Vasculogenic Female Sexual Dysfunction: the Hemodynamic Basis for Vaginal Engorgement Insuf®Ciency and Clitoral Erectile Insuf®Ciency

Home , Clitoral erection, Clitoris, Erectile tissue, Erection, Tumescence, Vaginal epithelium

Vasculogenic female sexual dysfunction: The hemodynamic basis for vaginal engorgement insuf®ciency and clitoral erectile insuf®ciency

K Park, I Goldstein, C Andry, MB Siroky, RJ Krane and KM Azadzoi

Department of Urology, Boston University School of Medicine and Boston Veteran's Administration Hospital, Boston, MA, USA

Objective: Organic female sexual dysfunction may be related in part to vasculogenic impairment of the hypogastric-vaginal/clitoral arterial bed. The aim was to develop an animal model of vaginal engorgement insuf®ciency and clitoral erectile insuf®ciency. Methods: Pelvic nerve stimulated vaginal engorgement and clitoral erection were achieved in control (normal diet, n 8) and atherosclerotic (balloon injury of aorto-iliac arteries and 0.5% cholesterol diet, n 7) New Zealand White female rabbits. After 16 weeks, novel hemodynamic variables including vaginal wall and clitoral blood ¯ow, vaginal wall and clitoral intracavernosal pressure, vaginal length, vaginal luminal pressure, blood levels of cholesterol and triglycerides, aorto-iliac angiography and vaginal wall and clitoral erectile tissue histology were recorded in the two groups. Results: Concerning pelvic nerve stimulated vaginal hemodynamic changes, there was signi®cantly less increase in blood ¯ow (ml/min/100 gm tissue), wall pressure (mmHg) and length changes (mm) in atherosclerotic (9.3 Æ 3.7, 4.8 Æ 3.8, 67.3 Æ 8.3) compared to control (13.9 Æ 4.5, 5.5 Æ 2.6, 74.1 Æ 10.0) animals respectively. Histologic examination of clitoral erectile tissue demonstrated cavernosal artery atherosclerotic changes and diffuse vaginal and clitoral ®brosis. Aorto-iliac angiography in atherosclerotic animals revealed diffuse moderate to severe atherosclerotic occlusion. Conclusions: Vaginal engorgement and clitoral erection depend on increased blood in¯ow. Atherosclerosis is associated with vaginal engorgement insuf®ciency and clitoral erectile insuf®ciency.

Keywords: female sexual function; vasculogenic female sexual dysfunction; vaginal engorgement; vaginal engorgement insuf®ciency syndrome; clitoral erection: clitoral erectile insuf®ciency

