|Home | About | Journals | Submit | Contact Us | Français|
We tested the hypothesis that extended-term (5-week) estrogen therapy would negatively impact voiding function in a postpartum, ovariectomized rat model.
Immediately after delivery, 30 primiparous Sprague–Dawley rats underwent intravaginal balloon dilation, followed by ovariectomy 1 week later. Cystometry at postpartum week 2 determined normal or abnormal voiding patterns. After randomization, one-half the normal and abnormal voiding rats received 5 weeks of estrogen therapy, while the remainder received placebo. Estrogen effect was determined by repeat cystometry and immunohistochemical analysis of the urethra and vagina.
Abnormal voiding increased from 60.0% to 73.3% in the estrogen- treated group and declined from 60% to 33% for the placebo group. Rats were then divided into 4 groups for comparison: normal voiding versus placebo (group 1), abnormal voiding versus placebo (group 2), normal voiding versus estrogen (group 3) and abnormal voiding versus estrogen (group 4). Bladder capacity, leak point pressure and maximum voiding pressure were most depressed in group 4. Estrogen treatment was associated with a significant downregulation of α1A and α1D-adrenoceptors in the urethral submucosa but an upregulation of nNOS in the urethral smooth muscle.
Extended-term estrogen therapy in a rat model of simulated birth trauma and ovariectomy resulted in a higher rate of incontinence. Immunohistochemical examination demonstrated significant downregulation of urethral α1A- and α1D-adrenoceptors and upregulation of neuronal nitric oxide synthase (nNOS) in the urethra of estrogen-treated groups. These studies question the use of hormone replacement therapy in the treatment of postmenopausal incontinence.
After detrusor overactivity, stress urinary incontinence (SUI) is the second most common type of urinary leakage in the elderly. SUI in women has an overall prevalence of approximately 10%–30% between the ages of 15 and 64 years and results in significant patient morbidity and billions of dollars in health care expenditures each year.1 SUI is defined as the involuntary loss of urine through the urethral meatus when bladder pressure exceeds urethral pressure in the absence of detrusor activity; risk factors include pregnancy, childbirth, parity, hysterectomy, menopause, obesity, diabetes, chronic respiratory disease and conditions related to impaired pelvic floor innervation.2
Hormonal changes linked with aging may also contribute to the development of urinary disorders. An increase in urinary symptoms, including incontinence, frequency, urgency and recurrent infection, has been reported after menopause.3 Several small scale clinical trials support the commonly held notion that exogenous estrogen may improve urinary function and decrease incontinence rates.4,5,6 Hormone replacement therapy (HRT) has thus prevailed as a leading clinical treatment to reverse genital atrophy and improve urethral closing pressure.
Recently, long-term studies involving thousands of women indicated that hormonal therapy neither improves continence in women with existing urinary incontinence nor prevents the development of incontinence in continent individuals. In the Heart and Estrogen–Progestin Replacement Study (HERS), a randomized, blinded study of 2763 postmenopausal women under age 80 years with coronary disease and intact uteri, the physiological effects of HRT were evaluated.7 Participants were randomly assigned to 0.625 mg of conjugated estrogens plus 2.5 mg of medroxyprogesterone acetate in 1 daily tablet or placebo and were observed for a mean of 4.1 years. Incontinence improved in 26% of the women assigned to placebo versus 21% assigned to the hormone group, while only 27% of the placebo group worsened, compared with 39% of the hormone group (p = 0.001). Based on significant findings evident by 4 months of therapy, the authors concluded that HRT is not recommended for the treatment of incontinence.7 Additionally, during clinical testing of levormeloxifene, a selective estrogen receptor modulator (SERM), women who received the drug had a marked increase in urinary incontinence, compared with those who took placebo (17% v. 4%) as well as an increase in voiding frequency, endometrial thickness and uterine enlargement. Secondary to the number and severity of adverse events, the trial was prematurely halted after 10 months.8
Over the past 7 years, our laboratory has investigated the influence of simulated birth trauma and hormonal deficiency on urinary continence in a postpartum, ovariectomized female rat model.9–11 We determined that prolonged labour and ovariectomy induced higher rates of SUI versus ovariectomy alone. Further, our results confirmed that muscular and neurological alterations in the urethra and pelvic floor likely play a role in the development of SUI.11
Although scientific data are available regarding the effects of female hormones on the vagina, uterus and bladder, examination of the urethra has largely been neglected to date. The objectives of this study were to evaluate the effect of estrogen replacement on voiding function in our postpartum, ovariectomized rat model and identify the immunohistochemical alterations of the urethra and vagina.
