Review For reprint orders, please contact: [email protected]
2014
Resection and transplantation for hepatocellular carcinoma: factors influencing surgical options
Rebecca M Dodson1, Jin He1 & Timothy M Pawlik*,1
7
7
2 1 7 /
uture
ion for factors
ture , UK
Abstract: The management of hepatocellular carcinoma within the Milan criteria and with well-compensated cirrhosis is a topic of debate. Recent surveillance programs in patients with hepatitis C and cirrhosis have allowed some patients to be diagnosed with early, potentially curable, disease via liver resection (LR), liver transplantation (LT) or liver ablation. LT has excellent outcomes with 5-year survival rates >70% for patients within the Milan criteria. However, its utilization is limited by increasing organ shortages. LR is also effective with 5–year survival outcomes between 50–70% and safe in light of advances in surgical technique, preresection optimization and patient selection. Patients with solitary tumors and well-preserved liver function are good candidates for LR, whereas LT is best reserved for patients with compromised liver function and multifocal disease. LT and LR should not be viewed as competing tools but as complementary tools in the current armamentarium to treat early hepatocellular carcinoma. Hepatocellular carcinoma (HCC) is the most common primary liver cancer and its incidence is increasing in the USA and worldwide (Figure 1) [1] . In the USA, the incidence is 4.9 per 100,000 and has been steadily rising at a rate of 4.5% per year for the last three decades, in part, as a result of the hepatitis C viral (HCV) epidemic [2] . Other predisposing factors for HCC include: hepatitis B virus (HBV), alcoholic liver cirrhosis, nonalcoholic steatohepatitis, obesity and diabetes. In fact, the proportion of patients with non-B non-C hepatitis (viral hepatitis that is HBV and HCV negative) is increasing. Other important risk factors for HCC include nonalcoholic steatohepatitis and metabolic syndrome [3] . While surgery remains the only curative treatment for HCC, there are a multitude of therapeutic options for HCC including liver resection (LR), liver transplantation (LT), liver ablation (LA), intra-arterial therapy, as well as systemic chemotherapy. Importantly, patients with HCC, unlike other patients with cancer, very often have two underlying problems: cancer and cirrhosis. There is an ongoing debate regarding the use of LR versus LT in patients with HCC, especially in the management of patients with early HCC and minimal or well-compensated cirrhosis. Currently, many surgeons advocate for resection as the best option for treatment of HCC in noncirrhotic patients, whereas in patients with concomitant HCC and cirrhosis, LT offers treatment of both underlying conditions with extirpation of the malignancy and replacement of the cirrhotic liver. Although, from an overall therapeutic standpoint, LT may be the best therapeutic option for patients with early HCC and cirrhosis, the relative scarcity of available organs and the long waiting time may make resection a more prudent alternative for some patients. The use of LR relative to LT in the management of early HCC entails the careful selection of candidates based on individual clinical factors, as well as data to suggest that LR can be performed safely with potential benefits similar to LT.
Keywords
• cancer • early stage • HCC • hepatocellular carcinoma • liver • resection • surgery • transplantation
Johns Hopkins University School of Medicine, Department of Surgery, 600 North Wolfe Street, Blalock 688, Baltimore, MD 21287, USA *Author for correspondence: Tel.: +1 410 502 2387; Fax: +1 410 502 2388; [email protected] 1
10.2217/FON.13.225 © 2014 Future Medicine Ltd
Future Oncol. (2014) 10(4), 587–607
part of
ISSN 1479-6694
587
Review Dodson, He & Pawlik 6
16 14
Survival
5
12 4
10
3
8 6
2
5-year survival (%)
Incidence rate (cases/100,000 persons)
Age-adjusted incidence rate
4 1
2
0 2007
2006
2004
2005
2003
2001 2002
2000
1999
1997 1998
1995
1996
1992
1993 1994
1990 1991
1988 1989
1987
1985 1986
1983 1984
1981 1982
1979 1980
1977 1978
1976
1973
1974 1975
0
Hepatocellular carcinoma diagnosis (year)
Figure 1. Hepatocellular carcinoma age-adjusted incidence and 5-year survival rates have increased for patients in the USA from 1973 to 2007. Reproduced with permission from [1].
