Pubic Bone Injuries after First Childbirth: Utility of MR in detection and differential diagnosis of structural injury

Abstract

Objective

Evaluate utility of MR in diagnosing structural injury in primiparous women with pelvic floor injury risks.

Methods

Observational study of 77 women with 3T MR imaging after delivery. Women (n=45) were operationally defined as High Risk for levator ani muscle tears (2nd stage labor> 150 minutes, anal sphincter tear, forceps, age> 35, birth weight > 4000 grams) or Low Risk (n=32): vaginally delivered without these risk factors (n=12); delivered by cesarean after 2nd stage labor >150 minutes (n=14), and cesarean without labor (n=6). All women were imaged using MR fluid sensitive sequences. Two musculoskeletal radiologists reviewed images for bone marrow edema, fracture, pubic symphysis measurements, and levator ani tear.

Results

MR imaging showed pubic bone fractures in 38% of women at High Risk and 13% of at Low Risk for pelvic floor injury (χ²(3)= 9.27, p=0.03). Levator ani muscle tears were present in 44% of the High Risk and 9% of Low Risk women (X2(3)=11.57, p=0.010). Bone marrow edema in the pubic bones was present in 61% of women studied across delivery categories. Complex patterns of injury included combinations of bone marrow edema, fractures, levator ani tears and pubic symphysis injuries. No MR documented injuries were present in 18% of women at High Risk and 44% at Low Risk (χ²(1)=6.2, p=.013).

Conclusions

Criteria identifying primiparous women at risk for pelvic floor injury can predict increased risk of bone and soft tissues changes at the pubic symphysis. Fluid sensitive MR imaging has utility for differential diagnosis of structural injury in postpartum women.

Introduction

Criteria identifying women at high risk for labor-related injuries, such as anal sphincter injury and damage to the levator ani muscles have been identified^(1-2). However, despite the frequent clinical occurrence of pubic pain after delivery, limited descriptions of pubic changes and relationships of bone and joint trauma to muscular injury have been reported in the literature. Complex combinations of osseous and soft tissue injuries may be occurring which could impact postpartum recovery for women, especially those at high risk for pelvic floor injury.

The reported incidence of levator ani muscle damage is as high as 15% at first births^(3-5). Damage to this important supportive musculature is associated with later long-term sequelae of vaginal and rectal prolapse^(4,6-8). Levator ani muscles can tear at their insertions on the pubic bones and do not recover after difficult vaginal births⁽⁹⁾. Injuries to the pubic bones and pubic symphysis, while known to occur, have been incompletely evaluated by MR. Yet such bone and soft tissue trauma might explain why some women have difficulty recovering postpartum due to refractive symptoms of pain or activity intolerance and the diverse clinical presentations of symphysis pubis dysfunction⁽¹⁰⁾ or symphysial pelvic dysfunction⁽¹¹⁾. These MR soft tissue and bony alterations may contribute to a better understanding of pregnancy related structural changes, thus representing an important clinical area of concern for childbearing women.

While evaluating levator ani muscle status after vaginal birth, we became aware of changes visible in the pubic bones in certain women. The purpose of this paper is to examine the MR detected occurrence of a) bone marrow edema in the pubic bone, b) evidence of fracture, and c) changes to the pubic joint in primiparous women, comparing women with criterion-based High and Low risk for pelvic floor injury during childbirth.

Materials and Methods

Patient Sample

The study sample consists of 77 women post first birth (Table 1) who completed a MR approximately six weeks postpartum. Women were recruited from a larger parent study on levator ani muscles and childbirth that our Institutional Review Board approved.

Table 1.

Characteristics of Labor and delivery for women in four delivery categories.

