Overview of hepatic resection

INTRODUCTION — Hepatic (liver) resection is needed to manage many types of pathology, malignant and benign. Planning hepatic resection requires consideration of the nature of the lesion and its location within the liver, the patient's anatomy, and the quality and volume of the liver tissue that will remain after resection (adequate future liver remnant). Perioperative outcomes for hepatic resection have improved due to better surgical techniques that take advantage of the segmental anatomy of the liver, improved techniques for control of bleeding, and improved intensive care. Hepatic resection that is performed in high-volume centers by specially trained hepatobiliary surgeons is associated with better outcomes [1-3].

Surgical resection of the liver and complications of liver resection will be reviewed here. Specific management of pathologies that indicate a need for liver resection and disease-specific outcomes related to liver resection are discussed in separate topic reviews. (See "Surgical resection of hepatocellular carcinoma" and "Surgical management of gallbladder cancer" and "Surgical techniques for managing hepatic injury" and "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma".)

LIVER ANATOMY AND PHYSIOLOGY — The liver is divided into two lobar segments (right and left) and further subdivided into eight (Couinaud) segments based upon vascular supply and bile duct distribution (figure 1). The segmental anatomy of the liver is the basis for the various types of anatomic hepatic resections (figure 2). The surgical anatomy of the liver and types of hepatic resection are discussed in detail elsewhere. (See "Open hepatic resection techniques", section on 'Surgical anatomy'.)

Liver function and regeneration after resection — The hepatocytes perform a variety of metabolic functions, including:

●Removing metabolic waste products, hormones, drugs, and toxins

●Producing bile to aid in digestion

●Processing nutrients absorbed from the digestive tract

●Storing glycogen, certain vitamins, and minerals

●Maintaining normal blood sugar

●Synthesizing plasma proteins, albumin, and clotting factors

●Producing immune factors and removing bacteria

●Removing senescent red blood cells from the circulation

●Excreting bilirubin

The expected volume of functional liver (ie, functional or future liver remnant) that is needed to maintain these important metabolic functions following liver resection depends upon the quality of the remaining liver tissue and its ability to regenerate. Liver regeneration is fundamental to the ability to perform more extensive hepatic resections. The mechanisms responsible for this capability are an area of active research [4-6]. In an animal model, angiogenesis inhibitors severely suppressed hepatic regeneration [5]. As more and more patients are treated with small molecule inhibitors, it is crucial that all members of a multidisciplinary team, both surgeons and medical oncologists, remain "on the same page" regarding current therapeutic needs and future therapeutic plans.

Although the exact amount of hepatic tissue regenerated will vary from patient to patient, healthy livers can regenerate significant amounts of liver within weeks to months after resection. In a series of 91 patients, liver volumes were measured before and after liver resection [7]. At six months postoperatively, the volume of liver regenerated was linearly related to the resected volume, but the liver volume had not yet reached total preoperative liver volume. Some patients have insufficient regeneration and may experience functional liver failure if the liver remnant is too small. One small study suggested that liver regeneration in patients with body mass index >30 may be slower than in others [8]. (See 'Contraindications' below.)

Preoperative portal vein embolization (PVE) prior to large-volume hepatic resection can stimulate liver hyperplasia [4]. We typically allow the liver to regenerate for four weeks after PVE before repeating liver imaging to calculate the predicted future liver remnant. If adequate future liver remnant (20 percent for completely normal liver, 30 percent for steatohepatitis or moderate chemotherapy exposure, or 40 percent for mild cirrhosis or major chemotherapy exposure) is confirmed on repeat liver imaging, we perform hepatic resection two weeks after reimaging (ie, four to six weeks from PVE). The techniques of PVE are discussed elsewhere. (See 'Preoperative PVE and other alternatives' below and "Preoperative portal vein embolization", section on 'Introduction'.)

An alternative to PVE is the ALPPS (associating liver partition and portal vein ligation for staged hepatectomy) procedure. (See "Preoperative portal vein embolization", section on 'PVE versus the ALPPS procedure'.)

INDICATIONS FOR HEPATIC RESECTION — Malignant tumor within the liver (primary or secondary) is the most common indication for hepatic resection. However, benign liver conditions, which can be congenital or acquired, may also require hepatic resection. Although hepatic trauma is most commonly managed conservatively, on occasion, hepatic resection may be required to definitively manage hemorrhage.

Malignancy — Patients with underlying conditions of the liver known to predispose to malignancy are generally screened at a regular interval for the development of cancer with imaging studies (eg, ultrasound, computed tomography, magnetic resonance) and serum markers (eg, alpha fetoprotein) [9,10]. In susceptible patients, a lesion that is clearly not a benign cyst should be considered malignant until proven otherwise, and dysplastic nodules are considered premalignant lesions and generally treated as malignant [11]. (See "Surveillance for hepatocellular carcinoma in adults".)

Hepatocellular carcinoma is the most common primary hepatic malignancy and can occur in the context of inherited (eg, hemochromatosis) or acquired (eg, chronic hepatitis C, alcoholic cirrhosis) preexisting conditions [12,13]. In one large series, cholangiocarcinoma was the second most common malignant tumor for which hepatic resection was performed [14]. (See "Epidemiology and risk factors for hepatocellular carcinoma" and "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma".)

