Home  |  Archive  |  AJA @ Nature  |  Online Submission  |  News & Events  |  Subscribe  |  APFA  |  Society  |  Links  |  Contact Us  |  中文版

Effects of castration and testosterone replacement on veno-occlusion during penile erection in the rat

Yu-Tian DAI1, Vivienne Stopper2, Ronald Lewis1, Thomas Mills1,2   

1Department of Surgery, Urology Section, 2 Department of Physiology and Endocrinology Medical College of Georgia, Augusta, GA 30912-3000, USA

Asian J Androl  1999 Jun; 1: 53-59

Keywords: castration; testosterone; cavernous sinus; penile erection; nitric oxide; androgens
Aim: To determine if androgens directly regulate veno-occlusion or if androgens act indirectly to maintain the penile structures which control outflow. Methods: Using CASTRATE and TESTO rats, measurement was made of mean arterial pressure (MAP), intracavernosal pressure (CCP), and intracavernosal flow (CCF) during erection resulting from stimulation of the autonomic innervation of the penis. CCP and CCF were also measured during saline infusion into the cavernosal sinuses before and after treatment with sodium nitroprusside (SNP, a nitric oxide donor drug) to fully relax cavernosal smooth muscle. Penile tissue was also collected to measure the content of actin and proline and hydroxyproline to determine if brief withdrawal of androgenic support led to changes in the number of smooth muscle cells or the collagen content of the tissue. Results: Infusion of saline into the cavernosal sinuses demonstrated that veno-occlusion was defective in CASTRATE rats while veno-occlusion was fully functional in TESTO animals. Furthermore, veno-occlusion could be induced in CASTRATE rats if they were first treated with SNP. This observation suggests that failure of veno-occlusion in the CASTRATE rats is due to a deficiency in the production of NO resulting in a reduction in the degree of relaxation of the penile smooth muscle. The measurements of smooth muscle actin and proline and hydroxyproline content of collagen showed that both were unaffected by castration and that the basic structure of the penis did not degenerate after one week without androgenic support. Conclusion: These results can be interpreted to mean that androgens control the veno-occlusive mechanism indirectly via a NO dependent mechanism and not by maintaining the structures of the penis which are essential to veno-occlusion.

1 Introduction

Dilation of the arterioles which carry blood into the cavernous sinuses is essential to the mammalian erectile response. Of equal importance, however, is veno-occlusion, the pressure dependent reduction in the rate at which blood flows out of the sinuses. It is the combination of the increased inflow and the decreased outflow which leads to the high intracavernosal pressures characteristic of penile erection[1]. Studies from this and other laboratories have demonstrated that the erectile response in rats is regulated, in part, by androgens with the hormone acting to control the synthesis and actions of nitric oxide, the principle arteriolar and cavernous smooth muscle relaxing neurotransmitter in erection[2-7]. Despite the importance of veno-occlusion to the erectile response, however, the need for androgenic support of this process has not been well studied although we have previously speculated that testosterone may play a role[8]. In our prior report, veno-occlusion appeared to be compromised in castrated animals as demonstrated by the need for a significantly higher rate of saline infusion into the cavernous sinuses to achieve veno-occlusive pressures in the castrated animals receiving no hormone replacement (CASTRATE) than in castrated animals which were implanted with a testosterone pellet (TESTO). In this experiment, however, no attempt was made to fully relax cavernosal smooth muscle prior to the saline infusion. More recently, we employed a laser Doppler flow meter to demonstrate that, during electrically induced erection, veno-occlusion occurred in TESTO rats but not in CASTRATE animals[9]. If, however, the nitric oxide (NO) donor drug sodium nitroprusside (SNP) was first injected, veno-occlusion did occur in CASTRATE rats. Based on this observation, we suggested that veno-occlusion may be an androgen-dependent process and mediated by NO. The present study was undertaken to confirm that veno-occlusion was an androgen-dependent process by measuring the intracavernosal pressure during infusion of saline into the intracavernosal sinuses at rates adequate to cause veno-occlusion. Furthermore, these measurements were made before and then again after complete relaxation of all penile smooth muscle by the administration of a NO donor drug. The studies also investigated the possibility that differences in veno-occlusion in CASTRATE and TESTO rats were due to differences in the structure of the penis specifically the actin content as a measure of the number of smooth muscle cells and the content of collagen, the protein which is responsible for the compliance of the tunica albuginea.

