The typical form of levator trauma, a unilateral avulsion of the pubococcygeus muscle off the pelvic sidewall, is clearly related to childbirth (see Figures 1-2, Dietz and Lanzarone, 2005, Kearney et al, 2006) and is generally palpable as an asymmetrical loss of substance in the inferomedial portion of the muscle, at the site of its insertion on the pelvic sidewall (Dietz et al., 2005, Kearney 2006). Bilateral defects (Fig 2) are more difficult to palpate due to the lack of asymmetry. Levator avulsion is usually occult but may occasionally be observed directly in women after major vaginal tears (see Figures 3 and 4). Main clinical markers of such trauma are tears of the vaginal sidewall (OR 6.34, CI 2.72- 14.75) and 3rd/ 4th degree anal sphincter tears (OR 6.49 (2.97-14.17) (Green et al., 2013).
The typical finding in cases of overt tears is that of a complete detachment of the vagina from the pelvic sidewall. On exploring this space one can detect the retracted muscle lateral to the anal canal, with the inferior pubic ramus more or less denuded of muscle, and the obturator fascia visible as the lateral limit of the defect (Dietz et al., 2007).
While this major form of maternal birth trauma has onjy very recently been rediscovered, Obstetricians have been aware of it in the past. The textbook illustration reproduced in Figure 5 is not quite correct in its representation of the site of rupture, but the association with Forceps and the fact that it is difficult to repair were clearly known in the 1930’s (see Fig 5). It is all the more surprising that current textbook omit all mention of avulsion.
Levator avulsion is common (about 20% in parous women, Dietz and Steensma, 2005, Dietz 2007), and associated with maternal age at first delivery- a worrying finding in view of the continuing trend towards delayed childbearing in western societies.The likelihood of major levator trauma at vaginal delivery more than triples during the reproductive years- from under 15% at age 20 to over 50% at 40 (Figure 6, Dietz 2006). Taken together with the increasing likelihood of Caesarean Section, it seems that the likelihood of a successful vaginal delivery without levator trauma decreases from over 80% at age 20 to less than 30% at age 40 (Dietz 2007).
Levator avulsion clearly is associated with anterior and central compartment prolapse and likely represents the missing link (or a large part of the missing link) between childbirth and prolapse, but the relationship with bladder dysfunction is not nearly so obvious. The larger a defect is, both in width and depth, the more likely are symptoms and/ or signs of prolapse. A levator avulsion seems to at least triple the risk of significant anterior and central compartment prolapse (see Figure 7), with much less of an effect on posterior compartment descent (Dietz and Simpson, 2008).
Not surprisingly, the effect on function is marked. Contraction strength as estimated by Oxford grading (Dietz et al 2007) and instrumented speculum (Kearney et al 2007) is reduced by about 1/3 which helps diagnose levator trauma. A patient with large cystocele, unilaterally poor levator contraction strength, and a history of a first baby born by Forceps at age 37 is very likely to have an avulsion (see tables 1 and 2 for a predictive model). Avulsion also results in a hiatus that is larger (by 20-30%), especially in the coronal plane, more distensible and less contractile (Abdool et al., 2009), and it can be highly asymmetrical (see also Figure 4 in Levator Hiatus). Video 1 shows severe ballooning in a patient with bilateral avulsion after rotational Forceps.
The most repeatable diagnostic technique at present is probably tomographic imaging (Dietz et al., 2008 and 2009). Our standard approach is to identify the plane of minimal dimensions (see above) and then to produce 8 slices at 2.5 mm interslice interval, from 2 below to 5 above the plane of minimal hiatal dimensions. In this way the region of interest, i.e., the insertion of the puborectalis muscle on the os pubis, is bracketed. If the plane of minimal hiatal dimensions is identified correctly one should see the inferior pubic rami meet one slice above the reference slice (see Figure 8). In this way the anterior aspects of one or two of the most inferior slices may be below the insertion of the puborectalis and suggest a bilateral defect. This artefact is due to the non-euclidean (warped) plane of the puborectalis muscle which means that the lowermost slices may partly be placed outside the muscle. For this reason we don’t evaluate slices below the plane of minimal hiatal dimensions. The minimal requirement for the diagnosis of a full avulsion is that the three central slices, ie., the slices at the plane of minimal dimensions plus the ones at 2.5 and 5 mm cranial to that plane, should be abnormal (Dietz 2010). If one is not quite sure whether to rate a given slice as abnormal than measurement of the levator urethra gap (LUG) can be of assistance (Dietz et al., 2008), see Figure 9.
As mentioned in the section on 2D imaging, the puborectalis muscle can be assessed by 2D ultrasound (Dietz and Shek, 2008). Figure 10 demonstrates normal findings in an oblique parasagittal plane, and Figure 11 shows 2D ultrasound and tomographic ultrasound imaging of a left sided avulsion. This technique is not as repeatable as 3D imaging but may be useful if 3D/ 4D imaging is not available.