Surgical treatment options HCC is currently the most common cause of death among patients with compensated cirrhosis. The early detection of HCC via surveillance programs has offered improved survival partly due to the ability to detect early-stage HCC that is amenable to curative therapy [4] . Therapies considered to be curative for HCC include LR, LT and LA [4] . The current literature lacks meaningful comparison between LR and LT, as no randomized controlled trials exist and prospective studies are unlikely to proceed due to the difficulty in randomizing patients to resection versus transplantation. Since no randomized-controlled clinical trials exist comparing the three major therapies, the accumulation of cohort studies have helped establish the current roles for each of these therapeutic options. However, these retrospective studies often fail to compare patients according to tumor stage and liver function. In general, LR may be the best option for patients with normal liver function or early fibrosis/cirrhosis. LR has been reported to be associated with a 5–year survival rate of up to 70%. However, tumor recurrence following LR is high, between 50–70% of patients will develop recurrence within 5 years of surgery.
588
Future Oncol. (2014) 10(4)
In addition, LR is not feasible in many patients with HCC, as only 5–10% of all HCC cases in the USA are amenable to surgical resection. Unfortunately, many patients are not candidates for surgical resection owing to portal hypertension and its stigmata: varices, splenomegaly and platelet count 70% and tumor recurrence rates are 12 months and drop-out rates as high as 20–50% may limit the availability of LT for all patients. LA is the other potentially curative therapeutic option for patients with HCC. In general, LA is reserved for patients with small lesions in patients with otherwise inoperable/unresectable disease, or those with small lesions and a poor performance status that would prohibit LR or LT. LA is also commonly used as a bridge therapy to LT. Other noncurative treatments that have been shown to improve survival for HCC include transarterial chemoembolization (TACE) and sorafenib [5,6] .
future science group
Resection & transplantation for hepatocellular carcinoma: factors influencing surgical options ●●Downstaging & neoadjuvant therapies
Neoadjuvant therapies in HCC are therapies that are given to patients that would otherwise could be treated with curative therapy to improve long-term outcomes, make definitive therapy safer, or bridge for transplantation. Neoadjuvant therapies include chemotherapy, TACE, intra-arterial radioembolization and ablation. For patients with resectable disease at presentation neoadjuvant intra-arterial therapies are not recommended due to risk of tumor progression or resultant liver failure [7,8] . Approximately 30–50% of patients on the waiting list receive neoadjuvant therapy in order to decrease dropout. Currently bridging neoadjuvant therapy is supported so long as it does not delay transplantation. In contrast to neoadjuvant therapy, downstaging refers to treatment in which the intent is to make an unresectable HCC resectable and has been accomplished in up to 24–90% of patients [9] . ●●Targeted therapies: sorafenib
After the approval of sorafenib as the standard of care for patients with advanced HCC, the Sorafenib Hepatocellular Carcinoma Assessment Randomized Protocol demonstrated an increased median survival and the time to radiographic by approximately 3 months [5] . The median overall survival (OS) was 10.7 months in the sorafenib group and 7.9 months in the placebo group (p < 0.001), while time to radiologic progression was 5.5 months in the sorafenib group and 2.8 months in the placebo group (p < 0.001). Currently the STORM trial, is testing sorafenib as an adjuvant therapy in 1100 patients with early HCC treated by resection or ablation to determine whether inhibiting kinase activity reduces the risk of HCC recurrence. Villanueva and Llovet review over 50 current molecular therapies under evaluation for HCC, which may lead to development of individualized targeted therapies [10] . Growing numbers
Review
of case reports describing the downstaging of unresectable HCC using sorafenib alone or in combination with locoregional therapy suggest promise in a minority of patients with advanced HCC [7,11–13] . Therapeutic decision-making in the treatment of HCC necessitates the consideration of both tumor- and liver-specific factors in the selection of treatment type (Table 1) . Liver-specific factors relate to both the quality and quantity of underlying liver. Although tumor-specific factors may be conducive to LR, unfavorable liver factors may preclude it – making LT a more attractive option. While a subset of noncirrhotic patients may tolerate a major hepatectomy, the degree to which patients with compromised hepatic function tolerate surgery will depend, in part, on the quality and quantity of the [14] . Cirrhosis & quality of the underlying liver ●●Child–Turcotte–Pugh classification
