A COMPARATIVE STUDY of the AZYGOS VENOUS SYSTEM in MAN, MONKEY, DOG, CAT, RAT and RABBIT by DAVID BOWSHER Department of Anatomy, University of Liverpool

[ 400 ]

A COMPARATIVE STUDY OF THE AZYGOS VENOUS SYSTEM IN MAN, MONKEY, DOG, CAT, RAT AND RABBIT BY DAVID BOWSHER Department of Anatomy, University of Liverpool

It has been generally considered that the azygos vein acts as a by-pass between the inferior and superior caval systems, and most research has centred on the connexions of its caudal end. In view of the correlation between changes in the pressures of the azygos vein and the cerebrospinal fluid, it was decided to investigate the functional value of the azygos venous system, and the importance of its connexions with the internal vertebral venous system.

HISTORICAL INTRODUCTION The azygos vein has excited interest since the earliest days of anatomical study, and is mentioned in the third century by Galen (ed. 1822). Vesalius (1555) wrote of it at some length, and alluded to its connexion with the internal vertebral veins and inferior vena cava. Eustachius (1722) shows, without comment, the hemiazygos arising from the left renal vein. Winslow (1776) gives a fairly accurate description of the vein, together with the branches draining the spinal canal into the intercostal veins. The definitive anatomy of the spinal (internal vertebral) veins and their connexions was established by the Paris school in the first half of the nineteenth century, and undoubtedly the most important of these writers was Breschet (1829). He pointed out that a large vein emerges from each thoracic intervertebral foramen and ascends over the body of the vertebra to join the azygos or hemiazygos vein, having first joined forces with the posterior intercostal vein; and that this 'vein of the intervertebral foramen' is larger than the intercostal vein which overlies it. The description of the azygos and internal vertebral venous systems in the text- book of Poirier & Charpy (1902) is based mainly on the work of Breschet (1829) and of Walther (1885). These authors, and Breschet independently of them, state that there are no valves in the internal vertebral veins or in the azygos or hemiazygos, except that there is sometimes an incompetent valve in the arch of the azygos. It should be mentioned, however, that there is a valve in the posterior intercostal vein just before it joins the vein of the intervertebral foramen. Research has centred mainly upon the tenuous connexions of the caudal end of the azygos system, starting with Lejars's description of the 'canal reno-azygo- lombaire' (1888) said to be present in 80 % of cases. Seib (1934) dissected 100 white and 100 negro cadavers and classified the inferior caval and renal origins of the azygos and hemiazygos veins. Analysis of his tables shows that in 83 % of cases the azygos is connected with the inferior vena cava or the left renal vein. Comparative study of the azygos venous system 401 The azygos system has also been studied by authors whose main interest was in the collateral circulation following obstruction of the inferior vena cava. Among the more important of these are Sappey & Dumontpallier (1861), Pleasants (1911) and Batson (1940). Recently, Robinson (1949) ha§ studied the collateral circulation in stillborn foetuses after ligation of the inferior vena cava, and has indicated the connexions between the azygos and internal vertebral systems. Coman & de Long (1951) have made vinilyte corrosion preparations in the rat showing the normal anastomoses between the azygos and internal vertebral systems. Herlihy (1948) has studied the internal vertebral venous plexus of the cat, demonstrating the topography and mentioning its connexions with the azygos venous system, to which he also refers in an earlier publication (1947). Much of the literature on the internal vertebral venous system has been reviewed by Harris (1941). MATERIAL AND METHODS The material consisted of four full-term stillborn (but otherwise normal) human foetuses, three macaque (Macaca mulatta) monkeys, three dogs, two cats, twenty albino rats, and six rabbits. All the material except the human was adult. In all cases the azygos vein was injected retrogradely with either 20% bismuth oxy- chloride ('Chlorbismol') or a dispersion of barium sulphate ('Micropaque'), pre- paratory to radiography and dissection. Four of the rats underwent operative ligature of the vena azygos at its junction with the superior vena cava. In the operative technique, positive pressure anaesthesia was provided by a modi- fication of the method of Porter & Small (1947), details of which will be published elsewhere. The rat azygos is left-sided (see Beddard, 1907), so a lateral thoracotomy was performed in the third left intercostal space and the third rib resected. A ligature was passed under the azygos vein at its point of entry into the superior vena cava, immediately medial to the arch of the aorta, and the vein tied off. The chest was then closed in layers without drainage. These rats were killed and injected at intervals between 10 days and 1 month after operation. Once the anaesthetic and operative technique had been perfected, the mortality was nil. In the earlier cases of the series, the animals were heparinized before being killed, in order to facilitate the flow of the injection material; but experience showed that, provided the injection was made immediately after death, this made no difference. In all cases of retrograde injection of the azygos vein, the following ligations were made before injection: (i) Inferior vena vava, cephalic to the renal veins, but below the liver. (ii) Inferior vena cava, between the diaphragm and heart. (iii) Both renal pedicles at the hilum. (iv) Both lung roots. (v) Inferior vena cava at its point of formation from the common iliac veins. The liver, gut and both lungs were removed before injection, and the abdominal vena cava between the two lower ligatures was removed after injection. In a few specimens (e.g. cat, PI. 3, figs. 7, 8), towards the end of the injection, the injection mass burst through into the surrounding connective tissue between the layers of the mediastinal pleura, but this did not affect the radiographic pattern. 402 David Bowsher

