| |
Liver transplantation in patients with hepatocellular carcinoma
|
| |
| |
Myron Schwartz *Adult Liver Transplantation and Hepatobiliary Surgery, The Mount Sinai School of Medicine, New York, New York
Liver Transplantation
Volume 10, Issue S2, Pages S81-S85
Published Online: February 2004
ABSTRACT
The preferred therapy for hepatocellular carcinoma (HCC) apparently confined to the liver is surgical removal of the tumor. If the location of the tumor and the functional status of the liver are such that resection with an adequate margin can be achieved with low likelihood of subsequent hepatic failure, liver resection is the preferred approach. When HCC apparently localized to the liver is diagnosed in a patient who, by virtue of tumor characteristics or diminished hepatic reserve, is not a candidate for liver resection, liver transplantation becomes a consideration. This work outlines the approach at The Mount Sinai Hospital to the diagnosis, evaluation, preoperative management, transplantation, and posttransplant follow-up in patients with unresectable HCC. The allocation of livers to patients with HCC is reviewed, and predictors of tumor recurrence and results of liver transplantation for HCC are discussed. Finally, the impact of viral hepatitis and of immunosuppression on transplant outcome are discussed.
Evaluation of Liver Function
While there are a number of true liver function tests, such as indocyanine green clearance, measurement of the lidocaine metabolite MEG-X after intravenous lidocaine injection, arterial ketone body ratio, and galactose elimination, in our experience none has proved more useful in sorting patients between resection and transplantation than the Child-Turcotte-Pugh classification in conjunction with an assessment of portal hypertension. Patients with Child's A cirrhosis without significant portal hypertension (either directly measured or utilizing the surrogate marker of platelet count > 100,000), in whom resection can be achieved without the loss of a significant amount of functioning liver parenchyma, are potential resection candidates. Patients with Child's A cirrhosis and portal hypertension, and good Child's B patients, may survive resection; they are, however, likely not to recover fully compensated liver function after surgery. Likewise, the patient with Child's A cirrhosis and no portal hypertension who has a small tumor located such that sacrifice of the right lobe of the liver would be required presents a significant risk of liver failure after resection. In circumstances such as these, transplantation should be considered.
Diagnosis of HCC
Liver imaging has improved dramatically over recent years, allowing for both accurate diagnosis and accurate staging. Either dual phase helical CT scan or MRI with gadolinium should be obtained routinely as part of transplant evaluation. In a patient with cirrhosis, a discrete enhancing liver mass that enhances during the arterial phase of the study and then becomes hypodense relative to the surrounding liver during the portal phase is extremely likely to be an HCC.
It is increasingly becoming accepted that biopsy is unnecessary in the diagnosis of HCC. Biopsy of HCC in cirrhotic livers entails a risk of bleeding (many patients are thrombocytopenic and deficient in coagulation factors) and of tumor dissemination. Furthermore, there is a significant incidence of false-negative results, due to sampling error or to the fact that the cytology of well-differentiated HCC as seen on fine-needle aspiration may be quite similar to that of normal liver. In our view, it is more accurate to declare discrete enhancing masses 2cm in cirrhotic livers to be HCC that to rely on biopsy, and at our institution biopsy for HCC is only rarely performed.
Predictors of Posttransplant Tumor Recurrence
The entire diseased liver is removed during liver transplantation. Recurrence of tumor after transplant is thus entirely the result of extrahepatic dissemination that has occurred before or during the transplant procedure. Possible routes of spread include direct extension, peritoneal (the result of tumor rupture), lymphatic, and hematogenous.
Direct extension to adjacent structures typically occurs in large, extensive tumors that are likely to recur; in and of itself, however, attachment of an HCC to diaphragm, abdominal wall, stomach, etc., does not preclude curative transplant if en bloc resection is possible. Similarly, HCCs that rupture are usually advanced and not suitable for transplant. Although concern is often raised about peritoneal spread in ruptured HCC, we have only rarely observed this phenomenon; blood-borne or lymphatic metastasis is far more commonly observed in following these cases. When a small (often exophytic) tumor presents with rupture, consideration of transplant should not be rejected out of hand.
Vascular invasion is a prerequisite for lymphatic and hematogenous spread. Among the variables that are easy to assess preoperatively, tumor size appears to correlate most closely with vascular invasion. Small, encapsulated HCCs (<3 cm diameter) rarely demonstrate vascular invasion. Once they reach 5 cm, the large majority of HCCs (particularly those that develop in the setting of hepatitis C cirrhosis) will demonstrate microscopic invasion into portal venules and lymphatics around the periphery of the tumor. These findings provide the scientific underpinning for the current organ allocation scheme in the U.S. that assigns priority for transplant to patients with HCC 5 cm.
