Surgical anatomy of the mid-vagina.

AIM: The mid-vagina (MV) represents Level II of the vagina. The surgical anatomy of the MV has not been recently subject to a comprehensive examination and description. MV surgery involving anterior and posterior colporrhaphy represents a key part of surgery for a majority of pelvic organ prolapse (POP).
METHODS: Literature review and surgical observations of many aspects of the MV were performed including MV length and width; MV shape; immediate relationships; histological analysis; anterior and posterior MV prolapse assessment and anterior MV surgical aspects. Unpublished pre- and postoperative quantitative data on 300 women undergoing posterior vaginal compartment repairs are presented.
RESULTS: The MV runs from the lower limit of the vaginal vault (VV) to the hymen. Its length is a mean of 5 cm. Its shape in section overall is a compressed rectangle. Its longitudinal shape is created by its anterior and posterior walls being inverse trapezoid in shape. Histology comprises three layers: (i) mucosa; (ii) muscularis; (iii) adventitia. MV prolapse staging uses pelvic organ prolapse quantification (POP-Q). Anterior MV prolapse can be quantitatively assessed using POP-Q while posterior MV prolapse can be assessed with POP-Q or PR-Q. Around 50% of both cystocele and rectocele are due to VV defects. POP will increase anterior MV width and length. Native tissue anterior colporrhaphy is the current conventional repair with mesh disadvantages outweighing advantages. Posteriorly, Level II (MV) defects are far smaller (mean 1.3 cm) than Level I (mean 6.0 cm) and Level III (mean 2.9 cm).
CONCLUSION: An understanding of the surgical anatomy of the MV can assist anterior and posterior colporrhaphy. In particular, if VV support is employed, the Level II component of a posterior repair should be relatively small.

INTRODUCTION

The middle section of the vagina (mid‐vagina [MV]) is otherwise known as Level II of the vagina (Figure 1). The surgical anatomy of the MV has not been recently subject to a comprehensive examination and description. The MV commences posteriorly at the lower end of the vaginal vault (VV), that is, 2.5 cm down the posterior wall from the vaginal apex in the posterior fornix. , It ends at the posterior aspect of the hymen, the latter included in Level II. The posterior mid‐vaginal length (PMVL) should be the same before and after hysterectomy as the uterus enters and is removed via the anterior fornix, with no significant immediate impact on the posterior compartment.

Figure 1

The mid‐vagina is Level II of the vagina

The mean total vaginal length (TVL—posterior fornix to the posterior aspect of the hymen) has been noted to be 7.5 cm in a series of 300 consecutive women undergoing posterior repair. Thus, the mean PMVL is 5.0 cm (7.5–2.5 cm). The PMVL represents 66.6%, that is, two thirds of the TVL. The total posterior vaginal length (TPVL—vaginal apex to anterior perineum) is a mean of 9.2 cm3 as it includes a mean of 1.8 cm for the length of the posterior vestibule. The MV (Level II) is thus the longest of the three vaginal sections.

Post‐hysterectomy, the anterior vaginal length (AVL) is that between the anterior fornix next to the cervix down to the anterior aspect of the hymen. , Post‐hysterectomy, the AVL is that between the anterior aspect of the vaginal cuff and the anterior aspect of the hymen.

AVL is on average 6.1± 1.3 cm in women with normal support and lengthened in women with a cystocele. As around 1.0–1.5 cm of the upper anterior vagina is involved in the makeup of the anterior fornix, that is, Level I vagina (VV ), the effective anterior mid‐vaginal length (AMVL) would be close to 5.0 cm, similar to the PMVL.

There are different methods of assessing vaginal width. Using magnetic resonance imaging (MRI) studies, vaginal width was noted to be largest in the proximal vagina (3.25 cm), decreasing as it passed through the pelvic diaphragm (2.78 cm), and smallest at the introitus (2.6 cm). Using vinyl polysiloxane casts, vaginal widths were found to be between 4.8 and 6.3 cm. A simple digital intraoperative assessment (lithotomy position) would support the width of the anterior and posterior walls being around two fingers (3.5 cm wide approximately) at the level of the hymen (inferior aspect of Level II). It expands to 2 ½ fingers (4.5 cm wide approximately) in the upper part of the MV. The lateral walls are around 1½ fingers (or 2.5 cm) in an anteroposterior direction (Figure 2).

