All patients with potentially resectable liver lesions referred to our institution from December 2005 to December 2007 were considered for inclusion in this study. Forty-six were enrolled but 32 were excluded for various reasons: multiple bilobar CLM (n = 13), HCC with ascites and/or total bilirubin level higher than 2.0 mg/dl (n = 15) and patients' refusal (n = 4) and 3 patients scheduled for liver resection were referred to high-volume liver centers due to the complexity of the planned surgical strategy.
Eleven consecutive patients thus underwent curative liver resection at our institution and were the subjects of the present study. There were 8 males and 3 females; median age was 74 years (62–82 years). American Society of Anesthesiologist classification was 3 (1–3) and Body Mass Index was 23.8 kg/m2 (19.8–33.8). Ethical approval was granted by the Medical Research Ethics Committee of our hospital, and informed consent was obtained from all patients.
Perioperative data were collected and analyzed prospectively. Regardless of whether the patients were presumed to have underlying liver disease, indications for operation were established on the basis of liver function reserve, tumor stage and volume of the remnant liver 
. In patients with bilirubin concentration below 2.0 mg/dl the surgical strategy was planned according to the flowchart published by Makuuchi and Kokudo 
. Indocyanine green retention rate at 15 min (ICGR15) was not evaluated, and in patients with total bilirubin below 1.0 mg/dl, portal vein embolization was indicated if remnant liver volume determined by CT was expected to be less than 50% of the total liver volume.
Fresh frozen plasma (FFP) was administered at a rate exceeding blood loss by 20% and to maintain the serum total protein level at 5.0 g/dl. Packed red blood cells were administered if intraoperative blood loss exceeded 1000 ml or the hemoglobin level fell to 10 mg/dl during surgery or to 8 mg/dl on postoperative days (PODs). The association of International Normalized Ratio (INR) above 1.6 and serum total bilirubin level above 3 mg/dl on POD5 was considered as an indication for multiple bacteriologic examinations and for chest and abdominal CT to rule out pneumonia, abscess or portal vein thrombosis 
. Total bilirubin in drain discharge was sampled on POD3, POD5, and POD7. Drains were removed on POD7 or when the total bilirubin level in the discharge fluid was lower than the one recorded on the previous POD. Chest and abdominal US examination were performed in the presence of clinical indications and were also routinely performed on POD2, POD4 and POD6.
The terminology for liver anatomy and resection is based on the Couinaud classification 
. Resection of at least 3 adjacent segments was considered a major resection. Complications were graded on a 1–5-scale according to a grading system published by Dindo et al. 
. Grade I is any deviation from the normal postoperative course treated with drugs such as antiemetics, antipyretics, analgesics, diuretics and electrolytes. This grade also includes wound infections opened at the bedside. Grade II complications are those requiring blood transfusions or pharmacological treatment with drugs other than those allowed for Grade I. Grade III complications require invasive procedures. Grade IV is life-threatening complications requiring intensive care unit management. Grade V complications result in death. Grades I and II are grouped as “minor” and Grades III–V are considered as “major” complications. Bile leakage was defined as continuous drainage fluid with a bilirubin concentration greater than 5 mg/dl for more than 7 days or when intra-abdominal collection bilirubin level was greater than 5.0 mg/dl after puncture 
. Hepatic failure was defined as a serum bilirubin concentration greater than 5.0 mg/dl, a prothrombin time rate below 50% for 3 or more consecutive days, or both 
. Postoperative mortality was defined as death within 1 month of operation or during the hospital stay. In cancer patients, abdominal US, blood samples for liver function tests and serum tumor markers were performed every 3 months, and contrast-enhanced CT imaging every 6 months. Cut-end recurrence was defined as recurrence in the bed of resection. Operation-related complications identified during hospitalization and follow-up were reported.