Introduction of the women had arousal or orgasmic dysfunctions.3 Similar to male sexual dysfunction, the prevalence of female sexual dysfunction has been Females have sexual dysfunction. Post-menopausal shown to increase with age and be associated with women often complain of discomfort with inter- the presence of vascular risk factors and the course, dryness of the vagina and diminished development of the menopause.4±6 Post-menopausal vaginal arousal.1,2 Studies comparing sexual dys- women and women with a history of vascular risk function in couples have revealed 40% of the men factors have been shown to have signi®cantly more had erectile or ejaculatory dysfunction whereas 63% complaints of self-reported female vaginal and clitoral dysfunctions than pre-menopausal women or women without vascular risk factors.5 Correspondence: Dr KM Azadzoi, Department of Urology, Despite the existence of widespread clinical B5-45, Boston VA Medical Center, 150 South Huntington pathology in both male and female sexual dysfunc- Avenue, Boston MA 02130. tions, there has been a virtual focus on the male in This manuscript was awarded the Jean-Paul Ginestie Prize at scienti®c investigations. Over the past decade, there the VII World Meeting of the International Society for has been widespread interest in the understanding Impotence Research in 1996. This work was supported by Grant DK 45087 from the of the physiology and neuroanatomy of male erectile Department of Health and Human Services and a grant from tissue, the pathophysiology of disease mechanisms Veterans Affairs Central Of®ce. inducing male erectile dysfunction and the diag- Received 1 June 1996; accepted 6 November 1996 nosis and treatment of male impotence syndromes. Vasculogenic female sexual dysfunction K Park et al 28 Over the same time period, however, there has been suf®ciency syndromes. In this model, supporting a dearth of research concerning the physiology of data from aortic and iliac artery arteriographic and female sexual dysfunction, the pathophysiologic vaginal spongy and clitoral erectile tissue histologic mechanisms of female sexual dysfunction and the studies were performed. It is the ultimate interest of development of diagnostic and therapeutic alterna- this research endeavor to hypothesize the existence tives for af¯icted female patients. of an underlying vascular pathology in female What is the basis for the apparent lack of interest sexual dysfunction (similar to the male) and to in the investigation and treatment of female sexual encourage young investigators to pursue needed dysfunction? While the answer remains complex new scienti®c knowledge in this developing ®eld. and multi-factorial, there have been at least three postulated explanations. There has existed a strong taboo and even a hostility against the laboratory Materials and methods study of female sexual function.7,8 There has existed a lack of support from the major research funding Experimental model agencies.7,8 There has existed a lack of a relatively inexpensive, reliable and appropriate animal model of female sexual dysfunction.7,8 In particular, the The New Zealand White female rabbit (3±3.5 kg) was historic dif®culty with the availability of a female selected as the animal model to examine prelimin- animal model was based on the hypothesis that the ary female vaginal and clitoral sexual physiology prime expression of female sexual dysfunction was and pathophysiology. The animals were divided orgasmic-related and thus the only appropriate into control (n 8) and atherosclerotic (n 7) model for laboratory investigation was the human.7 groups. Control animals received a regular diet There have been few previous studies with while atherosclerotic animals underwent balloon animals concerning female sexual responses.9±12 At injury of the common and external iliac arteries and the level of the pelvic organs, female sexual were placed on 0.5% cholesterol diet. response has been evidenced by the appearance of After 16 weeks, control and atherosclerotic vaginal lubrication along with tumescence of the animals were anesthetized with titrated intravenous clitoris.13 Both are physiologic phenomena asso- phenobarbital and hydrated with intravenous 0.9% ciated with localized autonomic nerve stimulated NaCl. An angiocatheter was placed into the carotid increases in blood ¯ow. Previous investigations of artery for measurement of systemic arterial pressure. such hemodynamic evens have thus far been Through a midline abdominal incision, the abdom- plagued by methodological problems. Measure- inal aorta and both iliac arteries were exposed. ments have been based largely on indirect maneu- Arterial blood ¯ow (ml/min) through the right and vers such as photoplethysmography, thermal left common iliac arteries was monitored by placing clearance and other temperature-based methods.7 an appropriately sized perivascular ¯ow sensor Direct measurement of both vaginal and clitoral around the respective artery. The ¯ow sensor was blood ¯ow using laser Doppler ¯ow probe technol- connected to an ultrasonic ¯owmeter (Transonic ogy following local pelvic nerve stimulation has not, Systems., Ithaca NY). to the best of our knowledge, previously been The pelvic nerve branch to the vagina and clitoris reported. was identi®ed and carefully dissected under the The overall goal of our study was to apply perivesical fat on the postero-lateral aspect of the recognized techniques for the scienti®c study of vagina as the nerve emerged from its associations male erectile dysfunction14,15 to initiate novel in- with the rectum, bladder and uterus. The clitoris vestigation into the physiology of female sexual and corresponding corporal bodies were exposed via function and the pathophysiology of female sexual a midline incision. dysfunction. One speci®c aim was to establish, for the ®rst time, preliminary data concerning a reliable and relatively inexpensive animal model for the Pelvic nerve stimulation physiologic investigation of pelvic-nerve stimulated vaginal engorgement and clitoral erection. In this A Harvard subminiature electrode was placed model, direct measurements of physiologic variables around the pelvic nerve (Figure 1). Unilateral pelvic such as vaginal and clitoral blood ¯ow, vaginal nerve stimulation was performed by square wave spongy and clitoral erectile tissue pressure, vaginal unidirectional stimulation (10 V, 16 Hz, 8 ms) deliv- length changes and vaginal luminal pressure were ered for 20 s with a Grass SD-9 stimulator (Grass able to be performed. A second speci®c aim was to Instruments, Quincy, Massachusetts). These stimu- develop, for the ®rst time, preliminary data con- lation parameters produced reliable and reproduci- cerning a reliable animal model of the pathophy- ble hemodynamic responses. Preliminary frequency siology of atherosclerosis-induced vaginal response studies revealed that maximal blood ¯ow engorgement insuf®ciency and clitoral erectile in- responses were observed at 16 Hz. This frequency of Vasculogenic female sexual dysfunction K Park et al 29 Measurement of vaginal luminal pressure

A dual lumen 9F rectal pressure catheter (Life-Tech) was ®lled with saline, and inserted into the distal 1/ 3 of the vaginal lumen. One port was connected to a pressure transducer to an ampli®er. The other port was ®lled with saline to increase balloon pressure to approximately 10 mmHg and then the port was occluded to create a closed system.

Measurement of vaginal (and penile) length changes during pelvic nerve stimulation