The quality of underlying liver can be assessed using the Child’s classification, the Model for End-Stage Liver Disease (MELD) score, and direct measurement of the portal venous pressure. The Child-Turcotte-Pugh classification (CTP) was developed as a tool to predict outcomes following shunt procedures. High CTP scores have been shown to correlate with high mortality following major abdominal surgery. Garrison et al. and Mansour et al., using the five parameters of CTP (total bilirubin, albumin, international normalized ratio, ascites and hepatic encephalopathy), found very comparable stratification of mortality following abdominal surgery, with dramatically increased mortality in CTP class B (CTP-B) and CTP class C (CTP-C) [15,16] . Specifically, Garrison et al. and Mansour et al. reported that patients who were CTP class A (CTP-A) had a 10% mortality, whereas mortality in patients with CTP-B and CTP-C liver function were 30–31% and 76–82%, respectively [15,16] . Current estimates
Table 1. Tumor- and liver-specific factors considered in treatment allocation. Tumor-specific factors
Liver-specific factors
Size Number Location Relation to adjacent structures Distant spread Vascular invasion Histologic grade
Cirrhosis/fibrosis Ascites Portal hypertension Thrombocytopenia Bilirubin Size of remnant liver Hepatitis etiology
future science group
www.futuremedicine.com
589
Review Dodson, He & Pawlik Table 2. Relationship between Model for End-Stage Liver Disease score and postoperative mortality. MELD score 0–7 (n = 351) 8–11 (n = 257) 12–15 (n = 106) 16–20 (n = 35) 21–25 (n = 13) ≥26 (n = 10)
Mortality % (patients at risk[n]) 7 days
30 days
90 days
1 year
5 years
10 years
1.9 (314) 3.3 (236) 7.7 (94)
5.7 (301) 10.3 (219) 25.4 (78)
9.7 (287) 17.7 (200) 32.3 (69)
19.2 (253) 28.9 (170) 45.0 (56)
50.7 (123) 58.5 (83) 69.5 (24)
72.6 (57) 78.1 (35) 87.3 (10)
14.6 (29) 23.0 (7) 30.0 (6)
44.0 (19) 53.8 (4) 90.0 (1)
55.8 (15) 66.7 (3) 90.0 (1)
70.5 (10) 84.6 (2) 100 (0)
94.1 (2) 92.3 (1) 100 (0)
94.1 (2) 100 (0) 100 (0)
MELD: Model for End-Stage Liver Disease. Reproduced with permission from [19].
of nontransplant operative mortality in patients with CTP-A, CTP-B, and CTP-C liver function are 5–10, 9–30 and 50–90%, respectively. As such, LR for HCC is largely relegated to patients who are CTP-A, whereas patients with HCC who are CTP-B or CTP-C are best served with LT. ●●MELD score
Although CTP has been shown to be highly predictive of mortality after resection of HCC, it is somewhat subjective and has been criticized as only an estimation of liver function. A subset of patients with presumably well-compensated/ preserved function designated as CTP-A can still develop significant hepatic decompensation after LR. For example, Bruix et al. found that >50% of CTP-A patients had signs of hepatic decompensation after resection of HCC and more than a third of these patients had not recovered by 90 days [17] . The MELD score has been shown to accurately predict long-term survival in patients with cirrhosis, as well as outcomes for cirrhotic patients undergoing resection for HCC and other major surgical procedures [18] . The MELD score was developed to predict the short-term outcome of patients undergoing the transjugular intrahepatic portosystemic shunt (TIPS) procedure for cirrhosis and is derived from international normalized ratio, total bilirubin and serum creatinine. The MELD score is more objective compared with the CTP classification, as it depends solely on quantifiable laboratory values. The MELD score has been used as a criterion in liver allocation by the United Network for Organ Sharing (UNOS) since 2002. In a large study of 772 cirrhotic patients, Teh et al. concluded that patients with a small HCC ≤5 cm and a MELD ≤8 tolerated LR well [19] . By contrast, the authors noted a MELD score ≥9 was a predictor
590
Future Oncol. (2014) 10(4)
of both increased perioperative mortality and worse long-term survival. Similarly, in their series, Cucchetti et al. reported that no patient with a MELD 1 cm from the tumor, these patients had similar DFS and OS as patients with macroscopic invasion. While macroscopic invasion can be determined preoperatively, microscopic invasion requires pathologic examination. Consequently, preoperative findings have been used to predict the presence of vascular invasion. Pawlik et al. definitely demonstrate that vascular invasion increases dramatically as tumor size increases
www.futuremedicine.com
595
Review Dodson, He & Pawlik
Tumors with microvascular invasion (%)
80 70 60
**
Low Intermediate
**
**
High
**
50 40
*
30 *
20 10 0 ≤3
3.1–5
5.1–6.5
>6.5
Tumor size (cm) Figure 6. Histologic grade stratified by tumor size and rate of microvascular invasion. Increasing tumor size was associated with microvascular invasion and higher grade. *p < 0.05 versus large (>5 cm) intermediate- or high-grade tumors. **p < 0.05 versus small (≤5 cm) intermediate- or high-grade tumors. Reproduced with permission from [28].
