Liver surgery is nowadays almost a standalone surgical subspeciality. Its development has been much slower than bowel surgery or orthopaedics probably due to enormous difficulties in understanding the anatomy and function of this spongy and bloody organ of the right upper abdomen and consequent reluctance of practitioners to deal with it. Moreover, for centuries, any health issue related to the liver was considered invariably lethal, making any treatment largely futile. However, liver scholars’ knowledge has gradually increased and surgeons have developed an inquisitive attitude towards the liver and the right upper quadrant organs in general. Progressive acquaintance with the hepatobiliary system has allowed step-by-step resolution of problems that have hindered surgical approach to the liver for ages. We will take inspiration from this progressive problem-solving to present evolution of hepatobiliary surgery from ancient times to the present days.

The word “liver” comes from the Anglo-Saxon word “lifere,” which might have been the origin of the verb “to live.” Therefore, liver is associated with life, which seems obvious considering that the liver has always been associated with blood (as it is undoubtedly the most blood-rich organ of the abdomen) and that blood and life are strictly related. Other etymologists have linked the word “liver” with the Indo-European root “leip-” meaning “fat,” which is close to the meaning and etymology of the Latin word “ficatum” (giving birth to the Italian word “fegato,” the Romanian “ficat,” the French “foie” and the Spanish “higado”), meaning “fatted with figs.” It might have derived from the ancient Roman habit of obtaining a special food from the liver of ducks fatted with dried figs, a sort of old-fashioned foie gras (in ancient Greek “hépar sykoton”). Words “hepatic” and “hepatology” come from the Greek “hépar,” a word connected to the concept of “pleasure” (and possibly with the word “hedoné,” pleasure) because the ancient Greeks believed that the blood-rich liver was the place of passions and feelings. As a matter of fact, in the theory of humours of Hippocrates, the bile (or “yellow bile”) was connected to the liver. It was responsible for a “bilious behaviour,” signifying a passionate and fiery temper [1,2]. Popular sayings of southern Europe call the liver a site of courage (“that woman has liver” = “she is a brave woman”) and bile a bad temper (“he is bilious” = “he is irritable, hysteric”), whereas the British identifies a coward as “chicken liver” and someone insignificant as “chopped liver.”

The concept of liver regrowth after trauma or excision was probably well known to ancient Greeks who conceived the myth of Prometheus, whose punishment after stealing fire from the Olympus and giving it to humanity was to be bound to a rock in the Caucasus and have his liver eaten every day by an eagle (symbol of Zeus). The liver regrew every night, making the punishment perennial. A similar, less known, myth was the one of Tityus, one of the many illegitimate sons of Zeus, who was punished by Zeus himself for trying to rape Leto, mother of Apollo and Artemis. Tityus was bound in Hades, the kingdom of death, where two vultures came daily to eat his liver. As in the myth of Prometheus, the liver regrew every night to perpetuate the torture [3]. In the legend of Prometheus, the liver, seat of life and courage, was chosen to symbolise the hero’s willpower to support mankind. Zeus tried to break Prometheus’ strength with the help of the eagle. However, Prometheus’ stubbornness was eternal, thus frustrating Zeus and his command. For the record, after many generations and countless cycles of liver disruption and regrowth, Prometheus broke free. Some sources claimed that it was Chiron, the centaur, who decided to replace Prometheus and die for him. Other sources mentioned that Heracles managed to kill the eagle and set Prometheus free and unbroken (Fig. 1) [4]. On the contrary, the choice of the liver for Tityus’ punishment was a consequence of his sin to follow his pleasure. There was clearly a different approach, as Prometheus was regarded as a hero and his liver was eaten by a mighty eagle, symbol of strength and power, while the criminal Tityus’ liver was torn by two vultures, scavengers of rubbish and carcasses [3].

Fig 1. Prometheus Bound. Oil on canvas. Begun c. 1611–1612, completed by 1618. Peter Paul Rubens (Flemish [active Italy, Antwerp, and England], 1577–1640), and Frans Snyders (Flemish [active Antwerp], 1579–1657). Philadelphia Museum of Art: Purchased with the W. P. Wilstach Fund, 1950, W1950-3-1. Photo by Alberto Otero Herranz, courtesy of the Prado Museum, 2015. Reproduced by permission.

