ORIGINAL ARTICLES

Gross and Microscopical Blood Supply of the Trachea

John R. Salassa, B.S., Bruce W. Pearson, M.D., a n d W. Spencer Payne, M.D.
ABSTRACT Twenty-one human tracheal specimens were perfused and dissected, 10 with conventional techniques and 11 with clearing and microdissection techniques. The lateral pedicles of the trachea and esophagus induct vessels from the inferior thyroid, subclavian, supreme intercostal, internal thoracic, innominate, and superior and middle bronchial arteries. These vessels are interconnected along the lateral surface of the trachea by an important longitudinal vascular anastomosis. From the 2 lateral longitudinal anastomoses the lateral and anterior tracheal walls receive their blood supply through transverse segmental vessels that run in the soft tissues between the cartilages. These transverse vessels interconnect the longitudinal anastomoses across the midline and feed the submucosal capillary network that arborizes richly beneath the endotracheal mucosa. The tracheal cartilages receive nourishment from the capillary bed applied to their internal surface. The esophageal arteries and their subdivisions that supply the posterior membranous wall of the trachea contribute almost nothing to the circulation of the cartilaginous walls.

human tracheal blood supply. Tracheal vessels are mentioned only briefly in Swigart and associates description [71 of the esophageal blood supply and in Michalewskis study [4] of the bronchial arteries. Our study supplies new information about the blood supply to the human trachea, including the origin of the tracheal arteries, their segmental distribution, gross anastomoses, and the tracheal microscopical blood supply.

Methods

Eleven fresh laryngotracheal postmortem specimens were suspended in saline at 50C and perfused with a bright red colloidal suspension of radiopaque medium (Colorpaque) in 5% gelatin through the inferior and superior thyroid arteries. After cooling and fixation, the specimens were dehydrated and cleared in toluene. They were sectioned in the midsagittal plane and immersed in silicone fluid (Dow Corning 710) for study with the aid of a Zeiss operating microscope. The immersed specimens were disToday, tracheal resective and reconstructive sected from the luminal side and from the exterprocedures have become common enough that a nal tracheal surface. Twelve hemispecimens description of the human tracheal blood supply, were subsequently embedded in celloidin, semore detailed than those given in standard rially thick-sectioned at 2 mm (some transanatomy texts [l, 2, 81, is needed. Faced with versely, others coronally), and studied under this problem, Miura and Grillo [51, using higher-power magnification to allow further colroentgenograms of specimens injected with ra- lation of the pathways taken by vessels within diopaque medium, studied the contribution the tracheal walls. Ten additional laryngotracheal postmortem of the inferior thyroid artery to the cervical trachea. They found that the inferior thyroid specimens were removed with a sufficient volartery supplied the cervical portion of the ume of surrounding tissue to permit study of the trachea, usually through three branches, the vessels approaching the trachea. Each specimen lowest of which is dominant in most instances. included the larynx, trachea, roots of the main We found no other definitive studies of the bronchi, strap muscles, thyroid, pharynx, esophagus, great vessels, and superior mediasFrom the Mayo Medical School and Mayo Clinic and Mayo tinal soft tissues. The specimens were flushed Foundation, Rochester, MN. with embalming fluid, cannulated, immersed in Presented at the Thirteenth Annual Meeting of The Society gelatin, and injected through the subclavian and of Thoracic Surgeons, Jan 24-26, 1977, San Francisco, CA. superior thyroid arteries after the gelatin had Address reprint requests to Mr Salassa, c/o Section of Publi- firmed. After fixation, gross dissection was carcations, Mayo Clinic, 200 First St SW, Rochester, ried out. Suitable hemiinjection specimens were MN 55901.
100

101 Salassa, Pearson, and Payne: Blood Supply of the Trachea

obtained in 17 instances. Three hemispecimens were unsuitable because of technical deficiencies.

