One of the first animal models to study the pathophysiological mechanisms of stress urinary incontinence associated with childbirth was developed 15 years ago by Lin et al. [80] and performed in rats using a vaginal distension method that introduced an intravaginal balloon filled with 2 mLl of saline using a transurethral 12 F catheter left in situ for 4 h. This murine model showed a significant decrease in the periurethral striated and smooth muscle and c-Fos immunostaining neurons in the L6 to S1 spinal cord segments, indicating that irritation causes nerve and muscle injury. This finding paved the way for the next studies using vaginal distension methods for animal models of stress urinary incontinence and recovery after childbirth [81]. Since then, many authors have modified the vaginal distension method in rats or mice using different sized catheters or balloons or by changing the distension time. However, the results have always been similar with short-term effects including urethral dysfunction, a decrease in urethral resistance, neuromuscular damage of structures responsible for continence with edema in the pelvic musculature, and overexpression of hypoxia-inducible factors and cellular markers of innate repair. The main limitation of this model is the short durability of the functional, structural, and biomechanical defects due to vaginal distension, with a recovery within 10 days [82] and up to 6 weeks [81], depending on the vaginal distension time. This limitation makes the vaginal distension method useful for mechanistic studies of childbirth-induced tissue injury and recovery, but it is challenging for testing the efficacy of stress urinary incontinence treatments beyond 6 weeks. The strength of this model is that it recapitulates childbirth injury which is one of the strongest known risk factors for the development of stress urinary incontinence. However, to the best of our knowledge, the long-term effects of simulated childbirth injury have not been investigated in animals. Of the women who have full remission of stress urinary incontinence after delivery, almost half redevelop stress urinary incontinence 5 years later, and many more develop stress urinary incontinence decades later in menopause. For this reason it is important to test the effect of age in vaginally distended rodents, since they should be expected to redevelop stress urinary incontinence as a consequence of their altered function or support.

Vaginal distension rat models have been used to study stem cell regeneration therapies. For example, autologous mesenchymal stem cells were shown to home to the urethra and vagina after tail vein injection [59]. Although this study showed improvement in continence as measured by leak point pressure 1 week posttreatment, the external urethral sphincter function as measured by electromyography was not improved at that time. Therefore, regenerative treatment investigations may be limited by the short durability exhibited in the vaginal distension models. Other authors have combined vaginal distension with bilateral ovariectomy in rats [67] to study tail vein or urethral injection of adipose-derived stem cells and showed that 66.7% of the animals had normal voiding function after 1 month of treatment as measured by four-channel conscious cystometry.

The vaginal distension technique has been established in larger animals as well. Burdzin’ska et al. developed the vaginal distension method in pigs and showed a decrease in maximal urethral closure pressure (50%) and functional urethral length (52%) 28 days post-vaginal distension [83]. The advantage of this animal model is that the external urethral sphincter of pigs and humans consists of a high percentage of slow-twitch myofibers and urodynamic evaluations can be carried out using the same procedure used for human patients, whereas this is not the case in rodents or dogs [83]. The disadvantage of larger animal models is the rate of animal growth during the study and the maintenance costs of larger animals compared to murine models. However, it is important to have both small and large animal models available since regulatory requirements that apply to novel therapies usually involve completion of a full nonclinical development program.

The pudendal nerve that controls external urethral sphincter activity can be injured during vaginal delivery. Among other authors, Kerns et al. developed a murine model to study external urethral sphincter dysfunction associated with stress urinary incontinence due to pudendal nerve damage during childbirth [84]. The pudendal nerve was accessed dorsally and then crushed bilaterally in the ischiorectal fossa. The electromyograms showed motor units undergoing typical denervation changes followed by regeneration and recovery. Pan et al. used a similar model [85] and demonstrated greater alterations in histological muscle atrophy in diabetic rats, as well as a reversible reduction in leak point pressure versus the control group. Recently, Castiglione et al. [86] used a rat model for pudendal neuropraxia to test betamethasone as a treatment for bladder dysfunction and showed that pudendal nerve crush leads to stress urinary incontinence with irregular micturition by involuntary muscle contractions.

This model has been used to test stem cells as a therapy in a rat model of fecal incontinence [87]. However, no significant differences in either resting or peak pressure of the anal sphincter following treatment at 10 days postinjection with mesenchymal stem cells were observed. Although this model is useful for investigating mechanisms of neuromuscular recovery and for testing neuroregenerative agents and can be used in conjunction with other pathological animal models to assess mechanisms of injury and recovery in populations at increased risk for stress urinary incontinence, its short-term durability could represent a handicap in cell-based therapy investigations.