The Etruscans, who were based in Central Italy, north of Rome, and whose civilisation had its height in the 6th century BC, had some knowledge of the hepatic anatomy and used it to make divinations, as demonstrated by the so-called “Liver of Piacenza” (Fig. 2). It is an Etruscan bronze artifact found in 1877 near Piacenza (Italy) and part of the exhibition at the Palazzo Farnese Museum in the same city. Dated back to the 2nd century BC, it was considered a model of a sheep’s liver used as a manual of “hepatoscopy,” or the art of exploring the liver of sacrificed animals to obtain previsions for the future. The bronze liver is subdivided into 16 marginal sections (paralleling the celestial sphere as believed by the Etruscan) and 24 internal sections. Each of these 40 sections has the name of one of the Etruscan deities. Furthermore, in the convex part of the “Liver,” there is a clear representation of the gallbladder and the caudate lobe. Similar objects were found in Iraq, where ancient people of Mesopotamia, the Sumerians, believed that the liver represented the origin of the Universe. An interesting clay model of a sheep liver from Babylonia is on display at the British Museum in London. It is divided into approximately 50 fields. It is supposed to be an aid to the teaching of divination [5].

Fig 2. Etruscan “Liver of Piacenza,” Museo Farnese, Piacenza (Italy).

The first mention of the liver as a mean to predict the future dates back to the Assyro-Babylonian era (around 2000–3000 BC), as testified by the Bible: “For the king of Babylon will stand at the fork where the two roads divide to seek an omen. He will shake the arrows, consult the household gods, and inspect the liver” (Ezekiel 21:21) [6]. In the Bible itself, there are 12 to 19 citations of the liver in version one [7]. Most of these are related to sacrifices. It mentioned that the fat surrounding the liver and kidneys of the animal was removed to be offered to God and burn it on the altar (Exodus 29:13,22–Leviticus 3:4,10,15–Leviticus 4:9–Leviticus 7:4–Leviticus 8:16, 25–Leviticus 9:19) [7]. The liver (and the smoke exhaling when it is burned) was also considered a cure for people affected by demoniac possession (Tobit 6:5-8,17–Tobit 8:2) [7]. The Old Testament also considers the liver as the seat of life, probably because of the knowledge that a direct trauma–in humans as well as animals–would cause death (Proverbs 7:23) [7]. As with many other items, mostly related to religion, divination through the liver was spread from Babylonia to the rest of the world, along with the first descriptions of animal liver anatomy. There were separate names for different parts of the liver [5]. Of note, the common bile duct was named as “the outlet” [5], signifying that ancient Babylonian priests had a rough knowledge that bile was produced by the liver and came out through the common bile duct.

The most popular theory about the origin of Etruscans, based on writings of Herodotus, stated that people who settled in Etruria came from the Near East with a first wave to colonize the place and a second wave of immigration to offer specialised manpower to the settlers [8]. Therefore, it was likely that the practice of liver divination reached western Europe through this route. Most of the ancient Roman haruspices were Etruscans in origin. Therefore, we were led to believe that the practice of “hepatoscopy” started in Babylonia and continued throughout the Roman Empire.

Hippocrates recommended incision and drainage of right upper quadrant abscesses. However, it was unclear whether he was able to distinguish between liver abscesses and gallbladder empyemas [9]. Claudius Galen, a Greek-Roman physician of Pergamon who looked after the health of Roman emperors Marcus Aurelius, Lucius Verus, Commodus, Septimius Severus, and Caracalla, obtained a huge knowledge of mammals’ anatomy by dissecting animals, but not humans as this practice was not allowed in the Roman Empire. He speculated that the liver had a role in processing food and producing blood [9]. However, in his writings, Galen mentioned anatomical concepts of Herophilus of Alexandria, considered the father of anatomical studies as he took advantage of the tolerance of the Greek world towards human body dissection, which was lately prohibited or at least strongly discouraged both in Christian and Islamic civilizations. Galen described lobes of human and animal liver and gave a detailed description of intra and extrahepatic vessels while proposing a first comprehensive physiologic theory [5]. Galen followed the Hippocratic theory of the “four humors.” The human body contained four humors: blood, phlegm, yellow bile and black bile, corresponding to the fundamental elements of the universe: air, water, earth, and fire. Good heath was due to a perfect balance of the four humors, while their imbalance caused disease. The liver was where blood was produced starting from food that reached the liver through the portal system as chyle. Yellow bile was removed from blood by the bile system, while blood reached the heart through the inferior vena cava and black bile reached the spleen through splenic veins. Water, or phlegm, was removed from the blood by the kidneys [5].

Long considered one of the seats of the soul and the source of blood, the liver has historically been considered “off-limits” by physicians who have feared its bloody and friable parenchyma. However, there are reports of occasional debridement of liver tissues after trauma by Celsus in ancient Rome and Paulus from Aegina in the Byzantine Empire [9]. This practice of liver debridement after war wounds continued for centuries. However, the study of human liver anatomy was long hindered by the moratorium on human dissection issued by the Catholic Church and by the dogmatic influence of Galen’s and Hippocrates’ writings. Vesalius based his De Humani Corporis Fabrica Libri Septem, published in 1543, on anatomical dissection of human specimens, but its description of human liver might have been still influenced by Galen’s words [5]. Looking at Vesalius’ illustrations as reported by McClusky et al. [5], we wondered if his anatomical studies were done on both animals and humans. In fact, while the depiction of the liver in situ definitely resembles the human liver, the isolated liver and portal system recalls the anatomy of pigs and dogs with multisegmented livers. Similarly, William Harvey, who first described the human circulation in his Exercitatio Anatomica de Motu Cordis et Sanguinis in Animalibus published in 1628, was heavy influenced by Galen’s description of liver and missed the concept of portal circulation, stating once again that blood, and hence vessels, originated from the liver [5].