Results The trachea begins at the lower border of the cricoid cartilage, 5 cm above the jugular notch, and runs inferior and posterior, ending at the tracheal bifurcation, 6 cm below the jugular notch. In the anatomical position, a cervical portion (upper 5 cm) and a thoracic portion (lower 6 cm) of the trachea are recognized.
Blood Supply of the Cervical Trachea THE INFERIOR THYROID ARTERY. The cervical trachea received its blood supply primarily from branches of the inferior thyroid artery that run in a medial direction behind the lower part of the common carotid artery. In the most common pattern (9 of 17 inferior thyroid arteries studied), 3 branches arose at various distances along the length of the artery, passed posterior to the carotid sheath, and approached the tracheoesophageal column laterally (Fig 1). Each branch divided terminally into tracheal and esophageal branches. The tracheal branches traveled anterior or posterior to the recurrent laryngeal nerve or on both sides of it. The first of the tracheoesophageal branches arose from the proximal half of the inferior thyroid artery, supplied the lowest portion of the cervical trachea, and gave only minor or no contributions to the esophagus. The second branch of the inferior thyroid artery supplied the trachea between the first and third branches and gave substantial contributions to the esophagus. The third branch arose at the site where the inferior thyroid artery penetrates the posterior aspect of the thyroid gland. This branch supplied the uppermost ends of the trachea and the esophagus. In 6 of 17 inferior thyroid arteries studied, only 2 tracheoesophageal branches were found (Fig 2). This pattern was otherwise identical to the most common pattern previously mentioned. In 2 of the 17 inferior thyroid arteries, a single vessel supplied the upper cervical trachea. The lower cervical trachea was dependent upon anastomoses between this single vessel and the blood supply of the thoracic trachea.

Fig 1 . Left anterior v i e w of vessels supplying the trachea. In this specimen the lateral longitudinal anastomosis links branches o f t k e inferior thyroid, costocervical trunk, and bronchial arteries.

In 2 specimens a tracheoesophageal branch from the subclavian artery supplied the lower cervical trachea (see Fig 2). This branch arose from the subclavian artery proximal to and separate from the origin of the thyrocervical trunk. The distribution of this branch was similar to that of the first branch of the inferior thyroid artery (see Fig 1). In each instance the inferior thyroid artery gave rise to only 2 tracheoesophageal branches. THE SUPERIOR THYROID ARTERY. The superior thyroid artery gave no direct branches to the trachea. However, it anastornosed with the inferior thyroid artery in and around the thyroid gland and partially filled the bed supplied by the inferior thyroid artery in specimens in which the inferior thyroid artery was not perfused.
THE SUBCLAVIAN ARTERY.

Blood Supply of the Thoracic Trachea The blood supply of the thoracic trachea was more variable than that of the cervical trachea

102 The Annals of Thoracic Surgery Vol 24 No 2 August 1977

Innominatea

em

bronchus

Fig3. Posterior viezv of bronchial vessels supplying the carina. In this specimen the anterior branch of the superior bronchial artery and main trunk of the middle bronchial artery supply the lower thoracic trachea.

Fig2. Right anterior v i e w of vesselssupplying the trachea. In this specimen the lateral longitudinal anastomosis links branches from the inferior thyroid, thesubclavian, the internal thoracic, and the superior bronchial arteries.

and arose from a greater number of arteries. It was constant only at the extreme distal trachea, which was always supplied by the bronchial arteries. The rest of the blood supply derived from the innominate-subclavian system. Tracheal arteries arose from the supreme intercostal artery in 6 specimens, a subclavian artery in 5, the right internal thoracic (mammary) artery in 2, and the innominate artery in 3. The superior or midBRONCHIAL ARTERIES. dle (or both) bronchial arteries supplied the extreme distal end of the trachea in all specimens (see Figs 1, 2). The superior bronchial artery arose from the right posterior surface of the proximal descending aorta in line with the right intercostal arteries (Fig 3) and divided into 2 branches. One branch passed posteriorly, behind the esophagus, to the right bronchus. A second branch, which traveled anteriorly to the left side of the esophagus and trachea, frequently ascended a short distance and then passed an-

