The arterial blood supply of the pancreas: a review. I. The superior pancreaticoduodenal and the anterior superior pancreaticoduodenal arteries. An anatomical and radiological study

The gross anatomy of the pancreatic blood supply has been subjected to numerous studies. The results of such studies, however, have never been summarized in detail, even in the most important textbooks. For this reason, a certain confusion was generated, especially regarding the interpretation of the nomenclature used to identify pancreatic arteries. This review summarizes more than two centuries of studies of the gross anatomy of the pancreatic blood supply, clarifies the arterial nomenclature, and underlines the aspects about which anatomists are not in agreement. Moreover, it supplies, as far as possible, documentary evidence for numerous observations previously reported only verbally. For this purpose, more than 200 references were directly consulted to provide the anatomical background of the topic, and more than 1000 angiograms were studied to support the review with original figures. The present paper, on the superior pancreaticoduodenal and anterior superior pancreaticoduodenal arteries, is the first of a series of articles dealing with the pancreatic blood supply.

[The arteries of the stomach]

In order to specify the gastric-arterial systematization, 65 fresh stomachs of adults deceased from non-gastro-duodenal causes, extracted by necropsy, allowed us to study the arteries of the stomach through the injection-corrosion technics. A selective injection for each arterial trunk has been made on 15 stomachs. This work allowed us to know better the arterial supply of the stomach and ut into obviousness important vascular anastomosis. While emphasizing on the substitute notion, the persons who wrote this summary describe regions poor in anastomosis, which correspond surgically with "critical" areas. They also emphasize how important it is to have an estimated score of the ischemia for gastric surgery, especially in the case of oeso-gastroplasty.

Tortuosity of the human splenic artery

Arantius (1571) was the first to describe tortuosity of the splenic artery. The present study investigated the variations in its tortuosity in man, and possible relationships with age, sex, and presence of atheroma. Twenty-nine cadaveric specimens and forty-four celiac angiograms were studied. The straight distance from the origin of the splenic artery, from the celiac trunk, to the point of commencement of the hilar branches was measured, as was the total length of the artery between these two points. The ratio of these two measurements is called the "index of tortuosity." The cadaveric arteries were then opened and graded for the presence of atheroma on a scale of 0 to 3. Marked variation in the index was found in both the cadavers and the angiograms. No definite relationship was found with sex. However, there was a suggestion of increasing tortuosity with age, although in one 10-year-old girl, marked tortuosity was demonstrated on angiography. No significant correlation was shown between increased tortuosity and the extent of atheroma. At present, there is apparently no satisfactory explanation for tortuosity of the splenic artery.

A comparative study of the tortuosity of the splenic artery

The splenic artery of four mammals--pig, greyhound, rhesus macaque, and olive baboon--was studied to determine its tortuosity and pattern of convolutions. The results were compared to those previously gathered in humans. It was found that the isolated cadaveric arteries showed a noticeable species variation in mean index of tortuosity. In the pig and dog, the artery was nearly straight throughout its length, with a very low index of tortuosity, not significantly different in the two species (P > 0.1). The artery was significantly more tortuous in the rhesus macaque and baboon than in both pig and dog, (Ps < 0.001), with a single large convolution present in the proximal one-third. In humans, the artery is commonly tortuous throughout its length and is significantly more tortuous than in the pig and dog (P < 0.001), but not significantly more tortuous than in the baboon (P > 0.1) and only just significantly more so than in the rhesus macaque (0.02 < P < 0.05). The speculation that the tortuosity of the artery may be related to habitual posture, being less in the pronograde pig and dog than in the partly orthograde rhesus macaque and baboon, is not supported by results in the wholly orthograde human.

Three-dimensional reconstruction of the pig sheathed artery

The structure of the sheathed artery and the number of branches of the sheathed capillary were investigated by three-dimensional reconstruction computer graphics in the pig spleen. The capillary branched 1 to 6 times in the capillary sheath in more than 90% of all samples examined. The sheathed artery varied in shape and size in accordance with the number of branches and direction of the capillary. The usual descriptions for the sheathed artery were suggested to be mostly in appropriate at least for that of the pig spleen.

The anatomical basis for laparoscopic splenectomy

The anatomy of the splenic blood supply is extremely variable. Two principal patterns, the distributed and the magistral types, include the majority of variations. These two types differ with respect to the length of the splenic trunk, the number of arterial branches and the proportion of the medial surface of the spleen occupied by the branches as they enter the organ. The suspensory ligaments of the spleen, including the sustentaculum lienis, are usually avascular except for the gastrocolic and lienorenal ligaments on the medial side. The tail of the pancreas touches the spleen in 30% of cases and lies within 1 cm of the spleen in 73%. Knowledge of these basic anatomic facts and the variations should contribute to the safety of laparoscopic splenectomy.