A good depiction of human liver anatomy was provided only in 1654 by Francis Glisson from Cambridge (Fig. 3) [10]. Like most of medical books written in that period, Glisson’s Anatomia Hepatis was written in Latin by George Ent, who translated Glisson’s English into that universally used noble language [11]. Ent was himself a scientist and an anatomist. He was a close friend of William Harvey, whose theory of circulation he defended strenuously. Francis Glisson was born in 1597 in Bristol. He graduated from Cambridge University, where he became a Regius Professor of Physics. Subsequently, he lived long years in Covent Garden in London, where he died in 1677. He too was a colleague and good friend of William Harvey. Despite his groundbreaking description of human circulation, it did not grasp the complexity of liver blood flow. On the contrary, in his Anatomia Hepatis, Glisson was the first to depict the internal anatomy of the liver through injection and corrosion casting methods. He also provided the first description of the bilioduodenal sphincter, 230 years before Ruggero Oddi. His name has been forever linked to the capsule of the liver and internal distribution of this capsule accompanying and encasing the portal triad when it enters the liver parenchyma at the porta hepatis and branches thereafter [12], despite this had been already described by Johannes Walaeus of Leiden [13]. Previously, in the 16th century, Leonardo Da Vinci had already grasped the complexity of the anatomy and vascular supply of the liver in some of his drawings which are now part of the Windsor Castle Royal Collection [14].

Fig 3. Francis Glisson by William Faithorne line engraving, published 1672. National Portrait Gallery, London. Reproduced by permission.

Continuous wars during the 17th and 18th centuries, as well as the perennial civil turmoil, gave the possibility to military and civil doctors to get acquainted with liver traumas. McPherson described the excision of small liver bits from an abdominal wound of a Hindu in 1846 and mentioned a previous case going back to 1688 of liver debridement from a sword wound [9]. A dr Massie referred the excision of a significant part of the right lobe of the liver of a 7-year-old boy who shot himself while cleaning a gun [15]. Other reports of post-trauma liver debridement have been published in those years [5], following the description of liver trauma by Ambroise Paré [5].

The advent of antisepsis, anaesthesia, and radiology in the last decades of 19th century allowed a more systematic approach to abdominal organs, including the liver. The introduction of ether anaesthesia in 1846 by William Morton, the description of chloroform anaesthesia in 1847 by James Simpson, and that of antisepsis with carbolic acid in 1867 by Joseph Lister paved the way to major surgeries [16]. Blood transfusion had already been introduced into clinical practice in 1818 by Blundell [17]. Jean François Reybard performed the first colonic resection in Paris in 1823. Jacques Lisfranc again from Paris published his first three rectal resections in 1833. Theodore Billroth in Wien did his first gastrectomy in 1881 and Alessandro Codivilla performed the first pancreatoduodenectomy in Imola (Italy) in 1889. In 1867, John Bobbs from Indiana performed the first cholecystolithotomy. A few years later, Theodore Kocher in Bern fashioned the first cholecystostomy. Langenbuch [18] performed the first successful cholecystectomy in 1882 in Berlin (Fig. 4). A few years later, he performed the first successful liver resection on a 30-year-old lady with a mass of the left lobe. Unfortunately, the operation was complicated by postoperative bleeding from a vessel of the hepatic hilus that required a reoperation, but the patient eventually recovered [5,9]. Tanabe [19] reported that the first liver resection was attempted in 1886 by A. Lius to remove a large adenoma of the left lobe, but the patient died postoperatively of haemorrhage.

Fig 4. The first page of Carl Langenbuch’s report of his first cholecystectomy. Adapted from the article of Langenbuch. Berliner klinische Wohenschrift 1882;48:725-727 [18].

The first liver resection in the United States was performed in 1891 at the Jefferson Medical College in Philadephia by William Keen (Fig. 5), who subsequently reported on his first three liver resections. The second and third ones were performed in 1897 and 1899, respectively, all with excellent results. In his analysis of the available literature at that time, liver resections had a 17% mortality rate [5].

Fig 5. William Keen. Smithsonian Institution Archives, Accession 90-105, Science Service Records, Image No. SIA2008-4586. Reproduced by permission.