Major morphological abnormalities of the levator ani affect surgical outcomes. There are now at least five studies showing relative risks of 2-4 for prolapse recurrence after anterior repair, Burch, hysterectomy and other urogynecological procedures (Dietz et al.2010, Model et al., 2011, Weemhoff 2011, Morgan 2011, Wong 2011). Avulsion and ballooning together are able to distinguish recurrence risks of between 12 and over 90%, and most of the effect of mesh on recurrence seems to be in women with avulsion (see Figure 1 in the Prolapse Surgery section). This implies that levator avulsion may help select women for mesh surgery; in fact, it may be considered reasonable to see avulsion as an indication for the use of anchored anterior compartment mesh.
Is there anything we can do to fix such major maternal birth trauma? Surely we should try, seeing that an avulsion triples the relative risk of significant cystocele and uterine prolapse. So far however, all attempts at intrapartum repair (Dietz et al. 2007) have failed. From a plastic surgery point of view, failure seems very likely, given the quality of the tissues and the fact that there is no opportunity for splinting or immobilisation. On the other hand, one would be very brave (or foolhardy) to suggest intrapartum mesh reinforce-ment of such trauma.
Direct repair at the time of prolapse surgery is feasible (Dietz 2011) as shown in Figure 12, through a lateral colpotomy at the level of the hymenal remnant. Results so far have been disappointing, probably mainly due to the fact that in many women the muscle is not just disconnected from its insertion, but also irreversibly overstretched.
It may be easier to limit surgical approaches to patients who need prolapse surgery in later life. A recently developed ‘puborectalis sling’ may compensate for such trauma and can reduce the size of the levator hiatus by at least 25% (Dietz et al. 2011). It remains to be seen what effect such an approach would have on prolapse recurrence.
Another major issue is prevention. We recently undertook a multicentre RCT at Nepean and Royal Prince Alfred Hospital in Sydney, testing whether it is possible to modify pelvic floor biomechanical properties by using the EPI-No device. Unfortunately, the recently analysed (unpublished) results are disappointing. In addition, it may be possible to modify birth canal distensibility during labour, but such an approach (the ‘Materna trial’) we have recently had to abandon due to engineering issues. This trial has been re-commenced in San Francisco and Houston (Baylor) after the device was redesigned.
Over the next few years levator trauma will become an issue that will influence anternatal education and informed consent. The most obvious consequence is that the weighing of pros and cons for vaginal birth after caesarean or VBAC has just become even more complex. There is anecdotal evidence suggesting that successful VBAC may be more dangerous to the pelvic floor than a first vaginal delivery. This would make sense, given that tissues are older and stiffer, the baby bigger and the uterus often more effective in contracting than with a first delivery. Anal sphincter trauma certainly seems to be more likely in VBAC.
Table 1: Model for the prediction of levator trauma (from: Dietz and Kirby, 2011; n=983). The resulting score is used in Table 2.
Table 2: Model for the prediction of levator trauma (aus: Dietz and Kirby, 2011; n=983).
Figure 1: Antepartum and postpartum findings in a patient with unilateral levator trauma after vaginal delivery. From: Dietz et al., Pelvic Floor Ultrasound. Springer London, 2007.
Figure 2: Antepartum and postpartum findings in a patient with blateral levator trauma after vaginal delivery. From: Dietz et al., Pelvic Floor Ultrasound. Springer London, 2007.
Figure 3: Levator trauma as seen in Labour Ward in a patient with large vaginal tear, on US (center) and on MR 3 months postpartum. Modified from: Dietz et al., , ANZJOG 2007; 47: 341-44
Figure 4: Levator avulsion in Delivery Suite: there is a gap between vagina and pelvic sidewall (A), and the fascia of the obturator internus is visible in B.
Figure 5: Illustration from an obstetric textbook published in 1938. The depiction of avulsion is not quite correct, as the rupture seems to generally occur at the muscle insertion.
Figure 6: The relationship between maternal age at first delivery and levator trauma. The dotted lines represent the risk of avulsion injury in patients with one or two vaginal operative deliveries (from: Dietz and Simpson, ANZJOG 2007; 491-495)
Figure 7: The proportion of unilateral and bilateral levator trauma in women with cystocele, uterine prolapse and rectocele.
Figure 8: Tomographic ultrasound imaging of a left-sided avulsion injury. The three central slices are the most diagnostic. To maximise repeatability it is suggested that the symphysis pubis be imaged open in the central left slice, closing in the central, and invisible (closed) in the right central slice, as shown by the arrows.
Figure 9: The levator- urethra gap as measured in a patient with intact pelvic floor, and in one with avulsion. The cut- off for the diagnosis of avulsion is 25 mm.
Figure 10: The puborectalis muscle can be imaged with 2D translabial ultrasound, if the transducer is oriented in an oblique parasagittal direction.
Figure 11: 2D (A is patient’s right, B is patient’s left) and tomographic ultrasound imaging of a right-sided avulsion.
Figure 12: Surgical reconstruction of an avulsed puborectalis muscle. Through a lateral colpotomy the ischiorectal fossa is entered, the muscle dissected off the vagina and reconnected to the inferior pubic ramus, using mesh reinforcement.