RESULTS In all the species studied, the azygos and hemiazygos veins were well filled with injection mass. The degree of filling of the posterior intercostal veins was variable, depending upon the force of the injection and the competence of their valves. In the cases in which they did fill, it could be observed by direct vision that they were the last vessels to be filled, and injection was immediately stopped if and when the veins filled. In the human foetuses (P1. 2, figs. 3, 4) the intercostal veins were never in any case filled. This is testimonial to the statement of Poirier & Charpy (1902) that venous valves are more competent in infancy. Also in each species can be seen, at least in the thoracic region, in dorso-ventral views, the 'vein of the intervertebral foramen' joining the posterior intercostal vein just prior to its termination. In most cases it can be seen that this vessel is larger than the posterior intercostal vein prior to this union. Three other features of importance were noted in all species. First, in no case, save in the monkey, was any segment of the inferior vena cava filled by the injection mass. Even in the monkey, this was probably due to the fact that the left renal vein had been ligated lateral to its suprarenal tributary. Secondly, in all species (though not all specimens examined), the left suprarenal gland was outlined not through the suprareno-renal vein, but by a vessel which issues from the first lumbar intervertebral foramen and joins the plexus of vessels issuing from the gland. Lastly, in all species some or all of the torn ends of the lumbar veins were filled. In the Primates and Carnivora examined the internal vertebral venous plexus was constantly filled. In the four species studied, the main vessels seen are two lateral longitudinal trunks, whose position in relation to the vertebral bodies can be estimated by comparison of dorso-ventral and lateral radiographs. Dissection showed that they lay in the vertebral canal, outside the dura mater in the epidural fat. They received a large number of tributaries which emerge from the dura in company with the nerve roots. Certain differences in these veins between the Primates and Carnivora can be observed. Thus, in the Primates (Pls. 1, 2, figs. 1-4) there are a large number of cross-connexions between these two lateral longitudinal trunks, and the communications with the intradural vessels appear to be much richer than in the case of the Carnivora. In the cat (PI. 3, figs. 7, 8) there are no cross-connexions between the lateral longitudinal trunks; this confirms the observations of Herlihy (1948). In the dog, the cross-connexions are very sparse (PI. 2, figs. 5, 6). In the Primates, the internal vertebral venous plexus connects at its lower end with the anterior sacral plexus (P1. 1, fig. 1; PI. 2, fig. 3), which lies on the anterior aspect of the bodies of the sacral vertebrae; this plexus is not apparent in the Carnivora. Inspection of the lateral radiographs (P1. 1, fig. 2; P1. 2, fig. 5; PI. 3, fig. 8) shows the lateral longitudinal trunks to be greater in diameter than the vena azygos itself. It should also be noted here that man is the only species examined which appears to possess ascending lumbar veins. In the normal rodent (P1. 4, figs. 9, 10) the lower end of the azygos venous system terminates at the level of the left suprarenal vein; and although the connexions of Comparative study of the azygos venous system 403 the azygos and hemiazygos veins with the veins of the intervertebral foramina in the thoracic region are apparent, the internal vertebral venous plexus is not filled. However, after operative ligation of the azygos vein, flow in this system is reversed, and hypertrophy of the main trunks and all its tributaries occurs. Although this operation was successfully carried out in the rat, the rabbit seems unable to stand the procedures involved, and no rabbits survived operation. Injection of the operated rat (P1. 4, figs. 11, 12) shows the presence of two lateral longitudinal trunks (less well defined than in Primates and Carnivora), with numerous tributaries, and also a dorsal longitudinal trunk, very evident in cleared specimens, which runs along the tips of the vertebral spines. Here again the lateral longitudinal trunks are larger (when filled under pressure) than the vena azygos major.

