Introduction

Urinary incontinence is a highly prevalent condition associated with significant patient morbidity and economic costs [1, 2]. In their lifetime, as many as 35 % of women will experience stress urinary incontinence (SUI), the involuntary leakage of urine with laughing, coughing, or sneezing, whereas as many as 46.9 % of women will experience urge urinary incontinence (UUI), the involuntary leakage with urge [2, 3]. Costs in the United States alone in the year 2000 have been estimated to be at least $20 billion [4]. As the population continues to age, it is anticipated that both the number of patients seeking care as well as the costs will continue to rise [5].

Standard treatment options for SUI include pelvic floor physical therapy, incontinence pessaries, or surgical options, such as periurethral bulking injections, midurethral slings, or retropubic urethropexies. Periurethral or transurethral injections, in which a bulking agent, such as calcium hydroxylapatite, is injected near or within the urethra, have relatively low success rates reported at approximately 40 % for cure and 70-80 % for improvement [6]. Patients often need multiple injections given the transient effectiveness. Complications are generally mild and include urinary retention, infection, and pain; however, more serious events, including abscess formation, extrusion into the urethra, and hypersensitivity to the injected material, have been reported [6, 7]. Finally, concern for particle migration after injection highlights additional problems with this treatment option [8]. Midurethral slings have very high reported success rates up to 80.8 %, but patients may have reservations about the placement of transvaginal mesh [9]. While providing superior results, mesh is associated with multiple risks including mesh exposure (rates ranging from 0.6-5.4 % after midurethral sling), pelvic pain, dyspareunia, infection, vaginal discharge, sinus tract formation, and the need for reoperation [10]. Thus, the many limitations and concerns with the use of mesh reinforce that novel treatment options and materials are warranted.

First-line treatment of UUI often includes antimuscarinic drugs in combination with behavioral modifications. Unfortunately, 35.1 % of patients who try these drugs will never have their prescription refilled, likely due to lack of efficacy or intolerable side effects [11]. Additional therapies for UUI include intravesical botulinum toxin injection, sacral neuromodulation, and percutaneous tibial nerve stimulation with reported success rates of 60-85 %, 67-80 %, and 54.5 %, respectively [12–15]. Botulinum toxin injection is associated with a high rate of urinary retention, underscoring the need for patients to be counseled about the risks and potential need for self-catheterization. Similarly, rates of urinary tract infection have been noted to be as high as 31 % [16]. Lastly, the effects of botulinum toxin are temporary with most women needing additional therapy within 1 year [17]. Sacral neuromodulation is a more invasive option involving the placement of an implantable pulse generator and permanent electrodes to stimulate the S3 sacral nerve root. It is associated with a high rate of reoperation (25–30.3 %), infection (5–7.9 %), and pain (9-27 %) and thus may be associated with significant morbidity [14, 18–21].

Given the variability in efficacies, the potential side effects, and the lack of options for women who have poor smooth muscle integrity, absence of coaptation, and have failed conventional therapies, researchers have been studying stem cell therapy as a potential therapeutic option for urinary incontinence. This review provides a brief overview of stem cell therapy and discusses the relevant literature regarding the use of various stem cells for treatment of both stress and urge urinary incontinence.

Stem Cell Therapy

During the past 60 years, significant advances have been made in the field of stem cells and tissue engineering. Stem cells are unique in that they have the potential of replicating, differentiating into various different cell types, and forming a population from just one cell. Stem cells may be totipotent, pluripotent, multipotent, or unipotent with the ability to differentiate into any cell type, including embryonal, any of the three germ cell layers, a select few cell types, or only self-renew, respectively [22]. They may be obtained from various sources, such as bone marrow, adipose tissue, or embryos, which often define their potency.

Embryonic stem cells, derived from the inner cell mass of embryos in the blastocyst stage, are ideal for therapy given their pluripotency and ability to proliferate undifferentiated for many passages [23]. With the benefits of pluripotency also come the risks of immunorejection and teratoma or teratocarcinoma formation. Human-derived embryonic stem cells, often obtained from embryos created with assisted-reproductive technology, have additional ethical concerns, which have limited their use [24]. However, Zhang et al. found that embryonic stems cells derived from arrested embryos retained multiple pluripotency markers similar to those obtained from developing embryos [25]. Thus, arrested embryos may be a source of stem cells with less ethical limitations. Other gestational tissues contain stem cells, such as amniotic fluid, placenta, and umbilical cord blood, and are less controversial. Amniotic fluid-derived stem cells, with a short doubling time of 36 hours, are able to differentiate into all three germ cell types [26]. Likewise, hematopoietic stem cells have been isolated from umbilical cord blood, whereas the Wharton’s jelly of the umbilical cord contains mesenchymal stem cells [27].