These advancements would have never been possible without studies on liver regeneration. In the late 19th century, Gluck [20] and Ponflick [21], working separately, demonstrated that liver function could be maintained after resection of the wide part of the parenchyma and that liver could regenerate after massive parenchymal resection [22].

In the meantime, knowledge of liver anatomy was somehow stagnant since writings by Francis Glisson of 1654. In 1764, Albrecht von Haller proposed a new subdivision of liver parenchyma in four lobes–right, left, quadrate, and caudate [5]. This paved the way for subsequent studies by Rex and Cantlie as it represented somehow a step forward from the traditional division of liver parenchyma in two lobes based on the falciform ligament. By means of cast-corrosion studies, in 1888, the Czech anatomist Hugo Rex demonstrated that portal vein branched at the porta hepatis into two main veins supplying two separate parts of the liver which did not resemble the two traditional lobes [23]. He gave his name to the blind ending of the left portal branch of the adult liver where it was divided into segmental branches for segments 3 and 4, whereas in the foetus, the Rex recess marked the insertion of the umbilical vein into the intrahepatic portal circulation. In 1898, a Scottish physician James Cantlie (Fig. 6), while studying liver atrophy on autopsy specimens, noticed that the line of segmental liver atrophy corresponded almost always to a plane passing from the gallbladder fossa to the right margin of the inferior vena cava. He concluded that the line that was named after him (Cantlie line) since then was the real midline of the liver [24].

Fig 6. James Cantlie by Walter Stoneman bromide print, 1918. National Portrait Gallery, London. Reproduced by permission.

Surgery of hepatic hydatid disease

Despite initial experiences of formal hepatic resections, the most frequent form of liver surgery was represented by surgical treatment of hepatic hydatidosis in the early ’70s of the 20th century. It started in 1880 when a British gynaecologist from Birmingham, Robert Lawson Tait (Fig. 7), drained a giant hydatid cyst of the liver and sutured its edges to the abdominal wall [25]. In the first half of the 20th century, surgery for Echinococcus of the liver was almost always done for massive and symptomatic liver cysts, often communicating with the biliary tree. A usual operation entailed an excision of the dome of the cyst followed by sterilisation of the hydatid cavity with iodine. Unfortunately, the passage of iodine (as well as other scolicidal solutions) into the biliary tree frequently caused secondary “caustic” sclerosing cholangitis with progressive liver failure [26]. For this reason, the practice of “sterilising” the cyst before surgery to prevent dissemination of the parasite has been strongly discouraged, favouring the safer and more radical total pericystectomy in particular after the 1979 Lecture at the Congress of the Italian Society of Surgery by Tagliacozzo et al. [27], who demonstrated that total excision of the hydatid cyst with its pericyst, a method that he developed when leading the Department of Surgery in Cagliari, could reduce complications and hospital stay.

Fig 7. Robert Lawson Tait by Elliott & Fry half-plate glass negative. National Portrait Gallery, London. Reproduced by permission.

Hepatic resections

Further to anatomical studies and early experiences of the last part of the 19th century, early days of the 20th century saw a slow and prudent development of resective liver surgery. In 1909, Hans Von Haberer, who was a pioneer in vascular and upper gastrointestinal surgery in Innsbruck, Graz, Dusseldorf and Köln, proposed to ligate the left hepatic artery before removing the left hepatic lobe [9]. In 1911, Wendel [28] gave credit to Cantlie’s new anatomical description of liver anatomy and performed a right hemihepatectomy following the Cantlie’s line after ligating right vasculobiliary elements at the hepatic hilum. Surgeons gradually started to appreciate advantages of performing resections following an almost avascular plane. Thus, major liver resections (right and left hepatectomies) were gaining wide diffusion. Vittorio Pettinari from Padua published his first experience with liver resections in 1939. However, it went almost unnoticed [29] until his lecture at the 16th Congress of the International Society of Surgery in 1955 [30]. After WWII, in 1945, Wangesteen performed a right hepatectomy after ligating the right portal triad [31]. In 1946, Donovan and Santulli [32] from New York and Hershey [33] from Wheeling, USA ligated the left portal triad branch before performing left hepatectomies. On October 16, 1951, Lortat-Jacob and Robert [34] in Paris performed an extended right hepatectomy following Wangesteen’s technique through a thoracophrenolaparotomy after preliminary vascular control at the porta hepatis, thus performing “reglée” hepatectomy, meaning a major liver resection guided by vascular ligation induced partial parenchymal ischaemia. This raised worldwide interest in resective liver surgery. Thus, a young Swedish surgeon from Lund named Bengmark [35] in 1961 convinced his director, Professor Ragnar Romanus, to help him get a financial support by Volvo to spend a few months in Padua and learn liver surgery from Prof Pettinari.

