Journal of Gastrointestinal Surgery

, Volume 16, Issue 8, pp 1516–1523

Role of Operative Therapy in Non-cirrhotic Patients with Metastatic Hepatocellular Carcinoma


  • Victor Zaydfudim
    • Division of Gastroenterologic and General SurgeryMayo Clinic
  • Rory L. Smoot
    • Division of Gastroenterologic and General SurgeryMayo Clinic
  • Clancy J. Clark
    • Division of Gastroenterologic and General SurgeryMayo Clinic
  • Michael L. Kendrick
    • Division of Gastroenterologic and General SurgeryMayo Clinic
  • Florencia G. Que
    • Division of Gastroenterologic and General SurgeryMayo Clinic
  • Michael B. Farnell
    • Division of Gastroenterologic and General SurgeryMayo Clinic
    • Division of Gastroenterologic and General SurgeryMayo Clinic
Original Article

DOI: 10.1007/s11605-012-1918-2

Cite this article as:
Zaydfudim, V., Smoot, R.L., Clark, C.J. et al. J Gastrointest Surg (2012) 16: 1516. doi:10.1007/s11605-012-1918-2



We investigated the role of operative therapy in non-cirrhotic patients who developed metastatic hepatocellular carcinoma (HCC).


This retrospective cohort study included consecutive non-cirrhotic patients with metastatic HCC after a prior hepatectomy treated between 1990 and 2009. Patients were stratified by operative therapy (resection, ablation, transcatheter therapy). Kaplan–Meier analyses with log-rank comparisons tested effects of operative therapy on overall survival (OS) and progression-free survival (PFS).


Of 195 non-cirrhotic patients treated for HCC during the study period, 98 [median age 65, interquartile range (IQR) 53–71; 55 % male] subsequently developed metastatic HCC (55 intrahepatic only). Median time to development of metastases after the index operation was 10 months (IQR 5–20 months); median number of metastases was 3 (IQR 2–7). Half of these patients (n = 50) underwent operative treatment of metastases; 20 (40 %) underwent metastasectomy, 18 (36 %) ablation, and 12 (24 %) transcatheter therapy. Operative therapy was associated with improved OS (p < 0.001). Resection or ablation was associated with improved PFS and OS compared to transcatheter therapy (all p ≤ 0.006). Nine patients (seven resection, two ablation) are disease free at a median of 50 months (IQR 24–80 months) posttreatment.


Resection and ablation are associated with an improved PFS and long-term OS and should be considered in select patients with metastatic HCC.


Hepatocellular carcinomaRecurrenceMetastasisHCCMetastasectomySorafenibOutcomeResectionAblation


The role of operative metastasectomy in patients with hepatocellular carcinoma (HCC) remains controversial.13 While the epidemiology and predisposing factors associated with HCC vary with geography, majority of the patients with HCC worldwide harbor underlying chronic liver disease. Chronic inflammation with resultant chronic liver injury and cirrhosis contributes to the development of HCC in approximately 80 % of cases.4,5 Consensus guidelines for patients with hepatic dysfunction have been developed to allocate surgical treatment strategies based on both tumor extent and underlying liver disease.6 However, in patients with well-preserved liver function, resection remains the most appropriate and effective treatment.6,7

Recent studies have examined the primary surgical outcomes among patients with HCC and preserved liver function. While operative mortality is low, 50–58 % of the patients develop recurrent HCC at 5-year follow-up.79 The presence of multiple hepatic tumors as well as vascular invasion at the time of original resection has been associated with a greater risk of postoperative recurrence. Systemic therapy options for advanced HCC have been traditionally ineffective.1012 The recent addition of sorafenib to chemotherapeutic regiments has modestly improved survival among patients with advanced HCC.13

Operative treatment remains an option for patients with preserved liver function and metastatic HCC. Previous single-institution reports demonstrated a benefit of resection of metastatic lesions among selected patients with intrahepatic or extrahepatic, predominantly pulmonary, recurrences.1416 In our study, we aimed to examine our experience with operative treatment in patients with preserved liver function and recurrent intra- and extrahepatic metastatic HCC. We hypothesized that an aggressive operative approach will improve overall survival (OS) and progression-free survival (PFS) among patients with metastatic HCC.