2 Materials and methods

2.1 Animals

Male Holtzman rats (Harlan Sprague Dawley, Inc. Indianapolis, IN, USA), 6 to 8 months of age, were used in these studies. Animals were castrated under ether anesthesia and implanted subcutaneously with a testosterone (Steraloids, Inc, Wilton, NH, USA) pellet (3 mm diameter, weight 3-5 mg, 1:1, testosterone: cholesterol-group designation: TESTO) or a control pellet of cholesterol only (group designation: CASTRATE). The animals were used in the studies 6 to 8 days after castration. We previously reported that one week of castration led to a significant decline in the magnitude of the erectile response and that prolonging this post-castration period to as long as 7 weeks did not lead to a further decline in the erectile response[8]. For all surgical procedures, rats were anesthetized with intramuscular ketamine (Ketaset, Fort Dodge Labs, Fort Dodge, IA, USA, 87 mg/kg body weight) plus xylazine (Sadazine, Fort Dodge Labs, Fort Dodge, IA, USA, 13 mg/kg body weight) with supplemental ketamine as needed. All procedures were performed in accordance with the Guiding Principles in the Care and Use of Animals approved by the American Physiological Society.

2.2 Measurement of the intracavernosal pressure (CCP)

The procedure used to measure the intracavernosal pressure (CCP) has been previously described from this laboratory[4,8,10,11]. Briefly, in this method, the animals are anesthetized, the left carotid artery is cannulated for continuous monitoring of mean arterial blood pressure (MAP). Once this cannula was in place, the right corpus cavernosum was cannulated by insertion of a 30 gauge needle attached to PE 50 tubing which had been drawn to a fine tip and connected to a pressure transducer to permit continuous monitoring of CCP (Statham P23AC transducers and Grass 7D polygraph from AstroMed, Grass Instrument Division, Braintree, MA, USA). The left corpus cavernosum was cannulated with a 30 gauge needle attached to pump for the infusion of saline at rates ranging from 50 to 1000 L/min into the cavernous sinuses. This cannula was also used for the delivery of 1-2 L of saline containing sodium nitroprusside (SNP, Sigma-Aldrich, St Louis, MO, USA).

2.3 Estimation of the  intracavernosal pressure at which veno-occlusion occurs.

A laser Doppler flow meter (Periflex PF3, PeriMed Stockholm, Sweden) was employed to measure the movement of blood in the cavernous sinuses during erection[9]. The FC 302 laser Doppler flow detection probe was positioned on the left corpus cavernosum at the base of the penis. The proper positioning of the probe was confirmed during ganglionic stimulation when the turbulence caused by partial veno-occlusion could be heard (audible output of the flow meter) as pulses marking the flow spurts during systole. To estimate the intracavernosal pressure (CCP) at which veno-occlusion takes place, the intracavernosal pressure (CCP) and intracavernosal blood flow (CCF) were continuously monitored before and during erection resulting from ganglionic stimulation at 5 V. Figure 1 shows a diagrammatic representation of the CCP and CCF during induced erection. To estimate the CCP at veno-occlusion, the CCP was determined when CCF was maximal (line A, Figure 1), when CCF had declined by 50% (line B) and at the minimal CCF during veno-occlusion (line C). Based on these measurements, we selected 55 mmHg (7.3 kPa), the CCP approximately half way between points B and C, for use in the subsequent saline infusion studies.

Figure 1. A representative tracing of the intracavernosal blood pressure (CCP) and intracavernosal blood flow (CCF) during erection induced by electrical stimulation of the major pelvic ganglion (horizontal black bar). Measurements were made in 6 TESTO rats to determine the CCP values at maximal CCF (A), at half maximal CCF (B) and at minimal CCF (C). Measured average values (meanSEM) were A=(71) mmHg (1.00.2) kPa, B=464 mmHg (6.10.5) kPa, C=(595) mmHg (7.80.6) kPa and based on these results, the value of 55 mmHg (7.3 kPa) was selected as the point at which veno-occlusion occurs.