Figure 1 Schematic diagram showing rabbit female genitalia and experimental model. Symphysis pubis is dissected to expose Vaginal length was measured by insertion of 7 mm vaginal tube and the uterus. Experimental setup for the measure- diameter plastic probe. This probe diameter was ment of vaginal and clitoral blood ¯ow and pressure is shown selected because the mean diameter of the rabbit within the box. penis has been recorded to be 9.7 mm in the proximal portion, 7.7 mm in the middle portion pelvic nerve stimulation was similar in magnitude and 6.1 mm in the distal portion. The plastic probe to that found in the male animals14 and was utilized was lubricated and introduced into the vagina until throughout these experiments. the feeling of resistance. Vaginal length was measured from the tip of the plastic probe to the vulval ori®ce before and during pelvic nerve stimulation. Measurement of vaginal wall blood ¯ow and clitoral These measurements were compared to the size of intracavernosal blood ¯ow the erect penis in a separate study utilizing New Zealand white male rabbits (n 5) prior and following pharmacologic stimulation. The penis and Vaginal wall blood ¯ow was measured by laser corresponding corporal bodies were exposed via a Doppler ¯ow probes placed into the vaginal muscu- midline incision. The length of the penis was laris layer within the spongy region of blood-®lled measured from the base of the penis to the top of spaces and vascular smooth muscle. Clitoral intra- glans. The diameter of the penis was measured in cavernosal erectile tissue blood ¯ow was measured three different locations equidistant from the base of with a similar laser Doppler ¯ow probe placed into the penile shaft. the corporal bodies (Figure 1). The ¯ow probes were connected to a laser Doppler ¯owmeter (Transonic Systems, Inc.) which was calibrated against an Injection of papaverine and phentolamine into the internal standard reading ¯ow in units of ml/min/ spongy vaginal muscularis tissue 100 gm of tissue. The laser Doppler probe uses the Doppler shift of a projected beam of laser light that registers on a To determine the hemodynamic effect of vasodilator photodetector. Static tissues will produce no Dop- agents, a 0.2 ml solution of papaverine hydrochlor- pler shift in wavelength but moving red blood cells ide (6 mg) and phentolamine mesylate (0.2 mg) was will produce a shift proportional to the red cell injected into the subepithelial layer of the vagina velocity.16 To determine the effect of acute arterial within the spongy muscularis tissue, approximately occlusion on hemodynamic parameters, aorta was 1 cm proximally from either the laser Doppler ¯ow clamped and released while stimulating the pelvic probe or wall pressure butter¯y needle locations. As nerve and the changes were recorded. a control, 0.2 ml saline solution was injected at the same site in the spongy vaginal muscularis tissue prior to the injection of vasoactive agents. Measurement of vaginal wall pressure and clitoral intracavernosal pressure Blood tests A heparinized 21-gauge butter¯y needle was inserted into the spongy tissue of the vaginal muscu- After hemodynamic studies, arterial blood was laris layer or clitoral intracavernosal erectile tissue obtained from control (n 4) and atherosclerotic and connected through a pressure transducer to an (n 5) animals. The blood levels of cholesterol and ampli®er (Figure 1). triglycerides were analyzed. Vasculogenic female sexual dysfunction K Park et al 30 Radiologic studies neutral buffered formalin and were allowed to ®x overnight at 4C. The tissues were routinely embedded in paraf®n and 3 mm sections were At 16 weeks, the control and atherosclerotic animals prepared. Tissues were stained with hematoxylin underwent angiography of the descending aorta and and eosin or Masson's trichrome stain. iliac arteries. Under titrated pentobarbital anesthe- sia, a Swan-Ganz catheter (4F) was passed through a carotid arteriotomy into the left ventricle. The Swan- Ganz balloon was in¯ated by air, the catheter was Statistical analysis ¯oated downstream and placed above the aortic bifurcation. Anogiography was performed following The results were expressed as mean Æ standard Conray 40 (3±5 ml) injection through the catheter deviation. Nonparametric analysis was performed. into the prospective ilio-hypogastric arteries. During Wilcoxon signed-rank test was used to compare each angiogram, care was taken to position the parameters within the group. Mann±Whitney U-test image intensi®er at the same height. A grid was was used to compare data between the groups. positioned at the level of the spine to provide Differences were considered statistically signi®cant correction for magni®cation error. at P < 0.05.

Histopathologic studies Results

The right and left common iliac arteries were dissected. In addition, the lower portion of the Control Group vagina was dissected from both lateral walls. The clitoris was completely removed and was carefully All eight animals survived 16 weeks and were marked and dissected as proximal, middle and available for investigation. distal. All tissues were coded, placed in 10%

Hemodynamic studies of the vagina: In the quiescent vaginal state, prior to pelvic nerve stimulation, the mean value of vaginal wall blood ¯ow for the eight animals was 5.9 Æ 2.6 ml/min/100 gm tissue. In the engorged vaginal state, following pelvic nerve stimulation, there was an immediate increase in vaginal wall blood ¯ow. The shape of the vaginal wall blood ¯ow changes following pelvic nerve stimulation resembled a bell-shaped curve (Figure 2). The mean peak vaginal wall blood ¯ow value for the eight animals was 13.9 Æ 4.5 ml/min/100 gm tissue (P < 0.05) (Table 1) and occurred a mean of 76.2 Æ 26.1 s following the onset of pelvic nerve stimulation. Following pelvic nerve stimulation, the Figure 2 Effect of pelvic nerve stimulation on vaginal wall blood vaginal wall pressure increased slowly, but signi®- ¯ow and vaginal wall pressure in a control animal. Pelvic nerve stimulation caused signi®cant increase in vaginal blood ¯ow and cantly in the eight animals from a mean value of vaginal wall pressure. 2.8 Æ 2.3 mmHg to a mean value of 5.5 Æ 2.6 mmHg

Table 1 Effect of pelvic nerve stimulation (PNS) between control and atherosclerotic group

Control group (n 8) Atherosclerotic group (n 5)

Before PNS After PNS Before PNS After PNS

Mean arterial pressure (mmHg) 69.3 Æ 11.7 70.3 Æ 11.3 82.6 Æ 6.1 82.8 Æ 6.1** Vaginal blood ¯ow (ml/min/100 gm tissue) 5.9 Æ 2.6 13.9 Æ 4.5* 5.1 Æ 3.1 9.3 Æ 3.7*,** Vaginal wall pressure (mmHg) 2.8 Æ 2.3 5.5 Æ 2.6* 3.9 Æ 3.2 4.8 Æ 3.8** Vaginal length (mm) 64.6 Æ 9.2 74.1 Æ 10.0* 63.8 Æ 8.0 67.3 Æ 8.3** Intraclitoral pressure (mmHg) 6.3 Æ 3.4 9.7 Æ 3.7* 7.9 Æ 3.8 9.4 Æ 4.6

Data is presented as mean Æ standard deviation. *Statistically signi®cant (P < 0.05) difference comparing parameters before and after PNS in the same group. **Statistically signi®cant (P < 0.05) difference between control and atherosclerotic group. Vasculogenic female sexual dysfunction K Park et al 31

Figure 3 Effect of pelvic nerve stimulation on vaginal luminal pressure. Vaginal luminal pressure was immediately diminished after pelvic nerve stimulation, and it was recovered slowly to the previous pressure value.

