Pelvic floor muscle injury during a difficult labor

Abstract

Pubovisceral muscle (PVM) injury during a difficult vaginal delivery leads to pelvic organ prolapse later in life. If one could address how and why the muscle injury originates, one might be able to better prevent these injuries in the future.

In a recent review we concluded that many atraumatic injuries of the muscle-tendon unit are consistent with it being weakened by an accumulation of passive tissue damage during repetitive loading. While the PVM can tear due to a single overstretch at the end of the second stage of labor we hypothesize that it can also be weakened by an accumulation of microdamage and then tear after a series of submaximal loading cycles. We conclude that there is strong indirect evidence that low cycle fatigue of PVM passive tissue is a possible mechanism of its proximal failure. This has implications for finding new ways to better prevent PVM injury in the future.

1.1. Introduction

Pelvic floor muscle injury occurs in 13 to 41% of the women who deliver vaginally^17,29, usually during a first birth²⁴. The injury often involves the proximal pubovisceral muscle (PVM) near its origin from the pubic bone¹⁷ (Figure 1). The PVM is a critical portion of the levator ani muscle complex that helps to hold the pelvic floor closed and support the pelvic organs; it is sometimes also called the pubococcygeal muscle. PVM injuries have been associated with the development of pelvic organ prolapse⁶ later in life, significantly affecting women’s quality of life⁹, and often resulting in the need for surgery. Unfortunately, this surgery has a failure rate of approximately 30%², so preventing the injury in the first place would be desirable. This raises the following question addressed in this paper: What do we know about injuries of the muscle-tendon unit in other parts of the body and how might this help inform our understanding of birth induced injury?

Figure 1 -.

A left inferior view illustrating the pubovisceral muscle (PVM), which originates behind the pubic bone and forms the most distal portion of the levator ani muscle. It is particularly stretched late in the second stage of labor as the fetal head is driven by strong 2–3 maternal pushes, at three minutes intervals, through the birth canal to crown. EAS: External anal sphincter; ICM: iliococcygeal muscle; PRM: Puborectal muscle; ©DeLancey.

A recent systematic review of 3,246 musculoskeletal injuries (including partial or complete tears, whether in the tendon, the musculotendinous junction, the tendon-bone junction, or the muscle belly; tendinopathy, whether in the tendon-bone junction or tendon midsubstance; tendon avulsions, defined as a complete detachment of the tendon from the bone while containing a small fragment of bone and musculotendinous strains if verified using magnetic resonance imaging, computed tomography scans, or surgery) found 35% of the injuries within the tendon, 28% in the musculotendinous junction, 18% at the tendon-bone interface, 13% within the muscle belly and 6% tendon avulsions that included a bony fragment³⁸. There was an age effect in that the mean ages at injury were 29 years for the muscle belly, 33 years for tendon avulsion, 34 years for musculotendinous junction, 47 years for tendon mid-substance injuries and 51 years for tendon-bone interface. Only 13% of all injuries were linked to evident trauma such as a fall, a work or car accident, or sports injuries sustained during collisions or tackles involving high tensile loading rates: 63% were tendon avulsion injuries, 15% being tendon mid-substance injuries, 8% musculotendinous junction injuries, 3% muscle belly injuries and 2% being tendon-bone junction injuries. Since the preponderance, 87%, of injuries were not associated with evident trauma and occurred in well-trained athletes (professional, collegiate and high-schoolers) and people previously or currently actively engaged in sports activity, we have argued in Vila Pouca et al³⁸ that many of these atraumatic injuries may therefore be due to passive tissue fatigue failure: namely repeated sub-maximal loading cycles causing an accumulation of microdamage that progressively weakens the passive tissue in the muscle-tendon unit until it fails under a normal sub-maximal loading cycle.

