The levator hiatus represents the largest hernial portal in the human body. What’s worse, it is traversed by three structures essential to life: the urethra for elimination of liquid waste, the anorectum for elimination of solid waste, and the vagina which is essential for reproduction. It is not surprising that enlargement of the levator hiatus may be associated with substantial disturbance of the function of several organ systems.
Levator hiatal dimensions can be determined on 3D ultrasound by identifying the plane of minimal dimensions, i.e., the plane which contains the minimal distance between the posterior symphyseal margin and the anterior margin of the levator ani loop immediately posterior to the anorectal angle (see Figure 1). Depth, width and area measurements (see Figure 2) seem highly reproducible (Intraclass correlation coefficients of 0.70 - 0.82) and correlate strongly with pelvic organ descent (Dietz et al., 2004, Dietz and Steensma, 2006). The hiatus can be measured by two different methods- both in a single axial plane (Dietz and Shek 2005) and in a rendered volume (Wong 2012), see Figure 3. Both methods are largely interchangeable, with the rendered volume method yielding slightly better repeatability.
While it is not surprising that hiatal area on Valsalva should be correlated with descent (as downwards displacement of organs may push the levator laterally), it is much more interesting that area at rest is associated with pelvic organ descent on Valsalva. This data constituted the first real evidence for the hypothesis that the state of the levator ani is important for pelvic organ support, even in the absence of levator trauma.
Relative enlargement of the hiatus on Valsalva, or rather distension or elongation of its muscular component, may be a measure of compliance or elasticity. The population distribution seems remarkably wide, with measurements from 6 to 36 cm2. Part of this seems to be due to varying degrees of reflex levator activation which is evident from a reduction in sagittal diameters during Valsalva. This necessitates biofeedback teaching which often markedly improves the quality of a Valsalva (see Figure 4). In about 20% of young nulliparous women however, this reflex levator co-activation seems impossible to overcome, even after extensive biofeedback teaching (Oerno et al., 2007).
By using ROC statistics we have proposed 25 cm2 as the limit of normal (Dietz et al, 2008). Ballooning (excessive distension of the hiatus) we define as mild (25-29.9 cm2), moderate (30-34.9 cm2), marked (35-39.9 cm2) and severe (40+ cm2). Video 1 shows normal hiatal dimensions on Valsalva; Video 2 demonstrates a case of severe ballooning in a patient with levator trauma and 3 compartment prolapse. Levator avulsion increases the size of the hiatus (by about 5-7 cm2), but there are many women in whom ballooning is congenital or due to intrapartum overdistension (‘micro-trauma’) rather than actual tears. The effect of childbirth on hiatal dimensions is only partly explained by avulsion- about half the effect, about 3.5 cm2, is due to irreversible overdistension (Shek et al., 2009). This is not surprising, given that if mammalian skeletal muscle is distended to over 150% its original length the likelihood of permanent dysfunction due to disruption of sarcomere structure is very high. We have recently been able to show that this degree of distension is very likely to be exceeded in a large proportion of primiparous women (Svabik et al., 2009).
Pelvic floor compliance or distensibility clearly deserves further study as it may be important both for the progress of labour and in the diagnosis and treatment of pelvic organ prolapse. Hiatal dimensions have been shown to be associated with length of second stage (Lanzarone et al., 2007) and delivery mode (Balmforth, 2003). It seems to correlate with digital assessment of resting tone (Thyer et al., 2007) and needs to be tested against biomechanical parameters of elasticity or compliance. Clearly, it would be of major interest as an outcome parameter in any studies intending to change levator biomechanical properties, e.g. any physiotherapy intervention targeted at urinary incontinence or prolapse.
As mentioned, an overdistensible hiatus may be congenital (Dietz and Shek 2005), but in many women it is the result of irreversible overdistension of the hiatus (‘microtrauma) in childbirth, and due to avulsion it may be markedly asymmetrical in some women (see Figure 5). We have defined microtrauma as an increase of more than 20% in hiatal area on Valsalva, and this seems to occur in about 30% of women giving birth vaginally (Shek et al., 2010). While we are currently unable to distinguish between congenital and acquired overdistensibility (‘ballooning’), it is very likely acquired in women with avulsion (see Video 2) . Regardless of aetiology, ballooning is a strong risk factor for prolapse recurrence after pelvic reconstructive surgery, with an additional 7% risk of recurrence conveyed by every cm2 of hiatal area (Rodrigo 2012).
Given that the hiatus is a hernial portal, it is likely that hiatal reduction should potentially be useful in pelvic reconstructive surgery. There seems to be little effect of conventional prolapse surgery on hiatal area, regardless of whether the procedure was successful or not (Andrew, 2011), suggesting that ballooning is more likely to be cause rather than a consequence of prolapse.
We have recently shown that it is possible to reduce the hiatus surgically (Dietz, 2011), and we are currently starting a randomised controlled trial to investigate the effect of such a procedure on prolapse recurrence (Figure 6). First results are promising.
Figure 1: Determination of the plane of minimal dimensions in the midsagittal plane (left), showing the corresponding oblique axial view on the right.
Figure 2: Hiatal dimensions measured in the axial plane. The sagittal and coronal diameters are 6.67 and 4.96 cm respectively, the area is 26.25 cm2, and circumference is 19.12 cm.
Figure 3: Two methods of measuring the levator hiatus: by rendered volume (A,B) and in the sectional axial plane (C,D).
Figure 4: The effect of levator co-activation on appearances in the sagittal (top row) and axial plane (bottom row). It is evident that both hiatal dimensions and bladder neck descent are optimised with levator relaxation (right images). Modified from: Dietz et al., Pelvic Floor Ultrasound. Springer London 2007.
Figure 5: Asymmetrical ballooning in a patient with right-sided avulsion injury.
Figure 6: Hiatal reduction from 35 cm2 to 22 cm2 3 months after insertion of a puborectalis sling. Midsagittal (A) and axial (B) view on Valsalva before anterior repair, transobturator sling and sacrospinous fixation; midsagittal (C) and axial (D) view on Valsalva 3 months after the procedure. S= symphysis pubis, B= bladder, L= levator ani.
Video 1: Normal hiatal dimensions on Valsalva.
Video 2: Massive ballooning in a patient with bilateral avulsion injury and three compartment prolapse.