Australian and New Zealand Journal of Obstetrics and Gynaecology 2015; 55: 70–75
DOI: 10.1111/ajo.12261
Original Article
Can we identify changes in fascial paravaginal supports after childbirth? Jordi CASSADO-GARRIGA, Vivien WONG, Kalai SHEK and Hans Peter DIETZ Department of Obstetrics and Gynaecology, Sydney Medical School Nepean, University of Sydney, Penrith, New South Wales, Australia
Background: Defects of anterior vaginal wall fascia are generally assumed to be factors in the aetiology of cystocele. However, to date, there is very little information on diagnosis by imaging. Aim: To document the appearance of vaginal fornices before and after childbirth using 4D ultrasound volume data sets as an aid in diagnosing paravaginal defects of the anterior vaginal wall. Materials and Methods: This study was performed by re-analysing data sets obtained in a previously published study involving ante- and postpartum pelvic floor assessment by ultrasound. Two hundred and two nulliparous women had been seen at a mean gestation of 37.2 weeks at two tertiary hospitals. One hundred and sixty-three returned 3 months postpartum. All the participants underwent an interview and 4D translabial ultrasound at both antepartum and postpartum appointments. The integrity of vaginal fornices and levator ani was assessed by tomographic ultrasound. Results: Vaginal fornices were assessed in both ante- and postnatal volumes, and loss of forniceal tenting was found in 85 patients (52%). On average, seven slices were affected (range, 1–16). On multivariate analysis, controlling for potential confounders, including partial/complete avulsion, loss of forniceal tenting remained independently associated with increased cystocele descent (P = 0.005). Conclusions: Vaginal childbirth is associated with loss of tenting of the vaginal fornices, independent of levator trauma, and also with impaired anterior vaginal wall support. This evidence suggests the existence of paravaginal defects and may imply a role for such defects in the causation of anterior vaginal wall prolapse. Key words: birth trauma, fascial supports, paravaginal defect, pelvic floor, pelvic organ prolapse.
Introduction Major morphological abnormalities of the levator ani muscle in the form of an avulsion of the puborectalis muscle from its insertion on the inferior pubic ramus are common after vaginal childbirth.1–3 This trauma is likely to be important in the development of female pelvic organ prolapse3–7 and seems to be a risk factor for prolapse recurrence after surgical correction.8 While fascial trauma in the posterior compartment is relatively easy to diagnose9 and appears to be associated with childbirth,10 the situation is much different for the anterior vaginal compartment. Fascial trauma in the anterior compartment is assumed to be either due to a central defect in anterior vaginal wall fascia under the bladder or due to a separation of anterior vaginal wall fascial structures from the pelvic sidewall, commonly termed a ‘paravaginal defect’. Such defects are difficult to diagnose clinically,11 not the least because of the marked effect of avulsion on anterior vaginal wall
anatomy, and it is not clear whether these fascial abnormalities can be diagnosed on imaging. Some practitioners assess paravaginal defects by correcting the presumed defect with support during Valsalva (for example, using a ring forceps) in the lateral vagina to see whether this reduces the prolapse, but the sensitivity and specificity of this test seem to be poor.12 There have been several attempts at studying paravaginal defects using ultrasound, either transabdominally13 or translabially.11 The results obtained have been inconclusive. We undertook a study using ante- and postnatal 4D ultrasound volume data sets as an aid in diagnosing paravaginal defects of the anterior vaginal wall. For this reason, we evaluated the appearance of the lateral vaginal fornices before and after childbirth, as a paravaginal defect would be expected to change forniceal appearances on imaging.
Materials and Methods Correspondence: Dr Jordi Cassad o-Garriga. Mutua Terrassa University Hospital, Placßa Doctor Robert, 5, 08221 Terrassa, Spain. Email: [email protected] Received 30 January 2014; accepted 18 August 2014.
