Three-dimensional ultrasound of pelvic floor: is there a correlation with delivery mode and persisting pelvic floor disorders 18–24 months after first delivery?

Authors

  • A. Falkert,

    Corresponding author
    • Krankenhaus Barmherzige Brüder – Frauenklinik St. Hedwig, Department of Obstetrics and Gynecology, University of Regensburg/Germany, Regensburg, Germany
    Search for more papers by this author
  • A. Willmann,

    1. Krankenhaus Barmherzige Brüder – Frauenklinik St. Hedwig, Department of Obstetrics and Gynecology, University of Regensburg/Germany, Regensburg, Germany
    Search for more papers by this author
  • E. Endress,

    1. Krankenhaus Barmherzige Brüder – Frauenklinik St. Hedwig, Department of Obstetrics and Gynecology, University of Regensburg/Germany, Regensburg, Germany
    Search for more papers by this author
  • P. Meint,

    1. Krankenhaus Barmherzige Brüder – Frauenklinik St. Hedwig, Department of Obstetrics and Gynecology, University of Regensburg/Germany, Regensburg, Germany
    Search for more papers by this author
  • B. Seelbach-Göbel

    1. Krankenhaus Barmherzige Brüder – Frauenklinik St. Hedwig, Department of Obstetrics and Gynecology, University of Regensburg/Germany, Regensburg, Germany
    Search for more papers by this author

Correspondence to: Dr A. Falkert, Krankenhaus Barmherzige Brüder, Frauenklinik St. Hedwig, Steinmetzstrasse 1-3, D-93049 Regensburg, Germany (e-mail: andreas.falkert@barmherzige-regensburg.de)

ABSTRACT

Objective

Three-dimensional (3D) transperineal ultrasound has been shown to be a reliable and reproducible method for visualization of morphological changes in the female levator ani muscle. The aim of this study was to evaluate the relationship between persisting pelvic floor disorders 18–24 months after first delivery, biometric measurements of the pelvic floor and mode of delivery.

Methods

In this prospective observational study, we recruited on their second day after delivery 130 primiparous women. All were Caucasian and their pregnancies had been singleton with cephalic presentation. 3D transperineal ultrasound was performed, with standardized settings, on the second day (results published previously) and 18–24 months after delivery. Volumes were obtained at rest and on Valsalva maneuver and biometric measurements of the levator hiatus were determined in the axial plane. Obstetric and constitutional parameters were obtained from our clinical files and, 18–24 months after the delivery, a standardized questionnaire was used to evaluate persisting pelvic floor disorders. Ultrasound measurements at 18–24 months were compared according to clinical symptoms of pelvic floor disorders and mode of delivery, including a subgroup analysis of vaginal (spontaneous vs operative vaginal) and Cesarean (primary i.e. elective vs secondary i.e. after onset of labor) delivery groups.

Results

Seventy-seven (59%) women had complete follow-up at 18–24 months. Biometric measurements showed a significantly larger hiatal area in the vaginal delivery group than in the Cesarean section group 2 years after delivery (P < 0.01), whereas subgroup analysis within the vaginal and Cesarean delivery groups did not show significant differences. Although there was no statistical correlation between delivery mode and persisting stress urinary incontinence, women with persisting stress urinary incontinence 2 years after delivery had a larger hiatal area than did women without this clinical symptom (P < 0.01). There were no significant differences in hiatal dimensions in women with bladder urgency or dyspareunia.

Conclusions

3D transperineal ultrasound, which is easily accessible, can provide useful information on morphological changes of the female pelvic floor. Women with a spontaneous or operative vaginal first delivery had a significantly larger hiatal area and axial distension than did women whose first delivery was by Cesarean section, even 2 years after delivery. Performing 3D ultrasound after the first delivery may help to identify women at high risk for persisting pelvic floor disorders.