Macroscopic invasion of the portal or hepatic veins is readily discerned on good-quality imaging. While segmental intrahepatic portal or hepatic vein invasion does not preclude transplant from a technical standpoint, the future appearance of metastases is highly likely. Resection, when feasible, may be warranted in such cases, as significant prolongation of life may result, but the high perioperative risk of transplantation and the donor organ shortage make transplant seem unreasonable to almost all workers in the field.
The significance of the number of tumors present is less clear. In patients transplanted with known HCC, it is common to find multiple smaller lesions upon pathological examination of the liver. The identification of multiple small lesions on pretransplant imaging studies, particularly in view of the increasing sensitivity of such tests, probably has no great significance. On the other hand, in patients with a large (>5 cm) tumor, the presence of multiple other smaller tumors, particularly in the same or adjacent liver segments (satellites), is evidence of vascular invasion with metastasis. In such cases the risk of unrecognized spread outside the liver is increased.
Lymph node metastasis is a contraindication to transplantation. Caution should be exercised in interpreting scans, however, as many patients with chronic liver disease (in particular hepatitis C) have large lymph nodes, often up to 2-3 cm, in the porta hepatis. In patients with HCC, hilar lymph nodes are routinely examined by frozen section at the time of transplant; if metastatic HCC is found, the procedure is aborted and the donor liver is instead transplanted into a back-up recipient.
Tumor grade in HCC increases as the tumor grows larger, with small lesions demonstrating well-differentiated histology and larger tumors typically demonstrating areas of moderate or poor differentiation. Occasionally, tumors are encountered that, despite their small size, exhibit poor differentiation. In the Mount Sinai series, tumor recurrence after transplant has been observed in only 4 of 120 cases with tumors 3 cm; in each case the tumor was poorly differentiated and despite its small size demonstrated microscopic vascular invasion.
Staging
In order to decide whether to allocate a scarce donor liver to transplant a patient with HCC, the extent of disease must be assessed as accurately as possible. Staging systems in current usage are not ideally suited to the assessment of transplant candidacy. The Okuda scheme, widely employed in Japan, attaches greater importance to the state of the primary liver disease than to the extent of tumor. Such a scheme is not useful in assessing transplant candidacy. The TNM system formulated by the American Joint Committee on Cancer is similarly ill-suited for application in transplant candidate evaluation: Stage 3 includes patients with lymph node metastasis who are clearly not transplant candidates, while some patients classified as Stage 4 have significant potential to be cured by a transplant. For this reason, the U.S. transplant community has adopted a new staging system based on current knowledge about risk factors for posttransplant tumor recurrence.
Allocation of Deceased Donor Livers to Patients With HCC
Because of historically poor results, until 1998 the allocation system established by the United Network for Organ Sharing (UNOS) did not assign priority to patients with HCC. In light of recent studies (chief among them that of Mazzaferro et al.) that have demonstrated that patients with unresectable T1 or T2 HCC are unlikely to experience tumor recurrence if transplanted expeditiously, the allocation scheme was modified in 1998. Since then, patients with T1 or T2 HCC receive waiting list priority on the basis of tumor. Currently in the U.S., allocation is on the basis of MELD, a scoring system that assigns patients a numeric index of the severity of their liver disease ranging from 6-40 based on prothrombin time (INR), creatinine, and bilirubin. Patients with T1 HCC are granted a MELD score of 20, and T2 HCC patients receive a score of 24; the scores for tumor patients rise every 3 months as long as the tumor does not grow to exceed priority criteria. While patients with more extensive tumors are not excluded from transplant candidacy, they are not eligible for priority on the waiting list, and thus (at least in most regions of the country), their likelihood of receiving a deceased donor liver is low.
Transplantation for Unresectable HCC Beyond T2
Reported results of transplant in patients with HCC beyond T2 have been worse than those of the general transplant population, with long-term survival between 40-50%. In view of the shortage of donor organs, these patients are not accorded priority by UNOS. On the other hand, transplant clearly provides the best treatment option for many of these patients. We have therefore offered transplantation to patients with HCC beyond T2 within the guidelines that follow (20): --No extrahepatic disease by CT of chest and abdomen, and bone scan;
--Patency of the main portal vein (segmental portal occlusion does not exclude patient) and of main hepatic veins;
--Multicentricity in one lobe does not exclude patient; if bilateral disease, tumor in the lobe with less involvement must be <5 cm;
--Total tumor volume cannot exceed 75% of total liver volume.
