The venous anatomy of experimental left varicocele: comparison with naturally occurring left varicocele in the human*

FERTLTY AND STERLTY Copyright" 1994 The American Fertility Society Vol. 62, No.4, October 1994 Printed on acid-free paper in U. S. A. The venous anatomy of experimental left varicocele: comparison with naturally occurring left varicocele in the human* Terry T. Turner, Ph.D.t:j: Stuart S. Howards, M.D. University of Virginia School of Medicine, Charlottesville, Virginia Objective: To determine the effect of experimental left varicocele on the anatomy of the veins serving the rat testis and to compare that anatomy to known patterns of vascular drainage from the human testis with and without varicocele. Design: Vascular maps were made of the effluent vessels from the rat testis in control animals and those with a 30-day experimental left varicocele. Consensus maps were arrived at and these were compared to published reports of the pertenent venous anatomy in humans with and without varicocele. Setting: Research laboratory Results: The major route of blood leaving the rat testis was confirmed to be the spermatic vein, but nine common collaterals were also found to exist. Four of these collaterals became more pronounced with experimental varicocele as did several dilated perineal veins. These latter vessels all led to the iliac vein. The vasculature of the rat experimental varicocele model shares some important anatomical features with human varicocele anatomy. Conclusions: Varicocele in humans and in the rat model causes a redistribution of blood flow from a route primarily out the spermatic vein to routes leading to the iliac vein. The redistribution is similar but not identical. Fertil Steril 1994;62:869-75 Key Words: Rat, varicocele, vasculature, anatomy, testis, human The pathophysiology of varicocele has been the object of much research, and theories developed from human studies (1, 2) have been tested in experimental models including primates (3), dogs (4), rabbits (5), and rats (6-9). Although experimental left varicocele in animal models is not identical to naturally occurring varicocele in humans, the human and the experimental varicoceles do result from similar forces. Received December 20, 1993; revised and accepted April 25, 1994. *Supported by grant HD18252 from the National nstitutes of Health, Bethesda, Maryland. t Departments of Urology, Anatomy, and Cell Biology. *Reprint requests: Terry T. Turner, Ph.D., Departments of Urology, Anatomy, and Cell Biology, University of Virginia School of Medicine, Charlottesville, Virginia 22908 (FAX: 804 924-8311). Departments of Urology and Physiology. Experimental left varicocele is induced by partial occlusion of the left renal vein (3, 6) that mimics the "nutcracker" phenomenon in humans, an anatomical compression of the left renal vein believed by some to be a major cause of naturally occurring varicocele (10, 11). n both man and laboratory animals, the partial occlusion of the left renal vein causes an increase in intravenous pressure lateral to the occlusion site, which is transmitted to the inserting left spermatic vein. This increase in V pressure causes the development of a varicosity of the spermatic vein and pampiniform plexus. The experimental left varicocele model has allowed inquiry into the pathophysiological consequences of varicocele beyond that possible with human patients. Using a wide range of techniques and protocols in experimental animals, various investigators have studied the consequences of experimental left varicocele on such physiological parameters Vol. 62, No. 4, October 1994 Turner and Howards Anatomy of experimental varicocele 869

as testicular blood flow (4, 6, 7), spermatogenesis (3, 5), endocrine function (3, 9), and immune response (8). The widespread use of such an experimental model mandates constant re-evaluation of the model as more comprehensive information becomes available; it was because ofthis that the present study was undertaken. t has been demonstrated in earlier studies that surgery to induce experimental left varicocele in the rat causes the development of a varicosity of the left spermatic vein (6, 12), and this varicosity extends to the pampiniform plexus. Thus, to the degree examined, the rat vasculature pertinent to experimental left varicocele has appeared similar to the human spermatic vasculature with varicocele; nevertheless, a comprehensive examination of the effluent vessels