Other animal models combine the two methodologies vaginal distension and pudendal nerve crush to mimic stress urinary incontinence after childbirth. Song et al. developed a murine model combining vaginal distension and pudendal nerve crush to study the long-term effects of damage associated with childbirth [88]. Six weeks after surgery, there was no difference in the leak point pressure data between the vaginal distension/pudendal nerve crush and the sham group. However, after 9 weeks distal nerve regions were diffuse and were innervated by tortuous and multiple axons, demonstrating that reinnervation of the area was still in progress and could explain female recurrent stress urinary incontinence after the recovery following the first childbirth.

In conclusion, reversible stress urinary incontinence models such as vaginal distension, pudendal nerve crush, or a combination of both may be useful to evaluate pathophysiological mechanisms associated with stress urinary incontinence after childbirth, urethral damage and recovery, and neuromuscular recovery to test potentially neuroregenerative treatments and to identify favoring situations for recurrence of stress urinary incontinence. Cell therapy investigations using the vaginal distension model would be interesting, but due to their reversibility in the absence of treatment, they must be combined with other stress urinary incontinence models to ensure structural and functional viability of the injected cells.

In 1999, Kato et al. developed a canine model to evaluate the effects of urethrolysis and periurethral nerve resection on the capacity and compliance of the bladder [89]. Their experiments laid the groundwork for others, such as Rodriguez et al., to establish a murine model of stable and durable urethral dysfunction by transabdominal urethrolysis that resulted in neuromuscular and functional differences between incontinent and sham-operated rats [90]. Urethrolysis was performed by making a lower abdominal midline incision. After the bladder and the urethra were identified, the proximal urethra was detached circumferentially by incising the endopelvic fascia, and the remaining urethra was detached from the anterior vagina and the pubic bone. They observed a significant decrease in retrograde urethral perfusion pressure. The mean leak point pressure and retrograde urethral perfusion pressure decreased to 9.8 cm H2O and 11.2 cm H2O, respectively, at 1 week post-urethrolysis, and these changes were maintained for up to 24 weeks. After urethrolysis, histological analysis showed a 65% reduction in urethral smooth muscle with a high correlation between retrograde urethral perfusion pressure and muscular atrophy. The number of apoptotic cells in the urethra and bladder was also significantly higher, with the most obvious apoptosis rate in the submucosa and muscle layers. Since then, others have improved this technique showing the reproducibility of the urethrolysis mechanism as an animal model for lasting stress urinary incontinence with decreased urethral resistance and connective tissue damage for at least 8 weeks and up to 12 weeks [91]. Recently, Skaff et al. developed a durable animal model of stress urinary incontinence in rabbits by using urethrolysis which resulted in a significant decrease in the leak point pressure (from 33 to 12 cm H2O) and a decrease in 22% of the smooth muscle density for up to 12 weeks according to histological analysis [91].

Although the urethrolysis model induces a severe type of incontinence that is not entirely representative of the complex etiologic factors associated with stress urinary incontinence, and it is not known if full physiological recovery to the normal preoperative status could occur in such a destructive model, these studies show how the alteration of anatomical structural components of the continence mechanism plays a key role in the development of stress urinary incontinence. The urethrolysis model could represent severe forms of incontinence associated with non-sparing prostatectomy during cystectomy or urethral intrinsic sphincter deficiency or apoptosis of the muscle that correlates with stress urinary incontinence and age. However this model may or may not be reflective of stress urinary incontinence in women over a lifetime.

Kinebuchi et al. used this methodology in a rat model of stress urinary incontinence which combined urethrolysis with cardiotoxin injection and tested the use of bone-marrow-derived mesenchymal stem cells as a cell-based therapy [92]. In this model cultured autologous bone-marrow-derived mesenchymal stem cells were injected into the periurethral tissues, and leak point pressure was measured before injury and during the follow-up for up to 13 weeks postinjection. Despite that the authors view that the bone-marrow-derived mesenchymal stem cells group transplanted cells survived and differentiated into striated muscle cells and peripheral nerve cells and that there was a clear trend toward recovery of leak point pressure in bone-marrow-derived mesenchymal stem cells-transplanted urethras, no significant effect was detected. Although this study did not result in improved continence, the significant reduction in smooth muscle and correlation with urethral pressure observed in this model [90, 91] suggests that it may be useful for investigators interested in muscular regeneration of the internal urethral sphincter composed of smooth muscle. This may be especially important since it has been suggested that the density of the circular smooth muscle decreases up to 50% with age which may account for age-related declines in urethral closure pressure, a predominant factor associated with stress urinary incontinence and aging in women more often than loss of urethral support.

