The anatomists described the various vascular systems of the liver, through meticulous dissections, and as of the seventeenth century, by vascular injection mouldings with corrosions. The classical teaching of the anatomy of the liver considered two lobes separated by the round ligament. The development of the biliary surgery has allowed a more precise approach to the various elements of the hepatic pedicle. Excepted for some specialists, the knowledge of the intrinsic anatomy of the parenchyma with the vascular systems, arterial, venous portal and hepatic veins was very limited. Surgical and digital technological progress profited from the fundamental work completed in France in the 1950s year.

Anteriorly the liver, which function to receive and distribute blood, was regarded as the central body, at the origin of life. Ambroise Paré [30] had a precise knowledge of the hepatic regeneration and considered that a part or a complete lobe could be resected without provoke the death. F Glisson, in Cambridge [10] described the segmental anatomy of the liver, but this work remained forgotten for nearly 300 years (Fig. 1). In 1863, within a competitive examination of anatomy, Sabatier, in Montpellier, made a children’s liver corrosion (Fig. 2).

Fig. 1
figure 1

Initial documents in 1654 of F Glisson (Anatomia hepati. London) (a), and drawings after dissection of the organization of the very precise first vascular structures of liver’s segmentation (b)

Full size image
Fig. 2
figure 2

Liver of fetuses with corrosion by Sabatier during a anatomy competition in 1863 (Conservatory Anatomy Montpellier)

Full size image

Starting from 1880, the practice of the surgery of abscesses opened the way to the resections with the first hepatectomy by Langenbuch [19] for an “hepatoptosis”. The first resection of a liver tumor was made by Luis [23] who excised a 370-gr. portion of the left lobe for an adenoma in 1886. The first left lateral supposed segmentectomy was performed by Keen [18]. Rex [35] investigated the distribution of interhepatic vessels and established hepatic segmentation, in mammals and man by the injection corrosion method. Cantlie in Hong-Kong [2] with vascular casts of the liver demonstrated that the main lobar scissura is oblique and extended from right to left and from the visceral to the parietal surface at an angle near 70°. He concluded that the main division between the right and left lobe extended from the gallbladder to the right side of the inferior vena cava postero superiorly. Therefore, the right and left livers individualized by the main portal scissura are independent as regards their portal and arterial vascularization and biliary drainage.

In 1889 about thirty hepatectomies were indexed [20]. Wendel [44] resected the first right lobe, and in 1945 approximately 220 hepatectomies were found in the literature. Since the experiments of Ponfick [32] we know that 65–70% of the hepatic mass could be removed without functional damage. Thirty years later, Cantlie’s work [7] was verified by Mclndoe, [26] and Hjortsjö [16] on 42 human livers, and reemphasized the concept of the bilaterality of the liver. Tôn Thât Tùng, in Hanoi, [41] described the role of the venous drainage of the liver as a functional lobar anatomy, Hardy [13] studied the distribution of interhepatic vessels, all aiming to reduce the risks associated with hepatic resection. Several methods of hepatic segmentation were proposed and Hjortsjö [16] concluded the division of the liver into 6 segments. These studies demonstrated that the left lobe was divided into a medial and lateral segment by the line of the falciform ligament. Goldsmith [11] described the anatomical planes through the liver parenchyma for performing a right hepatectomy, a left hepatectomy and a left lateral segmentectomy. Healey [14] with casts of the biliary tree, showed that the right lobe was divided into an anterior and a posterior segment. He proposed 4 segments with taking into account the caudate lobe and later revised their study and advocated the division of each segment into two-superior and inferior-resulting in 8 segments. He concluded that the left lobe was divided into a medial and lateral segment by the falciform ligament line [12].

Duckett [7] showed that vascularization by the chief branches of the portal venous and hepatic artery were not essential and only the integrity of the venous system and hepatic vein was crucial. The suppression of the portal contribution in a hepatic territory involved its atrophy with lubricating degeneration, whereas the suppression of the hepatic vein drainage resulted in the infarction of the corresponding zone. The venous system has an essential value: the portal vein and hepatic veins constitute, with the hepatic capillary network, a unique vascular formation. The problems arising from the hepatectomies depended on the various techniques of hemostasis of the slices of section of the parenchyma, slightly on the left of the sickle-shaped ligament. This difficulty was overcome after the knowledge of the segmentation, by the ligature of a large hepatic vein ensuring the vascularization of a zone of parenchyma much more important than that removed by the resection. Caprio [3] carried out the first left hepatectomy regulated with prime ligature of the pedicles. There after Tôn Thât Tùng [42], Hjortsjö [16] codified a complex hepatic segmentation. Sénèque [39] recalled in 1950 that one removed all the parenchyma whose vascularization was interrupted. Mikami [27] advocated hepatic segmentation with 12 divisions and demonstrated the possibility of systematic hepatic resection as a result of his investigation of the distribution of intrahepatic vessels. When considering hepatic resection for hepatocarcinoma, hepatic segmentation should be determined according to the mode of tumor progression. From their studies, each vein could be named according to Couinaud's classification [5], but some could not be classified in the system in relation to the tumor so as to facilitate safer and more radical hepatic resection and conserve the maximum remaining hepatic function. Furthermore, when he selected which segment to resect, the form of the flow and the branching should be evaluated using hepatic venography in every case pre-operatively, then, the area to be resected should be defined according to the tumor's location.

