Fascial and Muscular Abnormalities in Women with Urethral Hypermobility and Anterior Vaginal Wall Prolapse

Fascial and muscular abnormalities in women with urethral hypermobility and anterior vaginal wall prolapse

John O. L. DeLancey, MD Ann Arbor, Mich

OBJECTIVE: Our purpose was to assess the structural integrity of individual elements of the urethral and anterior vaginal wall support system. STUDY DESIGN: Notes were made during retropubic operations for cystourethrocele and stress incontinence in 71 women aged 52 ± 12.4 (SD) years. Vaginal support was assessed with the Baden-Walker system with the following average findings: urethra 1.9 ± 0.6, bladder 1.9 ± 1.0, apex 0.8 ± 1.1, upper posterior wall 0.3 ± 0.8, and rectocele 1.1 ± 0.7. The presence of the following features was noted: paravaginal defect, integrity of the pubic and ischial attachments of the arcus tendineus fascia pelvis (ATFP), appearance of the ATFP on the sidewall, and abnormalities in the pubococcygeal muscle. RESULTS: Paravaginal defects were present in 87.3% on the left and in 88.7% on the right. Detachment of the ATFP from the pubic bone was present in 1.4% (left) and 2.8% (right). The ATFP was detached from the ischial spine in 97.6% (left) and 95.1% (right). Remnants of the ATFP were present on the sidewall in 62% (left) and 63% (right). Of these, 9% extended one fourth the distance to the spine, 21% one half the distance, 3% three fourths the distance, and 17% all the way to the spine. The pubococcygeal muscle was visibly normal in 45% (left) and 39% (right). It showed localized atrophy in 22% (left) and 30% (right) and generalized atrophy in 22.5% (left) 30.0% (right). CONCLUSION: The ATFP usually detaches from the ischial spine, but not from the pubis; slightly less than half of these women have visibly abnormal levator ani muscles. (Am J Obstet Gynecol 2002;187:93-8.)

Key words: Urethral hypermobility, vaginal wall prolapse, fascial abnormalities, muscular abnormalities, paravaginal defect

Structural failure of the urethral support system is one of could separate from an intact ATFP. Alternatively, the lev- the important factors involved in the cause of stress urinary ator ani muscles that contract during cough could be incontinence. Richardson et al1 have observed that loss of damaged and their impulse lost. At present, the fre- anterior wall descent occurred because of lateral detach- quency with which any of these defects occurs in women ment rather than midline stretching in women with stress with stress incontinence is not known. incontinence and urethral hypermobility. They observed a The following study was undertaken to determine the paravaginal defect where the edge of the pubocervical fas- occurrence of these anatomic abnormalities visible in cia is displaced medially from the pelvic sidewall. urethral support system anatomy among women with ure- The urethral support system has several compo- thral hypermobility. nents.2,3 Damage to either the connective tissue elements or the muscles of this apparatus could result in increased Material and methods urethral mobility. For example, the arcus tendineus fascia Study population. Between June 1, 1996, and February pelvis (ATFP) could become detached from either the 1, 2001, notes were made during retropubic operations in pubis or the ischial spine and the pubocervical fascia 68 women operated on for cystourethrocele and stress urinary incontinence. In addition, there were 3 women with cystourethrocele in whom paravaginal defect repair From the Department of Obstetrics and Gynecology, University of Michi- was planned in conjunction with abdominal sacral gan Medical Center. Supported by grants No. DK 47516 and DK 51405 from the National colpopexy with urethral hypermobility to support the an- Institutes of Health. terior vaginal wall in women without stress incontinence, Presented at the Twenty-Second Annual Meeting of the American Urogy- making a total study population of 71. Women with prior necologic Society, Chicago, Ill, October 25-28, 2001. Reprint requests: John O. L. DeLancey, MD, Women’s Hospital L4100, surgery that could distort the anatomy in this area and 1500 E Medical Center Dr, Ann Arbor, MI 48109-0276. E-mail: De- women with genitourinary anomalies were excluded. [email protected] Stress incontinence was considered to be present in © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/6/125733 women with a primary complaint of urinary incontinence doi:10.1067/mob.2002.125733 when symptoms were associated with activity that in-

93 94 DeLancey July 2002 Am J Obstet Gynecol

Table II. Findings on Baden Walker assessment of vaginal support

Lower Upper Cx/ Upper Lower Grade anterior anterior apex posterior posterior

002395511 I142415637 II 46 25 5 5 18 III 8 12 7 0 2 IV 0 5 2 2 0 Not recorded 3 3 3 3 3 Total 71 71 71 71 71

Cx, Cervix.

Table III. Presence of paravaginal defect

Left Right

Paravaginal defect No. % No. %

Fig 1. Space of Retzius showing the left pelvic sidewall. ATLA, Present 62 87.3 63 88.7 Absent 5 7.0 4 5.6 Arcus tendineus levator ani; CL, Cooper’s ligament; IS, ischial Atypical 2 2.8 2 2.8 spine (not seen directly, put position indicated); PCM, pubococ- Not evaluable 2 2.8 2 2.8 cygeus muscle; PBA, pubic bone attachment of the ATFP; Total 71 100 71 100 PS, pubic symphysis (unusually prominent). © DeLancey.