2.4 Determination of the saline infusion rate required for veno-occlusion

The anesthesized rat was placed on a heating pad and the left corpus cavernosus was cannulated for continuoum measurement of CCP. The right corpus cavernosum received a cannula made from PE 50 tubing attached to a 28 gauge needle connected to a variable speed syringe pump for infusion of saline into the cavernous sinuses. The infusion rate was increased step-wise (50-800 L/min in 50 or 100 L/min steps) until an infusion rate was reached at which CCP rose to 55 mmHg (7.3 kPa) or higher within 3 min indicating veno-occlusion. Once the infusion rate needed for veno-occlusion had been determined for each animal, sodium nitroprusside (SNP-8 g/kg body weight) was injected in 1-2 L saline directly into the cavernous sinuses to achieve complete relaxation of all smooth muscle and the step-wise saline infusion procedure was repeated.

2.5 Determination of the smooth muscle actin content of penile tissue collected from CASTRATE and TESTO rats.

The whole penis was collected from CASTRATE and TESTO rats and rapidly cleaned including the removal of the dorsal vein and the corpus spongiosum. The distal portion was discarded and the remainder was cut into four pieces (left and right crus, left and right shaft) for quick freezing in liquid nitrogen and storage at -80. For actin, the left crus and left shaft were weighed and extracted in TRIS buffer containing 2% SDS using a polytron homogenizer. Following centrifugation to remove particulate matter, an aliquot was taken for determination of the total protein content using the BioRad DC protein assay and the remaining samples were stored at -20. For Western analysis, 10 g of total protein from each sample was loaded on to 12% polyacrylamide Ready Gels (Bio-Rad Laboratories, Hercules, CA, USA) and the proteins were separated by electrophoresis with standard molecular weight markers, a negative control of liver protein and a positive control for actin prepared from rat aorta. The separated proteins were transferred to a nitrocellulose membrane and actin was detected on the membrane by immunochemical staining using a mouse monoclonal first antibody followed by the biotinylated second antibody. Color was developed using the AEC stain (Fisher Scientific, Inc, Pittsburgh, PA, USA). The stained membranes were scanned and density of the actin staining quantified with an IS-1000 Digital Imaging System (Alpha Innotech Corp, San Leandro CA, USA). The amount of actin in each sample was determined relative to the standard prepared from rat aorta smooth muscle.

2.6 Determination of the proline and hydroxyproline content of penile tissues collected from CASTRATE and TESTO rats.

The right crus and right shaft from each penis (CASTRATE and TESTO animals) were weighed, homogenized in cold 5% perchloric acid (PCA) with a polytron homogenizer and the precipitated proteins collected by centrifugation. The protein pellet was washed several times with cold PCA and the final pellet was reconstituted in 4 mL of 6 mol/L HCl. The proteins were hydrolyzed for 24 h at 110 and then dried under a nitrogen stream. The resulting residue was reconstituted in acetonitrile: water (1:1) and derivatives of the amino acids prepared using PITC. The content of proline and hydroxyproline was determined by high performance liquid chromatography (Model 110A, Beckman Instrument Co, Fullerton, CA, USA) using an Econosphere C18 5 micron column and the resulting quantities expressed as nmol/mg tissue.

2.7 Statistical analysis.

Results are expressed as meanSEM. The experiments to determine the infusion rate required to achieve veno-occlusion, the actin content and quantities of proline and hydroxyproline in tissues from CASTRATE and TESTO rats were analyzed statistically by ANOVA and t test[12]. Statistical significance was set at P<0.05.

3 Results

In our prior reports[4,8], circulating levels of testosterone were approximately 1 g/L in castrated rats receiving a single pellet of 50% testosterone (TESTO) while in castrated control rats (CASTRATE), levels were less than 25 ng/L. In the present study, blood was not collected for testosterone measurement but, at necropsy, the sizes of the seminal vesicles, ventral prostate, the dorso-lateral prostate were grossly examined for size and apparent secretory activity. Since these are androgen sensitive tissues, their appearance served to confirm the treatment group to which each animal belonged (ie, CASTRATE or TESTO). 

To determine the intracavernosal pressure at veno-occlusion, measurements were made of CCP at maximal, minimal and half maximal flow (CCF) rates (Figure 1). Based on measurements made in 6 TESTO rats, veno-occlusion must occur when CCP is greater than CCP at maximum inflow of 71 mmHg (1.00.2 kPa) and greater than the half maximal flow rate (B) of 464 mmHg (6.10.5 kPa) but less than 595 mmHg (7.80.6 kPa) at C. For use in the intracavernosal saline infusion studies, the assumption was made that veno-occlusion occurred by the time CCP reached 55 mmHg (7.3 kPa).