In France at quite the same period (1951–1954) two anatomists surgeons EP. Rapp [34] in Montpellier with 134 livers’s corrosions and Cl. Couinaud [5] in Paris with 144 livers’s corrosions worked using the same method. After the operation by Lortat-Jacob [22], Couinaud [5], followed the same approach defining the sectors and the hepatic segments separated by spaces in which the hepatic venous return is carried out. A coherent system considering both hemodynamic and the anatomical level was defined.

All the original livers pieces (Rapp [34] and Couinaud [5] are exposed in the conservatory of the faculty of medicine (Montpellier), it’s certainly to our knowledge the most important collection on the corrosion anatomy livers in the world. Here we present their descriptive findings and some original pieces.

Methodology of corrosion

Rapp [34] employing the plexiglass in solution in acetone, for its first four livers (A.B.O.D.) developed the technique of injection. Due to the defects inherent in plexiglass, three “vinylchloride-acetate resins (VYHH)” in the form of white powder were used. The “vinylchloride-acetate resins” vinylchloride and vinyl acetate chemically mixed during the manufacturing, produced a new resin combining the best characters of the two components. The force, the hardness and the chemical resistance of the polymerized vinylchloride (VYLF) are allied with the plasticity and the extreme solubility of polymerized vinyl-acetate. It had the advantage for the injection of vessels of small caliber of giving solutions much more fluid than any other resin for a given concentration.

In the Rapp’s work [34], the first 50 livers were injected with a solution of VYHH-1, the following 84 with VYHH-3 and 5 livers of fetuses injected with VYLF, ideal indication, given the wealth of vessels requiring a viscosity as low as possible to be well injected.

The use of the plastics introduced by Schummer [38] with the monomeric vinylchloride easy to handle and allowing higher results allowed a better evaluation of the vascular anatomy by corrosion. The polymerization of the monomeric vinylchloride, required a high temperature and the addition of a catalyst made its use complex. Narat et al. [29] simplified the technique by employing the acetate of polyvinylchloride in an acetone solution. Thereafter Hill [15], Sanders [36], Trueta [43] improved the technique.

Techniques of corrosion by Rapp [34]

The studied livers taken 12–48 h after death were injected between day 8 and day 12 after cold conservation. The more or less long interval between death and injection had no consequence in obtaining injection mouldings. To preserve the exact shape of the liver, the parts as soon removed, and until corrosion, were laid out in suspension in water and water injection had the advantage, in case of leak, to cause an immediate gel of the plastic solution and to obtain its filling. Before the injection an acetone washing of the venous network eliminated the clots avoiding quick setting, in contact with the venous walls, of the first cubic centimeters injected. The portal system was injected by the trunk of the portal vein and by the hepatic vein system, either by catheterization and separated injection of the three large hepatic veins (livers A and B), or by the inferior vena cava cut through the basis of the auricle. Each liver was plunged in a large 20% hydrochloric acid container, avoided the contact with the walls, evaporation in the mass being slow. After a 2-day washing with a rather strong fountain at request, the corrosion being completed, the liver was immersed for 24 h in an alkaline aqueous solution neutralizing any trace of acid.

Results

The results of the analysis of the vascularization of the liver by Rapp [34] are of a remarkable precision combining total morphology, the variations and the intrinsic vascular segmentations directed towards the technics of hepatectomies. The casts obtained showed that the distribution of the pedicle including portal vein, hepatic artery and bile duct make it possible to divide the liver into a certain number of independent territories, lobes and segments, with well individualized pedicles, drained by vessels which are thrown in the principal hepatic veins located in the plans separating the main lobes (Fig. 3).

Fig. 3
figure 3

Result on methodology used with for each liver (a,b) and drawing vascular structures (c) based on the corrosion technique by Rapp. Liver C (28-year-old woman died of pulmonary tuberculosis with injection 24 h after death, system portal vein: 80 cc, supra-hepatic system: 100 cc). (Collection E. Rapp) A: Superior surface; B: Inferior surface; C: drawing (partial)

Full size image

Portal veins systematization

Classically the trunk of the portal vein, at the level of the hepatic hilum, is divided into two terminal sections, right and left. The left branch runs along the transverse scissura from right to left, continued by a direct segment from back to head, along the left sagittal scissura, and ends up into a round and blind portion (the Rex umbilicae recess). The right branch is very short, it is divided from its origin into two or three branches. The detailed study of the mode of division, of the principal branches of the trunk of the portal vein makes it possible to conclude that a formal symmetry exists between left and right portal systems. (Fig. 4).

Fig. 4
figure 4

Liver’s corrosion: Portal vein (VP) and main branches, RPV (right portal vein), RAPV (right anterior portal vein), RPPV (right posterior portal vein), LPV (left portal vein), LPVt (left portal vein transverse), LPVt/u ( left portal vein transverse to umbilical portion). a Branches of inferior vena cava (IVC), with right hepatic vein (IVC), medial hepatic vein (MHV), left hepatic vein (LHV), for the other branches there are a lot of variations (Collection E. Rapp). b Other piece with more precise branches after liver’s corrosion (Collection E. Rapp)

Full size image

On the left an umbilical sinus or recess described, and a right sinus from one sinus to other are the right and left portal branches constituting the “intersinusal vein”, which runs all along, the transverse scissura of the lower face of the liver. The trunk of the portal vein flows into this intersinusal vein. The existence of a right and a left sinus of equal value was undoubtedly ignored up to now, because if the left sinus is superficial, oblique in bottom and ahead, in an oblique plan obliques to medium to 40° on horizontal on medium, in the sagittal left fissura, relative to the umbilical fisura and ligament, the right sinus is slightly opposite, located in the full parenchyma of the right lobe, directed forward, in a horizontal stright plane. The two portal sinuses are not in the same spatial plane: due to the difference in orientation of the two right and left lobes.