of the rat testis with comparison with their analogues in the human has not been made, and no comprehensive study exists of the effect of experimental left varicocele on these vessels. Such a study is important in establishing the degree of similarity between the human varicocele and the common experimental left varicocele rat model. MATERALS AND METHODS Adult male Sprague-Dawley rats (450 to 550 g) were obtained from University Vivarium sources and maintained in a 12 hour:12 hour light:dark cycle with ad libitum food and water. Animals were anesthetized with urethane (10 mg/kg P) and prepared for micropuncture of the testicular surface veins of the left testis. Animals were administered 0.1 ml heparin solution (7,000 U/mL) through the femoral vein. The animal was then killed with an intracardiac injection of saturated KCl, and the testicular vasculature and the internal spermatic vein were exposed through their entire length. A sharpened, glass micropipette (100-ttm tip diameter) attached by cannula to a 10-mL syringe filled with 0.3% blue dextran (2 X 10 6 MW; Sigma Chemical Co., St. Louis, MO) was used to puncture one of the testicular surface veins. The testicular venous vasculature was perfused in the cranial direction under constant pressure until dye in the left spermatic vein had reached and filled the left renal vein. The perfusion rate was approximately 1 ml/min, and, typically, approximately 3 ml of the dye solution was used in the perfusion process. Upon completion of the perfusion, the testicular effluent vasculature, anastomosing vessels, and collaterals were further exposed by careful dissection. The entire venous tree was examined, including viewing under the dissecting microscope. Vascular maps that indicated the effluent vascular routes from the testis, the collaterals that existed to those routes, and the anastomoses that existed between routes were prepared. This was all traceable by following the path of the still intravascular blue dextran dye. Vessel identity was assisted by reference to a classical text of rat anatomy (13). n all animals, the diameter of the internal spermatic artery was measured at the level of the crossing ileolumbar vein. Diameter of major anastomosing vessels was also noted, but these vessels varied between animals. The above procedure was performed on a group of seven control rats and a group of seven rats that had a surgically induced experimental left varicocele 30 days previously (4, 6, 9). Briefly, the left renal vein was subjected to blunt dissection medial to the insertion of the left spermatic vein. A 4-0 silk suture was passed around the vessel and tied down over a 0.85-mm metal rod (4, 6, 9). The metal rod was retracted from the stricture, leaving the renal vein diameter reduced by approximately 50%. f collaterals to other abdominal vessels were detected upon visual examination of the midabdomen area, typically around and above the ileolumbar vein, the vessel was ligated to prevent it relieving the increase in intravessel pressure. This operation induces the significant dilatation of the left spermatic vein (12). Using the data maps from each animal, consensus maps were constructed of the rat testicular venous return with and without experimental left varicocele. These maps were compared with similar consensus maps of the effluent vessels from the human testis with and without varicocele. These latter maps were determined by survey of studies using direct examination, venography, scintigraphy, and cadaver studies (10, 14-20) along with reference to standard anatomy texts. RESULTS Venous Drainage of the Rat Testis The general pattern of the venous system serving the left rat testis is illustrated in Figure 1A and the nomenclature given in Table 1. The major route for blood leaving the left rat testis is the spermatic vein, but at the distal aspect of or just superior to the pampiniform plexus, several consistent anastomoses do exist. These anastomoses make direct connections between the left spermatic vein and the vesicular plexus of the seminal vesicles, the def- 870 Turner and Howards Anatomy of experimental varicocele Fertility and Sterility

