnant (FLR). Most experts agree that in the cirrhotic liver, an FLR of 40% of the total liver volume is the minimum required for safe resection. The presence of portal hypertension and thrombocytopenia can complicate the picture: both of these negatively affect the outcomes of resection. For these reasons, and based on review of multiple center-specific series, the AHPBA said in their 2010 consensus statement that liver transplantation is the preferred approach for patients with HCC and cirrhosis, or in patients with Child A cirrhosis in the absence of portal hypertension.