Like urethrolysis, electrocauterization achieves a similar outcome of stress urinary incontinence but via a different method. The most representative animal model of periurethral electrocauterization was developed by Chermansky et al. in which cauterization of the periurethral region 1 cm from the bladder neck to the upper face of the symphysis of the pubis for 30 s at an elevated temperature (up to 1204 °C) was performed in rats [93]. Sphincter function was assessed by measuring the leak point pressure through a suprapubic bladder catheter. Before the measurement, the spinal cord was transected at the T9–T10 level to eliminate the micturition reflex with increased intravesical pressure. Six weeks after surgery, cauterized rats with sphincter damage demonstrated a significant decrease in leak point pressure versus the sham-operated group, and differences remained for 16 weeks. Histological evaluation of the midurethra at 6 and 16 weeks post-electrocauterization showed an alteration in neuromuscular periurethral striated fibers, which was even more evident in the later evaluation. Although to date self-regeneration of striated muscles after electrocauterization has not been documented and therefore one may think that this is a good model to test cell-based therapies, we should note that this model mainly reproduces damage of the urethral sphincter that can occur accidentally during a surgical procedure, such as during prostatectomy in men or any procedure approaching the urethra and anterior vagina in women. Therefore, this type of stress urinary incontinence could represent a conventional model.

However, it is not fully representative of the complex etiologic factors associated with stress urinary incontinence, and like urethrolysis, it is unclear if full physiological recovery to the preoperative state occurs. Having said that, regenerative therapies have been tested in rat models of stress urinary incontinence induced by electrocauterization.

For example, Lim et al. used mononuclear cells isolated from human umbilical cord blood as a source of stem cells [94]. Four weeks after injection, leak point pressure values were significantly higher in the experimental group, and the external urethral sphincter of the experimental group remained well organized and intact, while the saline-treated control group showed severe degeneration and disruption of the sphincter muscle layers. However, the mechanism as to how the injected human umbilical cord blood mononuclear cells acted on the urethral sphincter is not explained in this work. Although they found human cells in the lamina propria and in the striated sphincter layers at 2 weeks, they could not find human cells at 4 weeks. In a previous study, intraurethral injection of muscle-derived cells improved urethral sphincter function in an electrocauterized rat model of intrinsic sphincter deficiency [95]. Despite that they showed improvement in sphincter function after 4 weeks and that the muscle-derived cells had integrated within the striated muscle layer of the cauterized tissue, they could not answer the question whether muscle-derived cells were bulking the urethra or promoting the functional recovery of the injured sphincter. In another study the use of muscle precursor cells isolated from limb myofibers of rats were used to regenerate the sphincter in an electrocauterized rat model of stress urinary incontinence and showed partial muscle regeneration in the context of denervation and fibrosis as demonstrated by histologic staining and immunofluorescence detection of beta-galactosidase expressed in the injected muscle precursor cells [96].

Similar models of direct sphincter damage by muscle resection have been developed as a model not only for stress urinary incontinence but also for fecal incontinence [97]. In 2007 Praud et al. developed three mouse models of stress urinary incontinence (freezing, longitudinal sphincterotomy, and periurethral injection) and observed that only sphincterotomy caused stress urinary incontinence as measured by a decrease in urethral closure pressure 21 days after surgery [98]. Eberli et al. described a durable canine model of sphincter deficiency by microsurgical abdominal sectioning of 25% of the periurethral sphincter (smooth and skeletal) muscle that resulted in a decrease in urethral profiles up to 7 months post-surgery and an absence of muscle cells [99]. Furthermore, in vivo pudendal nerve stimulation confirmed the loss of sphincter tissue function. Therefore, if this model [99] is reproducible in other animals and results in long-lasting stress urinary incontinence, it may represent a tool for the evaluation of methods reproducing sphincter function and intending to regenerate smooth and striated muscle of the internal and external urethral sphincter, respectively.