Using the value of 55 mmHg (7.3 kPa) as the pressure at which veno-occlusion occurs, saline was infused in a step-wise fashion into the cavernous sinuses until veno-occlusion occurred. Figure 2 showed the perfusion rate mL/(minkg) which was required for veno-occlusion and confirms our prior report that a significantly greater rate of saline infusion is required to achieve veno-occlusion in CASTRATE than in TESTO rats. After measuring these infusion rates, each rat was injected (intra cavernous injection) with sodium nitroprusside (SNP) to cause complete relaxation of all cavernosal and arteriolar smooth muscle in the penis and the step-wise saline infusion protocol was repeated. After injection of this NO donor drug, the infusion rates required for veno-occlusion were significantly reduced in both the CASTRATE and TESTO rats and there was no significant difference between the rates required for veno-occlusion in the two treatment groups.

Figure 2. Effects of castration and testosterone replacement on the rate of saline infusion (mLmin-1kg-1 body weight) required to activate the veno-occlusive mechanism in the rat penis. Infusion rates were determined before and after intra-cavernosal injection of sodium nitroprusside (SNP). Note that before SNP, the required infusion rate for veno-occlusion in CASTRATE rats is significantly higher than the rate in TESTO rats (*). After SNP, the rate in both groups is significantly reduced but there is no difference in the rates between CASTRATE and TESTO animals. Bars represents the meanSEM of 7 TESTO rats or 10 CASTRATE animals.

The results in Figure 3 showed that there was no difference in the actin content of either the shaft or the crus of the penis in CASTRATE and TESTO animals when measured by immunochemical blotting. Similarly, HPLC measurement made in hydrolyzed penile tissue samples revealed no difference between TESTO and CASTRATE rats in the proline or hydroxyproline (Figure 4) content of either the shaft or crus.

Figure 3. Western analysis of the content of actin of the shaft and crus portion of penis collected from TESTO and CASTRATE rats. Each bar represents the SEM of individual measurements made in duplicate on tissues from 4 CASTRATE and 5 TESTO rats. Statistical analysis reveals that there is no significant difference between CASTRATE and TESTO rats in either shaft or crus samples.
Figure 4. The content of proline and hydroxyproline of the shaft and the crus of penis collected from TESTO and CASTRATE rats. Determinations were made of both amino acids in duplicate by HPLC and presented here as the meanSEM of individual measurements made in tissues from 4 CASTRATE and 5 TESTO rats.

There is no significant difference in either amino acid between CASTRATE and TESTO in shaft and crus. There is, however, significantly more proline than hydroxyproline present in both tissue locations.