Figure 2

Compressed rectangular shape of the mid‐vagina with trapezoid shape for the anterior and posterior vaginal walls: (i) supero‐inferior length a mean 5 cm; (ii) superior width (adjacent to Level I—VV) is around 4.5 cm; (iii) inferior width (adjacent to the hymen and Level III) is around 3.5cm. Hymen is not included in the figure

GROSS ANATOMY OF THE MV

The vagina has been referred to as a fibromuscular “tube,” the latter defined as something “long, hollow and cylindrical and here flexible.”  From the measurements above, we see the vagina, not as a “tube,” but rather as a “compressed rectangle” (Figure 2). All walls are compressed, unlike the VV, which can be kept partially open by the presence of the cervix (Figure 1). As noted above, the anterior and posterior walls widen more proximally, similar to an inverse trapezoid (Figure 3), as has been described for the shape of the perineum.

Figure 3

Inverse trapezoid

These clinical findings agree with previous MRI findings , , that the anterior vaginal wall is trapezoid in shape with the narrow part at the pubis. We suggest the posterior wall is also inverse trapezoid in shape, narrow close to the hymen, and gradually widening up to the border with the VV (Figure 3). We support the theory that this proximal widening is due, in part, to the influence of the VV supports, the cardinal and uterosacral ligaments (CL and USL) , and in part to the bulk of the cervix pre‐hysterectomy.

The vagina comes from the Latin for “sheath,” an organ of copulation that receives the penis during sexual intercourse. The MV thus provides a major role in accommodating the penis during copulation. The average length of the erect penis is 13.0–14.0 cm (5.1–5.5 in.) The TVL was reported by Masters and Johnson to be 7–8 cm (identical to the 7.5 cm TVL above) increasing by around 53% to 11–12 cm in the stimulated state.

RELATIONSHIPS OF THE MV (Figures 4, 5, 6)

Anterior vaginal wall: Is in direct contact with the bladder base, bladder neck, and urethra.

Posterior vaginal wall: Is in direct contact with the rectum.

Lateral vaginal walls: Interface with the pelvic floor. Musculofascial extensions from the levator ani support the MV.

Axial view of the pelvis (Visible Human Project, National Library of Medicine, NLM. USA—with permission)

Sagittal view of the pelvis (Visible Human Project, National Library of Medicine, NLM. USA—with permission)

Coronal view of the pelvis (Visible Human Project, National Library of Medicine, NLM. USA—with permission)

HISTOLOGY OF THE MV

The MV is a fibromuscular structure with three layers: (i) mucosa; (ii) muscularis; (iii) adventitia. It is kept moist by mucus produced by the cervical glands. The histology is non‐controversial (Figure 7).

Figure 7

Vaginal histology : “HistologyGuide. com. T. Clark Brelje and Robert L. Sorenson, University of Minnesota, Minneapolis, MN” (with permission from Dr Brelje)

• Mucosa : Contains numerous transverse folds (rugae). Layers are: (a) Stratified squamous nonkeratinized epithelium with estrogen promoting the storage of glycogen in the upper and middle layers; (b) Lamina propria: dense irregular connective tissue rich in collagen and connective tissue. No glands are present.

• Muscularis : Composed of two indistinct layers of smooth muscle: (a) Inner circular layer; (b) Outer much thicker longitudinal layer.

• Adventitia : Inner layer of dense connective tissue and outer layer of loose connective tissue.

The vaginal wall is only a few mm thick, 2.5–3.5 mm in one series where measurement was by ultrasound. One study denied the MV was thinner in the presence of POP though another series shows a thinning from POP Stage I to II, then and an increase in thickness to Stage III. This is however strong. The combined muscularis and adventitia layers can support an additional layer of plicatory sutures during anterior colporrhaphy. This would occur after the insertion of the prevesical plicatory sutures and before vaginal skin closure. This would reduce the tension on the skin closure both intraoperatively and postoperatively. The first author has found it a useful additional surgical step, to increase the anterior colporrhaphy from two to three layers.

MID‐VAGINAL PROLAPSE

The MV is where all aspects of POP are first revealed. The two aspects of POP arising in the MV are cystocele anteriorly and rectocele posteriorly. These MV defects, being closer to the vaginal introitus, naturally are generally more obvious than Level I defects (uterine and VV prolapse). Studies have shown that around half of the cystocele and 55% of rectocele is due to vaginal vault descent. Other clinical studies support these findings. , Level I descent can be more difficult to assess clinically; traction of the Level I areas under anesthetic can be more accurate.

Gross staging of MV prolapse can be obtained clinically using the POP‐Q , (Figure 8)

Figure 8

POP‐Q staging of mid‐vaginal prolapse

Stage 0: No prolapse is demonstrated.

