The Oesophagus: A Muscular Tube With Four Dangerous Narrowings
Most people imagine the gullet as a drainpipe — food falls down it and gravity does the work. It does not. Swallow a mouthful while standing on your head and it still arrives in your stomach, because the oesophagus is a 25 cm muscular pump that grips a bolus and drives it uphill in a travelling wave. And along that pump there are four places where the tube is naturally squeezed narrow. Those four points are not trivia: they decide where a swallowed coin stops, where a swallowed corrosive burns deepest, where an endoscope meets resistance, and where a tumour will sit quietly until it is too late. Learn the tube, and you can predict the disaster before it happens.
A man of forty is eating grilled fish with his family. He laughs mid-swallow, and something sharp catches high in his throat — a sensation he can point to precisely, just above the notch at the top of his breastbone. He drinks water; it stays. He swallows bread; it stays. By the time he reaches the emergency department three hours later he is drooling slightly and refusing to swallow at all, because every attempt drives the point of a fish bone into the same spot. The endoscopist passes the scope, watches the centimetre markings on its shaft slide past the incisor teeth, and at fifteen centimetres — before the chest has even begun — the lumen closes to a slit and the bone is there, wedged crosswise. It is the narrowest point in the entire gastrointestinal tract, and it is where such things almost always stop. Two minutes with a pair of forceps and he goes home. The whole encounter was decided by a ring of muscle he has never heard of.
From the cricoid to the cardia: the course
One tube, three countries: the neck, the chest and a very short stay in the abdomen. The oesophagus begins where the pharynx ends, at the lower border of the cricoid cartilage — level with the sixth cervical vertebra, C6 — and runs down for roughly 25 centimetres to open into the cardia of the stomach at the level of T11. It is conventionally divided into three parts: a cervical part of about 5 cm, a thoracic part of about 18 cm, and an abdominal part of only 1 to 2 cm. In the neck it lies slightly to the left of the midline; it returns towards the midline around T5, then swings left again as it approaches the diaphragm. It descends first through the superior and then the posterior mediastinum, described in the mediastinum, hugging the vertebral bodies with the trachea and then the heart in front of it. Finally it pierces the muscular right crus of the diaphragm at **T10** — the oesophageal hiatus — accompanied by the anterior and posterior vagal trunks and the oesophageal branches of the left gastric vessels. Remember the openings in order — **T8 for the inferior vena cava, T10 for the oesophagus with the vagi, T12 for the aorta with the thoracic duct and azygos vein** — and note that the aorta does not pierce the diaphragm at all but passes behind it, which is why an aortic hiatal hernia is not a thing while an oesophageal one is common. Note too that the oesophagus is the most posterior of the thoracic viscera, pressed against the vertebral column: that is why it is approached surgically from behind or from the left, and why a probe swallowed into it can look forwards at the heart.
The four constrictions — and the centimetres an endoscopist counts
Measure everything from the incisor teeth, and the tube becomes a map. The oesophagus is not a uniform cylinder. Four structures press on it along its length, and each narrowing has a distance from the incisor teeth that every endoscopist knows by heart. **(1) The cricopharyngeal sphincter at about 15 cm** — the upper oesophageal sphincter, formed by the cricopharyngeus part of the inferior constrictor of the pharynx. This is the narrowest point in the whole gastrointestinal tract, and it is where swallowed foreign bodies — coins, fish bones, dentures, a lump of meat — most often lodge. **(2) The crossing of the arch of the aorta at about 22 cm**, where the arch described in the aorta and great vessels sweeps over and indents the left side of the tube. **(3) The crossing of the left main bronchus at about 27 cm**, where the bronchus of the bronchial tree crosses in front of it. These two are so close that they are often described together as a single broncho-aortic constriction. **(4) The diaphragmatic hiatus at about 40 cm**, where the right crus grips the tube on its way into the abdomen.