Initially, hepatic surgery was becoming a standalone subspeciality very slowly. In the early ’70s Stig Bengmark in Lund, Leslie Blumgart in Glasgow (and then London, at the Hammersmith), and Henri Bismuth in Paris–the so-called “B-team.” Bach, Beethoven and Brahms, of hepatobiliary surgery by Michael Trede–were establishing their own subspecialistic surgery units [35] attended by an ever-growing number of young surgeons who were eager to learn HB surgery.

Unfortunately, mortality of major liver surgeries was exceedingly high in the ’60s. Gradually, skills and knowledge improved. In 1983, Thompson et al. [36] from Los Angeles published his series of 138 liver resections with “only” 10% mortality. In 2000, Belghiti et al. [37] from Paris published his series of 747 patients with 1% mortality in non-cirrhotic patients.

Although hepatic surgery was prudently entering its age of maturity, its progress was still hindered by three main issues, namely, the identification of intrahepatic anatomy, the control of bleeding, and the risk of postoperative liver failure.

Hepatic anatomy

Modern hepatic resective surgery has its roots in the publication of Claude Couinaud’s revolutionary studies on liver anatomy in 1954 [38] and 1957 [39], introducing the concept of “functional independence” of sectors with segments separated by fissures and supplied by their own intrahepatic portal pedicles, with each draining into its own hepatic venous branch. It took 4 years and 140 corrosion-casting of “fresh” livers for Couinaud to establish his view of hepatic anatomy based on two hemilivers, four sectors, and 8 segments. Obviously, Couinaud’s liver anatomy is deeply embedded in the above-mentioned principles of Rex and Cantlie, whose trail Couinaud followed and expanded, along with results of dissections by McIndoe and Counseller [40] who, in 1927, confirmed that the human liver was divided in two halves on the basis of branching of the portal vein at the porta hepatis, not by the falciform ligament.

In the second part of his book, Couinaud classified resective procedures on the liver as “atipiques” and “réglées,” suggesting great prudence with atypic (or “wedge”) resections, where bleeding control was done with gross and possibly ineffective methods. “Réglées” (or “anatomical”) resections, on the contrary, should be performed following precise intersegmental planes with bleeding control assured by preliminary ligation and division of inflowing vascular pedicles. Couinaud’s studies represented a first turning point in the history of liver surgery as they moved the focus from external morphology of the liver to a proper functional anatomical map of the liver based on the internal distribution of the vasculobiliary tree (portal triad). The branching of the portal triad divides the liver into eight segments, which Couinaud numbered in a clockwise fashion resembling numbers of the central “arrondissements” of Paris.

Couinaud’s book “Le Foie” was a huge comprehensive 530-page monography. It was extremely difficult to read and understand. However, it had its merit as it paved the way for Henri Bismuth’s “anatomical surgery” of the liver [41] so that surgical resections could be based on the segmental anatomy of the liver and intrahepatic branches of the portal pedicle. This new concept decreased surgical risk of liver resections by reducing bleeding, bile leak, and parenchymal ischaemia. Although it was still a radical surgery, it allowed a parenchymal-sparing, thus expanding surgical indications to patients with hepatocarcinoma on cirrhotic liver and those with multiple metastatic lesions. Actually, other authors had preceded Couinaud’s work. However, they did not have the same fortune. In 1951, Hjortsjö [42] from Lund, Sweden, published results of cast corrosion and cholangiogram on 10 specimens, demonstrating that hepatic arteries and bile ducts had a segmental distribution. These results were confirmed by Elias and Petty [43] from Chicago in 1952 and by Healey et al. [44,45] from Houston in 1953. Segment 1 (S1) has an history of its own. Couinaud named the caudate lobe as “segment 1,” but Kumon in 1985 divided S1 into three portions: the Spigelian lobe, the paracaval lobe (or portion), and the caudate process. Each of them has its own blood supply with venous and biliary drainage [46]. Knowing the anatomy of S1 is crucial in the treatment of hilar cholangiocarcinoma.

The lack of reliable diagnostic imaging techniques hindered the progress of liver surgery for years.

The introduction of ultrasound scanning (USS) into clinical practice in the late ’70s–after a long experimental period starting in the ’40s when the Naval Medical Research Institute tried to apply the sonar technology to human body–allowed early detection of small liver lesions. It also helped the comprehension of intrahepatic anatomy as described by Couinaud. Up until then, palpation was the only method to identify liver masses. Computer assisted tomography (CAT) scan was introduced in 1972 by Sir Godfrey Hounsfield who worked at electric and music industry (EMI) laboratories. He had been appointed to develop radars and guided weapons. He shared the invention of the scanner (and the subsequent 1979 Nobel’s Prize) with physicist Dr. Allan Cormack. EMI, who was the record label of the Beatles and one of the main record producers, used revenues of the Beatles’ records to fund Hounsfield’s researches on “EMI scan,” later renamed as CAT scan. The first CAT scan was performed on a man with a brain tumor. It soon became an unreplaceable diagnostic instrument to study the liver and biliary tract.

