A rectocele is a herniation of the anterior rectal wall into the vagina, and it is typically observed during defecation.(1) The incidence of rectoceles in the general population is unknown. However, rectoceles have been identified in 20%–80% of women referred to pelvic floor clinics.(2) Although rectoceles have been reported in asymptomatic healthy individuals,(3) they may be most common in women with symptomatic pelvic floor disorders.(4) Rectoceles are believed to be caused by a defect in the rectovaginal septum that leads to bulging of the posterior vaginal wall, which can be identified on examination during the Valsalva maneuver.(5) This abnormality is associated with symptoms of pelvic floor disorders, such as constipation and incon-tinence.(6) Symptomatic rectoceles may manifest in a wide range of clinical presentations, and these symptoms can negatively impact patients’ quality of life.(5) Constipation is reported in 75%–100% of patients with rectoceles, although the referred and operated cases are often biased.(7,8) Additionally, depending on the clinical practice, the incidence of incontinence associated with rectoceles may be as high as one-third of the total number of rectocele cases.(9,10)

Although defecography is considered the gold standard imaging technique for evaluating rectoceles, it is costly and unpleasant for the patient, and it involves the use of radiation.(11) In colorectal units, the use of defecography in the diagnostic workup of women with clinically apparent rectoceles is limited because of resource implications and the inconvenience associated with the technique. Conversely, transperineal ultrasound (TPUS) has been demonstrated to be an acceptable alternative to defecography for evaluating rectoceles due to its safety, convenience, and minimal patient discomfort.(1,12,13) Dietz and Beer-Gabel (1) reported that a descent >10 mm is considered diagnostic of a rectocele on TPUS imaging.

A large rectocele is believed to be associated with a multitude of symptoms, including evacuation difficulty, constipation, rectal pain, and fecal incontinence. However, very few studies have demonstrated an association between the extent of rectoceles and clinical symptoms.(14) We hypothesize that rectoceles are caused by an anatomical defect, and wider defects cause larger protrusions. Hence, a rectocele essentially adds readily distensible volume to the lower rectum; therefore, increased clinical symptom severity may be reflective of a larger rectocele.

As TPUS is simpler and more convenient than other radiological alternatives, it could potentially replace other techniques for diagnosing patients with pelvic floor disorders. In the present study, we investigated the clinical features of symptomatic rectoceles, as measured by TPUS, and evaluated the association between rectocele size and the clinical symptoms of pelvic floor disorders.

This was a retrospective study using data obtained from a pelvic floor center between August 2020 and January 2021. The study was approved by the institutional review board of Seoul Song Do Hospital (No. 2021-03), and informed consent from the patients was waived because of the retrospective nature of the study. All method were implemented in accordance with relevant institutional guidelines and regulations. The decision to utilize a 3D pelvic floor ultrasound (including TPUS) was made by a colorectal surgeon who evaluated the patients’ symptoms in the outpatient clinic. All the ultrasounds were performed by an experienced pelvic floor surgeon. A total of 125 patients with defecation disorders, such as constipation and fecal incontinence, were included. Our assessment involved a medical history examination and questionnaire scores that were used to evaluate constipation and fecal incontinence. The preoperative questionnaire included the Cleveland Clinic Constipation Scoring System (CCCS, Wexner constipation score), Cleveland Clinic Incontinence Score (CCIS, Wexner incontinence score), Fecal Incontinence Severity Index (FISI), and Fecal Incontinence Quality of Life (FIQOL) scale.(15,16) The pelvic organ prolapse quantification (POP-Q) stage was evaluated for all patients in an outpatient clinic. All patients also underwent physiological tests, such as anal manometry, electromyography (EMG), and pudendal nerve terminal motor latency (PNTML).

The examinations were performed using a 3D ultrasound device (Flex Focus, endoprobe model 8838; B-K Medical, Herlev, Denmark) for the diagnosis of pelvic floor disorders and assessment of pelvic floor injuries. The ultrasound was performed with the patient in the semi-lithotomy position, immediately after voiding. The images consisted of transperineal 2D functional images, endovaginal functional 2D images, endovaginal 3D images, and endoanal 3D images. These images were acquired according to the five steps reported by Shobeiri et al.(17) Only cases with complete data and high-quality scans were included. Good-quality ultrasound scans were defined as those that enabled visualization of the levator plate and pubic bone. Images with partially visible levator plates were not used. We measured the levator ani deficiency (LAD) score, levator plate descent angle (LPDA), and minimal levator hiatus (MLH) on the recently acquired 3D pelvic floor ultrasound to evaluate the levator plate injury. The LAD score, introduced by Morgan et al.(18) divides both sides of the pelvic floor muscles into the puboanalis, pubovisceralis, and puborectalis, and assigns scores from 0 to 3 to each area (score range 0 to 18 points). The MLH described by Shobeiri et al.(19) measures the area within the levator ani muscle’s inner perimeter enclosed by the inferior pubic rami and the inferior edge of the symphysis. All pelvic floor disorders were diagnosed based on the American Society of Colon and Rectum Surgeons guidelines.(20)