70
This study was performed by re-analysing data sets obtained in a previously published study14,15 involving ante- and postpartum pelvic floor assessment by pelvic floor ultrasound. Between November 2006 and October 2008, 202 nulliparous women carrying a singleton pregnancy and aiming for a normal vaginal delivery were
© 2014 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists The Australian and New Zealand Journal of Obstetrics and Gynaecology
Childbirth and anterior fascial supports
seen at a mean gestation of 37.2 weeks (range 36– 38.3 weeks) at two tertiary hospitals. All were scheduled for a second assessment at least 3 months postpartum, which was attended by 163 women (81%). All the participants underwent an interview and 4D translabial ultrasound (Voluson 730 expert with RAB 8–4 MHz transducer), after voiding, in the supine position as previously described16 at both antepartum and postpartum appointments. These data sets were evaluated by the first author, blinded to all clinical data, at a later date. A test– retest series of forniceal scoring was carried out in 21 cases to evaluate inter-rater agreement (between first and senior author, both blinded to all clinical data). The fornices were assessed by tomographic ultrasound, with eight slices obtained from the plane of minimal hiatal dimensions (Fig. 1) to a distance of 12.5 mm above this plane, encompassing the area immediately caudal to the bladder neck (Fig. 2). The vaginal fornices were rated normal if it was possible to identify a triangular fornix, with the apex aiming at the os pubis (‘tenting’) as shown in Figure 2. Forniceal tenting was rated as present or absent (1 or 0), with a total score ranging from 0 to 16 for individual patients, before and after childbirth, resulting from eight single scores per side. As opposed to tomographic assessment of the levator ani muscle, the best view of the fornices is usually obtained either at rest or on submaximal Valsalva, which tends to put intact fornices under stretch, making them more prominent. The puborectalis muscle, in contrast, was assessed on maximal pelvic floor muscle contraction. Complete avulsion was diagnosed if the plane of minimal dimensions, and the two slices at 2.5 and 5 mm cranial to it, were abnormal.16 Partial avulsion was diagnosed if only 1 or 2 of these slices were affected. As a loss of paravaginal support would be expected to result in an increased descent of bladder neck and bladder, we used bladder neck descent (BND) and maximal cystocele descent on Valsalva as potential clinical correlates of a paravaginal defect, in an attempt to validate the diagnosis.
We used Minitab V. 13 (Minitab Inc., State College, PA, USA) and SPSS V 17 (SPSS Inc., Chicago, IL, USA) for statistical analysis. Cohen’s kappa was used to evaluate inter-rater agreement and to compare the scoring of slices before and after childbirth. Pearson’s correlations were used to test for any association between forniceal scores and bladder neck descent. Ante- and postpartum scores were compared using paired t-tests. Modelling was performed to adjust for tenting. As some outcome variables were binary, these models were logistic regressions, while continuous variables were linear models.
Results A test–retest series of forniceal scoring was carried out in 21 cases. Agreement between raters was found in 297/336 observations (88.4%); the kappa was 0.63 (95% CI 0.52– 0.73). All 202 patients were nulliparous; 92% were Caucasian. Demographic and delivery data are shown in Table 1. All data except those for the second stage of labour were normally distributed. One hundred and sixty-three women (81%) attended postnatal follow-up, on median 4.08 (interquartile range, 3.50–4.90) months after birth. With US tomography, complete avulsion was diagnosed in six patients and partial trauma (any abnormal slice in slices 3–8) in 24. The fornices were assessed in both ante- and postnatal volumes, and the absence of forniceal tenting was found in 65 patients (32.2%) before and in 101 (62%) after childbirth. This change is significant on X2 test (P < 0.001, OR: 3.4, 95% CI: 2.2–5.4). On average, seven slices were affected (range, 1–16) (Fig. 3). Ultrasound data such as bladder neck descent (BND), cystocele descent and area on Valsalva ante- and postpartum are shown in Table 2. After delivery, this equated to a 4-mm change in BND and a 4.7-mm change in cystocele descent, which was highly significant. Loss of forniceal tenting postpartum was associated with increased postpartum BND (P = 0.01) and increased maximum bladder descent (P = 0.006). Loss of tenting
(a)
(b)
Figure 1 Plane of minimal hiatal dimensions in the mid-sagittal (A) and axial (B) plane. S = symphysis pubis, B = bladder, U = urethra, V = vagina, A = anus, L = levator ani, R = rectum. The intact fornices are indicated by *. © 2014 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists
71
J. Cassad o-Garriga et al.
Figure 2 Fornices were assessed by tomographic ultrasound, with eight slices obtained from the plane of minimal hiatal dimensions to 12.5 mm above this plane. In the figure, the fornices are all intact (that is, showing a triangular appearance, with the apex aiming towards the os pubis) and indicated with (*).