Magnetic resonance imaging (MRI) scan was introduced in 1979 following the work by Paul Lauterbur from Stonybrook, USA and Peter Mansfield from Nottingham, UK. They too shared the Nobel Prize in 2003 for the invention of MRI. Positron emission tomography scan was developed by Gordon Brownell at the Massachusetts Institute of Technology in 1953. However, it was not introduced into clinical practice until in the ’70s when fluorodeoxyglucose was discovered.

Intraoperative USS introduced in 1983 by Makuuchi et al. [47] from Tokyo “rendered the liver transparent” [48] and allowed regulated segmental resections, defining what Bismuth et al. [48] described as “hépatectomie à la carte.” Initially, the technique of intraoperative USS did not get much popularity among some traditional surgeons who were more worried of learning a new skill than enthusiastic at the idea of finally mastering completely the anatomy of the liver during a resective procedure. On the contrary, the group of Bismuth et al. [49] in France and the team of Gozzetti et al. [50] in Bologna, Italy expanded and perfected the technique of intraoperative USS of the liver.

Bleeding control

The risk of bleeding hindered the progress of liver surgery for many years. Initial attempts with cautery, gross stitches or clamps were not particularly effective [9]. In 1908, John Hogarth Pringle of Glasgow (Fig. 8) [51] published what is now regarded as one of the milestones articles in liver surgery describing his manoeuvre of manual control over the hepatic pedicle. In his historical paper, Pringle described eight cases of patients admitted to the Glasgow Royal Infirmary with ruptured liver in 11 years. Three of these patients died immediately after admission, one refused operation and died after a couple of days, and four underwent laparotomy. Two of them died during the surgery while two died shortly afterwards. In three cases, compression of the hepatic pedicle allowed temporary reduction of bleeding from the hepatic lesion, but these patients died anyway. In one case, there was an unstoppable bleeding from the dome of the liver probably from the hepatic veins. Of interest, Pringle concluded that the best way to stop bleeding from the liver was to pass mass sutures close–but not too close–to the edge of the parenchymal wound [52].

Fig 8. James Hogarth Pringle. Adapted from the article of Obituary. Br Med J 1941;1:734 [51] with original copyright holder’s permission.

The so-called Pringle manoeuvre—i.e., hepatic pedicle clamping—was considered as an almost mandatory step in resective liver surgery [29]. As Pringle himself could see in his initial experience, the manoeuvre named after him was not able to control every haemorrhage from the liver but only those of portal origin, while the most troublesome and hard-to-control bleedings are from the hepatic veins, probably made even worse by the high central venous pressure maintained by anaesthetists in order to sustain a good cardiac output. Therefore, vascular control became crucial. Several different techniques were envisaged. The Pringle manoeuvre helped reduce bleeding originating from the small portal or arterial branches in the section surface of the liver. However, it was initially feared that a prolonged ischaemia would cause irreversible damage to the liver parenchyma. Teams of Gennaro Nuzzo in Rome and Renzo Capussotti in Turin demonstrated that a number of liver resections could still be performed without ischaemia. They also demonstrated that in selected cases, a prolonged clamping of the hepatic pedicle up to 120 minutes significantly reduced the amount of bleeding and the need for transfusions without being associated with permanent parenchymal injury or increased rate of complications [53,54]. Intermittent clamping gave equivocal benefits. In Italy, it was favoured by the group from Turin [52], although others preferred a continuous pedicle clamping [54].

Clavien et al. [55] of Zurich and Surinder Yadav of the Duke University of Durham, North Carolina, USA emphasised the risk of repeated reperfusion injury in intermittent clamping and the risk of bleeding during periods of reperfusion. They proposed the technique of preconditioning entailing an initial period of 10 minutes ischaemia and 10–15 minutes reperfusion before starting parenchymal transection under ischaemia for 30 minutes.

A European Survey by van der Bilt et al. [56] published in 2007 demonstrated that only 19% of respondents used routine pedicle clamping and 71% only if needed. Intermittent ischaemia was chosen by 65% of surgeons. That survey showed that routine clamping was more frequently applied by senior surgeons working in high volume centres.

Vascular control of the liver during parenchymal transection was made easier by the Pringle’s manoeuvre as it allowed a significant reduction of portal bleeding. However, control of bleeding from hepatic veins was much more challenging. Stucke in 1961 and Storm and Longmire in 1971 proposed the use of a clamp to compress the liver and achieve haemostasis [19]. However, it did not obtain wide acceptance. The use of clamps is now considered obsolete.