Adult-derived stem cells are not pluripotent, but some still maintain the ability to differentiate into different cell types [22]. Similarly, adult cells can be reprogrammed by transduction with retroviruses containing genes that improve growth potential and dedifferentiation to form induced pluripotent cells. Mesenchymal stem cells, which maintain multipotency, have been derived traditionally from bone marrow, but also may be derived from skeletal muscle, adipose tissue, and the majority of adult tissue types [28]. They are ideal candidates for regenerative therapy given the low risk for malignant differentiation, the ability for autologous transfer and thus elimination of the risk of rejection, and the absence of ethical controversy as seen in embryonic stem cells [29••]. Indeed, the majority of stem cell research in the lower urinary tract focuses on mesenchymal stem cells.

Stress Urinary Incontinence and Autologous Connective Tissue Cells

Urethral bulking, periurethral or transurethral injection of bulking agents to improve urethral coaptation, has been traditionally performed with nonimmunogenic and biocompatible substances, such as bovine collagen, calcium hydroxyapatite, or silicon particles with various degrees of success [30]. As potentially more efficacious options, various connective tissue cells have been studied as possible bulking agents [31, 32]. In one study, 32 women with intrinsic sphincter deficiency diagnosed by a Valsalva leak point pressure of less than 90 cm H20 received a single autologous ear chondrocyte periurethral or transurethral injection distal to the bladder neck [31]. At 1 year follow-up, 16 patients (50 %) were dry, whereas 10 reported sustained improvement (31.3 %). Of the 22 women with repeat leak point pressures, 9 (40.9 %) had increases greater than 30 cmH20. Few women reported adverse events and these mostly included urinary tract infections, vaginal bleeding, dysuria, urinary frequency, and urinary retention.

In a much larger study, including 123 women with urodynamically proven SUI, women received transurethral injection of autologous myoblasts and fibroblasts obtained from a biceps muscle biopsy [33]. The fibroblasts were mixed with collagen to prevent site migration. Seventy-nine percent (94/119) of the women were completely continent at one year, with improvement in Incontinence Scores from a median of 6 to 0. An additional 16 women reported substantial improvement. Periurethral surface electromyography measurements also were improved with increased rhabdosphincter contractility at rest (34.0 ± 11.0 to 45.1 ± 15.0 μV) and with voluntary contraction (43.1 ± 11.8 to 55.4 ± 15.3 μV). No adverse events were reported; however, the authors note that much further study is needed including multicenter studies and long-term follow-up to confirm this therapy as a potential treatment for SUI. A smaller study by the same authors, including 20 women with SUI treated with transurethral injection of myoblasts and fibroblasts, revealed that 16 (80 %) still reported cure at 2 years of follow-up [32].

Stress Urinary Incontinence and Muscle-Derived Stem Cells

A significant amount of research has focused on muscle-derived stem cells (MDSC) in the lower urinary tract. Myoblasts can be injected into either urethral or bladder wall tissue with subsequent formation of myotubes in the smooth muscle layer [34]. After formation of myotubes, these cells become postmitotic and thus have a low likelihood for growth or spread. In contrast to myoblasts, MDSC have the ability to survive longer than 48 hours [35]. Yokoyama et al. demonstrated that autologous MDSC obtained from gastrocnemius muscle biopsies from Sprague–Dawley rats and injected into the urethra or bladder were able to survive and differentiate into myotubes and myofibers [36]. When they compared this injection to that of a bovine collagen injection, rats injected with MDSC had persistence of the injected periurethral nodules at 30 days while those with collagen injection had significantly smaller nodules indicating loss of injected mass. They postulated that the potential effects of MDSC were not only the ability of these cells to differentiate into myotubes but that they also may contain satellite cells with the ability to differentiate into various cell types based on the surrounding environment. Other theories of mechanism of action include promotion of innervation and improvement in tissue function. In a follow-up experiment, MDSC also survived and differentiated into myotubules after being injected into murine bladders [37].