Patients meeting these criteria underwent chemoembolization and were placed on the transplant waiting list. We listed 80 such patients between October, 1991 and January, 1999. Of this group, 43 were eventually transplanted, while the others dropped off of the waiting list, most commonly due to tumor progression. Patients received 6 cycles of adriamycin posttransplant. With a mean follow-up of 55 months, 5-year patient survival was 44%; recurrence was seen in 52% of patients. Currently, since the allocation system does not grant priority for tumors >T2, the use of living donors has emerged as a way to successfully transplant these patients.
Tumor Management While Awaiting Transplantation
Because of the long waiting period for a cadaveric donor liver, it is important to treat HCCs in transplant candidates to prevent tumor progression. Modalities commonly employed today include percutaneous ethanol injection (PEI), chemoembolization (CE), and radiofrequency ablation (RFA).
PEI has a long track record and has been demonstrated effective in destroying small (2 cm) HCCs. For tumors between 2-4 cm, the efficacy of PEI falls off; beyond 4 cm, it is not useful. Because of its simplicity and effectiveness, it is the preferred method at many centers to treat small, solitary lesions. Percutaneous injection is limited by the accessibility of the tumor - lesions located high at the dome of the liver may be difficult to approach. Instillation of saline into the abdominal cavity prior to injection may facilitate visualization of such tumors. While concern has been expressed over possible seeding of tumor cells along the injection track, this has not been a common clinical occurrence.
CE is a well-established treatment for HCC that has been shown to significantly prolong life independent of transplant. The details of the method vary considerably among centers. The efficacy of this procedure depends on the fact that HCCs derive their circulation nearly entirely from the hepatic artery. The risk of CE rises with worsening liver function, and the procedure is relatively contraindicated in patients with Child's C cirrhosis. CE is typically employed when the tumor is >4 cm and when there are multiple lesions within an anatomic region of the liver.
RFA has rapidly come into favor, in many centers taking the place of PEI. RFA may be performed percutaneously, laparoscopically, or open, using ultrasound guidance. With current technology, the maximum size of induced necrosis by a single RFA application is in the range of 5-6 cm; to treat tumors requiring a larger volume of tissue destruction (i.e., >4 cm), overlapping spheres of necrosis must be created. Guidance in such instances is imprecise, and the likelihood of complete tumor ablation decreases. For tumors between 2-4 cm, we find that RFA more reliably achieves complete ablation than PEI, as judged by the absence of enhancement on imaging studies.
Viral Hepatitis, HCC, and Transplantation
The large majority of patients transplanted for HCC have viral hepatitis. Recurrence of hepatitis after transplant is an important problem that significantly affects overall prognosis. The issue with HBV has been largely overcome; with the routine use of passive immunoprophylaxis with hepatitis B immune globulin, reinfection occurs in only about 10% of cases, and antiviral drugs such as lamivudine and adefovir are effective in ameliorating its course should recurrent HBV develop.
Recurrence of hepatitis C is a much larger problem. Nearly 100% of patients who are HCV RNA(+) before transplant remain so afterwards; about 50% will have chronic hepatitis at 1 year, and 20% will have cirrhosis at 5 years. Five-year survival is reduced by 5-10% in patients with HCV as compared with patients transplanted for liver diseases that do not recur. Current therapy for HCV, with interferon and ribavirin, is difficult to manage in the posttransplant setting, and reported results are in the range of only 20-25% sustained clearing of HCV. Hepatitis C remains an important factor that must be considered in weighing the overall risk of transplantation in patients with HCV/HCC.
HCC and Posttransplant Immunosuppression
An early study comparing growth rates of metastatic lesions in transplant recipients to growth rates of recurrent tumors in patients after resection raised concern that posttransplant immunosuppression may accelerate the growth of HCC. This concern, however, was lessened by a 1992 study that plotted time to recurrence and to death in patients who underwent either resection or transplant for HCC and found no differences. There is now general agreement that the growth of HCC (as well as most nonvirally induced tumors) is unaffected by immunosuppression.
There are a number of effects of immunosuppressive drugs that have been demonstrated in vitro but have not been proven clinically relevant. In a study employing adjuvant chemotherapy posttransplant, it was hypothesized that cyclosporine may actually be beneficial, reversing expression of the Multiple Drug Resistance gene that limits sensitivity of tumors including HCC to doxorubicin. On the other hand, cyclosporine and tacrolimus are known to directly stimulate hepatic regeneration, and the question has been raised whether HCC might be similarly stimulated. The new agent, sirolimus (rapamycin), has been shown to inhibit hepatic regeneration, and it has been proposed on this basis that it may be a rational immunosuppressive agent for patients transplanted with HCC.
|
|
| |
| |
|
|
|