Patients and Measures

The study protocol was reviewed and approved by the Mayo Clinic Institutional Review Board. All consecutive patients treated at the Mayo Clinic (Rochester, Minnesota) for intra- and extrahepatic HCC recurrence without clinical and/or histologic evidence of cirrhosis between January 1990 and May 2009 were included in this retrospective cohort study. Patients with fibrolamellar HCC variant and mixed HCC cholangiocarcinoma were excluded.

All relevant demographic and clinical outcome measures were reviewed, including age at the time of diagnosis of metastatic disease, sex, as well as primary tumor size, number, vascular invasion, stage, and resection margin. The American Joint Committee in Cancer 7th edition was used to stage the primary tumor based on T and N stage of the original resection.17 None of the patients had metastatic disease at the time of primary resection. Patients were followed routinely with cross-sectional imaging. Metastases were categorized based on location of recurrent disease: lung, bone, brain, peritoneal surface, lymph node(s), adrenal, omentum, and skin. Temporal association between metastatic presentation and original diagnosis was ascertained. Operative treatment was categorized based on surgical approach: (1) metastasectomy—liver, lung, or other metastatic sites; (2) ablation—radiofrequency (RFA), cryoablation, or percutaneous ethanol injection (PEI); or (3) transcatheter therapy—transarterial chemoembolization (TACE), transarterial bland embolization (TAE), or selective internal radiation therapy (SIRT). Nonoperative patients were treated with systemic therapy at the discretion of medical oncologist.

Clinical follow-up of the study cohort was updated on August 2011. Progression-free survival was calculated as time between operation for metastatic disease and documented progression of residual disease, recurrent disease, or patient death from any cause, whichever occurred first. Overall survival was calculated as (1) the time from the primary HCC diagnosis and death from any cause among the entire patient cohort, and (2) the time from the diagnosis of metastatic disease and patient death from any cause among the patients with recurrent HCC.

Statistical Analyses

Summary data are reported as median values with interquartile range (IQR) or percentages. Continuous variables were compared using the Wilcoxon rank sum test. Categorical variables were compared with Pearson’s chi-squared test or Fisher’s exact test, as appropriate, and expressed as proportions. Kaplan–Meier survival analysis with the log-rank test for between-group comparisons was used to quantify the effects of recurrence, operative management, and operative approach on OS and PFS.18 Probability of a type I error of <0.05 was considered statistically significant. A Bonferroni correction was applied to pairwise tests between individual survival curves of OS and PFS estimates to adjust for multiple comparisons among patients treated with operative therapy.19 Log-rank p value < 0.017 (0.05/3) was considered statistically significant for these pairwise comparisons. STATA 10.1 (Stata Corporation, College Station, TX) statistical software was used for data management and analysis.


Demographic and Clinical Covariates

One hundred and ninety-five non-cirrhotic patients were treated for HCC during the study period, of whom 50 %—98 patients (median age 65, IQR 53–71; 55 % male)—developed metastatic HCC. Median follow-up time for the entire cohort since the index HCC resection was 36 months (IQR 15–70 months). Median time to development of metastases among the 98 patients with metastatic HCC was 10 months (IQR 5–20 months). OS was significantly better among patients who did not experience postoperative HCC recurrence (Fig. 1), log-rank p < 0.001. Of the 98 patients with metastatic disease, 50 (51 %) underwent operative treatment, while 48 were managed with systemic therapy alone. Median age at the time of diagnosis with metastatic disease, sex, primary tumor size, vascular invasion, tumor stage, and margin of primary resection did not differ between patients who did and did not undergo operative therapy of metastatic disease (all p ≥ 0.227). Demographic and clinical data for this cohort are summarized in Table 1. Patients with metastatic disease treated with an operative approach had a greater disease-free interval between the time of primary resection and the development of recurrence of HCC (Table 1, p = 0.031).
Fig. 1