Figure 5 This ®gure shows the effect of papaverine and phentolamine administration into the vaginal muscularis layer. Following injection remarkable increase in vaginal wall blood ¯ow and vaginal wall pressure was observed.

Injection of saline solution (0.2 ml) into the vaginal spongy muscularis layer did not show any increase in vaginal blood ¯ow and pressure. On the other hand, injection of the vasoactive agent papaverine and phentolamine solution (0.2 ml) into the spongy vaginal muscularis layer demonstrated increases in vaginal wall blood ¯ow compared to baseline values and increases in vaginal wall pressure comparable to those pressure values achieved by pelvic nerve stimulation (Fig- ure 5).

Figure 4 Effect of mechanical aortic occlusion on vaginal wall blood ¯ow. While stimulating the pelvic nerve, descending aorta Hemodynamic studies of the clitoris: Clitoral in- was clamped. Aortic clamping induced remarkable drop in tracavernosal blood ¯ow showed signi®cant in- vaginal wall blood ¯ow. After releasing the clamp vaginal wall creases following pelvic nerve stimulation, from a blood ¯ow showed complete recovery to previous baseline value. mean value of 2.04 Æ 0.8 ml/min/100 gm tissue at baseline ¯accid clitoral state to a mean peak value of 4.02 Æ 1.5 ml/min/100 gm tissue during clitoral erec- (P < 0.05) (Figure 2, Table 1). The mean delay in the tion (P < 0.05). The nerve stimulated clitoral intra- increase of vaginal wall pressure from baseline cavernosal blood ¯ow changes were similar to those values following the onset of pelvic nerve stimula- of the vaginal wall in that they were immediately tion was 48.0 Æ 20.8 s. timed to the onset of pelvic nerve stimulation and Vaginal luminal pressure was immediately di- they were bell-shaped (Figure 6A, Table 1). The minished with pelvic nerve stimulation, and recov- mean time from onset of pelvic nerve stimulation to ered slowly to the original pressure value over a the mean peak clitoral intracavernosal blood ¯ow period of minutes (Figure 3). was 41.0 Æ 24.8 s. Vaginal length also increased signi®cantly from Clitoral intracavernosal pressure also increased mean baseline values of 64.6 Æ 9.2 mm in the eight signi®cantly from a mean value of 6.3 Æ 3.4 mmHg in animals to 74.1 Æ 10.0 mm following pelvic nerve the ¯accid clitoral state to a mean value of stimulation (P < 0.05) (Table 1). In the separate male 9.7 Æ 3.7 mmHg in the erect state (P < 0.05) (Figure study, the mean ¯accid length of the male penis was 6B, Table 1). The mean time from onset of pelvic 29.4 Æ 1.9 mm and the mean erect penile length was nerve stimulation to the onset of clitoral intracaver- 46.8 Æ 3.6 mm. nosal pressure changes from baseline values was The effect of acute mechanical aortic occlusion on 46.0 Æ 22.7 s. There was no measurable increment of vaginal wall blood ¯ow revealed that aortic clamp- clitoral length following pelvic nerve stimulation. ing induced a sudden drop in vaginal wall blood ¯ow to zero blood ¯ow values. Following the releasing of the aortic clamp, vaginal wall blood Angiography and histology of the aorto-iliac ar- ¯ow showed complete recovery to previous baseline teries: In angiographic studies on control animals, values (Figure 4). there was no signi®cant differences in the luminal Vasculogenic female sexual dysfunction K Park et al 32

Figure 6 Effect of pelvic nerve stimulation on intraclitoral blood ¯ow and intraclitoral pressure in control animals. Pelvic nerve stimulation caused signi®cant increase in intraclitoral blood ¯ow (A) as well as a signi®cant increase in intraclitoral pressure (B).

Figure 8 Histology of the common iliac artery in the control (A) and atherosclerotic animal (B). In the control animals arterial Figure 7 Angiogram of aorto-iliac arteries in the control (A) and lumen, the intima and media were normal. In the atherosclerotic atherosclerotic (B,C) animal. (A) shows the normal angiogram. (B) animals arterial lumen was narrowed by atherosclerotic lesion, shows moderate occlusion of the common iliac, external and the media was thickened with in®ltration of foam cells (H-E stain, internal iliac arteries. (C) shows severe occlusion. The left and Â40). right external iliac arteries were totally occluded. The right common iliac artery showed aneurysm. Right and left internal iliac arteries showed diffuse atherosclerotic occlusion. Hemodynamic studies of the vagina: Pelvic nerve diameters of the right and left common iliac and stimulated increases in vaginal wall blood ¯ow in internal iliac arteries (Figure 7A). the ®ve atherosclerotic animals were signi®cantly Histologic examination of the right and left different from a mean value of 5.1 Æ 3.1 ml/min/ common iliac and internal iliac arteries also showed 100 gm tissue in the baseline quiescent vaginal state normal lumina, intima and media (Figure 8A). to 9.3 Æ 3.7 ml/min/100 gm tissue in the engorged vaginal state (P < 0.05). The mean baseline vaginal Histologic ®nding of the vagina and clitoris: Cross blood ¯ow values in the atherosclerotic group were section of the lateral vaginal wall showed smooth similar to those of the control group. The mean peak muscle, connective tissue and a vasculature with vaginal blood ¯ow values were however signi®- prominent venous channels (Figure 9). The corpora cantly smaller than those of control group. The cavernosa of the clitoris showed normal trabecular typical bell-shaped blood ¯ow increases following structure (similar to the penis) with smooth muscle, pelvic nerve stimulation in the controls were not connective tissue and cavernosal artery (Figure identi®ed in the atherosclerotic animals. In the 10A). atherosclerotic group, the blood ¯ow response appeared muted, ¯attened and damped compared to control patterns (Figure 11, Table 1). Pelvic nerve stimulated increases in vaginal wall Atherosclerotic group pressure were recorded from baseline values of 3.9 Æ 3.2 mmHg to engorged values of Five of the 7 atherosclerotic animals survived 16 4.8 Æ 3.8 mmHg. Pelvic nerve stimulated increases weeks. in vaginal length were observed from 63.8 Æ 8.0 mm Vasculogenic female sexual dysfunction K Park et al 33

