This open-label nonrandomized phase I clinical trial was registered on clinicalTrials.gov (#NCT00472069)
What's known on the subject? and What does the study add?
Cell therapy using muscle precursor cell (MPC) injections has shown promise for urinary incontinence due to intrinsic sphincter deficiency (ISD), but the cell-preparation process is complex and costly. Implantation of freshly isolated myofibres carrying MPCs, mainly satellite cells, was very efficient in repairing muscle damage in recent animal experiments.
In a phase I clinical trial, we investigated whether periurethral myofibre implantation generated local myogenesis and improved continence in 10 patients (five men and five women) with ISD. We found that myofibre implantation increased intraurethral pressure and periurethral electromyographic activity in patients with ISD. There were no serious side-effects.
To assess the safety of periurethral myofibre implantation in patients with urinary incontinence due to intrinsic sphincter deficiency (ISD)
To assess the resulting myogenic process and effects on urinary continence.
Patients and Methods
An open-label non-randomised phase I clinical trial was conducted in five men and five women with ISD (mean age, 62.5 years).
A free muscle strip from the patient's gracilis muscle was implanted around the urethra as a means to deliver locally myofibres and muscle precursor cells (MPCs).
Patients were assessed for collection formation and incomplete bladder emptying.
The maximum urethral closure pressure (MUCP) and concomitant periurethral electromyographic (EMG) activity were recorded before surgery and 1 and 3 months after surgery. Continence was assessed using the 24-h pad test and self-completed questionnaires, for 12 months.
There were no serious side-effects.
Continence improved significantly during the 12-month follow-up in four of the five women, including two who recovered normal continence. In the women, MUCP increased two-fold and de novo EMG periurethral activity was recorded. In the men, MUCP and EMG recordings showed similar improvements but the effect on continence was moderate.
The few patients enrolled could affect these results.
This is the first report of a one-step procedure for transferring autologous MPCs via myofibre implantation in patients with ISD.
EMG and urodynamic assessments showed improvement of periurethral muscle activity.
Further work is needed to confirm and improve the therapeutic efficiency of this procedure.
International Consultation on Incontinence Questionnaire
intrinsic sphincter deficiency
muscle precursor cell
mean power frequency
final MPF value
initial MPF value
maximal urethral closure pressure
root mean square
final RMS value
initial RMS value
striated urethral sphincter
University of California Los Angeles, Prostate Cancer Index
(stress) urinary incontinence
visual analogue scale
Urinary incontinence (UI) is a global health problem that affects ≈10% of all men and women [1, 2]. Intrinsic sphincter deficiency (ISD) refers to the deterioration of the striated urethral sphincter (SUS) and is associated with severe forms of UI . Implantation of the artificial urinary sphincter (AUS) AMS800® is the reference standard for treating severe cases in both men and women [3, 4]. However, this procedure is associated with significant long-term risks of urethral erosion and implant infection [4-6]. Less invasive procedures, e.g. the adjustable continence therapy device (Pro-ACT®) , sub-urethral compressive slings , and bulking agents [4, 9], are not sufficiently effective to correct severe UI.
ISD can be viewed as a focal muscle disorder and therefore as a candidate for cell therapy. Among the myogenic cell sources that might hold promise for strengthening the incompetent SUS, skeletal muscle precursor cells (MPCs) are the most extensively studied [10-14]. MPCs consist mainly of satellite cells, a small population of quiescent cells located beneath the basal lamina of adult myofibres. In various animal models of ISD, intraurethral transplantation of MPCs improved SUS contraction [11, 12, 14-17].