Worse overall survival associated with development of recurrent disease after primary resection of HCC

Table 1

Demographic and clinical covariates by operative treatment


Operative treatment (n = 50)

No operative treatment (n = 48)

p value

Age (years)a

64 (53–68)

66 (52–74)



26 (52)

28 (58)


Primary tumor size (cm)a

9.9 (6.5–13)

10 (6–14.3)


Number of primary tumorsa

1 (1–2)

1 (1–5)


Primary tumor vascular invasiona

14 (28)

12 (25)


Primary tumor stageb




23 (46)

16 (33)


6 (12)

4 (8)


10 (20)

17 (36)


8 (16)

8 (17)


3 (6)

3 (6)

Margin positive original resectionb

3 (6)

2 (4)


Time to development of metastases (months)a

16 (6–23)

7 (4–13)


aData reported as median (IQR)

bData reported as n (percent)

Characteristics of patients with HCC metastases are summarized in Table 2. The number of patients with liver metastases did not vary among patients in operative and nonoperative groups (p = 0.559); however, patients selected for an operation had fewer metastases with a smaller median size of individual metastatic lesions (both p ≤ 0.002). More patients in the operative group had intrahepatic recurrence without extrahepatic metastases compared to patients treated nonoperatively (72 vs. 40 %, p = 0.001). More patients with lung and lymph node metastases were treated with systemic therapy alone than selected for metastasectomy (both p ≤ 0.029). There was no difference in metastatic distribution of other involved sites (all p ≥ 0.088). The majority (88 %) of the patients selected for operative treatment had only one anatomic metastatic site; 12 % had two anatomic sites of metastatic involvement. Metastatic distribution was significantly different among patients treated with systemic therapy alone. In this group, 33 % of the patients had two and 11 % had three anatomic sites of metastatic HCC (p = 0.001).
Table 2

Characteristics of metastatic disease


Operative treatment (n = 50)

No operative treatment (n = 48)

p value

Liver metastases

39 (78)

35 (73)


Number of liver metastases

2 (1–4)

4 (2–11)


Size of largest liver metastasis (cm)

2.3 (1.5–3.2)

3.4 (2.3–5.6)


Intrahepatic recurrence without extrahepatic disease

36 (72)

19 (40)


Lung metastases

5 (10)

19 (40)


Peritoneal metastases

6 (12)

3 (6)


Bone metastases

2 (4)

7 (15)


Brain metastases

0 (0)

2 (4)


Lymph node metastases

1 (2)

7 (15)


Adrenal metastases

1 (2)

1 (2)


Omental metastasis

1 (2)

0 (0)


Skin metastasis

1 (2)

0 (0)


Number metastatic sites




44 (88)

27 (56)


6 (12)

16 (33)


0 (0)

5 (11)

Data reported as n (percent)

Operative Treatment of Metastatic Disease

Operative treatment of patients with metastatic disease is summarized in Table 3. Twenty patients underwent metastasectomy: 10 liver, 3 lung, 7 other metastatic sites (4 patients with peritoneal metastases, 1 omentum, 1 adrenal, and 1 skin). Margin-negative (R0) resection was achieved in all 10 patients undergoing hepatic metastasectomy and all 3 patients treated with lung metastasectomy. Among patients undergoing extrahepatic and extrapulmonary metastasectomy, five patients underwent an R0 resection, while two patients had residual microscopic HCC (R1 resection). Eighteen patients underwent ablation (13 RFA, 1 cryoablation, and 4 PEI), and 12 were treated with transcatheter therapy (6 TACE, 1 TAE, 5 SIRT). Patients treated with transcatheter therapy had more intrahepatic tumor burden compared to patients treated with resection or ablation. The transcatheter treatment group had a median of 5 liver metastases (IQR 3–7 metastases), compared to 2 liver metastases (IQR 1–3) among the resection group and 1 liver metastasis (IQR 1–2) among the ablation group (both p ≤ 0.005). Perioperative mortality was zero. One patient died as a result of an ischemic biliary stricture with subsequent recurrent cholangitis and biliary sepsis 5 months after TACE of the remnant left lateral segment which contained two metastatic lesions.
Table 3