This study was approved by the local ethical committee for experimentation (University of California, San Francisco, Institutional and Animal Care Use Committee) and complied with National Institutes of Health (NIH) regulations for the care and use of laboratory animals. Thirty primiparous pregnant 2-month-old Sprague–Dawley female rats at gestational day 16 (230–280 g) were housed at a constant 16°C room temperature and 47% humidity, with a 12-hour light-dark cycle. Immediately postpartum, animals underwent intravaginal balloon dilation under anesthesia and 1 week later underwent bilateral ovariectomy (Fig. 1). Two weeks postpartum, initial transvesical cystometry was performed. Based on the urodynamic findings, each rat was classified as having normal (NV) or abnormal voiding (AV). After cystometry, AV animals outnumbered the NV group (1812), and we felt that the comparison would be difficult after estrogen treatment. Therefore, the 2 groups (NV v. AV) were randomly divided in half and were cross-mixed, establishing 2 equal groups of 15 animals each. At the 3-week mark, 1 group underwent dorsal subcutaneous implantation of a 0.5-mg 17 β-estradiol pellet, while the other group received placebo. Estrogen or placebo pellet placement was repeated at 6 weeks to maintain constant drug bioavailability. Repeat cystometric results were obtained 7.5 weeks after balloon dilation, and the rats were again evaluated with urodynamic studies. For final outcome analyses, animals were grouped as follows 1) NV plus placebo; 2) AV plus placebo; 3) NV plus estrogen; 4) AV plus estrogen (Fig. 1). At sacrifice, the pelvic tissues, midurethra and vagina were harvested. The midurethra was chosen because it represents the thickest region of the urethra with both striated and smooth muscle components and has demonstrated the least anatomic variation in our previous studies.
Immediately postpartum, all rats underwent anesthesia using intraperitoneal injections of ketamine (90 mg/kg) and xylazine (10 mg/kg). To simulate prolonged labour, a modified distal-tip shortened 22 Fr Foley catheter was inserted transvaginally and inflated (3 mL sterile water). A 130-g weight was attached to the catheter's proximal end. The rat was placed in a fixed prone position at the edge of the mini operating table, leaving the caudal end of the animal hanging downward without touching the table. This allowed the weighted catheter to hang freely against gravity. The rat's colour, warmth and respirations were monitored without positional disturbance for 3 hours. This mechanism simulates the pressure effects of a large fetus on the pelvic floor during the second stage of labour and has been previously shown to create a consistent model of SUI.11 We used postpartum instead of virgin rats because it is more representative of the clinical situation for which most SUI occurs (after vaginal birth).
One week after balloon dilation, the rats were anesthetized as above. Using a small vertical lower midline abdominal incision, vascular pedicles were suture ligated and ovaries removed, followed by a 2-layer abdominal closure. We performed ovariectomy 1 week after vaginal dilation because performing both procedures at the same time increased the morbidity and mortality of the rats in our previous experiments.
Two weeks after delivery, under sedation with ketamine (100 mg/kg), a repeat lower midline incision exposed the bladder. A 27-gauge butterfly needle was inserted into the bladder dome, and all urine was evacuated. Using a 3-way stopcock, the needle was connected to both a Baxter Uniflow pressure transducer (Baxter Healthcare Corp., Irvine, Calif.) and a Harvard Model 22 infusion pump (Harvard, Millis, Mass.). Saline at 37°C was infused at 0.1 mL per minute after calibrating the pressure transducer to zero atmospheres. A customized Macintosh Quadra 800 (Apple Computer Inc., Cupertino, Calif.) with LabView 4.0 software (National Instruments Corp., Austin, Tex.) collected the data.
Voiding was considered normal if a detrusor contraction was followed by expulsion of saline from the meatus and with bladder emptying. The modified leak point pressure was determined at this point, and the instilled intravesical volume represented bladder capacity. A pattern of moderate incontinence was identified on cystometry if intermittent small leakage occurred with or without bladder contraction during filling, followed by a sustained detrusor contraction and voiding. Cystometric studies in animals with severe incontinence displayed a continued small amount of saline leakage at low detrusor pressures, without the bladder filling to capacity (Fig. 2). Urodynamics were repeated at 7.5 weeks postpartum. Because these studies were performed under ketamine sedation, we could not reliably differentiate overactive bladder from sphincter deficiency, as in the case of awake cystometry and therefore grouped them together as the AV group.