4 Discussion

Studies from this and other laboratories have shown that the erectile response in the rat is dependent on testosterone[4,7,11,13-15]. After a period of castration lasting 1 week, there is about a 50% decrease in the magnitude of the erectile response when expressed as the intracavernosal pressure relative to mean arterial pressure. If the castrated animals were treated with testosterone[4,13,15] or dihydrotestosterone[14] the erectile response was fully maintained. The mechanism by which androgens maintain erection has been widely investigated and many laboratories have shown that androgens act to support the synthesis of NO[4,5,7,16]. Androgens have also been reported to reduce the sensitivity of the erectile tissue to the constricting effects of sympathetic stimulation[10], may act via a non-NO dependent pathway[17] or act on central erectile mechanisms[18,19]. The full erectile response requires the activation of two processes: 1) relaxation of the smooth muscle in the cavernosal arterioles to permit greater blood flow into the cavernous sinuses. 2) relaxation of the smooth muscle in the cavernous sinuses so that as blood flows in, the sinuses fill and expand against the tunica albuginea. This sinusoidal expansion compresses and partially, or fully, occludes the venous system which carries blood out of the sinuses and the intracavernosal blood pressure rises under the force of the central arterial pressure. We recently reported measurements of the inflow rates during erection which showed that inflow is about twice as great in TESTO rats as in CASTRATE animals demonstrating that the relaxation of the arteriolar smooth muscle is under androgenic regulation[8]. The present study was undertaken to determine if the rate of outflow is under androgenic regulation was well. Erectile dysfunction due to defects in the veno-occlusive mechanisms is normally diagnosed by saline infusion directly into the cavernous sinuses at increasing rates until veno-occlusion is achieved and the flow required to maintain veno-occlusive pressure is determined[20]. If the flow to maintain is high, a venous leak or fistula through which blood is diverted out of the cavernous tissue, may be suspected. Veno-occlusive disease can also result from inadequately or excessively compliant tunica albuginea or from a failure of the cavernous sinuses to expand fully due to a reduced number of smooth muscle cells or to a deficiency in their capacity to respond to the vaso-dilatory neurotransmitters or reduced synthesis and release of the neurotransmitters[21,22]. Using a similar saline infusion technique, we have previously reported that approximately twice as high a saline infusion rate was required to cause veno-occlusion in CASTRATE than in TESTO rats and we suggested a veno-occlusive disorder in the CASTRATE animals[8]. However, in our prior studies, the cavernosal smooth muscle cells were not stimulated to relax prior to saline infusion; incomplete relaxation of these cells could limit the degree of expansion of the cavernous sinuses against the tunica albuginea and thereby prevent full veno-occlusion. The studies of Hatzichristou and coworkers[23] have stressed the importance of injection of vasodilator drugs into the cavernous sinuses prior to attempting to measure the pressure required for veno-occlusion. More recently, we investigated veno-occlusion using a laser Doppler flow meter to follow intracavernosal blood flow. Our studies showed that during electrically induced erection, veno-occlusion occurred in the TESTO rats but not in the CASTRATE animals[9]. Veno-occlusion could, however, be induced in the CASTRATE rats if they were first treated with SNP prior to saline infusion. In this prior study, we did not establish how the additional NO permitted veno-occlusion to occur. The NO could act directly on the arterioles to permit greater blood inflow and/or NO could cause a greater reduction in cavernosal smooth muscle tone and permit expansion of the cavernous sinuses and veno-occlusion. The results presented in Figure 2 confirm our prior report that a greater rate of saline infusion into the cavernous sinuses was required in CASTRATE than in TESTO rats to achieve high intracavernosal pressure. This figure also shows that after treatment with SNP, there was no difference in the saline flow rate required to achieve veno-occlusion in the CASTRATE and TESTO rats. Thus, the difference in veno-occlusion in the CASTRATE and TESTO animals disappears when the NO donor drug was used and this suggests that a deficiency in NO production or action is the underlying basis for the failure of veno-occlusion following castration. The target of the NO could be the arteriolar smooth muscle, the cavernosal smooth muscle or both.

Next, we sought to determine if differences in the physical properties of the penis such as the compliance of the tunica albuginea contribute to the difference in veno-occlusion between TESTO and CASTRATE animals. If the tunica becomes less compliant, the loss of elasticity could interfere with one of the mechanisms of veno-occlusions[1]. If the cavernosal smooth muscle is decreased or less reactive to neurotransmitters, the tissue would be less expansive and thereby reduce the compression of the outflow venous system and reduce veno-occlusion. Measurements were made of the total collagen content of the penile tissue by determining the amount of proline and hydroxyproline (Figure 4) in hydrolyzed proteins collected for CASTRATE and TESTO animals. These two amino acids constitute a significant portion of the collagen molecules which are present in the tunica albuginea and the trabecular matrix that makes up the framework of the erectile tissue. The failure to detect any difference in the quantities of either amino acid suggests that the collagen content is not changed after one week of castration and there is likely no important difference in the compliance of the tunica albuginea after this brief time period. Our studies were also designed to determine if differences in the veno-occlusive response in CASTRATE and TESTO rats were due to differences in the expression of actin protein in the cavernosal smooth muscle cells. Measurement of the quantity of this protein in the penis (Figure 3) shows no difference between the treatment groups indicating that an androgen dependent maintenance of the cavernosal smooth muscle population is likely not an essential part of veno-occlusion under these experimental conditions.

Thus, based on the results of this study, two conclusions appear to be warranted: First, androgens control the veno-occlusive mechanism during erection indirectly, by regulating the tone of the arteriolar and/or cavernosal smooth muscle. This action appears to be mediated by NO. Secondly, the major structural components of the penis essential to veno-occlusion are unaffected by a short period (one week) of castration even though the erectile response is markedly reduced over the same time period.


