Sarcoma

One of the largest surgical series of patients with sarcoma metastatic to the liver was composed of 56 patients who underwent liver resection (DeMatteo et al, 2001). These patients were selected from 331 patients with liver metastases from sarcomas who had been admitted to MSKCC during the years 1982 through 2000. Gastrointestinal stromal tumors (GISTs; Fig. 81C.4) and leiomyosarcomas were the most common histology; no perioperative deaths were reported in patients undergoing complete resection of the tumor, and the 5-year overall survival rate was 30% with a median of 39 months in completely resected patients. Patients who did not undergo complete resection had a 5-year survival of only 4% (Fig. 81C.5). A disease-free interval of less than 24 months was a significant adverse prognostic parameter for survival on univariate and multivariate analysis (Fig. 81C.6).

FIGURE 81C.4 Right hepatectomy for a liver metastasis from gastrointestinal stromal tumor.

FIGURE 81C.5 Disease-specific survival of patients with liver metastases from sarcoma who underwent complete resection (upper line, n = 56) versus other treatment (lower line, n = 275; P = .0001).

(From DeMatteo RP, et al, 2001: Results of hepatic resection for sarcoma metastatic to the liver. Ann Surg 234:540-548.)

FIGURE 81C.6 Disease-specific survival of patients after complete resection of liver metastases from sarcoma with a disease-free interval of more than 2 years (upper line, n = 32) versus less than 2 years (lower line, n = 24; P = .002).

(From DeMatteo RP, et al, 2001: Results of hepatic resection for sarcoma metastatic to the liver. Ann Surg 234:540-548.)

In a report from Pawlik and colleagues (2006b), 66 patients undergoing liver resection and/or ablation were included; the 5-year disease-free and overall survival rates were 16.4% and 27.1%, respectively. Patients in whom RFA was part of the treatment and patients who did not receive chemotherapy, mostly imatinib mesylate for metastatic GISTs, showed a significantly reduced survival.

The treatment strategy for patients with liver metastases from GISTs has changed since the development of the targeted agent imatinib mesylate, which achieves dramatic tumor response rates (Corless et al, 2004; DeMatteo, 2002) such that imatinib is now the first-line treatment. Resection is considered for patients when they reach the maximal response to imatinib, if all gross tumor can be removed and for those who have immediate or limited acquired resistance to the drug (Antonescu et al, 2005). Patients with mulitfocal resistance to imatinib seem not to beneftit from surgical intervention (DeMatteo et al, 2007).

Breast Cancer

In a small proportion (4% to 5%) of patients with breast cancer, liver metastases are the only sign of disseminated disease (Fig. 81C.7; Elias et al, 2003). Although it has not been formally proven that liver resection prolongs survival for selected patients with liver metastases of breast cancer, 5-year survival was 33% to 61% in more recently published series (Caralt et al, 2008; Elias et al, 2003; Vlastos et al, 2004). Patients with liver metastases from breast cancer receiving chemotherapy only rarely, if ever, survive 5 years (Wyld et al, 2003). Even with aggressive systemic chemotherapy, the median survival for patients with isolated liver metastases of breast cancer was only 23 to 27 months, with time to progression of 8 to 10 months in two more recent trials (Atalay et al, 2003). Resection for liver metastases in this setting may prolong survival for a subset of highly selected patients (Singletary et al, 2003).

FIGURE 81C.7 Liver metastasis of breast cancer. A, Computed tomographic scan shows the lesion (arrow) to have early peripheral contrast enhancement. B, The patient underwent resection of segments II and III; a gross photograph of the sectioned specimen is shown.

Table 81C.3 summarizes the published studies of liver resection for metastatic breast cancer. Meaningful comparisons between these studies cannot be made, because they are heterogeneous studies reporting small patient numbers with different inclusion criteria and treatment strategies. Valid criteria for selecting patients who might benefit from hepatic resection for metastases of breast cancer cannot be defined at this time. A common recommendation is that patients with extrahepatic disease should not undergo liver resection, even though some studies did not find a worse prognosis for such patients. Some authors also suggest that patients should first undergo systemic chemotherapy, and that only patients who do not progress should undergo liver resection. Formal randomized trials confirming this strategy have not been performed. Recently, a report regarding radiofrequency ablation (RFA) has been published; median survival for included patients was 30 months, but the follow-up time was only 19 months, and the local tumor progression rate was 25% (Meloni et al, 2009).