Pancreatic arteries and anatomicosurgical segments in dogs

The distribution of the arteries supplying the pancreas was studied in 22 adult dogs (7 females and 15 males) of mixed breed. The pancreas was divided into 2 lobes, 4 sectors and 6 anatomicosurgical segments. These segments are defined as territories recognizable anatomically and surgically removable. The division and subdivision of the pancreas and the segmental arteries were the following: Left Lobe: Omental sector: I. Proximal segment: splenic a. II. Middle segment: splenic a. III. Distal segment: gastroduodenal and right gastro-omental Interpapillary sector: IV. Juxtaduodenal segment: cranial pancreaticoduodenal a. Right Lobe: Distal sector: V. Distal segment: cranial pancreaticoduodenal a. Proximal sector: VI. Proximal segment: caudal pancreaticoduodenal a.

The whale (Odontoceti) spleen: a type of primitive mammalian spleen

Three spleens from two Odontoceti species were studied histo-anatomically. These spleens consisted of lymphatic nodules, the red pulp (broad sense), and the trabeculo-capsular system composed of the elasto-fibroleiomyocytic tissue. The periarterial lymphatic sheath (PALS) was unclear. Two layers, the intermediate zone and perivenous layer, were distinguishable in the red pulp (broad sense). The perivenous layer was narrow in width and consisted of venules and the intervascular reticular tissue rich in myeloid cells. The collecting and drainage veins were enclosed in this layer. The perivenous layer corresponds to the red pulp (narrow sense) of the common mammalian spleen and may be under involution in a process that probably relates to the remodelling of the intrasplenic vein. The pattern of the arteriovenous communication seemed to be closed, and no ellipsoids were noted around arterial terminals. The Odontoceti spleen has two venous drainage routes (hilar and capsular systems), suggesting a primitive state of evolution, and may be an additional example of the primitive mammalian spleen.

[Anatomical study on the branches of the celiac trunk (IV). Comparison of the findings with Adachi's classification]

The authors studied the mode of branching of the celiac trunk in 184 Japanese cadavers during dissecting practice at Gifu University from 1985 to 1988. The findings were as follows; 1) One hundred and sixty-six of the 184 cases belonged to Type I of Adachi's classification of variation in the arrangement of the branches of the celiac trunk. One hundred and twenty-nine cases belonged to Group 1, one case to Group 2, 10 cases to Group 3, 9 cases to Group 4, one case to Group 6, 2 cases to Group 7, 8 cases to Group 8 and 4 cases to Group 11. Two cases did not belong to any group. One of them had an anastomosis between the hepatic artery and the superior mesenteric artery. The other one had the middle colic artery arising from the hepatic artery. 2) Seven cases belonged to Type II, Group 12. 3) Two cases belonged to Type III. One belonged to Group 18. The other one had an accessory right hepatic artery arising from the superior mesenteric artery. 4) One case belonged to Type IV, Group 20. 5) Three cases belonged to Type V, Group 23. 6) Two cases belonged to Type VI. One case had a gastrolienal trunk and an accessory right hepatic artery arising from the superior mesenteric artery. The other one had a gastrolienal trunk and an accessory right hepatic artery arising from the gastrolienal trunk. 7) Of the 184 cases, three could not be classified according to Adachi. Two cases had a gastrolienal trunk and the hepatic artery arising from the abdominal aorta. One case had a lienomesenteric trunk and a gastrohepatic trunk.

Anatomic basis for conservative surgery of the spleen

100 pancreatico-splenic blocks obtained at autopsy were treated by the injection-corrosion method. The splenic artery ended in the lienorenal ligament in 64% of cases, in the pancreatic segment in 32% and in the hilar segment in 4%. In 84% of cases, the artery divided into two lobar arteries, upper and lower. In 16% the artery trifurcated. The pedicle was 20 to 60 mm long (mean 35 mm). The branches of the splenic artery were in three successive stages: the primary or lobar arteries were each associated with a lobe; the secondary or segmental arteries each supplied one of the constituent segments of a lobe; the tertiary or subsegmental arteries supplied the various areas which made up a splenic segment. Nine cases of extraparenchymal anastomoses were found. There were 32 intraparenchymal anastomoses, being either interlobar or intersegmental. There were a few interlobar anastomoses. The findings support the notion of upper and lower intersegmental and interlobar planes. The latter is almost avascular, while the former has limited vascularity. These planes make conservative surgery of the spleen possible.