Why insist on those numbers? Because everything slow and dangerous happens at a narrowing. A swallowed corrosive — a child drinking drain cleaner, an adult in a suicide attempt — sits longest where the lumen is tightest, so the burn is deepest there and the fibrous stricture that follows weeks later forms at exactly those levels. An endoscope meets resistance at those levels, and the temptation to push is exactly how an iatrogenic perforation happens, most often just above the cricopharyngeus. Tumours grow at those levels and obstruct earlier there than elsewhere. And a dilated structure in front of the tube can imitate a constriction: a hugely enlarged left atrium in mitral stenosis, or an aneurysm of the aortic arch, can press on the oesophagus and cause dysphagia with no oesophageal disease at all.
Picture a rubber hose threaded down the back of a crowded engine bay. Nothing about the hose itself is weak — but it has to squeeze past a bracket at the top, duck under a hot pipe, slip behind a belt, and finally pass through a grommet in the bulkhead. Four places where the neighbours pinch it. Now ask a mechanic where that hose will eventually crack, where grit will collect, where a blockage will form. He will not answer "anywhere"; he will name the four pinch points, because that is where the hose spends its life under pressure. The oesophagus is that hose, and its neighbours are the cricoid ring, the arch of the aorta, the left main bronchus and the diaphragm.
The wall: striated above, smooth below
The beginning of a swallow is your decision. The rest of it is not. The wall has the usual four layers minus one. The mucosa is lined with **non-keratinised stratified squamous epithelium** — a tough, abrasion-resistant surface built for hot tea, bread crusts and bone fragments, quite unlike the delicate columnar lining of the stomach. Beneath it lies a loose submucosa carrying mucous glands, a rich vascular plexus and, critically, a dense longitudinal lymphatic network. Outside that is the muscularis externa, in the usual inner circular and outer longitudinal arrangement — but with a twist found nowhere else: **the upper third is skeletal (striated) muscle, the middle third is a mixture, and the lower third is entirely smooth muscle**. That is why the first instant of swallowing is voluntary and everything after it is not. Once the bolus is past the upper sphincter, the wave takes over and cannot be recalled.
Two more facts about the wall carry enormous clinical weight. The first is the **Z-line**: at the gastro-oesophageal junction the squamous lining stops abruptly and gives way to gastric columnar epithelium in a ragged, visible circle. Chronic acid reflux can push that line upwards as the squamous epithelium is replaced by columnar epithelium with goblet cells — Barrett's metaplasia — and it is on that metaplastic mucosa that adenocarcinoma of the lower oesophagus arises. Squamous cell carcinoma, by contrast, is a disease of the upper and middle thirds and of smoking and alcohol. The second fact is a negative one: **the oesophagus has no serosa** over almost its whole length — only the short abdominal segment gets a peritoneal covering. There is no tough outer envelope to contain a tumour, so oesophageal cancer breaches its wall and spreads early; and there is nothing to seal a tear, so a perforation empties straight into the mediastinum.
The missing serosa explains two exam favourites at once. Ask why oesophageal cancer has such a poor prognosis and the usual answer — "it presents late" — is only half of it; the other half is anatomical. Without a serosal barrier, and with a submucosal lymphatic plexus that runs longitudinally rather than segmentally, tumour cells travel up and down inside the wall far beyond the visible edge, which is why surgeons take generous margins and why skip lesions occur. And ask why a ruptured oesophagus is so lethal and the answer is the same absence: gastric contents pour into the mediastinum, which has no barriers of its own, and mediastinitis follows within hours. One missing layer, two very different catastrophes.