One week after initial cystometry and randomization, a 0.5-mg 17 β-estradiol (21 d timed-release formula) or placebo pellet (Innovative Research of America, Sarasota, Fla.) was implanted in a subcutaneous pouch. After ketamine sedation, a 5-mm incision in the dorsal skin at shoulder height allowed pellet deposition. This procedure was repeated in all groups 3 weeks later to maintain constant estrogen or placebo levels. In our pilot studies, we did not notice any changes with shorter-term estrogen therapy, thus we implanted a second pellet at 3 weeks, when the first pellet's effects had terminated.
Upon completing the second cystometry at 7.5 weeks after delivery, animals were euthanized. The pelvic floor was exposed, and samples of the midurethra and vagina were removed.
Tissue specimens were fixed in cold 2% formaldehyde and 0.002% saturated picric acid in 0.1 μmol/L phosphate buffer (pH 8.0) for 4 hours, followed by overnight immersion in a buffer containing 30% sucrose for cryoprotection. The specimens were embedded in OCT Compound (Sakura Finetic USA, Torrance, Calif.) and stored at –70°C until use. Fixed frozen tissue specimens were cut at 5 μm thickness, mounted to SuperFrost Plus charged slides (Fisher Scientific, Pittsburgh, Pa.) and air-dried for 5 minutes. Endogenous peroxidase activity was blocked with 0.3% H2O2 methanol for 10 minutes. After rinsing, sections were washed twice in phosphate-buffered salene (PBS) for 5 minutes and incubated with 3% horse serum in PBS with 0.3% Triton X-100 (Rohm and Haas Co., Spring House, Pa.) for 30 minutes (room temperature) to eliminate nonspecific protein binding. After draining, the slides were incubated overnight at 4°C with goat polyclonal antibodies to α1A- or α1D-adrenoceptors (Santa Cruz Biotechnology, Santa Cruz, Calif; 1100) or mouse monoclonal antibody to neuronal nitric oxide synthase (nNOS) (BD Transduction Laboratories, San Jose, Calif; 1500) to examine potential changes in contractile (α adrenergic) and relaxant (nitric oxide [NO]) properties of the urethral sphincter. Slides incubated without antibodies in a similar fashion served as negative controls. After washing with buffer sections, slides were immunostained with the avidin-biotin-peroxydase method (Elite ABC, Vector Laboratories, Burlingame, Calif.), using diaminobenzidine as the chromagen, followed by hematoxylin counterstaining. For image analysis, 5 randomly selected fields per tissue, per animal, for each treatment group were photographed and recorded with a Retiga Qimage digital camera and ACT-1 software (Nikon Instruments Inc., Melville, NY). Images were quantified with Image-Pro Plus software (Media Cybernetics, Silver Spring, Md.). The ratio of the α1A- or α1D-adrenoceptor or nNOS stained pixels to all pixels of the whole picture (equal to 100%) provided a percentage. Different team members performed the functional study and histological evaluation, and evaluations were blinded until after the final analyses.
All values are expressed as mean and standard error of the mean. We used the unpaired 2-tailed t test for statistical analysis, using YSTAT 2002 (Igakutosho, Tokyo, Japan). For posttreatment continence status, we performed the chi-squared test. Comparisons were considered significant at p < 0.05.
Initial transvesical cystometric profiles performed 2 weeks after postpartum intravaginal balloon dilation and 1 week after ovariectomy identified 18 rats (60%) with AV patterns and 12 (40%) with normal voiding. Table 1 compares urodynamic findings. Repeat cystometry after estrogen treatment documented an increase in AV from 60% to 73.3% (from 9 to 11 rats), whereas the placebo group demonstrated a decline in AV from 60% to 33.3% (from 9 to 5 rats). The difference is statistically significant (χ2 of 4.821 and p = 0.028). In a subset analysis, estrogen-treated rats with abnormal voiding (group 4) more frequently displayed a higher severity of incontinence, compared with placebo-treated rats with AV (group 2), with continuous leakage hampering the ability to reach bladder capacity. Other parameters are summarized in Table 2, with statistically significant lower bladder capacity, maximal bladder pressure and modified leak point pressure noted in group 4.
We observed thickened epithelium and subepithelium, as well as increased amounts of smooth muscle fibres in the muscularis, for the estrogen-treated vaginal tissue, and we observed no general morphological differences for urethral specimens between estrogen- and placebo-treated groups.
Alpha adrenoceptors 1A and 1D were identified in both the urethra and the vagina. Estrogen therapy significantly downregulated α1A- and α1D-adrenoceptor expression in the endothelium and subendothelial regions of the urethra while increasing levels of both receptors in the vagina (Table 3, Fig. 3 and Fig. 4).
