The exact place of laparoscopy in liver surgery is not clearly determined (5). Surgery should begin with a careful exploration of the abdominal cavity to rule out peritoneal carcinosis or an unexpected bilobar involvement of the liver by metastases, which could be a contraindication for resection. The presence of metastatic lymph nodes in the porta hepatis and the coeliac region considerably worsens the prognosis but should not be considered as an absolute contraindication to resection if they can be completely removed because 5-year recurrencefree surviving patients have been reported in such cases (6). Intraoperative ultrasound (IOUS) should be performed in every case because it allows a precise mapping of the anatomical relations of the metastases to the main intraparenchymatous vascular pedicles and helps select the type of resection. IOUS can detect small intraparenchymatous lesions and thereby modify the extent of the initially planned operation (7). It may also be used to guide the fine-needle biopsy of doubtful lesions or to evaluate the degree of destruction of a metastasis treated by radiofrequency (RF) ablation.

If remnant liver parenchyma is normal, up to six of the eight anatomical segments can be resected without inducing postoperative liver failure. Liver resections can be divided in two groups: (a) anatomical resections removing one or several segments, and (b) atypical or wedge resections removing a portion of liver parenchyma surrounding a hepatic lesion. Resections removing three or more continuous segments are defined as major hepatic resections: right hepatectomy (segments V, VI, VII, and VIII); left hepatectomy (segments II, III, and IV); and extended right hepatectomy, also called right lobectomy (segments IV, V, VI, VII, and VIII) (8,9).

The aim of carcinologic surgery for liver metastases is to remove or destroy with ablation procedures all metastatic sites with a free clearance margin. The type of liver resection depends on the size, number, and location of the metastases; their relation to the main vascular and biliary pedicles; and the volume of the liver parenchyma that can be left in place after surgery. Superficial small metastases can be resected with wedge resections. Larger lesions often require major resections. It should also be kept in mind that a large resection may preclude further surgery in the case of intrahepatic recurrence.

If synchronous metastases are discovered at the same time as the primary cancer, it is usually preferable to perform the bowel resection during a first procedure because combined resection of both primary and liver metastases is associated with an increased mortality and morbidity rate (intra-abdominal fluid infection, vascular clamping deleterious for the viability of digestive sutures) when hepatic resection is a major resection (6). Usually, surgical resection of the liver metastases is delayed for 2 to 4 months after bowel resection, allowing observation of the response of liver metastases to systemic chemotherapy, which is often administered during the interval, and constituting an important prognostic factor.

In most recent studies, in-hospital mortality rates vary from 0% to 5% and are strongly influenced by perioperative blood loss, preoperative liver function, and extent of liver resection (Table 46.1). Reversible postoperative complications are observed in 25% to 40% of patients. Morbidity after hepatic resection is usually due to transient liver failure, hemorrhage, subphrenic abcesses, or biliary fistula. The mean hospital stay after liver surgery ranges from 10 to 15 days in the absence of complications.

Liver resection of colorectal metastases is associated with 3- and 5-year survival rates close to 40% and 30%, respectively (Table 46.2). After resection, recurrences are observed in two-thirds of patients and involve the liver in 50% of cases. In a large retrospective survey, 5-year survival was 28% in 1,568 patients who had a resection of isolated colorectal liver metastases and 15% in 250 patients who had resected liver and extrahepatic metastases. None of the 77 patients who had a palliative resection survived for 5 years (18).

Several studies have assessed factors influencing survival. Gender and the site of the primary tumor do not seem to influence the outcome. The stage of the primary tumor is associated with 5-year survival rate of 70% in stage I or II CRCs and 33% in stage III (6). Prognosis seems better in cases of metachronous metastases and small lesions, and when there are less than four lesions but the involvement of one or both lobes does not influence the outcome. CEA level is strongly correlated to the recurrencefree survival. A free margin of ≥1 cm offers better chances of avoiding recurrence, but several series have shown that a smaller margin did not affect survival (19). The type of resection does not seem to influence the prognosis provided that a clear margin is obtained. Blood transfusions could be associated with an adverse outcome but may reflect the surgical difficulties for the resection of large and numerous lesions. In large retrospective series of 1,568 patients with resected liver metastases from carcinoma, a multivariate analysis showed that age, size of the largest metastasis, CEA level, stage of the primary tumor, diseasefree interval, number of liver nodules, and resection margin >1 cm or <1 cm were only independent prognosis factors (18).

Preoperative chest radiograph combined with a CT scan is performed to detect lung metastases. Synchronous liver and lung metastases should not be considered a contraindication to hepatic resection, provided that both sites can be completly resected. The carcinologic principle for resection of lung metastases is similar to that for liver metastases. The primary tumor should have been totally resected with no evidence of local recurrence or other unresectable metastases. In one study, 239
patients were operated on for lung metastases of CRC, 43 (18%) had previously had synchronous liver metastases surgically resected. Seven patients (16%) underwent subsequent lung resection for recurrences, and the median survival from lung resection was 19 months (20). Similar results were reported by the Metastatic Lung Tumour Study Group of Japan, with 47 patients who underwent pulmonary and hepatic resection with 3-, 5-, and 8-year survivals of 36%, 31%, and 23%, respectively (21). Surgical resection of lung metastases can significantly prolong survival. Prognostic factors are similar to those associated with resection of liver metastases; age, gender, and type of resection have not not been shown to have an impact on survival. After resection of lung metastases, the lung is the first site of recurrence in 50% to 70% of cases, followed by locoregional recurrences at the site of the primary metastasis, and brain and liver metastases. Repeat lung resections can be considered in some cases because 5-year actuarial survival rates of 30% have been reported (22). Exploration of other sites of possible metastasis, such as brain by CT scan or bone by scintigraphy, is performed only if there is a clinical suspicion. Their presence is a contraindication to liver or lung resection because the prognosis depends on the evolution of these unresectable metastases.