A <!..-..., _j ' \ _,/ -~;.:::... BLA'--... _j -r ' ' \, / '... _... J', Figure 1 Vascular anatomy of experimental left varicocele in the rat. These illustrations are schematic and are intended to show vessels that are or could be routes of effluent blood flow from the left testis. To illustrate the vessels clearly, it was necessary to keep the scale and position of vessels only approximate. (A), Conventional venous system pertinent to left varicocele in the rat. Abbreviations are explained in Table 1. (B), Venous anatomy found in control rats of the present study. ncidence and approximate size of numbered collaterals are explained in Table 2. (C), Venous anatomy of rats with experimental left varicocele. Noticeably dilated vessels are shown in black. The effect on numbered collaterals in indicated in Table 2. erential (vasal) vein of the vas deferens, and the external iliac (Fig. la). Although standard anatomic descriptions were confirmed in the present study, considerable variability did exist between animals. There were nine common, but irregularly occurring, collaterals present in the control animals (Fig. lb) that conveyed dye injected into the left testicular vein. These anastomoses are numbered in Figure lb as follows: [1] a spermatic vein inferior vena cava (V C) shunt just inferior to the insertion of the left renal vein into the VC; [2] a duplication of the spermatic vein typically running from just below the crossing of the ileolumbar vein (midabdomen, not the pelvic ileo- Table 1 Veins Relevant to Venous Drainage of the Testis Abbreviation Vein Abbreviation Vein. C. R.V. A.V. 1.8. U.V..V. C.l..E. V.P. P.P. E.. nferior vena cava Renal Adrenal nternal spermatic Ureteric liolumbar Common iliac nferior epigastric Vesicular plexus Pampiniform plexus Exterior iliac 1.1. F.V. c.v. D.V. v.v. E.P. P.E..P. s.v. R.C. nferior iliac (hypogastric) Femoral Cremasteric (exterior spermatic) Deferential Vesicular Exterior pudendal Pseudoepigastric trunk nternal pudendal Scrotal Renal capsular Vol. 62, No.4, October 1994 Turner and Howards Anatomy of experimental varicocele 871

Table 2 rregular Collateral Veins in Control Rats and Those With Experimental Left Varicocele: Rate of Occurrence and Approximate Vessel Diameter (mm) Control ELV Symbol Occurrence Diameter Occurrence Diameter 1. 3/7 so.ol 7/7.:0:0.02 2. 2/7 so.ol 2/7 so.ol 3. 5/7 so.ol 4/7 so.ol 4. 2/7 so.ol 2/7 so.ol 5. 2/7 so.ol 0/7 NA* 6. 5/7 so.ol 0/7 NA 7. 4/7 so.ol 5/7.:0:0.02 8. 5/7 so.ol 3/7 so.ol 9. 4/7 so.ol 3/7 s0.02 * NA, not available. lumbar vein of humans) to the duplicated insertion in the renal vein; [3] a small group of collaterals connecting the ureteric vein to the renal capsule or perinephric fat; [ 4] a long anastomosing vessel between the ileolumbar vein and the insertion of the spermatic vein into the left renal vein; [5] small collaterals tracing into the abdominal fat, no discernible connection to the VC; [6] small collaterals tracing into the left abdominal fat, no discernible connections directly to other veins; [7] a shunt connecting the left spermatic vein with the left ureteric vein; [8] a shunt connecting the left spermatic vein with the juncture of the common iliac and the VC; and [9] a shunt connecting the pampiniform plexus to the cremasteric (external spermatic) vein. The rate of occurrence and general vessel size of these collaterals are noted in Table 2. The collateral vessels are typically 0.01 mm in diameter or smaller. n control animals, the diameter of the left spermatic vein at the level of the ileolumbar vein was 0.16 ± 0.02 mm. No direct collaterals to the contralateral testis existed, but communication with the right venous system via the vesicular plexus and the pudendal plexus (not shown) was evident. No dye ever traced to the contralateral testis. Thirty days after surgery to establish experimental left varicocele, the midabdominal spermatic vein diameter had enlarged to 1.47 ± 0.16 mm. The general venous architectural pattern remained essentially the same after experimental left varicocele but with a clear enlargement of several vessels via the shunts originating in the spermatic vein (Fig. 1C). The collaterals that had noticeably increased in caliber were those identified as collaterals 1, 2, 7, and 9 (Fig. 1C, Table 1). Collateral 7 had developed sufficiently to be the apparent origin of the enlargement of the left ureteric vein often seen after experimental left varicocele (Fig. 1C). Collateral 9 had increased in diameter as well but had no noticeable effect on the small receding vessel, the cremasteric vein. Rats with 30-day experimental left varicocele had dilated left spermatic veins, ureteric veins, deferential veins, vesicular plexi, and vesicular veins (Fig. 1C). As in control animals, no dye