CC BY-NC-ND
0Urological Science xxx (2015) 1—5
Contents lists available at ScienceDirect
Urological Science
journal homepage: www.urol-sci.com
Original article
Macroplastique outcome in women with stress urinary incontinence secondary to intrinsic sphincteric deficiency
Emily C. Rosenfeld a, Alana Christie b, Chasta D. Bacsu a, Philippe E. Zimmern a' *
a Department of Urology, UT Southwestern Medical Center, Dallas, TX USA b Department of Biostatistics, UT Southwestern Medical Center, Dallas, TX, USA
ARTICLE INFO ABSTRACT
Objectives: To evaluate short-term outcomes of Macroplastique (MPQ) in women with stress urinary incontinence (SUI) using patient reported outcome and three-dimensional vaginal ultrasound (3DUS). Materials and methods: After obtaining institutional review board approval, a chart review of non-neurogenic women that received MPQ for intrinsic sphincter deficiency (ISD) was extracted from a prospective database. Patients were divided into three groups: naïve (Group I), prior incontinence surgery (Group II), and both prior incontinence surgery and bulking agent (Group III). Women with urethral hypermobility were excluded. Baseline evaluation included a history, physical examination to confirm SUI, and questionnaires [Urinary Distress Inventory-6 (UDI-6), 1 quality of life (QOL) global score based on visual analog scale], and in select patients urodynamic studies and/or standing voiding cystogram. Patient follow up included repeat questionnaires scores and 3DUS to objectively assess MPQ volume. Success was defined as sufficient improvement after one injection so that a subsequent reinjection/ different SUI operation was not requested at the last follow-up visit. It was hypothesized that Group I would fare best.
Results: Fifty-nine women met the inclusion criteria. Success rate was 83% for Group I, 70% for Group II, and 69% for Group III (p = 0.54) at 9 months mean follow up. Fifteen patients underwent a second 3DUS during follow up with a stable volume, compared to the first study (4.5 ± 1.5 vs. 4.4 ± 1.5, p = 0.70), which confirmed stable volumes over time. Among the failures (N = 15), nine patients proceeded with reinjection; four patients had fascial slings, and two patients had artificial sphincters. Conclusion: As confirmed by 3DUS, Macroplastique appears efficacious as a primary treatment and as a salvage treatment for SUI due to ISD in the short-term.
Copyright © 2015, Taiwan Urological Association. Published by Elsevier Taiwan LLC. All rights reserved.
Article history: Received 9 December 2014 Received in revised form 28 January 2015 Accepted 3 February 2015 Available online xxx
Keywords: bulking agent female global score Macroplastique quality of life
stress urinary incontinence Urinary Distress Inventory-6
1. Introduction
Urethral bulking agents (UBA) are an attractive option for treating stress urinary incontinence (SUI) due to intrinsic sphincter deficiency (ISD), especially for patients desiring a less invasive treatment. Macroplastique (polydimethylsiloxane) has been used for the treatment of SUI with demonstrable efficacy as early as 1991 in Europe.1 In Macroplastique, silicone-based particles are incorporated within a polyvinylpyrrolidone (PVP) suspension and are > 100 mM in diameter, which effectively limits migration.
* Corresponding author. Department of Urology, UT Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-9110, USA.
E-mail address: Philippe.zimmern@utsouthwestern.edu (P.E. Zimmern).
Similar to our experience using collagen as a urethral bulking agent, we have incorporated three-dimensional transvaginal ultrasound (3DUS) imaging and patient self-report questionnaires into our follow-up management algorithm2,3 for Macroplastique injections. Three-dimensional vaginal ultrasound provides objective information about the volume of the bulking agent obtained and its configuration around the urethra. The use of 3DUS is not widespread or part of standard guidelines; however, it is a tool with low cost and no morbidity or invasiveness, which helps in deciding whether patients are suitable for observation, reinjection, or another modality of treatment for SUI. Three-dimensional vaginal ultrasound has been integral in our assessment of women receiving UBA such as Collagen since 3DUS became available at our center in 1999. Among the early studies reporting on Macroplastique, one series of 60 women included transurethral ultrasound findings in nine patients.4 The ultrasound hyperechoic nature of
http://dx.doi.org/10.1016/j.urols.2015.02.001
1879-5226/Copyright © 2015, Taiwan Urological Association. Published by Elsevier Taiwan LLC. All rights reserved.