Periurethral or transurethral injection of MDSC has been shown to be a possible treatment for SUI. In rats with denervated and atrophied urethras, periurethral injection of MDSC resulted in improvements in fast-twitch muscle contraction to 87 % of the muscle contraction seen in control animals [38]. In addition, there was minimal immunohistochemical evidence of inflammation, despite the presence of a significant amount of skeletal muscle fiber formation. Similarly, periurethral injection of MDSC in rats improved leak point pressures in denervated rats [39]. Compared with both control and sham injected rats, those injected with MDSC had higher leak point pressures at 1 week (25.2 ± 1.9, 28.6 ± 0.8, and 36.7 ± 2.3 cmH20, for control, sham, and MDSC injected rats, respectively) and 4 weeks (25.8 ± 2.5, 18.6 ± 5.2, 44.1 ± 6.6 cm H20, respectively). Histological examination of the MDSC injected rats revealed that the muscle fibers had variable orientation highlighting that injection likely served as a bulking agent as opposed to becoming a normal layer of the urethra. However, in a follow-up study, denervated rats were periurethrally injected with either MDSC or fibroblasts to determine if MDSC add additional benefits other than just bulking [40]. Denervated rats were either injected with saline, MDSC, fibroblasts, or a MDSC-fibroblast 50:50 mixture. Injection of MDSC significantly improved fast-twitch muscle contraction amplitude to 73 % of normal while fibroblast injection only improved it to 46 % of normal. Thus, the authors propose that the mechanism of benefit is different between MDSC and fibroblasts. Indeed, rats injected with fibroblasts also had a higher connective tissue mass at the injection site, highlighting that this may have more of a bulking effect.

In one of the only studies in humans, Carr et al. treated eight women with urodynamic-proven SUI [41••]. Autologous MDSC were obtained from thigh biopsy and transurethrally injected in up to four circumferential locations. Three women withdrew from the study after 1 month of follow-up; the remaining five all demonstrated improvement in stress incontinence as determined by pad tests and bladder diaries. The initial onset of improvement was between 3 to 8 months after injection, suggesting that the mechanism of action was improving muscle function as opposed to providing bulking. However, two individuals needed additional injections to achieve improvement and two went on to have midurethral slings placed. There were no reported adverse events. One other large human study involving periurethral injection has subsequently been retracted [42].

Stress Urinary Incontinence and Adipose Tissue-Derived Stem Cells

Adipose tissue-derived stem cells (ADSC) have been studied recently for lower urinary tract dysfunction given their easy accessibility and overall abundance [43, 44]. ADSC have been shown to differentiate into myoblasts and these myoblasts have subsequently been injected in to both the bladder smooth muscle and urethra. In a study by Fu et al., ADSC were obtained from 20 female rats with SUI and then induced ex vivo into myoblasts [45]. The myoblasts were then injected periurethrally with a microinjector into the urethral sphincter muscularis in the proximal urethra in ten of the animals; the remaining ten animals had injection of undifferentiated ADSC. To obtain leak point pressures, anesthetized animals underwent placement of an epidural catheter attached to an urodynamics flowmeter through a two centimeter cystotomy. At half-bladder capacity, increasing external pressure was placed on the animal’s abdomen until urine leaked; pressures were then measured. The average leak point pressure and the maximal bladder capacity at 3 months were higher in the group with myoblast implantation compared with the group with undifferentiated ADSC (41.12 ± 2.89 versus 32.52 ± 2.16 cmH20 and 2.09 ± 0.09 versus 1.88 ± 0.05 cc, respectively). A similar study was done in 28 female rats with induced SUI; 12 rats were urethrally injected with ADSC while 6 rats were intravenously injected with ADSC [46]. Cystometry revealed that injection of ADSC either transurethrally or intravenously was associated with less voiding dysfunction compared with control animals (33.3 % versus 80 %, p < 0.05) as well as significantly lower leak point pressures. Histologic analysis revealed rats injected with ADSC had higher elastin content than controls. Although the authors were unable to quantify how much of the intravenous ADSC were present in the urethra, with the use of 5-bromo-2-deoxyuridien labeled-ADSC, they were able to demonstrate migration of the intravenous cells to the urethra. Thus, the authors highlight the potential tracking of intravenously injected ADSC to the urethra as yet another area of future research.