teriorly over the base of the left main bronchus to the carina. The anterior branch supplied the carinal and paratracheal nodes and anastomosed with the middle bronchial artery below and the thoracic tracheal vessels above. The middle bronchial artery arose just below the superior bronchial artery and ran anteriorly along the left side of the esophagus to the left bronchus. This artery traveled around the medial aspect of the left bronchus and, in 3 instances, formed a major anastomosis at the carina with the superior bronchial and higher thoracic tracheal vessels. In 1 specimen, in which the superior bronchial artery did not supply the trachea (absent anterior branch), the middle bronchial artery anastomosed directly with higher thoracic tracheal vessels. The inferior (left)bronchial artery, when present, was severed when specimens were taken from the cadavers; therefore, it was assumed that it reached the bronchi distal to the tissue block. An indirect contribution to the lower trachea, by way of anastomoses from the inferior bronchial artery, thus could not be excluded. SUPREME INTERCOSTAL ARTERY. Tracheoesophageal arteries arose from the right supreme intercostal artery in 4 specimens and from the left supreme intercostal artery in 2. In 5 specimens this branch arose from the proximal centimeter or the base of the supreme intercostal

103 Salassa, Pearson, and Payne: Blood Supply of the Trachea

artery (see Fig 1). In 1 specimen the right supreme intercostal artery gave 4 branches from its proximal 5 cm; the first 2 supplied the trachea, and the last 2, the esophagus. The branch or branches from the supreme intercostal artery descended obliquely in the lateral wall of the superior mediastinum, approached the tracheoesophageal column laterally, and divided into esophageal and tracheal branches. The tracheal branches passed anterior to, posterior to, or on both sides of the recurrent laryngeal nerve and, in lower segments, the right vagus. SUBCLAVIAN ARTERY. The subclavian artery gave rise to a single tracheal vessel in 3 specimens and to 2 tracheal vessels in 1 specimen. This artery supplied areas in the lower twothirds of the trachea (see Fig 2 ) . It arose from the subclavian artery before the origin of the vertebral artery, passed through the posterolateral superior mediastinum, and approached the tracheoesophageal column laterally. Distally, this artery gave off small branches to the esophagus before it entered the trachea. In 1 specimen no esophageal branches arose. The relationship to the recurrent laryngeal nerve was variable. RIGHT INTERNAL THORACIC ARTERY. The right internal thoracic (internal mammary) artery gave origin to tracheal branches in 2 specimens (see Fig 2 ) . These branches arose from the first 3 cm of the internal thoracic artery, passed through the superior mediastinal tissues, approached the tracheoesophageal column laterally, and entered the lateral wall of the lower trachea. Tracheal and esophageal divisions were present. This artery was present only in the absence of a tracheal branch from the ipsilateral supreme intercostal artery (see Fig 2 ) . INNOMINATE ARTERY. A branch to the middle and lower thoracic trachea arose from the posterior surface of the innominate artery in 3 specimens. This branch was small compared with other tracheal arteries. In 2 specimens it passed directly posterior and inferior to enter the right lateral trachea. In 1 specimen the vessel passed posterior and inferior and entered the left lateral trachea. In all specimens this artery supplied only the trachea.

OTHER ARTERIES. No tracheal branches were found to arise from the common carotid artery, the vertebral artery, or directly from the arch of the ascending aorta. There was no consistent difference between the blood supplies of the right and left sides of the trachea.

Microscopical Blood Supply In brief review, segmental tracheoesophageal vessels arose from the inferior thyroid arteries (1 to 3 branches), the supreme intercostals, the subclavians, the right internal thoracic, and the superior and middle bronchial arteries. Lateral to the tracheoesophageal groove these major vessels divided into primary tracheal and primary esophageal branches (Fig 4). The primary tracheal branches passed directly to the lateral tracheal wall, contacting it about 0.7 to 1.5 cm anterior to the tracheoesophageal groove. The primary esophageal branches frequently sent off smaller secondary tracheal twigs to the posterior tracheal wall before they reached the esophagus.
THE LATERAL LONGITUDINAL ANASTOMOSIS.