(Figure 6) [28] . Specifically, vascular invasion nearly doubled in tumors >5 cm versus 5 cm; in the same study the authors noted that histologic grade was the most significant predictor of occult vascular invasion in tumors >5 cm. Therefore, tumor grade determined by preoperative fine needle aspiration (FNA) has been proposed by some as a means to help define patients for transplantation. For example, Cillo et al. report a 5-year actuarial survival of 75% and DFS of 92% in patients undergoing LT in well-to-moderately differentiated HCC, despite 38% of patients being beyond the Milan criteria
596
Future Oncol. (2014) 10(4)
[37] . As such, careful consideration of patients for LT for tumors >5 cm should include criteria such as solitary tumor, low tumor grade and low AFP levels. The University of California San Francisco (UCSF) group has advocated for extended criteria for transplantation especially with regard to size when these other factors are considered [38] . ●●Tumor number
Multinodular HCC (MNHCC) may be due to intrahepatic metastasis lesions or multicentric carcinogenesis. MNHCC has been shown to be a major independent prognostic factor in patients undergoing LR [23,39,40] . Wang et al. found significant differences in the long-term survival of solitary HCC and MNHCC with 1-, 3-, and 5-year survival rates of 88.0, 69.2 and 58.4% versus 86.1, 55.5 and 29.9%, respectively (p < 0.001) [39] . The median 1-, 3-, and 5-year DFS rates for patients with MNHCC was only 45.7, 29.2 and 18.4%, respectively. Factors particularly associated with a poor outcome after LR included AFP >400 ng/ml, total tumor size >5 cm, largest tumor size >5 cm, total tumor number >3 and presence of microvascular invasion. Wang et al. concluded that LR provided a superior benefit to ablation or chemoembolization among those patients with MNHCC with tumor number ≤3 and largest tumor size ≤5 cm
future science group
Resection & transplantation for hepatocellular carcinoma: factors influencing surgical options [39] . However, a high recurrence has also been reported for patients undergoing LR versus LT for MNHCC (46 vs 27%, respectively) [41] . As such, due to high recurrence after LR, MNHCC is considered a relative contraindication to LR in some centers. LT should be considered a better therapeutic option for MNHCC within the Milan or extended UCSF criteria.
AFP AFP has been used as a serum tumor marker since the 1970s as greater than 75% of patients present with levels above 10 ug/l. Multiple groups have associated elevated AFP levels with large tumors, bilobar involvement and portal thrombus in addition to prognostic value [42–46] . The UCSF group found levels >1000 ng/ml with increased risk of recurrence after LT [38] . Ma et al. found that preoperative serum level of AFP ≤20 ng/ ml, AFP 20–400 ng/ml and AFP >400 ng/ml differentiated clinical and pathological factors [42] . Patients with levels ≤20 ng/ml had higher cell differentiation, lower rates of microvascular invasion and higher DFS. Whereas AFP levels >400 ng/ml conferred poorer postoperative survival. Duvoux et al. found that AFP levels independently predicted tumor recurrence and correlated with both vascular invasion and cellular differentiation in patients undergoing LT [47] . In fact, a subset with AFP levels ≤100 ng/ml had low risk of recurrences with survival near 70%, whereas patients with AFP levels >1000 ng/ml were at high risk of recurrence and reduced survival. Treatment allocation The BCLC staging systems combines tumor stage, degree of liver function, the patient’s general condition/performance status, cancerrelated symptoms and treatment efficacy. These factors are combined with a treatment algorithm to form the BCLC system, which, in turn, provides treatment recommendations. A key criterion in the BCLC staging system is performance status, which has been shown to be a comprehensive determinant of overall health status, an indicator of treatment benefit, and a predictor of long-term survival in cancer patients. The BCLC is often cited in the treatment of patients with HCC (Figure 7) [4] . In general patients at a very early stage (Stage 0) and early stage (Stage A) are candidates for surgical therapy. In the absence of portal hypertension and normal bilirubin, solitary tumors are evaluated for LR.