The two branches thus formed are very variable, the trunk dividing more or less on the right or on the left. The interpretation whereby the trunk of the portal flows into the intersinusal vein transversely at an unspecified point makes it possible to explain easily all the variations of the mode of termination of this portal trunk portal. Once having specified that from the angle of union of the right and left sinus with the intersinusal vein right and left the dorsal lobar veins are born. This allows to study the various possibilities created by the progression of the meeting point of the trunk of the portal vein, from the median part of the intersinusal vein until its right end. This meeting point never moves towards the left half of the vein intersinusal vein: the left dorso lobar vein has no direct relationship with the termination of the trunk. This is why we did not represent it systematically.

If the portal trunk joins fully in the intersinusal vein, there is a fork, each right and left segment varying according to the exact place of the fork. If, moving right, the portal trunk arrives at the very right end of the intersinusal vein, on the same line that the right sinus, one will have an apparent trifurcation: the right branch corresponding to the dorso lobar vein, the median branch to the right sinus.

The right dorso lobar vein is divided from its origin into two branches, the cases of quadrifurcation are explained. The mode of termination of the portal trunk classified under the term of “special form” is explained as follows: with the progression to the right of the portal trunk, is associate a slip to the origin of the right dorso lobar vein. In one case, a mode of termination entirely confirms our point of view: the translation towards the right of the portal trunk in the right dorso lobar vein.

The concept of an intersinusal vein and of two right and left sinuses agrees perfectly with the theory by Hjortsjö [16], from existence of two right and left portal territories, individualized and separated by a clear cut scissural plan. From each sinus and the corresponding extreme part of the intersinusal vein are born the various elements ensuring the portal vascularization of the corresponding hemi-liver, and constantly whatever mode of termination of the trunk of the portal vein. While admitting the concept of a right portal branch vascularizing the right hemi-liver and of a left branch vascularizing the left hemi-liver, how can one interpret the systematization of the portal vein vascularization when there is no more junction of the trunk of the portal vein, but trifurcation, quadrifurcation, a special form of division or even a very atypical division as it is for the piece.

Each sinus ensures, with the corresponding extreme part of the intersinusal vein, the portal vascularization of the corresponding hemi-liver, these two hemi-livers, right and left, being separated by a scissural plan, called “principal scissural” by Couinaud [5]. It should be specified how each sinus ensures its role as a “vascular distributor”. At first sight the vascular distribution of each hemi-liver can appear complex: actually, beside many secondary branches, of large elements are constantly found at the origin of one system. This system is a remarkably constant from one piece to the other, whereas the variations of detail are numerous (Fig. 5).

Fig. 5
figure 5

a–d Variations on portal vein (Collection E Rapp)

Full size image

Being established by principle that often, this being particularly true for the right hemi-liver much bulkier than the gland, a single schematized branch will be duplicated or tripled or surrounded by a bunch of small additional veins, one can state that each sinus ensures its role of portal contribution through three systems of veins.

The external zone, on the right of the right sinus, on the left of the left sinus, is vascularized by two veins: one originating from the external angle of union of the sinus with the intersinusal vein: “right or left dorsal lobar vein, the other originating from the lobar anterior extremity on its external edge: “right or left ventral lobar vein”.

- The median intersinusal zone, is irrigated by two veins each originating from the anterior extremity of each sinus, on its internal edge: “right or left medial sagittal recurring veins”. They present a general curve leading them, backward and upward to the hepatic convexity.

These three venous systems, ensure the portal vascularization of each hemi-liver. Given the least volume and the general orientation of the left hemi-liver, these three veins are generally well individualized, easily to see. On the right, on the contrary, the variations in the details are more frequent. The parenchymatous mass being more important, the branches are often more numerous.

Here, on the contrary we want to establish a “general systematization”, this one being necessary to the development of any analytical work: before speaking about the exact number of vascular branches originating from such or such place, should no one know what they correspond to so that the whole is understandable.

These two territories of portal irrigation door are separated by a scissural plan that no portal element crosses, and as Hjortsjö insisted [16], no arterial or biliary element: this is why the term of “principal scissura” can be regarded as particularly relevant. This relevant scissural plan presents a double slope in the frontal plan, it proves slightly oblique from top to bottom and from left to right, going schematically from the anastomosis of the medial sagittal hepatic vein in the inferior vena cava at the top, with the medium part of the small cystic cavity at the bottom; according to the general morphology of the liver, this scissural plan will form with the sagittal plan an open angle in the bottom and on the right, from 0 to 10°.

In the horizontal plane, the main scissura presents most often an orientation from to back to front and from right to left, this angle formed with the sagittal plan varying from 0° to 30°, its sinus being opened to the left. The variations of these two angles do depend on the hepatic morphology: any relative increase of the volume of the right liver compared with the left liver entails a decrease of the first one and an increase of the second.