The liver is a common site for metastasis from solid tumors. In selected patients with focal or isolated disease, resection of liver metastases is associated with low rates of major perioperative morbidity (approximately 3 percent) and mortality (approximately 4 percent) with good long-term results [15-19]. Neuroendocrine lesions, primarily from the foregut, are another source of metastases that respond well to liver resection [20,21]. Resection of nondigestive endocrine, and noncolorectal, non-neuroendocrine tumor metastases (eg, breast, sarcoma, genitourinary, melanoma), have also been reported [22-24]. (See "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy" and "Metastatic gastroenteropancreatic neuroendocrine tumors: Local options to control tumor growth and symptoms of hormone hypersecretion".)

Surgical treatment of gallbladder cancer involves resection of the gallbladder and involved tissues to obtain a tumor-free margin. However, fewer than one-half of patients are candidates for resection at the time of diagnosis because of advanced disease. Provided there is no evidence of disease elsewhere, options for hepatic resection in patients with gallbladder cancer include extended cholecystectomy (en bloc resection of the gallbladder and a rim of liver), re-resection of portions of segments IVb and V, and, less commonly, right hemihepatectomy. (See "Surgical management of gallbladder cancer".)

Benign disease

●Simple cysts, hemangiomas, adenomas, and focal nodular hyperplasia comprise the majority of benign hepatic lesions [25-28]. Symptomatic lesions causing pain or discomfort can be resected with minimal margins, often laparoscopically [29]. Most asymptomatic lesions can be managed conservatively and will not require resection; however, some asymptomatic lesions, such as large or giant hemangiomas, and adenomas larger than 4 to 5 cm, warrant resection when anatomically feasible [27]. (See "Hepatic hemangioma" and "Hepatocellular adenoma" and "Diagnosis and management of cystic lesions of the liver".)

●Bacterial hepatic abscesses are generally treated with broad-spectrum antibiotics for four to six weeks and percutaneous drainage with or without irrigation [30,31]. Surgical resection may be necessary for source control in patients who are extremely ill but is rarely required and not considered the standard of care in most patients. In one study, aggressive local surgical resection for pyogenic abscess was found to be beneficial in patients with an APACHE II score of ≥15 [32]. (See "Pyogenic liver abscess".)

●Amebic liver abscesses are usually treated effectively with metronidazole without the need for surgical intervention, biopsy, or drainage [33]. However, liver resection may be an option for the few very large abscesses where rupture is a concern, for patients who do not respond to medical treatment, or if the diagnosis is unclear [33]. (See "Extraintestinal Entamoeba histolytica amebiasis", section on 'Amebic liver abscess'.)

●Hepatic resection is also effective treatment of intrahepatic stone disease when accompanied by biliary stricture or segmental atrophy [34,35]. The management of these patients is individualized based upon the location of stricture(s) and atrophic regions.

Trauma — Although the management of liver trauma is primarily conservative, liver resection may be needed to control hemorrhage from higher-grade (grade IV, V) liver injuries. Angioembolization for vascular liver injuries is a safe and effective alternative to surgery for lower-grade injuries in many institutions.

For traumatic liver injuries resulting in devascularization of portions of the liver, the devascularized regions are often resected during a planned take-back to the operating room 24 hours after initial trauma exploration.

The grading and management of liver injuries and approach to the management of liver trauma in adults are discussed elsewhere. (See "Management of hepatic trauma in adults", section on 'Surgical management' and "Management of hepatic trauma in adults", section on 'Hepatic embolization' and "Surgical techniques for managing hepatic injury".)

CONTRAINDICATIONS — Patients with severe underlying functional liver disease (eg, cirrhosis, nonalcoholic steatohepatitis [NASH], chemotherapy related) are not candidates for major liver resection. For patients with less severe disease, the degree to which the underlying liver disease constitutes an absolute versus relative contraindication to hepatic resection depends upon the anticipated volume of liver remaining after resection (ie, future liver remnant [FLR]), the presence of medical comorbidities, and resources available in the event of perioperative liver failure, such as the availability and proximity of liver transplantation. Model for end-stage liver disease (MELD) scores do not directly impact decision making related to liver resection but may be useful in counseling a patient when choosing between liver resection and transplant [36,37]. In these cases, perioperative mortality after resection, death from primary disease, and risks associated with transplant need to be taken into account [38-41]. (See "Model for End-stage Liver Disease (MELD)".)

In retrospective reviews evaluating outcomes of patients undergoing liver resection, primarily resection of hepatocellular carcinoma or colorectal metastases, the risk of death increases with decreasing volumes of the future liver remnant [38,39,42-46]. For patients with normal liver function, FLR <20 percent increases the risk of liver failure and death following major hepatic resection [38,39]. In a review of 300 patients undergoing extended hepatectomy, multivariate analysis found a significantly increased risk of mortality if the FLR was ≤20 percent (odds ratio 3.18, 95% CI 1.34-7.54) [39]. The incidence of postoperative liver insufficiency and death was significantly greater in those with an FLR <20 percent compared with FLR 20.1 to 30 percent or FLR >30 percent (liver insufficiency 34 versus 10 and 15 percent, respectively; death 11 versus 3 and 2 percent, respectively). Patients with mild-to-moderate underlying functional liver disease have further increases in the risk of liver failure and death if the future liver remnant is inadequate, but there are no firm guidelines to define what constitutes "inadequate" for specific populations [47-49].