Figure 9 Histology of the vagina in the control animal. Cross section of the lateral vaginal wall shows normal structure of smooth muscle, connective tissue and vasculature. (Masson's trichrome stain, Â40).

to 67.3 Æ 8.3 mm (P 0.066). In both vaginal wall pressure and vaginal length measurements these changes were not considered signi®cant (Table 1). Comparing with similar values in the control group, vaginal blood ¯ow, vaginal wall pressure and vaginal length changes following pelvic nerve stimulation in the atherosclerotic group were signi®cantly lower (P < 0.05) (Table 1).

Hemodynamic studies of the clitoris: Clitoral intracavernosal pressure changed from 7.9 Æ 3.8 mmHg to 9.4 Æ 4.6 mmHg after pelvic nerve stimulation (Table 1). In the atherosclerotic compared to the control group, the rise in clitoral intracavernosal pressure following nerve stimulation tended to be lower but the mean value in 5 animals did not reach signi®cance.

Blood test: In the atherosclerotic animals compared to controls, blood levels of cholesterol (14 Æ 9 and 1005 Æ 501 mg/dl respectively) and triglyceride (55 Æ 18 and 127 Æ 119 mg/dl respectively) were remarkably greater. Figure 10 Histology of the clitoris in the control (A) and atherosclerotic animal (B,C) at the proximal part. (A) shows normal trabecular structure of smooth muscle, connective tissue and a normal cavernosal artery. (B) shows thickened cavernosal Angiography and histology of the aorto-iliac ar- artery with an increase in size and number of collagen bundles teries: Angiographic study demonstrated athero- compared to control. (C) shows nearly occluded cavernosal artery sclerotic occlusive disease and showed moderate and severe ®brosis. (Masson's trichrome stain, Â100). (3 of 5 cases, Figure 7B) to severe (2 of 5 cases, Figure 7C) atherosclerotic lesions in common iliac arteries and internal iliac arteries. Histologic ®nd- ings revealed intimal thickening associated with collagen bundles as compared to the control group. in®ltration of foam cells (Figure 8B). The stroma of the corpora also appeared to be more edematous than control. Of note, as compared to control (A), the cavernosal artery in atherosclerotic Histologic ®nding of the clitoris: The corpora clitoris animals was frequently thickened (B) or occasion- contained an increase in size and number of ally obstructed (C) (Figure 10). Vasculogenic female sexual dysfunction K Park et al 34 pidly expand the physiological/pathophysiological knowledge data base of female sexual dysfunction. The ultimate objective would be to treat in the future all sexual dysfunctions, male and female, as a `couple's disease', much like infertility. It is anticipated that once new information on female sexual physiology and pathophysiology is available, future management strategies such as duplex Doppler investigations and pharmacological administration of vasodilators, considered routine in impotent men will emerge in af¯icted women. One major impediment to the generation of Figure 11 Vaginal wall blood ¯ow and vaginal wall pressure in scienti®c data on female sexual function and an atherosclerotic animal. Although pelvic nerve stimulation dysfunction has been the lack of a reproducible, caused signi®cant increase in vaginal blood ¯ow, peak vaginal blood ¯ow during pelvic nerve stimulation was signi®cantly readily available and relatively inexpensive animal smaller compared with the control group. Pelvic nerve stimula- model. In this study, we have preliminary devel- tion failed to increase in vaginal wall pressure. oped an animal model with the New Zealand White female rabbit for pelvic nerve-stimulated vaginal Discussion and clitoral function studies. The New Zealand White male rabbit has been widely utilized for the study of erectile function in man18 and comparison Atherosclerotic vascular disease of the hypogastric- between the sexes to similar pelvic nerve stimulated cavernous arterial bed in the male, with risk factors physiologic events, such as blood ¯ow, smooth of increasing age, diabetes, hypertension, heart muscle tone, geometric and pressure changes can disease, cigarette smoking and hypercholesterole- be made. The vagina in New Zealand White female mia, is considered a major pathophysiology for has a unique structure from other mammals includ- organic vasculogenic erectile dysfunction, af¯icting ing human (Figure 1).19 In contrast to the human an estimated 10±20 million American men.17 It is vagina which is 70±90 mm in length,20 the `com- possible that atherosclerotic vascular disease of the plete' rabbit vagina measures about 130±140 mm. In hypogastric-vaginal/clitoral bed in post-menopausal the rabbit, however, the typical strati®ed squamous women will be found to result in similar female vaginal epithelium was found in only the lower 1/3 organic vasculogenic sexual dysfunction syndromes of the vagina from the level of the pubic symphysis to be called `vaginal engorgement insuf®ciency' and up to the vestibulum.19 The remaining 2/3 of the `clitoral erectile insuf®ciency'. These syndromes vagina serves as a common conduit for the two will consist of impaired hemodynamic responses cervices and two uteri in the female rabbit.19 In this to sexual efferent autonomic pelvic nerve stimula- study, the `effective' vaginal length was about tion in women with similar vascular risk factor 