DISCUSSION It would appear that the azygos venous system is the main intermediary between the superior caval system and the internal vertebral venous system. It is a fact of no little importance that it is the superior caval system which is mainly concerned. Many previous investigators (e.g. Seib, 1934) have been concerned with the connexions between the lower end of the azygos and hemiazygos veins with the inferior caval system. It is shown anatomically that these connexions are tenuous and of no functional importance. The most important piece of evidence in this connexion is the effect of ligation of the vein in the rat. There is no enlargement of any azygo-inferior-caval anastomoses; nor is there any effect upon the posterior intercostal vessels (PI. 4, figs. 11, 12), which shows that in the thoracic cage of the rat internal vertebral connexions are of paramount importance, a point which was not stressed by Halpern (1953). Breschet (1829) was the first to claim that flow in the azygos and internal vertebral venous plexuses may be in either direction, and further attention has been drawn to this by Franklin (1937). The recent work of Collins, Weinstein, Norton & Webster (1952) has demonstrated the collateral circulation following ligature of the inferior vena cava, and this has been studied by Robinson (1949). Previous to these workers, Sappey & Dumont- pallier (1861) and Pleasants (1911) had studied the collateral circulation in such cases. They all stress the enlargement of the internal vertebral venous plexus and of the azygos major. In all such cases it would appear that inferior caval blood is first transferred to the internal vertebral venous plexus and thence to the azygos system. Batson (1940) has stressed the importance of the internal vertebral venous plexuses in metastatic phenomena. It seems that by means of the connexions between the azygos and internal vertebral venous systems, an explanation may be found for the frequency of the occurrence of metastases from bronchial carcinoma to the adrenal gland; it is intended further to investigate this particular problem. Lastly, the role of this venous plexus in the control of cerebro-spinal fluid pressure should be noted. This has already been the subject of some attention by Herlihy (1947) and Bowsher (1953). The filling of the lateral longitudinal sinuses and their tributaries causes an increase in cerebro-spinal fluid pressure; this filling can in its turn be caused by compression of the intrathoracic azygos veins, which squeezes blood into the internal vertebral venous plexus; and this appears to be the 404 404DaiBoheDavid Bowsher mechanism of the respiratory variation in cerebro-spinal fluid pressure. Similarly, a rise in pressure is caused by inferior caval obstruction and this is compensated by adjustments in the production and absorption of cerebro-spinal fluid. It is also important to note that the venous drainage of the adrenal is not solely into the renal vein on the left side. Attempts to measure the blood flow through, or hormonal output from, the left adrenal gland by cannulation of the left adrenal vein (e.g. Vogt, 1944) are probably not entirely accurate.

SUMMARY 1. The literature on the anatomy of the azygos vein and its connexions is reviewed. 2. The azygos venous system and its connexions, particularly with the internal vertebral venous plexus, is described in man, monkey, cat, dog, rat and rabbit. 3. It is particularly stressed that the azygos venous system is functionally a part of the internal venous vertebral system, draining this latter into the superior caval system. Various consequences of this function are discussed. This research was carried out during the tenure of a John Rankin Research Fellow- ship in Human Anatomy at the University of Liverpool and was aided by a grant from the Medical Research Council. I wish to express my thanks to Prof. R. G. Harrison, at whose suggestion this investigation was undertaken, and to Mr L. G. Cooper and Mr C. Fitz-Simon for their technical assistance.