The reason for this proposition is that microdamage accumulation, in the form of collagen unravelling has been demonstrated at the origin of a passive, Type 1 collagen structure, the anterior cruciate ligament (ACL), following repetitive sub-maximal loading⁴, and in the form of collagen denaturation when a tendon reaches its yield point^14,42. Microdamage accumulation under repetitive loading has also been reported at the musculotendinous junction of rabbit leg muscles (for example²⁵) and it has led to complete structural failure of tendon in vitro (for example,²⁶) as well as in vivo¹. These examples of damage mechanisms weakening a tensile structure under repetitive sub-maximal loads until it fails are consistent with the phenomenon of low-cycle tissue fatigue^a whereby, in biological tissue, the rate of damage accumulation exceeds the rate at which the tissue can repair itself.

What is the mechanism of low cycle tissue fatigue^b? It is widely known in material science that when a large enough tensile load is applied to a structure, the structure will fail in one loading cycle due to the stress or strain in the material exceeding that which it can withstand without failing (Figure 2). In low-cycle tissue fatigue^b, however, two or more sub-maximal loading cycles, neither of which are intense enough to induce failure by themselves, cause a sufficient number of defects to appear, coalesce and accumulate with each cycle so as to weaken the material to the extent that it fails (tears) partially or even completely under the next sub-maximal cycle (Figure 3). An additional characteristic of low cycle fatigue behavior is that the lower the load being applied repetitively, the greater the number of loading cycles the structure will withstand before failing, and vice versa¹⁶. Let us say the second stage of labor lasts, say, two hours in nullipara¹¹, and a maternal push is timed to coincide with the uterine contraction every 3 minutes¹² this would require up to 40 pushes by the nullipara. If coached, the maternal push can be performed up to 2–3 times per contraction¹⁰, which would require up to 80–120 pushes by the nullipara. Any structure that fails under 2 to 120 loading cycles due to an accumulation of microdamage weakening can be described as failing due to low cycle material fatigue^c.

Figure 2 -.

Schematic illustration showing the relationship between material stress^d and strain^e as a tissue fails under an applied displacement (a) or applied load (b) load in a single loading cycle at each of three loading rates. The top row at left is the case when a linearly increasing displacement is applied to the material, at right (b) is the case when the same material is loaded instead by a linearly increasing tensile force. In the top row of (a) and (b) the three different curves in each plot are for a high strain rate (blue), a medium strain rate (pink) and a low strain rate (yellow). The lower panels for a) and b) show the corresponding relationships between the tensile stress induced in the tissue and its strain. Note the effect of loading rate on the value of the tensile stress at failure (indicated by the arrow): the lower the rate of loading, the lower the stress, but the longer the length at which the material fails.

Figure 3 -.

Schematic representation of how low cycle fatigue can cause the failure of a soft tissue material subjected to cyclic sub-maximal stretch loading. (a) A cyclic linear increase and decrease in stretch is applied with the peak value less than the value that would cause failure in the first cycle. The lower figure shows the time course of the failure occurring at a lower stress than the peak stress induced by the first loading cycle, due to an accumulation of microdamage in the tissue. (b) A cyclic linear increase and decrease in tensile force is instead applied to the same material, with the peak force value being less than that required to cause failure in the first cycle. The lower panel shows the time course of the peak tensile strain induced by each loading cycle until failure occurs during the fifth loading cycle due to an accumulation of microdamage. Pushing during the second stage of labor loads the pelvic floor tissue most closely to the example shown at right in (b).

If, as we have described above, low cycle fatigue can occur in a highly organized, fibrous structure as specialized as the ACL, as well as at the origin and insertion of leg muscles, we hypothesize that low-cycle fatigue can also occur at or near the origin of the PVM during a difficult labor. The right and left PVM are the last pelvic floor muscles to be engaged by the fetal head late in labor and, more importantly, they are the portion of the pelvic floor muscles that are stretched more than any other pelvic floor muscles³. The PVM of the nullipara have never been called upon before to resist a tensile force or elongation of the magnitude induced by the fetal head as it elongates the muscle every three minutes towards the end of the second stage of labor (Figure 1) and despite the protective adaptations that occur during pregnancy in preparation for childbirth, it is still subjected to an extraordinary loading. Therefore, the aim of the present paper was to determine whether there is indirect evidence that would support the hypothesis of low cycle fatigue causing PVM failures during a difficult or long labor of the type described by Tracy et al.³⁵.

