Major liver resections, perioperative issues and posthepatectomy liver failure: A comprehensive update for the anesthesiologist

Abstract

Significant advances in surgical techniques and relevant medium- and long-term outcomes over the past two decades have led to a substantial expansion in the indications for major liver resections. To support these outstanding results and to reduce perioperative complications, anesthesiologists must address and master key perioperative issues (preoperative assessment, proactive intraoperative anesthesia strategies, and implementation of the Enhanced Recovery After Surgery approach). Intensive care unit monitoring immediately following liver surgery remains a subject of active and often unresolved debate. Among postoperative complications, posthepatectomy liver failure (PHLF) occurs in different grades of severity (A-C) and frequency (9%-30%), and it is the main cause of 90-d postoperative mortality. PHLF, recently redefined with pragmatic clinical criteria and perioperative scores, can be predicted, prevented, or anticipated. This review highlights: (1) The systemic consequences of surgical manipulations anesthesiologists must respond to or prevent, to positively impact PHLF (a proactive approach); and (2) the maximal intensive treatment of PHLF, including artificial options, mainly based, so far, on Acute Liver Failure treatment(s), to buy time waiting for the recovery of the native liver or, when appropriate and in very selected cases, toward liver transplant. Such a clinical context requires a strong commitment to surgeons, anesthesiologists, and intensivists to work together, for a fruitful collaboration in a mandatory clinical continuum.

Key Words: Liver resection, Chronic liver disease, Preoperative assessment, Vascular clamping, Intraoperative hemodynamic monitoring, Postoperative intensive care unit, Posthepatectomy liver failure, Artificial liver support

Core Tip: Aim of our review is, highlighting surgical anatomy of the liver, types of liver resection, and systemic consequences of surgical maneuvers, to provide the anesthesiologists involved in liver surgery with the expertise for a proactive management of the perioperative period. We will address cardiovascular consequences of vascular clamping and declamping, intraoperative hemodynamic monitoring, postoperative treatment of posthepatectomy liver failure, the use, if and when appropriate, of artificial support(s) and, in very selected cases, the rescue indication to liver transplant.

INTRODUCTION

Over the past 20 years, major (or complex) liver surgery has experienced a remarkable increase in both volume of activity and relevant improvements in medium- to long-term results. This progress is attributed to a better understanding of hepatic anatomy and physiology, significant surgical technical advancements, the high volume of activity in specialized centers, and the increased anesthesiologic expertise spanning the entire perioperative period[1,2]. However, morbidity reaching up to 30% and mortality ranging from 2.5% to 3.5% rates remain considerable. These figures are even higher in oncologic surgery performed on cirrhotic patients, with morbidity and mortality increasing to 50% and 8%, respectively.

For anesthesiologists and intensivists involved in these procedures, a comprehensive understanding of the surgical anatomy of the liver and the types of hepatic resections is crucial for an appropriate interpretation of the intraoperative (patho)physiologic profile of the patients, aiming at proactive management of cardiorespiratory, renal, and metabolic changes during the various phases of surgery. Such an approach is also mandatory to manage the early postoperative period and to contribute to reducing postsurgical complications, particularly the feared posthepatectomy liver failure (PHLF)[3,4]. This review covers pre- and intraoperative anesthesia management strategies, early postoperative complications including PHLF, and procedures and interventions in high-dependency units (HDUs) and intensive care units (ICUs). As recently highlighted by Sparrelid et al[3], in the context of PHLF, where no effective means currently exist to support the remnant liver, “the best way to treat PHLF is to prevent its occurrence”.

SURGICAL ANATOMY OF THE LIVER: THE RELEVANT ANATOMIC FEATURES

Lobes

According to classical anatomy, the liver is divided by the falciform ligament into two main lobes—the larger right lobe (75% of the liver mass), and the smaller left lobe (25% of the mass)[5,6]. The right lobe includes the quadrate lobe, located between the gallbladder and the fissure for the ligamentum teres hepatis, and the caudate lobe (Spiegel’s lobe), situated between the fissure for the ligamentum venosum Arantii and the inferior vena cava (IVC). In 1897, Sir James Cantlie introduced a novel anatomical division of the liver, defining two parts of similar size based on an imaginary line along the middle hepatic vein (MHV). This line extends from the gallbladder fossa to the external border of IVC, dividing the liver into a right lobe (60%) and a left lobe (40%), each independently vascularized from the right and left portal branches.

Segments and sectors

In the late 1950s, Claude Couinaud proposed an innovative perspective on liver anatomy[7], which was extensively applied to liver surgery by Henry Bismuth[8]. Couinaud’s classification divides the liver into eight functional segments (I-VIII), based on the distribution of blood vessels and bile ducts (the segmental pedicle). Each segment is functionally independent and is supplied by the third division branch of the portal vein (PV). Segment IV was further subdivided into IVa and IVb. There are four liver sectors—right anterior, right posterior, left medial, and left lateral—defined by their external morphology and hepatic vein distribution. These sectors, demarcated by imaginary planes passing through the three main hepatic veins and divided into segments by the portal plane, are vascularized by the second division branch of the portal vein. The Couinaud classification for liver segments and sectors is now widely accepted and used in clinical practice for liver surgery and pre-operative radiological assessments[5,6].

Biliary system

Bile, produced by liver cells, drains through a network of small ducts called bile canaliculi. These canaliculi merge to form the intrahepatic bile ducts. The right and left hepatic ducts converge to form the common hepatic duct, which then combines with the cystic duct to create the common bile duct. The gallbladder stores and concentrates bile produced by the liver.

Ligaments

Within the abdomen, the liver is” anchored” to various structures by ligaments. The falciform ligament separates the right and left anatomical lobes, while the coronary ligaments attach the liver to the diaphragm. These ligaments also help maintain the liver’s anatomical position.

Takasaki classification of liver anatomy

An innovative concept in liver surgical anatomy was proposed by Japanese surgeon Takasaki[9]. This concept, involving extra-hepatic Glissonean pedicle isolation based on Laennec’s capsule, has gained widespread popularity over the past decade, particularly in modern minimally invasive liver surgery enabling optimal control and isolation of the pedicles in every kind of liver surgery, preventing parenchymal damage, and minimizing bleeding problems. Unlike the traditional eight-segment Couinaud’s classification, Takasaki’s system divides the liver into four functional segments (three segments plus the caudate lobe), each fed by secondary portal branches. The tertiary portal branches form a variable number of “cone units” (ranging from 22 to 24) that receive vascularization from the tertiary portal and arterial branches. The extra-hepatic Glissonean approach, now considered the gold standard in liver resections for segmental resections and living donor liver procurements, has been widely adopted in recent years[10].

LIVER VASCULARIZATION

The hepatic artery (HA) and PV provide the double blood supply to the liver, accounting for 20%-25% of the cardiac output. The total hepatic blood supply approximates 1500 mL/min: 300 mL from the HA (delivering 50% of the oxygen to the liver) and 1200 mL from the PV flow, which provides the remaining 50% of the oxygen supply[11-13]. HA and PV divide into right and left branches at the porta hepatis. PV pressure, typically ranging from 5 to 10 mmHg, is higher than the pressure in IVC. Portal hypertension is defined as a PV pressure > 10 mmHg or a PV-IVC gradient > 4 mmHg. Hepatic venous drainage includes three large intrahepatic veins - the suprahepatic veins (right, middle, and left)- draining into the IVC just below the diaphragm. The MHV, receiving branches from both right and left veins, demarcates the junction between the right and left lobes, each receiving vascularization from the corresponding right and left vascular pedicles. The caudate lobe possesses its unique venous drainage system[5,6]. The liver, whose venous return drains into the right atrium, is extremely sensitive to obstructions in venous outflow, sometimes leading to venous hypertension[13]. Relevant conditions include right heart failure, Budd-Chiari syndrome, or changes in venous drainage due to both cardiac (acute or subacute right ventricular congestion) and non-cardiac surgery[13]. Prolonged obstruction of hepatic venous outflow leads to an increase in intrahepatic blood volume associated with moderate-to-severe hepatic dysfunction. Furthermore, the liver serves as a significant blood reservoir: during acute hemorrhage, ortho-sympathetic-mediated vasoconstriction can increase circulating blood volume by approximately 25% of the blood loss[12].

The anatomo-functional unit of the liver is the hepatic lobule, hexagonally shaped and centered around the central vein (a terminal branch of the hepatic vein, also known as the centrolobular vein). Portal spaces (portal triad), containing arterial, portal, and biliary branches, are situated at each corner of the hexagon[11]. From a functional and metabolic perspective, three zones are recognized: Zone I (periportal), surrounding the portal space, is the most oxygenated and the most resistant to hemodynamic insults: it is responsible for intense metabolic and synthetic activity (such as glycogenolysis and gluconeogenesis), and has a high regenerative capacity. Zone III (pericentral), contiguous with the central vein, is hypo oxygenated and highly susceptible to hemodynamic insults and hypoxia. In contrast to Zone I, its regenerative capacity is low. Zone III is the primary site for ketogenesis and phase 1 and 2 drug and substance metabolism, including glycuronoconjugation and detoxification processes[11,12].