65 mm, just slightly larger than the erected rabbit exposure. It is likely that future epidemiologic penile length of approximately 50 mm in length. studies will identify these organic vasculogenic The human vagina receives arterial blood supply female sexual dysfunction syndromes to exist in from the vaginal artery, the vaginal branch of the millions of aging American women whose present- uterine artery, the internal pudendal artery, and the day symptoms of vaginal discomfort with coitus, vaginal branches of the middle rectal artery. All of dryness, diminished vaginal size and diminished the above arteries form a network of anastomoses sexual arousal are not well-appreciated and are located in the adventitia and extensive numbers of presently managed empirically with either hor- arteriovenous anastomoses exist.20,21 Direct record- mones (with risks of breast cancer) and/or lubrica- ing of vaginal wall blood ¯ow was able to be made in tion. this model. In the quiescent vaginal state prior to The overall ®eld of organic female sexual dys- pelvic nerve stimulation, the mean vaginal blood function is in great and urgent need of basic ¯ow was recorded as 5.9 ml/min/100 gm. In the laboratory investigation. These last few decades vaginal engorged state following pelvic nerve sti- have witnessed the application of modern molecular mulation, the mean vaginal blood ¯ow was recorded biologic, biochemical, physiological, neuroana- as 13.9 ml/min/100 gm tissue, an increase in blood tomic, engineering, histopathologic and numerous ¯ow of 136%. In contrast, baseline and erect arterial other methodologies to the study of male erectile blood ¯ow into the rabbit penis is 8 Æ 3 ml/min/ dysfunction. Although there has been little previous 100 gm tissue and 25 Æ 4 ml/min/100 gm (213% scienti®c evaluation of pelvic nerve stimulated increase) respectively following vasoactive agent vaginal engorgement and clitoral erectile mechan- administration.15 In addition, the delay from pelvic isms, there is no logical reasons why the multiple nerve stimulation to peak blood ¯ow in the vagina available technologies cannot be applied and ra- (approximately 75 s) is 1.75 times the delay in the Vasculogenic female sexual dysfunction K Park et al 35 penis. This may re¯ect the voluminous spongy layer phase of sexual excitement associated with vaginal of the vaginal wall compared to the relatively low smooth muscle relaxation.23 It is well known that volume penile corpora cavernosa. nitric oxide is an important neurotransmitter asso- The vaginal structural anatomy re¯ects functional ciated with corporal smooth muscle relaxation in needs. The prime function of the pelvic nerve penile erection.24,25 To our knowledge, there has stimulated vagina is to prepare for painless penile been no report demonstrating the role of nitric oxide penetration. There is no tunica albuginea to help as a neurotransmitter in the vagina, but there is trap blood as there is no purpose to achieve vaginal evidence that nitric oxide may be related in the rigidity. The histologic structure of the human clitoral erection.26 vagina consists mucosa, muscularis and ®brosa.21 The clitoris is the homologue of the penis. It is a The larger blood vessels run in the deeper portion of cylindrical, erectile organ composed of the glans, the mucosa, which give off branches into capillary corporal body and the crura. The corporal body is networks in the stroma and muscularis.21 These surrounded by a ®brous sheath, tunica albuginea, capillaries empty into the veins that form a plexus of which encases cavernosal tissue consisting of broad venous channels in the muscularis.21 In the sinusoids and surrounding smooth muscle.27 The rugae, there are large veins which give them the clitoris responds to sexual excitement by tumes- spongy character of male erectile tissue.21 The outer cence and erection, although this does not occur part of the vagina, the ®brosa, consists of dense with the degree of pressure elevation as found connective tissue with many coarse elastic ®bers.21 during penile erection. In our observations, pelvic Following pelvic nerve stimulation there is an nerve stimulation increased clitoral blood ¯ow increase in arterial vaginal blood in¯ow which is approximately 100% from 2 to 4 ml/min/100 gm initially not matched by rapid venous drainage, thus tissue. The characteristics of the clitoral blood ¯ow allowing for the vaginal spongy muscularis layer to increases paralleled those of the male. They were become engorged with blood.7 Vaginal wall pressure immediately timed on the onset of pelvic nerve increases document this vascular engorgement. The stimulation and there was a characteristic 41 second basis for increase venous out¯ow resistance en- delay to peak blood ¯ow values. abling this vascular engorgement is not yet appre- Pelvic nerve stimulation also resulted in in- ciated. Vaginal engorgement is related to vaginal creased clitoral intracavernosal pressure. In the lubrication which is known as one of the ®rst pelvic rabbit model, the mean baseline clitoral intracaver- responses to sexual stimulation.13 nosal pressure was 6.3 mmHg which increased by The vagina is associated with three layers of 54% to 9.7 mmHg following pelvic nerve stimula- muscle; the vaginal smooth muscle, the striated tion. These results are nearly the same as was