Introduction

Despite the multiple advances in the different fields of human biology, and when it seems that today human beings can be represented by an infinitesimal trace of their genome, there are still regions of human anatomy that cause controversies, disagreements and misunderstandings. This has been happening for decades or even centuries with the anatomy of the gastroesophageal junction, its constituent elements and those which surround it. Therefore, if the anatomy is ignored or misunderstood, it is harder to understand how things work normally or when we deal with its dysfunction and disease and, finally, the consequences, for better or worse, of our interventions.

In this article we will review some embryologic, anatomic and physiologic aspects that seem relevant and of interest for the surgeon who deals with this region.

The diaphragm

The diaphragm is an elliptical muscle made of a thin, flat musculotendinous structure which separates the thoracic cavity from the abdominal wall. It has a central tendon with two domes of striated muscle that rise to each side. The diaphragm is attached to the xiphoid process at the front, laterally to the inner surface of the costal cartilages of the seventh to twelfth ribs, and to the transverse processes of vertebra L1, to the bodies of the upper three lumbar vertebrae and also the intervening intervertebral disc at the back.

Embryologically, the diaphragm originates from four structures: the septum transversum, the pleuro-peritoneal membranes, the dorsal mesentery of the esophagus, and the mesoderm of the body wall. During the fourth week of development, the mesodermal septum transversum incompletely separates the pericardial and peritoneal cavities and begins to emigrate following the caudal direction of the embryo's growth and pulling the phrenic nerve branches along with it. At the eighth week, the septum transversum reaches its thoracolumbar location, but does not close the entire area because the communication remains through the pleuro-peritoneal canals; in this period, the right and left pleuro-peritoneal membranes fuse with the septum transversum and with the dorsal mesentery of the esophagus, closing the communication between the pleural and abdominal cavities. The central tendon and the more central muscles originate from the septum transversum. The posterior lateral muscles originate from the wall. The dorsal mesentery of the esophagus will provide muscle fibers that will develop the crura. This diverse origin determines the costal and crural diaphragm.

Blood flow is supplied by the pericardiophrenic vessels, internal thoracic artery and branches of the thoracic aorta and abdominal aorta.

The right and left phrenic nerves provide motor innervation to the diaphragm and to the crural region. It is important to know the distribution of the phrenic nerves so as not to lesion them when the diaphragm is sectioned. Incisions can be performed at the periphery of the diaphragm near the costal attachment or from the anterior portion of the esophageal hiatus toward the central tendon to advance into the mediastinum during the Pinotti's maneuver^1-4.

The hiatus

Many structures pass through the diaphragm, but the most important are the aorta, the vena cava and the esophagus. Each structure has its own opening. The aortic hiatus is located at the level of vertebra T12 or L1 and is bounded posteriorly by the body of the vertebrae, laterally by the diaphragmatic crura and anteriorly by the median arcuate ligament which joins both crura. The aorta, thoracic duct, azygos vein and hemiazygos vein pass through this hiatus.

The vena caval foramen is situated at the level of vertebra T8 and placed in the central tendon to the right of the midline so that its margins are tendinous. It transmits the inferior vena cava, the right phrenic nerve and some lymphatic vessels.

The esophageal hiatus is situated at the level of vertebra T10 and placed in front and a little to the left of the aorta. As opposed to the other openings, the esophageal hiatus is completely made up of muscle. According to some authors, there are different types of arrangements of the right and left crus in the formation of the esophageal hiatus, but in 60% of cases the right crus divides into two bundles to encircle the esophagus and the left crus joins the left branch of the right crus, so that the left crus usually looks firmer and thicker (Fig. 1). The esophageal hiatus is traversed by the vagus nerves, branches of the left gastric artery and veins and of the left phrenic nerve.

Figure 1 sophageal hiatus. The cura are observed with the esophagus emerging. 

The crural diaphragm has a minor respiratory role compared to the costal diaphragm, but is more involved in gastroesophageal functions, such as swallowing, vomiting, and preventing reflux.

The crural part relaxes when the esophagus distends with a food bolus for easy passage from the esophagus to the stomach in coordination with the esophageal peristalsis, thus allowing food transition across the hiatus.

Emesis requires a complex coordination of the respiratory muscles, abdominal muscles, diaphragm and gastroesophageal junction. During the expulsive phase, the costal diaphragm keeps contracting to produce positive abdominal pressure, whereas the crural diaphragm relaxes to allow the gastric contents to be ejected upwards^1,2. Some authors consider the crural diaphragm of great importance in the prevention of gastroesophageal reflux, while others think of it as a second sphincter⁵. It seems reasonable to think that the crural diaphragm acts as external support due to its muscular structure and location within the gastroesophageal junction, preventing reflux when the abdominal pressure increases; however, in our opinion, it would not seem to have a determining anti-reflux role under normal conditions of rest or during the interdigestive period.

The phrenoesophageal ligament

The abdominal esophagus is the segment 2 to 4 cm long of the esophageal tube, which can be exposed and encircled, located between the diaphragmatic hiatus and the gastroesophageal junction. The esophagus is characterized by the lack of serosa and the longitudinal fibers of its external layer. The cardia is the point where the esophagus connects to the stomach and is recognized as the oblique boundary between the esophagus and the gastric serosa that extends from the angle of His to the lesser curvature, proximal to the cardio-tuberous vein.