E.C. Rosenfeld et al. / Urological Science xxx (2015) 1—5
Macroplastique was recognized at that time, and had a generally good outcome for at least 1 year after the procedure when it completely encircled the urethra.5
In line with these early reports, the aim of our study was to study the short-term outcomes of Macroplastique (MPQ) in women with SUI secondary to ISD by using validated questionnaires (i.e., patient reported outcome) and 3DUS imaging (i.e., objective outcome). Furthermore, the outcomes of three subgroups were analyzed to broaden the relevance of this new study. Group I comprised naïve patients. The remaining two groups were more complicated: Group II, comprised patients with a history of previous anti-incontinence surgery and Group III comprised patients with previous anti-incontinence surgery and failed prior collagen injection. We hypothesized that Group I would fare best.
2. Materials and methods
After receiving approval from the UT Southwestern Institutional Review Board (IRB), we performed a review of prospectively collected data on women with SUI who received MPQ. Inclusion criteria were women with SUI secondary to ISD, a well-supported urethra, and at least one postoperative follow-up visit with questionnaires and 3DUS imaging. Exclusion criteria included neuro-genic bladder, prior pelvic radiation therapy, and women with urethral hypermobility.
Success was defined as sufficient improvement after one injection so that a subsequent reinjection/different SUI operation was not requested by the patient at their last follow-up visit. "Cur-e"—the ability to remain dry and not wear any pads—was defined as a score of 0 on Question 3 of the Urinary Distress Inventory-6 (UDI-6) in relation to SUI.6 Patients were divided into three subgroups: naïve (Group I), prior anti-incontinence surgery (Group II), and prior anti-incontinence surgery + collagen injection (Group III).
Baseline evaluation included medical history, physical examination, UDI-6, and one global quality of life visual analog scale (VAS) that asked "If you were to spend the rest of your life with your urinary condition just the way it is now, how would you feel about that?" (with a "0" answer being "pleased" and "10" being "terrible.")7 In addition, some patients received further workup such as office cystoscopy, voiding cystourethrogram and/or multichannel urodynamic studies if the patients were considered for MPQ reinjection or surgical sling placement. Women with SUI secondary to ISD and a well-supported urethra were identified primarily by examination findings (Aa point at -3) with demonstrable SUI during stress maneuvers. Other tools used to confirm a well-supported urethra in our practice included an angle of < 30° on a true lateral voiding cystourethrogram during straining views. Aurodynamic study, when obtained, provided a Valsalva leak point pressure. The value of the Valsalva leak point pressure to confirm ISD is typically < 60 cmH2O; but higher values can occur, which indicates a lesser degree of ISD severity. Baseline urodynamics with a noninvasive flow test and a postvoid residual by bladder scan was obtained at baseline and in follow up. Of note, patients who had received previous collagen injections had a 3DUS prior to receiving MPQ, which provided collagen volume and configuration at baseline.
The MPQ injections were performed in an outpatient basis by the same urologist (PZ) with the patient under light anesthesia (i.e., intravenous sedation or laryngeal mask airway intubation). In general, in accordance with our prior protocol on Collagen injection, two injections (2.5 cc x 2 or approximately 5 mL) were delivered at the 3 o'clock and 9 o'clock positions to obtain luminal coaptation.
The first follow-up visit was 6—8 weeks after the MPQ injection procedure and included questionnaires and 3DUS imaging.2 Subsequent follow up continued on an annual basis and included questionnaires with or without 3DUS.
ZÈ S'
• MACRO
TOTAL = 6.0 CC
Fig. 1. Macroplastique on 3DUS has a very hyperechogenic appearance along the urethra (U) on (A) longitudinal and (B) transversal images in a naïve patient who was dry after one injection. 3DUS = three-dimensional vaginal ultrasound.