	High Risk Vaginal (n=45)	Low Risk Vaginal (n=12)	Cesarean Section after long 2nd stage (n=14)	Cesarean Section, no labor (n=6)
Time post delivery to MRI (days± sd)	43.3 (15.2)	44.8 (15.8)	42.8 (20.8	43.6 (9.3)
Age	30.9 (6.4)	25.7 (6.0)	28.0 (5.1)	30.6 (6.2)
Birth Variables
Second stage ≥ 150 minutes (%)	50%	0%	100%	0%
Birthweight ≥ 4000 grams (%)	11%	0%	14%	0%
Anal sphincter tear	22%	0%	0%	0%
Forceps delivery	4%	0%	0%	0%
Epidural placed	73%	100%	86%	0%
Second stage average (mins-sd)	166.5 (137.3)	88.2 (40.9)	301.6 (78.6)	0

One group consisted of 45 primiparous women whose birth events included factors known to be associated with levator ani muscle injuries (categorized the High Risk group). Our definition of high risk was based on the data available in 2007, when the study was initiated, and these factors were for levator ani injury, not fracture. Criteria defining High Risk include prolonged second stage labor (>150 minutes), 3^rd or 4^th degree lacerations, instrument assisted delivery (vacuum or forceps), maternal age 35 years or older, baby’s birth weight greater than 4,000 grams, and unusually short 2^nd stage (less than 30 minutes of pushing)⁽³⁾. The operationally-defined Low Risk group (n = 32) consisted of: a) women who delivered vaginally without these risk factors for levator ani muscle injury (n = 12), b) women with a long 2^nd stage (>150 minutes) delivered by cesarean (n = 14), and c) women who delivered by planned cesarean with no labor (n = 6). These risk factors were selected for the parent study to enhance sampling for those with major levator ani injury, as based on Kearney⁽³⁾. For instance, the cutpoint of 150 minutes for second stage labor was chosen as a compromise between the average 93-minute second stage shown in primipara without any levator injury and the average 195-minute second stage in primipara with major levator injury.

Subjects were recruited by reviewing birth records for all primiparous women at a tertiary level maternity care hospital. Subjects meeting High Risk criteria were recruited between May2007 and September 2009. The comparison Low Risk group was recruited in the same institution from April 2009 to September 2009. After chart review indicating likely eligibility, women were mailed an informational invitation shortly after their hospital discharge, followed by a telephone call of invitation to participate. If interested in participating, they were screened during the phone conversation. Exclusion criteria included women less than 18 years of age, non-English speakers, women with infants in intensive care units, women unable to make the commitment for repeat visits to the hospital, or those without childcare. For all women in the study, informed consent was obtained and participants were compensated for their time. None of the authors clinically managed the women in the study.

MR Imaging

Women were invited to be imaged after delivery (mean ± sd 43.6 ± 15.8days). The time frame of 2 - 7 weeks postpartum was selected to allow for resolution of typical post-traumatic acute blood and non-specific soft tissue edema, which may mask the visualization of structural changes. After a few weeks with resolution of blood and edema, the remaining structural injuries are more likely to be clinically significant. However, seeing muscle tears acutely helps to tie any later muscle atrophy to the acute event. Fracture healing follows a well-documented cascade of processes. Mineralization, endochondral ossification and new woven bone formation only begin 2 weeks after a fracture. Bone remodeling at the fracture site proceeds for at least 12 weeks depending on the size of the fracture, motion at the fracture site and bone quality^(12-14). Woven bone formation is also a response to fatigue loading of bone and can be seen after even one episode of such loading⁽¹⁵⁾. New bony remodeling whether bone marrow edema or fracture, can continue to be detected by MR and bone biopsy for many months even after clinical recovery has occurred⁽¹⁵⁾.