Most studies of MPC therapy used suspensions of autologous cells obtained from muscle biopsies, processed by enzymatic digestion, and expanded in vitro before injection [18-20]. This approach has major drawbacks that limit its efficiency. One of these drawbacks is the complex and costly process required for MPC preparation, which requires a dedicated cell-therapy unit and raises safety concerns. This drawback limits the ability to make MPC therapy available to the large number of patients with ISD. Furthermore, there is now increasing evidence that enzymatic separation of satellite cells from their parental myofibres and exposure to culture conditions dramatically diminishes their myogenic potential [21-23]. Collins et al.  previously showed in an animal model of irreversible muscle injury that the implantation of a single myofibre with about seven non-manipulated satellite cells could generate >100 new myofibres. In this study, separating satellite cells from their natural environment by enzymatic digestion resulted in massive cell death after injection, presumably because of cell membrane damage .
In a pig model of ISD, we previously investigated a method of MPC delivery designed to circumvent the above-mentioned problems [12, 24]. Freshly isolated muscle strips containing myofibre and natural satellite cells were implanted around the urethra. The rationale for this procedure is the remarkable myogenic potential of satellite cells maintained in their natural microenvironment [25-29]. The myofibres rapidly underwent degeneration and necrosis, and the satellite cells underwent in situ activation and differentiation into myoblasts . Within 1 month, the myoblasts fused to form myotubes that replaced the parental myofibres. Myofibre implantation also produced a neurotrophic paracrine effect that promoted the sprouting of urethral cholinergic nerve endings, resulting in the formation of functional motor units .
To the best of our knowledge, autologous myofibre transplantation has not been investigated in human patients as a means of improving local muscle function. Here, we report the results of a preliminary phase I clinical trial resulting from the translational research conducted in the pig mentioned above. In 10 patients with ISD we implanted myofibre strips around the urethra at the vicinity of the incompetent SUS to deliver non-manipulated satellite cells with their preserved environment. The objective of the present study was to assess the safety of the degeneration/regeneration process after transplantation (primary endpoint) and to investigate the resulting effects on urinary continence, urodynamic parameters, and periurethral electromyographic (EMG) activity (secondary endpoints).
Patients and Methods
This open-label nonrandomised phase I clinical trial was sponsored by the Assistance Publique des Hopitaux de Paris, approved by the appropriate ethics committee (Comité de Protection de la Personne Ile-de-France IX), and registered on clinicalTrials.gov (#NCT00472069). Written informed consent was obtained from all patients before study enrolment. The study was conducted between April 2007 and October 2008.
We enrolled 10 patients, five women and five men. Patients aged 40–75 years were eligible to participate if they had stress UI (SUI) related to ISD after failure of physical therapy (pelvic floor exercises). SUI was defined as a positive cough provocation test and a 24-h pad test >20 g. ISD was defined as a maximum urethral closure pressure (MUCP) of <40 cmH2O [30, 31] by urodynamic testing in both genders and, in women, by a negative Ulmsten test and a Q-tip test >40 ° indicating urethral immobility.
Patients were excluded if they had previous pelvic radiation therapy, haemostasis disorders, genetically determined muscular disease, UI of neurological origin, a positive Ulmsten test and Q-tip test >40 ° in women, incomplete bladder emptying, a maximum flow urinary rate <12 mL/s (debimetry), detrusor overactivity, a bladder capacity <300 mL (cystometry) and an urethral stricture (urethrocystoscopy).
The myofibre implantation procedure consisted of periurethral implantation of a muscle strip harvested from the left gracilis muscle and measuring 5 cm × 1 cm × 2 mm. General anaesthesia was used and a catheter was inserted into the urethra. Two fragments of the muscle strip were frozen and kept in a tissue bank.
In women, the urethra was approached via a 2-cm incision in the anterior vaginal wall and the muscle strip was wrapped in an Ω-shape around the distal third of the urethra in the first patient and around the mid-urethra in the following four patients.
In men, the strip was implanted around the membranous urethra approached via a perineal incision. The bulbospongiosus muscle was dissected out from the central perineal tendon, and the muscle strip was wrapped in an Ω-shape around the membranous urethra. In both genders, the ends of the muscle strip were secured using a 2/0 polyglactin (Vicryl®) suture without tension. The urethral catheter was removed on the following day.