future science group
Review
Candidates are evaluated for transplantation if the patient has three nodules
2014
Resection and transplantation for hepatocellular carcinoma: factors influencing surgical options
Rebecca M Dodson1, Jin He1 & Timothy M Pawlik*,1
7
7
2 1 7 /
uture
ion for factors
ture , UK
Abstract: The management of hepatocellular carcinoma within the Milan criteria and with well-compensated cirrhosis is a topic of debate. Recent surveillance programs in patients with hepatitis C and cirrhosis have allowed some patients to be diagnosed with early, potentially curable, disease via liver resection (LR), liver transplantation (LT) or liver ablation. LT has excellent outcomes with 5-year survival rates >70% for patients within the Milan criteria. However, its utilization is limited by increasing organ shortages. LR is also effective with 5–year survival outcomes between 50–70% and safe in light of advances in surgical technique, preresection optimization and patient selection. Patients with solitary tumors and well-preserved liver function are good candidates for LR, whereas LT is best reserved for patients with compromised liver function and multifocal disease. LT and LR should not be viewed as competing tools but as complementary tools in the current armamentarium to treat early hepatocellular carcinoma. Hepatocellular carcinoma (HCC) is the most common primary liver cancer and its incidence is increasing in the USA and worldwide (Figure 1) [1] . In the USA, the incidence is 4.9 per 100,000 and has been steadily rising at a rate of 4.5% per year for the last three decades, in part, as a result of the hepatitis C viral (HCV) epidemic [2] . Other predisposing factors for HCC include: hepatitis B virus (HBV), alcoholic liver cirrhosis, nonalcoholic steatohepatitis, obesity and diabetes. In fact, the proportion of patients with non-B non-C hepatitis (viral hepatitis that is HBV and HCV negative) is increasing. Other important risk factors for HCC include nonalcoholic steatohepatitis and metabolic syndrome [3] . While surgery remains the only curative treatment for HCC, there are a multitude of therapeutic options for HCC including liver resection (LR), liver transplantation (LT), liver ablation (LA), intra-arterial therapy, as well as systemic chemotherapy. Importantly, patients with HCC, unlike other patients with cancer, very often have two underlying problems: cancer and cirrhosis. There is an ongoing debate regarding the use of LR versus LT in patients with HCC, especially in the management of patients with early HCC and minimal or well-compensated cirrhosis. Currently, many surgeons advocate for resection as the best option for treatment of HCC in noncirrhotic patients, whereas in patients with concomitant HCC and cirrhosis, LT offers treatment of both underlying conditions with extirpation of the malignancy and replacement of the cirrhotic liver. Although, from an overall therapeutic standpoint, LT may be the best therapeutic option for patients with early HCC and cirrhosis, the relative scarcity of available organs and the long waiting time may make resection a more prudent alternative for some patients. The use of LR relative to LT in the management of early HCC entails the careful selection of candidates based on individual clinical factors, as well as data to suggest that LR can be performed safely with potential benefits similar to LT.
Keywords
• cancer • early stage • HCC • hepatocellular carcinoma • liver • resection • surgery • transplantation
Johns Hopkins University School of Medicine, Department of Surgery, 600 North Wolfe Street, Blalock 688, Baltimore, MD 21287, USA *Author for correspondence: Tel.: +1 410 502 2387; Fax: +1 410 502 2388; [email protected] 1
10.2217/FON.13.225 © 2014 Future Medicine Ltd
Future Oncol. (2014) 10(4), 587–607
part of
ISSN 1479-6694
587
Review Dodson, He & Pawlik 6
16 14
Survival
5
12 4
10
3
8 6
2
5-year survival (%)
Incidence rate (cases/100,000 persons)
Age-adjusted incidence rate
4 1
2
0 2007
2006
2004
2005
2003
2001 2002
2000
1999
1997 1998
1995
1996
1992
1993 1994
1990 1991
1988 1989
1987
1985 1986
1983 1984
1981 1982
1979 1980
1977 1978
1976
1973
1974 1975
0
Hepatocellular carcinoma diagnosis (year)
Figure 1. Hepatocellular carcinoma age-adjusted incidence and 5-year survival rates have increased for patients in the USA from 1973 to 2007. Reproduced with permission from [1].