Melanoma

Melanoma recurs after potentially curative resection in approximately one third of patients, with almost every organ being at risk (Allen & Coit, 2002). Hepatic metastases are diagnosed in about 10% to 20% of patients with stage IV melanoma, even though it is well documented that most patients have liver metastases on postmortem examination. Patients with liver metastases as the initial site of metastatic disease have a very poor prognosis, with a median survival of 4.4 months (Rose et al, 2001). Most patients have unresectable disease, owing to extrahepatic disease or disseminated hepatic metastases. This point was shown by a study from the John Wayne Cancer Institute and the Sydney Melanoma Unit (Rose et al, 2001). During the years 1971 through 1999, 26,204 patients with melanoma were seen in these institutions, and 1750 patients (6.7%) had liver metastases. Only 34 patients underwent surgical exploration for attempted liver resection, and hepatectomy was performed in 24 patients. Of these 24 patients, 12 had synchronous extrahepatic disease, and 18 patients could be rendered disease free surgically. The 10 patients who underwent exploration only had a median survival of 4 months; the overall survival of all patients with liver metastases treated nonoperatively was 6 months. Overall survival in the patients who underwent complete resection was 28 months, with a median disease-free survival of 12 months. One patient achieved 10-year survival, and the 5-year overall survival was 29%. Complete gross resection and histologically negative margins were associated with an improved disease-free survival, and complete gross resection showed a trend for improved overall survival (P = .06).

From this experience, one might conclude that extrahepatic disease per se is not a contraindication for resection, as long as all disease is resectable. Selection of patients is crucial, which is well documented in this study. Only 1.4% of all patients with liver metastases underwent hepatic resection, and these patients had a median disease-free interval of 58 months before presenting with hepatic metastases. Several other small series in the literature mostly comprise fewer than 10 patients undergoing liver resection for metastatic melanoma, with median survival times of 10 to 51 months (Herman et al, 2007; Rose et al, 2001). Repeat hepatic resection might be beneficial for some patients who can be rendered disease-free surgically (Mondragon-Sanches et al, 1999).

Gastric and Pancreatic Cancer

Most reports on liver resection for NCNN metastases of gastrointestinal origin are based on patients with gastric cancer. Patients undergoing hepatic resection for metastatic gastric cancer are highly selected, which is apparent in the report of Ochiai and colleagues (1994), who treated 6540 patients with gastric cancer. Liver resection for metastases was performed in only 30 of these patients (0.46%). Table 81C.4 summarizes more recently published data; there have been few long-term survivors.

Only a handful of long-term survivors have been reported after the resection of liver metastases from pancreatic ductal adenocarcinoma (Detry et al, 2003). In a report from Shrikhande and colleagues (2007), 11 patients underwent combined pancreatic and liver resection for metastatic pancreatic cancer, with a median survival of 11.4 months. A recent review on this topic summarized 103 patients who underwent liver resection for metastastic pancreatic cancer; median survival ranged from 5.8 to 11.3 months (Michalski et al, 2008). Consequently, surgery remains highly controversial and unlikely to benefit most of these patients. For other primary malignant tumors of the pancreas that show a less aggressive tumor biology, such as solid pseudopapillary tumor of the pancreas, resection of liver metastases might be justified (Martin et al, 2002).

Renal Carcinoma

About 10% of patients with renal tumors develop liver metastases, and they have a dismal prognosis: fewer than 10% survive beyond 1 year, and only about 2% to 4% develop hepatic disease that is amenable to complete resection. Several series in the literature report fewer than 10 patients undergoing liver resection for metastatic renal tumors, with survival exceeding 5 years in about 13% of patients (Alves et al, 2003). In one series, 14 patients with liver metastases from primary renal tumors underwent partial hepatectomy (Alves et al, 2003). Median survival was 26 months, with a 3-year survival of 26%. Patients with a disease-free interval less than 24 months and patients with a diameter of the metastases exceeding 5 cm showed a worse prognosis. In patients with recurrence after hepatic resection, survival was better for patients who could undergo a repeat hepatectomy. In a more recent report on 31 patients, 5-year overall survival was 39% for the whole patient group and 50% for margin-negative patients (Thelen et al, 2007). In patients with hepatic metastases from renal tumors who are candidates for a complete resection, surgical exploration may be justified. Other therapies should also be considered, such as hepatic artery embolization and targeted molecular therapy, such as with sunitinib.