"Intermediate zone" of mammalian spleens: light and electron microscopic study of three primitive mammalian species (platypus, shrew, and mole) with special reference to intrasplenic arteriovenous communication

The intermediate zone (IZ) of nonperfused and perfused spleens in three species of primitive mammals (shrew, mole, platypus) was studied morphologically. The IZ is a tissue zone consisting of plexiform vessels, probably venous capillaries, and is located transitionally between the white and red pulp. The IZ is separated from the white pulp by the arterial net (AN), in which the white pulp arteries terminate. Development of the IZ differs between the three species examined being distinctive in the platypus and shrew. The IZ is thin in the mole spleen. A closed type of arteriovenous (A-V) anastomosis was demonstrated in or around the IZ in the two Insectivora species examined. In the shrew spleen, peripheral arterial branches running within the IZ anastomose with the AN around the follicle. The AN anastomoses eventually with venous plexiform vessels of the IZ around the nonfollicular area of the white pulp to form a closed system. In the mole spleen, A-V anastomoses were noted between white pulp arteries (follicular and AN) and veins of the red pulp, either by direct communication or through fenestrated IZ vessels compatible with the plexiform vessels of the shrew spleen. A-V anastomosis in the IZ is probable, but not confirmed, in the platypus spleen, as analysis was limited to a nonperfused specimen. Well-developed ellipsoids were noted around arterial terminals of the IZ in the shrew spleen. Ellipsoids were also noted around all arterial terminals of the mole spleen directed to the red pulp. Most ellipsoids of the mole spleen appeared located within the IZ. No ellipsoids were present around arterial terminals of the IZ in the platypus spleen. Closed circulation was noted in terminals of the pulp artery in spleens of all three species. All pulp arteries of the mole spleen are postellipsoid segments of white pulp (AN and follicle) arteries. No ellipsoids were found around terminals of the pulp artery (penicillar artery) in shrew and platypus spleens. The IZ is probably homologous to the perilymphatic sinusoid (vein) of the lungfish spleen and may be regarded as part of the red pulp. The IZ may be representative of primitive mammalian spleens that have closed circulation. The marginal zone (MZ) of common mammalian spleens is probably a modified IZ by differentiation (remodelling) of the intrasplenic vein. In this process, withdrawal of venous vessels from the IZ occurred, leaving a lymphoreticular zone with open circulation (MZ). The marginal sinus reported in some mammalian spleens is probably a modified AN formed during this process. Possible morphological alterations of the spleen in vertebrate phylogeny are discussed.

Accessory arteries supplying the human transverse colon

An accessory colic artery which arose from the superior mesenteric artery more proximally than the first jejunal artery and met the marginal artery at the splenic flexure was observed in 32 of 65 subjects (49.2%). Considering its constant morphological features, this accessory colic artery was given the new nomenclature 'superior left colic artery (arteria colica sinistra superior; CSS)'. The unusual middle colic artery arising from the celiac trunk was supposed to be caused by the anastomosis between the dorsal pancreatic artery from the CSS and that from the celiac trunk. An additional epiploic branch derived from the pancreatic branch (posterior epiploic branch) also sent an accessory colic artery in 2 subjects. But judging from its origin and area of distribution which overlaps with that of the middle colic artery, this accessory colic artery is not equivalent to CSS.

Pancreatic autotransplantation in the pig: variations in epigastric arterial blood supply

A major problem in pancreatic autotransplantation is the vascular supply of the left segment, which serves as the graft. We therefore examined the arterial blood supply of the epigastric organs in 36 German landrace pigs. In 19 pigs (anatomical variation 1) there was a pancreatic branch of the splenic artery supplying the left segment. After splenectomy the splenic artery could be used as the graft-supplying vessel. In 14 pigs (anatomical variation 2) the arterial branch for the left segment was the first branch out of the hepatic artery distal to the coeliac trunk. In these cases a splenohepatic bypass was performed to ensure delivery of the blood to the epigastric organs. Thereafter, the proximal part of the hepatic artery could be used as the vessel supplying the graft. The left segment could not be used for autotransplantation in 3 pigs (anatomical variation 3) because of the atypical vascular supply. During the 6-month observation period, no malfunction of the epigastric organs was evident.

[Intraorganic structure of blood vessels and ducts of the pancreas]

Zones and smaller structural unit, segments, were distinguished anatomically in the pancreas on the basis of the distribution of the vessels in it and formation of its duct. The detected features of the vessels and ducts of the human pancreas make it possible to reduce the traumatic character of the operative interventions on this organ and develop radically new surgical procedures in the treatment of diseases and traumas of the pancreas.

Surgical anatomy of the human spleen

Post-splenectomy sequelae are now well recognized, and conservative splenic surgery is widely advocated. However, controversies exist regarding splenic surgical anatomy. We studied 127 human spleens using anatomical dissection and a sequential injection method involving both radiology and corrosion casting, with the primary aim of examining segmental splenic anatomy and blood supply. The existence of well-defined splenic segments was confirmed and these ranged in number from 3 to 7 with a mean of 4.3. Each segment had its own arterial supply and venous drainage. The segments were separated from each other by avascular planes. Subsegments with independent blood supply were also identified. The splenic artery was found to divide into two branches; in all cases these further divided into segmental arteries supplying the central segments of the spleen. The polar segments were supplied by segmental vessels of highly variable origin and size. This study aims to highlight the importance of identifying these segmental vessels at operation when splenic conservation is considered.