We use the following approach:

●We consider cirrhotic patients with Child-Pugh (table 1) class C and Child-Pugh class B with an FLR <40 percent as absolutely unresectable [42]. Other Child-Pugh B (liver remnant >40 percent) and some Child-Pugh A patients may be relatively unresectable. Portal vein embolization (PVE) to achieve FLR >40 percent can, in select cases, allow for a reasonable attempt at resection. If FLR does not increase to over 40 percent in this patient population, then we do not offer resection. In some centers, indocyanine green (ICG) clearance is performed to assess hepatic function. Normal ICG clearance at 15 minutes (ICG₁₅) is 90 percent; clearance of <60 percent indicates an extremely high risk of postresection liver failure and mortality. An alternative to absolute FLR is the kinetic growth rate (per week) after PVE. Higher kinetic growth rates generally denote individuals who will safely tolerate major hepatectomies. (See "Preoperative portal vein embolization", section on 'Mechanism and physiology'.)

●We do not offer hepatic resection to patients with NASH who have an FLR <30 percent [42]. There are few data on the safety of hepatic resection in patients with NASH, but the FLR should be higher than in patients with normal liver tissue. It may be reasonable to stratify these patients as medium-risk or high-risk based upon the presence of additional comorbidities such as insulin resistance, cardiovascular disease, or severe obesity [46]. For NASH patients with an FLR >30 percent and no other major comorbidities, it may be reasonable to offer liver resection.

●For patients who are candidates for hepatic resection but who are deemed to have an inadequate FLR, we suggest preoperative PVE. Preoperative PVE increases the volume of the FLR and may permit subsequent hepatic resection. (See 'Preoperative PVE and other alternatives' below and "Preoperative portal vein embolization", section on 'Introduction'.)

Additional contraindications to hepatic resection include comorbidities precluding/limiting safe anesthesia or location of disease near major vascular or biliary structures that would preclude a margin-negative resection. Documented extrahepatic disease is a contraindication to hepatic resection for many malignancies, but not for all. Inferior vena cava invasion is generally considered a contraindication to surgical intervention; however, hepatic vein invasion away from the inferior vena cava, although suggestive of aggressive tumor biology, is not an absolute contraindication. Disease-specific contraindications to hepatic resection are discussed in separate topic reviews.

PREOPERATIVE IMAGING — We obtain a preoperative imaging study in all patients prior to elective hepatic resection to evaluate the potential margins of resection, which will determine the volume of the future liver remnant (ie, anticipated volume of liver remaining after resection) (table 2) [50]. Considerations for specific diseases are discussed elsewhere. (See "Surgical resection of hepatocellular carcinoma" and "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy".)

It is important to obtain the highest-quality imaging available. High-quality imaging is generally defined as thin cuts (1 to 3 mm); coronal reconstructions; and includes arterial, portal venous, and venous phases [51]. At some institutions, this will be multiphase computed tomography (CT) and, at others, magnetic resonance imaging (MRI) with and without contrast; either is a reasonable choice provided the study is performed with appropriately thin-slice images and a properly timed intravenous contrast bolus. The most experienced radiologist available with the chosen modality should provide the interpretation, and the surgeon should also be comfortable interpreting the images. Although thin-slice CT is more commonly used, MRI may surpass CT in clinical utility as it becomes more time efficient, cost effective, and readily available [52].

We generally determine future liver remnant using CT volumetry in consultation with a radiologist. MRI volumetry provides similar information. The volume of the remnant liver (FLV) will be according to the equation FLV = (future liver volume / total estimated liver volume) × 100 [53]. Although there are numerous equations used to estimate liver volume [53,54], we prefer to utilize the equation method to calculate the total estimated liver volume and not measure the total liver volume with imaging.

Multiphase CT includes noncontrast, arterial, venous, and portal phases and should also include volumetric analysis of the future liver remnant. The arterial phase should identify important arterial anatomy and any aberrant vessels (such as replaced arteries) and define the relationship of intrahepatic tumor to critical vascular or biliary structures. Cut thickness should be approximately 1 mm. Omitting the portal venous phase increases the risk of missing small, isodense hepatocellular carcinoma lesions [55].

Although other imaging modalities such as ultrasound, positive emission tomography (PET), and single photon emission computed tomography (SPECT) may be useful adjuncts in management, poor resolution limits their usefulness for defining anatomy in planning hepatic resection.

PREOPERATIVE PVE AND OTHER ALTERNATIVES — Preoperative portal vein embolization (PVE) is a valuable adjunct to major liver resection, particularly for right-sided tumors. PVE initiates hypertrophy of the anticipated future liver remnant (FLR) and may allow for a more extensive resection or allow staged, bilateral resections. (See "Preoperative portal vein embolization" and "Preoperative portal vein embolization", section on 'Adjunctive techniques'.)

PREOPERATIVE EVALUATION AND PREPARATION — The evaluation of the patient undergoing hepatic resection involves medical risk assessment and a determination of the location of the lesion using imaging studies, expected margins of resection, and the volume of the residual liver remnant. Together, these will determine if resection is feasible and, if so, the extent of the resection. Some patients may benefit from portal vein embolization. These considerations are discussed in the sections above. (See 'Preoperative imaging' above and 'Preoperative PVE and other alternatives' above and "Preoperative portal vein embolization", section on 'Introduction'.)