Stage I: Most distal portion of the prolapse is more than 1 cm above the level of the hymen.

Stage II: The most distal portion of the prolapse is situated between 1 cm above the hymen and 1 cm below the hymen.

Stage III: The most distal portion of the prolapse is more than 1 cm beyond the plane of the hymen but everted at least 2 cm less than the total vaginal length.

Stage IV: Complete eversion or eversion at least within 2 cm of the total length of the lower genital tract is demonstrated.

ANTERIOR MID‐VAGINAL POP ASSESSMENT , — CYSTOCELE ,

There is only one schema for quantifying anterior MV prolapse—POP‐Q. , This can be used either clinically or intraoperatively with the limitation that no instrumentation can be used clinically (Figures 9A, B).

Figure 9

(A) (Left)  cystocele seemingly without major uterine or vaginal vault descent ; (B) (Far right)  cystocele largely related to vaginal vault descent (NB cuff scar) ,

• Point Aa: , A point located in the midline of the anterior vaginal wall three 3 cm proximal to the external urethral meatus (shortened definition).

• Point Ba: , A point that represents the most distal (i.e., most dependent) position of any part of the upper anterior vaginal wall from the vaginal cuff or anterior vaginal fornix to Point Aa (shortened definition).

• Point C: , A point that represents either the most distal (i.e., most dependent) edge of the cervix or the leading edge of the vaginal cuff (hysterectomy scar) after total hysterectomy (Figure 10).

Figure 10

POP‐Q Schema : (Anteriorly) Points Aa and Ba as well as Point C are relevant to anterior mid‐vaginal prolapse; (Posteriorly) Points Ap and Bp as well as Point C are relevant to posterior mid‐vaginal prolapse

POSTERIOR MID‐VAGINAL POP ASSESSMENT—RECTOCELE ,

There are two schemas for assessing posterior MV prolapse: (i) POP‐Q ,  and (ii) PR‐Q ,

• Point Ap: , A point located in the midline of the posterior vaginal wall 3 cm proximal to the hymen (shortened definition).

• Point Bp: , A point that represents the most distal (i.e., most dependent) position of any part of the upper posterior vaginal wall from the vaginal cuff or posterior vaginal fornix to Point Ap (shortened definition) (Figures 11A, B).

Figure 11

(A) (Left ) Rectocele seemingly without major uterine or vaginal vault descent; (B) (Right middle ) Rectocele largely related to vaginal vault descent (NB: cuff scar)

PR‐Q assessment (intraoperative) , ,

• MVL (vaginal vault undisplaced) , , : Laxity of the vaginal mucosa (anterior traction) midpoint in the vagina super‐posteriorly and in the midline with the VV held in an undisplaced position (similar to that after VV fixation).

• Recto‐vaginal fascial laxity (RVFL) , , : Laxity of the rectovaginal fascia (anterior traction) midpoint in the vagina super‐posteriorly (mucosa opened) and in the midline with the vaginal vault held in an undisplaced position (Figures 12A, B).

Figure 12

(A) (Left) , , : Mid‐vaginal laxity (vaginal vault undisplaced), (B) (right) , , : Recto‐vaginal fascial laxity

SURGICAL ASPECTS OF ANTERIOR MID‐VAGINAL POP DEFECTS

Anterior colporrhaphy involves the reduction of the bulk of the prolapse (tissue reduction). There are not the convenient ligaments (e.g., sacrospinous ligament—SSL) that can be used for posterior repairs, though anterior approaches incorporating the SSL or USL are being used. These techniques can provide combined Level I and II support. Native tissue surgery involving a multilayer repair is, however, the traditional and still most popular approach. The excision of the redundant anterior vaginal wall is generally considered as part of the tissue reduction process.

Section 8 of Haylen et al. contains a summary of current surgical techniques for repairing anterior MV POP defects. Newer techniques for anterior colporrhaphy are regularly trialed. ,

For around 15 years from 2005, mesh was employed in an attempt to enhance the strength of the anterior colporrhaphy. It is not a feature of this article to explore this aspect except to note two summary conclusions. The first of these is that “polypropylene anterior compartment mesh offers improved objective and subjective outcomes compared with native tissue repair; however, these benefits must be considered in the context of increased morbidity associated with the mesh.” A second conclusion is this, “current evidence does not support the use of a mesh repair compared with a native tissue repair for anterior compartment prolapse owing to increased morbidity.”