The size of each rectocele was measured using the transperineal 2D images. The presence of a rectocele was determined based on the maximal Valsalva (Fig. 1). The best of at least three attempts was used, with no attempt to standardize the intra-abdominal pressure. The rectoceles were measured following the methods described by Dietz et al.(21,22) The maximal caudal displacement of the rectal ampulla was determined using the maximal Valsalva.(21) The maximum rectocele depth was determined by assessing the entire Valsalva maneuver and selecting the volume that reflected the rectocele pocket at its deepest, as descending stool may sometimes flatten the diverticulum at the maximal Valsalva.(22) A significant rectocele size of ≥10 mm was clearly associated with symptoms of obstructed defecation.(23) Enterocele is defined as a hernia into the pouch of Douglas or between the rectum and the vagina, and it contains small bowel. A cystocele was diagnosed on ultrasound if any part of the bladder leached ≥10 mm below the symphysis pubis. During the Valsalva maneuver, the anorectal angle does not open and sometimes becomes narrower diagnosed with anusmus. Internal intussuscep-tion was defined as an invagination of the rectal wall into the distal lumen.(20)

The patients were assigned to one of the following three groups based on the rectocele depth: no rectocele (<10 mm), ≥10 mm rectocele, and ≥15 mm rectocele. Dietz et al.(24) suggested that a cut-off of 15 mm may be optimal for the diagnosis of a significant rectocele; hence, a rectocele with a depth ≥15 mm that satisfied the criteria for a significant rectocele was categorized into the largest rectocele group. Based on the studies of Dietz et al.(24), the patients were divided into 3 groups based on the diagnostic criterion for rectocele of 10 mm on 2D transperineal ultrasound and the cut-off value of 15 mm for requiring surgical treatment.

Numbers and percentages were computed for the categorical variables. The continuous variables are expressed as the means and standard deviations. To determine the differences in the patient demographics, physiological test results, ultrasound results, and rectocele clinical symptoms among the three groups, we used an analysis of variance for continuous data and the Student’s t-test for non-continuous data. Statistical significance was set at P < 0.05. All the statistical analyses were performed using SPSS software (version 22.0 for Windows; IBM-SPSS, Inc. Chicago, IL).

A total of 125 women fulfilled the inclusion criteria and were included in the analyses. All the patients underwent 3D pelvic floor ultrasound imaging, which included transperineal 2D imaging. The mean patient age was 64.75 ± 12.14 (range: 27–86) years, and the median parity was 2.46 ± 1.09 (range: 1–7). A total of 43 participants (34.4%) had no rectocele, 50 (40.0%) had a rectocele ≥10 mm in size, and 32 (25.6%) had a rectocele ≥15 mm. There was a significant difference in POP-Q stage between the 3 groups (P = 0.020). There were no statistical differences between the three groups in terms of age (62.12 ± 13.42 vs. 65.40 ± 12.16 vs. 67.28 ± 9.74 years, respectively; P = 0.169), parity (2.35 vs. 2.40 vs. 2.72, respectively; P = 0.303), and type of delivery (P = 0.322) (Table 1). The results of the anal manometry, EMG, and PNTML analyses are summarized in Table 2. There was a significant difference in the maximum tolerable volumes (MTVs) between the no rectocele, ≥10 mm rectocele, and ≥15 mm rectocele groups (129.44 ± 38.47 vs. 140.44 ± 43.51 vs. 158.13 ± 59.21, respectively; P = 0.045), but there were no differences between the three groups for any other anal manometry characteristic. There were no differences among the three groups regarding the EMG and PNTML analyses. The ultrasound results are presented in Table 3. The mean rectocele sizes were 3.87 mm (SD ± 3.97) and 12.28 mm (SD ± 1.51) in the no rectocele and ≥10 mm rectocele groups, respectively. The mean rectocele size in the ≥15 mm rectocele group was 18.97 mm (SD ± 3.92). There was a significant difference in the prevalence of anismus among the no rectocele, ≥10 mm rectocele, and ≥15 mm rectocele groups (44.2% vs. 28.0% vs. 12.5%, respectively; P = 0.011), but there were no differences in the incidences of enteroceles (2.3% vs. 0.0% vs. 6.3%, respectively; P = 0.196), cystoceles (11.6% vs. 18.0% vs. 28.1%, respectively; P = 0.189), and internal intussusception (51.2% vs. 64.0% vs. 78.1%, respectively; P = 0.056). There was no difference in internal anal sphincter injury (14.0% vs. 4.3% vs. 8.3%, respectively; P = 0.184). There were significant differences in the LAD scores (10.07 ± 4.44 vs. 10.56 ± 3.86 vs. 12.84 ± 3.64, respectively; P = 0.010) and MLH (15.63 ± 3.01 vs. 15.83 ± 2.54 vs. 17.89 ± 2.78, respectively; P = 0.001) among the three groups, as measured using the 3D pelvic floor ultrasound images.

