Expanding Treatment Options for Stress Urinary Incontinence in Women

At least 13 million adults in the United States experience urinary incontinence (UI).¹ Most are women, in whom the prevalence of "bothersome incontinence" increases from 2% between the ages of 20 and 24 years to 9% at 50-54 years and to 16% at 85 years.² The annual direct costs of UI in the United States are estimated at $12.4 billion for women and $3.8 billion for men (in 1995 US dollars),³ similar to estimates for osteoporosis, arthritis, Alzheimer disease, human immunodeficiency virus, and AIDS.¹ However, these estimates are conservative because the majority of individuals with UI are unknown to their physician and thus are neither evaluated nor treated. In addition, neither the substantial indirect costs nor the attendant medical or psychosocial morbidity are included in estimates of its impact. Despite these considerations, UI remains relatively neglected by clinicians and researchers alike¹ even though this condition is generally responsive to therapy.

The type and prevalence of UI vary by sex and age. Stress UI is characterized by instantaneous leakage with physical strain such as occurs with coughing or sneezing. Urge incontinence is characterized by a sudden sensation of the need to void, which occurs moments prior to leakage. Mixed incontinence is diagnosed when both of these conditions are present.⁴ In women, stress UI is the most common type until approximately age 65 years, when the combined prevalence of urge UI and mixed UI surpasses it.⁵ In men, stress UI is rare except after radical prostatectomy and, when present, its mechanisms, treatment, and responsiveness differ from those in women.⁶ Other forms of UI, such as overflow (manifest as dribbling from a constantly overfilled bladder), are much less common. Functional UI (inability to get to the bathroom because of cognitive or mobility limitations), although commonly diagnosed in older adults, is a problematic category that probably is a rare cause of UI but a common contributor to it in such patients.⁴

The mechanisms underlying stress UI are inadequately understood, as are those that mediate its therapeutic response. Multiple factors are involved, including muscle and ligament support of the bladder base, absence of bladder spasms (detrusor overactivity) during bladder filling, integrity of neurological innervation, and preservation of the urethra's endothelial cushion and sealing ability, which involve urothelial, vascular, connective tissue, and neurological components.⁷ However, even with a mildly compromised sphincter mechanism, continence often can be maintained by keeping bladder volume below the threshold at which leakage occurs. This can be accomplished by adjusting fluid intake and voiding frequency and by minimizing physical stress through treating a cough. Thus, response to therapy reflects not only the intervention being tested but also patient selection, evaluation, and control of the other contributing factors.

Previously accepted therapies for UI have lost some of their appeal. For instance, surgical intervention, such as colposuspension or suburethral sling, was considered the definitive treatment for stress incontinence until recently. Reported cure rates approximated 90%, but longer follow-up revealed that cure rates decreased to between 60% and 80% after between 3 and 5 years.^{8 - 9} These long-term cure rates are disappointing for a condition that predominantly affects middle-aged women, whose life expectancy is measured in decades, especially since cure rates after a previous operation are lower than after the initial surgery.⁹ Nonetheless, surgery remains a woman's best chance of cure in 2003.

The efficacy of pharmacotherapy has been only modest.¹⁰ Moreover, because it traditionally comprised α-adrenergic agonists, with or without hormone therapy, enthusiasm has waned as the hazards of both α-adrenergic agonists and hormone therapy have become apparent.^{10 - 11} New pharmacological approaches have been proposed; for example, duloxetine inhibits reuptake of norepinephrine and serotonin and is believed to increase neural input to the urethral sphincter. But clinical studies of duloxetine are still limited and its therapeutic role is not yet clear.¹⁰ Consequently, other approaches have generated interest, including exercises assisted by mechanical means such as vaginal cones,¹² removable intraurethral devices ("plugs"),¹² and potential applications of gene therapy,¹³ and stem cell implants.¹⁴ For each, either experience has not yet matched the promise or the treatment is still investigational.