As noted above and by Weber and Walters, the vagina has three layers: mucosa, muscularis, and adventitia; there is no vaginal “fascia.” Vaginal support is provided by the underlying levator ani muscles and by lateral connective‐tissue attachments at the arcus tendineus fasciae pelvis or “white line.” Dissection during anterior colporrhaphy splits vaginal muscularis, and repair involves plication of the muscularis and adventitia (not vaginal “fascia”) in the midline. , Paravaginal repair restores the lateral attachments to the pelvic sidewall at the white line.

Cystocele formation increases the AVL (Figure 1). Similarly repair of cystocele repair reduces the AVL by 28% in one series. It also appears to increase the MV width by 28% from 3.2 ± 0.9 cm in those with normal support to 4.1 ± 1.0 cm in those with prolapse.

There appears to be a dearth of studies of the anatomical effects of anterior colporrhaphy using POP‐Q quantitative markers.

A recent review of 40 randomized controlled trial on anterior colporrhaphy showed not only wide differences in each step of the procedure but also in perioperative care, anesthesia, anatomical structures used, and surgeons' experience, all important factors influencing the outcome of the operation. Standardization of surgery is cited as important with exact and reproducible descriptions of the procedure and assessments of surgical performance. We believe this is possible in the posterior compartment, as discussed below.

SURGICAL ASPECTS OF POSTERIOR MID‐VAGINAL DEFECTS

There are more definitive quantitative data on the effects on posterior MV POP defects (rectocele) by posterior native tissue vaginal repair, with and without VV support. Mesh is also currently not recommended in the posterior compartment.

Table 1 outlines the preoperative and postoperative (unpublished) data from a cross‐sectional study of 300 consecutive posterior repairs (PRs), mostly following prior or concomitant hysterectomy and concomitant anterior colporrhaphy. All preoperative POP‐Q, and PR‐Q measurements used in the study were taken immediately before the PR, that is, at the end of any concomitant prolapse surgeries. Postoperative measurements were taken at the end of the PR surgery. At this point, vaginal skin closure prevented measurement of the postop RVFL. There were no patients excluded from a PR.

Table 1

Pre‐op and post‐op results for (i) PR‐Q and (ii) POP‐Q surgical markers

	Pre‐op				Post‐op
Variable	Mean	SD	Min	Max	Mean	SD	Min	Max
PR‐Q posterior prolapse markers (cm)
PG ‐ Level III	2.9	1.0	0.3	6.0	0.0	0.0	0.0	0.0
PVVD (overall)—Level I	6.0	2.0	0.3	15.0	0.2	0.5	0.0	2.0
PVVD (SSC performed)—Level I	6.5	2.0	4.5	15.0	0.06	0.1	0.0	1.0
PVVD (no SSC performed)—Level I	3.5	1.0	0.3	5.5	0.6	0.4	0.0	2.0
MVL, undisplaced—Level II	1.3	0.6	0	3.5	0.2	0.1	0.0	0.7
Recto‐vaginal fascial laxity (RVFL)—Level II	1.1	0.7	0	4.0	n/a		n/a	n/a
POP‐Q posterior prolapse markers (cm)
Point C ‐ Level I	−0.9	2.3	−8.0	8.0	−6.3	1.4	−2.5	−10.0
Point Ap ‐ Level II	1.0	1.4	−3.0	5.0	−2.9	0.3	−1.0	−3.0
Point Bp ‐ Level II	1.0	1.5	−3.0	6.0	−2.9	0.3	−1.0	−3.0
Genital hiatus (GH) ‐ Level III	3.7	0.9	1.5	6.5	2.6	0.7	1.5	4.5

Abbreviations: Min, minimum; Max, maximum; MVL, mid‐vaginal laxity; Pre‐op, preoperative; Post‐op, postoperative; PG, perenial gap; PVVD, posterior vaginal vault descent; SD, standard deviation.

Table 2, using the same group of 300 women in Table 1 (unpublished) shows (i) the PR‐Q defect marker; (ii) the action used in repairing the defect; (iii) the percentage of times the action was used; (iv) the result of that action; (v) equivalent POP‐Q result of that surgical action.