Several recent publications have identified adverse effects of HRT in a diverse group of disease processes, including stroke,12 breast cancer,13,14 osteoporosis8 and urinary incontinence.7 Further, the Women's Health Initiative (WHI) trial, evaluating the effects of postmenopausal estrogen plus progesterone therapy was prematurely halted secondary to multiple reported adverse trends in the treatment arm.12–14 The results of our experiment align with these aforementioned studies, indicating that extended term estrogen therapy (> 5 wk) worsened voiding dysfunction in our postpartum intravaginal balloon-dilated, ovariectomized rat model.
Using this model, we previously reported that birth trauma simulated by weighted ballooning and ovariectomy appears to contribute to SUI.9,11 We have also demonstrated that these rats rarely have UI during pregnancy or after ovariectomy alone without vaginal distension. Although the scenario does not create a true replica of prolonged labour as seen in humans, we believe this model is suitable for studying the urinary continence mechanism in the rat.
In our study, cystometric curve and the observation of urine (saline) leakage pattern clearly identified rats with normal voiding patterns and those with abnormal leakage. Since the cystometry was performed under ketamine sedation, we could not determine whether the AV pattern was due to pure sphincter insufficiency or mixed with overactive bladder, we included all rats with abnormal voiding in the AV group. Future studies with conscious cystometry will be required to further study the AV patterns.
Using immunohistochemistry, we observed a significant downregulation of α1A- and αD-adrenoceptor expression in the endothelium and subendothelial regions of the urethral submucosa in rats treated with estrogen. Conversely, there was an increase in α1A- and αD-receptor immunostaining in vaginal tissues. This differential regulation of adrenoceptor in the urethra and vagina is intriguing and warrants further investigation (Table 3).
Urinary continence is believed to be maintained by sympathetic impulses through α1-adrenoceptors in the bladder neck and urethral smooth muscle.15,16 Binding and molecular biological studies established the α1A-adrenoceptor to be the predominant subtype in the urethra, responsible for the contractile responses of the urethra via elevated intracellular calcium levels to adrenergic agonists.17–19 Taki and colleagues20 demonstrated that α1-adrenoceptors play an important role in maintaining the resting urethral tonus of the female urethra. They divided the urethra into segments and administered noradrenaline (NA), an α1 agonist, demonstrating a concentration-dependent contraction with a peak amplitude in the middle to proximal urethra, while acetylcholine produced contractions only in the proximal urethra and bladder neck. Applying these results to our experiment, a downregulation of α1-adrenoceptors would contribute to a significant decrease in contraction and tone of urethral smooth muscle.
The 4 reported α1-adrenoceptor subtypes include α1A, α1B, α1D and α1L.21 In the proximal female urethra, the expression of α1-adrenoceptor subtype mRNA has a reported ratio of 90010 for α1A, α1B and α1D, respectively.22 However, Nishimatsu and colleagues23 reported that phenylephrine, a nonselective α1-adrenoceptor agonist, induced contractions through the α1L receptor, and not through the α1A subtype. The role of each of these α1-adrenoceptor subtypes in relation to urethral contractile function has yet to be delineated. Regardless, our results suggest that downregulation of both α1A and α1D adrenoceptors may represent one mechanism by which impaired contraction of the urethra can lead to incontinence.
NO, a neurotransmitter that causes relaxation of urethral smooth muscle, is produced by NOS, an enzyme located at several sites, including neurons.24,25 Recent studies suggest that adrenergic and NO-mediated nerves may function in a complex interactive manner and not as independent agents.17 In the rabbit urethra, released NO suppresses NA release from adrenergic nerve terminals; the reverse has also been reported.26 Perhaps our finding of upregulated urethral nNOS in the estrogen group results in smooth muscle relaxation by both local effect and α1-adrenoceptor downregulation with a net decrease in NA release.
Further investigation is needed to determine the intricate mechanisms causing SUI. Both molecular study and physiological animal experiments are needed to advance our knowledge and improve clinical management. Identifying the mechanism by which estrogen modulates the α1 adrenoceptor will be valuable not only to incontinence research but also to a broad spectrum of diseases.
Extended term estrogen therapy in a rat model of simulated birth trauma and ovariectomy resulted in a higher rate of incontinence. Immunohistochemical examination shows significant downregulation of urethral α1A and α1D adrenoceptors and upregulation of nNOS in the urethra of estrogen-treated groups. These studies question the use of HRT in the treatment of postmenopausal incontinence.
Acknowledgements: This work is supported by NIDDK/NIH grants: R01 DK069655-01 and P50 DK64538. Dr. A.J. Bella is the American Urological Association Foundation Robert J. Krane Scholar and a Royal College of Physicians and Surgeons (Canada) Detweiler Travelling Fellow.
This article has been peer reviewed.
Competing interests: None declared.