INTRODUCTION

Pregnancy and childbirth have been shown to be major etiological factors for pelvic floor disorders in women[1, 2]. It has been suggested that the strain of the gravid uterus and hormonal changes during pregnancy lead to connective tissue remodeling and disruption of normal pelvic floor function. Additional traumatic damage may result during vaginal delivery (particularly during vacuum or forceps extraction)[3, 4]. The prevalence of urinary incontinence among women in western countries has been estimated to be between 12 and 38% and increases with age[5, 6]. Pelvic organ prolapse is a common problem and it is estimated that women have a 12% lifetime risk (by age 80) of undergoing surgical treatment for urinary incontinence or prolapse[7, 8].

Until recently, magnetic resonance imaging (MRI) was the only imaging method capable of assessing the levator ani muscle. However, cost, access and contraindications (e.g. metallic implants) may limit the scope for adoption of this diagnostic method in clinical practice.

For more than 10 years, translabial or transperineal two-dimensional (2D) ultrasound has been used to assess bladder neck position and movement during Valsalva maneuver and coughing[9-11]. Three-dimensional (3D) transperineal ultrasound has been shown to have many advantages over conventional imaging modalities[12-14]; for example, it can provide the multiplanar view and direct imaging of the entire levator hiatus (axial plane), which were previously the domain of MRI. Ultrasonography is clinically more convenient and accessible than MRI, and can be used safely throughout pregnancy and puerperium. The technique to acquire and interpret ultrasound data can be learned easily by interested clinicians and offline analyses of the stored 3D volume are possible at any time. Using 4D ultrasonography, i.e. real-time imaging, the levator hiatus can be followed during maneuvers and provide qualitative and quantitative information on muscle function[15]. Furthermore, traumatic injuries to the levator ani muscle (e.g. levator avulsion) can be visualized easily by 3D/4D transperineal ultrasound, with higher sensitivity than digital examination[16].

The aim of this study was to evaluate the relationship between persisting pelvic floor disorders (i.e. incontinence, pelvic organ prolapse, dyspareunia) 18–24 months after the first delivery, 3D transperineal ultrasound biometry of the pelvic floor and mode of delivery.

METHODS

This study was performed at a German level I perinatal center with 2000 deliveries per year. Over an 8-month period (January–August 2008), 130 Caucasian primiparae with singleton pregnancy with cephalic presentation were recruited during their hospital stay on the second day after delivery. All women gave informed consent to participate and ethics committee approval was obtained. Women with diagnosed congenital urogenital malformations were excluded. The patients underwent a semi-structured interview about their history and continence status. Different obstetric parameters (maternal body mass index, age, mode of delivery, duration of second stage labor, episiotomy, maternal injuries, birth weight, neonatal head circumference) were obtained from the clinical files.

All ultrasound volumes were acquired on the second day after delivery by one of two experienced sonographers using a GE Voluson 730 Expert ultrasound system (GE Healthcare, Munich, Germany) equipped with a 4–8-MHz curved array 3D/4D ultrasound transducer (RAB 4-8p / obstetric). After voiding her bladder, the patient was placed in a supine position. The probe was covered with a sterile latex-free condom and placed on the perineum in the sagittal plane. The field of view angle was set to a maximum of 70° in the sagittal plane and the volume acquisition angle to 85° in the axial plane; preset ‘obstetrics 2./3. trimenon’ was used. Two 3D volumes (one with the patient at rest and one during Valsalva maneuver) were recorded, with each automatic image acquisition taking about 4 s. These results, obtained 2 days after delivery, have been published previously[17].

For the purposes of the current study, the same women were invited for a follow-up examination 18–24 months after delivery. In addition to a second transperineal 3D ultrasound examination (two 3D acquisitions, one at rest and one during Valsalva), each woman answered a standardized questionnaire on the typical symptoms of pelvic floor disorders (urinary/fecal incontinence, dyspareunia). We performed a regular vaginal examination with the woman in a supine position, at rest and during Valsalva, to look for pelvic organ prolapse (quantified according to POP-Q classification[18]).