Prior to hepatic resection, a frank conversation should be undertaken with the patient and their loved ones regarding the potential benefits of hepatic resection and complications, particularly the possibility of liver failure for those at risk. The risk of bile duct injury and bile leak and the potential need for repair, which might require an additional procedure, such as a hepaticojejunostomy or other biliary-enteric anastomosis, should also be discussed. (See 'Complications' below and 'Mortality' below.)

The patient with malignancy should understand that the procedure usually begins with diagnostic laparoscopy and intraoperative ultrasound, and if unresectable hepatic or extrahepatic lesions are found, the procedure will be terminated without hepatic resection.

Nutrition — From the initial time when patients are evaluated for surgery, we recommend a high-protein diet with increased physical activity (prehabilitation) prior to surgery. (See "Overview of prehabilitation for surgical patients".)

In some patients, a short-course low-calorie and low-fat diet immediately before surgery may benefit liver resection. In a randomized trial, patients with a body mass index ≥25 kg/m² who consumed an 800 kcal, 20 gram fat, and 70 gram protein diet for one week before liver resection incurred less intraoperative blood loss (452 versus 863 mL), and their livers were judged to be easier to manipulate by surgeons, compared with those who received no special diet [56]. The mechanism of action may be decreased liver glycogen and water content as there was no difference in the degree of liver steatosis in this study. However, a consequence of this may be overall decreased protein intake, which has not been well studied.

Medical assessment — Assessment of operative risk prior to liver resection includes establishing the severity of liver disease and the presence of other medical comorbidities. The majority of hepatic resections are performed under elective circumstances for which there is adequate time for risk assessment and optimization of the patient's medical status. Preoperative medical assessment is discussed elsewhere. (See "Assessing surgical risk in patients with liver disease" and "Evaluation of cardiac risk prior to noncardiac surgery" and "Evaluation of perioperative pulmonary risk" and "Preoperative medical evaluation of the healthy adult patient".)

Prior to hepatic resection, a complete blood count, serum chemistries, and liver function tests, albumin, and coagulation studies should be obtained. In addition, we require hepatitis serologies for all patients, and a recent colonoscopy, as well as tumor-specific markers, including CA19-9, carcinoembryonic antigen (CEA), and alpha-fetoprotein (AFP). Patients found to have liver dysfunction may not tolerate the extent of resection that would otherwise be indicated on imaging studies. (See 'Contraindications' above.)

Risk stratification for hepatic resection depends upon properly identifying those with liver tissue abnormalities. Core liver biopsy should be performed if there is any concern about nonalcoholic steatohepatitis (NASH), chemotherapy-induced steatohepatitis, or liver fibrosis. (See "Approach to liver biopsy".)

Prophylactic antibiotics — A decision to administer antibiotics in patients undergoing hepatic resection should take into account the indication for the procedure (eg, tumor, infection) and any additional procedures (eg, cholecystectomy, colectomy) that will be performed. When hepatic resection is being performed without surgery on other organs, and without known or suspected infectious conditions, hepatic resection can be considered clean surgery.

In general, prior to clean, uncontaminated hepatic resection, we give antibiotics directed against skin flora (table 3) within one hour prior to the incision [57]. In the placebo arm of one small, randomized trial, patients did not receive any antibiotics in the perioperative period [58]. Wound infection rates were similar to those reported using prophylactic antibiotics, suggesting that routine prophylaxis prior to clean hepatic resection may not always be necessary. Perioperative antibiotics generally suffice, but patients undergoing combined surgeries and those with complications may require ongoing antimicrobial therapy [59]. (See "Antimicrobial prophylaxis for prevention of surgical site infection following gastrointestinal procedures in adults".)

For hepatic resection associated with cholecystectomy, we give antibiotics according to the presence of any biliary tract complications with antibiotics directed against skin flora if the gallbladder is normal (table 3), and against biliary flora if there are complications such as obstruction, infection, or prior biliary tract instrumentation.

When resection of hepatic metastases is combined with surgery of other gastrointestinal structures (eg, colon resection) (table 3), antibiotic prophylaxis should be directed toward any likely sources of contamination.

For hepatic resection being performed to control infection, antibiotics appropriate to the diagnosis should be continued. (See "Pyogenic liver abscess", section on 'Antibiotic therapy'.)

Thromboprophylaxis — Patients undergoing major liver resection are at risk for thromboembolism (table 4) due to the nature of the surgery (major open surgery >45 minutes) and intraoperative maneuvers that may lead to vena cava compression. The presence of malignancy increases the risk [60]. We place intermittent pneumatic compression devices prior to induction and continue their use until the patient is ambulatory. Preoperative pharmacologic thromboprophylaxis, most often heparin (5000 units subcutaneous injection), is recommended for moderate- and high-risk patients based on the patient and the liver function. Our default approach is to give heparin 5000 units subcutaneously prior to induction of anesthesia and after epidural placement for open operations. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients".)

Anesthesia and analgesia — Hepatic resection is generally performed under general anesthesia with or without epidural anesthesia/analgesia. We prefer to use epidural analgesia in all open operations. The choice of anesthesia and the effects of anesthesia in patients with liver dysfunction are discussed elsewhere. (See "Overview of anesthesia" and "Anesthesia for the patient with liver disease".)