1.2. What distinguishes muscle/ligament origins and insertions?

The origin and insertion of tendons and ligaments are known to be highly developed regions of transition between these strong, flexible structures and the more rigid bone to which they attach. Since tendon and ligament share a similar hierarchical structure, with highly aligned collagen fibrils, fibers, fiber bundles and fascicles, the interfaces between tendon or ligament and bone have been described as indistinguishable²³. So, we shall here term them as the “collagen-bone junction (CBJ)”.

The CBJ, also known in some cases as an enthesis, can be fibrous or fibrocartilaginous in nature (Figure 4). A fibrous CBJ is characterized by dense fibrous tissue at the interface with perforating mineralized collagen fibers and a structure that becomes ever more mineralized as one nears the bone²⁸. Other CBJ have fibrocartilage interposed between the collagen tissue and the bone, such as occurs in the anterior cruciate ligament, thereby creating a continuous gradient in mechanical properties, such as the elastic modulus, from uncalcified collagen to calcified bone²⁸.

Figure 4 -.

Examples of medial, central and lateral region histology of the pubic origin of the PVM. Each row is from a different donor. Histological images showing the pubic origin of the levator ani from medial (column 1), central (column 2), and lateral (column 3) areas with the orientation picture below. All samples shown are stained in Masson’s trichrome, and scale bars are 5mm for each row of images. Medial: The medial levator ani muscle fibers originate from multiple slips attaching in an enthesis to the pubis; oblique interface between the pubic bone (PB) and the levator ani muscle (LA) can also be observed. The thickness of the LA is greater than in other areas. N/A denotes missing data due to a technical issue during harvesting process. Central: The central portion originates from the PB in a single aponeurotic attachment, which is noticeably thinner than medial portion. The obturator internus muscle (OI) can be seen lateral to the LA. Lateral: The levator arch (LArch) appears as a dense blue connective tissue attaching to the PB and forming the lateral margin of the pubic origin of the LA. Note that relative preponderance of the three portions varies by individual. ©DeLancey

Muscles are typically connected to their tendons via finger-like projections of muscle fascicles which interdigitate with collagen tissues³⁷.

Muscle and ligament origins and insertions characteristically include junctions between dissimilar connective tissues, the active muscle on the one hand, and the passive collagenous connective tissue on the other, each having significantly different tensile material properties. From an engineering point of view, the locations where dissimilar materials interdigitate are typically locations where mechanical stresses concentrate^f. In other words these junctions sustain higher stresses even when the whole muscle-tendon unit is uniformly loaded, making them more prone to damage³⁷. This fact logically makes muscle origins and insertions a region of interest as a potential site for damage to accumulate at and this could eventually accumulate to the point of causing musculoskeletal injury.

1.3. Characteristics of the origin of the pubovisceral muscle

The PVM arises bilaterally from the pubic bone as a thin parallel fibered muscle which thickens to attach, more centrally and caudally, to the perineal body, the vaginal wall, and anal intersphincteric groove².

Its origin on the dorsal aspect of the pubic ramus is long and thin, almost eyebrow shaped (Figure 5), with the thicker region located more medially, then thinning and extending laterally about 30 mm⁸. In the medial and central portions, a fibrous aponeurosis of thin passive collagenous tissue arises from the synovium over the bone, shorter in the medial region and longer in the central region, to attach to the pubovisceral muscle fibers⁸. The lateral portion of the origin attaches to the ventral portion of the arcus tendinous levator ani (levator arch) that originates from the pubic bone about a centimeter lateral to the pubic symphysis, and passes medial to the obturator fascia to insert on the ischial spine, the latter serving as the nominal boundary between the PVM and iliococcygeal portions of the pelvic floor muscle⁸.