Analysis of all stored volumes was conducted offline by an experienced investigator. Measurements were performed as described by Dietz et al.[13], in the axial plane at the level of ‘minimal hiatal dimensions’. The following sonographic parameters were assessed: (1) maximum diameters of the levator hiatus (anteroposterior and transverse), with the woman at rest and during Valsalva (Figure 1a); (2) area of the levator hiatus at rest and during Valsalva (Figure 1b); (3) pubovisceral muscle thickness (to left and right of rectum) at the level of maximum muscle thickness (Figure 1a). Levator avulsion was diagnosed on tomographic ultrasound imaging (Figure 2) according to Dietz et al.[19]. Ultrasound examination and offline analysis of the stored volumes were performed by the same physicians 18–24 months after delivery as had done them on day 2.

Figure 1.

Biometric measurements on two-dimensional ultrasound (axial view) of the levator ani muscle. (a) Hiatal diameter. Calipers show transverse diameter (Caliper 1), anteroposterior diameter (Caliper 2) and levator muscle thickness (Calipers 3 and 4). (b) Hiatal area.

Figure 2.

Tomographic ultrasound imaging demonstrating unilateral levator avulsion.

Statistical analysis was performed after normality testing (Kolmogorov–Smirnov) using PASW Statistics 18.0 (IBM, Ehningen, Germany). Pearson's correlation coefficient (r) was used to compare normally distributed continuous variables. Mean values were compared by ANOVA (more than two groups) or t-test (two groups). Cross tables were evaluated by chi[2] (two groups each) or Phi/Cramer-V (more than two groups). P < 0.05 was considered statistically significant.

RESULTS

Initially, 130 primiparae, all of them Caucasian with singleton pregnancies with cephalic presentation, underwent 3D ultrasound assessment of the pelvic floor on the second day after delivery. Satisfactory volume datasets and delivery data were obtained in all cases[17].

The mean age was 29.8 (range, 16–42) years and the mean body mass index was 23.2 (range, 17–40) kg/m2. The mean gestational age at delivery was 39.3 (range, 34–42) weeks. The mean birth weight was 3343 (range, 1795–4930) g and the mean head circumference was 34.7 (range, 30–39) cm. Thirty-nine of the 130 (30.0%) women were delivered by Cesarean section (11 of them elective Cesarean), 77 (59.2%) had a spontaneous vaginal delivery and 14 (10.8%) had an instrumental vaginal delivery[7]. For the purpose of analysis, no differentiation was made between the different types of operative vaginal delivery (forceps vs vacuum extraction) due to the small numbers that would have resulted in each category.

The median duration of second-stage labor in the group of women who delivered vaginally was 89.1 (range, 7–278) min. An episiotomy was performed in 46.1% of all vaginal deliveries.

At the 18–24-month follow-up, 77 (59.2%) women could be re-evaluated. None of these women was pregnant or had delivered a second child since initial examination. Of the 77, 50 (64.9%) had a spontaneous vaginal delivery, eight (10.4%) had an instrumental vaginal delivery and 19 (25%) were delivered by Cesarean section (five of them elective Cesarean). There were no statistical differences concerning the relative group sizes according to delivery mode at follow-up compared with on the second day after delivery. During the first examination (second day after delivery), none of the participants had symptoms of prolapse or fecal incontinence but 8/130 (6.2%) women reported de novo stress urinary incontinence. At the time of the follow-up examination (18–24 months after delivery), 28/77 (36.4%) women reported persisting stress urinary incontinence (seven of them daily), 33 (42.9%) women had urgency symptoms (16 of them daily) and 32 (41.6%) women reported dyspareunia problems (three of them regularly).