MR images were completed on a 3 Tesla, Philips Achieva, Philips Medical System, Eindhoven, Netherlands; 8 channel cardiac coil positioned over the pelvis and included coronal, axial, and sagittal proton density-weighted sequences (relaxation time (TR) 2107 ms; echo time (TE) 30 ms; number of slice averages (NSA) 2; slice thickness 4mm, gap 1mm; field of view (FOV) 20 cm, matrix 256 × 256). For better definition of the anterior pelvic floor anatomy, additional 2 mm proton density-weighted sequences in the axial, coronal, and sagittal planes were obtained (relaxation time (TR) 2107 ms; echo time (TE) 30 ms; number of slice averages (NSA) 2; slice thickness 2mm, gap 0.2mm; field of view (FOV) 18 cm, matrix 256 × 256). Additional fluid sensitive imaging in axial and coronal planes through the anterior pelvic floor were obtained with either proton density-weighted fat saturation (relaxation time (TR) 2355 ms; echo time (TE) 30 ms; number of slice averages (NSA) 2; slice thickness 2mm, gap 0.2mm; field of view (FOV) 18 cm, matrix 256 × 256) or short tau inversion recovery (STIR) sequences (relaxation time (TR) 5987 ms; echo time (TE) 60 ms; number of slice averages (NSA) 2; slice thickness 2mm, gap 0.2mm; field of view (FOV) 18 cm, matrix 256 × 256). The MR images were reviewed by two board-certified radiologists with fellowship training in musculoskeletal imaging. They used a standard monitor of the Picture Archiving and Communication System (PACS) and were blinded for obstetric and delivery factors. The radiologists agreed on 98% of the categories, disagreements were between adjacent grades. Consensus was achieved by reviewing the images together and discussing the findings. One radiologist graded 22 women more than 6 months before the final grading sessions with 2% to 7% disagreement between the presence or absence of the mildest grades of fractures and bone marrow edema.

Pubic Bone Evaluation

Bone marrow edema was assessed by noting the presence of increased signal within the marrow of each of the pubic bones as compared to the ischial tuberosity on either proton density-weighted fat saturation or STIR sequences. If present, the increase in signal was recorded as symmetric or asymmetric and for each side the location of increased signal was noted.

Fractures were defined as linear decreased signal on the proton density-weighted sequence associated with matching focal increased signal on the proton density-weighted fat saturation or STIR sequences. To be documented as present, this finding had to be visible in two imaging planes. Fractures were recorded as either subcortical or cortical fractures. Location of the fracture was noted for each side.

The pubic symphysis was assessed for changes. The pubic joint capsule was measured with electronic calipers at a workstation from anterior superior pubic cortex to the undersurface of the superior pubic ligament in the coronal plane 2mm proton density-weighted sequence. Measurements of pubic symphysis width were made at the midpoint of the joint (cephalocaudad) in the axial plane 2mm proton density-weighted sequence.

Levator Ani Evaluation

The levator ani muscles were evaluated for muscle fiber loss in two or more adjacent 2 mm sections in both axial and coronal planes proton density-weighted sequences. Each side of the muscle was graded separately. A muscle tear was recorded if greater than 20% of the expected muscle volume was absent. Evaluation of muscle volume or bulk was based on overall appearance of the muscle in all three planes, an estimate common in clinical and research practice and not a detailed quantification on individual images. Similarly 20% to 25% is a general threshold for decision-making. This paper focused on injury of the pubic symphysis and not an analysis of levator ani injury.

Demographics, Labor and Delivery Variables

Labor and delivery variables were obtained by hospital chart review conducted by a certified nurse midwife.

Statistical Analysis

Analyses were conducted using SPSS version 11 (SPSS Inc. Chicago Il). Descriptive statistics were used to characterize the sample and distributions of injuries. Chi square tests examined differences in the distribution of injury among groups. ANOVA was used to examine differences by injury and number of days to imaging time, and for pubic symphysis measures across risk groups. All tests of statistical significance were set at a predetermined level of p<0.05.

Results

Risk Factors

The distribution of women among criterion based High Risk and Low Risk labor and delivery categories is shown in Table 1. As illustrated in this Table, the most frequently occurring risk factor among the High Risk group was extended duration of the second stage of labor delivery (50%). Birth variables present in lower-risk deliveries included 14% of infant birth weights greater than 4,000 grams.

The distribution of MR-detected post-birth pubic injuries among these risk groups is summarized in Table 2. The statistical significance of differences in extent of injuries is also reported.

Table 2.

Summary of MR-detected pubic injuries in women among at-risk delivery groups.

