A simplified method for evaluating the anatomical axis of the upper two-thirds of the vagina on MRI: A hospital-based cross-sectional study

2.1 MR image analysis

Pelvic MRI scans were conducted in both of our hospitals using a 1.5-T magnet with patients in the supine position. The measurements from the MRI scans were obtained using WebViewer (Greenlander Information Technology, Version 1.0.0.53449, Hangzhou, China) on the sagittal plane, either at the midline or close to the midline. A fat-saturated T2-weighted fast recovery fast spin-echo sequence was utilized, with a repetition time (TR) of 4100–4200 ​ms and an echo time (TE) of 102–109 ​ms. The measurements were performed independently by a gynecologist-obstetrician and a radiologist, both blinded to the study hypothesis.

The sagittal diagram of the pelvis in Fig. 1 illustrates several key anatomical landmarks. Point P corresponds to the most inferior aspect of the pubic symphysis, while point C indicates the most distal edge of the cervix. Point D represents the apex of the tented-up posterior fornix. Two distinct reference lines were employed (Fig. 1A and B). The first line, PS3L, was drawn from point P to the midpoint of the third sacral vertebra. The PCL was the second line from point P to the last coccygeal joint.⁶

A perpendicular line was drawn from the bladder neck (Point B) to the PS3L, and this line was extended into the vaginal area, and the midpoint of the vaginal portion of this line was designated as Point b. Points b, C, and D were expected to represent the distal, middle, and apical regions of the upper vagina, respectively. The perpendicular distances from Points b, C, and D to both reference lines were measured (Fig. 1B).

Furthermore, the angles between the axis of the upper vagina (a line connecting point b to point D) and both the PS3L and PCL, as well as the angle between the PCL and PS3L, were determined (Fig. 1C). All measurements were recorded in centimeters or degrees. If an anatomical landmark was positioned above a reference line or an angle rotated counterclockwise, it was assigned a positive value. Conversely, if an anatomical landmark was situated below a reference line or an angle rotated in the clockwise direction, it was assigned a negative value.

2.2 Statistical analysis

The data analysis was conducted using IBM SPSS Statistics version 23. The normality of distribution for continuous variables was assessed using the Kolmogorov-Smirnov test. Descriptive variables like baseline data and MRI outcomes are presented as percentages (n (%)), median (interquartile range (IQR)), and mean (standard deviation (SD)). Spearman's correlation coefficient was used to demonstrate the potential effect of influencing factors (age, BMI, parity, previous abdominal surgery, and pregnancy) on the position of vaginal points referring to PS3L. To better observe the sample, we conducted a rough age stratification based on menopausal age and parity. And found that cases below the age of 30 were predominantly those who had not completed their reproductive journey, individuals between 30 and 49 years were considered within the reproductive age group, and 50 years was the average age of menopause. So we stratified the patients into three categories:≤ 30 years, between 31 and 49 years, and≥ 50 years.Between-group comparisons of continuous data were performed using a two-independent t-test or a Mann-Whitney U test, while differences in categorical data were evaluated using a chi-square test. A significance level of P ​< ​.05 was considered statistically significant.

3.1 Patient characteristics

A total of 3,545 consecutive patients who underwent pelvic MRIs were included in the initial review. Among these patients, 2,931 individuals met the exclusion criteria, leaving 614 cases for the final analysis (Fig. 2). Of the 614 cases, 333 patients underwent MRI for benign and malignant uterine lesions, 179 for adnexa diseases, 80 for cervix lesions, and 22 for other factors such as vulvar diseases and abdominal wall conditions. The median age of the study sample was 43 years (range: 17–76 years), and the median body mass index was 22.4 ​kg/m² (range: 15.6–38.8 ​kg/m²). Among the patients, 390 (63.5 ​%) had laparoscopic or transabdominal surgery records.

3.2 Assessment of the vaginal axis using PCL and PS3L

The median distances from the distal, middle, and apical points of the upper vagina to the pubococcygeal line (PCL) were 0.4 ​cm [IQR, 0.0–0.7 ​cm], 2.1 ​cm [IQR: 1.7–2.5 ​cm] and 3.1 ​cm [IQR: 2.5–3.7 ​cm], respectively (Table 1). The PCL and the vaginal axis formed an acute angle with a median of 29.0° [IQR: 23.0–34.0°].

The median distances from the distal, middle, and apical points of the upper vagina to the line connecting point P to the midpoint of the third sacral vertebra (PS3L) were -0.5 ​cm [IQR: -0.9–0.0 ​cm], 0.0 ​cm [IQR: -0.4–0.6 ​cm], and -0.2 ​cm [IQR: -0.9–0.0 ​cm], respectively (Table 1). The PS3L was nearly parallel to the vaginal axis, and the median angle between the PS3L and the vaginal axis (PS3L-vaginal angle) was 0.0° [IQR: -4.0–7.0°] (Table 1).

3.3 Vaginal axis in different age subgroups

Correlation analyses for the relative positions of the three studied points concerning PS3L in relation to age, BMI, parity, vaginal delivery, cesarean section, and previous abdominal surgery showed that only the relative position of the b to PS3L exhibited a weak negative correlation with age (r ​= ​-0.49, p ​< ​.001), parity (r ​= ​-0.36, p ​< ​.001), and vaginal delivery (r ​= ​-0.40, p ​< ​.001) (Table 2). Age exhibits a positive correlation with parity and vaginal delivery (p ​< ​.001).