muscles encircling the introitus such as ischioca- reported in the dog model.12 The physiologic vernous and bulbospongiosus, and the striated difference in corporal veno-occlusion in men and muscles of the pelvis such as Levator ani.7 In our women is likely to be due to the different functional studies, the vaginal muscularis layer underwent roles between the penile and clitoral corpora relaxation following pelvic nerve stimulation. This cavernosa: rigidity and creation of a cylindrical was evident by the immediate diminished vaginal insertion tool in men versus tumescence, support luminal pressure and by the approximately 1 cm and glans extrusion in women.28 elongated vaginal length. Both effects of smooth It is well documented that sexual dif®culties in muscle relaxation assist with the vaginal function of women increase sharply following menopause.1,2 painless penile penetration. In particular, vaginal Anatomic changes of the post-menopausal vagina elongation can be explained by active and passive wall have been associated with estrogen depriva- motions. The active motion may be caused by tion.13 In particular, the post-menopausal female's vaginal smooth muscle relaxation and vaginal vaginal tissues are thinner, the vagina barrel is engorgement with blood. The passive motion may shortened, and the uterus becomes atrophic and be related to the contraction of levator ani muscle ®brous.13 An atherosclerotic animal model of im- via pudendal nerve stimulation. Levator ani muscle potence with the New Zealand White rabbit has contraction is known to cause widening of the been well established.29 This model has recently vaginal introitus, vaginal elongation and ballooning demonstrated that ischemia-induced corporal veno- of the upper vagina as well as uterine elevation.22 occlusive dysfunction might be the result of altera- In the human, vaginal nerves are derived from the tions in corporal smooth muscle relaxation and/or uterovaginal plexus. Sympathetic, parasympathetic changes in the structure and ®broelastic properties and afferent ®bers pass through this plexus.20 of erectile tissue.30 Vasoactive intestinal peptide containing nerve ®bers We preliminary developed a similar atherosclero- are abundantly located in the vaginal wall and tic animal model to study pathophysiology of vasoactive intestinal peptide may play an important vasculogenic female sexual dysfunction. There were role as a neurotransmitter in the increased vaginal several noted observations from this atherosclerotic blood ¯ow and vaginal lubrication during the initial animal model. First, the vaginal blood ¯ow in- Vasculogenic female sexual dysfunction K Park et al 36 creases following pelvic nerve stimulation were female sexual function and the pathophysiology of signi®cantly less than those of the control group. female sexual dysfunction. The female clitoris and Second, the vaginal wall pressure increases follow- vagina undergo physiologic changes during sexual ing pelvic nerve stimulation were not signi®cant in stimulation. Engorgement and elongation of vagina the atherosclerotic animals. Third, the vaginal and clitoris depends on increased vaginal and length changes following pelvic nerve stimulation clitoral blood ¯ow. Atherosclerosis inhibits vaginal were signi®cant decreased compared to those of the and clitoral engorgement and elongation and leads control group. Fourth, the clitoral intracavernosal to ®brotic histologic features comparable to those pressure increases following pelvic nerve stimula- observed in menopausal states. The existence of tion were not signi®cant in the atherosclerotic vaginal engorgement insuf®ciency and clitoral erec- animals. Such preliminary data supports the ex- tile insuf®ciency syndromes appear likely. Organic istence of atherosclerosis-induced vasculogenic fe- vasculogenic female sexual dysfunction may play a male sexual dysfunction, that is, the functional signi®cant role in the genesis of a couple's sexual changes which occur in the vagina and clitoris dysfunction. It is anticipated that future manage- following pelvic nerve stimulation are interfered ment strategies for af¯icted women with vasculo- with by arterial vascular disease in the hypogastric- genic sexual dysfunction, as they do for men with vaginal/clitoral arterial bed. similar vasculogenic sexual dysfunction, may in- In this atherosclerotic model, elevated levels of clude duplex Doppler investigations and pharma- blood cholesterol and trigcerides were observed. In cologic administration of vasodilators. addition, aorto-iliac angiography and histology con®rmed the presence of arterial occlusive disease in the hypogastric-vaginal/clitoral arterial bed (Fig- References ure 7B,C, 8B and 10B,C). Especially in the clitoris, atherosclerotic lesions in the cavernosal artery were well documented and loss of corporal smooth 1 Semmens JP, Semmens EC. Sexual function and the meno- muscle tissue with replacement by collagen con- pause. Clin Obstet Gynecol 1984; 27: 717±723. 2 Mongomery JC, Studd JW. Physiological and sexual aspects of nective tissue was well demonstrated. Atherosclero- the menopause. Brit J Hosp Med 1991; 45: 300±302. tic vaginal wall ®brosis was more dif®cult to 3 Frank E, Anderson C, Rubinstein D. Frequency of sexual establish in this model primarily because in the dysfunction in `normal' couples. N Engl J Med 1978; 299: 111± control animal, vaginal smooth muscle bundles 115. 