Table I. Patient characteristics and test results Five women had previously had an abdominal hysterectomy and 3 had a vaginal hysterectomy (not performed for Straining cotton prolapse). At the time of the study operation, 35 women swab had 42 operations performed in addition to retropubic Age Parity MUCP angle surgery for stress incontinence. These included abdominal Average 52 3 48 67.7 hysterectomy, 19; posterior colporrhaphy, 9; abdominal SD 12.4 1.38 15.8 16.8 sacral colpopexy, 4; vaginal hysterectomy, 3; anterior-poste- Minimum 28 0 21 30 rior colporrhaphy, 3; anal sphincteroplasty, 2; sacrospinous Maximum 78 8 89 100 ligament suspension, 1; and tubal ligation, 1. MUCP, Maximal ureteral closure pressure. Once the space of Retzius was fully opened for paravaginal defect repair, the status of the fascial and muscular elements of the urethral support system were creases abdominal pressure. Each woman confirmed that observed and noted. the physical finding of stress urinary incontinence Presence or absence of paravaginal defect. For the pur- demonstrated during a full bladder cough test was the poses of this study, we define paravaginal defect as the type of incontinence for which she sought treatment. A medial displacement of the vaginal wall and pubocervical cotton swab test and urethral pressure study with a later- fascia from its normal line of attachment to the pelvic ally oriented 8F dual microtip catheter was obtained. sidewall at the ATFP. The attachment between the pubo- A cystometrogram was also performed. cervical fascia and pelvic wall normally lies along a line Patient characteristics and test results are seen in Table I. from a point 1 cm above the inferior border of the pubic Pelvic support was assessed with the Baden Walker sys- bone to the ischial spine.1 The lateral margin of the pub- tem4 in all patients because the pelvic organ support sys- ocervical fascia was identified by three anatomic land- tem had not been introduced at study initiation (Table marks: (1) the longitudinal blood vessels that run along II). A score of 1 indicated descent of the relevant portion the superior lateral sulcus of the vagina, (2) the edge of of the vaginal wall half way to the hymen; 2, to the hymen; the bladder and perivesical fat, and (3) the palpable lat- 3, <4 cm below the hymen; and 4, ≥4 cm below the eral edge of the pubocervical fascia at the lateral margin hymen. Points evaluated are the lower anterior vaginal of the bladder. Decisions on what constitutes normal wall (urethrovesical junction), upper half of the anterior anatomy were based on my experience with normal con- vaginal wall, cervix or vaginal apex, upper posterior vagi- tinence support anatomy.2,3,5 nal wall, and lower posterior vaginal wall. In three indi- Status of the ATFP. The ATFP is a fibrous band viduals scores were not recorded. stretched between attachments at the pubic bone and the Volume 187, Number 1 DeLancey 95 Am J Obstet Gynecol

Fig 2. Examples of defects seen during surgery showing the ven- Fig 3. Conceptual diagram showing the mechanical effect of de- tral half of the arcus. A, Normal anatomy. B, Paravaginal defect tachment of the ATFP from the ischial spine. Top, The trape- without any remnant of the ATFP left on the pelvic wall. C, Rem- zoidal plane of the pubocervical fascia. Left, The attachments to nant that extends one fourth the way to the ischial spine (not the pubis and the ischial spines are intact. Right, The connection shown). D, Split insertion to the pubic bone and remnant that to the spine has been lost, allowing the fascial plane to swing goes at least three fourths the way to the spine. © DeLancey. downward. Bottom, The anterior vaginal wall as they would be seen with a weighted speculum in place. Left, Normal anterior vaginal wall support; right, the effect of dorsal detachment of the ischial spine (Fig 1). The anatomic characteristics evalu- arcus from the ischial spine. © DeLancey. ated were (1) the anterior attachment of the ATFP to the pubis, (2) the posterior attachment of the ATFP to the ischium near the spine, and (3) presence of arcus rem- for the normal connection between this structure and the nants on the pelvic sidewall. The details of each periosteum of the ischium just cephalad to the ischial assessment follow. spine. When this defect was suspected, the pubocervical Status of the ATFP’s attachment to the pubic bone. The fascia approximately 8 cm from the external urinary mea- normal tendon-like attachment of the ATFP to the pubic tus was held against the ischial spine. Normalization of bone was examined. This point occurs consistently 1 cm anatomy by this maneuver confirmed the presence of de- lateral to the midline at the lower portion of the pubic tachment as a contributing cause of paravaginal defect bone approximately 1 cm above the arcuate pubic liga- (Fig 3). ment. Abnormalities were classified in the following way. How far is the ATFP visible along the pelvic sidewall The arcus was considered detached from the pubis when from the pubic bone toward the ischial spine? A portion of no attachment could be seen even when traction was the ATFP often remained on the pelvic wall between the placed on the ATFP. A split was present when the ATFP pubis and spine. When present, it was always seen at the had two definite components that arose from the bone pubic end and was visible for a variable distance toward (Fig 2), one going to the lateral margin of the endopelvic the spine. The extent of its visibility was expressed as a fascia and one to a remnant of the ATFP remaining on proportion of the distance between the pubic bone and the pelvic wall. The attachment was considered to be frag- the spine and expressed in quarters. For example, if the mented when the attachment was separated into more arcus was visible for 2 cm and the distance from the pubis than two parts. to the spine was 8 cm, this was expressed as one quarter of Posterior attachment of the ATFP to the spine. Early in the distance (Fig 2). the study, while we were evaluating the ATFP, it became Status of the pubococcygeal muscle. A portion of the evident that the separation between the pubocervical fas- pubococcygeal muscle that connects the vagina to the cia and the pelvic wall involved detachment of the ATFP pubic bone contributes to the urethral support system.2 It from the ischial spine. Once this became evident, its sta- is visible through the space of Retzius and its status was tus was specifically noted in the last 41 women by looking examined. Abnormalities were classified as a generalized 96 DeLancey July 2002 Am J Obstet Gynecol