Although this model does not represent the multifaceted etiology of stress urinary incontinence, one needs to consider that a major postoperative complication of radical prostatectomy is damage to the muscle-nerve-blood vessel units around the urethra. Vaginal delivery also causes varying degrees of neuromuscular damage. Such a microsurgical approach to remove the sphincter muscle offers the advantage that it produces durable damage and that the surgical techniques are relatively simple and reproducible. In addition, it can generate a direct muscle injury without pronounced nerve damage therefore potentially allowing a better direct assessment of a muscle regenerative therapy with easier electromyographic measurements. However, like the other destructive models, it is not known if full physiological recovery to the normal preoperative status can occur.

Animal models of incontinence by transection of the pubourethral ligament, which removes the structural support of the urethra, and/or transection of the pudendal nerve generated stress urinary incontinence for up to 1 month. Kefer et al. compared the two methods and transected the pubourethral ligament between the middle urethra and posterior symphysis of the rat and the pudendal nerve via a dorsal incision and the bilateral ischiorectal opening [100, 101]. After 28 days, a decrease in leak point pressure was observed in the rats with pubourethral ligament or pudendal nerve transection compared to the sham-operated group. Following pubourethral ligament transection, histologic studies demonstrated an absent pubourethral ligament for up to 4 weeks post-injury. Therefore, the long-term durability of this model is unclear at this point. Applications directed at treating hypermobility of the urethra could benefit from the use of this model. Pubourethral ligament could have similar effects to the urethrolysis model, since transectioning the ligaments may damage vascular, nerve, and possibly muscle structures, and could also be combined with other models to investigate cell therapies, such as transection of the pudendal nerve, favoring the neurogenic atrophy and decreased periurethral neurofilaments.

Bilateral pudendal nerve transection has been used in several animal models as a mechanism of a stress urinary incontinence model due to indirect periurethral sphincter damage. In 2003 Kamo et al. demonstrated the importance of the pudendal nerve in the continence mechanism in a murine model by demonstrating an 80% decrease in the muscle response at the midurethral level when transectioning the bilateral pudendal nerves and the nerves to the iliococcygeus and pubococcygeus muscles, but not by transection of the visceral branches of the pelvic nerves and hypogastric nerves [101]. Their data also showed that in stress conditions such as sneezing, the contribution of pudendal nerve-mediated striated muscle activity in the external urethral sphincter to the continence mechanism may be greater than that of iliococcygeus and pubococcygeus muscle activity. Peng et al. investigated the effect of the unilateral versus the bilateral transection of the pudendal nerve in rats [102]. After 6 weeks they observed a significant decrease in leak point pressure and striated muscle atrophy in rats that underwent unilateral transection of the pudendal nerve, which was even more prominent in the bilateral transection of the pudendal nerve group based on voiding efficiency. Other non-rodent animal models of stress urinary incontinence have been developed as well, including cats, dogs, and more recently nonhuman primates. Badra et al. developed a model for stress urinary incontinence in premenopausal female primates by abdominal bilateral transection of the pudendal nerve and cauterization of the pudendal nerve [103]. Urodynamic investigations were performed before injury and 3, 6, and 12 months after injury. Electromyography prior to necropsy showed decreased levels in the transection of the pudendal nerve group versus the sham group. Histological and inmunohistochemical analysis showed a decrease in smooth and striated muscle fibers, increased collagen, and decreased vascularization in injured animals. Cystogram results were consistent with those results showing greater amplitude and incompetence of the bladder outlet and the external urethral sphincter.

This model offers the advantage that it may be representative of neurogenic damage that occurs in radical prostate surgery or urethral denervation and/or pudendal nerve and muscular damage that probably occurs during vaginal delivery [102]. These studies demonstrate the feasibility of using this model in cell therapy approaches investigating regeneration of neuromuscular tissue. It also demonstrates that transection of the pudendal nerve affects the structural and functional anatomical capacity of the periurethral urethral sphincter in a human-like animal model.

Despite the success of such stress urinary incontinence animal models, most if not all cell therapy regenerative assessments have been performed to date in murine models (rats or mice) using different cell types such as human or allogeneic skeletal muscle stem cells [104, 105], autologous or allogeneic adipose-derived stem cells [106, 107], or allogeneic bone-marrow-derived mesenchymal stem cells [66, 108]. These aspects of regenerative cell therapies will be discussed in the next section.