The clinical symptoms of the rectoceles are summarized in Table 4. There were no differences among the groups in terms of the subjective symptoms of fecal incontinence, constipation, and urinary incontinence in comparison to the rectocele size. The incontinence and constipation symptom scores increased going from the no rectocele group to the ≥10 mm rectocele group to the ≥15 mm rectocele group, and the quality of life worsened (Fig. 2). Additionally, the values for the CCIS (6.00 ± 4.95 vs. 8.62 ± 5.77 vs. 11.08 ± 5.63, respectively; P = 0.004), FIQOL (13.72 ± 4.19 vs. 13.42 ± 4.35 vs. 10.38 ± 3.88, respectively; P = 0.006), FISI (18.83 ± 17.67 vs. 25.15 ± 17.34 vs. 33.42 ± 15.49, respectively; P = 0.010), and CCCS (7.50 ± 6.26 vs. 8.65 ± 5.31 vs. 13.11 ± 5.90, respectively; P = 0.006) were significantly different.

In this study, we attempted to validate the TPUS method by defining the prevalence of symptomatic rectoceles in women with pelvic floor disorders and correlating ultrasound findings with the symptoms of these disorders. Using TPUS and objective scoring, we showed that constipation and fecal incontinence symptoms were severe in patients diagnosed with rectoceles, and the symptoms in the ≥15 mm rectocele group were more severe than those in the ≥10 mm rectocele group. On the other hand, there was no significant difference between the 3 groups in subjective symptoms. Symptomatic rectoceles were common in the women presenting with symptoms of pelvic floor disorders in this study, as they were found in 65.6% of the enrolled patients. Guzman Rojas et al.(25) used TPUS to detect anatomic abnormalities in patients with pelvic floor dysfunction and found that more than half of the subjects were diagnosed with true rectoceles. Our results are consistent with this true rectocele prevalence. Radiological diagnosis of rectoceles is more common than sonographic findings in symptomatic women.(14,26) In our study, 80% (100/125) of the symptomatic rectoceles were observed using defecography, which was more than the 65.6% (82/125) observed with TPUS. The difference in the diagnostic rates between the two tests may be the result of contrast media retention in the patients with anismus or abnormal EMG results.

Although the true prevalence of incidental asymptomatic rectoceles in the general population is unknown; patients with rectoceles often suffer from constipation and fecal incontinence. Constipation is reported in 75%–100% of patients, although the series of referred and operated cases are also often biased.(27) Moreover, depending upon the clinical practice, the incidence of fecal incontinence associated with rectoceles may be as high as one-third of the total cases.(28) In this study, 72% of the patients visiting the pelvic floor center had fecal incontinence, and 66% had constipation. Of the patients with these symptoms, 67% were diagnosed with symptomatic rectoceles. For the patients diagnosed with symptomatic rectoceles using TPUS, the symptoms of constipation and fecal incontinence were severe, and the quality of life was diminished (Appendix 1). Moreover, the symptoms worsened as the extent of the rectocele increased. In a previous study, a rectocele with a depth of 15 mm was classified as a significant rectocele;(24) therefore, we used this criteria to assign our patients to either a ≥10 mm or a ≥ 15 mm group. The results of this study confirm that the depth of the rectocele is related to the severity of the constipation and fecal incontinence symptoms.