HEPATIC RESECTION: INDICATION AND SURGICAL TECHNIQUES

Currently, two are the approaches to liver resections: the conventional “open” approach and the “minimally invasive” technique, this latter, which includes both laparoscopic and robotic methods, on the rise[14-17]. The main indications for liver resection are summarized in Table 1. Compared to the traditional open approach, minimally invasive liver resections are associated with reduced blood loss, less postoperative pain, much rare ascitic decompensation, lower incisional hernia rates, faster postoperative recovery, and shorter hospital stays. The absence of the modified Makuuchi or Mercedes incision also correlates with fewer respiratory complications. However, morbidity and mortality rates remain largely unchanged[1,2,14-17].

Table 1 Indications for liver resection.

Benign solid tumors (selected cases)

Adenoma (single/multiple)

Hemangioma

Focal nodular hyperplasia

Inflammatory pseudotumor

Malignant solid tumors

Liver metastasis

Primary tumors of the liver and biliary tract

Hepatocarcinoma

Cholangiocarcinoma

Cholecystic tumor

Primary hepatic sarcoma

Angiosarcoma

Cystic lesions

Cystadenoma/cystadenosarcoma

Simple epithelial cyst

Hepatic polycystosis

Pyogenic liver abscess

Amoebic abscess

Echinococcus cyst

Biliary tract diseases

Primary sclerosing cholangitis

Caroli’s disease

In selected cases, for both benign and malignant lesions, liver transplantation may be the best treatment option. The indications for transplant in hepatocellular carcinoma on cirrhosis (HCC) are well-established. Nowadays, patients with cholangiocarcinoma may benefit from transplantation under strict selection criteria, while prospective trials dealing with liver transplantation for colorectal metastases are currently underway[14,16,17]. In major liver resections, ensuring a sufficient future liver remnant (FLR) is crucial. Pre-operative liver volumetry, calculated via computerized tomographic (CT) scan or magnetic resonance imaging (MRI), is mandatory to prevent fatal PHLF. The parenchymal function can be dynamically assessed using indocyanine green (ICG) clearance. A successful strategy to increase FLR volume involves portal and hepatic vein embolization of the liver portion to be resected, usually performed by interventional radiologists. In selected cases, a two-stage hepatectomy may be conducted. The first stage involves resecting malignant lesions from the lobe that will be preserved, followed by ligation and sectioning of the main portal branch of the contralateral lobe to induce parenchymal hypertrophy. The second stage entails a planned major liver resection after achieving sufficient hypertrophy. However, single ligation of either the left or right main portal branch alone has not proven effective in inducing hypertrophy in the contralateral hemiliver. The success rate for achieving sufficient contralateral lobe hypertrophy to perform a two-stage hepatectomy is about 60%[17,18]. Consequently, in recent years, an innovative and challenging technique - ALPPS (Associating Liver Partition and Portal vein ligation for Staged Hepatectomy) - has been proposed[19,20]. This two-stage technique increases the treatment options for otherwise unresectable tumors. In the first stage, the portal branch ipsilateral to the hemiliver to be removed is ligated. Any lesions in the hemiliver to be preserved are resected, and a parenchymal transection along the Cantlie line is performed. Hypertrophy of the lobe to be preserved is usually achieved within about two weeks. The second stage, usually 15 to 20 days later, is undertaken following a CT scan volumetry and is performed when appropriate hypertrophization of the FLR is confirmed. In this second stage, the hemiliver is removed by transecting the artery, the biliary duct, and the hepatic vein. Although ALPPS is effective in inducing hypertrophy of the remnant liver, it is associated with high surgical complications and mortality rates[17-20]. Extreme technical and surgical expertise, along with mandatory multidisciplinary collaboration, are required. As evident, the anesthesiologist should master the surgical technical issues to respond to the challenges imposed by such types of hepatic resections, close collaboration with the surgeon being mandatory to optimize perioperative strategies[17-20].

RESECTIVE TECHNIQUES

Anatomic resection is defined as the resection of one or more hepatic segments, sectioning the segmental Glissonian arterial and portal pedicle, and resecting all the dependent parenchyma[4,17,21-25]. Anatomic resections are associated with shorter surgical times and reduced blood loss compared to non-anatomical resections. Hepatic resections are classified as minor (involving up to two segments) or major (involving more than two segments). In non-anatomical “wedge” resections, the main goal is to preserve liver parenchyma. The liver resection is strictly guided by the oncological margins of the lesion, aiming at the preservation of as much remnant parenchyma as possible while ensuring adequate inflow and outflow.

The most common major liver resections include (Figure 1): Right hepatectomy or right bisectionectomy (segments 5-6-7-8, with or without the caudate lobe). Anatomic right hepatectomy typically preserves the MHV for proper segment IV (S4) outflow. Left hepatectomy (segments 2-3-4, with or without the caudate lobe). Right extended hepatectomy or right trisectionectomy (segments 4-5-6-7-8). In extended resections, the FLR must be properly assessed before surgery to avoid PHLF, often associated with an excessive reduction of the remaining liver mass, unable to support the functional needs and leading to a “small for size” syndrome.

Figure 1 Surgical anatomy of the liver and resection lines. Couinaud segmental liver anatomy and the normal portal venous structures biliary tract structures. Resection lines (---) and Hepatic segments (Arabic numbers) resected during major hepatectomies Right, Left and extended) (from Njoku DB, Chitilian HV, Kronish K. Hepatic Physiology, Pathophysiology, and Anesthetic Considerations. In Miller’s Anesthesia Michael A. Gropper, Ronald D Miller, Neal H. Cohen Lars I. Eriksson Kate Leslie, Jeanine P. Wiener-Kronish NINTH EDITION. Elsevier 2020: 420-443[27]; and with permission).

The (sometimes mandatory) removal of the MHV during a right hepatectomy is an example of the possible functional issues encountered during liver resection. Without collateral veins, acute congestion in segment 4 and segmental atrophy may occur, reducing the FLR. Technically, all “open” liver resections, and especially minimally invasive procedures, are carried out under ultrasound guidance[21]. Large vessels can be sutured and ligated using vascular staplers. Parenchymal transection may be performed using a Cavitron Ultrasonic Surgical Aspirator (CUSA) and/or other devices based on ultrasound or radiofrequency technology. Hemostasis can be achieved with bipolar forceps, clips, and stitches or with topical hemostatics (tissue adhesive or fibrin sealants).

Left liver resection

This involves the resection of segments II, III, and IV.

Non-anatomic resections

Non-anatomic resections are performed without following the segmental vascular pedicle, as in “wedge resection”, in case of small superficial lesions or when the anatomy of the lesion does not allow other approaches. Anatomic resections, allowing ligation of the vascular pedicle, are associated with shorter surgical times and reduced blood loss. Extremely complex resections, defined according to the Brisbane classification, are beyond the scope of this review[26].

MAJOR LIVER SURGERY AND THE ANESTHESIOLOGIST: TOWARDS A PROACTIVE STRATEGY

Modern liver surgery demands a multidisciplinary approach. The anesthesiologic commitment to “perioperative medicine” in major liver surgery should span the entire surgical period, encompassing preoperative evaluation, intraoperative management, and intensive or semi-intensive postoperative care[1,27]. Close collaboration with surgeons, hepatologists, interventional radiologists, and digestive endoscopists is mandatory. Combined surgical and anesthesiologic strategies aim for optimal intraoperative management, which includes minimizing blood loss, reducing operative time, all features able to lower lowering postoperative complication rates and ease postoperative rehabilitation. In this context, the Enhanced Recovery After Surgery (ERAS) strategy, when appropriately implemented and applied, is an extremely effective clinical pathway[28-30].

Among the surgical maneuvers and manipulations anesthesiologists must be familiar with due to their potential relevant hemodynamic impact are: (1) The surgical “manipulation” of the liver, aiming at the best exposure of the anatomical structures and able to acutely reduce the venous return from the IVC, leading to severe, abrupt arterial hypotension; and (2) selective or total vascular clamping before resection, aimed at reducing bleeding[1,27]. Identifying the surgical resection plane is a significant technical challenge, particularly in the presence of distorted anatomy due to neoplastic lesions and/or cirrhosis-induced changes, and/or bleeding, constant communication between the surgeon and the anesthesiologist allows for the implementation of proactive strategies to anticipate and/or mitigate significant hemodynamic consequences. If foreseeable during the preoperative planning phase, dedicated technical solutions should be planned. Intraoperative ultrasound (US), CUSA Dissector, Ligasure, Argon beam coagulation, and topical hemostatic agents (tissue adhesive and fibrin sealants) are among the innovative surgical hemostatic solutions used to identify vascular structures and reduce surgical bleeding. The anesthesiologist should be familiar with and master selective (Pringle maneuver) or total vascular exclusion techniques (supra- and infrahepatic clamping of the IVC; aortic clamping), or, in extreme cases, might indicate the use of extracorporeal venovenous bypass to maintain venous return during procedures involving the right atrium or the IVC (usually because of neoplastic atrial or caval thrombi) or in cases of acute juxta-atrial injuries[1,27]. These strategies, if to be considered, should be discussed during the planning of the surgical procedure with the support of cardiac anesthesiologists, cardiac surgeons, and perfusionists. The proactive approach includes anticipating every predictable intraoperative problem and considering appropriate solutions in advance, to be readily available during surgery.