As was noted 2 days after the first delivery[17], even 2 years later women with a vaginal delivery had a significantly larger hiatal area at rest and on Valsalva, compared with women delivered by Cesarean section (P < 0.01, Table 1). Similarly, as was reported 2 days after the first delivery[17], there were no statistical differences in the subgroup analysis (spontaneous vs operative vaginal and elective vs secondary Cesarean section, Figure 3a). Compared with the results on the second day after the first delivery[17], the relationship between delivery mode and hiatal dimensions also persisted 2 years later, although the significance level was lower (P < 0.0001 vs P < 0.01, Figure 3a). When comparing the levator ani dimensions to the clinical symptoms 2 years after the first delivery, women with persisting stress urinary incontinence had a significantly larger hiatal area during Valsalva than did women without this problem (P < 0.01, Table 2 and Figure 3b). However, there were no statistical correlations between delivery mode and persisting urinary stress incontinence or urgency 2 years after the first delivery. There was also no statistical correlation between levator hiatus dimensions 18–24 months after the first delivery and dyspareunia problems. Furthermore, there was a highly significant correlation between mode of delivery and levator avulsion (P < 0.0001), with avulsions only in women delivered vaginally.

Table 1. Biometric parameters of the levator hiatus in women with singleton pregnancy 18–24 months after their first delivery, according to mode of delivery
ParameterMode of deliveryPa
Spontaneous vaginal (n = 50)Operative vaginal (n = 8)Primary Cesarean (n = 5)Secondary Cesarean (n = 14)
  1. Data are given as mean ± SD.

  2. a

    ANOVA. NS, not significant.

Levator thickness (mm)10.0 ± 1.410.0 ± 1.110.0 ± 0.79.6 ± 1.1NS
At rest     
Hiatal diameter (mm)     
Transverse39.7 ± 4.640.8 ± 4.637.2 ± 4.235.6 ± 4.2< 0.05
Anteroposterior56.2 ± 6.156.5 ± 7.054.0 ± 5.450.9 ± 7.1NS
Hiatal area (cm2)16.5 ± 2.716.9 ± 2.914.7 ± 1.613.4 ± 2.9< 0.01
On Valsalva     
Hiatal diameter (mm)     
Transverse44.1 ± 6.246.5 ± 9.639.8 ± 4.838.1 ± 4.4< 0.05
Anteroposterior57.5 ± 7.758.7 ± 10.351.8 ± 3.150.5 ± 6.2< 0.05
Hiatal area (cm2)19.3 ± 5.322.1 ± 8.715.1 ± 0.914.5 ± 2.4< 0.01
Figure 3.

Hiatal area (mean ± 2 SD) on Valsalva maneuver, in 130 women with singleton pregnancy 2 days following their first delivery (image) and in 77 of these women 18–24 months after the delivery (image), according to mode of delivery (a) and persistence of stress urinary incontinence (b). (a) Hiatal area was significantly different according to mode of delivery both 2 days after delivery (P < 0.0001, ANOVA) and at 18–24-month follow-up (P < 0.01, ANOVA). (b) Hiatal area was significantly greater in women with persistence of stress urinary incontinence at 18–24-month follow-up (P < 0.01) but not 2 days post-delivery (ANOVA). *t-test. NS, not significant.

Table 2. Biometric parameters of the levator hiatus in women with singleton pregnancy 18–24 months after their first delivery, according to persisting stress urinary incontinence
ParameterStress urinary incontinencePa
Present (n = 28)Absent (n = 49)
  1. Data are given as mean ± SD.

  2. a

    t-test. NS, not significant.

Levator thickness (mm)10.3 ± 1.19.8 ± 1.3NS
At rest   
Hiatal diameter (mm)   
Transverse39.9 ± 5.138.1 ± 4.2NS
Anteroposterior56.9 ± 6.154.4 ± 6.8NS
Hiatal area (cm2)16.5 ± 3.015.4 ± 2.8NS
On Valsalva   
Hiatal diameter (mm)   
Transverse45.5 ± 7.841.1 ± 4.9< 0.01
Anteroposterior58.6 ± 7.854.3 ± 7.6< 0.05
Hiatal area (cm2)20.1 ± 6.816.8 ± 3.8< 0.01

We did not find an association between levator ani avulsion (n = 9, all of them unilateral avulsions) and urinary incontinence/urgency/dyspareunia, although three (one third) of the patients with levator avulsion were shown to have a prolapse with cystocele ≥ Stage II according to POP-Q-classification[18]. No other constitutional (e.g. maternal age, body mass index) or obstetric (e.g. duration of second-stage labor, birth weight, gestational age at delivery) factor had any statistical correlation to the measured levator dimensions 18–24 months after delivery.