ever traced to the contralateral testis. Venous Drainage of the Human Testis A consensus map of the venous effluent of the human testis is illustrated in Figure 2A. The testicular vasculature drains primarily via the internal spermatic vein, but consistent anastomoses exist with the following: [1] the cremasteric vein that is contributary to the inferior epigastric that is contributary to the external iliac; [2] the scrotal veins that are con tributary to the external pudendal that is contributary to the great saphenous vein; and [3] the deferential vein (often involving the vesicular plexus) that is contributary to the internal iliac. Left varicocele causes a dilation of the left spermatic vein and pampiniform plexus, and variably, the anastomosing vessels in the three main accessory routes mentioned above. The accessory vessels consistently noted in a variety of studies (14-20) in the order of their general (though not universally accepted) prominence are the cremasteric vein, the external pudendal vein, and the deferential vein (Fig. 2B). DSCUSSON t should be noted in the beginning that individual variation is one of the major difficulties of the study of venous anatomy in both humans and laboratory animals. Consensus maps are valuable because they indicate major, consistently appearing routes of vascular flow, but their weakness is that they simplify a natural complexity. This is indicated in the rat by the variability of the shunts and collateral vascular routes found in the present study (Table 2). Also, it has commonly been the case that the experimental left varicocele operation does not cause all animals to develop a varicocele (4). This is due, at least in part, to the variability of the vasculature and the existence of collaterals not detected by cursory visual exam. Failure to develop a varicocele after the experimental left varicocele operation has typically occurred in fewer than 10% of cases. n the past, development of the varicocele 872 Turner and Howards Anatomy of experimental varicocele Fertility and Sterility

Adrenal B Figure 2 Vascular anatomy of left varicocele in man. (A), Conventional pattern of veins that potentially serve as routes of effluent blood flow from the left testis. Abbreviations are explained in Table 1. (B), Vessels consistently noted in previous studies (see text) as being major effluent routes in men with varicocele are shown in black. after experimental left varicocele surgery was determined by examination for dilation of the left internal spermatic vein (4, 6, 9). However, in the present study, a more detailed examination was carried out with new emphasis on the pelvic and infrainguinal anastomoses that eventually contribute to the iliac veins, i.e., those anastomoses involving the deferential vein, the vesicular vein, and the internal and external pudendals (Fig. 1A). Note was also made of previously undetected, relatively consistent anastomoses that exist in the rat (Fig. 1B, Table 1) and have extended previous observations of the vascular effects of experimental left varicocele by demonstrating major dilations of the anastomo- ses leading to the iliac veins mentioned above. We have also documented the common existence of a ureteric-spermatic shunt that accounts for the dilation of the ureteric vein often seen in rats after experimental left varicocele. Four consistent differences exist between the rat and human veins serving the testis (Figs. 1 and 2). First, in the rat there is no evidence of a connection between the pampiniform plexus and the greater saphenous vein via the external pudendal vein as exists in the human (Fig. 2A). Rather, in rats the external pudendal joins the inferior epigastric to form a psuedoepigastric trunk that inserts into the external iliac (Fig. 1A). Second, in the rat, the Vol. 62, No.4, October 1994 Turner and Howards Anatomy of experimental varicocele 873