REFERENCES BATSON, O. V. (1940). The function of the vertebral veins and their role in the spread of metastases. Ann. Surg. 112, 138-149. BEDDARD, F. E. (1907). On the azygos veins in the mammalia. Proc. zool. Soc. Lond. 181-223. BOWSHER, D. (1953). The cerebrospinal fluid pressure. Brit. med. J. 1, 863-865. BRESCHET, G. (1829). Recherches anatomiques, physiologiques et pathologiques sur le system veineux, et specialement sur les canaux veineux des os. Thesis, Paris. COLLINS, C. G., WEINSTEIN, B. B., NORTON, R. 0. & WEBSTER, H. D. 81952). The effects of ligation of the inferior vena cava and ovarian vessels on ovulation and pregnancy in the human being. Amer. J. Obstet. Gynec. 63, 351-358. COMAN, D. R. & DE LONG, R. P. (1951). The role of the vertebral venous system in the metastasis of cancer to the spinal column. Experiments with tumor-cell suspensions in rats and rabbits. Cancer, 4, 610-618. EUSTACHIUS, B. (1722). Tabulae Anatomicae, Tabula Iv, p. 11. Amstelcedami: Wetstenios. FRANKLIN, K. J. (1937). A Monograph on Veins. London: Bailliere, Tindall and Cox. GALEN, C. (ed. KUHN) (1822). De Usu Partium Corporis Humani. Vol. III, Book VI, chap. 14. Leipsig: Cnobloch. HALPERN, M. H. (1953). The azygos vein system in the rat. Anat. Rec. 116, 83-93. HARRIS, H. A. (1941). A note on the clinical anatomy of the veins, with special reference to the spinal veins. Brain, 64, 291-300. HERLIHY, W. F. (1947). Revision of the venous system: the role of the vertebral veins. Med. J. Aust. 1, 661-672. HERLIHY, W. F. (1948). Experimental studies on the internal vertebral venous plexus. From Essays in Biology, pp. 151-163. Presented to A. N. Burkitt. Sydney: University Press. LEJARS, F. (1888). Les voices de sfirete de la veine rbnale. Bull. Soc. Anat. Paris, 43, 504-511. PLEASANTS, J. HALL (1911). Obstruction of the inferior vena cava, with a report of eighteen cases. Johns Hopk. Hosp. Rep. 16, 363-548. POIRIER, P. & CHARPY, A. (1902). Traits d'Anatomie Humaine, 2nd ed., Tome II, Fasc. 3. Paris: Masson et Cie. Comparative study of the azygos venous system 405 PORTER, C. B. & SMALL, J. T. (1947). A method for intrathoracic operation on the rat. Proc. Soc. exp. Biol., N.Y., 64, 239-241. ROBINSON, L. S. (1949). The collateral circulation following ligation of the inferior vena cava (injection studies in stillborn infants). Surgery, 25, 329347. SAPPEY, C. & DUMONTPALLIER, V. A. A. (1861). Note sur un cas d'obliteration de la veine cave inferieure avec circulation collaterale; suivie de faits analogues demonstrant qu'il existe trois principales varietes d'oblit6rdtion de cette veine. C.R. Soc. Biol., Paris, 3rd ser., 3, Sect. mom., 135-155. SEIB, G. A. (1934). The azygos system of veins in American whites and American negroes, including observations on the inferior caval venous system. Amer. J. Phys. Anthrop. 19, 39-163. VESALIUS, A. (1555). De Humani Corporis Fabrica, Lib. III, p. 461. Basel: Oporini. VOGT, M. (1944). Observations on some conditions affecting the rate of hormone output by the suprarenal cortex. J. Physiol. 103, 317-332. WALTHER, C. (1885). Recherches Anatomiques sur les Veines du Rachis. Thesis, Paris. WINSLOW, J. B. (1776). An Anatomical Exposition of the Structure of the Human Body, section v. Transl. by G. Douglas. 5th ed. (revised). London: J. F. Rivington.

EXPLANATION OF PLATES PLATE 1 The normal pattern of the azygos and internal vertebral venous systems in the monkey (Macaca mulatta). After ligation of the inferior vena cava cephalic to the renal veins, but below the liver, the inferior vena cava between the diaphragm and heart, both renal pedicles at the hilum, both lung roots, and the inferior vena cava at its point of formation from the common iliac veins, and removal of the liver, the gut below the diaphragm and both lungs, the azygos vein was injected retrogradely with barium sulphate suspension ('Micropaque'), and the abdominal vena cava between the two lower ligatures was then removed prior to radiography. Fig. 1. Dorso-ventral view. A, vein of intervertebral foramen; B, lateral longitudinal internal vertebral vein; C, suprarenal vein; D, anterior sacral venous plexus. Fig. 2. Lateral view. A, azygos vein; B, lateral longitudinal internal vertebral vein; C, suprarenal vein, showing connexion to internal vertebral system; D, lumbar vein.

