Surgical resection, an effective treatment for metastatic liver cancer, is considered likely to benefit patients in the overall course of the disease process. Surgical resection can only be considered when the following requirements are met:

The 5-year survival rate of patients undergoing resection for metastatic liver cancer is between 25 and 49 % [2–4]. Recurrence and metastasis occurs in 70–80 % patients undergoing operation despite the fact that surgical resection is considered the most effective option among available therapies. Reoperation can be considered in recurrent cases when surgical requirements are met. Patients with recurrent metastases in the remnant liver, if operated, can obtain outcomes comparable to those resulting from initial hepatectomy [5, 6]. Compared with initial hepatectomy, the incidence of complications and hospital death after second hepatectomy is not significantly different, although prolonged operation time and increased bleeding volume are likely to occur during the second hepatectomy.

Past conventional beliefs hold that the presence of metastases in each hemiliver, the existence of more than three metastatic nodules, a diameter of metastases beyond 5 or 10 cm, and the coexistence of extrahepatic metastases are each contraindications for hepatectomy [7, 8]. However, these are not currently considered absolute contraindications. The assessment of patient safety prior to operation and the potential success of radical resection should be given priority when surgical resection is considered in suitable cases. According to Vauthey et al. [9], surgical resection for metastatic liver cancer in non-cirrhotic patients is feasible when liver function is normal, and 25–30 % of remnant liver by volume can be preserved.

Extrahepatic metastasis has long been viewed as a contraindication for surgical resection. In 2003, Elias et al. [10] reported that negative resection margin (R0) occurred in 77 of 111 (69 %) of patients who underwent surgical resection for metastatic liver cancer originating from the colon or rectum while concomitantly undergoing operations for extrahepatic metastases. The 5-year survival rate was reported to be 29 %, suggesting that extrahepatic metastases do not compromise 5-year survival outcome. Surgical resection is generally thought to be suitable in patients with metastatic liver cancer and extrahepatic metastases when R0 is predicted and extrahepatic metastases are responsive to chemotherapy as well as postoperative treatment is effective and available. However, the outcome of patients with hilar lymph node metastasis, especially in cases with retroperitoneal lymph nodes involvement, is worse than single extrahepatic metastasis. It is controversial whether surgical resection is contraindicated in patients with more than three metastatic nodules or whether an upper number limit of nodules should be used as a contraindication to surgery. Some argue that the number of metastatic nodules is a significant predictor, claiming that patients with three or four metastatic nodules are inappropriate for surgical resection because hepatectomy would elevate mortality and reduce 5-year survival rate. However, some research indicates that the number of metastatic nodules is unrelated to patient outcome. Moroz et al. [11] reported that 5-year survival rate was not significantly different between 22 cases with 4–7 nodules (39 %) and 91 cases with 1–3 nodules (30 %) among 123 patients undergoing hepatectomy for metastatic liver cancer from the colon or rectum. This study also found that local infiltration and recurrence after operation are independent of the overall outcome of patients with metastatic liver cancer. Several research studies with large sample sizes have demonstrated that mortality and the incidence of complications after surgery do not increase as the number of metastatic nodules grows [12, 13]. These studies suggest that three or more than three metastatic nodules are not an absolute contraindication to surgical resection.

In conclusion, hepatectomy for metastatic liver cancer is indicated only when the primary tumor has already been removed or is able to be removed simultaneously. Advanced age alone is not a contraindication. Senile patients with cardiac and lung comorbidities cannot tolerate surgical trauma and should therefore not be considered for hepatectomy. A presurgical understanding of the size, position, and blood supply of the tumor, as well as the relationship of the tumor with hepatic conduits, facilitates establishing an accurate surgical strategy.

No consensus has been reached as to whether hepatectomy should be performed simultaneously with or following the operation for the primary tumor when metastasis and primary lesion are detected concurrently. Advocates for concomitant hepatectomy believe that patients who postpone hepatectomy risk losing the opportunity for surgical resection because hepatic metastases may transfer again. In contrast, advocates for postponed hepatectomy believe that mortality and the incidence of complications will increase and that removal of micrometastases may be incomplete when concomitant hepatectomy is performed. Postponed hepatectomy is recommended to be performed 2–3 months after the operation of the primary tumor, due to the belief that concomitant hepatectomy would increase surgical complications and mortality [14]; furthermore, there may be concealed micrometastases that make scheduled concomitant hepatectomy and the goal of radical cure unachievable [15]. Growing evidence indicates that concomitant hepatectomy is as safe as staged hepatectomy [16] and leads to comparable outcomes [17] with current improved surgical technique and detection methods. For this reason, the timing of surgical resection should depend on patient conditions such as physical tolerance for undergoing surgical procedure, the sites and size of primary and metastatic cancer lesions, and whether incision facilitates exposure during hepatectomy.

For patients whose hepatic metastases come from gastrointestinal tumors and who are suitable for concomitant hepatectomy, hepatic metastases should be removed prior to the removal of the primary gastrointestinal lesion. This is indicated for several reasons. First, the low central venous pressure that is usually adopted to decrease bleeding during hepatectomy impairs splanchnic blood flow. To decrease the duration of low perfusion to gastrointestinal anastomoses, gastrointestinal surgery is performed directly following hepatectomy, thus allowing for the rapid recovery of central venous pressure after these procedures have been successfully performed. When the gastrointestinal procedure is performed before hepatectomy, hilar occlusion can result in gastrointestinal anastomotic engorgement that impairs anastomotic healing. Additionally, hepatectomy is not as prone to contamination as gastrointestinal surgery, and surgical instrument replacement is therefore not required when moving from a hepatic to gastrointestinal surgical site. Several incision sites can be considered in concomitant hepatectomy, including a median incision between the xiphoid process and pubic symphysis and a subcostal oblique incision combined with a right paramedian incision.

Arguments exist as to whether negative resection margin (R0) should be achieved in surgical resection for metastatic liver cancer. It is highly debatable whether R0 is required and what the optimal distance between the cutting edge and the tumor in hepatic metastases resection should be if R0 is determined to be necessary. It is generally believed that R0 contributes to reduced recurrence of cutting edge and intra- and extrahepatic metastasis, as well as contributing to improved outcome overall. However, current studies show that positive resection margin does not correspond with an increased risk of cutting edge and intra- and extrahepatic recurrence, nor does positive resection margin relate to recurrence-free survival and overall survival [18]. Some believe that a margin greater than 1 cm can improve survival, while others argue that the difference in outcome between short and distal margin cases is insignificant when negative resection margin is assured. As a result, a margin <1 cm should not be viewed as a contraindication on preoperative evaluation [19]. Further studies are required to determine the minimal margin that guarantees a negative resection margin.