DISCUSSION

This study demonstrates clear differences in postpartum levator hiatus distension according to the different modes of delivery in primiparae. These morphological changes seem to persist and can be reproducibly demonstrated by 3D transperineal ultrasound even 2 years after the first delivery. Women with vaginal or vaginal operative birth had a significantly larger hiatal area and axial distension than did women who delivered by Cesarean section, especially during Valsalva maneuver.

Concerning the hiatal dimensions, there were no significant differences found in subgroup analyses (spontaneous vs operative vaginal and primary vs secondary Cesarean). Not surprisingly, compared with the results on the second day after first delivery[17], significance levels were lower in all measurements after 2 years.

All cases with visible levator avulsion 18–24 months after delivery were in the group of women delivered vaginally. According to the literature, levator avulsion is found in 10–36% of women after first delivery[20-22]. In our study, there was a unilateral levator avulsion at the time of the follow-up examination in 18% of the patients with vaginal delivery, one third of whom already showed clinical signs of pelvic organ prolapse. However, we did not find a relationship between dimensions of the levator hiatus and/or levator injury and the length of second-stage labor or other obstetric or constitutional maternal parameters.

About one third of the patients at follow-up had persisting stress urinary incontinence 2 years after their first delivery, and fewer than 10% of these reported daily incontinence problems. All of these women had a significantly larger hiatal area and pelvic floor hypermobility than did the women without symptoms. Urge urinary incontinence and dyspareunia can occur during pregnancy and after delivery as a result of morphological and hormonal changes of the pelvic floor. Whereas stress urinary incontinence is more common after vaginal delivery, de-novo urgency is reported more frequently after Cesarean section[23].

Dyspareunia may affect up to 75% of women within the first months after their first delivery, with persisting pain during sexual intercourse in up to 14% of cases, with a strong correlation to delivery mode[24]. In our study, urgency/dyspareunia problems 18–24 months after the first delivery showed no significant correlation to delivery mode and/or hiatal dimensions.

Possible explanations for traumatic pathogenesis in pelvic tissue include alterations in the matrix of connective tissue and damage to the pelvic fascia, which may lead to a loss of contractility of the levator ani complex[25-27]. This may arise as a result of damage to the muscle from distension, obstetric trauma or neuropraxia of the pudendal nerve[28, 29]. Also, pre-existing individual differences in the biomechanical properties of the pelvic floor and the connective tissue may result in differences in tissue rebuilding during pregnancy and delivery and have an impact on development of pelvic floor disorders. It should be noted that, while nerve injury and tissue damage are statistically more likely to occur during vaginal delivery, they do still occur in women delivered by Cesarean section.

Possible limitations of this study include the small number of patients in certain sub-groups (e.g. operative-vaginal delivery group), the 41% loss to follow-up and the fact that the ultrasound examiners were not blinded to mode of delivery (this was impossible as they could obviously see during examination the scars of Cesarean section, perineal tearing and episiotomy). Also, the length and intensity of Valsalva maneuver during 3D volume acquisition may have an influence on biometric hiatal measurements. Recent literature suggests that a longer maneuver is required for a more precise assessment of pelvic organ prolapse and levator ballooning[30].

In conclusion, this study demonstrates that persisting morphological changes of the female pelvic floor can be imaged by 3D ultrasound techniques 2 years after first delivery. Larger hiatal dimensions were significantly correlated to vaginal delivery and persisting stress urinary incontinence. In our opinion, this easily accessible diagnostic tool may be helpful to identify young mothers with a high risk for pelvic floor dysfunction. Performing hiatal measurement after the first delivery may also have an influence on specific counseling or coaching of women with pelvic floor disorders during their second or third pregnancy and throughout their life.

Ancillary