Surgical treatment options HCC is currently the most common cause of death among patients with compensated cirrhosis. The early detection of HCC via surveillance programs has offered improved survival partly due to the ability to detect early-stage HCC that is amenable to curative therapy [4] . Therapies considered to be curative for HCC include LR, LT and LA [4] . The current literature lacks meaningful comparison between LR and LT, as no randomized controlled trials exist and prospective studies are unlikely to proceed due to the difficulty in randomizing patients to resection versus transplantation. Since no randomized-controlled clinical trials exist comparing the three major therapies, the accumulation of cohort studies have helped establish the current roles for each of these therapeutic options. However, these retrospective studies often fail to compare patients according to tumor stage and liver function. In general, LR may be the best option for patients with normal liver function or early fibrosis/cirrhosis. LR has been reported to be associated with a 5–year survival rate of up to 70%. However, tumor recurrence following LR is high, between 50–70% of patients will develop recurrence within 5 years of surgery.
588
Future Oncol. (2014) 10(4)
In addition, LR is not feasible in many patients with HCC, as only 5–10% of all HCC cases in the USA are amenable to surgical resection. Unfortunately, many patients are not candidates for surgical resection owing to portal hypertension and its stigmata: varices, splenomegaly and platelet count 70% and tumor recurrence rates are 12 months and drop-out rates as high as 20–50% may limit the availability of LT for all patients. LA is the other potentially curative therapeutic option for patients with HCC. In general, LA is reserved for patients with small lesions in patients with otherwise inoperable/unresectable disease, or those with small lesions and a poor performance status that would prohibit LR or LT. LA is also commonly used as a bridge therapy to LT. Other noncurative treatments that have been shown to improve survival for HCC include transarterial chemoembolization (TACE) and sorafenib [5,6] .
future science group
Resection & transplantation for hepatocellular carcinoma: factors influencing surgical options ●●Downstaging & neoadjuvant therapies
Neoadjuvant therapies in HCC are therapies that are given to patients that would otherwise could be treated with curative therapy to improve long-term outcomes, make definitive therapy safer, or bridge for transplantation. Neoadjuvant therapies include chemotherapy, TACE, intra-arterial radioembolization and ablation. For patients with resectable disease at presentation neoadjuvant intra-arterial therapies are not recommended due to risk of tumor progression or resultant liver failure [7,8] . Approximately 30–50% of patients on the waiting list receive neoadjuvant therapy in order to decrease dropout. Currently bridging neoadjuvant therapy is supported so long as it does not delay transplantation. In contrast to neoadjuvant therapy, downstaging refers to treatment in which the intent is to make an unresectable HCC resectable and has been accomplished in up to 24–90% of patients [9] . ●●Targeted therapies: sorafenib
After the approval of sorafenib as the standard of care for patients with advanced HCC, the Sorafenib Hepatocellular Carcinoma Assessment Randomized Protocol demonstrated an increased median survival and the time to radiographic by approximately 3 months [5] . The median overall survival (OS) was 10.7 months in the sorafenib group and 7.9 months in the placebo group (p < 0.001), while time to radiologic progression was 5.5 months in the sorafenib group and 2.8 months in the placebo group (p < 0.001). Currently the STORM trial, is testing sorafenib as an adjuvant therapy in 1100 patients with early HCC treated by resection or ablation to determine whether inhibiting kinase activity reduces the risk of HCC recurrence. Villanueva and Llovet review over 50 current molecular therapies under evaluation for HCC, which may lead to development of individualized targeted therapies [10] . Growing numbers
Review
of case reports describing the downstaging of unresectable HCC using sorafenib alone or in combination with locoregional therapy suggest promise in a minority of patients with advanced HCC [7,11–13] . Therapeutic decision-making in the treatment of HCC necessitates the consideration of both tumor- and liver-specific factors in the selection of treatment type (Table 1) . Liver-specific factors relate to both the quality and quantity of underlying liver. Although tumor-specific factors may be conducive to LR, unfavorable liver factors may preclude it – making LT a more attractive option. While a subset of noncirrhotic patients may tolerate a major hepatectomy, the degree to which patients with compromised hepatic function tolerate surgery will depend, in part, on the quality and quantity of the [14] . Cirrhosis & quality of the underlying liver ●●Child–Turcotte–Pugh classification