Arterial segmentation in the spleen

The authors have conducted a corrosion cast study of the human spleen to systematize the branches of the splenic artery and the vascular territories irrigated by them. The existence of two primary divisional or lobar branches of the splenic artery and their respective territories or lobes is confirmed. The secondary or segmental branches and their vascular territories were studied; an average of 6 segmental territories was found. These findings are compared with those reported in the literature.

Arterial segmentation and subsegmentation in the human spleen

The segmental nature of the arterial tree of the human spleen was analyzed in 181 subjects of both sexes, who had died of various accidental causes. Based on the observation of the pattern of the terminal and polar splenic branches, selective arteriographs and corrosion casts, and taking account of the ideas reported in the literature, we proposed that the spleen is divided in arterial segments and subsegments. Segments are the territories corresponding to both the primary branches of the splenic artery (primary segments) and the polar arteries (polar segments). In 92.82% of the cases there are two primary segments and in 7.18% three primary segments. Associated to these, in 29.28% there is a superior polar segment, in 44.75% an inferior polar segment, and in 10.49% both superior and inferior polar segments are present. Thus, the number of segments varies from two to five. Occasionally, two or three inferior polar segments can be present. Subsegments are the territories corresponding to the extrasplenic subdivisions of the primary branches and the polar arteries. According to the number of arterial subdivisions, the subsegments can be of second, third, fourth, fifth or sixth order. The last branches of the splenic artery (penetrating arteries) are all subsegmentary in nature and supply hilar or polar subsegments. Anastomoses between extrasplenic branches of the splenic artery were observed in 19.89% of the cases. Sometimes, thin anastomotic bridges could be observed between arterial splenic compartments.

[Arterial blood supply of the upper pole of the spleen in main artery occlusion]

Vascular dissections of the splenic-, gastric- and pancreatic region, corrosion specimens of the upper splenic pole and arteriographies of the superior polar artery after trunc lienal artery occlusion in human served as an answer to the question of the arterial supply of the upper lienal pole. The present results point out the importance of the upper polar artery and the existence of intersegmental anastomosis. The latter gain importance in case of trunc splenic artery occlusion. The short gastric vessels are of subordinate value.

[Functional anatomy of the human spleen]

For surgery of the spleen the variability of the splenic artery and the lack of major intrasplenic anastomoses are important. The latter enables segmental resections. Lymphocyte subsets, macrophages and dendritic cells, which are necessary for immune responses of the spleen, show a specific localisation in the splenic compartments. The spleen is the most important organ for lymphocyte recirculation. Intracellular particles and membrane defects of red cells are removed from the erythrocytes as they pass through the pulp cords of the red pulp. A knowledge of the functional anatomy of the spleen helps to interpret the effects of splenectomy and splenic transplantation.

Anatomic characteristics of the pancreatic arteries. Radiologic observations and their clinical significance

Major complications of operations that remove or transplant a portion of the pancreas appear related to the arterial blood supply. Hemorrhage is a result of inadequate control of the appropriate principal arteries, and ischemia and its consequences occur when the arterial blood supply to portions of preserved pancreas has been inadvertently interrupted. This radiographic study of the principal arteries of the pancreas demonstrates that these vessels may be visualized and their origin determined in the vast majority of cases. The application of data gained from preoperative arteriography can potentially decrease the incidence of major complications associated with removal or transplantation of portions of the pancreas.

Venous segmentation of the human spleen. A corrosion cast study

Partial splenectomy is possible because of vascular segmentation of this organ. Although interest has focused upon arterial supply, this study shows similar venous segmentation. Emerging splenic veins in excess of 1.7 mm at the hilum of ten unembalmed spleens were injected with different colors of modified Batson's compound, and corrosion casts were made. In four additional spleens, alternate veins and arteries were injected. Ten spleens had two primary segments (lobes); four had three lobes. Half of these lobes were further subdivided into two segments. Lobes and segments extended through the full thickness of the spleen. Relatively avascular interlobar planes were always approximately perpendicular to the longitudinal axis of the spleen, whereas intersegmental planes were more variable. Both interlobar and intersegmental planes frequently corresponded to marginal notches. When arteries and veins were both injected, similar segmentation was noted. Venous drainage did not overlap arterial segments, indicating that veins are intrasegmental. These results indicate that the surgical unit of the spleen can be based upon surgically accessible vessels at the hilum of the spleen. Identification of segmental vessels could permit salvage of a larger amount of splenic tissue than would be possible if only lobar vessels were identified.