The patients remained under close observation in the hospital for 6 days after the procedure to ensure the detection of any early adverse events. A physical therapist conducted an evaluation 1 month after the procedure to look for abnormalities in pelvic floor command. At baseline (1 month before implantation) then 1 (M1), 3 (M3), and 12 months (M12) after implantation each patient underwent a urinary continence assessment. A urodynamic study and endourethral surface EMG recording were performed at baseline, M1 and M3.
The primary endpoint was the occurrence of adverse events associated with the myofibre implantation procedure. Special attention was directed to detecting clinical symptoms of perineal haematoma or abscess formation; incomplete bladder emptying defined as a residual volume of >20% of the voided volume; and dysuria defined as a maximal urinary flow rate of <12 mL/s. The perineum and thighs were examined every day during the hospital stay then at M1, M3, and M12. At these time points, post-void residual urine volume was measured using ultrasonography, body temperature was recorded, and pain intensity was scored on a 0–10 visual analogue scale (VAS). An urethrocystoscopy was performed before and 3 months after implantation to check the absence of urethral stricture.
The secondary endpoints were chosen to reflect the effects of myofibre implantation on urinary continence, urodynamic parameters, and periurethral EMG activity. Continence was assessed clinically using the 24-h pad test as previously described , in combination with the following self-completed questionnaires: Contilife  in women, International Consultation on Incontinence Questionnaire (ICIQ)-UI Short Form  in both genders, and the University of California Los Angeles, Prostate Cancer Index UCLA-PCI  in men.
A response to myofibre implantation was defined as a >50% decrease in the 24-h pad test result at M3 and a complete response as absence of pad use at M12. Patients with persistent UI 6 months after myofibre implantation were given the option of conventional surgery for ISD. If conventional surgery was performed, special attention was given to the intraoperative detection of any local changes associated with myofibre implantation.
Urodynamic testing was performed using a Duet Encompass system (Medtronic, Minneapolis, MN, USA) with a three-lumen multi-perforated catheter for cystometry and urethral pressure profilometry (9 F, 400 mm, Mediwatch Plc, Rugby, UK). A three-electrode probe (FSR-03; Inomed, Teningen, Germany) 1 mm in diameter and 30 cm in length was attached to the pressure catheter using sterile adhesive tape for concomitant EMG signal recording (Fig. 2A).
All tests started with flow measurements and ultrasonographic estimation of post-void residual urine volume. Next, MUCP and periurethral EMG were recorded concomitantly, both at rest and during sphincter contraction (urethral pressure profilometry). The pressure catheter was then placed at the site where the MUCP was obtained. The bladder was filled at 70 mL/min, up to 100 mL then 200 mL. The patient was asked to produce maximal sphincter contraction during 30s at both filling volumes, and EMG activity was recorded during the contractions (Fig. 2B,C). Filling was then continued until the maximal bladder capacity was achieved.
The EMG signals were characterised in the time and frequency domains by computing the root mean square (RMS) and mean power frequency (MPF) on 500-ms processing windows with 50% overlap. The mean values were computed over 2-s windows to eliminate artifacts due to the shortcuts present in some of the recordings.
During urethral profilometry, the computation window was centred on the MUCP. During cystometry, the signal that showed the greatest stability for 22 s of the total 30-s contraction time was kept for further analyses. The mean initial RMS and MPF values were measured during the first 2 s of the contraction (RMSi and MPFi) and mean final values during the last 2 s of the contraction (RMSf and MPFf) (Fig. 2C). The differential values (Δ, as percentages) were computed as [(f–i)/i]·100) to assess increases or decreases in RMS or MPF values during contractions.
The changes in clinical, urodynamic, and EMG results across the three time points (preoperative, M1, and M3) were assessed using repeated-measures anova. A P < 0.05 was considered statistically significant. In cases of significance, Tukey-Kramer multiple comparison post-tests were performed to compare M1 and M3 values to the preoperative value. In the first female patient (patient #1), a technical problem during the preoperative investigation precluded the analysis of EMG data and technical artifacts during one of the urodynamic tests precluded the analysis of MUCP during contraction. Therefore, only nine patients were included in the statistical analyses of contraction MUCP and of all EMG parameters.