Figure 5 -.

Characteristic features of the origin of the right pubovisceral muscle (PVM). Medial and central portions attach to the pubic bone by an aponeurosis, while the lateral portion attaches to the levator arch (a.k.a arcus tendineus fascia pelvis). Laterally the structure is thin enough to visualize the forceps placed under it for demonstration purposes. ©DeLancey

1.4. Is there evidence of low-cycle fatigue damage occurring in soft tissues?

Passive connective tissues, consisting principally of Type 1 collagen, are known to exhibit low-cycle fatigue damage under repetitive strenuous loading. As we have noted in the Introduction, molecular, cellular and ultrastructural evidence of low-cycle fatigue damage accumulation has been demonstrated in the Type 1 collagen of the femoral enthesis of anterior cruciate ligament after repeated sub-maximal loading⁴. This has been observed clinically⁴¹ as well as experimentally under large repetitive sub-maximal knee loading cycles¹⁵. Likewise, in tendon, another passive connective tissue structure also with a high Type 1 collagen content, microdamage in the form of collagen fibril disruption has been reported at only 6% strain^{g 32}. In addition, the resistance to fatigue damage is known to decrease with age, both in the interfascicular matrix and in the fascicles³⁴, a fact that might be important in understanding why older nullipara have a significantly higher PVM injury risk⁷. So, if a Type 1 collagen structure with the generous cross-sectional dimensions of ligaments or tendons can accumulate fatigue damage under repetitive strenuous loading, it seems reasonable that heavily loaded sheet-like structures, such as aponeuroses and fascia connecting the PVM to adjacent structures can also exhibit an accumulation of microdamage. It is noteworthy that passive structures like a ligament or tendon are not the only structures to exhibit fatigue damage under repetitive loading.

In striated muscle, for example, damage accumulation near the musculotendinous junction has also been reported under repetitive submaximal loading, with histology showing areas of fiber rupture and hemorrhage near the musculotendinous junction²⁰ and even incomplete disruption following a strenuous cyclic loading history³³. Recently, the microdamage accumulation in the muscle has been described as a disruption in the myofibrils, whereby sarcomeres lose their structural integrity³⁴. So, active structures such as striated muscle can also exhibit low-cycle fatigue damage as well.

Finally, it well known that even semi-rigid materials like bone can also accumulate fatigue damage, causing the bone to be susceptible to stress-fractures³⁹, a phenomenon that has recently been shown to occur in the pubic bone during a delivery¹⁷.

So, in summary, there is strong evidence in the literature that passive collagenous structures such as ligament and tendon, active structures such as muscle, and even bone can all accumulate microdamage under repetitive sub-maximal loading. Moreover, the mechanism of fatigue damage seems similar in these structures in that damage starts in the microstructure, progressing under continued strenuous loading cycles to larger length scales of organization until it can affect the macroscopic properties or even cause structural failure. This is consistent with fatigue damage seen in engineering materials, where the first signs of fatigue damage are microscopic imperfections or flaws that can grow into macroscopic cracks under repeated loading, these in turn can propagate to a critical size until the structure partially or even completely fails¹⁶. The main difference with the engineering materials, is that biological tissues have the ability to heal, grow and remodel. So, if the rate of damage accumulation is less than the rate at which the tissue can repair itself in a timely manner, it would not be prejudicial. In fact, it could even be beneficial if the microdamage triggers metabolic and structural adaptations that, to reverse it, cause muscle or tendon growth⁵ consistent with Davis’s law. However, a combination of high sub-maximal loads, many repetitions and/or short recovery times, will eventually lead to irreversible microdamage progression which in turn can cause irreversible macroscopic defects thereby weakening the tissue to cause partial or even complete failure.