Certain authors wanted to equate this principal scissura with the portal junction portal, this one thus having a topographic value. It is obvious that this concept, sometimes exact, will often prove to be false given the many variations of the topography of the termination of the trunk of the portal vein. As Hjortsjö [16], and Couinaud underlined [5], this scissura is followed from one end to the other by the medial sagittal hepatic vein which stands not only as a fixed reference mark, but also, as a determining factor; thus, the basic problem of the general systematization of the veins of the liver appears: two vascular systems whose very vascular zones are delimited by elements of the other system.

This systematization of the venous portal network proves to be simple, clear and symmetrical. It provides a simple and satisfactory interpretation for all the variations met, not easily explainable differently. It will prove very useful by its perfect and easy integration in the overall systematization of the veins of the liver.

Hepatic vein systematization

If the portal circulation divides the liver into two vascular territories with the “right portal lobe” and “left portal lobe”, hepatic vein circulation will divide it into three vascular territories “right hepatic lobe”, “right hepatic medial lobe” and “left hepatic lobe”, each one of these lobes being completely and only drained by one of the three large hepatic veins: right, sagittal, medial and left. These three veins flow into in the inferior vena cava at the upper part of its hepatic segment, the meeting point being on the anterior faces and the right and left neighboring zones of the lateral face of the vein. Sometimes the three veins flow into a single opening, most often into two openings: one for the right hepatic vein, the other for the left hepatic vein which very quickly is divided into two branches: the medial sagittal vein and the left hepatic vein itself; in certain cases, each vein presents a total autonomy, flowing together separately in to the inferior vena cava (Fig. 6).

Fig. 6
figure 6

Variations about organization of collateral supra-hepatic veins (Collection E. Rapp)

Full size image

It is necessary to point out the frequent presence of additional hepatic vein originating from the whole right face of the hepatic segment of the inferior vena cava and intended for the right lobe. These additional veins topographically in sub and retroportal situation, acquire sometimes an enormous importance, ensuring the venous drainage of most of the right lobe: there is always a balance between the volume of these additional right hepatic veins and the volume of the right hepatic vein. These variations do not interfere in the construction of a general systematization: as numerous and varied they may as numerous and varied they may be in the same injection moulding, they always remained perfectly within the general framework of the systematization of the unit which proves besides extremely simple.

The right hepatic vein has its origin towards the anterior extremity of the right lobe in its external part, in general more runs posteriorly upward to the right edge of the liver than to the cavity. It runs posteriorly, upward and to the left, until his confluence with the anterior face or with the right edge of the inferior vein cava towards behind, in top and the left until its meeting on the anterior face of the lower vena cava. It follows a general curve drawing a right convexity. It ensures the drainage of a parenchymatous zone whose exact limits are determined by certain elements of the portal system. This segment, or right hepatic lobe, does not correspond to the morphological right lobe but has a corner shape with a left summit at the level of the inferior vena cava and with a right external base: its volume can enormously vary from one case to case.

The left hepatic vein is practically symmetrical to the preceding one. From its origin to the anterior part of the left lobe it runs upward backward, to the right towards the lower vena cava, thus describing a convex curve to the left. However, much more often than on the right, this vein will present variations the principal and the most frequent of which an early division into two branches, one for the anterior part of the morphological left lobe, the other one for the posterior part: they are the two veins of segments III and II of Hjortsjö [16] and Couinaud [5]. Thus constituted, the hepatic left vein drains the parenchymatous territory located to the left of the left sagittal fissure, but here to the exact limit of this “left hepatic lobe” is fixed by an element of the portal system, this lobe having shape a corner with a right summit at the level of the inferior vena cava, and an anterior left base.

There remains a median parenchymatous zone in the corner shape with a posterior apex at the level of the vena cava and an anterior base: it is the “median hepatic lobe”, the territory of drainage of the medial sagittal hepatic vein coursing from front to back providing branches running on either side of its axis, towards the convexity of the liver.

This systematization of the hepatic veins is simple, and classic. The three hepatic lobes are separated, as were the two portal lobes, by two clear scissural plans that no hepatic venous element crosses. As in the main fissure no macroscopic portal element passed through, in the same way in the two scissuras separating the right and median, and median and left hepatic lobes, no macroscopic hepatic element passes through these scissural zones are on the other hand these scissural zones are crossed by portal arterial and biliary elements: this is why, in contrast to the portal scissural plan, we will call them right and left secondary fissures. We will study their exact value with the overall systematization of the veins of the liver. He found some special pieces like Riedel lobe, or individual caudate lobe, or metastatic lesions. (Figs. 7, 8, 9).

Fig. 7
figure 7

a–b Redel’ s lobe on two cases with liver’s corrosion (Collection E. Rapp)

Full size image
Fig. 8
figure 8

Injection corrosion of caudate lobe (Collection E Rapp)

Full size image
Fig. 9
figure 9

Liver’s corrosion with two metastases (yellow arrows) (Collection E Rapp)

Full size image

General systematization

For Rapp the liver is divided, from the vascular point of view, into five lobes: two portal lobes, right and left, and three hepatic lobes, right, median and left: we think there are not 5 hepatic vascular lobes, but two lobes and three lobes”, or two portal blocks and three hepatic blocks”, each block of a system being used to anchor the blocks of the other system: each one of those sharing in its own way totality of the liver, this leads to a superimposition of the portal and hepatic territories. It is the basic essential condition necessary for any hepatic surgery of resection, having to strictly respect the two venous systems of any parenchymatous zone left in place.