All patients included in this study were referred to our institution with a diagnosis of severe UI due to ISD and were candidates for AUS implantation. The mean age was 62.5 years.
In the five men, the cause of ISD was radical prostatectomy (RP). None of the men had a history of surgery to treat UI. Of the five women, three (patients #1, #4, and #8) had a history of pelvic surgery: tension-free vaginal tape surgery had been performed unsuccessfully in two (#1 and #8) and the remaining patient (#4) had severe ISD after partial resection of the bladder trigone and urethra for removal of synthetic mesh implanted previously to treat a cystocele.
No serious adverse events were detected during the 12-month follow-up. No fluid collections formed in the perineum. The mean VAS pain scores at the thigh and perineum were 0/10 and 2.3/10, respectively, in women and 2.0/10 and 2.2/10, respectively, in men. None of the patients had a fever (defined as a temperature of >38 °C). In patient #9, incomplete bladder emptying resolved spontaneously after five days of self-catheterisation. No post-void residual volume or dysuria was noted in the other patients. Postoperative urethrocystoscopy did not show any sign of urethral stricture.
Table 1 reports the results of the clinical assessments. In the overall population, the 24-h pad test value and ICIQ score decreased significantly from baseline to M3. The improvements were more marked in the women than in the men.
Table 1. The results of the clinical assessments of continence.
Patient # (gender)
Overall population (N = 10), mean (sd)
Women (N = 5), mean (sd)
Men (N = 5), mean (sd)
P-values obtained by repeated-measures analysis of variance (rm-anova) comparing preoperative data to data after 1 month (M1) and 3 months (M3) are in bold type when they indicate a significant difference (P < 0.05) and underlined when P < 0.1. The significance of Tukey-Kramer post-tests comparing M1 and M3 values to the preoperative value is given as *P < 0.05; **P < 0.01; and ***P < 0.001. At M12, values are reported only for patients without other surgical procedures for UI during follow-up.
Of the five women, all but one (#1) had marked improvements in their symptoms after myofibre implantation. Their subjective perception of UI, as assessed by the ICIQ score, improved throughout the 12-month follow-up period. In addition, all Contilife questionnaire subscores (daily and effort activities and self-image) improved significantly from baseline to M3. In patient #1 whose symptoms did not improve, urodynamic testing showed a marked improvement, with MUCP normalisation at rest (see below). Less than 1 year after myofibre transplantation, she received an Pro-ACT® device with a good effect on the symptoms. In this patient, the myofibre were implanted around the distal third of the urethra. We hypothesised that the discrepancy between MUCP normalisation and symptom persistence was ascribable to excessively distal myofibre implantation, in an area that did not contribute to continence. Consequently, we implanted the myofibres more proximally in the other four women, around the mid-urethra, at the site of the SUS. This change was associated with better outcomes. Of the four other women, in whom the myofibres were implanted around the mid-urethra, two no longer used pads at M12, and the other two had 46% and 90% improvements in the 24-h pad test vs baseline.
In the men, the effects of myofibre implantation on the clinical symptoms were less marked. Nevertheless, MUCP and EMG activity showed similar increases to those in the women. Post hoc tests failed to detect any significant improvements vs baseline in the ICIQ and UCLA-PCI urinary function scores at M3, despite some improvements in the urinary bother items (P = 0.038) and pad test (P = 0.867, NS). For example, patient #5 had a four-fold decrease in the 24-h pad test (150 g at M3 vs 600 g at baseline), but this did not result in perceived symptom relief, as shown by the absence of changes in the ICIQ and UCLA-PCI scores. Because of the persistent severe UI, four of the five men had further surgical procedures, consisting of AUS implantation (patients #2 and #6), a sub-urethral sling (patient #5), or a Pro-ACT® device (patient #10). In all four patients, these procedures were followed by marked symptom relief. During surgery, no specific changes related to myofibre implantation were noted. The remaining man (patient #3) continued to have severe UI but declined AUSimplantation.