The primary tracheal arteries branched up and down over 3 or 4 interspaces on the lateral aspect of the trachea. In so doing, they connected with the primary tracheal arteries from above and below. This produced an important, sometimes irregular, but complete longitudinal tracheal anastomosis on the lateral wall of the trachea (see Figs 1, 2). This lateral longitudinal tracheal anastomosis was present in 14 of 17 gross hemisections studied. Its caliber was similar to that of the primary tracheal arteries it connected (approximately 1to 2 mm in diameter).
THE TRANSVERSE INTERCARTILAGINOUS AR-

Each soft tissue space between the tracheal cartilages received intercartilaginous arteries from the lateral longitudinal anastomoses. These arteries ran anteriorly, then medially to sink deeper and deeper into the tracheal wall (see Fig 4). By the time they reached the anterior midline to anastornose with the arteries from the opposite side, they had reached the submucosa. Smaller intercartilaginous arteries turned posteriorly from the lateral longitudinal tracheal anastomoses and ended at the junction of the cartilaginous and membranous tracheal walls. We were unable to demonstrate meaningTERIES.

104 The Annals of Thoracic Surgery Vol 24 No 2 August 1977

__

Coronal section of tracheal wall.. .

Fig4. Semischematicviezu of the tracheal microscopical blood supply. 7ransverse intercartilaginous arteries derived f r o m the lateral lo ngi t u din a1 anastomosis penetrate the soft tissues between each cartilage to supply a rich vascular network beneath the endotracheal mucosa.

ful linkage between these smaller posterior transverse intercartilaginous arteries and the secondary tracheal twigs that stemmed from the primary esophageal vessels (see Fig 4). Thus, in the soft tissues between each cartilage, a small transverse intercartilaginous artery was found. Often there were 2 such arteries running side by side. From these vessels the entire blood supply of the lateral and anterior tracheal walls was derived. The posterior tracheal wall was supplied by the secondary tracheal twigs from the primary esophageal vessels. Longitudinal anastomoses from one level to another in the muscular posterior tracheal wall were well developed. They

were intimately adherent to the posterior wall and spanned several segments. A rich and completely interanastomotic capillary bed was present throughout the entire endotracheal submucosa. The high-power microscopical appearance resembled a large expanse of chicken wire, the segmental or linear orientation of larger vessels being completely lost. In the anterior and lateral tracheal walls this submucosal capillary plexus was supplied by 4 or 5 short feeder vessels.from each transverse intercartilaginous artery. The submucosal capillary bed was intimately related to the endotracheal surface of the tracheal cartilages. Until this point, the tracheal cartilages had been bypassed by the blood vessels penetrating its walls. There were no important capillary plexi on the external surface of the tracheal cartilages. In the posterior membranous tracheal wall this submucosal capillary plexus was supplied by feeder vessels from the secondary tracheal twigs.

105 Salassa, Pearson, and Payne: Blood Supply of the Trachea

Comment Our findings support those of Miura and Grillo [5], that the cervical trachea is supplied primarily by branches of the inferior thyroid artery, and those of Michalewski [4], that the superior and middle bronchial arteries supply branches to the trachea and tracheobronchial lymph nodes. Swigart and colleagues [7], in their study of the blood supply to the esophagus, mentioned tracheal vessels only incidentally. They stated that the internal mammary, costocervical trunk, and subclavian arteries are each a source of an exceptional esophageal (tracheal) artery. We found that tracheoesophageal branches from these vessels are not ekceptional but are frequent, variable, multiple, and bilateral. We found no studies describing the microscopical blood supply of the human trachea. Sobin and associates description [6] of the tracheal microscopical blood supply in mammals does not apply to man. The fine but important longitudinal anastomoses along the lateral tracheal wall are better developed in cats and dogs than they are in man (examples of which-Fig %we have studied in our laboratory); consequently these animals may tolerate larger degrees of experimental tracheal mobilization.