Which are the connections, the precise limits of the various zones of portal contribution portal and hepatic vein drainage? The study of the scissural plans reveals the interdependence of the two systems: each one plays the role of topographic determinant of the other, each scissural portal being centered by a hepatic vein element and vice versa. The portal scissural plan portal, or principal scissura, is directed from back to front, along to the medial sagittal hepatic vein which, at the bottom of the scissura, plays the role of determining factor. This portal scissural plan is doubly oblique on the sagittal plane, the two created angles varying according to the general morphology of the liver. It will be possible to know the general slope of this scissural plan by the sole morphological study. The two hepatic scissural plans or right and left scissuras correspond to two tangent plans at the left edge of the lower vena cava and passing one through the right portal sinus (secondary right scissura) portal, the other by the left portal sinus (left secondary scissura). These two portal sinuses correspond to a hepatic intersegmentary portal inter segmentary sinus as the medial sagittal vein corresponds to a portal intersegmentary hepatic vein. According to the general morphology of the liver, the angle, delimited by the two secondary scissural plans may vary from 35° to 90°.

Anastomoses between the hepatic veins

The input of this study is the frequency and the multiplicity of anastomoses, often of large caliber, existing between the hepatic veins. The multiplicity of anastomoses between the various hepatic veins is opposed to the absence of anastomosis between the portal branches. The system of the hepatic veins ensures the return circulation, and the anastomotic multiplicity of the efferent vessels is a general fact from the anatomical point of view, to which the hepatic circulation does not make an exception. On another hand, the portal vein is not a venous system, it is an efferent vessel whose distribution is strictly terminal. The portal system and the hepatic veins derive from the same embryonic venous trunks split up by the proliferation of the hepatic spans. Half the time the large venous trunks present an anastomotic system, rich and abundant once every fourth time. The casts which were used as a basis for this study were built with viscous enough solutions to eliminate the jumble of very fine branches. From a practical point of view, the traumatic or surgical, involuntary or not interruption, of a large hepatic trunk, its isolated thrombosis, do not seem to out of the return circulation in its territory of origin, at least in most case.

Systematization of the spieghelian portal veins [34]

For the classic authors, the right portal branch right receives the cystic vein and then divides into two or three secondary branches which are distributed in the right lobe and the Spieghel’s lobe. Cordier [4] in a study on the vascular pedicles of the square lobe and of the Spieghel’s lobe, concluded that the portal vascularization was carried out the middle by median short branches originating from the right and left primary trunks of the portal vein at the left extremity by the spieghelian recurring vein coming from the recessus of Rex, at the right extremity by the vein of the right anterior angle of Spieghel’s lobe,which sometimes arises with the vein from the right posterior angle of the left lobe.

Huard [17] on 40 livers, described two systems: one symmetrical, where the branches left and right branches of the portal vein giving two or three collateral, including a principal on each side; the other one asymmetrical, where the left branch of the portal vein gives four to six branches “including three principal constants”. Couinaud [5] on 10 vascular injection mouldings, reported that: the number of the portals vein of the Spieghel’s lobe varies from 2 to 4 (2: 6 times, 3: 3 times, 4: I time) from the transverse segment of the portal vein, sometimes from the posterior face of the junction, seldom from the portal trunk itself, most often between the origin and the elbow of the left portal vein; the right pedicle originates from the very initial part of the right portal vein. None of these authors gave the percentages of various systematizations. For Rapp [34] on 134 casts, it arises that many variations are possible, the most frequently encountered from existing in only 15% of the cases. The various possibilities are divided into seven groups, maximum number possible, since these Spieghelian veins portals may have only three distinct origins: the portal trunk, the right portal branch and the left portal branch. He provided precise details on the number of Spieghelian portals veins: 1 vein: 6%, 2: 27%, 3: 40%, 4: 25%, 5: I %, 6: I % and on their origin: portal trunk 41%, portal left branch 91%, portal right 25%.

Systematization of the spieghelian lobe hepatic veins branches

The venous drainage of the Spieghel’s lobe is variously assessed without much precision and percentage. For Sappey [37] “a few branches emanated from the Spieghel’s lobe drain into this part of the vena cava which corresponds to the gutter of the liver. Poirier [31] specified: “the large right hepatic vein, the bulkiest, draws its origin from the right lobe and receives an important affluent from the Spieghel’s lobe the large right and left hepatic veins each receives a part of the venous blood from the square lobes and from the Spieghel’s lobe. For Huard [17], beside small veins entering into the inferior vena cava through the staged foraminula, there exist two Spieghelian principal and constants hepatic vein: one superior draining into the inferior vena cava, just below the insertion of the large hepatic right and left veins, and the other one medium entering the inferior vena cava in its medium part.

Couinaud [5] describes a large Spieghelian vein which enters into the antero-right side of the inferior vena cava, half way up its hepatic bed, at its highest extremity. Only once, in a very wide lobe, the three Spieghelian veins drained into the principal body of the left hepatic vein, between sagittal vein and left lobar vein.