Operative treatment of metastatic HCC


Operative treatment (n = 50)











Radiofrequency ablation




Alcohol ablation


Transcatheter therapy


Transarterial chemoembolization


Transarterial bland embolization


Selective internal radiation therapy


aOther includes: four peritoneal metastases, one omentum, one adrenal, one skin

Overall and Progression-Free Survival

OS was better among patients treated with operative intervention compared to systemic therapy alone (Fig. 2, log-rank p < 0.001). Patients treated with metastasectomy or ablation had better OS (Fig. 3a) and PFS (Fig. 3b) compared to patients managed with transcatheter-based therapies (both log-rank p ≤ 0.001). Pairwise median, 1-year, and 5-year comparisons for both OS and PFS among patients treated with metastasectomy, ablation, and transcatheter therapies are summarized in Table 4. After adjusting for multiple pairwise comparisons, patients treated with either resection or ablation had better PFS and OS compared to transcatheter therapy (all log-rank p ≤ 0.006). There was no difference in PFS and OS between patients treated with resection (n = 20) or ablation (n = 18; both log-rank, p ≥ 0.621). Median time to progression for patients treated with resection or ablation was 17 months, compared to 5 months for patients treated with TACE, TAE, or SIRT (both log-rank p ≤ 0.003, Table 4). OS was better among patients treated with transcatheter therapy compared to systemic therapy alone (median OS 21 vs. 9 months, log-rank p = 0.004). OS did not differ between operatively treated patients with liver only metastases and patients with extrahepatic metastatic HCC (p = 0.427). Among the 40 patients with HCC progression after treatment of metastatic disease, sites of disease progression were: 32 liver, 14 lung, 6 bone, and 1 peritoneum. Nine patients (seven resection, two ablation) are disease free at a median of 50 months (IQR 24–80 months) follow-up.
Fig. 2

An improvement in overall survival associated with operative treatment of metastatic HCC compared to systemic therapy alone
Fig. 3

a An improvement in overall survival associated with resection or ablation of metastatic HCC. b An improvement in progression-free survival associated with resection or ablation of metastatic HCC

Table 4

Median, 1-year, and 5-year overall and progression-free survival among patients treated with resection, ablation, or transcatheter therapy


Median (months)

1 year (%)

5 years (%)

p value

Progression-free survival



 Resection vs. ablation

17 vs. 17

74 vs. 56

21 vs. 14


 Resection vs. transcatheter

17 vs. 5

74 vs. 0

21 vs. 0


 Ablation vs. transcatheter

17 vs. 5

56 vs. 0

14 vs. 0


Overall survival



 Resection vs. ablation

54 vs. 76

95 vs. 94

50 vs. 72


 Resection vs. transcatheter

54 vs. 21

95 vs. 67

50 vs. 0


 Ablation vs. transcatheter

76 vs. 21

94 vs. 67

72 vs. 0


aBonferroni-corrected log-rank p value ≤0.017 (0.05/3) is considered statistically significant to adjust for multiple pairwise comparisons.


Currently, standard treatment for metastatic HCC remains controversial. Recent consensus recommendations suggest systemic therapy with sorafenib, supportive care, or a clinical trial as care pathways for patients with metastatic disease.1 Historically, systemic therapy has been largely ineffective in treatment of advanced HCC.1012,20 While sorafenib has been shown to improve OS and PFS in recent phase II and III trials among patients with advanced disease, the improvements in survival have been modest with a median survival benefit of 3–6 months.13,21 Combination therapy with sorafenib in addition to traditional systemic therapy or TACE has, so far, resulted in substantial toxicity without improved survival.2123 Sorafenib was approved for treatment of patients with unresectable HCC in November 2007; as such, it was not available during majority of our study period. While we use sorafenib for patients with unresectable HCC, we do not currently recommend routine sorafenib administration to patients with margin-negative resection of primary or metastatic HCC.