For open hepatic resection, epidural analgesia is a useful adjunct due to the extent of the incisions and also to aid pain management in the postoperative period, which improves pulmonary mechanics and decreases the risk for pulmonary complications. We have found that the benefits of epidural analgesia for postoperative pain outweigh the risks associated with liver resection, but the choice to proceed requires close collaboration with the medical providers managing the epidural catheter, usually an acute pain consulting service or the anesthesia providers. (See "Overview of anesthesia" and 'Complications' below and "Continuous epidural analgesia for postoperative pain: Technique and management".)

Warming devices (fluid warmers, external circulation devices) are routinely used given the extent and duration of exposure of the abdominal cavity and the frequent need for fluid therapy and transfusion. For major hepatic resections, invasive blood pressure and central pressure monitoring may aid anesthetic management. At the authors' institution, we do not routinely directly monitor central venous pressure; we place two to three large-bore peripheral intravenous lines and an arterial monitoring line. We routinely monitor pulse pressure variation as a marker for overall fluid status with a goal of having patients on the lower end of euvolemia. (See "Intraoperative fluid management" and "Anesthesia for the patient with liver disease", section on 'Anesthesia for hepatic resection'.)

HEPATIC RESECTION — Hepatic resection is performed in a stepwise fashion starting with laparoscopy and ultrasound imaging to evaluate the potential for complete resection and to define the relevant anatomy. (See "Open hepatic resection techniques", section on 'Staging laparoscopy and use of ultrasound'.)

Type and extent of resection — The type of hepatic resection chosen depends upon the location of the lesion(s), the ability to provide an adequate future liver remnant, and, for malignant disease, a tumor-negative margin. The types of hepatic resection include wedge resection, segmental resection (sectionectomy, sectorectomy), hepatectomy (right or left), and extended hepatectomy (right or left) (figure 2). Each of these (except wedge resection) constitutes anatomic resection based upon the segmental anatomy of the liver according to Couinaud (figure 1), but alternatively, nonanatomic resection can be performed. For most resections, cholecystectomy is performed followed by dissection of the porta hepatis to isolate and control the vascular and ductal structures. Thereafter, the liver tissue is divided to the extent that is mandated by the type of liver resection being undertaken, and hemostasis achieved prior to abdominal closure. The general techniques used for hepatic resection and the manner in which specific resections are performed are discussed elsewhere. (See "Open hepatic resection techniques", section on 'General techniques' and "Open hepatic resection techniques", section on 'Specific resections'.)

Anatomic resection, rather than nonanatomic resection, may be preferred because of improved perioperative outcomes and long-term survival [61-64]. However, nonanatomic resection may be needed if anatomic resection will result in inadequate residual liver volume to support recovery, such as in the patient with cirrhosis. More bleeding may occur with nonanatomic resection, which is a disadvantage of this technique; however, inflow occlusion can be used to reduce blood loss. (See "Open hepatic resection techniques", section on 'Vascular control'.)

Bilobar metastatic malignant disease is typically found in two variants: small lesions in fewer than four segments and larger lesions diffusely spread throughout the liver. Small lesions can be resected with similar outcomes using two partial anatomic or nonanatomic hepatic resections, provided sufficient healthy liver remains [65]. Another option is resection of one lesion and radiofrequency ablation of the second [66,67]. Portal vein embolization can also be used to induce hypertrophy of the parenchyma, allowing bilateral or staged resections [66,68]. (See 'Preoperative PVE and other alternatives' above and "Preoperative portal vein embolization", section on 'Introduction'.)

Tumor invasion into the diaphragm may require concurrent resection of a portion of the diaphragm. However, it is not clear if this extended procedure provides a long-term benefit [69-71]. If the resection can be accomplished en bloc, we believe it is reasonable to resect such a tumor. Mesh repair of the defect has been reported, but the oncologic consequences are unknown [70]. (See "Surgical treatment of phrenic nerve injury".)

Resection margins — The margin of resection that is needed depends upon the indication for liver resection. Benign lesions, such as hemangiomas, adenomas, complex cysts, and fibronodular hyperplasia, can be excised by enucleation or resection with limited margins.

Data to guide the margin of hepatic tumor resection are relatively sparse. We suggest a margin of at least 1 cm when resecting malignant tumor. A margin of 2 cm may be even more desirable for aggressive tumor biology or tumors expected to have peritumoral small or microscopic satellite foci of cancer (eg, cholangiocarcinoma). However, a wide margin is frequently not possible to achieve. As an example, for tumor in proximity to major vascular structures, a resection margin of only a few millimeters may be all that can be achieved.

For resecting colorectal metastases, some have suggested that a negative margin as little as 1 mm (compared with a traditional 1 cm margin) is adequate, but this is controversial [72-75]. Studies evaluating outcomes using greater or lesser margins have been confounded by a greater proportion of patients for whom a lesser margin was used because of multiple metastases or synchronous bowel disease [74]. It is possible that other types of metastatic lesions could be treated similarly to colorectal metastases, but until more evidence is available for other specific types of liver metastases, we prefer at least a 5 mm tumor-free margin, if at all possible. (See "Hepatic resection for colorectal cancer liver metastasis", section on 'Margins'.)

We use a 1 cm margin for hepatocellular carcinoma or cholangiocarcinoma because of lower recurrence rates with this margin and slightly prolonged rates of survival [76-80]. (See "Surgical resection of hepatocellular carcinoma".)