The quality of underlying liver can be assessed using the Child’s classification, the Model for End-Stage Liver Disease (MELD) score, and direct measurement of the portal venous pressure. The Child-Turcotte-Pugh classification (CTP) was developed as a tool to predict outcomes following shunt procedures. High CTP scores have been shown to correlate with high mortality following major abdominal surgery. Garrison et al. and Mansour et al., using the five parameters of CTP (total bilirubin, albumin, international normalized ratio, ascites and hepatic encephalopathy), found very comparable stratification of mortality following abdominal surgery, with dramatically increased mortality in CTP class B (CTP-B) and CTP class C (CTP-C) [15,16] . Specifically, Garrison et al. and Mansour et al. reported that patients who were CTP class A (CTP-A) had a 10% mortality, whereas mortality in patients with CTP-B and CTP-C liver function were 30–31% and 76–82%, respectively [15,16] . Current estimates
Table 1. Tumor- and liver-specific factors considered in treatment allocation. Tumor-specific factors
Liver-specific factors
Size Number Location Relation to adjacent structures Distant spread Vascular invasion Histologic grade
Cirrhosis/fibrosis Ascites Portal hypertension Thrombocytopenia Bilirubin Size of remnant liver Hepatitis etiology
future science group
www.futuremedicine.com
589
Review Dodson, He & Pawlik Table 2. Relationship between Model for End-Stage Liver Disease score and postoperative mortality. MELD score 0–7 (n = 351) 8–11 (n = 257) 12–15 (n = 106) 16–20 (n = 35) 21–25 (n = 13) ≥26 (n = 10)
Mortality % (patients at risk[n]) 7 days
30 days
90 days
1 year
5 years
10 years
1.9 (314) 3.3 (236) 7.7 (94)
5.7 (301) 10.3 (219) 25.4 (78)
9.7 (287) 17.7 (200) 32.3 (69)
19.2 (253) 28.9 (170) 45.0 (56)
50.7 (123) 58.5 (83) 69.5 (24)
72.6 (57) 78.1 (35) 87.3 (10)
14.6 (29) 23.0 (7) 30.0 (6)
44.0 (19) 53.8 (4) 90.0 (1)
55.8 (15) 66.7 (3) 90.0 (1)
70.5 (10) 84.6 (2) 100 (0)
94.1 (2) 92.3 (1) 100 (0)
94.1 (2) 100 (0) 100 (0)
MELD: Model for End-Stage Liver Disease. Reproduced with permission from [19].
of nontransplant operative mortality in patients with CTP-A, CTP-B, and CTP-C liver function are 5–10, 9–30 and 50–90%, respectively. As such, LR for HCC is largely relegated to patients who are CTP-A, whereas patients with HCC who are CTP-B or CTP-C are best served with LT. ●●MELD score
Although CTP has been shown to be highly predictive of mortality after resection of HCC, it is somewhat subjective and has been criticized as only an estimation of liver function. A subset of patients with presumably well-compensated/ preserved function designated as CTP-A can still develop significant hepatic decompensation after LR. For example, Bruix et al. found that >50% of CTP-A patients had signs of hepatic decompensation after resection of HCC and more than a third of these patients had not recovered by 90 days [17] . The MELD score has been shown to accurately predict long-term survival in patients with cirrhosis, as well as outcomes for cirrhotic patients undergoing resection for HCC and other major surgical procedures [18] . The MELD score was developed to predict the short-term outcome of patients undergoing the transjugular intrahepatic portosystemic shunt (TIPS) procedure for cirrhosis and is derived from international normalized ratio, total bilirubin and serum creatinine. The MELD score is more objective compared with the CTP classification, as it depends solely on quantifiable laboratory values. The MELD score has been used as a criterion in liver allocation by the United Network for Organ Sharing (UNOS) since 2002. In a large study of 772 cirrhotic patients, Teh et al. concluded that patients with a small HCC ≤5 cm and a MELD ≤8 tolerated LR well [19] . By contrast, the authors noted a MELD score ≥9 was a predictor
590
Future Oncol. (2014) 10(4)
of both increased perioperative mortality and worse long-term survival. Similarly, in their series, Cucchetti et al. reported that no patient with a MELD 1 cm from the tumor, these patients had similar DFS and OS as patients with macroscopic invasion. While macroscopic invasion can be determined preoperatively, microscopic invasion requires pathologic examination. Consequently, preoperative findings have been used to predict the presence of vascular invasion. Pawlik et al. definitely demonstrate that vascular invasion increases dramatically as tumor size increases
www.futuremedicine.com
595
Review Dodson, He & Pawlik
Tumors with microvascular invasion (%)
80 70 60
**
Low Intermediate
**
**
High
**
50 40
*
30 *
20 10 0 ≤3
3.1–5
5.1–6.5
>6.5
Tumor size (cm) Figure 6. Histologic grade stratified by tumor size and rate of microvascular invasion. Increasing tumor size was associated with microvascular invasion and higher grade. *p < 0.05 versus large (>5 cm) intermediate- or high-grade tumors. **p < 0.05 versus small (≤5 cm) intermediate- or high-grade tumors. Reproduced with permission from [28].