Table 1 Demographic and delivery data (n = 202)
Discussion
Demographic and delivery data
Levator avulsion, that is the traumatic disconnection of the levator ani muscle from its insertion on the inferior pubic ramus, is one of the most significant forms of birth trauma in women and has substantial consequences for the affected mother.17 However, it is clear that vaginal childbirth causes other forms of trauma, that is damage to nervous structures and fascia, and fascial trauma can be demonstrated by imaging in the posterior vaginal compartment where it affects the rectovaginal septum or ‘Denonvillier’s fascia’.10 To date, there is little information on the role of the fascia of the anterior compartment. It is difficult to determine the role of fascial defects in the genesis of anterior compartment prolapse, especially as their identification on clinical examination is likely to be confounded by levator trauma. Paravaginal defects and their repair were first described in the 1970s.18 Because of the difficulty of identification, prevalence figures vary greatly depending on the authors.12,19 It is not clear as to whether paravaginal defects truly exist and whether they are a significant factor in the pathogenesis of cystocele. As the anterior compartment is most prone to recurrence after surgical treatment, it seems reasonable to suggest that fascial injury should be investigated further.20,21 In this study, we have shown that forniceal tenting can be reproducibly imaged on 4D pelvic floor ultrasound,
Gestational age (control before childbirth) Age BMI Gestation at delivery Prelabour caesarean section (CS) CS in first stage CS in second stage Normal vaginal delivery Vacuum Forceps Epidural Episiotomy First stage Second stage Birthweight
37+2 (36–38+3 weeks)† 26.2 (18.1–45 years)† 31.6 (22.1–56.5 kg/m2)† 40+0 (36+6–42+4 weeks)† 14 (7%) 37 (18%) 11 (5%) 107 (53%) 22 (11%) 11 (5%) 90 (45%) 38 (19%) 456 (60–1465 min)† 54 (0–330 min)‡ 3483 (2010–4850 g)†
†Mean. ‡Median.
was associated with vaginal delivery (P = 0.0009) and BMI (P = 0.03), see Table 3. On multivariate analysis, loss of forniceal tenting remained independently associated with increased cystocele descent (P = 0.005), but not with BND (Table 3). 72
© 2014 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists
Childbirth and anterior fascial supports
Figure 3 Peripartum change in forniceal appearances. The left group of images shows intact fornices antepartum, the right group findings after a normal vaginal delivery at term. The (*) indicates loss of forniceal tenting, which is most marked in the bottom right slice.
Table 2 Ultrasound data comparing ante- and postpartum data Parameter Bladder neck descent (mm) Lowest bladder position (mm) Hiatal area on Valsalva (cm2)
Antepartum (SD)
Postpartum (SD)
23.3 (9.2) 6.98 (9.3) 22.7 (6.5)
27.3 (10.9) 2.18 (11.5) 22.7 (7.1)
Mean difference
P value
DOI: 10.1111/ajo.12261
Original Article
Can we identify changes in fascial paravaginal supports after childbirth? Jordi CASSADO-GARRIGA, Vivien WONG, Kalai SHEK and Hans Peter DIETZ Department of Obstetrics and Gynaecology, Sydney Medical School Nepean, University of Sydney, Penrith, New South Wales, Australia
Background: Defects of anterior vaginal wall fascia are generally assumed to be factors in the aetiology of cystocele. However, to date, there is very little information on diagnosis by imaging. Aim: To document the appearance of vaginal fornices before and after childbirth using 4D ultrasound volume data sets as an aid in diagnosing paravaginal defects of the anterior vaginal wall. Materials and Methods: This study was performed by re-analysing data sets obtained in a previously published study involving ante- and postpartum pelvic floor assessment by ultrasound. Two hundred and two nulliparous women had been seen at a mean gestation of 37.2 weeks at two tertiary hospitals. One hundred and sixty-three returned 3 months postpartum. All the participants underwent an interview and 4D translabial ultrasound at both antepartum and postpartum appointments. The integrity of vaginal fornices and levator ani was assessed by tomographic ultrasound. Results: Vaginal fornices were assessed in both ante- and postnatal volumes, and loss of forniceal tenting was found in 85 patients (52%). On average, seven slices were affected (range, 1–16). On multivariate analysis, controlling for potential confounders, including partial/complete avulsion, loss of forniceal tenting remained independently associated with increased cystocele descent (P = 0.005). Conclusions: Vaginal childbirth is associated with loss of tenting of the vaginal fornices, independent of levator trauma, and also with impaired anterior vaginal wall support. This evidence suggests the existence of paravaginal defects and may imply a role for such defects in the causation of anterior vaginal wall prolapse. Key words: birth trauma, fascial supports, paravaginal defect, pelvic floor, pelvic organ prolapse.