PREOPERATIVE EVALUATION

Liver surgery, encompassing advanced and complex procedures[17,26], falls into the category of intermediate or high-risk surgical operations, with mortality/major cardiovascular event rates ranging from 1%–3% to ≥ 5%[31,32]. A wise selection of candidates is crucial, addressing both the general comorbidities of the patient and the function of the remnant liver in cases of cancer pathology and HCC[1,17,27]. The time-honored Child-Pugh (CP) classification[33,34] is still in use. Prevalent in clinical practice is now the MELD score with its latest versions, relevant for the mortality prediction scores[35,36]. The online calculator available on the Mayo Clinic website (https://www.mayoclinic.org/medical-professionals/transplant-medicine/calculators/meld-model/itt-20434705) provides MELD values and associated mortality and morbidity risks[35]. Among tools to assess liver function on the rise are the “static” ALBI score[37-41] and the “dynamic” indocyanine green clearance (IGC)[17,42-45]. The ALBI score provides prognostic information about mortality and postresection liver failure in HCC patients[17,37-41]. IGC[42,43] and volumetric CT preoperatively predict liver function in both cancer and cirrhotic patients. Methods used to assess residual function after liver resection include: (1) Volumetric assessment (% future liver remnant volume,), with a remnant volume being sufficient if at least 20%–30% of the native volume (or 40% in cases of chemotherapy); and (2) functional assessment, primarily based on IGC clearance, with preoperative R15 (retention rate at 15 minutes) values above 15%-20% indicating a relative surgical contraindication[23,43,44]. Very recently, in a preliminary report, the intraoperative use of ICG clearance tested in patients undergoing staged hepatectomy documented a more accurate prediction of the functional capacity of the FLR than standard preoperative tests[45]. A major limitation in using ICG clearance could be serum bilirubin levels > 6 mg/dL, able to introduce a significant bias to the accuracy of the R15 value, leading to an overestimation of the liver function deterioration[17,43]. Due to the possible overlap of R15 values in patients with and without postoperative liver dysfunction, the ICG clearance test should not be used as a single assessment tool, but together with other clinical and surgical issues relevant to the outcome[44]. As addressed by Angeli-Pahim et al[17], another interesting non-invasive technique (even if not yet extensively used and still to be deepen), recently introduced to measure liver function, is the liver maximum capacity (LiMAx) breath test: It determines enzymatic liver function by measuring the hepatic cytochrome P450 1A2 metabolism of a ¹³C-labelled substrate, methacetin, whose ¹³CO2 byproduct is used to calculate a ¹³CO2: ¹²CO2 ratio: low values are indicative of a reduced metabolic rate, associated with impaired liver function and correlated with patients’ functional liver volume. The ¹³C-methacetin test is currently under scrutiny also during major liver resection[46]. In Table 2 an overview of preoperative tests to estimate RLF (when possible), complexity of the text and if validated for PHLF[3]

Table 2 Overview of preoperative tests to estimate adequate remnant liver function Adapted from Sparrelid et al[3].

Test	Agent used	FLR volume	TL function	FLR function	Distribution	FLR function after	Complexity	Validated for PHLF
Volumetry		Yes	No	No	No	–	–	++
Laboratory scores		No	No1	No	No	–	–	+
ICG test	Indocyanine green	No	Yes	No	No	–	+	++
LIMAX	13C-methacetin	No	Yes	No	No	–	+	+
LIMAX+volumetry	13C-methacetin	Yes	Yes	Yes	No	+	++	+
ICG+volumetry	Indocyanine green	Yes	Yes	Yes	No	+	++	++
HBS	99mTc-Mebrofenin	Yes	Yes	Yes	Yes	++	++	++
RLE-MRI	Gadoxetic Acid	Yes	Limited2	Limited2	Yes	+2	+	+
DCE-MRI	Gadoxetic Acid	Yes	Yes	Yes	Yes	++	+++	–

¹Only sensitive in low liver function/cirrhosis.

²Not absolute.

ICG: Indocyanine green; HBS: Hepatobiliary scintigraphy; RLE: Relative liver enhancement; MRI: Magnetic resonance imaging; DCE: Dynamic contrast-enhanced; FLR: Future liver remnant; TL: Total liver; PHLF: Posthepatectomy liver failure; +: Low; ++: Medium; +++: High.

The holistic evaluation of a candidate is the primary role of the anesthesiologist after the surgeon has indicated the type of resection and assessed the technical feasibility based on the predicted function of the residual liver (remnant volume or dynamic residual liver function) to avoid systemic complications or PHLF. The presence of cirrhosis, if properly assessed, is no longer an absolute contraindication due to advancements in surgical techniques[26,37], increased anesthesiologic expertise in the perioperative period[47], high volume activity, and multidisciplinary collaboration in the indication and scheduling of surgery. Clear communication of risks and benefits to the patients and their relatives and adherence to the surgical proposal are crucial in such a complex setting[48]. Decision-making support is now improved not only by the subjective ASA classification[31,32], but also by more objective evaluation methods based on clinical data supported by large databases such as ACS NSQIP, SURPAS, VOCAL – Penn Risk score (all developed in the United States), or SORT (developed in the United Kingdom)[49-54]. These tools of surgical risk assessment, available online, integrate clinical and physiological data using extensive databases, provide more objective assessments of morbidity and mortality, are validated in the literature, widely accepted in clinical practice, and strongly encouraged by several national and international scientific societies. To be underlined, no comprehensive and exhaustive model still exists[55]. In this context, ERAS in liver surgery should be approached with greater confidence and mandatory multidisciplinary support[56,57]

Factors such as advanced age (> 75 years), comorbidities (particularly metabolic syndrome, NASH, NAFLD, now renamed according to the new definitions which include the word “metabolic” MESH and MEFLD, respectively], diabetes, ischemic heart disease, chronic obstructive pulmonary disease, renal dysfunction, the nutritional status, and frailty), and pharmacological treatments (including antihypertensives, anticoagulants, antiplatelets, and new biological and immunological agents) are frequently prescribed to liver resection candidates. Even very advanced age (well above 80 years) is no longer a contraindication even in major liver surgery, and preoperative assessment should encompass both surgical and anesthesiologic issues to optimize preoperative conditions and prepare for intra- and postoperative strategies[57-60]. Defining the candidate’s frailty and sarcopenia is pivotal, due to the significant impact on both morbidity and mortality[61,62]. The adoption of physical and nutritional prehabilitation programs is a natural consequence of such assessment, is mandatory in ERAS protocols[61], and should be actively pursued.

CARDIOVASCULAR RISK ASSESSMENT

The most recent reviews of preoperative cardiovascular assessment for non-cardiac surgery[31,32], confirm the necessity of a thorough clinical and functional assessment for the high-risk surgical procedures. The risk is determined by both the type of surgery and the presence of comorbidities[53]. Functional assessment using the “subjective” definition of metabolic equivalents (METs) or the more objective Duke Activity Status Index (DASI) questionnaire (cutoff score 34) should now be integral to preoperative anesthesiologic evaluation[53,61,62]. For candidates with suboptimal stress tolerance (usually METs < 4 or DASI score < 34, according to the latest version) or with present or strongly suspected coronary artery disease (CAD), including silent CAD, a baseline echocardiogram followed by provocative testing (stress echocardiogram, myocardial scintigraphy) or, preferably, anatomic imaging [Coronary artery CT scan (CCTA) and coronary artery calcium score] may be considered[31,32,58,63]. Unfortunately, consensus on the optimal pathway for preoperative cardiac assessment in this setting is still lacking, as is in liver transplant candidates[63]. Risks and benefits, as above underlined, should always be discussed with the patients and their relatives. A positive anamnestic history or significant risk factors for ischemic heart disease, or a positive provocative test, deserve: (1) A thorough cardiac evaluation (CCTA vs coronary angiography) possibly involving a cardiologist familiar with perioperative issues; and (2) the determination of the best strategy. In cases of CAD deemed to be treated, the implantation of second-generation drug-eluting stents, followed by a shorter (but safe) dual antiplatelet therapy, is available[64,65]. A multidisciplinary approach and cardiologic advice are mandatory. In urgent contexts (e.g., liver cancer not allowing for a 4-week delay), possible options, always with cardiologist’s advice, include: (1) Preoperative drug optimization (mainly beta-blockers and statins)[31,32,64]; (2) the definition of an intraoperative strategy addressing hemodynamic, metabolic, and pharmacological issues using tailored invasive or minimally invasive monitoring[58,63]; and (3) postoperative ICU observation.