Urge Urinary Incontinence and Stem Cells

Much less is known about the potential benefits of stem cell therapy on UUI. Huang et al. specifically studied ADSC in rats with overactive bladder [47•]. Thirty male rats were fed high-fat diets to decrease bladder blood flow with resultant denervation and ultimately detrusor overactivity; ten rats fed regular diets served as controls. Rats were then either injected in the detrusor with saline or autologous ADSC or had injection of ADSC into the tail vein. Among rats with overactive bladder, those treated with intradetrusor ADSC had longer micturition intervals (368 ± 66.8 seconds versus 143 ± 28.7 seconds, p < 0.01) and greater mean volume per void (0.628 ± 0.122 mL versus 0.23 ± 0.05 mL, p < 0.01) compared with saline injected rats. Histologic data demonstrated improvement in microvessel and neuronal content of the detrusor muscle. Given the limited available data on stem cell therapy and urge incontinence, significantly more study is needed to determine if stem cells may be a potential treatment option.

Concerns with Stem Cells

Aside from the ethical issues discussed earlier, other major concerns with stem cell technology exist. The true risks of stem cell therapies remain undefined but may include complications as a result of overproduction with resultant neoplasia and migration. These complications likely will not be fully appreciated until more research in humans has been completed. To date, most cell transplantation studies are performed in patients with dismal prognoses (end-stage diabetes, advanced neurodegenerative diseases, spinal cord injury and cardiac failure) and limited life expectancies. In contrast, women seeking treatment for incontinence, a quality of life issue, are relatively healthy and most anticipate a normal life expectancy. Thus, in theory, transplanted cells and their potential complications will be present for decades or longer.

Concerns with Current Research

While the past two decades have provided early encouraging results in the use of stem cells for treatment of urinary incontinence, the majority of the studies involved animal models only and had limited discussions about the possible mechanism of action or target populations. Various animal models for SUI are routinely used, yet all cannot completely capture the complex process of micturition and urinary leakage, and thus rely on surrogate markers such as leak point pressures [48]. Similarly, it is impossible to capture intent in animals so the behavioral component of continence cannot be ascertained [49]. Lastly, animals often have to be anesthetized for advanced urologic testing adding additional sources of confounding. Thus, although research in animals is necessary, it is unclear how promising results with stem cell therapy will translate into human use.

Perhaps the most basic tenant of regenerative medicine is to define the basic pathophysiology of the problem needing repair. Unfortunately, the mechanism by which the urethral sphincters maintain continence is not well understood. Similarly, the available studies give limited explanations of how stem cell therapy improves continence. Theories thus far have included periurethral bulking alone, improvement in sphincter muscle or surrounding tissue function, and promotion of innervation [36, 41••]. A more thorough understanding of both the pathophysiology of incontinence and in the potential mechanisms of action of stem cells is needed. Lastly, the phenotype of the target population is not clearly defined. Ideal candidates for stem cell therapy include those with urethral hypomotility, a low-pressure urethra (i.e., maximal urethral closure pressure less than 20 cm H20), those who have failed previous incontinence surgery, the elderly, and those with mesh intolerance.

New technologies are extremely appealing to all involved: enthusiastic scientists, ambitious clinicians, and industry eager for commercial gain. Such a combination creates a potentially dangerous scenario in which a patient desperate to improve her condition becomes an easy target for unforeseen complications in the event that the technology is introduced too quickly and without good scientific evidence or oversight. Thus, clear goals and expectations and thorough understanding of therapy is crucial for patient safety.

Conclusions

Currently available treatment options for both stress and urge urinary incontinence have variable efficacies and potential adverse events, and thus there is a need for novel treatment options that may ultimately provide better outcomes at lower risks. Preliminary studies with stem cells in mostly animal models have shown this is a promising area for future research and may indeed eventually yield the ideal therapy for stress or urge incontinence. However, substantial research must be done specifically focusing on pathophysiology, mechanism of action, and safety before routinely offering this therapy.