Reproductive Tract Tumors

Effective chemotherapeutic regimens are available for most reproductive tumors. Resection is only one component of a multimodal approach to the treatment of liver metastases from these tumors. The development of liver metastases is a well-defined adverse prognostic factor for patients with germ cell tumors (Gholam et al, 2003). Rivoire and associates (2001) attempted to define guidelines for the resection of liver metastases from germ cell tumors. These authors examined 37 patients who had undergone liver resection for metastatic germ cell tumors. All patients had received cisplatin-based chemotherapy before surgery. Median survival was 54 months with an overall 5-year survival rate of 62%. The authors defined three prognostic factors associated with a worse outcome: 1) pure embryonal carcinoma in the primary tumor, 2) liver metastasis greater than 3 cm, and 3) presence of viable residual disease after chemotherapy. Because no patient with liver tumors less than 1 cm had viable disease, the authors recommended a nonsurgical approach for these patients. Men with liver tumors greater than 3 cm in diameter represent a high-risk group that may not benefit from partial hepatectomy, but resection was recommended for the other subgroups.

Hahn and coworkers (1999) presented data regarding 57 patients undergoing liver resection for metastatic testicular cancer after systemic chemotherapy. In 48 patients, concomitant cytoreductive procedures for extrahepatic disease were performed. The overall 2-year survival rate was 97.1%. Pathologic analysis of resected specimens showed either a benign lesion or only necrotic tumor in 58% of specimens. Three of five patients with active disease and persistently elevated serum markers died during follow-up, underlining the importance of response to chemotherapy as a predictor of outcome.

Patients with metastatic ovarian or fallopian tube carcinoma usually do not present with isolated liver metastases, because the disease is generally diffuse within the abdomen and pelvis. Cytoreductive surgery that reduces disease to less than 1 cm when combined with chemotherapy is an accepted treatment approach. For these diseases, liver resection may be necessary because of true liver metastases or secondary liver involvement by peritoneal disease. A median overall survival of 62 months after hepatic resection has been described with this approach in 24 patients, with 18 patients having extrahepatic disease at the time of hepatectomy (Yoon et al, 2003). In this study, complete resection of all gross disease was possible in 21 patients, whereas in three patients, tumor debulking to less than 1 cm was performed. Merideth and colleagues (2003) reported 26 patients who underwent liver resection for metachronous metastases from ovarian carcinoma; cytoreduction was suboptimal (residual tumor = 1 cm) in five patients. Median disease-related survival was 26.3 months, and a disease-free interval exceeding 12 months and optimal cytoreduction were associated with improved outcome. Other reports also justify hepatic resection in metastatic ovarian carcinoma (Bristow et al, 1999; Chi et al, 1997; Lim et al, 2009). Unlike most other tumors, ovarian cancer is usually confined to the liver surface, and although extensive involvement can be seen, intraparenchymal hepatic metastases are uncommon (Fig. 81C.8).

FIGURE 81C.8 Axial computed tomographic image of a large ovarian metastasis to the left liver. Note that the tumor is extending into the liver from without, with no penetration is evident through the liver capsule.

Liver resection for metastases from cervical and endometrial cancer have been reported in the literature with an overall survival of 7 to 50 months (Kollmar et al, 2008; Tangjitgamol et al, 2004). Selected patients may benefit from hepatectomy; however, because of the small number of published cases, no general conclusions can be drawn from the available data.

Other Primary Tumors

Resection of liver metastases of lung cancer has been reported, and in selected patients, long-term survival has been achieved. Di Carlo and associates (2003) summarized the available data from the literature. Liver resection was performed in 14 patients with liver metastases from lung cancer, and two patients lived longer than 5 years.

A recent report summarized the results of hepatic resection for metastatic squamous cell carcinoma from various primary sites (anus, head/neck, lung, esophagus, and others) (Pawlik et al, 2007). The median overall survival was 22 months, with synchronous disease, metastasis size greater than 5 cm, and positive surgical margins being adverse prognostic parameters.

Patients presenting with liver metastases from an unknown primary tumor are a challenge to manage, because median overall survival is approximately 5 months. Liver resection or ablative therapy might be appropriate for some patients, in whom all disease can be destroyed or removed, but a median disease-free survival of only 6.5 months has been reported (Hawksworth et al, 2004).