Cirrhotic cardiomyopathy, recently redefined, is increasingly diagnosed in liver surgery candidates and warrants particular attention, particularly in cases of moderate-to-severe diastolic dysfunction[66]. Due to the increasing age of candidates (now often well above 80 years) and broader surgical indications, valvular heart diseases, hypokinetic or dilated cardiomyopathies, complex cardiac arrhythmias, and the presence of cardiac pacemakers or implanted cardiac defibrillators are not uncommon. Intraoperative strategies with a proactive approach should include appropriate knowledge of the cardiac devices and tailored hemodynamic monitoring. Monitoring devices able to anticipate intraoperative arterial hypotension, a major risk factor for increasing perioperative morbidity and mortality, are now available, even if still under hot discussion (particularly during liver surgery, due to sudden surgical manipulations/maneuvers able to induce sudden critical hypotension)[67-69].

ASSESSMENT OF ISSUES RELATED TO RESPIRATORY AND GAS EXCHANGE ALTERATIONS

Protocols for preoperative evaluation of respiratory function and gas exchanges in major liver surgery are similar to those used for major abdominal surgery[70-72], despite the higher incidence (20% to 45%) of postoperative respiratory complications in liver surgery, particularly when performed as an open procedure[71]. Clinical history, physical examination, and thorough review of arterial blood gases (ABG) are pivotal. Functional cardiorespiratory assessments, such as Cardiopulmonary Exercise Testing (CPET), or the 6-minute walking test, along with a frailty assessment, are becoming mandatory and should be incorporated into the ARISCAT scores[72]. Imaging and respiratory function tests should be performed only if indicated by these assessments. Lung ultrasound is increasingly used during the preoperative evaluation and as a point-of-care (POC) tool[73]. Respiratory issues related to changes in respiratory exchanges typical of cirrhotic patients and advanced liver disease (hypoxia), associated with an increased incidence of postoperative respiratory complications, must be addressed. In cases of hypoxia (SaO₂ < 97%), conditions associated with Chronic Liver Disease (CLD) such as hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PoPH) should be ruled out, obstructive (common in the elderly) or restrictive (associated with ascites or pleural effusion) pathologies being much more common in the average candidate[74,75].

HPS is characterized by varying degrees of hypoxia (mild: PaO2 < 80 mmHg; moderate: PaO₂ 60-80 mmHg; severe: PaO₂ 50-60 mmHg; extremely severe: PaO2 < 50 mmHg). It is associated with an increased alveolar-arterial oxygen gradient (> 15 mmHg in room air) due to intrapulmonary vascular dilatation (shunt). Symptoms include dyspnea at rest or upon exertion, with platypnea (dyspnea and desaturation in orthostatic position) and orthodeoxia (reduction of PaO2 from supine to orthostatic position) present in 25%-30% of cases. Screening and further investigation are required for SaO2 < 96% in room air. At echocardiography, echocontrastography with bubbling and late passage of bubbles into the left atrium is diagnostic. While there is no definitive treatment, Methylene Blue may be considered for refractory hypoxia (anecdotal reports)[58,74].

PoPH, associated with portal hypertension, is present in 2%-5% of CLD patients and involves anatomic changes in the pulmonary vascular bed and increased circulating pulmonary vasoconstricting agents (e.g., endothelin-1)[58,63,75]. Diagnosis is based on the presence of mean pulmonary pressure (mPAP) > 25 mmHg and pulmonary vascular resistance > 240 dyne/s/cm^-5 with central venous pressure (CVP) < 5 mmHg and pulmonary wedge pressure < 15 mmHg. PoPH is classified as mild (mPAP 25-35 mmHg), moderate (mPAP 35-45 mmHg), or severe (mPAP > 45 mmHg). In cases where transthoracic echocardiogram (TTE) estimates systolic pulmonary pressure > 45 to 50 mmHg, consultation with a cardiologist and further investigation with right heart catheterization are mandatory. Among the available treatment options are phosphodiesterase-5 inhibitors, prostaglandins, endothelin-1 receptor antagonists, and nitric oxide for the perioperative period. Intra- and postoperative anesthesia strategies, developed in collaboration with the cardiologist and the surgeon, are mandatory. The perioperative risk is significantly increased in severe pulmonary hypertension due to acute right heart failure[58,63,75].

ASSESSMENT OF RENAL FUNCTION

The occurrence of acute kidney injury (AKI) after major liver surgery can be multifactorial and is a significant contributor to postoperative mortality[76-80]. Factors such as intraoperative hypovolemia, blood loss, massive transfusion, prolonged arterial hypotension, and the use of nephrotoxic substances (including contrast media) can worsen an already suboptimal renal function, due to the CLD. Hepatorenal syndrome (HRS, now HRS - AKI), is characterized by intense renal vasoconstriction, requires careful evaluation with hepatologists, and involves treatment with terlipressin and albumin[55,81]. Before the development of new AKI criteria[77], HRS was divided into two types, with different prognoses, type 1 having the worst outcome[55,81]. A renal function profile, including serum creatinine and glomerular filtration rate, along with serum and urinary electrolytes (particularly sodium and potassium, with Na < 125 mEq/L not uncommon in CLD patients), is necessary to quantify actual renal function and classify any acute insult using the AKI/KDIGO classification[77]. Serum creatinine may present interpretative concerns in sarcopenic patients and/or in the hyperbilirubinemic cirrhotic patient, limitations to be considered when assessing preoperative renal function. Often, blood urea nitrogen (BUN) is discrepant with serum creatinine (elevated BUN vs “normal” creatinine), and might provide a more accurate description of renal dysfunction. A predictive score of postresection renal damage developed by the Clavien’s group, though not widely used, could be an interesting predictive tool to finalize intraoperative hemodynamic monitoring and subsequent postoperative treatment[78]. The best approach to protect renal function during the perioperative period includes maintaining mean arterial pressure > 65 mmHg, to ensure renal perfusion pressure - and tailoring fluid balance and use of pressors implementing appropriate invasive hemodynamic monitoring (invasive arterial pressure and minimally invasive cardiac output if indicated). The use of mannitol and dopamine, although still considered, lacks evidence. Dopamine may be considered for its beta effect at 4-5 ug/kg/min in cases of reduced cardiac output to increase cardiac output and renal perfusion. In cases of extreme vasodilation (high cardiac output with arterial hypotension), vasoconstrictors such as norepinephrine, vasopressin, and terlipressin are the drugs of choice[55].

EVALUATION OF THE HEMOSTATIC PROFILE

The hemostatic profile of a non-cirrhotic patient undergoing liver surgery is expected to be normal. In cirrhotic patients, however, the profile is not “naturally” anticoagulated , but instead “rebalanced,” with coexisting procoagulant (increased factor VIII and von Willebrand factor, reduced ADAMTS-13, reduced natural anticoagulants like Antithrombin, Protein C, Protein S) and anticoagulant features (reduced synthesis of coagulation factors, hypopiastrinemia, impaired platelet function, hypofibrinogenemia)[82-85]. The coagulation profile is often activated (increased d-dimer), with the risk of consumption coagulopathy. Increased fibrinolysis and endogenous heparin-like products may also be observed[85]. Cirrhotic patients can be prone to thrombosis, as shown by portal and/or mesenteric thrombosis. Prolonged PT/INR and aPTT/R do not necessarily imply an increased hemorrhagic risk, but rather an “instability” of the hemostatic balance, which can shift towards a prothrombotic or prohemorrhagic state depending on the type of insult (proinflammatory, surgical, infectious, traumatic)[82-85]. The use of viscoelastic tests (TEG or ROTEM) is nowadays pivotal in clinical practice to define the hemostatic profile, to guide interventions, to provide information on the type of defect, enabling both the in vitro and vivo assessment of treatment effectiveness[86,87]. This has led to the abandonment of the “prophylactic” use of fresh frozen plasma (FFP) and the administration of blood products and blood components only on specific indications[65,86,87]. The transfusion strategy, both in terms of hemoglobin correction (transfusion of packed red cells, PRCs) and FFP use, plays a significant role in perioperative patient blood management. Currently, a “restrictive” transfusion philosophy is strongly supported[65]. The opportunity for objective and rapid feedback using multiparametric monitoring, combining cardiac, respiratory, metabolic, and hemostatic profiles, should be strongly considered. The intraoperative anesthesia strategy should aim at reducing blood loss and transfusion needs, both critical factors in the onset of PHLF. For an extensive and recent review of this topic, see Intagliata and Shah[84] and Roberts[85].