Therapies using mesenchymal stem cells have been used in stress urinary incontinence rodent models of vaginal distension and various strategies of bilateral transection of the pudendal nerve. Dissaranan used bone-marrow-derived mesenchymal stem cells in a vaginal distension rat model to investigate homing of mesenchymal stem cells to pelvic organs after i.v. injection and recovery from simulated childbirth injury [59]. However, a limitation of this study was that their assessments were short term (after 1 week postinjection) and mesenchymal stem cells from passage 16 were used. Interestingly though, they did show that the vaginal distension model disrupted the smooth and striated muscle of the urethral sphincter and that mesenchymal stem cells preferentially engrafted in the smooth muscle of the urethra and vagina. Although there was an improvement in leak point pressure, the external urethral sphincter function was not improved. While this study suggests that mesenchymal stem cells can home to the damaged tissue and may aid in some level of recovery in continence, other methods of injection such as direct injection into the sphincter may produce different results.

Corcos et al. and Kim et al. tested periurethral injection of mesenchymal stem cells in rats that underwent bilateral transection of the pudendal nerve and reported an improvement in urinary continence (Valsalva leak point pressure) as measured by cystometry at 4 weeks [66, 108]. Mesenchymal stem cells survived and were viable in the injected area based on immunohistochemistry analysis. Histopathologic regeneration of striated muscle was observed by desmin and myosin staining, respectively. However this approach is far from optimal since it is well known that fluorescent dyes coming out of transplanted cells can be transferred to surrounding cells and can mistakenly be taken as evidence for differentiation. Other authors have evaluated the use of biomaterials combined with mesenchymal stem cells for sphincter regeneration. Du et al. evaluated submucosal injection at the level of the urethra and bladder neck of an alginate/calcium gel alone or in combination with mesenchymal stem cells or muscle-like predifferentiated cells in vitro with 5-azacytidine in a rat stress urinary incontinence model that used a bilateral transection of the pudendal nerve model combined with transectioning of the nerves innervating the iliococcygeus/pubococcygeus muscle [109]. The three experimental groups showed an improvement in leak point pressure at 1, 4, and 8 weeks posttreatment compared to the control group, and immunohistochemical analysis resulted in desmin- and alpha-SMA-positive cells in all three groups, without observing clear and obvious specific muscle regeneration. In an analysis performed at 8 weeks after injection, the gel still maintained a certain volume, and therefore it cannot be ascertained whether the improvement in leak point pressure was due to a bulking effect since there was no group injected with cells alone. A remarkable fact is that addition of the calcium alginate composite gel seemed to stimulate angiogenesis with a higher density of newly formed microvessels in the cell plus gel group, which was not observed in the gel group alone.

In a combined urethrolysis and cardiotoxin injection stress urinary incontinence female rat model, bone-marrow-derived mesenchymal stem cells differentiated into striated muscle cells and peripheral nerve cells in vivo, but only resulted in a small improvement in urodynamics [92]. A recent study tested different combinations of what was defined by the authors as “early” muscle, neuron, and endothelial cells that were differentiated from human amniotic fluid stem cells for 7–21 days in vitro prior to injection in a transection of the pudendal nerve female mouse model [110]. Histological and immunohistochemical analysis revealed that regeneration of the sphincter muscle was accelerated in the muscle group through fusion with damaged cells, as well as activation of the local muscle progenitor cells, but the enhancement was less pronounced and the regenerated muscle quality was lower compared to the muscle/neuron/endothelial cell group. The muscle/neuron and muscle/neuron/endothelial cell groups showed increased muscle bundles in the urethral sphincter, while the triple combination group showed areas of connective tissue masses at the urethral sphincter and higher nerve and endothelial cell marker expression. Real-time PCR results also demonstrated higher levels of myogenic, neuronal, and endothelial marker expression in the muscle/neuron/endothelial cell group. At 4 weeks, the leak point pressure and closure pressure was significantly higher for each of the three cell injection groups versus the control which the authors propose may be due to tissue edema and increased outflow resistance due to over injection. This suggests that in addition to the effect on myogenesis, mesenchymal stem cells may have a synergistic effect on neurogenesis, the formation of neuromuscular junctions and angiogenesis during regeneration.

The limitation of these studies was a short-term follow-up (4–8 weeks) for functional analyses. Another consideration is the injury models used. For example, if a denervated urethral sphincter (e.g., after bilateral transection of the pudendal nerve) is treated with a cell-based therapy, can the regenerated tissue become innervated in a physiological manner? Reinnervation of muscle tissue after injury in such a model is necessary for functional recovery to occur. Mesenchymal stem cells are known to generate neurotrophic factors and can promote endogenous neuronal growth [111, 112], and the use of biomaterials or delivery products may enhance such an effect, but are they enough to reinnvervate the tissue?