The lower sphincter that is not a sphincter
Dissect for it and you will not find a ring. It is a conspiracy of four mechanisms. The lower oesophageal sphincter is physiological, not a discrete anatomical thickening like the pylorus. What keeps acid in the stomach is a combination of four arrangements working together. First, the **crural sling**: the muscular right crus of the diaphragm wraps around the hiatus and pinches the tube every time you inhale or strain, an external clamp that tightens exactly when abdominal pressure rises. Second, the **angle of His** — the acute angle at which the oesophagus enters the stomach — so that a distending gastric fundus presses a flap of mucosa against the opening like a one-way valve. Third, the short **intra-abdominal segment**: because that last centimetre or two lies below the diaphragm, the same abdominal pressure that would push contents upwards also squeezes the tube shut. Fourth, the **mucosal rosette**, loose folds of mucosa that plug the lumen at rest. Add a tonic contraction of the circular smooth muscle, relaxed on cue by the vagus ahead of an arriving bolus, and the seal is complete.
Break any part of that arrangement and you get disease. In a **sliding hiatus hernia** the gastro-oesophageal junction slides up through the hiatus into the chest: the intra-abdominal segment is lost, the angle of His is flattened, and the crural sling no longer grips the right place — so reflux follows, and with it heartburn, nocturnal cough and, over years, Barrett's change. In a **rolling (para-oesophageal) hernia** the junction stays put while the fundus herniates alongside it, so reflux is less of a problem but strangulation is a real one. And at the opposite extreme is **achalasia**: degeneration of the ganglion cells of the myenteric (Auerbach's) plexus in the lower oesophagus means the sphincter never relaxes and peristalsis is lost, so the tube above it dilates enormously and a barium swallow shows the classic smooth, tapering "bird's beak".
Neighbours: what lies in front, behind and beside
In the neck, the trachea lies directly in front, and in the groove between the two runs the recurrent laryngeal nerve on each side — which is why oesophageal surgery and thyroid surgery share the same feared complication of a hoarse voice. The lobes of the thyroid and the carotid sheaths lie laterally, the prevertebral fascia and longus colli behind. In the chest, the trachea and then the left main bronchus lie in front above; below that the pericardium and, immediately in front of the lower thoracic oesophagus, the **left atrium**. Behind it are the vertebral bodies, the thoracic duct, the azygos vein, the right posterior intercostal arteries and, lower down, the descending thoracic aorta. On the left lie the arch of the aorta, the left subclavian artery and the left pleura; on the right, the azygos vein arching forward and the right pleura, which comes so close that a right-sided pleural effusion is a classic sign of oesophageal perforation.
One relation above all others is worth memorising. The left atrium sits immediately anterior to the lower thoracic oesophagus, separated from it by little more than pericardium and a film of connective tissue. That single fact created an entire diagnostic technique: transoesophageal echocardiography, in which an ultrasound probe swallowed into the oesophagus looks forward at the left atrium with no lung or rib in the way, and can see a clot in the left atrial appendage or a vegetation on a valve that a probe on the chest wall would miss entirely. It works in reverse too. A left atrium hugely dilated by long-standing mitral stenosis presses backwards on the oesophagus and produces dysphagia — a patient whose difficulty swallowing is a cardiac sign. In the same anatomical neighbourhood, an enlarged left atrium can stretch the left recurrent laryngeal nerve under the aortic arch and cause hoarseness, a link explored further in the nerves of the thorax.
Nerves, arteries and veins — supplied in thirds
The nerve supply comes from the vagus and the sympathetic trunks, which weave together on the surface of the tube as the **oesophageal plexus**. The striated upper part is driven by branches of the recurrent laryngeal nerves; the smooth lower part receives parasympathetic preganglionic vagal fibres that synapse in the myenteric plexus in the wall — the same ganglion cells that die in achalasia. As the tube rotates during development, the left vagus comes to lie on the front and the right vagus on the back, so the trunks that pass through the hiatus at T10 are named **anterior (mostly left) and posterior (mostly right) vagal trunks**. Sympathetic fibres from the thoracic chain and greater splanchnic nerves are largely vasomotor and carry pain; their route explains why oesophageal pain — spasm, reflux, rupture — is felt centrally behind the sternum and can be indistinguishable from cardiac pain.