Table 2

PR‐Q markers, guidelines for repair action, percentage of cases repair action used; results of surgical action (PR‐Q); results of surgical action (POP‐Q)

PR‐Q marker	Our guideline for repair action	Percentage of cases action used	Result of action	POP‐Q equivalent level
PG 4	Excise PG 4	100% 4	100% excised 4	30% ↓ GH 4
PVVD 2 , 3	SSC if PVVD > 5 cm	84% SSC (mean preoperative 6.0 cm to mean 0.2 cm postoperative)	SSC performed— 99%↓PVVD. No SSC— 83% ↓PVVD	Point C from mean minus 0.9 to mean minus 6.3
MVL—vault undisplaced	Vaginal skin excision just under half MVL bilaterally	96% (67% of whom requiring excision of under 0.5 cm bilaterally)	85% mean reduction in MVL (1.3 cm preoperative to 0.2 cm postoperative)	Points Ap, Bp from 1.0 to minus mean 2.9
RVFL	Plicatory sutures if RVFL > 0.5 cm	76% (mean preoperative 1.1 cm)	Optimize plication though postoperative RVFL not applicable

Abbreviations: GH, genital hiatus; MVL, mid‐vaginal laxity; PG, perineal gap; RVFL, recto‐vaginal fascial laxity.

The initial conclusion (from Table 1) is that the MV defects (Level II) are the smallest defects, provided the VV is reduced to its undisplaced position. The VV (Level I) defects (posterior vaginal vault descent [PVVD] , —6 cm, Point C—minus 0.9 cm) are the largest defects. The Level III defects (perineal gap [PG] , —2.9 cm, genital hiatus [GH] , —3.7 cm) are the next largest. The Level II defects (MVL—1.3 cm, RVFL—1.1 cm, Ap—1.0 cm, Bp—1.0 cm) are relatively small, again provided the VV descent is reduced. The cut‐off used for the performance of a sacrospinous colpopexy (SSC) for VV support was a PVVD of 5 cm or more.

The second conclusion (from Table 2) is that there are high cure rates (99% reduction in PVVD—6.5–0.06 cm mean) for Level 1 defects in those having VV support using a  SSC. Also from Table 2, there is a 100% cure of the Level III defect by excision of the PG and repair.

Specific to Level II (Tables 1 and 2), the MVL was initially small (mean 1.3 cm) reducing to a mean of 0.2 cm (85% reduction) with under 0.5 cm bilateral skin trimming in 67% of cases. RVFL was again small (mean 1.1 cm), 76% requiring prerectal plicatory sutures.

These results confirm that the “cure” of a “rectocele” does not need to involve extensive dissection or layered repair of a perceived large Level II defect of rectovaginal fascia (septum) , , and accompanying extensive excision of vaginal skin laxity, with the increased risk of dyspareunia. If the Level I and III defects were simultaneously addressed, repair work at Level II involved smaller dissection and relatively limited repair work. Whilst vaginal skin excision was required in 96% of cases, in the majority (67%), it was limited to under 0.5 cm bilaterally, in our view, reducing the risk of dyspareunia.

CONCLUSIONS

The MV represents Level II of the vagina between the VV (Level I) and the vaginal introitus (Level III). The MV runs from the lower limit of the VV to include the hymen. Its length is a mean of 5 cm. Its shape in section overall is a compressed rectangle. Its longitudinal profile is created by its anterior and posterior walls being inverse trapezoid in shape, approximately 3.5 cm wide at the hymen and 4.5 cm at the border with the VV. Lateral walls are approximately 2.5 cm in the anteroposterior direction. Histology comprises three layers: (i) mucosa; (ii) muscularis (iii) adventitia. Immediate relations are pelvic floor musculature and lateral connective tissue to the sidewall of the pelvis, bladder anteriorly, and the rectum posteriorly.

MV surgery involving anterior and posterior colporrhaphy represents a key part of surgery for a majority of POP. MV prolapse staging uses POP‐Q. Anterior MV prolapse can be quantitatively assessed using POP‐Q while posterior MV prolapse can be quantitatively assessed with POP‐Q or PR‐Q. Around 50% of both cystocele and rectocele are due to VV defects.

POP will increase anterior MV width and length. Native tissue anterior colporrhaphy is the current conventional repair with mesh disadvantages outweighing advantages.

Posteriorly, Level II (MV) defects are far smaller (mean 1.3 cm) than Level I (mean 6.0 cm) and Level III (mean 2.9 cm). An understanding of the surgical anatomy can assist anterior and posterior colporrhaphy. In particular, if VV (Level I) support is employed and suitable perineorrhaphy performed (Level III), the Level II component of a posterior repair should be relatively small.

The MV (Level II) here is the same area, a mean of 5 cm in length, as the proposed combined Levels II and III in an earlier description. Level III currently refers to the vaginal introitus, below the hymen, incorporating anterior and posterior vestibules and perineum , (Figure 1).

AUTHOR CONTRIBUTIONS

Bernard T. Haylen, Dzung Vu, Audris Wong, and Sarah Livingstone contributed to the development concept, writing, and review of the paper.

CONFLICT OF INTEREST

The authors declare no conflict of interest.