(Figure 6) [28] . Specifically, vascular invasion nearly doubled in tumors >5 cm versus 5 cm; in the same study the authors noted that histologic grade was the most significant predictor of occult vascular invasion in tumors >5 cm. Therefore, tumor grade determined by preoperative fine needle aspiration (FNA) has been proposed by some as a means to help define patients for transplantation. For example, Cillo et al. report a 5-year actuarial survival of 75% and DFS of 92% in patients undergoing LT in well-to-moderately differentiated HCC, despite 38% of patients being beyond the Milan criteria
596
Future Oncol. (2014) 10(4)
[37] . As such, careful consideration of patients for LT for tumors >5 cm should include criteria such as solitary tumor, low tumor grade and low AFP levels. The University of California San Francisco (UCSF) group has advocated for extended criteria for transplantation especially with regard to size when these other factors are considered [38] . ●●Tumor number
Multinodular HCC (MNHCC) may be due to intrahepatic metastasis lesions or multicentric carcinogenesis. MNHCC has been shown to be a major independent prognostic factor in patients undergoing LR [23,39,40] . Wang et al. found significant differences in the long-term survival of solitary HCC and MNHCC with 1-, 3-, and 5-year survival rates of 88.0, 69.2 and 58.4% versus 86.1, 55.5 and 29.9%, respectively (p < 0.001) [39] . The median 1-, 3-, and 5-year DFS rates for patients with MNHCC was only 45.7, 29.2 and 18.4%, respectively. Factors particularly associated with a poor outcome after LR included AFP >400 ng/ml, total tumor size >5 cm, largest tumor size >5 cm, total tumor number >3 and presence of microvascular invasion. Wang et al. concluded that LR provided a superior benefit to ablation or chemoembolization among those patients with MNHCC with tumor number ≤3 and largest tumor size ≤5 cm
future science group
Resection & transplantation for hepatocellular carcinoma: factors influencing surgical options [39] . However, a high recurrence has also been reported for patients undergoing LR versus LT for MNHCC (46 vs 27%, respectively) [41] . As such, due to high recurrence after LR, MNHCC is considered a relative contraindication to LR in some centers. LT should be considered a better therapeutic option for MNHCC within the Milan or extended UCSF criteria.
AFP AFP has been used as a serum tumor marker since the 1970s as greater than 75% of patients present with levels above 10 ug/l. Multiple groups have associated elevated AFP levels with large tumors, bilobar involvement and portal thrombus in addition to prognostic value [42–46] . The UCSF group found levels >1000 ng/ml with increased risk of recurrence after LT [38] . Ma et al. found that preoperative serum level of AFP ≤20 ng/ ml, AFP 20–400 ng/ml and AFP >400 ng/ml differentiated clinical and pathological factors [42] . Patients with levels ≤20 ng/ml had higher cell differentiation, lower rates of microvascular invasion and higher DFS. Whereas AFP levels >400 ng/ml conferred poorer postoperative survival. Duvoux et al. found that AFP levels independently predicted tumor recurrence and correlated with both vascular invasion and cellular differentiation in patients undergoing LT [47] . In fact, a subset with AFP levels ≤100 ng/ml had low risk of recurrences with survival near 70%, whereas patients with AFP levels >1000 ng/ml were at high risk of recurrence and reduced survival. Treatment allocation The BCLC staging systems combines tumor stage, degree of liver function, the patient’s general condition/performance status, cancerrelated symptoms and treatment efficacy. These factors are combined with a treatment algorithm to form the BCLC system, which, in turn, provides treatment recommendations. A key criterion in the BCLC staging system is performance status, which has been shown to be a comprehensive determinant of overall health status, an indicator of treatment benefit, and a predictor of long-term survival in cancer patients. The BCLC is often cited in the treatment of patients with HCC (Figure 7) [4] . In general patients at a very early stage (Stage 0) and early stage (Stage A) are candidates for surgical therapy. In the absence of portal hypertension and normal bilirubin, solitary tumors are evaluated for LR.
future science group
Review
Candidates are evaluated for transplantation if the patient has three nodules