E.C. Rosenfeld et al. / Urological Science xxx (2015) 1—5
The 3DUS was performed by the same imaging team consisting of a senior ultrasound technician and a radiologist with longstanding experience with collagen measurements by ultrasound.7
A small finger-size vaginal probe was introduced in the patient in supine position. The probe was positioned underneath the urethral meatus and moved alongside the vaginal wall in longitudinal and transversal directions. The MPQ is very echogenic and easy to identify (Fig. 1) on each side of the urethral lumen using the Philips iU22 ultrasound system (Philips, Andover, MA, USA) and a broad band 3D volume probe (V9-3) with frequency range of 9 mHz—3 mHz. This mechanical 3D probe performs automatic acquisition of volume data sets. Measurements are performed transversally and axially to provide a volume calculation. The data collected from the 3DUS included the total UBA volume (including MPQ and Collagen, if previously injected) and the configuration (i.e., circumferential versus asymmetric).
Descriptive statistics were provided using the mean and standard deviation for continuous variables and frequencies and percentages for categorical variables. Fisher's exact test (for categorical variables), the Student t test (for continuous variables, 2 subgroups), and analysis of variance (ANOVA; for continuous variables, 3 subgroups) were used to determine differences between subgroups. Post hoc t tests were completed for any significant ANOVA comparisons. The paired t test was used to find if there were any differences between the pre- and postoperative VAS-QOL scores and the 3DUS volume changes between the first and second visit after injection. A mixed model was used to ascertain whether the 3DUS volume after final injection decreased significantly over time after being controlled for by the subgroups and number of injections. All statistics were completed using SAS 9.3 (SAS Institute Inc., Cary, NC, USA).
3. Results
From a prospective database of 72 patients from July 2011 to December 2013 who received MPQ, 59 women met inclusion
criteria (Table 1). No serious adverse effects were reported. Early urinary retention lasting < 48 hours occurred in five women. At a mean follow up of 8.5 ± 7.2 months, the success rate was 83% for Group I, 70% for Group II, and 69% for Group III. The cure rate was 17% for Group I, 9% for Group II, and 38% for Group III. The mean age (p = 0.21) and BMI (p = 0.28) were comparable across the three groups, as was the mean duration of follow up (p = 0.46; Table 1). Group II had a significantly higher parity than Group I (p = 0.046). For naive patients, the mean volume of MPQ injected was 4.43 mL (range, 2.5—5 mL) with the mean volume measured on first 3DUS after MPQ at 3.85 (range, 1.77—5.91). This gave a volume retention rate of 87%. At this first 3DUS study, 72% of studies indicated a circumferential configuration for MPQ.
Among the 15 failures, there was no significant difference between the groups—four failures were from Group I; seven failures from Group II, and four failures from Group III (p = 0.59) (Table 2). The pretreatment UDI-6 scores for Question 3 in the success group were generally more favorable than those of the failure group. Among the failures, nine patients proceeded with reinjection; four patients with slings; and two patients with artificial urinary sphincters. The mean time between the first and the repeat injection was 10 months (range, 3—19), whereas the mean time between the first injection and a secondary sling was 8.5 months (range, 2—16).
Fifteen patients underwent a second 3DUS during the follow up (mean interval of 9 months after the first 3DUS) with a stable volume, compared to first study (4.5 ± 1.5 vs. 4.4 ± 1.5, respectively; p = 0.70). Four patients with a prior collagen injection also underwent a second 3DUS with similar findings (5.3 ± 1.3 vs. 4.9 ± 1.4; p = 0.35). In a mixed model that controlled for the number of injections and prior urological history, time since the last injection was not a significant factor in ultrasound volume after the last injection (mean follow up, 6.5 months; p = 0.46).
Parameters related to voiding such as Question 5 of the UDI-6, maximum flow, or postvoid residuals were not affected by MPQ injection across the three groups.
Of patients that had VAS-QOL recorded during their pre- and postoperative visits, there was significant improvement whether
Table 1
Baseline group comparison, based on prior stress urinary incontinence surgical history.