4 Goldstein MK, Teng NN. Gynecologic factors in sexual were routinely widely separated by connective dysfunction of the older women. Clin Geriatr Med 1991; 7: tissue (Figure 9). In the atherosclerotic vaginal wall, 41±61. there was the impression of increased connective 5 Sadeghi-Nejad H et al. Impotence is a couple's disease: studies tissue but because of the wide separation of smooth in female sexual dysfunction. J Urol 1996; 155: 677A. muscle and connective tissue under control condi- 6 Slob AK et al. Sexuality and psychophysiological functioning in women with diabetes mellitus. J Sex Marital Ther 1990; 16: tions, any increase in connective tissue due to 59±69. atherosclerosis has been thus far complicated to 7 Levin RJ. The mechanisms of human female sexual arousal. accurately quantify. Annu Rev Sex Res 1992; 3: 1±48. This preliminary atherosclerotic model may be 8 Levin RJ. The physiology of sexual function in women. Clin Obstet Gynecol 1980; 7: 213±251. developed in the future in a more sophisticated 9 Allen ML, Lemmon WB. Orgasm in female primates. Am J manner. To make an atherosclerotic animal model in Primatol 1981; 1: 15±34. the female (as compared to the male) hormonal 10 Dixon AF. Genital sensory feedback and sexual behaviour in milieu should be considered. Women in the repro- male and female marmosets (Callitrix jacchus). Physiol Behav ductive age group are protected from coronary artery 1986; 37: 447±450. 31 11 McKenna KE, Chung SK, McVary KT. A model for the study of disease when compared to men. Estradiol is sexual function in anaesthetized male and female rats. Am J known to exert a protective effect against the Physiol 1991; 261: R1276±R1285. development of atherosclerosis, but the mechanism 12 Diederichs W, Lue TF, Tanagho EA. Clitoral response to of this hormonal action is unknown. Oophorectomy cavernous nerve stimulation in dogs. Int J Impot Res 1991; 3: 7±13. diminishes both circulating estradiol concentration 13 Weisberg M. Physiology of female sexual function. Clin Obstet and basal release of nitric oxide to levels seen in Gynecol 1984; 27: 697±705. male rabbits.32 In the future, we plan to perform 14 Azadzoi KM, Vlachiotis J, Pontari M, Siroky MB. Hemody- oophorectomy prior to inducing balloon and diet- namics of penile erection: III. measurement of deep intraca- associated arterial endothelial injury. vernosal and subtunical blood ¯ow and oxygen tension. J Urol 1995; 153: 521±526. 15 Azadzoi KM et al. Relationship between cavernosal ischemia and corporal veno-occlusive dysfunction in an animal model. J Conclusion Urol 1997; 157: March (in press). 16 Stern MD. In vivo evaluation of microcirculation by coherent light scattering. Nature 1975; 254: 56±58. 17 Feldman HA et al. Impotence and its medical and psycholo- An animal model has been preliminarily developed gical correlates: results of the Massachusetts male aging study. that facilitates the investigation of the physiology of J Urol 1994; 151: 54±61. Vasculogenic female sexual dysfunction K Park et al 37 18 Stief CG et al. The rabbit as a model for neurologic studies of oxide synthase isoforms in human clitoris. J Urol 1996; 155: the lower genitourinary tract. World J Urol 1990; 8: 233±236. 625A. 19 Barberini F, Correr S, De Santis F, Motta PM. The epithelium 27 Goss CM. The external genital organs. In: Gray's Anatomy of of the rabbit vagina: a microtopographical study by light, the Human Body, 29th edn. Lea & Febieger: Philadelphia, transmission and scanning electron microscopy. Arch Histol 1973, 1330±1332. Cytol 1991; 54: 365±378. 28 Pescatori ES, Engelman JC, Davis G, Goldstein I. Priapism of 20 Moore KL. The female genital organs. In: Clinically Orientated the clitoris: a case report following trazodone use. J Urol 1993; Anatomy. 2nd edn. Williams & Wilkins: Baltimore, 1985, 368± 149: 1557±1559. 372. 29 Azadzoi KM, Goldstein I. Erectile dysfunction due to athero- 21 Copenhaver WM, Kelly DE, Wood RL. The female reproduc- sclerotic vascular disease: the development of an animal tive system. In: Bailey's Textbook of Histology. 7th edn. model. J Urol 1992; 147: 1675±1681. Williams & Wilkins: Baltimore, 1978, 645±692. 22 Sha®k A. Vagino-levator re¯ex: description of a re¯ex and its 30 Azadzoi KM, Siroky MB, Goldstein I. Study of the etiologic role in sexual performance. Eur J Obstet Gynecol Reprod Biol relationship of arterial atherosclerosis to corporal veno- 1995; 60: 161±164. occlusive dysfunction in the rabbit. J Urol 1996; 155: 1795± 23 Wagner G. Aspects of genital physiology and pathology. 1800. Semin Neurol 1992; 12: 87±97. 31 Kannel WB, Hjortland MC, McNamara PM, Gordon T. 24 Rajfer J et al. Nitric oxide as a mediator of relaxation of the Menopause and risk of cardiovascular disease. Ann Int Med corpus cavernosum in response to nonadrenergic, noncholi- 1976; 85: 447±452. nergic neurotransmission. N Engl J Med 1992; 326: 90±94. 32 Hayashi T, Fukuto JM, Ignarro LJ, Chaudhuri G. Basal release 25 Snyder SH. Nitric oxide: ®rst in a new class of neurotrans- of nitric oxide from aortic rings is greater in female rabbits mitters? Science 1992; 257: 494±496. than in male rabbits: implications for atherosclerosis. Proc 26 Calvin DC et al. Immunohistochemical description of nitric Natl Acad Sci 1992; 89: 11259±11263.