The "optimal" margin of resection for managing gallbladder cancer is also not well defined. For primary resection, a 5 mm tumor-free margin may be reasonable; however, if re-resection is being undertaken, a 1 to 2 cm tumor-free margin should be the goal, especially if there is any evidence of satellite tumor nodules. (See "Surgical management of gallbladder cancer", section on 'Extended cholecystectomy'.)

Minimally invasive approaches — Given the increase in effectiveness of systemic therapy for colorectal hepatic metastases and other primary liver malignancies, tumors may be small enough to be resected with one of the minimally invasive approaches if the anatomic location allows for such resection [81-83]. Surgeon experience and anatomic location usually determine the utility of minimally invasive liver resection (or ablation). In highly select cases, we perform liver wedge resection, left lateral sectionectomy, and left hepatectomy in a minimally invasive fashion. With increasing experience, we now also consider performing anatomic resection of segment IV, V, or VIII depending on the distance between the tumor and vessels to the posterior right or lateral left sector as well. (See "Minimally invasive liver resection (MILR)".)

POSTOPERATIVE CARE — Following hepatic resection, patients can typically be extubated in the operating room. The patient should be admitted to a monitored setting, the nature of which depends upon the extent of the operation and hospital facilities (eg, intensive care unit, step-down type unit with telemetry).

Intensive care is often needed for one or two days to monitor hemodynamics, glucose levels, coagulation parameters, and electrolytes. In the immediate postoperative period, hyperglycemia and coagulation abnormalities are relatively common. We aggressively monitor and control blood sugar and correct abnormal coagulation parameters, as described below. Hypophosphatemia, potentially related to increased phosphate uptake by regenerating liver cells, occurs in nearly all patients following major hepatic resection and should be corrected as needed [84]. (See "Hypophosphatemia: Evaluation and treatment".)

For minor hepatic resection and uncomplicated major resections, the diet can usually be resumed on the first postoperative day and advanced as tolerated, and ambulation is encouraged starting on postoperative day 1. Discharge is generally possible between postoperative days 4 and 6.

Glycemic control — Postoperative hyperglycemia is common following major hepatic resection, and insulin resistance after liver resection can make adequate blood glucose control challenging. We obtain routine blood glucose measurements every one to two hours for the first two to three days postoperatively and control blood glucose with an insulin drip to maintain blood glucose within a normal range (typically 90 to 130 mg/dL), until longer-acting insulin is effective at controlling blood glucose levels. In general, we do not advocate overly tight control due to the increasing evidence that hypoglycemia may be as harmful as hyperglycemia [85-87]. (See "Management of diabetes mellitus in hospitalized patients" and "Glycemic control in critically ill adult and pediatric patients".)

Coagulopathy and hemorrhage — The incidence of posthepatectomy coagulopathy is difficult to determine. Factors that contribute to coagulopathy following hepatic resection include hemodynamic instability, intraoperative blood loss, preexisting hepatic dysfunction, acute liver injury in the remnant liver tissue, and hypothermia [88-91].

We correct any coagulation abnormalities as they are identified (intraoperative, postoperative) with early and aggressive component transfusion. However, recombinant Factor VII should not be used when there is ongoing intraoperative blood loss. One trial that randomly assigned patients undergoing hepatic resection to blood transfusion or recombinant Factor VII did not find any significant differences in outcomes between the treatments [92]. (See "Use of blood products in the critically ill" and "Clinical use of plasma components" and "Platelet transfusion: Indications, ordering, and associated risks", section on 'TTP or HIT' and "Recombinant factor VIIa: Administration and adverse effects" and "Massive blood transfusion".)

FOLLOW-UP — When the patient is ready for discharge, it is important to develop a strategy to manage potential complications for patients who have travelled to a regional center of excellence to undergo hepatic resection. In one retrospective review of 1281 patients, 14.4 percent of patients required readmission and 6.8 percent required reoperation [93].

The patient should return to see their surgeon one to two weeks after the surgery to evaluate wound healing and their overall recovery. The patient should also follow up with their primary clinician and/or medical oncologist within one month of discharge regarding long-term treatment plans and options for adjuvant therapy. We make an appointment to see the patient again at 3 to 6 months postoperatively and again at 12 months to reassess the surgical incision, the patient's weight, and their liver function. For patients with malignancy, we continue to follow up every six months to assess for recurrent disease or new metastatic disease to the liver.

In cases of early recurrence or the unfortunate occurrence of a positive margin, whether or not to proceed with a second (ie, re-resection) operation becomes a complex decision that is rarely encountered. The same considerations that are discussed above for primary resection should also be applied. Re-resection can be considered for patients with primary liver tumors or metastatic liver tumors, but only for patients with no metastatic disease outside the liver. There must be adequate hypertrophy of the liver after the first resection, and the resection should be anatomically feasible based upon the location of the recurrent disease. Fundamentally, prior to re-resection, the patient's liver function (and other comorbidities) as well as functional status should have returned to baseline. We prefer to wait at least three months; however, if the tumor biology is more aggressive in nature, we will wait six months or longer to allow for a trial of chemotherapy or biologic therapy, as indicated.