Figures 3, 4 and 5 report the urodynamic and EMG results. The detailed results for each patient are presented in the Supplementary Table S1. Overall, myofibre implantation resulted in a two-fold increase in MUCP at rest and during maximal voluntary contraction at M3 vs baseline (Fig. 3). The MUCP improvement was significant in patients of both genders, even in the patients with limited symptom relief (e.g. patient #1: MUCP at rest, 86 cmH2O at M3 vs 33 cmH2O at baseline). The maximum bladder capacity was not significantly modified between M0 and M3 [mean (sd) 386.8 (11) vs 413.4 (12) mL, P = 0.71).
Among EMG parameters during urethral profilometry, RMS is the main temporal parameter used to quantify myoelectric activity amplitude. RMS was not significantly modified at rest but increased significantly during maximal voluntary contraction (Fig. 4). After myofibre implantation, RMS during cystometry increased significantly after bladder filling at all volumes studied (Fig. 5). These changes were significant or almost significant in both genders. In contrast, MPF, one of the main frequency parameters used to quantify localised muscle fatigue, showed no change at rest or during contraction.
The present study reports preliminary results of an original procedure for periurethral MPC transfer aimed at restoring sphincter-like muscle activity in patients with ISD. The procedure was well tolerated. Moreover, the improvements in both MUCP and periurethral EMG activity suggest that the satellite cells transferred via the muscle strip underwent myogenic transformation and produced contractions, as previously observed in a pig model .
We obtained concomitant urodynamic and endourethral EMG recordings to assess the function of the muscle tissue derived from the myofibre implants. This combined approach allowed us to position the EMG electrodes optimally at the level of the implant, in a highly standardised manner across recording sessions . Endourethral surface EMG is associated with greater measurement accuracy and repeatability, compared with intrasphincteric needle EMG. Needle EMG records only part of the sphincter activity and, consequently, does not allow a quantitative analysis of sphincter EMG activity, in contrast to endourethral recordings. In addition, the needles cannot be positioned exactly at the same sites during serial recording sessions. Perineal surface electrodes can also be used [37, 38], but record a global EMG signal that lacks muscle selectivity.
The increase in EMG activity reflected by the change in RMS values from baseline to M3 strongly suggests a muscle regeneration process around the urethra with the development of new functional motor units. RMS is the main temporal parameter used to quantify the amplitude of myoelectric activity. Although RMS estimates may show considerable interindividual variability, the present finding that RMS values increased in the overall population constitutes sound evidence that myofibre implantation produced some degree of functional improvement.
An unresolved issue is whether the periurethral implantation of gracilis muscle myofibres, which are fast-twitch and rapidly fatigable, can generate slow-contracting fatigue-resistant myotubes similar to sphincter myofibres . In EMG signal analyses, muscle fatigue is characterised by an MPF decrease, which is chiefly ascribable to slowing of the myofibre conduction velocity. The present finding that MPF remained unchanged may suggest that the regenerated myofibres remained of fast-twitch type at M3. However, only a histological study could provide information on this issue.
Overall, the urodynamic and EMG improvements were similar in both genders, whereas significant clinical improvements occurred only in the women. A possible explanation for this discrepancy is that the increase of urethral pressure needs to be higher in men than in women to restore continence. The absence of significant subjective improvement in the men as compared with women despite a decrease in urine leakage and similar effects on MUCP could also be attributed to more severe 24 h-pad test values at baseline (294 vs 140 g) resulting in insufficient effect of myofibre implantation. Interestingly, the first woman enrolled in the study, in whom the myofibres were implanted around the distal urethra, showed the same pattern of postoperative results as the men, that is, MUCP normalisation at rest with no symptomatic improvement. In contrast, the other four women underwent myofibre implantation around the mid-urethra, near the SUS, and showed both urodynamic and symptomatic improvements. Thus, the implantation site may have a major effect on clinical outcomes, with implantation near the SUS providing the best results. In the men, the surgical approach to the SUS required more extensive dissection of the perineum, which may have resulted in damage to the perineal muscles and nerves or in improper positioning of the muscle strip. We are currently developing an endoscopic approach to facilitate myofibre implantation at the level of the SUS, without causing injury to surrounding structures.