Fig5. B l o o d s u p p l y o f n c l ~ ~ a r c d ctrachea. nt I J I t h e cat trachea the longitudinal arid transverse vessels arc exceptionally well defined: transverse intercartilaginous artery (A);cleared cartilage (B); lateral longitudinal anastomosis ( C ) .

Lateral Pedicles One of the aims of our study was to understand better the nature and contents of the lateral tracheal pedicles referred to by Grillo [3]. We found that the lateral pedicle consisted of an irregular sheet of connective tissue passing from the deep surface of the aorta and the innominate and subclavian arteries, and bearing branches from the thyrocervical, costocervical, subclavian, and internal thoracic arteries to the tracheoesophageal structures. Distally, this short lateral vascular pedicle splits to supply the trachea and esophagus. It extends the entire length of the trachea and bears 3 to 7 primary tracheal arteries. The recurrent laryngeal nerves and, in lower tracheal segments, the descending trunk of the right vagus lie enmeshed in the terminal ramifications of this pedicle. The esophageal and tracheal blood supplies are similar, but the trachea can be separated

from the esophagus without serious injury to the tracheal blood flow, because the tracheoesophageal blood supply in the lateral.pedicles splits and runs separately to each organ instead of to the esophagus and then to the trachea. The few twigs that do run such a course contribute to the posterior tracheal wall only. Their loss is tolerable because of the rich longitudinal blood supply in the posterior muscular wall. Collateral vessels from the larynx and bronchi, and perhaps the endotracheal submucosal plexus, appear to be sufficient to supply this longitudinal system of the posterior wall when the esophagus is freed completely from the trachea.

Arterial Anastomoses There are five areas of anastomosis in the tracheal blood supply: (1)the lateral longitudinal anastomosis links the segmental primary tracheal vessels along the lateral tracheal walls; (2) the transverse intercartilaginous arteries

106 The Annals of Thoracic Surgery Vol 24 No 2 August 1977

tween bronchial vessels and tracheal vessels descending from the subclavian system. Third, the trachea may be freed from the aorta by severing the bronchial arteries, but the circulation of the distal trachea will then depend on the integrity of the tracheal vessels from the subclavian system and the lateral longitudinal anastomoses adherent to the lateral tracheal wall. Fourth, if the cervical trachea is mobilized and not resected, the thyroid gland should, if possible, be left in contact with the trachea. The contribution from the most superior branch of the inferior thyroid artery, as well as numerous tiny vessels Cartilages passing directly from the thyroid gland to the The ability of the trachea to remain patent de- trachea, will be preserved. Fifth, the esophagus pends upon the integrity of the supporting may be separated from direct contact with the tracheal cartilages. Interestingly, the cartilagi- trachea without compromising tracheal vascunous portions of the trachea (lateral and anterior larity, provided the approach is not through walls) have a blood supply independent from the lateral pedicles. During segmental tracheal that of the muscular posterior tracheal wall. The resection the posttracheal space is accessible tracheal cartilages seem to depend on diffusion from above and below after the resected segfrom the submucosal capillary plexus on their ment has been removed. luminal aspect for their nutrition. There is no The variations among patients in their ability major capillary plexus on the external surface of to tolerate circumferential tracheal mobilization the trachea. Thus intraluminal compression of may depend on the integrity of the lateral lonthe tracheal mucosa leaves the underlying carti- gitudinal tracheal anastomosis. This system is lage essentially devoid of nutrition and highly variable among normal persons and may be susceptible to ischemic necrosis. compromised in patients whose trachea or paratracheal regions have been injured or scarred Operative Applications (that is, after transmural tracheal necrosis, For the surgeon who seeks to preserve the mediastinitis, or a previous paratracheal operatracheal blood supply during tracheal mobiliza- tion. However, should the longitudinal system tion procedures, we suggest that the following remain intact, the loss of a few of the primary five principles be considered in conjunction with tracheal arteries may be tolerable in view of the the location of the lesion and the extent of numerous and interanastomotic vessels feeding mobilization and resection required. First, and into the system. most important, is Grillos admonition to leave the lateral vascular pedicles intact. This broad Conclusions sheet could be buttonholed by blunt dissection Nutrition of the tracheal cartilages is dependent if necessary, but division, particularly close to on preservation of the lateral tracheal vascular the lateral tracheal wall, is inadvisable. Second, pedicles. The lateral pedicles are derived from anterior dissection may be done close to the branches of the inferior thyroid, supreme intercervical trachea without compromising the costal, subclavian, internal mammary, innomiblood supply. Lower in the thoracic trachea, nate, and bronchial arteries. These vessels interhowever, the blood supply is more anteriorly connect along the lateral tracheal wall to form an disposed (and the trachea more posteriorly di- important longitudinal tracheal anastomosis. rected). Close dissection of the paratracheal tis- From this anastomosis, the transverse intercarsues and the carinal and pretracheal lymph tilaginous arteries of each tracheal interspace nodes will sacrifice potential anastomoses be- area are derived. These transverse intercar-