Injury	High Risk Vaginal (n=45)	Low Risk: Vaginal (n=12)	Low Risk: Cesarean Section after long 2nd stage (n=14)	Low Risk: Cesarean Section, no labor (n=6)	Test of Statistical Significance Statistical test (df) p-value
Pubic bone fracture	17 (37.8%)	0 (0.0%)	4 (28.6%)	0 (0.0%)	χ2(3) = 9.27, p=0.03
Bone marrow edema	30 (66.7%)	5 (41.7%)	9 (64.3%)	3 (50.0%)	χ2(3) = 2.86, p= 0.41
Pubic symphysis
Capsule height (mm+sd)	2.24 (1.7)	1.58 (1.0)	2.21 (1.4)	2.30 (1.4)	F (3, 76)=, .658, p-0.58
Width	2.98 (0.7)	3.00 (0.4)	3.64 (1.9)	3.06 (1.0)	F (3,76) =, 2.34, p=0.08
Levator ani tear	20 (44.4%)	2 (16.7%)	1 (7.1%)	0 (0.0%)	χ2(3) = 11.57, p=0.01
No MR findings*	8(17.8%)	6 (50.0%)	5 (35.7%)	3 (50.0%)	χ2(1) = 6.2, p = .013

^*For the presence/absence of MR-documented injuries, we compared the High Risk group to the combined Low Risk Groups, to facilitate comparisons with results in existing research literature⁽³⁾

Pubic Bone Fracture

As indicated in Table 2, pubic bone fractures were present in 21 of the 77 women. Injuries varied by group with significantly higher proportion of fractures (17of 45 women) occurring in the High Risk group (χ² (3) = 9.27, p = 0.03). Only 4 of the 32 women in the Low Risk group showed pubic bone fracture; all 4 were in the group of 14 women who delivered by cesarean after long 2^nd stage labor. All fractures originated within the posteroinferior and/or para-symphyseal region of the pubis bone (Figure 1). Although most fractures were unilateral, 3 women had a fracture in both pubic bones. There were two cortical fractures, one of these extended along the inferior pubic ramus (Figure 2).

Fig. 1.

26-year-old woman with pubis fracture 4 weeks from cesarean section after prolonged labor. (A) Axial proton-density MR image and (B) coronal proton-density-weighted MR image with fat saturation show right inferior pubic bone marrow edema (B) and fracture (arrows at the symphysis in a posteroinferior location.

Fig. 2.

34 year-old woman with fracture and levator ani tear after “high risk” vaginal delivery. (A) Axial proton density MR image and (B) proton density-weighted image with fat saturation show focal intense left inferior pubic ramus bone marrow edema (B) and cortical fracture (white arrows).

Pubic Bone Marrow Edema

Table 2 illustrates that bone marrow edema was present in 47 of the 77 women, with bilateral edema in 31 and unilateral edema in the remaining 16 women (8 on right side only and 8 on left side only). There was no statistically significant difference in presence of bone marrow edema by risk groups (χ² (3) = 2.86, p = 0.41). There was also no statistically significant difference between the presence or absence of bone marrow edema and the number of days between childbirth and imaging (44.4 ± 17.8 vs. 43.1 ± 14.6 days respectively). While the numbers are small, there does appear to be a possible pattern of mild anterior pubic bone edema seen only with the cesarean section women. This was the only pattern seen with the elective cesarean women and was not seen in the vaginal delivery groups.

Pubic Symphysis Results

The height of the pubic symphysis capsule had a mean ± sd of 2.14 (1.5) mm. As illustrated in Table 2, the differences among risk groups did not differ significantly (F (3, 76) = .658, p = 0.58). The width of the pubic symphysis also did not differ significantly across groups (mean± sd of 3.10 ± 1.3 mm; F (3,76) = 2.34, p=0.08). There was one exceptional case with a pubic symphysis width of 12 mm and extensive bone marrow edema (Figure 3). In this case, the length of second stage was 7 hours 39 minutes before delivery by cesarean section. Another case of pubic symphysis injury involved distension of the superior joint capsule to 7 mm with rupture of the capsule and fluid dissecting along the superior pubic ramus into the pelvis adjacent to the levator ani muscles (Figure 4). She had bilateral bone marrow edema and bilateral levator ani muscle tears representing a Low Risk vaginal delivery complicated by operative postpartum management for hemorrhage.

Fig. 3.

20-year-old woman with pubic symphysis abnormalities delivered by cesarean section after 7 hours and 39 minutes of labor. Coronal proton density-weighted MR image with fat-saturation shows abnormal widening, torn fibrocartilaginous disc, and fluid in symphysis (arrow).

Fig. 4.