The diaphragmatic hiatus, the abdominal esophagus and the cardia are hidden because they are covered by a fibroelastic membrane called the prenoesophageal ligament. In the abdomen, this structure is covered by the peritoneum (Fig. 2).

Figure 2 Phrenoesophageal ligament covered by peritoneum containing fat pad and hiding the gastroesophageal junction. 

The phrenoesophageal ligament is made up of fibroblasts, collagen and elastic fibers and appears to arise from both the endothoracic fascia and the transversalis fascia, which fuse at the level of the hiatus; an upper leaf is attached to the esophageal wall, 2 to 4 cm proximal to the hiatus, while the lower leaf is attached to the serosa of the gastric fundus. It usually has a fat pad within its thickness that contains the vagus nerves (Fig. 3).

Figure 3 Schematic representation showing the architecture of the phre noesofageal ligament and its relation with the hiatus and the ele ments of the gastroesophageal junction 

This phrenoesophageal ligament provides a seal between the thoracic and abdominal cavities at the level of the hiatus and is also an attachment of the esophagus to the hiatus, yet flexible enough to move during swallowing.

The phrenoesofageal ligament can be approached from the lesser curvature or from the angle of His, which seems more convenient, to clear the gastroesophageal junction, and from there it can be dissected medially, releasing the left crus and protecting the anterior vagus nerve.

Some authors state that the phrenoesophageal ligament is an important anti-reflux barrier and probably has an association with hiatal hernias^6,7.

Lower esophageal sphincter or gastroesophageal sphincter

Few structures in human anatomy have been more discussed and still remain matter of debate than the existence of an anatomic sphincter at the gastroesophageal junction (GEJ)8, even though studies using esophageal manometry have demonstrated a sphincteric mechanism in the region^9,10 and which is otherwise the main barrier to gastroesophageal reflux¹¹. Until recently, the great problem has been to demonstrate an anatomic structure in the distal esophagus that coincides with the classic, and arbitrary concept of sphincter, conceived as circular or ring like bands of muscle fibers, that can easily be discerned by palpation and separated from the adjacent muscles by connective tissue septa¹². Such a structure has never been demonstrated at the GEJ although some authors over the course of history have believed or imagined that they have found it⁸. The absence of such a structure led to support the lack of an anatomic sphincter at that level, but the paper by Fike et al. in 1956⁹, which used manometry to confirm the existence of a sphincter at the level of the GEJ, made specialists propose the unusual and unprecedented interpretation, still prevalent in our days, of the presence of a physiologic yet not anatomic sphincter. In 1977 Winans further challenged the classic concept of the circular sphincter, and of "physiologic" sphincter, when he demonstrated the existence of a manometric asymmetry of the lower esophageal sphincter¹³.

In 1979, the anatomic studies of the GEJ performed by Liebermann et al.¹⁴ solved the problem by demonstrating that the lower esophageal sphincter is not a muscular ring but rather two muscle bundles that encircle the lumen: the clasp fibers at the lesser curvature and the oblique sling fibers at the greater curvature. The description indicates a particular arrangement and an increase in the amount of fibers of the internal muscle layer of the GEJ. The fibers of the inner muscular layer do not form a ring or circular muscle around the entire perimeter of the cardia; instead, they form a layer of semicircular fibers or clasps oriented transversely, embracing the lesser curvature at the GEJ. These clasps fibers are inserted firmly into the submucous connective tissue at the margin of contact with the oblique fibers. The oblique sling fibers, which surround the greater curvature (Fig. 4) are a muscular bundle of 3 cm width and cover an area that starts 1.5 cm above the angle of His and ascend forming part of the distal end of the esophagus, surrounding the GEJ as a stole. Both arms (anterior and posterior) run parallel to the lesser curvature in direction to the antrum. The end of the clasp fibers meets at almost a right angle the lateral margin of the sling fibers.

Figure 4 Arrangement of semicircular “clasp” fibers at the lesser curvature and oblique “sling” fibers at the greater curvature. These muscle bands constitute the anatomic structure of the gastroesopahgeal sphincter. There are no fibers surrounding the perimeter of the gastoresopha geal joint¹⁵. 

Therefore, the LES is not an annular sphincter, but rather made up of two muscle bundles, which are acting complementary to close the lumen: the clasp fibers and the oblique or cling muscular fibers. Several studies by our team confirm these findings and definitively rule out the existence of a muscular ring¹⁵.

Understanding the anatomy of the lower esophageal sphincter with its particular muscular structure, eliminating the concept of an annular or only physiologic sphincter is very important, because it allows understanding why surgical actions, not only on the cardia but also on the gastric body, can impact on the sphincteric action. The site of Zaijer's cardiomyotomy, and currently peroral endoscopic myotomy (POEM) in achalasia is not irrelevant¹⁶ and distal partial gastrectomy or sleeve gastrectomy can seriously impact and compromise sphincteric competence¹⁷.