Clinical, ultrasonographic, and functional outcomes after obstetric anal sphincter injury primary repair: A single‐center experience

To present clinical and instrumental sequelae after obstetric anal sphincter injuries (OASIS), evaluating correlations between intrapartum severity of lesions, postpartum symptoms, and sonographic and manometric findings; outcomes during subsequent deliveries were also evaluated.


| INTRODUC TI ON
Obstetric trauma is the leading cause of anal incontinence in women.
Rates of obstetric anal sphincter injuries (OASIS) range widely from 0.6% to 8% among different populations depending on the quality of obstetrical care, cesarean and instrumental delivery rates, and detection capability. 1,2Currently, we are facing a widespread trend towards a rise in OASIS rates. 3Risk factors include ethnicity, obesity, birth weight over 4 kg, persistent occiput posterior position, nulliparity, induction of labor, prolonged second stage of labor, shoulder dystocia, and instrumental delivery, as well as previous anal sphincter injury. 1,4,5OASIS are associated with short-and long-term morbidity, which can have physical and psychological effects, and seriously affect quality of life.In many cases, anal incontinence symptoms are transient and spontaneously tend to regress in the first days/weeks after delivery, and 1 year after OASIS primary repair most women are asymptomatic. 6However, almost 40% of women will suffer from early anal incontinence after OASIS. 6,7Moreover, a certain proportion of patients will develop symptoms with aging and menopauserelated estrogen deprivation.As a consequence, identifying those women at greater risk of maintaining or developing symptoms of anal incontinence would allow the development of cost-effective programs of secondary prevention.In addition, as most of the women who sustained OASIS are primiparous, it is of the utmost importance to stratify the risk of OASIS recurrence and define the mode of delivery in the subsequent pregnancy. 4Interestingly, previous experiences have shown that the relationship between the severity of the clinical degree of OASIS and symptoms of anal incontinence is inconstant. 8,9As a consequence, proper OASIS assessment may benefit from evaluation using instrumental tools.These mainly involve endoanal ultrasonography (EAUS) and anorectal manometry (ARM). 10,11However, correlations of instrumental findings between these diagnostic tools are unclear, and evidence is even more limited when considering women in the postpartum period. 12 aimed to present clinical and instrumental sequelae after OASIS in our institution, evaluating correlations between intrapartum severity of lesions, postpartum symptoms, and sonographic and manometric findings.As a secondary aim, we wanted to evaluate outcomes during subsequent deliveries.

| MATERIAL S AND ME THODS
We conducted a retrospective study, evaluating all women referred to the maternity ward of Monza and Brianza Child and Mother Foundation (Monza, Italy) between 2015 and 2020.Specifically, we analyzed all consecutive women who sustained an OASIS in the period of interest.In our hospital, we follow a protocol for the treatment, management, and future counseling of women with OASIS.In particular, at discharge from the maternity department, we routinely schedule (1) a 1-month follow-up visit in the pelvic floor outpatients to evaluate symptoms of anal incontinence and instruct patients in At the time of delivery, anal sphincter tears were classifiedaccording to RCOG-into IIIA, involving less than 50% of the external anal sphincter (EAS) thickness; IIIB, involving more than 50% of the EAS thickness; IIIC, involving the internal anal sphincter (IAS); and IV degree in case of additional involvement of the anorectal mucosa. 7Primary OASIS repair and postoperative management were performed following RCOG recommendations.During the urogynecologic follow-up visit, the clinical outcomes were evaluated in terms of anal incontinence and fecal urgency through the validated St. Mark's score filled out by the patient. 13This is a validated questionnaire based on the type and frequency of anal incontinence (gas, liquid, or solid) and its impact on daily life, including the need to wear a pad or plug, the use of constipating medication, and the lack of ability to defer defecation for 15 min.Each type of anal incontinence and its impact on daily life were scored from 0 (never) to 4 (daily); the need to wear protection and the use of medication were scored 0 (no) or 2 (yes), and the lack of ability to defer defecation was scored 0 (no) or 4 (yes).This corresponds to a total score ranging from 0 (complete continence) to 24 (complete incontinence).For analysis, three severity subgroups were formed-as previously-based on St. Mark's incontinence score: 0-4 (mild), 5-8 (moderate), and more than 8 (severe). 