In engineering materials, ultrasound is routinely used to detect signs of microdamage accumulation so action can be taken to avert a disastrous failure. For biological tissues, scanning acoustic microscopy (SAM) can provide high-quality microscopy of tissues and cells comparable with light microscopy without the need for staining¹⁸ as well as provide data on the elasticity of tissues, thereby measuring mechanical alterations¹⁹. The marked advantage of SAM is the ability to inspect the sample subsurfaces layer by layer simultaneously with good penetrating power of the ultrasonic waves while scanning the material surface, for detecting micro defects⁴⁰. The disadvantage is that it still requires flat ~10–15μm samples¹⁸; however, with the advent of better imaging modalities, it might be possible to develop preventative screening for biological tissues and find the threshold for detrimental or irreversible microdamage accumulation.

1.5. Can the origin of the pubovisceral muscle be susceptible to low-cycle fatigue damage in a difficult vaginal birth?

In the second stage of labor of a typical vaginal birth it is known that spontaneous uterine contractions occur approximately every 3 min, increasing the pressure near the fetal head from 2.6 kPa to 8.5 kPa. Maternal pushes (occurring 2–3 times per uterine contraction if instructed) are added to facilitate the fetus expulsion, increasing this pressure to 19 kPa³. This gives expulsive forces on the fetal head of 24 N at rest, of 74 N at the peak of uterine contraction and of 101 N at the peak of each maternal push (considering and fetal cross-sectional area of 9.2 cm²). So, at peak, this gives a combined effort of 175 N. In case of maternal exhaustion, when the women can no longer make progress, vacuum device is applied to the fetal head for an additional traction force of 113 N, typically up to four pulls, or if obstetric forceps have to be used, the additional traction force can reach 200 N³. Besides these forces, the fetal head induces extraordinary maternal tissue stretch, which many studies have shown to be especially large in the PVM portion of the pelvic floor. So, despite the pregnancy induced alterations in the pelvic floor to facilitate birth, levator ani injuries are prevalent after vaginal delivery.

Magnetic resonance images (MRI) taken weeks after a difficult delivery in nullipara showed pubic bone edema in 66% of the cases, pubic bone fracture in 29%, muscle edema in 90% and low-grade or greater tear in the pubovisceral muscle in 41%¹⁷. Months later, MRIs of the same women showed the edema to have subsided, but the previously torn PVM was no longer visible at its normal attachment site on one or both sides¹⁷. This edema is evidence of injury and the ensuing reparative process. Shek and Dietz²⁹ hypothesize that minor defects to the PVM portion of the levator ani could cause an irreversible increase in the hiatal area and called this a ‘microtrauma’ injury. 4D Transperineal ultrasound (TPUS) was performed between 36 and 38 weeks gestation and 3–4 months after vaginal delivery and found a complete detachment of the PVM in 13% of the cohort and an irreversible enlargement >20% of the hiatal area on Valsalva in 29%²⁹. Despite the term ‘microtrauma’, the authors actually describe phenomenon consistent with low-cycle fatigue, such as microdamage and permanent/plastic deformation. Recently, Da Silva et al^30,31 attempted to correlate the findings of TPUS with cadaveric observations and verified that injuries showing a complete detachment of the PVM under TPUS (also designated as “avulsion” in the literature) were truly still attached to the pubic bone, although appearing significantly narrower. Histology performed on the cadaveric PVMs showed tortuous muscle fibers and loss of striation, which interestingly is also consistent with the hypothesis of microdamage accumulation, despite the lack of a complete tear. These are only a small percentage of the studies reporting levator ani injuries after a vaginal delivery which, despite being extensive reported, still have poorly understood pathomechanics. Nevertheless, these results demonstrate just how strenuous the pelvic floor tissue and pubic bone loading is during birth, as well as showing the proximal PVM to be a region that is commonly injured in a difficult delivery. Exactly where in the PVM it is injured is still unknown – is it where the aponeurosis attaches to the bone, the aponeurosis itself, the junction between the aponeurosis and the muscle, or the muscle itself?