Introduction Major morphological abnormalities of the levator ani muscle in the form of an avulsion of the puborectalis muscle from its insertion on the inferior pubic ramus are common after vaginal childbirth.1–3 This trauma is likely to be important in the development of female pelvic organ prolapse3–7 and seems to be a risk factor for prolapse recurrence after surgical correction.8 While fascial trauma in the posterior compartment is relatively easy to diagnose9 and appears to be associated with childbirth,10 the situation is much different for the anterior vaginal compartment. Fascial trauma in the anterior compartment is assumed to be either due to a central defect in anterior vaginal wall fascia under the bladder or due to a separation of anterior vaginal wall fascial structures from the pelvic sidewall, commonly termed a ‘paravaginal defect’. Such defects are difficult to diagnose clinically,11 not the least because of the marked effect of avulsion on anterior vaginal wall
anatomy, and it is not clear whether these fascial abnormalities can be diagnosed on imaging. Some practitioners assess paravaginal defects by correcting the presumed defect with support during Valsalva (for example, using a ring forceps) in the lateral vagina to see whether this reduces the prolapse, but the sensitivity and specificity of this test seem to be poor.12 There have been several attempts at studying paravaginal defects using ultrasound, either transabdominally13 or translabially.11 The results obtained have been inconclusive. We undertook a study using ante- and postnatal 4D ultrasound volume data sets as an aid in diagnosing paravaginal defects of the anterior vaginal wall. For this reason, we evaluated the appearance of the lateral vaginal fornices before and after childbirth, as a paravaginal defect would be expected to change forniceal appearances on imaging.
Materials and Methods Correspondence: Dr Jordi Cassad o-Garriga. Mutua Terrassa University Hospital, Placßa Doctor Robert, 5, 08221 Terrassa, Spain. Email: [email protected] Received 30 January 2014; accepted 18 August 2014.
70
This study was performed by re-analysing data sets obtained in a previously published study14,15 involving ante- and postpartum pelvic floor assessment by pelvic floor ultrasound. Between November 2006 and October 2008, 202 nulliparous women carrying a singleton pregnancy and aiming for a normal vaginal delivery were
© 2014 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists The Australian and New Zealand Journal of Obstetrics and Gynaecology
Childbirth and anterior fascial supports
seen at a mean gestation of 37.2 weeks (range 36– 38.3 weeks) at two tertiary hospitals. All were scheduled for a second assessment at least 3 months postpartum, which was attended by 163 women (81%). All the participants underwent an interview and 4D translabial ultrasound (Voluson 730 expert with RAB 8–4 MHz transducer), after voiding, in the supine position as previously described16 at both antepartum and postpartum appointments. These data sets were evaluated by the first author, blinded to all clinical data, at a later date. A test– retest series of forniceal scoring was carried out in 21 cases to evaluate inter-rater agreement (between first and senior author, both blinded to all clinical data). The fornices were assessed by tomographic ultrasound, with eight slices obtained from the plane of minimal hiatal dimensions (Fig. 1) to a distance of 12.5 mm above this plane, encompassing the area immediately caudal to the bladder neck (Fig. 2). The vaginal fornices were rated normal if it was possible to identify a triangular fornix, with the apex aiming at the os pubis (‘tenting’) as shown in Figure 2. Forniceal tenting was rated as present or absent (1 or 0), with a total score ranging from 0 to 16 for individual patients, before and after childbirth, resulting from eight single scores per side. As opposed to tomographic assessment of the levator ani muscle, the best view of the fornices is usually obtained either at rest or on submaximal Valsalva, which tends to put intact fornices under stretch, making them more prominent. The puborectalis muscle, in contrast, was assessed on maximal pelvic floor muscle contraction. Complete avulsion was diagnosed if the plane of minimal dimensions, and the two slices at 2.5 and 5 mm cranial to it, were abnormal.16 Partial avulsion was diagnosed if only 1 or 2 of these slices were affected. As a loss of paravaginal support would be expected to result in an increased descent of bladder neck and bladder, we used bladder neck descent (BND) and maximal cystocele descent on Valsalva as potential clinical correlates of a paravaginal defect, in an attempt to validate the diagnosis.
We used Minitab V. 13 (Minitab Inc., State College, PA, USA) and SPSS V 17 (SPSS Inc., Chicago, IL, USA) for statistical analysis. Cohen’s kappa was used to evaluate inter-rater agreement and to compare the scoring of slices before and after childbirth. Pearson’s correlations were used to test for any association between forniceal scores and bladder neck descent. Ante- and postpartum scores were compared using paired t-tests. Modelling was performed to adjust for tenting. As some outcome variables were binary, these models were logistic regressions, while continuous variables were linear models.