COMPLICATIONS — Complications of any kind after hepatic resection occur in up to 40 percent of patients without cirrhosis, and at a higher rate in patients with some degree of cirrhosis [18,44,88,94,95]. In a review of 13,558 hepatobiliary operations from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database, overall perioperative morbidity was 18 percent for the resection of benign hepatic lesions and 21 percent for resection of hepatic malignancy [96]. Morbidity was highest for patients undergoing extended hepatic resection at 33 percent compared with 25 percent for hemihepatectomy, and 21 percent for partial hepatectomy (ie, segmentectomy or sectorectomy).

Major complications occur in approximately 10 to 20 percent of patients [88,89,95,97] and include bile leak, pulmonary complications, acute kidney injury, and liver failure. Advanced age and the presence of metabolic syndrome are associated with an increased risk for complications following hepatic resection [98,99].

Bile leak — Bile leak occurs in fewer than 10 percent of patients following hepatectomy. The International Study Group of Liver Surgery (ISGLS) has defined bile leakage as bilirubin concentration in drainage fluid at least three times the serum bilirubin concentration on or after postoperative day 3, or as the need for radiologic or operative intervention resulting from biliary collections or bile peritonitis [100]. Bile leakage following hepatic surgery is further classified as Grade A, B, or C. Grade A bile leakage causes no change in patients' clinical management. A Grade B bile leakage requires active therapeutic intervention but is manageable without relaparotomy, whereas for Grade C bile leakage, relaparotomy is required.

In retrospective reviews, independent risk factors associated with clinically relevant bile leakage (B, C) include drain fluid-to-serum total bilirubin concentration, prolonged operative time, resection for hepatocellular carcinoma, repeat hepatectomy, and extended left hepatectomy [101,102]. The majority of bile leaks can be managed with endoscopic decompression and percutaneous drainage [102]. The main causes of intractable bile leak are latent biliary stricture due to prior treatments and intraoperative hepatic duct injury during repeat hepatectomy [101]. (See "Overview of endoscopic retrograde cholangiopancreatography (ERCP) in adults".)

Pulmonary complications — Due to the altered respiratory physiology from the extent of the incision and retraction needed for surgical exposure, pulmonary complications following open upper abdominal surgery are common. In a retrospective study of 555 patients undergoing hepatic resection, pleural effusion and pneumonia occurred in 40 and 22 percent, respectively [103]. On multivariate analysis, independent risk factors for pleural effusion included prolonged surgery, right hepatectomy, neoadjuvant chemotherapy, and bilateral subcostal (ie, Chevron) incision. Risk factors for pneumonia included intraoperative blood transfusion, diabetes, and atrial fibrillation. The diagnosis and treatment of postoperative pulmonary complications are discussed elsewhere. (See "Overview of the management of postoperative pulmonary complications".)

Ascites — Ascites is common in patients with liver disease and is often found postoperatively [104]. Severe or increased amounts of ascites should alert the clinician to the possibility of portal vein thrombosis [105], or that liver failure may be ensuing. Routine aspiration or drainage is not recommended.

Thrombotic complications — Portal vein thrombosis and hepatic artery thrombosis are regarded as uncommon but serious potential complications of hepatic resection and may be related to technical issues during the operation. In a study of 222 patients without preoperative portal vein thrombosis who underwent hepatectomy [106], the incidence of postoperative portal vein thrombosis was 9.1 percent. Multivariate analysis identified right hepatectomy as a significant independent risk factor for main portal vein thrombosis (odds ratio 109, 95% CI 11-2906). Patients with peripheral portal vein thrombosis had a significantly longer duration of the Pringle maneuver than patients without portal vein thrombosis (76 versus 43 minutes). Patients were selectively anticoagulated depending on extent of thrombus, effect on portal flow, and risk for bleeding. Among those who were anticoagulated, portal vein thrombosis resolved in all patients after a mean treatment period of 1.6 months (total mean duration of therapy 4.6 months).

Symptoms of portal vein thrombosis are often vague and may be obscured by postoperative pain, but sharp increases in liver function tests should raise suspicion for portal vein thrombosis [105,106]. If there is any concern for a thrombotic complication, we suggest imaging evaluation (Doppler ultrasound or computed tomography of the abdomen) to assess portal venous and/or hepatic artery flow. Portal vein or hepatic artery thrombosis may require reoperation. (See "Acute portal vein thrombosis in adults: Clinical manifestations, diagnosis, and management".)

Liver failure — The most severe complication of hepatic resection is posthepatectomy liver failure (PHLF) [88]. One clinically useful definition of PHLF is impairment in the liver's ability to maintain its synthetic, excretory, and detoxifying functions as characterized by an increased international normalized ratio (INR) and hyperbilirubinemia on or after postoperative day 5 [107]. Mortality related to PHLF can be as high as 70 percent [14,108-110]. The main risk factors for PHLF are underlying functional liver disease and an insufficient volume of the residual liver remnant [111,112]. In small, experimental studies, elevated portal venous or hepatic venous pressures were found to predict PHLF [113,114]. However, preoperative pressures are not readily available, and thus, proper patient selection and preparation remains the most effective means of prevention. The importance of an adequate future liver remnant cannot be overly stressed. Management of PHLF is primarily supportive [107,110]. (See "Acute liver failure in adults: Management and prognosis".)