Naïve (n = 23) Prior surgery (n = 23) Prior surgery + collagen (n = 13) p
Age (y) 64.1 ± 8.1 65.0 ± 11.3 70.1 ± 10.2 0.21
Parity 1.9 ± 1.0 2.8 ± 1.3 1.9 ± 1.3 0.046
BMI 26.0 ± 5.3 28.6 ± 6.3 27.2 ± 3.9 0.28
Duration of follow up (mo) 7.2 ± 5.9 8.7 ± 7.4 10.3 ± 8.8 0.46
Preoperative VLPP
Mean ± SD 53.6 ± 17.2 75.1 ± 40.4 65.4 ± 28.5 0.22
<60 4/13 (31) 8/13 (62) 4/7 (57) 0.29
>60 9/13 (69) 5/13 (38) 3/7 (43)
(n = 19) (n = 18) (n = 9)
Qmax 15.5 ± 7.7 13.8 ± 8.5 20.1 ± 15.8 0.30
Postvoid residual
No residual 16/19 (84) 15/18 (83) 6/9 (67) 0.56
Residual 3/19(16) 3/18 (17) 3/9 (33)
UD16 Total score 11.4 ± 3.6 10.4 ± 3.9 9.5 ± 4.2 0.59
UD16 Q2 — Urge Incontinence
0 or 1 2/10 (20) 5/18 (28) 3/8 (38) 0.79
2 or 3 8/10 (80) 13/18 (72) 5/8 (63)
UD16 Q3 — Stress incontinence
0 or 1 0/10 (0) 1/19(5) 2/8 (25) 0.17
2 or 3 10/10(100) 18/19 (95) 6/8 (75)
UD16 Q5 — Emptying
0 or 1 5/10 (50) 15/19 (79) 6/8 (75) 0.29
2 or 3 5/10 (50) 4/19(21) 2/8 (25)
Data are presented as n/N (%) or mean ± SD. UD16 score of 0,1 = none or rare; Score of 2, 3 BMI = body mass index; Q2 = Question 2; Q5 = = moderate or severe. = Question 5; Qmax = XX; UD16 = Urinary Distress 1nventory-6 i; VLPP = Valsalva leak point pressure.
E.C. Rosenfeld et al. / Urological Science xxx (2015) 1—5
Table 2
Differences between the patients that only received one injection and had no other surgery versus patients that had a repeat injection or reoperation.
Success group (n = 44) Failure group (n = 15) P
Prior History
Naïve 19/23 (83) 4/23 (17) 0.59
Prior surgery 16/23 (70) 7/23 (30)
Prior surgery + prior collagen 9/13 (69) 4/13 (31)
Preoperative VLPP (cm H2O)
Mean ± SD 62.6 ± 31.4 71.9 ± 31.3 0.49
<60 11/16(69) 5/16(31) 0.22
>60 15/17 (88) 2/17 (12)
(n = 36) (n = 10)
Qmax (mL/s) 15.8 ± 10.5 15.5 ± 8.8 0.94
Postvoid residual (mL)
No residual 29/37 (78) 8/37 (22) 1.00
Residual 7/9 (78) 2/9 (22)
UDI6 Total 10.2 ± 4.0 11.2 ± 3.4 0.44
UDI6 Q2 — Urge Incontinence
0 or 1 8/10 (80) 2/10 (20) 0.69
2 or 3 18/26 (69) 8/26 (31)
UDI6 Q3 — Stress Incontinence
0 or 1 3/3 (100) 0/3 (0) 0.54
2 or 3 23/34 (68) 11/34 (32)
UDI6 Q5 — Emptying
0 or 1 19/26 (73) 7/26 (27) 0.70
2 or 3 7/11 (64) 4/11 (36)
Data are presented as n/N (%) or mean ± SD.
UDI6 score of 0 or 1 = "none" or "rare"; UDI6 score of 2 or 3 = "moderate" or "severe."
Q2 = Question 2; Q5 = Question 5; Qmax = XX; UDI6 = Urinary Distress Inventory-6; VLPP = Valsalva leak point pressure.
the procedure was classified as a failure or success. The success group reported a 3.3 ± 3.7 (p = 0.0003) improvement in the absolute value score. The failure group reported a 2.7 ± 3.4 (p = .03) improvement. The difference between the VAS-QOL scores for the success and failure groups were significantly different at the pre-operative visit (7.2 ± 2.6 vs. 9.4 ± 1.3, respectively; p = 0.002) and postoperative visit (3.8 ± 2.9 vs. 6.7 ± 3.2, respectively; p = 0.02), although the improvement in the score was not significantly different (p = 0.65).
There was no loss to follow up. Twenty-three patients had a follow up > 1 year. Of these, 13 of 23 (56%) patients only had one injection and were observed because of cure or significant improvement that did not require additional therapy. By contrast, seven of 23 (30%) patients had a repeat injection, two of 23 (9%) patients had a sling placed, and one of 23 (4%) patients had an AUS.