[1] Andersson KE, Wagner G. Physiology of penile erection. Physiol Rev 1995; 75: 191-236.
[2] Mills TM, Lewis RW. The role of androgens in the erectile response: a 1999 perspective. Mol Urol: A Translational Approach 1999; submitted.
[3] Shabsigh R. The effects of testosterone on the cavernous tissue and erectile function. World J Urol 1997; 15: 21-6. 
[4] Reilly CM, Zamorano P, Stopper VS, Mills TM. Androgenic regulation of NO availability in rat penile erection. J Androl 1997; 18: 110-5.
[5] Chamness SL, Ricker DD, Crone JK, Dembeck CL, Maguire MP, Burnett AL, et al. The effect of androgen on nitric oxide synthase in the male reproductive tract of the rat. Fertil Steril 1995; 63: 1101-7.
[6] Mills TM, Reilly CM, Lewis RW. Androgens and penile erection: a review. J Androl 1996; 17: 633-8. 
[7] Penson DF, Ng C, Cai L, Rajfer J, Gonzalez-Cadavid NF. Androgen and pituitary control of penile nitric oxide synthase and erectile function in the rat. Biol Reprod 1996; 55: 567-74. 
[8] Mills TM, Stopper VS, Wiedmeier VT. Effects of castration and androgen replacement on the hemodynamics of penile erection in the rat. Biol Reprod 1994; 51: 234-8.
[9] Mills TM, Lewis RW, Stopper VS. Androgenic maintenance of inflow and veno-occlusion during erection in the rat. Biol Reprod 1998; 59: 1413-8.
[10] Reilly CM, Stopper VS, Mills TM. Androgens modulate the  adrenergic responsiveness of vascular smooth muscle in the corpus cavernosum. J Androl 1997; 18: 26-31.
[11] Mills TM, Wiedmeier VT, Stopper VS. Androgen maintenance of erectile function in the rat penis. Biol Reprod 1992; 46: 342-8. 
[12] Winer BJ. Statistical principles in experimental design. 2nd ed. New York: McGraw-Hill; 1971.
[13] Garban H, Marquez D, Cai L, Rajfer J, Gonzalez-Cadavid NF. Restoration of normal adult penile erectile response in aged rats by long-term treatment with androgens. Biol Reprod 1995; 53: 1365-72.
[14] Lugg JA, Rajfer J, Gonzalez-Cadavid NF. Dihydro-testosterone is the active androgen in the maintenance of nitric oxide-mediated penile erection in the rat. Endocrinology 1995; 136: 1495-501.
[15] Giuliano F, Rampin O, Schirar A, Jardin A, Rousseau JP. Autonomic control of penile erection: Modulation by testosterone in the rat. J Endocrinol 1993; 5: 677-83. 
[16] Zvara P, Sioufi R, Schipper HM, Begin LR, Brock GB. Nitric oxide mediated erectile activity is a testosterone dependent event: a rat erection model. Int J Impotence Res 1995; 7: 209-19. 
[17] Reilly CM, Lewis RW, Stopper VS, Mills TM. Androgenic maintenance of the rat erectile response via a non-nitric-oxide-dependent pathway. J Androl 1997; 18: 588-94.
[18] Heaton JP, Varrin SJ. Effects of castration and exogenous testosterone supplementation in an animal model of penile erection. J Urol 1994; 151: 797-800.
[19] Hart BL, Wallach SJ, Melese-D'Hospital PY. Differences in responsiveness to testosterone of penile reflexes and copulatory behavior of male rats. Horm Behav 1983; 17: 274-83.
[20] Lewis RW. Venogenic impotence: Diagnosis, management and results. Problems Urol 1991; 5: 567-76.
[21] Christ GJ. The penis as a vascular organ. The importance of corporal smooth muscle tone in the control of erection. Urol Clin NA 1995; 22: 727-45. 
[22] Pickard RS, King P, Zar MA, Powell PH. Corpus cavernosal relaxation in impotent men. Br J Urol 1994; 74: 485-91.
[23] Hatzichristou DG, Saenz de Tejada I, Kupferman S, Namburi S, Pescator
i ES, Udelson D, et al. In vivo assessment of trabecular smooth muscle tone, its application in pharmaco-cavernosometry and analysis of intracavernous pressure determinants. J Urol 1995; 153: 1126-35.

Correspondence to Thomas M Mills, PhD, Professor.
Fax +1-706-721 7299
E-mail: tmills@mail.mcg.edu
Received 1999-03-08     Accepted 1999-05-08