14Anorectal manometry was performed in other medical centers.The traditional methods for ARM involves multi-channel water-perfused catheters, which take the average pressure at multiple intervals.In our study, equipment with eight-channel catheters was used.Pressures are recorded after the catheter has been placed into the anorectum for a few minutes.Anal canal pressures are recorded while the patient is relaxed (resting pressures) and with the sphincter closed (squeeze pressures).The maximum resting pressure (MRP), the maximum squeeze pressure (MSP), and the endurance of sphincter contraction (ESC) were recorded, and the squeeze pressure increment (SPI) was calculated as the increase in pressure above the resting pressure during voluntary squeeze effort.We considered values of MRP less than 40 mm Hg as parameters of deficit of basal sphincter tone, and values of MSP greater than 100 mm Hg, SPI less than 20 mm Hg, and ESC less than 20 s as parameters of deficit of voluntary contraction. 15l 3D-EAUS were carried out by an expert urogynecologist (MF) employing a Flex Focus 400 (BK Medical) equipped with an anorectal 3D 2052 probe.This is a mechanical multifrequency transducer with a built-in 3D mover to provide a 360° field of view of the anal canal.All patients were examined in gynecologic positions.The proximal, middle, and distal portions of the anal canal in the three orthogonal planes were evaluated, obtaining the whole volume of the sphincter complex in a single three-dimensional image.The integrity of the IAS and EAS was then assessed.The normal sonographic aspect of the IAS is a hypoechogenic ring in the proximal and middle anal canal, with a thickness of 2-3 mm, and the EAS appears as a hyperechoic ring surrounding the IAS in the middle anal canal and nearby anal mucosa in the distal portion of the anal canal, having a thickness between 7 and 9 mm (Figure 1).In contrast, obstetric defects are seen as irregularities in the echogenicity or as thinning of the IAS and the EAS, usually involving the anterior (obstetric) portion of the sphincter complex.A significant defect was defined as any full-thickness deficit in IAS and/or EAS (Figure 2) greater than a 30° angle in circumferential extension and more than 50% in longitudinal extension. 16Starck sonographic score was calculated to assess the severity of the damage of both EAS and IAS. 17 According to this score, the length (no, ≤1/2, >1/2, whole canal), the depth (no, partially, totally thinned), and the size (no, ≤90°, 91-180°, >180°) of ultrasound alterations in the two rings were measured.
The counseling for delivery mode definition in case of subsequent pregnancy was performed according to RCOG guidelines. 4In particular, symptomatic women and/or patients who were found to have a sonographic defect of the external sphincter of more than 30° and/or a maximum SPI of less than 20 mm Hg were considered to have a substantial compromise of anal sphincter function and were offered the possibility to undergo an elective cesarean section.
Data were extracted from dedicated database software for antenatal care, intrapartum monitoring, and postpartum care and collected by an assigned physician at every patient access and periodically reviewed by a senior consultant.The data used were already available for the analysis for all patients as part of the clinical report of the Obstetric Department of Monza and Brianza Child and Mother Foundation.Given the retrospective nature of the study, the management of each patient was not modified by the study, so it was considered exempt from institutional review board approval.Statistical analysis was performed using JMP version 9 (SAS).
Data were summarized using descriptive statistics and reported as mean ± standard deviation for continuous variables and as absolute frequency (relative frequency) for non-continuous ones.Student ttest was performed to evaluate differences in questionnaire scores after the procedure compared with baseline.A P value less than 0.05 was considered statistically significant.

| RE SULTS
During the study period, 15 200 women gave birth at the maternity department of San Gerardo Hospital, of whom 12 267 had a vaginal birth.In total, 146 women (1.2%) sustained an anal sphincter injury (Figure 3).Population characteristics, pregnancy, labor, and delivery are shown in Table 1.Seventy-six (52.1%) patients sustained a IIIA, 48 (32.9%) a IIIB, 17 (11.6%)a IIIC, and 5 (3.4%) a IV degree obstetric tear.Despite scheduled appointments, 39 (26.7%) patients did not perform any kind of clinical evaluation in our hospital and were lost at follow up (Figure 1).Of the remaining 107 patients, 15 did not undergo either 3D-EAUS or ARM.
Eighteen patients underwent only 3D-EAUS, 18 only underwent ARM, and 56 women completed the follow up with both instrumental evaluations.After primary repair, the following complications were observed.Two women showed early development of a small (<1 cm) vaginorectal fistula, probably as the result of unintended rectal mucosa transfixion with sutures during the primary repair (one IIIA and one IIIC).Both were successfully managed with seton treatment.Two women experienced dyspareunia after primary repair.One of them required a surgical procedure of colpoperineoplasty, and the other was successfully treated through the incision of a stenosing vulvar scar in the outpatient setting.Instrumental outcomes are reported in Table 3. ARM findings were completely normal in 21 (28.4%)patients.MRP was abnormal in 11 (14.7%)women, MSP in 41 (55.8%) women, and ESC in 45 (60.4%) women.Based on EAUS findings, EAS and IAS had a completely regular sonographic aspect in 35 (47.3%) and 55 (74.3%) patients, respectively.The mean Starck score was 3.9 ± 4.4.In most cases, alterations were mild, and in only 14 (18.9%)patients did it reach 8 points or more.