Postoperative period and the ICU: A controversial issue

While intraoperative anesthesia management is well-established[1,27,47,58], there remains a significant debate regarding the elective indication for postoperative intensive care observation, ideally to be planned during preoperative evaluation[88-93]. The logistics, layouts, and organization of each institution, even if seldom considered, play a crucial role in this context. Due to the lack of standardized protocols, postoperative observation in a dedicated surgical ICU has not yet reached a consensus so far and is often decided on a case-by-case basis[56-58,88-92]. Key areas of discussion among surgeons and anesthesiologists include:

Patient condition and operative risk: The patient’s overall condition, including frailty and sarcopenia, as well as the definition of operative risk and specific surgical technical challenges, contribute to defining the “high risk” status of a patient, to be defined in the most objective way and pivotal to decide where to care for the patient in the early postoperative period[53,93].

Potential postoperative organ failures: The chance of organ failures arising from intraoperative complications needs careful consideration[1,47,58,61].

Post-anesthesia care unit facilities: The availability of a post-anesthesia care unit (PACU) equipped for both invasive and non-invasive ventilation, invasive hemodynamic monitoring, and anesthesiologic and nursing care is pivotal. An “overnight PACU” for extended care or in case of late discharge from the operatory room is critical. Stable cardiorespiratory, renal, and metabolic parameters, along with optimal pain control, are essential for a safe discharge to the ward or step-down units[1,47,56,88-93].

Stepdown units availability: The presence of “stepdown units” (Surgical High Intensity Care wards) with facilities for non-invasive ventilation (High Flow, CPAP, or NIV), monitoring of vital parameters (EKG, respiratory rate, SaO2, and in advanced settings, invasive arterial blood pressure and CVP, as indicated), and POC analyzers for rapid (and if required, frequent) hematocrit/hemoglobin, glucose, creatinine, electrolytes, ABG/Acid-base equilibrium [ABE], and hemostatic testing tests are crucial[89-92]. Patients in these units can receive tailored medical and nursing care during the first postoperative days. Effective management in the first 12-24 postoperative hours can significantly impact the subsequent hospital stay and potentially PHLF.

MAJOR LIVER SURGERY: ANESTHESIOLOGIC MANAGEMENT

In major liver surgery, key objectives of intraoperative anesthesia management include maintaining adequate anesthesia depth and analgesia, core temperature control, cardiometabolic stability (stable circulatory, respiratory, and metabolic profiles), appropriate fluid balance, implementation of active strategies to minimize blood loss[1,27,47,58,65]. Anesthesiologic interventions should be driven by the understanding of the surgical maneuvers liver manipulations, vascular clamping and declamping and their hemodynamic and metabolic consequences. Intraoperative monitoring should be tailored towards identifying and pre-emptively addressing cardiorespiratory and metabolic changes, ideally anticipating issues like hypotension, hypoperfusion, and peripheral oxygen debt[47,53,58,93,94]. Hemostatic monitoring, as above underlined, is critical, particularly in CLD patients. Transfusion policy (PRCs/FFP/Platelets) should consider the actual hemostatic profile, based on both “static” (laboratory or point-of-care) and “dynamic” (VETs) data[47,58,65,86,87].

Pharmacological management in liver surgery, for both CLD and non-CLD patients, has recently been reviewed[1,47,58]. Evidence suggests that ischemic preconditioning and even “postconditioning” (adaptation to ischemic injury and remodulation of ischemia-reperfusion syndrome) can be facilitated by volatile anesthetics like sevoflurane, isoflurane, and desflurane[95-97] even if lights and shadows are present[98]. However, there are also recent favorable experiences with the use of totally intravenous anesthesia in major liver surgery (propofol/remifentanil)[99,100]. To note, no advantage of either technique on neoplastic recurrence has been confirmed as per the most recent review[100]. Thus, rather than an “ideal drug combination,” the “best-known combination of drugs” should be chosen. Preference should be given to drugs eliminated by the kidneys and/or those whose discontinuation of action depends on redistribution, including propofol, fentanyl, sufentanil, remifentanil, and cisatracurium. Concerning the use of muscle relaxants, intraoperative neuromuscular monitoring (NMM) should be considered mandatory: While the use of rocuronium appropriately antagonized by sugammadex can be strongly suggested, NMM should always be used[1,47,58,101].

HEMODYNAMIC ISSUES AND CONSEQUENCES OF VASCULAR OCCLUSIONS

During liver resection surgery, common causes of hemodynamic instability include blood loss during isolation, dissection, and resection, liver mobilization maneuvers (“dislocation”), and compressions or distortions of the IVC, able to impair venous return. Gas embolization, though rare, can occur in cases of very low CVP/extreme hypovolemia during exposure of large venous vessels[27,102,103]. Anesthesiologists involved in liver surgery should master the hemodynamic changes associated with the different vascular clamping maneuvers[27,103] (Figure 2).

Figure 2 Vascular occlusion techniques in hepatic surgery to reduce hemorrhage during hepatic resection. A: Pringle maneuver to occlude hepatic arterial and portal venous inflow to the liver; B and C: Selective hepatic vascular exclusion involves clamping of the vessels perfusing the hemi-liver which is being resected; D: Total hepatic vascular exclusion, clamping the inferior vena cava above and below the liver along with the hepatoduodenal ligament; E: Variant technique combining clamping of the infrahepatic Inferior vena cava with a clamp across the hepatoduodenal ligament (from Njoku DB, Chitilian HV, Kronish K. Hepatic Physiology, Pathophysiology, and Anesthetic Considerations. In Miller’s Anesthesia Michael A. Gropper, Ronald D Miller, Neal H. Cohen Lars I. Eriksson Kate Leslie, Jeanine P. Wiener-Kronish NINTH EDITION. Elsevier 2020: 420-443[27] ; and with permission).

The Pringle maneuver, which involves occlusion of the hepatic vascular pedicle (the portal triad) (inflow occlusion), aims at reducing blood loss, although its effects on morbidity and mortality are unproven at best (but not associated with an increased rate of PHLF)[104,105]. The Pringle maneuver may contribute to acute liver injury and could potentially delay regenerative activity. It is associated with a reduction in venous return and CO of about 15%, a change usually well tolerated due to an increase in orthosympathetic tone. Systemic hypotension is uncommon, a modest increase in systolic blood pressure may occur, due to increased systemic vascular resistance. Clamping can be continuous (15 to 30 min) or intermittent, the latter involving clamping periods of 10 to 15 min followed by declamping periods of 5 to 10 min. During declamping, hepatectomy is usually interrupted, with the section slice compressed with laparotomy patches to ensure hemostasis. With intermittent clamping, a condition akin to preconditioning is created. The duration of clamping should be reduced by about 50% in cirrhotic livers[103-105].

While selective vascular occlusion includes clamping of the inflow of the resected hemiliver, total clamping involves occlusion of the infrahepatic and suprahepatic IVC, with a significant blood pressure drop (40 to 50 percent reduction, leading to critical hypotension) and a substantial decrease in venous return and cardiac output (up to 50 percent) with compensatory tachycardia. The duration of clamping can be up to 100 minutes but should be limited to no more than 50 min in CLD patients. Measures in response to the hemodynamic changes induced by total clamping include moderate fluid administration and the use of catecholamines (norepinephrine) to maintain mean arterial pressure (MAP) ideally above 65 - 70 mmHg. Very rarely, and usually only in extreme emergencies (uncontrolled blood loss), aortic clamping should be considered[27,47,58,103].

Multidisciplinary planning of surgery should anticipate the use of high impact clamping techniques, allowing the anesthesiologist to be prepared, and in close coordination with the surgeon. While overly restrictive fluid strategies early during surgery should be avoided, managing increased venous pressure in the splanchnic district due to excessive fluid therapy (high CVP and risk of bleeding from “back flow” in the presence of high splanchnic venous pressure) is crucial[106]. Hemodynamic guidance with dynamic parameters (stroke volume variation (SVV), pulse pressure variation (PPV), minimally invasive cardiac output (CO), stroke volume (SV) and even CVP (measured in mmHg and not in cm H₂O!!) in complex cases is essential for managing rationally the resection phase[103,106-109]. Thus, as above alluded to, the importance of cardiac function assessment in the preoperative period and the need for tailored intraoperative hemodynamic monitoring to anticipate tolerance of vascular exclusion are paramount. In cases of extended clamping, understanding the systemic cardiovascular consequences through hemodynamic monitoring is vital, making the use of a Swan-Ganz catheter or, if available, transesophageal echocardiography (TEE), crucial[110].

HEMODYNAMIC MONITORING

Perioperative hemodynamic optimization, including appropriate setting of mechanical ventilation, involves protocols for fluid challenge using changes in CO, SV, and/or dynamic. A fluid challenge of 100-250 mL, followed by evaluation of changes in dynamic parameters, CO, or SV (preferably indexed), can assess “fluid responsiveness” and contribute to optimizing circulating blood volume, positively impacting outcomes[94,107,110-120].

In liver surgery, minimally invasive monitoring of CO, left heart function, and dynamic parameters can largely address the controversy around maintaining low CVP (< 5 mmHg) during liver dissection, a standard technique to reduce blood loss, debated if used without an appropriate rationale[1,47,58,110-116,118].