The use of adipose-derived stem cells or derivatives has been evaluated in different murine models of stress urinary incontinence using vaginal distension or transection of the pudendal nerve models [67, 106, 113, 114]. In most of the cases, the functional assessment after adipose-derived stem cells injection was evaluated by cystometric parameters (bladder capacity, leak point pressure, and retrograde urethral perfusion pressure) and reported a significant increase in these values in the cell-injected groups compared to control groups. However, the muscle regenerative capacity or the in vivo mechanisms of these cell sources to achieve such results are not well defined. The incontinence models most commonly used with this cellular source were investigated using the vaginal distension model as shown by Obinata et al. [114]. Although their group demonstrated a recovery in muscle atrophy starting at 2 weeks, which was fully evident at day 28, it is difficult to assess to what extent muscle tissue regeneration of the damaged tissue was achieved since muscle regeneration markers included collagen I/III, elastic fibers, and alpha-SMA which do not guarantee the regeneration of a quality and morpho-functionally competent smooth and/or striated muscle. What is evident in these models is the cell viability and the paracrine capacity of adipose-derived stem cells at the injection site.

Zhao et al. [106] and Obinata et al. [114] reported cell survival and viability of adipose-derived stem cells in the injection area. In a vaginal distension stress urinary incontinence model, Zhao et al. [106] demonstrated how the use of a biomaterial-derived hydrogel (polylactic-co-glycolic acid, PLGA), in conjunction with cell therapy (mature adipocyte-derived dedifferentiated fat cells), can be used to generate microparticles (growth factor delivery “vehicles”) that allow the introduction of certain factors such as nerve growth factor to improve the microenvironment and generate neurofilaments. However, it is not clearly defined if the improvement in bladder capacity of this group (mature adipocyte-derived dedifferentiated fat cells + PLGA/nerve growth factor) versus control is due to the self-improvement of the cellular microenvironment in vivo or an increased bulking effect. Li et al. [67, 113] observed that some adipose-derived stem cells may exhibit a vascular endothelial growth factor paracrine activity. Moreover this group showed increased vascular density and increased expression of p-ERK1/2, which could induce favorable changes in the extracellular matrix and cause trophic effects of immunomodulation and cell survival, thus improving the microenvironment for regeneration. Mature adipocyte-derived dedifferentiated fat cells alone have also promoted functional recovery from spinal cord injury-induced motor dysfunction in rats [115], suggesting reinnervation may occur with adipose-derived stem cells or mature adipocyte-derived dedifferentiated fat cell treatment.

However, the same question arises as with mesenchymal stem cells; can cell-based therapy lead to innervation in a physiological manner after bilateral transection of the pudendal nerve or pudendal nerve crush (which may in fact be more representative of postpartum stress urinary incontinence)? Some groups have suggested that factors such as brain-derived neurotrophic factor may need to be delivered with the stem cells to lead to such an effect.

Cell therapy with muscle-derived cells and muscle-derived stem cells has been tested in murine models of stress urinary incontinence by pudendal or sciatic denervation [105, 116]. Kwon et al. showed that injection of muscle-derived cells, fibroblasts, or both can produce a functional improvement of continence in leak point pressure [116]. However, electromyographic activity demonstrated that only the muscle-derived cells had the ability to differentiate into myotubes and myofibrils and integrated into the sphincter tissue. This model also suggests that a transient bulking effect, but not morphological and functional regeneration, could occur by injection of fibroblasts since the improvement reported was dose dependent and a linear increase in leak point pressure at increasing fibroblast doses occurred, except at high doses of fibroblast injection which caused urinary retention. In order to improve the disadvantages of using muscle-derived stem cells including migration and absorption, Xu et al. studied the effect of adding a biodegradable fibrin glue to the periurethral muscle-derived stem cells injection therapy [105]. They observed that although both groups injected with muscle-derived stem cells or muscle-derived stem cells + fibrin glue showed a significant improvement in the leak point pressure, only the muscle-derived stem cells implanted in combination with the fibrin glue appeared to integrate into the host tissue and result in an increase in survival and a tendency to form multinucleated myofibrils. These results demonstrated that the addition of biomaterials such as fibrin may not be necessary to generate an early functional improvement based on the analysis of the leak point pressure, but may promote the implantation and differentiation of muscle-derived stem cells into striated muscle.