The severity of OASIS-in terms of clinical involvement of the IAS-was associated with worse morphologic findings on EAUS in terms of total Starck score (P < 0.001).In contrast, we did not find an impact of the severity of obstetric lesion as diagnosed in the delivery room on symptom presence and severity, and ARM out-

| DISCUSS ION
Few papers are available about clinical, sonographic, and functional outcomes in the postpartum population after OASIS.Our study aimed to evaluate clinical and instrumental sequelae after OASIS, evaluating correlations between intrapartum severity of the lesion, postpartum symptoms, and sonographic and manometric findings.
Moreover, we wanted to evaluate outcomes during subsequent pregnancies in terms of mode of delivery and perineal outcomes.
We found that primary repair is associated with satisfying outcomes, in terms of minimal rate of complications (3.7%) and high rate of asymptomatic patients (78.5%).When symptoms were reported, they were usually mild, whereas severe symptoms were reported only by 1.9% of patients and associated with the missed diagnosisand repair-of an IAS injury.Instrumental tools were more likely to show some kind of alteration compared with symptoms, with ARM as the diagnostic tool demonstrating the highest rate of abnormal findings, up to 71.6% of patients.The clinical involvement of the IAS in the damage experienced during childbirth was poorly related to outcomes, being associated only with Starck score but not with the presence or severity of symptoms and ARM findings.In contrast, the presence and severity of symptoms showed a strong correlation with instrumental outcomes, in terms of MRP, SPI, Starck EAS, IAS, and total scores, and circumferential extension of defect for both EAS and IAS.We also observed a direct association with all manometric parameters and sonographic findings, with the exception of SPI, the abnormalities of which were not predictable based on EAUS results.

F I G U R E 3 Population selection flowchart.
Overall, outcomes after primary repair were satisfactory, with 78.5% of patients being completely asymptomatic.[20] Complications were limited, and dyspareunia was reported by only 1.9% of women, which was significantly lower than previous reports. 18,21However, we also recorded a 1.9% rate of early rectovaginal fistula, probably as the result of unintended rectal mucosa transfixion with sutures during the primary repair of IIIA and one IIIC injuries.Although obstetrical etiology still represents a major cause of fistula, even in high-resource settings, this type of complication may at least in part be prevented with adequate hands-on courses aimed to improve primary repair technique skills. 22 our series, the severity of the tear in terms of involvement of the IAS was not associated with the severity of symptoms.Although it might seem logical to assume that an increased grade of tear should be associated with increased severity of anal incontinence, previous reports showed controversial data. 8,9,23Underestimation of the severity of damage, such as missing an IAS defect, and quality of the primary repair represent major confounding factors when evaluating the relationship between the degree of OASIS and the presence/severity of symptoms.These data, in our opinion, stress once again the importance of improving the training of caregivers through courses and hands-on experience to increase proper identification and repair of OASIS, whose positive effects have been previously demonstrated. 24strumental data demonstrated excellent correlation with symptoms, confirming their crucial role in diagnostic assessment after OASIS.Many patients with endosonographic sphincter defects do not suffer from anal incontinence symptoms.This means that the anal sphincter complex has a certain grade of functional reserve, and that other structures (e.g., the puborectalis muscles) may play a compensatory role.However, all considered 3D-EAUS parameters were associated with both symptoms and ARM alteration, including the absolute circumferential extension of EAS and IAS defect and the presence of significant sphincter damage.Among all considered ultrasound parameters, the Starck score showed the best diagnostic performances, being constantly associated with symptom presence   As a consequence, 3D-EAUS findings can be used to identify asymptomatic women in whom PFMT should be considered as a tool to improve muscle function and prevent future anal incontinence.