39 year-old woman with pubic symphysis capsule rupture and bilateral levator ani tear after “low risk” vaginal delivery. Coronal proton density-weighted with fat saturation images (A) at mid symphysis, (B) posterior to the symphysis, and (C) sagittal plane right para-symphysis image show in (A) and (C) distension of the superior and inferior joint capsule with rupture (black arrow) along the right side. Fluid tracks into both levator ani muscles (B), which are partially torn (white arrows).

Levator Ani Tears

Among the study’s 77 participants, 54 women had no tear and 23 had at least one levator ani tear (Table 2). Of these 23 women, 14 had bilateral tears. The distribution of women with tears varied by delivery groups (Table 2) with a significantly higher proportion of women (20 of 45) occurring in the High Risk group compared to 3 of 32 in the Low Risk groups (χ² (3) = 11.57, p = 0.01). All the levator ani muscle tears involved at least the anterior fibers of the muscle insertion along the posterior surface of the pubic bones and were incomplete tears in that at least some muscle volume was detectable although the increased signal from associated muscle edema and surrounding tissue injury limited evaluation in some cases (Figure 5).

Fig 5.

39-year-old woman 7 weeks after “high risk” vacuum assisted vaginal delivery. (A) Axial proton density MR image and (B) axial proton density-weighted image with fat saturation demonstrating left levator ani muscle tear with fluid within the torn muscle (arrow).

Injury Combinations

To explore the complexity of osseous and soft tissue findings, we reviewed the images for various types of injury combinations. Pubic bone fractures never occurred in isolation; they were always accompanied by pubic bone marrow edema. In 9 of the 21women with a fracture, levator ani muscle tear was also present. Nineteen women showed an isolated finding of bone marrow edema and 8 women showed an isolated finding of levator ani tear. Potential clinically significant injuries to the pubic symphysis included widening with bone marrow edema (Figure 3) and capsular rupture (Figure 4). Of the 77 women in the study, 22 women (29%) had no recorded MR findings, i.e. no areas of bone marrow edema, no fractures, no levator ani tears and no abnormal widening or capsular distension of the pubic symphysis. The distribution of women without MR findings varied by group (Table 2) and were significantly different in the distribution of findings comparing the High Risk women, 8 of 45 (18%), to the combined Low Risk women, 14 of 32 (44%) (χ² (1) = 6.2, p = .01).

Discussion

This study’s investigation with fluid sensitive MR sequences identified patterns of injuries among primiparous women. Despite the relative novelty of this approach in obstetrics research, interpreting the meaning of our results can be guided by considering both mechanisms associated with labor and delivery as well as the growing body of such MR studies of the pubic region evaluating sports-related trauma^(16-19).

While MR- detected pubic symphysis injuries occurred in all groups, they were more frequently associated with labor and delivery risk factors for pelvic floor muscle injury. Bone marrow edema, also called “bone marrow contusion” and “bone bruising,” indicates non-specific stress injury within bone. Fractures occur when stress exceeds trabecular (subcortical) or cortical strength. Conventional radiography and MR without fluid sensitive sequences do not image bone marrow edema and subcortical fractures. Bone marrow edema was present in all delivery groups. Fractures were more frequent among High Risk women and in women delivered by cesarean section after prolonged labor, suggesting labor plays an important role.

The pubic symphysis functions as the central anterior attachment for abdominal wall, thigh, and pelvic floor (levator ani) musculature, spreading forces generated in the torso and lower extremities across the pelvis. The joint capsule is supported anteriorly, inferiorly and superiorly by an aponeurosis blending abdominis rectus and adductor tendons, the pubic arcuate ligament and the superior pubic ligament^(16-17). Hyaline cartilage covers each articular surface that abut a central fibrocartilaginous disc. This reinforced complex makes the pubic symphysis a strong anchor but susceptible to distraction⁽¹⁸⁾.