Other strategies have included the isolation and implantation of muscular precursor cells from muscle biopsies. Some results suggest that the use of myoblasts could be a potential treatment for damaged urethral sphincter muscle regeneration. Praud et al. performed periurethral injections of skeletal myoblasts in a stress urinary incontinence murine model of sphincterotomy and reported a significant improvement in urethral closure pressure [98]. However, there was only fusing of myoblasts and regeneration of new myofibers when the cells were grafted into normal striated skeletal muscle, which was not observed in the incontinent animals. Due to a large similarity between human and porcine rhabdosphincter muscle cells, Mitterberger et al. investigated myoblast injection in a non-incontinent pig model [117]. They showed cell dose-dependent functional effects. Thus, injection of high cell numbers (7.8 × 10⁷) led to a clear increase in urethral pressure values at 3 weeks compared with low cell dose groups (4.4 × 10⁷). However, the mechanisms underlying these effects were not explained. In the histological analysis, integration and differentiation of myoblasts was observed, as well as an increase in the formation of new myofibrils in the high cell dose injected groups. Likewise, Yiou et al. demonstrated that periurethral injection of muscle precursor cells in a murine stress urinary incontinence model caused by electrocauterization promoted the formation of myotubes and neuromuscular junctions [118]. Muscle precursor cell injection resulted in formation of myotubes and 40% restoration of the sphincter function 1 month after the injection.

While most studies on cell therapy using myoblasts injected into the bladder and urethra were performed in murine models, recently Eberli et al. investigated the use of muscle precursor cells in a sphincterotomy canine model of stress urinary incontinence [99]. Their group showed that injection of autologous muscle precursor cells, with a low concentration of collagen used as a cell carrier to provide a microenvironment for the muscle precursor cells, significantly decreased urethral pressure values and was structurally and functionally viable after 6 months. Histological analysis revealed the capacity of not only skeletal muscle development but also nerve development, demonstrating that the injected muscle precursor cells are able to survive in vivo for an extended period of time. Thus, they showed histologically how injected muscle precursor cells grow in clusters and form new myofibrils starting 1 month after the treatment and at 3 and 6 months new muscle fiber formation occurred, although thinner and less organized than native muscle tissue. Moreover, staining with S100b showed that nerve fibers of different sizes were present in between the newly formed muscle tissue in the cell-treated group. Urodynamic evaluations showed an improvement in static, stress, and detrusor leak point pressure, electromyographic parameters demonstrated an improvement in pressure during stimulation of the pudendal nerve, and anatomical parameters (cystourethrogram) showed a urethrovesical morphology in the muscle precursor cell-injected group comparable to normal anatomy at 1, 3, and 6 months postinjection. Again, this highlights the importance in survival and integration of the cells using a cell carrier in the context of cell therapy.

Kim et al. evaluated the feasibility of using human muscle precursor cells that were injected periurethrally in a sciatic sectioning murine model of incontinence [119]. They showed homing of the muscle precursor cells to the sphincter by immunofluorescence using human-specific antibodies. An improvement in leak point pressure at 4 weeks posttreatment was reported.

Lim et al. used human umbilical cord blood as a treatment for stress urinary incontinence [94]. A significant improvement in leak point pressure 4 weeks after cell injection was reported, as well as homing of DiI-labeled injected cells to the lamina propria and the muscular urethral sphincter at 2 weeks postinjection. However, 4 weeks after injection human cells could not be detected in the urethral tissue, perhaps because the iron oxide label of the injected cells was already metabolized at that point or because the injected cells were phagocytized and lost. Histologically, desmin staining in the control group showed a markedly disrupted sphincter muscle with atrophic muscle layers and collagen deposition, while the sphincter muscle was restored without damage in the cell-treated group. Despite a functional improvement in the human umbilical cord blood group, the mechanism as to how the injected human umbilical cord blood mononuclear cells act on the urethral sphincter is not explained in this work.