shunt between the pampiniform plexus and the cremasteric vein is inconstant and is a minor vessel (Fig. 1B, Table 1). Third, in the human, there is no main ureteric vein as in the rat, but small twigs of ureteral veins do sporatically join the distal internal spermatic vein (Fig. 2). Fourth, in humans, a lateral division of the spermatic vein anastomoses with the renal capsular vein and receives twigs from the perinephric fat (Fig. 2A). n the rat, it is the ureteric vein that receives twigs from the perinephric fat (Fig. 1A). Having stated these differences, it is important to note the important similarities between the gonadal venous systems of the rat and human. These are that the primary effluent vessel in both testes is the internal spermatic vein via the pampiniform plexus, and that the secondary effluent vessel, through its tributaries, is the iliac vein. The effects of varicocele on the effluent vessels of the testis appear similar in rats and humans, although variability of reported findings in the human remain problematical. Harrison ( 14), for example, reported that varicocele was a dilatation of the cremasteric veins while the common belief has been that the dilatation is of the pampiniform plexus ( 17, 20-22). nvestigators have also varied in their reports of which vessels other than the pampiniform plexus are the prominent effluents in varicocele. Hill et al. (16) and Wishahi (20), for example, have reported that the major secondary effluent route is via the external pudendal veins, whereas Segmund et al. (17) reported the most prominent secondary route to be the cremasteric vein. n like manner, several authors have reported a variety of secondary routes (14-20), with little agreement as to their order of importance. Some investigators remark on collateral vessels to the contralateral testicular vasculature (15) as if it is a common occurrence. Others report contralateral anastomoses in a minority of cases (16), whereas others maintain that there are no anastomoses with the contralateral vasculature (20). Despite these inconsistencies, it is possible to construct a consensus map in reasonable agreement with the majority of reports, and this map illustrates the prominence of the cremasteric vein, the external pudendal vein, and the deferential vein as accessory routes of blood flow in humans with varicocele (Fig. 2B). n the rat, the prominent secondary effluent vessels are the deferential vein, the vesicular vein, and the iliospermatic shunt. Although there is some variance in these routes between rat and human, the important similarity is that these secondary routes all terminate in the iliac vein in both humans and in rats. These findings demonstrate that, in general, the vasculature of the rat experimental left varicocele model reacts to compression of the left renal vein in much the same manner as in the human. Varicocele in either species causes a similar, though not identical, redistribution of blood flow from a route out the spermatic vein to routes ending in the iliac vein (Figs. 1C and 2B). This similarity and physical response at the level of the vasculature further strengthens the appeal of the rat experimental left varicocele model as an appropriate approach for the study of the pathophysiology of varicocele and brings attention to the iliac vein as a significant contributor to effluent testicular blood flow. The possible role of the iliac vein in varicocele has been addressed by others (23) but has received little general attention. Also, the expansion of the several effluent vessels illustrated in Figure 1 is consistent with the fact that testicular blood flow increases, rather than decreases, after experimental left varicocele (4, 6). The expansion of several possible effluent routes in the human (Fig. 2) and the finding of increased arterial flow velocity by Mellinger (24) are consistent with the same phenomenon occurring in the human with varicocele. Whether or not varicocele collaterals in the human are similar to collaterals 1 to 5 (Fig. 1B) in the rat is not known, but if they are, they would not interfere with the effectiveness of any of the surgical approaches currently used to correct a human varicocele. This is not the case with collaterals 6 to 9, which might remain open after a high ligation of the internal spermatic vein. These collaterals may be similar to the perforating veins some surgeons describe in the inguinal canal. These surgeons maintain that an inguinal or subinguinal approach is necessary to eliminate these collaterals and ensure that the varicocele is corrected (15). Thus, it is critical to determine whether or not these veins are important in the human, and if, when patent, they cause a persistent varicocele or if they merely allow normal venous drainage of the testis. n addition, several surgeons describe independent cremasteric veins that drain the testis but are not collaterals of the internal spermatic vein (25). We did not find that the rat testis is commonly drained by a cremasteric venous system, but such veins would need to be ligated if they do, in fact, contribute to varicocele formation in the human. Becket al. (25) found gubernacular veins in the majority of their human patients, but this is clearly 874 Turner and Howards Anatomy of experimental varicocele Fertility and Sterility