Rapp [34] gave details on the systematization of the Spieghelian hepatic veins on 134 casts, showing a very frequent essential type, where the venous drainage of the Spieghel’s lobe is done directly towards the inferior vena cava, the Spieghelian hepatic veins entering its left anterior side; type of the cases, where the Spieghel’s lobe remains latero, not overlapping the posterior face of the inferior vena cava. Another type, observed in a relatively frequent way, but with certain variations, where the spieghelian hepatic veins drain at the same time into the left hepatic vein and into the inferior vena cava. This possibility is observed when the Spieghel’s lobe has a more or less expansion of the parenchyma developed towards the line, behind the inferior vena cava. In a case no branch run to the inferior vena cava, all the Spieghelian system draining towards the left hepatic vein. In no case a Spieghelian hepatic vein drained into the right hepatic vein. Rapp brought details on the number of Spieghelian hepatic veins: 1 vein: 51%, 2: 37%,3: 10%, 4: 2% and on their mode of termination into the inferior vena cava: 99%, left hepatic vein: 14%, right hepatic vein: 0%.

If the external morphology of the liver with its four lobes does not correspond to the internal vascular structure, at least it has a direct action on the systematization of the venous system: complex systematization at first sight although quite simple, each of the two systems representing the intersegmental element of the other. This systematization of the veins of the liver will have a direct practical consequence: a portography, displaying the two sines, and the morphological appearance of the liver will be the two necessary and sufficient elements to know the exact topography of the hepatic scissural plans. The principal scissura itself will be easily located, the medial sagittal vein coursing the bisectrix of the median hepatic lobe.

Contribution of Couinaud about systematization

He was 30 years when André Delmas in Paris asked him to study arterial anastomoses between the right and left liver in the hilum. He advised him that good results could not be obtained by dissection, and that injection casts were needed. He perfected previously established technic using polyvinyl acetone. Following the analysis of the vascular elements he proposed a systematization intended to codify hepatectomy.

Couinaud’s methodology [5] was comparable to Rapp’s [34] with livers injected on the third day post mortem with an acetone solution of polyvinyl and corroded in hydrochloric acid. The tiny veins draining into the cava trunk have seen neglected. His descriptive studies are comparable to Rapp’s but differ from them by a systematization intended for the realization of hepatectomies. We can ask why Couinaud’s work had been during his life and actually worldly cited about liver anatomy? We can propose an explanation due to the fact that in Montpellier at this period the surgeons, where not truly specialized in liver surgery. In contrast in Paris many surgeons were especially interested in liver pathology. When actually we take in account for Rapp and Couinaud the problems encountered firstly on technical point of view, secondly the number of human cadaver’s donators and thirdly the time to obtain such result with the exceptional quality on liver anatomy, one must be amazed.

Couinaud published his result in 1958 on 30 livers [6] “Segmental boundaries are defined by the hepatic veins. Segmental anatomy enables segmental resection based on vascular inflow and outflow to liver parenchyma and has been useful in sublobar resectional surgery. A distinct advantage of segmental anatomy in an operative classification of liver injuries is its simplicity and consistency. In scheme the liver is divided into eight segments based on end distribution of the portal vein.

This description is still the international reference especially for liver surgery [6], although some prefer the Brisbane 2000 classification [40]. This segmentation allows to perform anatomical liver resection. Better knowledge of liver anatomy and improvement in surgical techniques also allow non-anatomical resection. It is the distribution of the portal vein which is used as a guide for the distribution of the other glissonian elements. The division of the portal vein into right and left branches makes it possible to distinguish a right and a left liver, largely exceeding the extent of the anatomical left lobe.

The hepatic vein territories are separated by scissuras by a plan which may be called the principal scissura or the great scissura. At the lower side of the liver this scissura follows from front to back the bottom of the bed of the gall bladder, whose right aspect belongs to the right liver and the left aspect to the left liver. The plan of the scissura the route of which has just been at the lower side of the liver draws with this face a varying angle of 55° to 90° open to the left.

All that is on the left is irrigated by the left elements of the glissonian pedoncle (left lobe, square lobe and left side of the bed of the gallbladder, caudate lobe), all that is on the right is irrigated by the right elements. In the scissural plan, there is a large hepatic vein, the large sagittal, which receives affluents from the right and left and flow into the origin of the left hepatic vein. It drains the whole left lobe which must be considered as the left hepatic sector situated in the left portal scissura. Its trunk is perfectly well-shaped at the posterior extremity of the liver and crosses the very posterior part of the Arantius’s fissure, follows the upper edge of the caudate lobe to flow into the left inferior cava trunk. The trunk of the left hepatic vein is always formed at the posterior extremity of the left hepatic of the left hepatic vein sector and is linked with the medium vein, to form the joint base which drains into the face antero-left aspect of the inferior vena cava. In the most frequent type (20/30) there exists a large main sickle-shaped vein, with a right concavity arriving from the anterior point or from the anterior anterior part to the left edge of the left lobe, closer to the left scissura that left edge of the liver, moving towards the posterior extremity, of the fissure of Arantius. It receives a relatively important lateral quota with great variations. Whatever the multiplicity of the veins, no difficulty is to be feared at the time of the resection of the left lobe, because the left scissura is perfectly delimited from the anatomical point of view, and by following it from front to back one respects the elements from segment IV, even when they get to the left vein, and one is sure to meet behind the trunks coming from the left lobe which one can identify perfectly and bind. The left lobe forms a hepatic entity, comparable from the portal point of view formed by the fusion of the left lateral lobe and the part of the left paramedian lobe, located on the left from the round ligament, lobed livers; it is cut transversely by the left portal scissura, corresponding to the left interlobar scissura of the lobed livers. In this type of liver, the return flow of is ensured by 2 distinct veins corresponding to each formation lobar and joining posteriorly.