In addition, ARM findings-in terms of MRP and SPI-were found to be related to symptom presence and severity.MRP represents the tonic activity of the sphincter complex, arising mostly (75%-85%) from the IAS fibers, but also from the EAS.As a consequence, MRP impairment may occur with both IAS and EAS obstetric trauma.Our data confirmed that the extension of sonographic defects of both IAS and EAS-evaluated through Starck score-were directly related to abnormal MRP.We also demonstrated an association between sonographic findings-in terms of Starck score-and functional parameters MSP and ESC.Previous correlation studies have reported controversial associations between EAUS and ARM. 12,15,28,29A study by Nielsen et al. 12 failed to demonstrate any correlation between endosonographic findings and anal canal pressures.Similarly, Sultan et al. 28 did not find any relationship between the manometric resting or squeeze pressures in the anal canal, and the internal or external sphincter thickness.In contrast, Titi et al. 29 reported that certain EAUS parameters-such as the presence of an EAS defect and its length, EAS maximum thickness, IAS ring quality-can be predictive of anal sphincter function at ARM.Similarly, Parangama et al. 15 in a cohort of 30 patients demonstrated that patients with documented internal sphincter defects have significantly reduced MRP.The authors concluded that both manometry and ultrasound should be offered to patients with a clinical history suggesting anal sphincter pathology.We did not find correlations between SPI and any of the considered sonographic parameters.As previously established, the SPI was calculated as the increase in pressure above the resting pressure during voluntary squeeze effort. 30As squeeze pressure represents the contribution from muscles that are under volitional control, it is influenced by both EAS and puborectalis muscles, which surround the distal and proximal parts of the anal canal, respectively.In particular, a major role of the puborectalis muscle in conditioning a significant increase in the cranial part of the anal canal has been previously demonstrated. 31,32Consequently, it may be reasonable to assume that relevant damage to the puborectalis muscles, which cannot be reliably investigated on 3D-EAUS, may affect SPI even in the absence of a significant EAS sonographic defect.
This may explain why in our series SPI was significantly associated with symptom presence and severity, without being related to sonographic findings, meaning that SPI may represent an independent predictor of anal sphincter impairment.This functional parameter may be particularly interesting when performing counseling about the mode of delivery in pregnancy after OASIS.RCOG guidelines state that after OASIS, if the woman is symptomatic or shows abnormally low anorectal manometric pressures and/or endoanal ultrasonographic defects, an elective cesarean section may be considered. 4sed on this, Scheer et al. 30 have developed an operative protocol for management that considers EAS sonographic defects and SPI as the functional parameter for counseling.Specifically, their protocol involves offering elective cesarean section to women with an EAS defect greater than 30° and/or a maximum squeeze pressure increment of less than 20 mm Hg.In conclusion, the majority of patients (78.5%) were asymptomatic after primary OASIS repair, and severe symptoms were exceptional (1.9%).ARM and 3D-EAUS were more likely to show some kind of alteration compared with symptoms.MRP, SPI, Starck score for EAS and IAS, and total scores, and circumferential extension of defect for both EAS and IAS were directly associated with symptom presence and severity.Moreover, there was a significant correlation between ARM and 3D-EAUS findings-being functional parameter abnormalities related to the extension of sonographic defects-with the exception of SPI, which may represent an independent predictor of anal sphincter impairment.
pelvic floor rehabilitation; (2) a 6-month anorectal manometry; (3) a 6-month three-dimensional (3D) EAUS; and (4) an outpatient evaluation during the subsequent pregnancy to counsel women about the mode of delivery, based on symptoms and instrumental findings according to Royal College of Obstetricians and Gynecologists (RCOG) guidelines.
Symptom presence and severity according to St. Mark's incontinence score are reported in Table2.The majority of women (84 patients, 78.5%) did not report any symptoms.The remaining 23 (21.5%) patients reported at least one anal incontinence-related symptom, with incontinence of gas being the most reported disorder; incontinence of solid stools was reported by only two (1.9%) patients.Symptomatic patients had a mean score of 3.2 ± 3.0 F I G U R E 1 Optimal sonographic aspect of middle anal canal after obstetric anal sphincter injury repair.F I G U R E 2 Sonographic aspect of middle anal canal following missed diagnosis of obstetric anal sphincter injury in the delivery room: major deficits of both external and internal anal sphincters are visualized from 8 o'clock to 1 o'clock.points according to St. Mark's scale, and only two patients reported severe symptoms (score > 8).Both of them were diagnosed as IIIB in the delivery room, but instrumental findings suggested a misdiagnosed involvement of the IAS, with severe alteration of MRP and sonographic extensive defect of the IAS.