not the case in rats, and the role of these veins in the development of varicocele is uncertain. nterestingly, the reported failure rates of various surgical approaches (high ligation, inguinal, scrotal) for treatment of varicocele are similar. This provides circumstantial evidence against the clinical importance of perforating collaterals or gubernacular veins in the development of varicocele. Finally, this has been a study of anatomy. We have not measured actual blood flow through the various testicular venous routes described and, thus, have not verified that any specific vessel dilated by varicocele contributes specifically to testicular pathophysiology. Acknowledgment. The authors appreciate the technical assistance of Ms. Katherine J. Brown. REFERENCES 1. Verstoppen GR, Steeno OP. Varicocele and the pathogenesis of the associated subfertility. Androlgia 1978;10:85-102. 2. Turner TT. Varicocele: still an enigma. J Urol 1982; 129:695-9. 3. Kay R, Alexander NJ, Baugham WL. nduced varicoceles in Rhesus monkeys. Fertil Steril 1979;31:195-9. 4. Saypol DC, Howards SS, Turner TT, Miller ED. The influence of surgically induced varicocele on testicular blood flow, temperature, and histology in adult rats and dogs. J Clin nvest 1981;68:39-45. 5. Snydle FE, Cameron DF. Surgical induction of varicocele in the rabbit. J Urol1983;130:1005-9. 6. Green KF, Turner TT, Howards SS. Varicocele: reversal of testicular and temperature effects of varicocelectomy. J Urol1984;131:1208-11. 7. Nagler HM, Lizza EF, House SD, Tomashefsky P, Lipowsky HH. Testicular hemodynamic changes after the surgical creation of a varicocele in the rat. J Androl 1987;8:292-8. 8. Shook TE, Nyberg LM, Collins BS, Mathews S. Pathological and immunological effects of surgically induced varicocele in juvenile and adult rats. Am J Reprod lmmunol 1988;17:141-4. 9. Turner TT, Evans WS, Lopez T J. Gonadotroph and Leydig cell responsive in the male rat: effects of experimental left varicocele. J Androl 1990;11:555-62. 10. Sayfan J, Halevy A, Oland J, Nathan H. Varicocele and left renal vein compression. Fertil Steril1984;41:411-7. 11. Mali WPM, Oei HY, Arndt JW, Kremer J, Coolsaet BLRA, Schuur K. Hemodynamics of varicocele. Part. Correlation among the clinical, phlebographic and scintigraphic findings. J Urol 1986;135:483-8. 12. Turner TT, Jones CE, Roddy MS. Experimental varicocele does not affect the blood-testis barrier, epididymal electrolyte concentrations, or testicular blood gas concentrations. Bioi Reprod 1987;36:926-32. 13. Greene EC. Anatomy of the rat. New York: Hofner Publishing Co., 1963. 14. Harrison RG. The anatomy of varicocele. Proc R Soc Med 1966;59:763-5. 15. Coolstaet BLRA. The varicocele syndrome: venography determining the optimal level for surgical management. J Urol 1980;124:833-9. 16. Hill JH, Hirsch AV, Pryor JP, Kellett MJ. Changes in the appearance of venography after ligation of varicocele. J Anat 1982;135:4 7-52. 17. Segmund G, Gall H, Bahren W, Wetterau V. Hemodynamics of varicoceles: venous shunting in grade and grade varicoceles. Ann Radio! (Paris) 1989;32:24-8. 18. Mali WPT, Oei HY, Arndt JW, Kremer J, Coolstaet BLRA, Schur K. Hemodynamics of the varaicocele.. Correlation among the results of renocaval pressure measurements, varicocele scintigraphy, and phlebography. J Urol 1986; 135:489-93. 19. Wishahi M. Detailed anatomy of the internal spermatic vein and ovarian vein. Human cadaver study and operative spermatic venography. J Urol 1991;45:780-4. 20. Wishahi MM. Anatomy of the spermatic venous plexus (pampiniform plexus) in men with and without varicocele: intraoperative venographic study. J Urol1992;147:1285-9. 21. Saypol DC. Varicocele. J Urol1981;2:61-71. 22. Pryor JL, Howards SS. Varicocele. Urol Clin North Am 1987;14:497-513. 23. Mali WPTM. The varicocele: a haemodynamic and spermatologic study. Dordrecht: CG Printing, 1984. 24. Mellinger BC. Testicular artery blood flow velocity in men with varicocele. J Urol 1991;145:282A. 25. Beck EM, Schlegel PN, Goldstein M. ntraoperative varicocele anatomy-a macroscopic and microscopic study. J Urol 1992;148:1190-4. Vol. 62, No.4, October 1994 Turner and Howards Anatomy of experimental varicocele 875