The medium or sagittal hepatic vein is a large vessel occupying the posterior part of the main portal scissura separating the right and left livers, following the edge of the bed of the gall bladder and joining the left hepatic vein to form the common trunk undermines which drains into antero-left aspect of the vena cava, very close to the higher edge of the posterior face of the liver. It drains the medium or median hepatic sector. The origin comes from the junction of affluents coming from the right liver and the left liver (respectively, territories of irrigation of the portal right and left veins, located on either side of the main portal scissura.

This origin, is variable closer to the inferior than the upper one. The main affluents flow into the right and left sides or into the upper face of the main trunk. Among the right affluents, beside a multitude of small branches of secondary importance, some are bulkier, the left elements are less important and the branches coming from the lower face of the medium hepatic vein are insignificant.

The right hepatic veins, corresponding to the right, the hepatic sector, are triple and constantly include a very bulky right upper hepatic vein, located in the right portal scissura, which flows into the side antero-right aspect of the cava trunk. It draws as a whole a sickle -shaped line with a left concavity closer to the right hepatic scissura than to the right edge of the liver. It receives many affluents which terminate on its various faces, but the main one flow into its right edge the most important are 2 to 5, and the bulkiest finish in the posterior half of the right sector.

The posterior elements are transverse, while anterior are oblique backward and on the left. Sometimes they can gather into one or more large trunks before flowing into in the hepatic vein. From the surgical point of view, the upper right hepatic vein always receives affluents until its termination into the vena cava. Out of a series of 100 parts, the vein of the posterior face of the liver flew 6 times directly into the cava trunk. The group of the vessel and its affluents is variable with 3 great types testifying to a great variability. The medium and lower right hepatic veins are of small caliber sometimes far more developed (able to reach the anterior edge of the liver) are located at the lower face of the liver, flow into the vena cava, at the lower part of the hepatic gutter below main portal branches. The dorsal sector, or caude lobe, drained by several venules, one of these being bulkier, directly flow into the cava trunk. One of the characteristics of these vessels, when they reach a certain development, is their situation hypoportal, whereas all the other hepatic veins are located above portal branches. The medium veins have a transverse direction, while the lower have an antero-posterior direction. They flow into the cava trunk at the level of the lower part of the hepatic gutter.

The left hepatic scissura located between the left and the medium hepatic vein, corresponds to the left anteroposterior fissure of the liver and to the insertion of the sickle-shaped ligament. It corresponds very exactly to the left longitudinal fissure of the lower face of the liver and to the point of attachment of the sickle-shaped ligament and separates the left hepatic sector (the classic a left lobe) from the median hepatic sector. Its anterior extremity anterior end corresponds to the insertion of the round ligament, while its posterior extremity coincides with that of the fissure of Arantius [24].

The right hepatic scissura located between the medium veins hepatic and right-hand sides, is much more variable in its anatomical route, due to the reciprocal variations of the extent of the sectors it separates. It is less constant in its route than the preceding one. The caudate lobe or dorsal sector is the only portion of the liver, where portal and hepatic sectors superimpose themselves, really forming a lobe in sense of Gans [9]. It is separated from the other part of the liver by the right and left dorsal scissura. Its fixed posterior extremity corresponds to a point located between the orifice of the upper right and sagittal hepatic veins. Its anterior extremity varies, on 9 pieces it corresponds to right edge of the gallbladder bed, on 2 pieces the anterior and right angle of the liver, on one piece it terminates to the union of the 3/4 posterior and 1/4 anterior of the right edge of the liver. Neither right portal scissura nor anterior extremity are marked by a precise morphological accident. It draws a concave curve, more clearly accentuated in its posterior aspect on 6 injection mouldings, due a more important development of the right posterior affluents of the sagittal vein (veins of the segment portal VIII segment)”.

Practically, the surgeon must regard as the hepatic scissura the plan which contains the insert portal pedicle with two hepatic pedicles. The constant left hepatic scissura, is located thank to the point of attachment shaped ligament, the more variable right scissura, located as a whole at a short distance from the right edge of the gallbladder is centered on the paramedian right portal pedicle. It seems far from probable that the resections carried out within these limits compromise the hepatic circulation in the remaining parenchyma (Fig. 10).

Fig. 10
figure 10

Liver’s corrosion according to Couinaud

Full size image

The hepatic segmentation by Couinaud

This distribution makes it possible to distinguish two territories: that of the upper and left hepatic vein, paramedian territory, and that of the inferior and right hepatic vein, external territory. These two zones are separated by a virtual scissural plan, the right scissura, inclined in a variable way on the lower face of the liver according at an angle of 25°–75° and spotted anterior by a line joining a quite frequent notch from the anterior edge frequent of the inferior previous to the right edge of the lower vena cava. If one draws a transverse line following the right and portal veins, portal axis, these territories are divided into VIII segments: segments I to IV form the left liver (Couinaud redefined the caudate lobes as segment I), segment V: territory of the anterior the previous collateral and terminal branches of upper and left hepatic vein, segment VI: territory of the anterior veins of the transverse segment of the inferior and right hepatic vein, segment VII: territory irrigated by the distal segment of inferior and right hepatic vein, segment VIII: territory of the posterior terminal.