Three parts, three sets of vessels — and at the bottom, a hazard. Arterial supply follows the three parts exactly. The cervical oesophagus is fed by the **inferior thyroid artery**; the thoracic oesophagus by direct **oesophageal branches of the descending thoracic aorta** together with branches of the bronchial arteries; and the abdominal oesophagus by the **left gastric artery** (from the coeliac trunk) and the **left inferior phrenic artery**. Venous drainage mirrors it: the upper part drains to the inferior thyroid veins, the middle part to the azygos and hemiazygos system, and the lower part to the **left gastric vein** — which is a tributary of the portal vein. Lymphatic drainage runs in both directions along that longitudinal submucosal plexus: the upper third to deep cervical nodes, the middle to superior and posterior mediastinal nodes, the lower to left gastric and coeliac nodes. Because the plexus is continuous, a lower tumour can seed nodes in the neck and an upper one can seed nodes in the abdomen — the reason nodal staging of this tube is so unforgiving.
Now put the last two facts side by side. At the lower end of the oesophagus, veins draining into the **left gastric vein (portal)** meet veins draining into the **oesophageal tributaries of the azygos (systemic)** in the submucosa. In health this portosystemic anastomosis is a trivial crossroads carrying almost nothing. In cirrhosis, when portal pressure rises and blood cannot get through the liver, it becomes a major detour: portal blood is forced backwards up the left gastric vein and out through those thin submucosal channels, which dilate into **oesophageal varices** — thin-walled, high-pressure veins lying just under a squamous mucosa that is rubbed by every swallow. When one ruptures the bleeding is torrential and the patient vomits frank blood. Nowhere else in the body does a liver disease cause a fatal haemorrhage from the gullet, and the reason is a venous watershed a few centimetres long.
The coin: a three-year-old swallows a coin; the chest X-ray shows it face-on at the level of the clavicles, held at the cricopharyngeus at 15 cm, and it is removed endoscopically. The tumour: a man of sixty-five reports six months of difficulty swallowing meat, then bread, then finally liquids — progressive dysphagia for solids before liquids, with weight loss, the classic story of a narrowing lumen rather than a motility disorder. The bird's beak: a woman of thirty has trouble with both solids AND liquids from the outset, regurgitates undigested food at night and has a barium swallow showing a dilated oesophagus tapering to a smooth point — achalasia, not cancer. The heartburn: a lorry driver with years of retrosternal burning after meals, worse lying flat, whose endoscopy shows the Z-line displaced upwards with salmon-coloured tongues of columnar mucosa — Barrett's, now needing surveillance. The rupture: a man vomits violently after a heavy meal and develops sudden severe chest pain, breathlessness and crackling under the skin of his neck — Boerhaave's syndrome, a full-thickness tear of the lower left posterolateral wall, spilling into the mediastinum.
The pattern of dysphagia tells you the mechanism before any test does. Difficulty with solids that later progresses to liquids means the lumen is getting narrower — a stricture or a tumour, and the shorter the history the more sinister. Difficulty with solids AND liquids from the very beginning, often with regurgitation of old food, means the muscle or its nerves have failed — achalasia or another motility disorder — because a paralysed tube handles water no better than steak. And difficulty in the first second of the swallow, with coughing, choking or nasal regurgitation, is not oesophageal at all; it is oropharyngeal, and it points at the striated muscle and its cranial nerves — a stroke, myasthenia, or a failure of the cricopharyngeus to relax.
- The oesophagus runs ~25 cm from the lower border of the cricoid cartilage (C6) to the cardia of the stomach (T11), in three parts: cervical (~5 cm), thoracic (~18 cm) and abdominal (1–2 cm).
- It descends through the superior and posterior mediastinum and pierces the diaphragm at T10 with the anterior and posterior vagal trunks — remember T8 IVC, T10 oesophagus, T12 aorta.
- Four constrictions, measured from the incisors: cricopharyngeus 15 cm (narrowest point of the whole GI tract), aortic arch 22 cm, left main bronchus 27 cm, diaphragmatic hiatus 40 cm.