Invasive arterial pressure monitoring is considered mandatory to track critical and abrupt variations, mainly hypotensive episodes associated with increased complications[1,47,58,103,117]. The recent application of machine learning algorithms to continuous intraoperative arterial pressure monitoring can anticipate hypotensive episodes, a crucial topic still under debate, particularly during liver surgery due to sudden surgical manipulations/maneuvers[67-69,117]. Conversely, CVP monitoring aiming at low values (< 5 mmHg), is now questioned. CVP values can be reduced implementing different strategies (increased diuresis, restrictive fluid management, anti-Trendelenburg position, use of vasodilators, post-induction phlebotomy) but able to induce side effects[111,113,115,116,118]. Zhu et al’s observation on postoperative hyperbilirubinemia associated with “low” CVP values and splanchnic hypoperfusion should be considered[115]. Values of SVV or PPV > 13% (normal < 9%, with a grey zone between 9 and 13%) are more reliable than the absolute value of CVP in maintaining a relative hypovolemic state while keeping CO within normal limits[112,114-116]. Dynamic parameters, although with “grey zones”, allow better assessment of hypotension (hypovolemia vs vasodilatation vs contractile dysfunction) and the rational use of fluids, vasopressors, and inotropes[114,115,119,120]. Interestingly, the use of dynamic parameters has been combined with peripheral venous pressure monitoring (via an antecubital vein) with satisfactory results[112,114-116]. In less complex cases, and in patients with a normal cardiac profile, the use of PPV derived from invasive arterial pressure is appropriate, while the cardiac index and the stroke volume index (SVi), as used in goal-directed hemodynamic therapy, could be game changers in high-risk patients[53]. The proper setting of mechanical ventilation (particularly tidal volume not less than 7-8 ml/kg) is crucial for the reliable interpretation of dynamic parameter changes[47], although the “grey zone” (between 9% and 13%) may introduce uncertainty in the interpretation of hemodynamic data in up to 25% of cases[114]. An alternative to dynamic parameters, again using heart-lung interaction, is the use of SVi variation after a fluid challenge, a “bedside” application of the Frank-Starling curve[53,107,119,120]. Noninvasive devices based on volume clamp or bioreactance techniques for arterial pressure monitoring, cardiac output monitoring, and fluid responsiveness are available, but more robust studies are needed to support their use in major liver surgery. For an updated review on hemodynamic monitoring, please refer to Pinsky et al[119] and Martin et al[120].

In recent reviews, the advantages and limitations of the Swan-Ganz catheter in non-cardiac surgery have been discussed[121,122]. In major liver surgery, its use should be limited to special or extremely complex situations or in cases of PoPH, in any case rare even in cirrhotic patients[110]. The use of intraoperative TEE can aid in interpreting hypotension episodes[110,123]. TEE provides direct vision and information on cardiac structures and functions optimizing the management of “relative hypovolemia” during dissection. The presence of esophageal varices is not an absolute contraindication to TEE, with minimal complication incidence in recent reports[110,123]. However, its use is limited by high costs, mandatory training, not yet widespread availability, and lack of evidence of superiority over routine monitoring. Nevertheless, TEE could represent the future of intraoperative monitoring for complex surgical cases, and it could be combined with the Swan-Ganz catheter[110]. The ideal monitoring in liver surgery is probably yet to be defined. As above underlined, continuous and reliable invasive arterial pressure monitoring is mandatory. The application of dynamic parameters and minimally invasive monitoring of cardiac output (when indicated) may be appropriate for volume optimization during central phases of resection, replacing CVP. Point-of-care systems for glycemia, electrolyte, ABE/ABG analysis allow reliable perioperative respiratory and metabolic monitoring, a “holistic” approach, crucial for managing the complex intraoperative period of liver resection.

Continuous central temperature monitoring is another relevant issue during liver resections, to make real-time adjustments to prevent hypothermia, contributing to improved outcomes. Maintaining the patient's optimal core temperature is paramount for a successful surgical outcome and for the patient's comfort and safety[118].

FLUID BALANCE MANAGEMENT

In modern anesthesia practice during major abdominal surgery, including liver resection, the restrictive fluid therapy is championed as a mainstay of anesthetic management: At the end of surgery, the best outcomes are associated with even or only slightly positive balance[118,124-129]. The key aspects of intraoperative fluid management are quantity and quality of fluids. The strategy involves targeted fluid administration and metabolic changes monitoring, while avoiding the “fixed hourly aliquot” in favor of balancing a wise restrictive policy during liver resection with a more liberal approach in the terminal phase of surgery, mainly implementing goal-directed therapy[126-128], left ventricular diastolic dysfunction potentially playing a role in early postoperative complications[129]. The choice of the “ideal fluid” is still debated. Crystalloids are commonly used, with a preference for balanced solutions (acetate perhaps better than lactate as bicarbonate- forming substrate)[125,126]. However, attention should be paid to the onset of postoperative metabolic alkalosis, common after major surgery in general and after liver resections, and compensated for by respiratory acidosis (secondary to hypoventilation). In such cases, a targeted use of saline solution might be appropriate. The use of colloids is controversial[130-132]. Hydroxyethyl starches may have a role in elective surgery, but recent discussions about potential adverse effects on kidney function and hemostatic profiles limit their use: conflicting evidence exists in surgery in non-critically ill patients[131]. Gelatins, often used in current practice, may alter coagulation and cause allergic reactions. Albumin solutions (addressing the different tonicity, 4%, 5%, 20%) might be suitable in the perioperative period of surgery on cirrhotic patients[132].

POSTOPERATIVE ANALGESIA

Open resective surgery historically involves significant postoperative pain. Thoracic epidural blocks (TEB) have become popular, reducing respiratory complications and systemic sympathetic response when appropriately used. However, widespread adoption of TEB has been limited by concerns about perioperative hemodynamic instability and risk of spinal hematoma (secondary to possible postoperative hemostatic alterations): recent ERAS guidelines no longer recommend TEB due to the evidence of effective alternatives[1,103,116,133-138]. Hence, there is growing interest in alternative techniques such as continuous local infiltration with wound catheters, “single shot” subarachnoid analgesia with local anesthetics and intrathecal opioids, and patient-controlled analgesia[136,138]. Transversus abdominis plane and erector spinae plane blocks are promising alternatives[103,137]. The choice of perioperative analgesia largely depends on the underlying liver disease, the patient’s condition and comorbidities, and the type of surgery[138].

NOTES ON THE EPIDEMIOLOGY AND PATHOPHYSIOLOGY OF PHLF

The constant increase in the number of liver resections (close to 3000 per year in United Kingdom) should be attributed to the extremely positive results, to the improved surgical techniques, and to a targeted perioperative medicine[116]. However, major liver surgery remains, as already underlined, a high-risk procedure, with crude mortality rates between 2 and 5% (up to 10% in more complex cases) and morbidity ranging from 20% to 40%. Common postoperative complications include infection(s), bleeding, biliary leakage, renal and respiratory or cardiac issues. In this specific context, the use of the Comprehensive Complication Index[139] together with NSQIP, as addressed by Donadon et al[26] and Wang et al[140] are promising and of extreme interest.

PHLF is the most feared complication within the first 30 postoperative days, particularly following resections for HCC or hilar cholangiocarcinoma, with mortality rates of 30% and 40%, respectively. The incidence of PHLF—varying from 1% to 32% in the literature—is influenced by the definitions, the types of surgery, and case mix, with a realistic incidence ranging from 8% to 12%[140-143]. Recent advancements in artificial intelligence (AI) techniques are being explored to reduce the occurrence of PHLF[144]. PHLF is usually characterized by rapid functional deterioration (3-5 d after surgery) of the liver remnant sustained by a complex interplay among various factors[3,4,140-143]. The relevant risk factors for PHLF are patient-related (preoperative comorbidities, preexisting liver disease, and liver functional reserve) and/or surgery-related (the extent of liver resection). At the base of the functional (and volumetric) recovery of the liver remnant is the regenerative response induced by major hepatectomy and mainly based on a compensatory hyperplasia of the residual hepatocytes. According to Sparrelid et al[3] who extensively revised the item, the “physiological” regeneration of the liver after hepatectomy is supposed to be triggered by an increase in PV pressure, secondary to increased portal inflow. The reduced liver volume produces vascular shear stress and increased intrahepatic vascular resistance. The shear stress changes the perfusion of liver sinusoidal endothelial cells, leading to the release of NO and other hepatotrophic factors, together with cytokines [tumor necrosis factor alpha (TNF-α), interleukin 1 (IL-1), and IL-6] released from Kupffer cells. Instead, an overwhelming increase in portal pressure might lead to a dysfunctional regeneration due to a ‘small-for-flow’ syndrome (excessive flow for the actual liver volume)[3]. Excessive shear stress induces an over expressed inflammatory response, followed by neutrophil recruitment into the liver. The main effects are inhibition of liver regeneration, parenchymal necrosis, and hepatocyte apoptosis. The mechanisms involved in this altered response are activation of specific pathways [among others, the hepatocyte growth factor/c-Met, epidermal growth factor (EGF)/epidermal growth factor receptor (EGFR), and transforming growth factor alpha/EGFR pathways]. Such an “inflammatory storm,” induced by the surgical trauma, is amplified by ischemia-reperfusion injury, via generation of reactive oxygen species, release of inflammatory mediators, and activation of apoptotic pathways[3]. The dysregulated release of proinflammatory cytokines and chemokines is responsible for tissue damage, reduced regenerative capacity, and consequent dysfunction of the liver remnant. Further ischemic-hypoxic damage is associated with microvascular thrombosis and endothelial dysfunction of the surgically disrupted liver microcirculation, both able to impair hepatic microcirculation. Then, PHLF includes both liver injury and dysfunction. The dysfunctioning liver is associated with reduced synthetic capacities (reduced synthesis of albumin, clotting factors, and natural anticoagulants), impaired metabolic (hyperbilirubinemia) and detoxification (hyperammonemia) functions, ascites, an altered hemostatic profile, and hepatic encephalopathy. Systemic complications and multiple organ dysfunction syndrome occur in moderate-to-severe PHLF, including acute renal and respiratory failure, bacterial and fungal infections, and superimposed sepsis/septic shock.