Additional operative therapy remains a potentially attractive option for selected patients with metastatic HCC. Previous single-institution studies have suggested an improvement in overall survival among patients with intrahepatic recurrence with repeat hepatic resection, ablation, or TACE when compared to nonoperative treatment.2427 In addition, extrahepatic metastasectomy has been described for patients with pulmonary and intra-abdominal metastases from HCC.16,28,29 The results of our study corroborate these findings. Our findings that patients treated with resection or ablation of intrahepatic recurrence had significantly better survival rates compared to patients treated with TACE or systemic therapy alone are similar to the recent findings published by others.27 As expected, patients selected for resection or ablation have a lesser tumor burden compared to patients treated with transcatheter-based therapies. In our study, the median number of metastases in the liver remnant was significantly greater for patients treated with TACE, TAE, SIRT, or systemic therapy compared to patients treated with metastasectomy or ablation. Despite the differences in the tumor burden, however, transcatheter-based therapies were associated with an improved OS compared to systemic therapy alone.

Our study was specifically limited to patients without cirrhosis. None of the patients in our study were listed for liver transplantation after diagnosis of metastatic disease. Liver transplantation is the preferred treatment for patients with compromised liver function and HCC. While the median size of the primary HCC within our study population was 10 cm, tumor downstaging has been described among patients with liver dysfunction, but outside of the Milan criteria.6,30 Despite strict selection criteria, recurrent HCC after transplantation occurs in approximately 10–30 % of the patients, with isolated hepatic recurrence in 15–20 % of the liver transplant recipients. Surgical re-resection and ablation of metastatic HCC are established as standard treatment among transplant recipients.31 Overall 5-year survival among these patients treated with re-resection or ablation after liver transplantation was 55 months and is similar to 5-year survival of 50 months for resection and 72 months for ablation groups in our study.

Characteristics of the primary HCC, such as tumor size, vascular invasion, and stage, did not vary between patients with metastatic HCC treated operatively from those patients treated with systemic therapy alone. Tumor burden among those patients who were offered an operative approach for metastatic disease, however, was considerably less. This patient group had smaller, fewer lesions compared to patients treated nonoperatively and were more likely to present with isolated intrahepatic metastases without extrahepatic disease. In addition, the disease-free interval among patients with metastatic disease undergoing an operative intervention was approximately 9 months greater than among patients treated with systemic therapy alone. Despite the described heterogeneity, our data suggest a benefit to operative treatment, in particular with use of resection or ablation, in surgical candidates.

The current study has several limitations. First, our study was retrospective and included patients treated under different adjuvant therapy protocols for metastatic HCC. Nevertheless, our survival rates are in congruence with the previously published reports demonstrating a benefit to an aggressive operative intervention for selected patients with potentially resectable metastatic HCC. In addition, the selection bias among patients with metastatic disease treated with an aggressive operative approach has been well described. These patients are likely to have less aggressive tumor biology with an associated reduction in aggressive tumor behavior when compared to patients who were offered systemic therapy without additional operative treatment. As we have described, tumor burden and disease-free interval were more favorable among patients treated with resection, ablation, or transcatheter therapy. Selection bias is difficult to control among a small population of heterogeneous group of patients with differential tumor burden and temporal distribution. Nevertheless, despite the possibility of selection bias, our data demonstrate the possibility of long-term survival among a select group of patients with metastatic HCC.

In conclusion, operative treatment can be performed safely in selected, non-cirrhotic patients with potentially resectable metastatic HCC. Operative therapy, including transcatheter modalities, is associated with an improved overall survival compared to systemic therapy alone. Operative resection and ablation of metastatic disease result in equivalent PFS and OS and provide greater PFS and OS than transcatheter approach. Despite an aggressive operative approach, success of sorafenib demonstrates the need for ongoing development of targeted molecular therapies to improve treatment options for patients who will benefit from operative therapy.

Copyright information

© The Society for Surgery of the Alimentary Tract 2012