4. Discussion
We report our early experience with MPQ, which indicates an overall success rate of nearly 70% and a slightly higher success rate in the naive subgroup. In a subset of women who underwent repeat 3DUS testing over time, no significant loss of volume was noted. Even though absolute cure or strict dryness after MPQ is only achieved in a minority of patients, MPQ provided durable SUI symptom improvement in most of our patient population. No changes in Question 5 of UDI-6 related to voiding, maximum flow, or postvoid residual were noted, which implies no effect on voiding function.
Several series on bulking agents have been reported in the literature such as a recent large series of 514 women treated with a variety of bulking agents within the past 13 years.8 However, limited data is available on Macroplastique and the use of an objective outcome measure such as the one employed in this study with 3DUS. A study by Hegde et al,9 which used 3DUS to evaluate MPQ outcomes at 15 minutes after the procedure, demonstrated that a better clinical outcome was associated with MPQ location in the proximal urethra and with a circumferential periurethral distribution. In the Hedge study (N = 100) divided treatment-naive patients into groups based on their clinical outcomes: patients with a good outcome (Group A; N = 72) and patients who either worsened or did not improve after injection (Group B; N = 28). The mean follow up was 19.32 weeks in Group A and 17.77 weeks in Group B.
In our study, we did not perform ultrasound to determine immediate results, but to evaluate the durability of the injection in volume and configuration around the urethra in different clinical subgroups of women. All studies were performed approximately 6—8 weeks after the injection procedure as part of the first follow-up visit. We have previously reported on the use of 3DUS, and focused on evaluating the mid- and long-term results of Collagen injections for SUI.2,3 These studies together demonstrate that circumferential distribution of an injectable agent is a more important predictor of favorable outcome than the injected volume and that 3DUS is a simple tool to serially and objectively evaluate UBA distribution.
Variable improvement and cure rates for MPQ have been reported in contemporary literature (Table 3)12—14. A recent meta-analysis by Ghoniem and Miller10 described cure rates of 45% (short-term outcome) and 36% (long-term outcome), whereas the improvement rates were 75% (short-term outcome) and 65% (long—term outcome). Variability between reports is associated with indications, success definition, and duration of follow up. Our
Table 3
Literature review on Macroplastique.
Study Design Follow-up period (mo) N Improvement rate (%) Cure/dry rate (%) Success determinant Adverse effects
Ghoniem et al15 Randomized, single-blind 12 mo 247 61.5 36.9 Stamey grade 24% UTI, 9% dysuria, 9% urgency, 8% frequency, 7% retention
Ghoniem et al13 Case series 24 mo 67 84 67 Stamey grade None reported
Maher et al14 Randomized controlled trial 12 mo 23 60 n/a Patient report 5% voiding dysfunction
Plotti et al16 Case series 12 mo 24 42 42 Urodynamic assessment None reported
Tamanini et al12 Case series 60 mo 15 33.3 40 Urodynamic assessment 10.3% transient urinary retention; 4.8% loss of PDMS through injection site
Zullo et al17 Prospective cohort study 12 mo 27 33 44 Urodynamic assessment None reported
PDMS = XX.
E.C. Rosenfeld et al. / Urological Science xxx (2015) 1—5
present short-term findings of improvement ranging 69—83% are consistent with these literature findings.
The MPQ deserves special attention as a salvage treatment post—anti-incontinence procedure because it has been under-reported in the literature. Lee-Lewis and Anderson11 reported a 34.8% cure rate and a 77% improvement rate in women treated for recurrent SU1 after a prior mid-urethral sling (MUS). 1n a study by Gumus et al,1 which compared women with and without a history of an anti-incontinence operation, the women who had undergone a prior procedure for SU1 were more satisfied with their condition post-MPQin a long-term follow up. Different expectations between the two groups may have affected their response rates. Our own study suggests the same adaptation of women to their incontinence condition after a prior failed anti-incontinence procedure with improvement but not complete cure being well accepted.
Long-term favorable outcome has been reported by Tamanini et al12 at 12 months, 24 months, and 60 months, which indicated the relative durability and stability of the MPQ. Ghoniem et al13 likewise reported that 84% of patients maintained significant Stamey grade improvement at their 12—24 month assessments.11 However, Maher et al14 reported a continence rate of only 22% post-MPQ at the 60-month follow up; however, these lower values for long-term findings were obtained from a nonvalidated questionnaire.