unite the circulation of the right and left sides; (3) the tracheoesophageal arteries in the lateral pedicles connect the esophageal and tracheal circulations by virtue of their common derivation; (4) anastomoses in and around the thyroid gland supplement the circulation to the cervical trachea, with contributions from the superior thyroid artery; and (5) the rich anastomoses in the paracarinal nodes interconnect the circulation of the upper thoracic trachea with the superior, the middle, and possibly the inferior bronchial arteries.

107 Salassa, Pearson, and Payne: Blood Supply of the Trachea

tilaginous arteries in turn give ris,eto a continuous submucosal capillary plexus. Connections from the posterior tracheal wall to the cartilaginous wall exist only through the submucosal capillary plexus. The tracheal cartilages and their support of the tracheal wall ultimately depend on the integrity of the submucosal tracheal capillary plexus.

7. Swigart LL, Siekert RG, Hambley WC, et al: The esophageal arteries: an anatomic study of 150 specimens. Surg Gynecol Obstet 90:234, 1950 8 . Wanvick R, Williams PL: Grays Anatomy. 35th edition. Philadelphia, WB Saunders Company, 1973

Discussion
(Newark, NJ): This paper points out the necessity of preserving as much of the blood supply to the proximal and distal segments as possible in order to obtain a permanent adequate airway. We recently reported our experimental data (in the Journal of Thoracic and Cardiovascular Surgery) regarding the feasibility of homologous tracheal transplants. Our results showed that a five-ring transplant inevitably failed within a short time from dissolution of the cartilaginous rings. Although three-ring transplants were successful, the cartilaginous rings were again dissolved, but the proximal and distal anastomoses were pulled together because of the short distance. However, when the trachea was transected proximally and distally over a ten-ring segment, with the blood supply left intact along the membranous portion and the tracheal ends then resutured, the cartilage was not dissolved. Our experiments corroborate observations from the Mayo Clinic that long segments of the trachea must not be denuded when a primary anastomosis is performed.
DR. WILLIAM E. NEVILLE

References
1. Gardner ED, Gray DJ, ORahilly R: Anatomy: A

2.

3.
4.

5. 6.

Regional Study of Human Structure. Fourth edition. Philadelphia, WB Saunders Company, 1975 Grant JCB: An Atlas of Anatomy; by Regions: Upper Limb, Abdomen, Perineum, Pelvis, Lower Limb, Vertebrae, Vertebral Column, Thorax, Head and Neck, Cranial Nerves, and Dermatomes. Sixth edition. Baltimore, Williams & Wilkins Company, 1972 Grillo HC: Reconstruction of the trachea: experience in 100 consecutive cases. Thorax 28:667,1973 Michalewski K: Topography of the bronchial branches of the aorta. Folia Morphol (Warsz) 28:417, 1969 Miura T, Grillo HC: The contribution of the inferior thyroid artery to the blood supply of the human trachea. Surg Gynecol Obstet 123:99, 1966 Sobin SS, Frasher WG Jr, Tremer HM, et al: The microcirculation of the tracheal mucosa. Angiology 14:165, 1963