Among the models still considered for PHLF is the “50-50 criteria on postoperative day 5” proposed by Belghiti’s group in 2005, based on bilirubinemia > 50 umol/L [> 3 mg/dL] and prothrombin time < 50%, and validated in 2009 is a reliable predictor of post-hepatic resection mortality[145,146]. The International Study Group of Liver Surgery (ISGLS, 2011)[147] provides the reference standard for PHLF, using three degrees of severity (A, B, C) based on INR prolongation, bilirubin increase, elevated lactate, and hepatic encephalopathy levels. Level C PHLF, requiring artificial support of organ failures, mandates ICU observation, level B being considered for HDU observation; mortality rate (absent in group A) is 12% for level B, reaching 54% for level C[147].

As mentioned above, perioperative factors associated with PHLF have been identified[3,148-156], and relate mainly to the patient’s comorbidities, liver remnant volume, intraoperative issues, and early postoperative complications , including sepsis/septic shock (Table 3).

Table 3 Perioperative risk factors for posthepatectomy liver failure (adapted from Sparrelid et al[3]).

Patient-associated
Sex	Risk double in males, especially males with HCC
	Female hormones show proliferative effect in animal models, inhibiting effect of testosterone on immune system
	NAFLD, lower postoperative risk than other chronic liver diseases, higher incidence in postmenopausal women
Age	Still unclear, possible changes in bile flow and acute-phase protein production
	Age-related sinusoidal pseudocapillarization, rescue in animal models through serotonin agonist injection
Sepsis	Bacterial endotoxins decrease cytokine production needed for liver regeneration
	Kupffer cell and hepatocyte function in liver regeneration inhibited
Metabolism	Insulin induces expression of IGF and HGF
	High BMI and malnutrition associated with PHLF
Other	Serum bilirubin, low platelets, insufficient renal function, cardiopulmonary disease, associated with PHLF
Liver-associated
Steatosis	Leads to changes in the hepatic microenvironment and higher risk for ischaemia–reperfusion injury
Neoadjuvant chemotherapy	Chemotherapy-associated liver injury and steatohepatitis are known complications after neoadjuvant chemotherapy
Fibrosis grade	Functional liver tissue reserve is reduced, patients often present with several comorbidities
Cholestasis	Jaundice increases morbidity after surgery; in animal models, bile duct ligation leads to reduced growth factor expression
Portal hypertension	High preoperative portal pressure in cirrhosis associated with increased risk of PHLF
Surgery-associated
Future liver remnant	‘Small-for-flow’ syndrome negatively impacts hepatic haemodynamics
	Increase in portal pressure leads to altered hepatic microcirculation and hepatocyte damage
Blood loss	Leads to intravascular fluid shifts, introduction of bacterial endotoxins into the hepatic microenvironment
	Increased risk of sepsis, coagulopathy and PHLF
Surgical technique	Vascular occlusion can cause ischaemia–reperfusion injury and in increases PHLF risk
	Long Pringle manoeuvre leads to increased oxidative stress and overshooting inflammatory response
	Extensive vascular resection can cause PHLF

HCC: Hepatocellular carcinoma; NAFLD: Non-alcoholic fatty liver disease; IGF: Insulin-like growth factor; HGF: Hepatocyte growth factor; PHLF: Posthepatectomy liver failure.

Residual liver function after resection is critical in the onset of PHLF. While rapid regeneration occurs within two to four weeks in a normal liver, even with resections up to 75% of the native liver mass (60% if normalized for body weight), the situation is different for a cirrhotic liver. Major resections in cirrhotic patients should be limited to CHILD A patients, without portal hypertension and with platelet counts > 100000/uL while MELD score > 11 predicts a high likelihood of PHLF. More problematic are CHILD B patients, for whom only minor resections are should be advisable. However, a recent international multicenter trial was able to propose a model to include very selected CHILD B patients to liver resection with an extremely interesting 5 years survival rate (> 47%) and easily available prediction models for overall survival (https://childb.shinyapps.io/survival/) and disease-free survival[150]. Algorithms using R15 indocyanine green (ICG) values (R15 ICG > 15% indicating a risk of PHLF) remain important predictors of short- and long-term mortality after liver resections for HCC. Clinical (CP, MELD, ALBI, APRI+ALBI) and functional assessments of FLR with imaging or dynamic techniques are well-established, but their application varies among centers[42,149,151-156].

PHLF: DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS

The diagnosis of PHLF is based on a range of altered liver function tests, including increased lactate, altered lactate clearance kinetic, hyperbilirubinemia, ascites, altered hemostatic profile, impaired renal function, and encephalopathy. The onset of PHLF can vary, typically occurring days to weeks post-surgery, with normalization expected within 1 to 3 weeks[141]. Differentiating PHLF from a septic complication can be challenging, as sepsis may coexist or even be the underlying cause[141,147]. Factors such as endotoxin release from visceral ischemic injury, bacterial translocation, and loss of macrophage function can contribute to increased susceptibility to sepsis or septic shock[152].

In the differential diagnosis, the so called “critical” patient may present with similar signs. A thorough clinical history, both remote and recent, is essential in guiding clinical judgment and supporting the differential diagnosis. Intraoperative factors that favor ischemic injury and thus PHLF of the remnant liver include hepatic venous congestion, arterial hypotension, significant blood loss and transfusion requirements, and prolonged periods of vascular occlusion, leading to ischemia-reperfusion injury. These factors contribute to ischemic hepatitis and exacerbate the “small-for-size” condition, the remaining liver volume being inadequate to support the required function. Hyperflow syndrome (hyperperfusion syndrome), more typical of a true “small-for-size” syndrome and reported in the early postoperative phase, is another factor that should be considered, albeit potentially confounding[146]. Other contributing factors to hepatic congestion and “stiffness” (secondary to venous stasis rather than to fibrosis)[141] include overly aggressive mechanical ventilation with high positive end-expiratory pressure (PEEP > 12-15 cm H₂O), fluid overload, and right ventricular dysfunction (as might be observed in cholestatic liver failure after cardiac surgery).

While reliable predictors of PHLF (such as biomarkers or instrumental investigations) are eagerly awaited, the prognostic evaluation of PHLF remains challenging[147-149]. In 2013, Hyder et al[153] proposed a composite integer-based risk score based on postoperative day 3 INR, bilirubin, creatinine, and complication grade (area under the curve 0.89; a score of ≥ 11 points with sensitivity and specificity of 83.3% and 98.8%, respectively).

In addition to scores and ICG, liver stiffness measurement obtained by ultrasound is now proposed and is awaiting validation as independent predictors of PHLF and residual function[154]. In a recent review, no convincing biomarker(s) assessment is available, as yet[152]. Predictive nomograms aiming at reliable risk assessment should represent an important clinical option, particularly if available prior to surgery[153-157]. Shen et al[155] in 2019 were able to develop a nomogram to predict PHLF based on retrospective preoperative data (portal hypertension, bilirubin, creatinine) and intraoperative factors (intraoperative blood loss). This nomogram, categorizing three risk groups, largely outperformed both the ALBI score and MELD, the need for intraoperative data being its major limitation. Two very recent studies are worth to be addressed[4,157]. Baumgartner et al[4] proposing a preliminary model that includes clinical risk factors and bilirubin on postoperative day one (POD1) found a high discriminatory potential for PHLF grade B/C in major resections (AUC 0.87). Having a discriminatory potential for major morbidity and 90-day mortality as the established ISGLS criteria for PHLF grade B/C, the score could be implemented starting from POD1, thus allowing an earlier identification of patients at high risk of adverse outcomes, potentially optimizing the early postoperative management. Even more interesting should be the APRI+ALBI multivariable model. Santol et al[156] studying more than 12000 patients from the NSQIP database generated a multivariable model to predict PHLF B + C using aspartate aminotransferase to platelet ratio (APRI) combined with albumin-bilirubin grade (ALBI). Its performance was compared to indocyanine green clearance (R15 or ICG PDR) and albumin-ICG evaluation (ALICE). The model including easily available preoperative parameters (APRI+ALBI, age, sex, tumor type and extent of resection) was able to predict PHLF B+C with an AUC of 0.77. With a PHLF B+C predicting capacity equal to the matched, more expensive, and complicated tests and the availability of a smartphone application, APRI+ALBI, if validated, could become an important step forward in this context.