Results A test–retest series of forniceal scoring was carried out in 21 cases. Agreement between raters was found in 297/336 observations (88.4%); the kappa was 0.63 (95% CI 0.52– 0.73). All 202 patients were nulliparous; 92% were Caucasian. Demographic and delivery data are shown in Table 1. All data except those for the second stage of labour were normally distributed. One hundred and sixty-three women (81%) attended postnatal follow-up, on median 4.08 (interquartile range, 3.50–4.90) months after birth. With US tomography, complete avulsion was diagnosed in six patients and partial trauma (any abnormal slice in slices 3–8) in 24. The fornices were assessed in both ante- and postnatal volumes, and the absence of forniceal tenting was found in 65 patients (32.2%) before and in 101 (62%) after childbirth. This change is significant on X2 test (P < 0.001, OR: 3.4, 95% CI: 2.2–5.4). On average, seven slices were affected (range, 1–16) (Fig. 3). Ultrasound data such as bladder neck descent (BND), cystocele descent and area on Valsalva ante- and postpartum are shown in Table 2. After delivery, this equated to a 4-mm change in BND and a 4.7-mm change in cystocele descent, which was highly significant. Loss of forniceal tenting postpartum was associated with increased postpartum BND (P = 0.01) and increased maximum bladder descent (P = 0.006). Loss of tenting
(a)
(b)
Figure 1 Plane of minimal hiatal dimensions in the mid-sagittal (A) and axial (B) plane. S = symphysis pubis, B = bladder, U = urethra, V = vagina, A = anus, L = levator ani, R = rectum. The intact fornices are indicated by *. © 2014 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists
71
J. Cassad o-Garriga et al.
Figure 2 Fornices were assessed by tomographic ultrasound, with eight slices obtained from the plane of minimal hiatal dimensions to 12.5 mm above this plane. In the figure, the fornices are all intact (that is, showing a triangular appearance, with the apex aiming towards the os pubis) and indicated with (*).
Table 1 Demographic and delivery data (n = 202)
Discussion
Demographic and delivery data
Levator avulsion, that is the traumatic disconnection of the levator ani muscle from its insertion on the inferior pubic ramus, is one of the most significant forms of birth trauma in women and has substantial consequences for the affected mother.17 However, it is clear that vaginal childbirth causes other forms of trauma, that is damage to nervous structures and fascia, and fascial trauma can be demonstrated by imaging in the posterior vaginal compartment where it affects the rectovaginal septum or ‘Denonvillier’s fascia’.10 To date, there is little information on the role of the fascia of the anterior compartment. It is difficult to determine the role of fascial defects in the genesis of anterior compartment prolapse, especially as their identification on clinical examination is likely to be confounded by levator trauma. Paravaginal defects and their repair were first described in the 1970s.18 Because of the difficulty of identification, prevalence figures vary greatly depending on the authors.12,19 It is not clear as to whether paravaginal defects truly exist and whether they are a significant factor in the pathogenesis of cystocele. As the anterior compartment is most prone to recurrence after surgical treatment, it seems reasonable to suggest that fascial injury should be investigated further.20,21 In this study, we have shown that forniceal tenting can be reproducibly imaged on 4D pelvic floor ultrasound,
Gestational age (control before childbirth) Age BMI Gestation at delivery Prelabour caesarean section (CS) CS in first stage CS in second stage Normal vaginal delivery Vacuum Forceps Epidural Episiotomy First stage Second stage Birthweight
37+2 (36–38+3 weeks)† 26.2 (18.1–45 years)† 31.6 (22.1–56.5 kg/m2)† 40+0 (36+6–42+4 weeks)† 14 (7%) 37 (18%) 11 (5%) 107 (53%) 22 (11%) 11 (5%) 90 (45%) 38 (19%) 456 (60–1465 min)† 54 (0–330 min)‡ 3483 (2010–4850 g)†
†Mean. ‡Median.
was associated with vaginal delivery (P = 0.0009) and BMI (P = 0.03), see Table 3. On multivariate analysis, loss of forniceal tenting remained independently associated with increased cystocele descent (P = 0.005), but not with BND (Table 3). 72
© 2014 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists
Childbirth and anterior fascial supports
Figure 3 Peripartum change in forniceal appearances. The left group of images shows intact fornices antepartum, the right group findings after a normal vaginal delivery at term. The (*) indicates loss of forniceal tenting, which is most marked in the bottom right slice.
Table 2 Ultrasound data comparing ante- and postpartum data Parameter Bladder neck descent (mm) Lowest bladder position (mm) Hiatal area on Valsalva (cm2)
Antepartum (SD)
Postpartum (SD)
23.3 (9.2) 6.98 (9.3) 22.7 (6.5)
27.3 (10.9) 2.18 (11.5) 22.7 (7.1)
Mean difference
P value