- Those four points are where foreign bodies lodge, where corrosive strictures form, where an endoscope meets resistance and is perforated, and where tumours obstruct early.
- Mucosa = non-keratinised stratified squamous, changing abruptly to gastric columnar at the Z-line; acid reflux drives Barrett's metaplasia there and adenocarcinoma on top of it.
- Muscularis externa: striated in the upper third, mixed in the middle, smooth in the lower third — so only the start of a swallow is voluntary.
- There is NO serosa except over the short abdominal part — a major reason oesophageal cancer spreads early and a perforation leaks straight into the mediastinum.
- The lower oesophageal sphincter is physiological: the crural sling of the right crus, the angle of His, the compressed intra-abdominal segment and the mucosal rosette.
- Relations: trachea and recurrent laryngeal nerve in front above; descending aorta, thoracic duct and azygos vein behind; the LEFT ATRIUM immediately in front of the lower thoracic part.
- Nerves: the oesophageal plexus from both vagi (left → anterior trunk, right → posterior after rotation) plus sympathetic fibres; myenteric ganglion cells fail in achalasia.
- Arteries by thirds: inferior thyroid; oesophageal branches of the thoracic aorta plus bronchial arteries; left gastric and left inferior phrenic.
- The lower end is a portosystemic anastomosis — left gastric (portal) meeting oesophageal tributaries of the azygos (systemic) — so portal hypertension produces oesophageal varices.
- Treating the lower oesophageal sphincter as an anatomical ring like the pylorus. Dissection shows no discrete thickening — the seal is a physiological conspiracy of the crural sling, the angle of His, the intra-abdominal segment and the mucosal rosette, which is exactly why a hiatus hernia breaks it.
- Putting the oesophageal hiatus at T12. The oesophagus passes at T10 with the vagi; T12 is the aortic hiatus (with the thoracic duct and azygos), and T8 is the caval opening.
- Confusing inhaled with swallowed foreign bodies. An INHALED object goes right, because the right main bronchus is wider, shorter and more vertical; a SWALLOWED object stops at the cricopharyngeus at 15 cm, and it is the LEFT main bronchus that indents the oesophagus at 27 cm.
During an upper gastrointestinal endoscopy the operator feels resistance and sees a pulsatile extrinsic indentation on the anterior wall of the oesophagus at 27 cm from the incisor teeth. Which structure is responsible for this normal constriction?
- A ~25 cm muscular tube from the cricoid cartilage (C6) through the superior and posterior mediastinum, piercing the diaphragm at T10 with the vagal trunks, to the cardia at T11; three parts — cervical, thoracic, abdominal.
- Four constrictions from the incisors — cricopharyngeus 15 cm (narrowest in the GI tract), aortic arch 22 cm, left main bronchus 27 cm, diaphragmatic hiatus 40 cm — and they dictate where foreign bodies lodge, strictures form and tumours obstruct.
- Stratified squamous mucosa changing at the Z-line (Barrett's, adenocarcinoma), muscle striated above and smooth below, and no serosa — so cancer spreads early and perforation floods the mediastinum.
- The lower sphincter is physiological (crural sling, angle of His, intra-abdominal segment, mucosal rosette); the lower end is a portosystemic watershed whose failure in portal hypertension gives oesophageal varices.
- Drake RL, Vogl AW, Mitchell AWM. Gray's Anatomy for Students — Thorax: the oesophagus and posterior mediastinum.
- Moore KL, Dalley AF, Agur AMR. Clinically Oriented Anatomy — The oesophagus: constrictions, relations and the oesophageal hiatus.
- Netter FH. Atlas of Human Anatomy — Oesophagus in situ; arteries and veins of the oesophagus.
- Last RJ. Last's Anatomy: Regional and Applied — The mediastinum and the oesophagus.
- Snell RS. Clinical Anatomy by Regions — Oesophageal constrictions, varices and perforation.
- TeachMeAnatomy — The Oesophagus; The Gastro-oesophageal Junction.