Goldsmith [11] considered left lobe as segments II and III. The quadrate lobe is segment IV and subdivided by its portal blood supply into segments IVA lying superiorly and IVB interiorly. The right liver is segments V (anteroinferiorly), VI (posteroinferiorly), VII (posterosuperiorly) and VIII (anterosuperiorly). Couinaud proposed the caudate lobe with the part to the left of the inferior vena cava (segment I) and right (segment IX). Hepatic venographies [25] were performed selectively in 42 patients with hepatocarcinoma. The findings were evaluated from anterior and lateral views. Thirty-nine right hepatic vein could be identified and the existence of one branch as the first ramification was found in 36 cases (92.3%). The first branches of the right hepatic v. could be classified into veins (V7) running from segment VII and those (V8) running from segment VIII. (A V 7) was identified in 26 cases (72.8%) and a V8 was identified in 10 cases (27.8%). The vena hepatica dorsalis (6) (V8) running from segment VIII was recognized in 10 cases. The middle and left hepatic vein were identified in 31 cases and 33 cases, respectively. Two main types of middle vein (one with no principal branching and the other with a branching) were found in 11 cases (37.9%) and 12 cases (41.4%), respectively. The first branch of the middle hepatic vein (V8) running from segment VIII was identified in 10 cases (32.3%). These results indicate that anatomical consideration of the hepatic vein in each patient is necessary when performing hepatic resection.

General concepts in hepatic segmentation have been discussed for many years. Recently hepatic segmentation based on the distribution of the portal vein has been proposed. It is from this perspective that we now evaluate findings of hepatic venographies and consider hepatic segmentation based on the hepatic vein. The vein running from segment VII (according to Couinaud's classification [6] directly to the inferior vena cava was identified in 1 case. A general survey on segmentation liver had been done by Poston [33]. Lee et al. [21] showed the existence of significant circulation between the middle hepatic vein and right hepatic vein in four out of five liver donors, mainly located in segments V and VIII. The internal anatomy of the liver raises many issues. From the surgical point of view the arterial system, according to the experiments by Duckett and Montgomegny [8], can be regarded as secondary, as well the biliary system the systematization of which may be precised by per-operative opacifying. In a study, Morel [28] found that in 62% of cases there were communicating veins between the right hepatic vein and the medial hepatic vein, located in segments V and VIII.

We agree with the complimentary paper by Belghiti [1] "the work that Claude Couinaud published in 1957 « The liver: anatomical and surgical studies» was a revolution for surgeons. This precise knowledge of anatomy has fostered the development of liver surgery and the French pre-eminence of this surgery owes much to our impregnation of his work. The greatest strength of this "anatomist surgeon" was his ability to adapt his anatomical descriptions to the evolution of liver surgery and the tremendous upheavals of radiological explorations. In his latest book "Tell me more about liver anatomy", his deepening of anatomical descriptions finds a justification with the development of the transplantation of a part of the liver. The need to transplant children with a reduced or shared graft, and the possibility of transplanting adults with a hemi-liver requires the removal of a perfectly vascularized and drained graft (Fig. 11) The lack of cadaveric grafts that can be used in adults has led to the development of transplantation of a right or left liver either from a sharing of a cadaveric graft or from a sample from a living donor. The Couinaud segments being the foundation of this nomenclature, the right liver (segments V to VIII) is designated right liver or right hemiliver and right hepatectomy right hepatectomy or right hemihepatectomy. The left liver (segments II to IV, including or not I) is designated left liver or left hemiliver and left hepatectomy left hepatectomy or left hemihepatectomy. The right lobectomy disappears in favor of extended right hepatectomy or extended right hemihepatectomy.The left lobectomy disappears in favor of bisegmentectomy 2–3 or left lateral sectionectomy. Sectionectomy replaces all sectorectomies thus defining the exeresis of segments V and VIII in right anterior sectionectomy, and excision of segments VI and VII in right posterior sectionectomy. It is difficult to predict whether the term sectionectomy will supplant the term bisegmentectomy. The term segmentectomy applies to all Couinaud segments; II or III segments excision is being defined in bi or trisegmentectomy. By titled his article "Error in the topographical diagnosis of liver damage," Couinaud illustrates the enormous changes that radiological explorations have brought to us. He pointed out some errors in surgical terminology; Thus, the left side sector is only made up of segment II and not the left lobe (segments II and III). The left lobectomy should, therefore, not be called left lateral sectionectomy, but this term corresponds stricto sensu to a more complicated intervention and creating a larger slice. In this bubbling article, Couinaud reappropriates his anatomy by redefining the angles of obliquity of the splits and their variations, calculating the average width of segments VI and VII and their variations according to the morphotype of each individual. He was an anatomist but also a visionary and shows that the anatomy of the liver can change."

Fig. 11
figure 11

Comparative morphology between liver in adult’s and newborn’s (3 livers). It shows the necessity for the use of a lobe (segment) for transplantation in newborn (Collection E. Rapp)

Full size image

Conclusion

Two “surgeons-anatomists”, Rapp in Montpellier and Couinaud in Paris, provided at the same period precise anatomy useful for the hepatic surgery and to the tremendous upheavals of radiological investigations. Although discussed, the fundamental anatomy of liver is classic with 8 segments as autonomous territories having a vascular afferent pedicle and a venous and biliary efferent drainage. These autonomous segments supported the surgical approach to the liver with the possibility of resecting a hepatic territory in a carcinological perspective which saving vascularized parenchyma. It will be in the future history about liver morphology to specify the contribution of these two “surgeons-anatomists”. The collection of the original livers corrosions may be seen in the “Conservatory of Anatomy” in Montpellier (France).