Our study has unique strengths. 1t was a prospective study on women with different SU1 backgrounds ranging from naïve to prior sling procedure with or without an additional bulking agent. The data was collected and analyzed by a neutral investigator not involved in patient care. The 3DUS procedure was performed by the same team of technicians and radiologists with longstanding expertise of more than a decade in 3DUS interpretation after collagen injection.
The limitations of this study include its short-term follow-up period and its relatively limited study size. However, it is one of the largest cohorts reported to date. 1n addition, because 3DUS was not performed immediately at the time of the injection under anesthesia, it was not possible to determine if failures were because of substance migration or extrusion, or operative/technical failure.
5. Conclusion
MPQ is apparently an effective treatment option in the short-term in women with naïve or complicated 1SD-related SU1. A complete cure is only achieved in a minority of patients, although a significant improvement was observed by many women. Three-dimensional vaginal ultrasound can be used to assess and
monitor MPQ injection results over time, and has demonstrated stable volumes in a subset group.
Conflicts of interest
References
1. Gumus II , Kaygusuz I, Derbent A, Simavli S, Kafali H. Effect of the Macroplastique Implantation System for stress urinary incontinence in women with or without a history of an anti-incontinence operation. Int Urogynecol J 2011;22:743—9.
2. Poon CI, Zimmern PE. Role of three-dimensional ultrasound in assessment of women undergoing urethral bulking agent therapy. Curr Opin Obstet Gynecol 2004;16:411—7.
3. Defreitas GA, Wilson TS, Zimmern PE, Forte TB. Three-dimensional ultrasonography: an objective outcome tool to assess collagen distribution in women with stress urinary incontinence. Urology 2003;62:232—6.
4. Harriss DR, Iacovou JW, Lemberger RJ. Peri-urethral silicone microimplants (Macroplastique) for the treatment of genuine stress incontinence. Br J Urol 1996;78:722—8.
5. Radley SC, Chapple CR, Mitsogiannis IC, Glass KS. Transurethral implantation of macroplastique for the treatment of female stress urinary incontinence secondary to urethral sphincter deficiency. Eur Urol 2001;39:383—9.
6. Uebersax JS, Wyman JF, Shumaker SA, McClish DK, Fantl JA. Short forms to assess life quality and symptom distress for urinary incontinence in women: the Incontinence Impact Questionnaire and the Urogenital Distress Inventory. Continence Program for Women Research Group. Neurourol Urodyn 1995;14: 131—9.
7. Lemack GE, Zimmern PE. Identifying patients who require urodynamic testing before surgery for stress incontinence based on questionnaire information and surgical history. Urology 2000;55:506—11.
8. Mohr S, Siegenthaler M, Mueller MD, Kuhn A. Bulking agents: an analysis of 500 cases and review of the literature. Int Urogynecol J 2013;24:241—7.
9. Hegde A, Smith AL, Aguilar VC, Davila GW. Three-dimensional endovaginal ultrasound examination following injection of Macroplastique for stress urinary incontinence: outcomes based on location and periurethral distribution of the bulking agent. Int Urogynecol J 2013;24:1151—9.
10. Ghoniem GM, Miller CJ. A systematic review and meta-analysis of Macroplastique for treating female stress urinary incontinence. Int Urogynecol J 2013;24:27—36.
11. Lee-Lewis H, Anderson DM. Absence of inflammation and pneumonia during infection with nonpigmented Yersinia pestis reveals a new role for the pgm locus in pathogenesis. Infect Immun 2010;78:220—30.
12. Tamanini JT, D'Ancona CA, Netto NR. Macroplastique implantation system for female stress urinary incontinence: long-term follow-up. J Endourol 2006;20: 1082—6.
13. Ghoniem G, Corcos J, Comiter C, Westney OL, Herschorn S. Durability of ure-thral bulking agent injection for female stress urinary incontinence: 2-year multicenter study results. J Urol 2010;183:1444—9.
14. Maher CF, O'Reilly BA, Dwyer PL, Carey MP, Cornish A, Schluter P. Pubovaginal sling versus transurethral Macroplastique for stress urinary incontinence and intrinsic sphincter deficiency: a prospective randomized controlled trial. BJOG 2005;112:797—801.