PLATE 2 The normal pattern of the azygos and internal vertebral venous systems in the fresh stillborn full-term human foetus and in the dog, as shown by radiography after the ligations described in the legend to Plate 1. Fig. 3. Human foetus. Dorso-ventral view. A, ascending lumbar vein; B, lateral longitudinal internal vertebral vein; C, anterior sacral plexus. Fig. 4. Human foetus. Lateral view. A, azygos vein; B, lateral longitudinal internal vertebral vein. Fig. 5. Dog. Lateral view. A, azygos vein; B, lateral longitudinal internal vertebral vein; C, suprareno-vertebral vein; D, lumbar vein. Fig. 6. Dog. Dorso-ventral view. A, lateral longitudinal internal vertebral vein; B, azygos vein; C, suprareno-vertebral vein. PLATE 3 The normal pattern of the azygos and internal vertebral venous systems in the cat. In the upper part of the figures it can be seen that the injection mass has burst through the azygos vein into the surrounding connective tissue (mediastinal pleura), but the lower part of the vein is intact. Radiography after the ligations described in the legend to P1. 1. Fig. 7. Dorso-ventral view. A, azygos vein (which has burst into surrounding connective tissue); B, lateral longitudinal internal vertebral vein; C, suprareno-vertebral vein. Fig. 8. Lateral view. A, azygos vein, which has burst into surrounding connective tissue; A', lower (unburst) portion of azygos vein; B, lateral longitudinal internal vertebral vein; C, suprareno-vertebral vein; D, lumbar vein. 406 David Bowsher

PLATE 4 The normal pattern of the azygos and internal vertebral venous systems in the rabbit and rat (figs. 9, 10). The pattern of the azygos and internal vertebral venous systems in the rat 1 month after operative ligation of the azygos vein at its entry to the superior vena cava (Figs. 11, 12). Radiography after the ligations described in the legend to P1. 1. Fig. 9. Dorso-ventral view. A, azygos vein; B, vein of intervertebral foramen; C, lateral longitudinal internal vertebral vein; D, suprareno-vertebral vein. Fig. 10. Dorso-ventral view. A, vein of intervertebral foramen; B, azygos vein; C, lateral longitudinal internal vertebral vein; D, suprarenal vein; E, suprareno-vertebral vein. Fig. 11. Dorso-ventral view. A, vein of intervertebral foramen; B, lateral longitudinal internal vertebral vein; C, connecting vessel between the two lateral trunks. Fig. 12. Dorso-ventral view. A, lateral longitudinal internal vertebral vein; B, suprareno- vertebral vein; C, lumbar vein. Journal of Anatomy, Vol. 88, Part 3 Plate 1

-- A: - ~ -'

BOWSHER-COMPARATIVE STUDY OF THE AZYGOS VENOUS SYSTEMI Journal of Anatomy, Vol. 88, Part 3 Plate 2

BOWSHER-CoMPARATIVE STUDY OF THE AZYGOS VENOUS SYSTEM Journal of Anatomy, Vol. 88, Part 3 Plate 3

A .s | ." | _ . 5..!,,,:, _ yea: !

A' .. z

B P

C-.-

(Or

BOWSHER COMPARATIVE STUDY OF THE AZYGOS VENOL S SYSTEM Journal of Anatomy, Vol. 88, Part 3 Plate 4

*.. " "|111" 1111||l tl_E*-I.",aI*_ilI| _I_ 11||*__C:its B - Lo i! s =@r / tW _ .-_}'_bsr.:K7g_1g!_.!u_A L--z.:IFVaI^;34R ;_i.IEI_1|111i

,i je ,D _¢.eB >'"0 5 *] _e,C _ t ri-rs s 7i\i_,--.l ...... V -r..ii.t'1E_ ffi^ ,,<5\ )s1 X. g .tt r _ l,, 1;,.=;s ..i_a...... _ Ersi ..s sj_ilnjw..s.s t : L _ D ;______.ls. :: ,v'j Z z :s}'DE\> s.g'. li ZM _ sz W ,: irX,-

_or..: .r¢:_ 7

.t.^_F __ _NF.:bi X_ t_ES

__! X *t *Lt1_ ' _ v74 >** _

____. A R! * i _ ..w_.,'s '&._!- 6 4|$ J _4 i,!, swa_ -- K_KJ .--- 8 it^ i./' -E_>iS,b, 5,,.w _ i ?d ^ A.. =' s /t'S 5 _t.

tK_ A . to ,_w,ts /g_ B____tY # s., FS*,¢,,._';. ',;wti- F's_-x rv _. _L: - *Stow,'S,,e BOWSHER-CONIPARATIVEj_C'S_/*__iL!...e,.....STUDY'OF THEeAZYGOS_Ew11iNtsi<_ca_T..VENOUSn_.jSYSTEM...Fg'),SiI_Fn,f_L