Ton That Tung’s technique of parenchymal finger-fracture with totally intraparenchymal or subglissonian vascular control was revamped by Lin et al. [57] in the late ’50s. This technique was probably misunderstood. Many surgeons despised it, thinking that it was raw and non-anatomical. Ton That Tung’s resections were always along anatomical planes. The technique was not particularly straightforward. It needed a great experience mostly to control the inevitable bleeding. It was replaced by Bismuth’s Kelly clamp crushing technique [41] and subsequently by other instruments using ultrasounds [58] or high-pressure water [59]. Although not a resective device, the argon-beam coagulator is also useful for controlling bleedings from the raw surface of the liver originating from the portal system during and after transections. Other methods such as Lin’s clamp and Habib’s sealer [60] did not reach worldwide acceptance. However, intraparenchymal control of both portal and hepatic vessels was still associated with significant bleeding. Lortat-Jacob et al. [61] have proposed double control of the glissonian pedicle and the correspondent hepatic vein before the parenchymal transection. Unfortunately, controlling hepatic veins via an extraparenchymal approach may not be straightforward. It carries the risk of injuring the inferior vena cava.

In 1959, Quattlebaum and Quattlebaum [62] tried to standardize the technique of major liver resections, proposing the following five steps: “(1) Adequate exposure through a large thoraco-abdominal incision. (2) Complete mobilization of the liver by dividing all its peritoneal attachments. (3) Dissection of the porta hepatis with individual ligation and division of the structures entering the involved lobe. (4) Division of the liver substance with a blunt instrument, rather than by sharp incision. (5) Ligation of smaller vessels with fine silk or cotton; control of oozing from the liver surface by Gelfoam packs and omentum; avoiding heavy mattress sutures through masses of liver substance.”

Child et al. [63] demonstrated that monkeys could survive after portal vein ligation, while dogs, cats, and rabbits could not. Based on this finding, they ligated the portal vein in two patients with advanced cancer and found that their patients survived the operation. Mays [64] demonstrated that the hepatic artery could be safely ligated and indeed suggested hepatic artery ligation in trauma, hepatocellular carcinoma, haemobilia, and liver angiomas.

Bismuth [41] originally proposed an extrahepatic suprahilar control of the portal pedicle leading to the segment or sector to be excised, while the hepatic vein was controlled within the parenchyma. A sectorial portal pedicle could be reached via suprahilar approach after dissection of the hilar plate. However, segmental pedicles could be identified only via a transparenchymal approach, preferably under ultrasound control, as suggested by Makuuchi et al. [65]. The segmental portal branch could be injected with methylene blue or selectively occluded with a balloon to demarcate the segment to be excised [66,67].

In case of sizeable lesions close to the hepato-caval confluence, the risk of bleeding from the hepatic veins is quite high and a total vascular exclusion (TVE) of the liver is still needed, associating the Pringle manoeuvre to the clamping of the supra and infrahepatic inferior vena cava and eventually of the infradiaphragmatic aorta, with or without hypothermic perfusion as proposed by Bernhard et al. [68], Heaney et al. [69], and Fortner et al. [70]. Huguet et al. [71] at Hopital St. Antoine in Paris perfected the technique of TVE.

Unfortunately, still a number of patients, particularly those with hypovolaemia and hypotension, could not tolerate the massive reduction of preload due to TVE. Therefore, Elias et al. [72] and Cherqui et al. [73] from Paris reconsidered Lortat-Jacob’s selective clamping of hepatic veins outside the liver parenchyma for maintaining the caval flow. Although very attractive, this technique’s critical point, the extrahepatic preparation of the right hepatic vein, is a tricky and dangerous manoeuvre. It is possible only with division of the posterior ligament of the inferior vena cava described by Makuuchi et al. [74] of the University of Shinshu and then Tokyo.

Several techniques have been proposed to improve tolerance of the liver to ischaemia. Ischaemic preconditioning has already been mentioned. Veno-venous bypass can occasionally be used to reduce haemodynamic instability and splanchnic congestion after TVE. Schrock et al. [75] described an atrial-caval shunt consisting of a long tube inserted through the right atrium and pushed down into the infrahepatic vena cava while the suprahepatic and infrahepatic vena cava were closed over the tube with tapes. Although progress of liver transplantation techniques has brought external veno-venous bypass into the common practice of resective liver surgery [76,77], it is still only used in very selected cases. Liver hypothermic perfusion was proposed by Fortner et al. [70] to reduce blood loss and improve the tolerance of liver to ischaemia.

Completely ex-situ bench liver surgery described by Pichlmayr et al. [78] and partial disconnection of the liver with division of the suprahepatic inferior vena cava proposed by Delrivière and Hannoun [79] can be useful in very selected cases. They require a great experience in liver surgery. In Italy, the group of Forni and Meriggi [80] in Pavia continued the French experience of bench liver surgery using techniques of liver transplantation such as hypothermic liver perfusion and veno-venous pump from the IVC and portal vein to the superior vena cava.