comes.Correlations between symptoms and instrumental findings are shown in Table4.The functional evaluation demonstrated that both abnormal MRP and SPI were associated with the presence and severity of anal incontinence symptoms.Based on sonographic findings, a significant defect of IAS, the circumferential extension of the IAS defect, and the total IAS Starck score were related to the presence of symptoms.Moreover, the EAS Starck score and the extension of EAS circumferential extension of the defect were significantly associated with the presence of symptoms, as well as cumulative IAS and EAS score at Starck classification.In addition, IAS, EAS, and total Starck scores, IAS and EAS circumferential damage extension, and the presence of a significant IAS defect were directly related to symptom severity.The relationship between ARM and ultrasound findings is shown in Table5.All considered sonographic alterations were directly associated with ARM abnormalities.In particular, Starck IAS, EAS, and total score and the extension of EAS circumferential damage were predictors of abnormal MRP, MSP, and ESC, but none of them was related to the alteration in SPI.Thirty-two women in the period of interest became pregnant again and were counseled for the delivery mode according to RCOG recommendations, based on symptoms and available instrumental data.After counseling, 8 (25%) patients underwent elective cesarean section at c.39 weeks of pregnancy.The remaining 24 (75%) were admitted for vaginal delivery and successfully had a natural birth.Following subsequent vaginal delivery, 15 patients had no perineal tear or a less than second-degree tear according to RCOG classification; the remaining nine patients had a paramedian episiotomy for obstetric indication (no "prophylactic" episiotomies were performed for previous OASIS).Notably, no patients suffered from a recurrence of OASIS.
and severity and with abnormality in MRP, MSP, and ESC evaluated with ARM.This may be because in the Starck classification the extension of the defect is considered in three dimensions, focusing also on depth and length of the defect, instead of considering only degrees of injuries.Moreover, the total score potentially allows comparisons of the defects over time and between different centers, making the Starck classification a valuable diagnostic tool.In addition, the Starck score may indicate appropriate treatments-such as helping to define asymptomatic women in TA B L E 1 Population characteristics, pregnancy, labor, and delivery.a

TA B L E 2 a
Symptoms according to St. Mark's incontinence score (107 patients).a Data are presented as mean ± standard deviation.whompelvic floor muscle training (PFMT) may be particularly indicated.Anal incontinence may develop several years after OASI, with a risk directly related to the immediate Starck score after childbirth.25 Specifically, Tejedor et al.26 recommended all patientsincluding asymptomatic patients-with an isolated defect on the IAS and/or Starck score less than 8 to start PFMT.This beneficial role has been confirmed by a recent study showing that PFMT started within the first month postpartum may be effective in reducing the medium-term (16-24 months) functional consequences of OASIS.27 OASIS according to RCOG recommendations resulted in excellent results, with 75% of women admitted to vaginal delivery and none suffering from a recurrence of anal sphincter damage.This confirms that endoanal ultrasound and anal manometry findings are of the utmost importance to make proper recommendations of subsequent delivery mode after OASIS.However, women's experience and sexual function may also play a major role in patient preference for the mode of delivery.Strengths of our study include the large number of participants, the multimodal evaluation including quality of life questionnaire, 3D-EAUS and ARM, and the follow up including outcomes of subsequent pregnancies.A limitation is represented by the retrospective design.
Data are presented as mean ± standard deviation or as number (percentage).
TA B L E 3 Instrumental findings for ARM findings (74 patients) and EAUS findings (74 patients) according to Starck classification.aAbbreviations:ARM,anorectalmanometry; EAS, external anal sphincter; EAUS, endoanal ultrasonography; ESC, endurance of sphincter contraction; IAS, internal anal sphincter; MRP, maximum resting pressure; MSP, maximum squeeze pressure; SPI, squeeze pressure increment.aDataare presented as mean ± standard deviation or as number (percentage).TA B L E 4Correlations between symptoms and instrumental findings.a a Correlation between EAUS and ARM findings.a 30Following this protocol, in a cohort of 59 patients with previous OASIS, they reported a 30.5% cesarean delivery (23.7% elective) rate and 6.8% anal sphincter injury recurrence.Moreover, no changes in symptoms, quality of life, and ARM were recorded following subsequent vaginal delivery or cesarean section.These data are consistent with the findings in our study, in which counseling and management of subsequent pregnancies after TA B L E 5Abbreviations: ARM, anorectal manometry; EAS, external anal sphincter; EAUS, endoanal ultrasonography; ESC, endurance of sphincter contraction; IAS, internal anal sphincter; MRP, maximum resting pressure; MSP, maximum squeeze pressure; SPI, squeeze pressure increment.a Data are presented as mean ± standard deviation or as number (percentage).