Given that no definitive “point of no return” has been established for PHLF[149], and in the absence of a single prognostic biomarker, an integrative approach with multiparametric models and multidisciplinary strategies are pivotal to anticipate, prevent, or aggressively manage PHLF: AI and machine learning approaches , now being explored, could become the game changer in predicting and preventing PHLF[144]. Mastering intraoperative issues during anesthesia should close the safety loop.

TREATMENT OF PHLF

The International Study Group of Liver Surgery (ISGLS) classification[147] guides the definition and logistics of PHLF treatment, emphasizing the importance of identifying organ failures to determine the most suitable clinical care setting (ward, high-intensity/semi-intensive areas, or ICU) for the optimal treatment. The implementation of preoperative (or very early postoperative) predictive models might play, if validated and extensively used, a relevant role in PHLF prevention, anticipating facilities and settings to treat the acute liver failure or its prodromes, of which PHLF is considered a phenotype by many experts[141].

According to Soreide and Deshpande there are “remediable” causes of PHLF, and multidisciplinary approaches play a crucial role[141]. Obstructive jaundice should be managed with the collaboration of interventional radiologists and/or digestive endoscopists (biliary stenting). Aggressive and timely ultrasonographic and radiological diagnostics, including CT scans and angiographies, are vital to rule out vascular obstructions in the arterial and portal inflow and venous outflow tracts[141,149]. In such scenarios, the role of the intensivist is crucial, to (artificially) support vital functions while managing thrombectomies, thrombolysis, and anticoagulation therapies.

Excluding infectious causes, such as perihepatic collections or abscesses, is pivotal for the differential diagnosis. This process should include aggressive and invasive ultrasonographic and radiologic workups and, if needed, invasive radiologic or endoscopic procedures to have relevant and reliable biological samples for immediate microbiological cultures[157]. Diagnosis of infectious complications includes the use of biomarkers such as procalcitonin (PCT) and C-reactive protein, bearing in mind their potential “noninfectious” elevation in the early postoperative period and the importance of trend analysis rather than single point values. In cases of definite or strongly suspected infections (evidenced by leukocytosis, increased PCT, tachypnea, hypotension requiring vasopressors, worsening renal failure, or sensory changes not solely attributable to hepatic encephalopathy), appropriate imaging and blood and/or collected samples should be sent for cultures. Broad empirical antibiotic coverage should be started promptly, followed by targeted treatment based on isolated microorganisms, as per the 2021 Surviving Sepsis Campaign guidelines[157]. Antibiotic use should be closely monitored, ideally implementing antibiotic stewardship and therapeutic drug monitoring, with appropriate de-escalation therapy and discontinuation measures when appropriate[158,159]. PCT levels and clinical trends should guide the treatment process as per the Surviving Sepsis Campaign guidelines[157].

THE ARTIFICIAL SUPPORT OF ORGAN FAILURE ASSOCIATED WITH PHLF

According to recent reviews in the ICU setting, there are several recommendations/suggestions for artificial organ support in the context of PHLF, in large part mutuated from ALF/ACLF maximal intensive treatment [141,160,161]:

Cardiovascular support

The use of balanced crystalloids and albumin for hypovolemia, particularly in case of hypoalbuminemia, is the first line treatment. The presence of relative hypovolemia due to vasodilation deserves vasoconstrictors (noradrenaline or terlipressin), targeting MAP of 65 -75 mmHg. The use of inotropes or vasoconstrictors should be guided by invasive hemodynamic monitoring, including TTE, transpulmonary thermodilution for volumetric and dynamic parameters monitoring, extravascular lung water in cases of capillary damage and interstitial edema, and central venous oxygen saturation (ScvO₂).

Respiratory support

Intubation and mechanical ventilation[162] is mandatory in cases of advanced encephalopathy (West Haven level 3)[163] and respiratory failure. Noninvasive ventilation may be suitable for conscious patients able to protect the airways, moderate to severe hepatic encephalopathy limiting its use[163]. Protective endotracheal intubation and mechanical ventilation are recommended at WH level 3, with careful use of PEEP (less than 12 cm H₂O) to avoid obstruction of venous return able to worsen liver failure (hepatic congestion)[162].

Renal function

Early initiation of continuous extracorporeal renal replacement therapy (CRRT) is advisable for controlling not only fluid balance but also ABE, dys-ionias, and hyperammonemia. The use of citrate in CRRT is feasible if the total/ionised calcium ratio is closely monitored. Terlipressin may be considered before CRRT. Loop diuretics or spironolactone should not delay the initiation of extracorporeal renal replacement[160].

Hemostatic profile

Dynamic monitoring (VETs provided by TEG or ROTEM), alongside laboratory tests, should guide transfusion policy, avoiding the “by default” overuse of FFP. Corrections should be based on viscoelastic tests, (extremely useful in such a setting) and clinical data, not limited to lab tests like PT/INR and aPTT/R. Thrombocytopenia (platelet counts below 50000/uL,) and hypofibrinogenemia( fibrinogen levels below 100 mg/dL or 150 mg/dL according to some guidelines) should be corrected[65,84-87].

Hepatic encephalopathy

Management should consider hyperammonemia as a primary cause of the neurological deterioration associated to liver failure, differentiating between acute and chronic settings. Advanced stages (West Haven level 3) coincide with the loss of airway protection and the need for orotracheal intubation. In the acute non cirrhotic postsurgical patient the most feared complication could be cerebral edema, usually not present in the CLD patient[163].

Interventional or pharmacological interventions

In cases of excessive portal flow, interventions such as splenic artery embolization[164] or the use of octreotide[165], possibly combined with beta-blockers to reduce portal vein pressure, may be considered (multidisciplinary advice of course mandatory). The use of N- Acetyl Cysteine during liver resections or in the case of PHLF is still debated and not supported by evidence[166].

Extracorporeal treatments for PHLF

Treatments like plasmapheresis, artificial replacement (molecular adsorbent circulating system - MARS - Prometheus, single pass albumin dialysis -SPAD- or the sorbent CYTOSORB), and bioartificial replacement (experimental at the moment), even if appealing, lack so far of strong evidence in PHLF[160,161,167]. In the PHLF setting, only MARS was systematically evaluated: however, even if safe and feasible, no evidence to support its routine use was provided by a recent systematic review[167]. As underlined by Göth et al[167], a better characterization of disease-driving and the dimension of molecules to be removed could be crucial to improve the use of artificial liver-support therapy also in PHLF. As a matter of fact, these treatments aim to bridge recovery of native liver function or, in non-reversible cases, serve as a bridge to liver transplantation[167-169].

Liver transplantation

In cases where PHLF finds no solutions and there are no oncological contraindications or no” too sick to be transplanted” issues, liver transplantation may be considered an extreme resolution, as evidenced in case reports and recent reviews[169].

CONCLUSION

The increasing surgical indications for neoplastic liver diseases, regardless of whether the patient has cirrhosis and often involving elderly candidates, mandate advanced and proficient expertise of the anesthesiologist(s) involved in the entire perioperative period of liver surgery. Effective preoperative evaluation (including prehabilitation and nutritional assessments), intraoperative proactive anesthesiologic strategies to foresee and manage intraoperative problems, and ERAS protocols are key to reduce/contain preventable perioperative complications. Appropriately staffed semi-intensive surgical units are crucial immediately following liver surgery. ICU monitoring in the immediate postoperative period remains a topic of intense debate and should not be viewed as a last resort, but rather as an opportunity to tailor the care for selected high-risk patients. However, due to the extended surgical indications with the encouraging results even in older patients, the incidence of PHLF is rising. In these cases, treatment(s) should be tailored to each case according to the degree of severity. The maximal intensive treatment should be offered for the most severe cases. The use of artificial liver supports (if and when indicated), requires further investigation, while the transplant option should not be seen as the last ditch, but rather a viable option in very selected cases.

ACKNOWLEDGEMENTS

We thank Ernestina Mazza, MD (Niguarda Hospital, Milan), Federico Piccioni, MD (Humanitas Hospital, Milan) and Raffaella Reinecke, MD (SanRaffaele Hospital, Milan) for their notes, remarks and fruitful suggestions; and Gloria Innocenti, BS, Niguarda Hospital Librarian, whose help was key in literature research.