What can we learn from studying injury sites in musculotendinous units elsewhere in the body? Our review³⁸ showed that bony avulsions and collagen-bone junction injuries are typically linked to older individuals. Since the age of first time mothers most commonly ranges from 24 – 29 years old (data from 37 developed countries of Europe, East Asia and USA²⁷), this makes the aponeurosis, musculotendinous junction, or muscle belly the most likely injury locations in younger women. MRI obtained after difficult first time deliveries shows signs of edema at the origin of the pubovisceral muscle¹⁷, which makes it more likely that an aponeurosis avulsion or musculotendinous junction occurred. In our review of muscle-tendon injuries in other parts of the body, avulsion injuries were frequently linked with an evident trauma³⁸ would certainly be consistent with the kind of injury sustained during an instrumented delivery with forceps when the head is often delivered by a single pull lasting a few seconds during which the applied traction force can reach 200 N²¹. But in the case of a difficult non-instrumented delivery, it could typically involve more than one stretch of the PVM at the end of second stage of labor. For the same reason, the rate of loading is an order of magnitude slower, for example, than a single fall onto an outstretched arm, which is a frequent cause of triceps brachii tendon avulsions in a tenth of a second²². The strains and loads acting on the pelvic floor are very significant, however the pelvic floor is relatively prepared for childbirth, so it is not expected that the strains and loads normally exceed the ultimate failure strain and/or stress^h without an instrumented delivery. An injury can happen, and in that case the injury would occur due to that single forceful load exceeding the ultimate failure properties of the tissues. However, in a typical labor, it is more likely that as the mother pushes hard every three minutes to deliver the fetal head¹², causing high but sub-maximal strains (which in the caudal part of the PVM, can reach to three times its original length¹³), microdamage starts to develop from the start, quickly progresses and eventually causes a partial or even complete failure, given that there is no time for any damage to be repaired.

Thus, we postulate that a prolonged or difficult second stage of labor, involving multiple and forceful pushes, can cause damage to accumulate at the weakest link of the proximal PVM, most likely the aponeurosis itself, or the junction between it and the muscle. In some cases, the damage accumulation can mean that a forceful push that ordinarily would not cause injury if it were the only push, can cause a complete tear because of the recent repetitive history of large sub-maximal loads in a structure that on day-to-day activity, is never called upon to withstand such large repeated loading cycles.

This argument does not preclude the possibility that a single forceful push, on its own, and without low-cycle fatigue, could tear the PVM from the bone, its CBJ or even the muscle belly, due to the extraordinary stretch required (vide supra) exceeding the ultimate strain in and/or stress in the structure. Rather, this paper augments the “single push” failure mode by adding a second mechanism of PVM failure: namely low cycle fatigue failure of the PVM after repetitive submaximal pushes. If future studies confirm this second possibility, it should lead to new ways of preventing this injury in the labor and delivery room during a difficult birth. Given that these tissues are viscoelastic^35,36, there is a tradeoff: making each push long enough to maximize tissues stretch during each push in order to facilitate fetal head descent, but not inducing maternal exhaustion with too lengthy or strenuous pushing, and knowing that too strenuous pushing near the end of labor increases the risk of partial or even complete material fatigue of the PVM. Given these factors, it might prove fruitful to identify the optimal pushing strategy, building on the work of Lien et al¹².

1.6. Conclusions

1. Evidence from musculoskeletal injuries elsewhere in the body point to the aponeurosis or the musculotendinous junction as being the weak links in pelvic muscle placed under large repetitive loading in young women of childbearing age.

2. This supports our hypothesis that the PVM may accumulate sufficient microdamage during the multiple forceful pushes near the end of the second stage of labor to weaken the tissue to the point of a partial or complete tear of the structure during the next push.

3. The clinical significance of these observations is that minimizing the number or intensity of such loading cycles during a vaginal birth would reduce the risk of irreversible PVM injuries, especially if the loading rate can be reduced to allow time for tissue viscoelastic behavior to aid in tissue elongation.