Ever since Kegel's early success with pelvic muscle exercises, investigators have pursued behavioral methods such as pelvic muscle training and biofeedback for treatment of UI. Unfortunately, when performed correctly, such therapy is time-consuming and labor intensive. Several groups have tried to streamline the process, but with mixed results.¹²

Based on 2 decades of pioneering work, Burgio and colleagues have developed briefer methods that have proven effective for urge incontinence in older women.¹⁵ They then devised written materials that were also effective for that group.¹⁶ In this issue of THE JOURNAL, this research group reports the results of applying these techniques to stress incontinence in younger women.¹⁷

The goal of the current study by Goode et al was to determine the relative benefit of behavioral methods over written instruction, and the marginal benefit of another long-standing but inadequately evaluated intervention, pelvic floor electrical stimulation (PFES). The investigators enrolled 200 women with moderate-to-severe stress or mixed incontinence (with stress as the predominant pattern) in an 8-week randomized controlled trial of 4 visits with behavioral training, 4 visits with behavioral training plus home PFES, or self-administered behavioral treatment using a self-help booklet. None of the women had dementia and their ages ranged from 40 to 78 years. Each intervention provided improvement, but the intensive behavioral or PFES approaches were significantly more helpful than the written instructions (about a 70% reduction in UI episodes for the 2 interventions vs 50% for the control booklet). These short-term results are notable, especially given the severity of the condition and the concurrence of urge incontinence in such a large proportion of patients.

Several factors should be considered in evaluating the generalizability of the study results. The women in this trial had fairly severe stress UI, with half reporting 10 or more episodes per week; this may be explained by the high prevalence of concurrent urge UI (about two thirds compared with rates half as high in most surgical series). Also, participants in this study were of higher socioeconomic status than is typically seen in clinic populations, were better educated, and were better motivated since they had to complete daily bladder diaries for weeks. Although the exclusion criteria are quite reasonable and the diagnostic testing strategy is similar to that used in most clinical practices, the behavioral intervention and the degree of clinician expertise and patient commitment it required differ substantially from what is practiced in the community or even in most academic centers. Moreover, the technique the investigators selected for PFES is reasonable, but their conclusions apply to the specific modality and parameters tested, applied precisely as described.

The study raises several questions. For instance, knowing whether the results apply equally well to patients in the 3 urodynamic groups (stress UI alone, stress UI with demonstrable detrusor overactivity, and those with symptoms of mixed UI but without demonstrable detrusor overactivity) is potentially important. Unfortunately, small cell sizes preclude these subanalyses. Such knowledge would be useful since, if the response of patients with concomitant detrusor overactivity differed, both the behavioral and stimulation regimens could be better tailored to this entity. In addition, it would be useful to know whether improvement in UI was mediated by urethral strengthening because the role of sphincter strength likely differs between individuals. According to a preliminary report, Miller et al¹⁸ found that women with stress UI may improve within days simply by learning how to time the tightening of their pelvic floor in response to a cough. Also, it would be useful to know whether treatment response differed according to the severity of UI. The authors found no difference according to the frequency of leakage but we do not know whether it differed according to the amount of leakage. Finally, it would be useful to know how durable the improvement will be over time, the characteristics that predicted response to each form of therapy and whether they differed by intervention, and the physiological mechanisms that mediated the response to each intervention.

What are the implications for clinical practice? Many are insightfully described by the authors themselves. First, although UI is highly treatable, most patients with UI do not mention it to their physician; thus, it is important to ask about it routinely. Second, unlike the individuals enrolled in this trial, most patients experience much milder degrees of leakage. Once any transient causes of incontinence are eliminated (eg, use of α-blockers),⁴ their UI may be more amenable to simpler interventions such as physical maneuvers¹⁹ or adjustment of fluid intake and voiding frequency, as well as reduction of physical stress such as cough. Third, even among the patients in this study, in whom UI was moderate to severe, nearly half experienced at least a 50% improvement by following written instructions and by keeping regular bladder diaries, and nearly 1 in 6 became completely continent on this regimen. This simple approach could help countless numbers of incontinent women. Fourth, the success of behavioral therapy requires a highly motivated patient and access to a therapist with the requisite expertise, equipment, commitment, and persistence.¹² Since in many areas a stimulation device may be more readily available than an expert therapist, it is tempting to advise a device instead. However, if it is not applied as in this study (ie, in addition to the careful behavioral approach), the success rate will likely be much lower. Fifth, other effective therapies are available for women with stress UI, including pelvic muscle exercise assisted by vaginal cones, intravaginal devices, and surgery. Moreover, surgical approaches continue to improve, and exciting research is ongoing in pharmacology,¹⁰ tissue engineering,^{13 - 14} and device development.¹²

Thus, clinicians and patients now have several choices for treatment that are effective and feasible for most patients. With additional research, better understanding of the mechanisms mediating UI should emerge and with it the hope for still more effective tools to help the hundreds of millions of individuals who experience UI worldwide.²⁰