Recently, human amniotic fluid stem cells have been proposed as a stem cell source for various cell therapies and tissue engineering. Thus, Kim et al. used amniotic fluid stem cells in a bilateral transection of the pudendal nerve murine model of stress urinary incontinence and demonstrated an improvement in the leak point pressure and closure pressure (using the vertical tilt/intravesical pressure clamp model) in the cell-transplanted group after 4 weeks of injury [120]. In this work, FACS analysis showed a lower expression of HLADR in the human amniotic fluid stem cells, and their injection did not stimulate CD8 T cell infiltration into the injected area or tumor formation after 8 weeks postinjection, suggesting that these cells are immunologically tolerated and/or result in an immunosuppression effect and have low tumorigenic potential. By immunohistochemistry using a human-specific antihuman nuclei antibody, they confirmed that the injected cells were able to survive; however, at days 3–7 cell migration from the injection site to the periurethral area was reported, and at day 14 a loss of transplanted cells was observed. This is especially interesting since in vivo cell tracking was performed and showed homing of the injected cells and cell clustering at the injection site during the first 10 days postinjection. Moreover, they reported muscle regeneration in the periurethral region in the experimental group by H&E staining and confirmed this result by MyoD antibody staining and real-time PCR of early and late striated muscle cell marker expression. Alpha-bungarotoxin for the acetylcholine receptor expression was reported to be similar between the experimental and control group. The experimental group also showed higher neurogenic gene expression (nestin, vimentin, neurofilament, microtubule-associated protein 2, β III tubulin, glial fibrillary acidic protein) than the control group. Subsequently in 2014, the same authors published the use of a triple combination treatment of muscle, neuron, and endothelial cells derived from human amniotic fluid stem cells as a cell therapy in the same mouse model of stress urinary incontinence described above [110]. This time they established four experimental groups: (1) amniotic fluid stem cells, (2) a single-cell group containing early muscle progenitor differentiated cells, (3) a double-cell group containing muscle/endothelial or muscle/neuron progenitor cells, and (4) a triple-cell-combination group with muscle/neuron/endothelial progenitor cells. Functional evaluation of stress urinary incontinence as evaluated by leak point pressure and closure pressure reported a significant improvement in all experimental groups compared to the control group, with the triple-cell-injected group showing superlative improvement compared to other experimental groups. Again CD8 T cell evaluations demonstrated a low immunoreaction in the experimental groups versus the control group. The cell tracking method used in the previous work showed the best result in terms of homing for the triple-cell-combination injection. Using the same techniques as in the previous work, muscle regeneration and increased formation of blood vessels were reported in all groups but to a higher extent in the triple-cell-injected group. This work reinforces their previous work and adds the option of combining human progenitor cells obtained from human amniotic fluid stem cells, suggesting that mesenchymal stem cells may have a synergic effect on myogenesis, neuromuscular junction formation, and angiogenesis during the regeneration of the damaged urethral sphincter.

Finally, Bandyopadhyay et al. reported the effectiveness of human myoblasts injection in a botulinum-A toxin injection murine model of stress urinary incontinence [121]. Significant improvement in the recovery of urination volume and muscle atrophy was observed. Following myoblast injection, GFP-positive myoblast cells and urethral staining for human desmin demonstrated homing of the cells to the urethra. They also demonstrated formation of myotubes and an increase in the thickness of the periurethral muscle in the treatment group compared to the control group.

Different types of stem cells have been used for stress urinary incontinence treatment in humans as well as in animal models including skeletal muscle derived (myoblasts, satellite cells, muscle progenitor cells, and muscle-derived stem cells), bone marrow stem cells (bone-marrow-derived mesenchymal stem cells), human umbilical mononuclear cells, adipose-derived stem cells, and modified/sorted stem cells which show various efficacy [23]. Skeletal muscle-derived cells, including myoblasts, muscle-derived stem cells, and fibroblasts, were the most commonly used in recent clinical trials [123–133]. Adipose-derived regenerative cells [134, 135], cord blood stem cells [136], and total nucleated cells and platelets [137] were tested less frequently. Almost authors have not reported any significant advantage from the use of any particular type of cells thus far.

The number of cells used for the treatment of stress urinary incontinence differed considerably between the studies and ranged from 0.6 mln to 1020 mln cells. However, it is still unclear by which route the implantation of cells should take place: peri- or transurethral, injected into the sphincter or submucosa, or both? Should injections be performed in single or repeated doses? How long after the first series of injections should the following one be performed? Which are better to implant: freshly isolated or expanded in vitro cells? These questions remain unanswered.

Recent review article [138] showed that the clinical outcomes of cell therapy in stress urinary incontinence based on 616 patients indicated that mean cure and improvement rates over a 12 month follow-up were 37.2 ± 29.7% and 33.1 ± 14.3% (patients), respectively (Table 10.2). A comparative analysis of cross-sex outcomes indicated that the cure rate was significantly higher in female (41 ± 30.7%) compared to male patients (28.7 ± 31.5%) (p < 0.01).