The gradual development of complex liver surgery requested strong collaboration with anaesthesiologists. When massive haemorrhage was predicted, anaesthetists were asked to keep the patient hypervolaemic with a very high central venous pressure to compensate for blood loss. Unfortunately, this increased the risk of bleeding from the hepatic veins, while bleeding from the portal system could be controlled by pedicle clamping. Consequently, reducing the central venous pressure before parenchymal transection became a common practice [81], thus emphasizing the strong need of a tight cooperation between surgeon and anaesthetist during a liver surgery.

Postoperative liver function

The risk of postoperative liver failure has always been an obstacle to the development of biliary surgery in particular in the cohort of patients with hepatocellular carcinoma on chronic liver disease, who anyway represent a large percentage of those needing liver resection. The improvement of preoperative evaluation of liver function through dynamic tests such as the bromsulphalein test and the indocyanine green test and progresses in pre- and post-operative optimisation of the nutritional status of the patient also through the administration of short-branched chain amino-acids were crucial in reducing surgical mortality due to hepatic global dysfunction. However, the risk of hepatic failure was still in direct relation with the amount of healthy parenchyma removed. Predicting the volume of the future liver remnant through an accurate CAT study with parenchymal volumetric assessment allowed a good estimate of postoperative function of the liver. In the case where this investigation gave a future liver remnant volume less than a safe 30%–40%, embolization of the portal branch or branches supplying the part of parenchyma to be excised proposed by Makuuchi et al. [82,83] in 1984 and 1990 and Kinoshita et al. [84] in 1986 allowed atrophy of the embolised volume and compensatory hypertrophy of future liver remnant. This technique was developed in Europe by the group of Azoulay et al. [85]. It allows major resections of healthy parenchyma, as it happens in patients with hilar cholangiocarcinoma or multiple metastases.

Associating liver partitioning and portal vein ligation for staged hepatectomy (ALPPS) was developed in Germany. It was initially reported with a three-case series by Baumgart et al. [86] from Mainz as a poster at the 9th E-AHPBA Congress in Cape Town. Other case report followed [87]. This technique was subsequently applied worldwide [88].

The problem of postoperative liver failure was particularly worrying in patients with liver cirrhosis who were unable to tolerate major resections. This stimulated several methodological improvements, including non-surgical treatment of liver lesions. In 1986, Livraghi et al. [89] introduced US-guided ethanol injection of hepatocellular carcinomas (HCC) on cirrhotic livers. A multicentre trial published in 1995 demonstrated that this technique was equivalent to surgery in terms of survival but was much less morbid and better tolerated [90]. This revolutionary technique prompted the performance of screening USS in cirrhotic patients to identify small HCC nodules that could be treated without surgery. In a small cohort of patients where ethanolization was not indicated, other strategies could be considered, including chemoembolization, originally introduced as a presurgical step by the team of Oi et al. [91] in Osaka in 1983 and then brought to Europe by the group of Rossi et al. [92] in Milan to treat non-operable HCC in 1989.

However, patients with borderline operable HCC in cirrhotic liver may still benefit a surgical approach with preoperative portal embolization, ALPPS or, even better, liver transplantation. In fact, in the ’80s and ’90s, cirrhotic patients with HCC got away much better with transplantation than with surgery as with a single operation, both cancer and the basic condition could be treated. History of liver transplantation is an exciting chapter of the history of surgery. However, it goes beyond the scope of this paper. Similarly, the history of surgery of portal hypertension is so wide that it may deserve a paper of its own.

The history of hepatic surgery is a tale of a progressive resolution of problems encountered day by day in the effort of providing hope of cure to patients with liver masses or trauma. Knowledge of anatomy and pathophysiology of the liver has been the lighthouse guiding surgeons in this difficult course. Every young surgeon approaching liver surgery invariably faces the same problems. They must tackle the same difficulties: (1) perfect anatomical knowledge, (2) thorough understanding of liver physiology, (3) control of parenchymal bleeding, (4) role of liver ischaemia, and (5) residual liver function after major resection. The surgical armamentarium is full of possibilities ranging from intraoperative ultrasound to hepatic inflow occlusion and from TVE to preoperative selective portal embolization and liver partitioning. The surgeon must be aware of what knowledge and technology available to him or her nowadays and what stands behind those opportunities in terms of struggle and research to be able to offer—paraphrasing an old adage—“the right technique to the right patient at the right time.”

None.

No potential conflict of interest relevant to this article was reported.

Conceptualization: GDT, FG, RM, GN. Data curation: SA, RC, AD, JD, DDN, FD, AG. Methodology: GDT, RC, GN. Visualization: GDT, SA, FD, RC. Writing - original draft: GDT. Writing - review & editing: All authors.