The interpretation of the present results should take into consideration the complex pathophysiology of ISD in both genders. Several hypotheses have been put forward to explain post-RP SUI including: (i) direct damage to the urinary sphincter [40, 41]; (ii) damage to its innervation ; and (iii) a decrease in functional urethral length . MRI studies in patients with post-RP SUI showed evidence of periurethral fibrosis that involved the SUS . In females with ISD, histopathological studies also showed fibrosis associated with loss of sphincter myofibres and features of denervation . The histopathological characteristics of SUS deterioration may vary among patients with ISD. However, according to these previous studies, it can be assumed that most patients in the present study presented some degree of SUS fibrosis and denervation. It is noteworthy that myofibre implantation generates motor units in this environment.
It should be noted that all 10 patients enrolled in the present study were initially candidates for AUS implantation and that this procedure was finally performed in only two patients. In three other patients, the continence improvement provided by myofibre implantation was sufficient to allow successful minimally invasive surgical treatment (male sling or Pro-ACT® device), obviating the need for AUS implantation. The present results also establish that myofibre implantation does not preclude subsequent surgical procedures for UI. Myofibre implantation may prove useful in patients with severe ISD as an adjuvant or neoadjuvant treatment combined with minimally invasive surgery, with the goal of avoiding AUS implantation. The present results suggest that symptom relief may be sustained for at least 1 year after implantation in female responders. These results need confirmation in large controlled studies with longer follow-ups.
As discussed in the introduction, one of the main advantages of myofibre implantation compared with cultured MPC grafting is greater simplicity, with the entire procedure being performed in one stage in the operating room. The development of methods for MPC transfer should also take into consideration the theoretical risk of tumour formation. There is still a concern whether human adult progenitor cells may transform into a malignant cell type in case of prolonged in vitro culture [44, 45]. Chromosomal alterations have been reported in mesenchymal stem cell cultures . To our knowledge the myogenic process after free muscle grafting has never been associated with tumour formation [46, 47]. The transfer of non-manipulated MPCs maintained in their environment as performed with the myofibre implantation procedure should avoid the chromosomal alteration observed with extensive cell culture processes. However, safety studies are required to assess the long-term risk of tumour formation in patients receiving cell therapy for post-RP UI.
We thank Ms Dalila Bitari and Mr Cedric Viallette for their technical support in monitoring the study and Ms R. Cousse for providing physical therapy (pelvic floor exercises) after myofibre implantation.
This study was funded by the Assistance Publique des Hopitaux de Paris (Fond d'amorçage de biothérapie, protocol #06005). The study was also supported by the Association Française contre les Myopathies. The funding sources had no role in the study design, data collection or analysis, decision to publish, or manuscript preparation.
Description of each author's contribution:
All authors contributed substantially to conceiving the study and revising the manuscript. The final version of the manuscript was approved by all authors.
Prof. R Yiou: chief investigator: study design, surgical procedures, and writing of the manuscript.
Dr J.Y. Hogrel: EMG study design, conduct, and analysis; writing of the manuscript.
Dr C.M. Loche: performed and analyzed the urodynamic studies; redaction of the manuscript.
Prof. F.J. Authier: participated in designing of the study.
Dr P. Lecorvoisier: participated in designing of the study.
Ms P. Jouanny: responsible for monitoring the study.
Dr F. Roudot-Thoraval: participated in designing of the study.
Prof J.P. Lefaucheur: EMG study design, statistical analysis, and writing of the manuscript.
Conflict of Interests
The authors have no competing interests to declare.