Stress-related bone injuries present clinically as localized pain exacerbated by activity and relieved by rest. Clinical findings can be non-specific with insidious, diffuse, or radiating pain^(16-21). Treatment emphasizes analgesic medication and, in athletes with chronic pain and/or pubic bone marrow edema or stress related fractures, up to 3 months of rest without athletic weight bearing activities^(15-16,22). Most stress injuries clinically heal in 6-8 weeks although the pubic rami may need 2 to 5 months⁽¹³⁾. Imaging and clinical findings of pelvic stress fractures usually resolve in 6 months⁽¹³⁾. In a MR study of consecutive conscripts with pelvic stress pain, 137 of 340 had bone injuries⁽²⁰⁾. A similar study of 100 consecutive soccer players with pubic/groin pain, 91 had pubic symphysis bone edema without any in the matched controls⁽¹⁶⁾. Chronic pain can be associated with these pubic findings^(15-16,20). Although bone marrow edema is generally considered benign, cartilage loss seen at follow-up suggests joint line bone marrow edema may indicate long-term cartilage damage pre-disposing to early osteoarthrosis^(23-24).

The pubic symphysis undergoes acute distractive stress during childbirth. Pubic diastasis is a known complication especially in women without surgical delivery for difficult vaginal births. Hormonally induced ligamentous laxity during pregnancy is stated to permit increased joint motion. A CT study within 24 hours of uncomplicated vaginal delivery compared with controls showed widening⁽²⁵⁾. In our study, such laxity may have resolved by the time of our measurements. In a study comparing nulliparous controls to symptomatic and asymptomatic women within 2 weeks of vaginal or cesarean delivery, para-symphyseal bone marrow edema was documented by MR only in the post-partum women (13 of 19)⁽²⁶⁾. In a study with 56 vaginally delivered women, 35 symptomatic and 21 asymptomatic, para-symphyseal bone marrow edema was noted in 86% and 76% of the women respectively at two weeks⁽²⁷⁾. In both studies, the width of the symphysis was significantly larger 2 weeks postpartum than nulliparous women⁽²⁶⁾ or controls⁽²⁷⁾. No fracture information was provided in either study. The pubic symphysis was not evaluated in postpartum MR studies including cesarean-section with fluid sensitive sequences⁽²⁸⁾ and two without fluid sensitive sequences including serial examinations⁽²⁹⁾ and cesarean delivery⁽³⁰⁾.

Our findings of pubic fractures and capsular ruptures extend the previous MR observations of bone marrow edema^(26-27). As most injuries were along the posterior and inferior para-symphyseal region, distractive forces may be most pronounced in this region. The levator ani muscle’s pubic portion inserts here and is known to undergo the most stress and lengthening during vaginal delivery⁽³¹⁾. As all of our delivery groups demonstrated pubic bone marrow edema, we speculate that distractive forces can occur without passage of the infant head beneath the joint. The mild anterior edema seen in elective cesarean section women may be secondary to anterior abdominal musculature alterations from cesarean surgery or post-operative biomechanics. However, our High Risk women had the most injuries, including combinations of fractures and muscle tears. Fewer women in our High Risk group, compared to our Low Risk group, had no MR findings; this indicates the risk group classification criteria also identified risk for pubic symphysis injury.

A limitation to our study was not systematically documenting pain data or symptoms associated with pubic trauma. Changes occurring before childbirth such as possible symphyseal separation or resolving before imaging were not detected. The number of women did provide sufficient power to detect injury patterns across the four risk groups but was too small to estimate overall prevalence of fractures or symphyseal ruptures.

This study demonstrates our capability to evaluate women for pubic symphysis osseous and soft tissue injury using fluid sensitive MR sequences. Criteria used to separate primiparous women into High Risk and Low Risk groups for pelvic floor muscle injury also may predict vulnerability to pubic symphysis injury. Evaluation of the relative contribution of soft tissue injury verse bony injury as sources of post-traumatic pain is difficult at any location. In post-partum women, multivariate analysis of a larger study population with clear symptom definition would be needed to clarify the relationship of clinical findings with pubic changes seen on MR. Antepartum ultrasound could evaluate pubic symphyseal widening and capsular injury. These imaging modalities may be used in the presence of suspected muscle and bone injury⁽³²⁾ and for pubic symphysis associated dysfunction syndromes. Chronic pubic pain and osteoarthrosis may result from childbirth related bone and joint injuries. Selective use of MR to evaluate women at risk may offer the ability to refine and individualize clinical decision-making to enhance prevention and recovery strategies.