Historically, defecography has been the most accurate method available for investigating rectocele formation.(26) However, the association between the size of the rectocele, as determined by defecography, and the clinical symptoms is controversial. Several studies have demonstrated a significant correlation between rectocele size and the degree of rectocele contrast evacuation during straining.(27,29) Conversely, Crapps(30) found that rectocele size did not correlate with symptoms. Additionally, in a study of 73 patients with rectoceles, Yang et al. (31) did not observe a correlation between symptoms and rectocele size. Unfortunately, there are few studies investigating the relationship between rectocele size and clinical symptoms. In a previous study, an analysis using TPUS revealed that true rectoceles are related to obstructed defecation.(32) In the current study, we found that symptomatic rectoceles detected with TPUS showed a significant relationship with constipation and fecal incontinence. Moreover, the size of the rectocele and the severity of these symptoms are also related.

Rotholtz et al.(33) suggested that manometry could be useful for predicting the presence of significant rectoceles in constipated patients. They confirmed that objective changes in the rectal first sensation, capacity, and compliance were directly related to the presence of a significant rectocele. However, their study included men or nulliparous women and only patients with symptoms of constipation. Symptomatic rectoceles are mostly found in parous women, and constipation and fecal incontinence symptoms are both common in these patients. In our study, the maximal tolerable volume increased with the size of the rectocele, whereas the minimal sensory volume did not change. These findings suggest that anal manometry has a limited role in predicting the presence of symptomatic rectoceles in parous women with PFDs. However, the increase in MTV according to the increase in the size of rectocele seems to be closely related to fecal incontinence. Therefore, further studies are required to confirm the role of significant rectoceles and anal manometry in patients with PFDs.

To date, the etiological mechanism of rectoceles remains unclear. However, based on the results of the current study, the etiology of rectoceles is similar to that of other pelvic floor disorders, where mechanical and connective tissue abnormalities may play a role. Dietz and Steensma(34) suggested that childbirth may play a role in the etiology of rectoceles. The theory that the birth process could enlarge rectovaginal septum defects provides the basis for surgical management in the form of defect-specific repair, and it presupposes traumatic stretching of posterior vaginal wall structures during passage of the fetal head.(35) Recently, various methods of expressing pelvic floor injuries using pelvic floor ultrasound have been developed. In our study, the LAD score, LPDA, and MLH were measured to confirm pelvic floor injuries. Of these measurements, it is noteworthy that the size of the rectocele and the area of the MLH exhibited a positive correlation. One of the reasons that the MLH provides the most sensitive reflection of rectovaginal septum defects is the increase in the MLH width due to the widening of the puborectalis muscle. Further anatomical studies on this aspect should clarify the etiology of rectoceles.

Increases in rectocele size are weakly associated with body mass index (BMI). A weak association between BMI and rectoceles was also noted in the Women’s Health Initiative Hormone Replacement Therapy Clinical Trial in menopausal women.(36) However, another study showed that an increase in the BMI was negatively associated with levator trauma.(37) In this study, half of the 82 patients who were diagnosed with symptomatic rectoceles underwent surgical repairs. The BMI could only be measured, however, in the patients hospitalized for surgery due to the lack of medical records for the patients seen in outpatient clinics. Although only parous women participated in the study, the relationship between rectocele size and the BMI of patients who underwent surgery was sub-analyzed, and no difference was found in the BMI (Appendix 2). Further studies should be conducted to evaluate the association between BMI and various forms of PFDs.

The present study has several strengths. First, to our knowledge, this study is the first to explore the relationship between the extent of rectoceles and the severity of clinical symptoms. Second, a reliable questionnaire was used to assess the patients’ subjective symptoms. Third, symptomatic rectoceles were diagnosed using TPUS rather than through clinical examinations. Fourth, both the ultrasound assessment and data evaluation were performed by individuals who were blinded to the clinical information, thereby avoiding observer bias. However, our study has several limitations. First, this was a retrospective study. Although TPUS is part of the standardized work-up for symptomatic rectoceles at our institution, it was not performed for all patients who visited our hospital, and the relatively small sample size and retrospective nature of the study allowed for the potential introduction of bias. Additionally, the study population consisted solely of East Asians; therefore, the results may not be generalizable to other ethnic groups.

TPUS is a valuable method for the anatomical evaluation of symptomatic rectoceles. Increases in the depth of symptomatic rectoceles, as measured using TPUS, correspond to increases in the severity of clinical symptoms. This retrospective study confirmed that the size of rectoceles, as determined using TPUS, is associated with the severity of constipation and fecal incontinence. However, a study involving a larger cohort is required to further evaluate this association.