MORTALITY — Perioperative mortality following hepatic resection is 1 to 3 percent at high-volume centers [88]. In a review of 13,558 patients who underwent hepatobiliary surgery from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database, perioperative (30 day) mortality was 2.1 percent for patients undergoing hepatic resection and was similar for benign and malignant lesions [96].

A retrospective database review in Taiwan identified 13,159 patients who underwent liver resection. Overall mortality was 3.9 percent [115]. Preexisting renal disease and cirrhosis-related complications were the strongest predictors of mortality. A risk score was developed based upon known risk factors for mortality. Mortality for those with scores ≤3 was <2 percent, whereas for a score >3, mortality was between 7 and 15 percent. The components of the score included:

●0 points:

•Benign disease

•Less than lobectomy

●1 point:

•Age (>65 years)

•Lobectomy or more

●2 points:

•Ischemic heart disease  

•Heart failure

•Cerebrovascular disease  

•Malignancy as indication for surgery

●3 points:

•Preexisting cirrhosis-related complications

●4 points:

•Preexisting renal disease

In another series of over 100 resections for hepatocellular carcinoma, the presence of cirrhosis (95 percent Child-Pugh A patients) similarly increased operative mortality from 1 to 14 percent [44]. Mortality of Child-Pugh Class C (table 1) patients ranges from 30 to 100 percent, depending upon the extent of liver resection [36,37]. Nonalcoholic steatohepatitis (NASH), which is highly associated with obesity, diabetes, and inflammation, is also associated with increased perioperative morbidity and mortality following liver resection [45].

Long-term survival following hepatic resection depends upon the pathology treated and, for malignancies, the ability to achieve a negative margin. As examples, hepatic resection of colorectal metastases is associated with overall 1, 5, and 10 year survival rates of 93, 47, and 28 percent, respectively [116]. Overall, five-year survival following resection of hepatocellular carcinoma ranges from 60 percent for small, nonfibrotic lesions (T1F0) to 10 percent for large, fibrotic lesions (T3F1) [117]. The prognosis of these malignancies and the outcomes of the various benign pathologies treated with hepatic resection are discussed in separate topic reviews.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Liver resection and ablation".)

SUMMARY AND RECOMMENDATIONS

●Indications for hepatic resection – Hepatic (liver) resection is used to manage many types of liver pathology. Malignant tumor within the liver (primary or secondary) is the most common indication for hepatic resection. Benign liver conditions that require hepatic resection are usually symptomatic and can be congenital or acquired. Hepatic trauma is most commonly managed conservatively but on occasion will require hepatic resection to definitively manage hemorrhage. (See 'Indications for hepatic resection' above.)

●Preoperative evaluation – The preoperative evaluation of patients undergoing hepatic resection involves medical assessment and imaging studies to determine the location of the lesion, potential margins of resection, and the expected volume of functional liver following resection (ie, future liver remnant). Together, these will determine if resection is feasible and, if so, the type and extent of the resection. (See 'Preoperative evaluation and preparation' above.)

●Contraindications – For patients with future liver remnant <20 percent of the total volume of the liver, we suggest not performing hepatic resection (Grade 2C). Perioperative morbidity and mortality is significantly increased with lesser volumes. The risk of liver failure and death is further increased for patients with underlying liver dysfunction (eg, cirrhosis, steatohepatitis), such that an even larger volume of future liver remnant is required. (See 'Contraindications' above and 'Liver function and regeneration after resection' above.)

●Preoperative portal vein embolization – For patients with a future liver remnant that contraindicates hepatic resection, we suggest preoperative portal vein embolization (PVE) (Grade 2C). PVE stimulates liver hyperplasia in the remnant liver. In patients with certain malignancies, PVE has improved survival and reduced rates of postoperative liver failure following more extensive hepatic resections. (See 'Preoperative PVE and other alternatives' above and "Preoperative portal vein embolization", section on 'Disease-specific outcomes' and "Preoperative portal vein embolization", section on 'Introduction'.)

●Prophylactic antibiotics – We recommend prophylactic antibiotics prior to incision in patients undergoing hepatic resection (Grade 1B). The specific agent given should take into account the indication for the procedure (eg, tumor, infection) and any additional surgical procedures (eg, cholecystectomy, colectomy) that will be performed (table 3). (See 'Prophylactic antibiotics' above.)

●Resection margins – For malignant liver lesions, we suggest a margin of at least 1 cm rather than a lesser amount, whenever possible (Grade 2C). A margin greater than 2 cm may be desirable for aggressive tumor biology, but this is frequently not possible, and for some tumors, a lesser margin may be appropriate. Benign lesions (hemangiomas, adenomas, complex cysts, fibronodular hyperplasia) can be excised by enucleation or resection with limited margins. (See 'Resection margins' above.)

●Morbidities and mortality – At high-volume centers, surgical mortality following hepatic resection is 1 to 3 percent and is highest among patients with underlying liver disease. In patients with no evidence of diffuse liver disease and minimal comorbidities, liver failure is uncommon following limited resections (ie, adequate liver remnant). Long-term survival depends upon the specific pathology for which the hepatic resection was performed. Complications of any kind occur in up to 40 percent of patients without cirrhosis, with a higher incidence in patients with some degree of cirrhosis. Major complications include bile leak, coagulopathy and postoperative bleeding, hyperglycemia, and liver failure. (See 'Complications' above and 'Mortality' above.)