There were 120 patients aged ≤ 30, 282 between 31 and 49, and 212 patients aged ≥ 50. Body mass index (BMI) values were generally low in all three subgroups, with limited differences observed among the groups. In the oldest group, 92.9 ​% (197/212) of patients had experienced vaginal delivery, with 49.1 ​% (104/212) multiparous. Conversely, in the youngest group, 79.2 ​% (95/120) were nulliparous (Table 1).

The vaginal marker points in the oldest group were slightly lower than those in the younger groups, whether referring to the PCL or the PS3L (p ​< ​.001) (Table 1 and Fig. 3). Moreover, both the PCL-vaginal angle and the PS3L-vaginal angle were slightly larger in the oldest group compared to the two younger groups (p ​< ​.001) (Table 1). The vaginal axis in the youngest groups more closely conformed to the PS3L, appearing parallel to it with the marker points nearly aligned along the line, resulting in median PS3L-vaginal angles of zero degrees (Table 1 and Fig. 4).

4.1 Main findings

This hospital-based cross-sectional study focused on the sagittal plane orientation of the upper two-thirds vagina, excluding individuals with prolapse or known anatomical variations. The findings revealed that the axis of the upper vagina was positioned approximately 2–3 ​cm above the levator ani plane represented by the pubococcygeal line (PCL), forming a 29-degree angle with the PCL. Furthermore, the vaginal axis is closely aligned with a line extending from the inferior border of the pubic symphysis to the third sacral vertebra (PS3L). The median distances from the vaginal marker points to the PS3L ranged from −0.5 ​cm to 0.0 ​cm, and the median angle between this line and the vaginal axis was 0.0°. These observations were consistent across all age subgroups, with the younger groups showing greater conformity to the PS3L.

4.2 Interpretations

In this study, the MRI was not done under maximal Valsalva. Although prolapses are assessed during maximal Valsalva, this study is to find a normal vaginal axis, and the pelvic structure is more likely to be in situ under rest. When assessing the position of the vagina, it is crucial to understand its supporting structures. The vagina was categorized into three levels according to DeLancey's classification.² The cardinal and uterosacral ligaments commonly support the upper third of the vagina; the middle third is laterally attached to the arcus tendineus fasciae of the pelvis; the lower-third merges with the urethra and attaches to the pubocervical fascia.² However, recent evidence has shown that the cardinal ligament functions as a mesentery structure comprising vessels and nerves.^8–10 Thus, the uterosacral ligament and pubocervical fascia are the primary ligament-like structures supporting the upper two-thirds vagina. In other words, it fuses at the inferior border of the pubic symphysis anteriorly and ends at the S2-S4 sacral segments posteriorly, with the axis closely approximating the PS3L. However, the measurements still showed variation in these parameters, possibly because the uterosacral ligament is attached anteriorly mostly but not exclusively on the posterior aspect of the cervicovaginal junction, posteriorly broadly to the first three sacral vertebrae and variably to the fourth sacral vertebra.^11,12 In addition, the definitive role of paravaginal support in the middle third of the vagina and its contribution to the development of prolapse are still unknown.^13,14

The study population comprised individuals of varying ages, parity, BMI, and delivery types. Additionally, many participants had undergone abdominal surgical interventions, and some patients had missing surgical records. However, it is essential to note that correlation analyses showed that only the relative position of the bladder neck to PS3L exhibited a weak negative correlation with age, parity, and vaginal delivery. It has been reported that aging, multiparty, and especially previous vaginal delivery are high-risk factors for pelvic organ prolapse.¹⁵ Generally, the three vaginal marker points are well followed by the PS3L in all age subgroups, especially in young nulliparous women. In the multiparous that was over 50 years old but had no POP history and were not diagnosed as POP, the vaginal axis was slightly lower. A moderate degree of prolapse in continent women based on MRI has been reported due to its ability to measure actual pelvic organ descent.^16,17

4.3 Limitations

Our study has several limitations. Ideally, for studying the natural axis of the normal vagina, prospective recruitment of healthy, young, nulliparous women for MRI would be the ideal approach. However, due to limitations in research funding, our study utilized a retrospective study involving MRI cases from hospitals. To address this limitation, we implemented strict exclusion criteria. Through these efforts, we aimed to exclude potential cases with abnormal pelvic organ positions to the best of our ability, ensuring a relatively “normal” sample for analysis. However, based on the current literature review, there is no evidence to suggest that those exclude potential factors would influence the vaginal axis. The upper limits of the measurements are primarily based on our observations during an extensive review of imaging cases. The retrospective design of the study introduces inherent limitations. Secondly, the MRI scans were conducted with patients in the supine position, which may not fully reflect the positions of pelvic organs in the standing position. This positioning limitation should be considered when interpreting the results. Lastly, the demographics of the study population, including the relatively low BMI and the specific ethnic group (Mongolian), may restrict the generalizability of the findings to other populations. Further studies involving a more diverse population and a normal sample are warranted to validate our findings.