Table IV. Attachment of the ATFP to the pubic bone Table V. Detachment of the ATFP from the ischial spine

Left Right Left Right

ATFP to pubic Posterior detachment attachment No. % No. % from ischial spine No. % No. %

Normal 56 78.8 53 74.6 Detached 40 97.6 39 95.1 Split 12 16.9 10 14.0 Attached 0 0 1 2.4 Fragmented 1 1.4 4 5.6 No paravaginal defect 1 2.4 1 2.4 Detached 1 1.4 2 2.8 Total 41 100 41 100 Not evaluable 1 1.4 2 2.8 Total 71 100 71 100 atrophy when all the normally visible red muscle was re- How far is the ATFP visible on the pelvic sidewall from placed by white fibrous tissue. The term “localized atro- the pubic bone? The distance that the arcus could be seen phy” was used when the presence of a well-defined strip of on the pelvic sidewall from its pubic to ischial ends is re- atrophic muscle was seen within visibly normal muscle. ported in Table VI as the proportion of the distance where the arcus was visible. There were 17 women in Results whom some portion of the arcus could be seen all the way Presence and absence of paravaginal defect. Most of to the spine. In these women, however, the majority of the these women had a paravaginal defect (Table III). Of the arcus had pulled away from the sidewall with only a few 71 women, 62 had bilateral defects, with 2 having unilat- fibers still remaining on the pelvic wall. No connection eral right defects and 1 a left defect. In 3 women the between the vaginal wall and the arcus at this level was anatomy was atypical or could not be adequately assessed. present. Three women with urethral hypermobility and stress Status of the pubococcygeal muscle. The status of the incontinence did not have a paravaginal defect on either pubococcygeal muscle attached to the pubis and visible side. A nulliparous woman with a cotton swab angle of 90 above the arcus tendineus fascia pelvis is shown in Table degrees and a cystouretrocele at the hymen had a defect VII. in urethral support below the ATFP and the other, a mul- Slightly more than 50% had visible abnormalities in tiparous woman, had similar findings. The third woman this muscle, and these abnormalities were evenly split be- had a complete avulsion of the pubovaginal portion of tween a generalized atrophy that involved all the visible the pubococcygeus muscle from the inner surface of the muscle and localized defects that involved a strip of mus- pubic bone. In two obese women, sufficient visibility to re- cle that was missing. In several women, the loss of muscle liably evaluate a defect could not be obtained without was so great that the superior aspect of the perineal mem- going beyond the dissection normally carried out during brane was visible from the space of Retzius. the operation. Detachment of the ATFP from the pubic bone. No Comment woman had a bilateral detachment of the ATFP from the The ATFP is a band of dense regular connective tissue pubic bone and detachment from one side was rare stretched between the pubic bone and the ischial spine. (Table IV). In many women the attachment was split, with The pubocervical fascia forms a trapezoidal layer span- a portion connecting to the pubocervical fascia and the ning the area between the two arcus tendineae. We have other portion attaching to the remnant of the arcus on found that dorsal detachment of the arcus from the spine the pelvic wall (Fig 2, D). In a number of women with por- is associated with anterior vaginal wall descent. The me- tions of the ATFP on the pelvic wall, the slit occurred chanical concept relating this detachment to descent of some distance from the attachment to the pubic bone so the urethra and anterior vaginal wall can be seen in Fig 3. that there are more women with ATFP on the pelvic wall The finding that the arcus is detached from the spine is than there are those with a split attachment (Figure 2, C). consistent with the pattern found in the portion of the Posterior detachment of the ATFP. In contrast to the al- arcus remaining on the sidewall. It was invariably the ven- most universal preservation of the connection between tral part nearer the pubic bone that remained visible. We the ATFP and the pubic bone, the posterior connection did not see a single instance with arcus attached at each to the ischial spine was usually lost in those 41 women end and missing in the middle. The paravaginal defect, evaluated for this feature who had paravaginal defects therefore, most often arises because the connection be- (Table V). In 37 of the 38 women with posterior detach- tween the arcus and the spine becomes detached, allow- ment, the posterior detachment occurred on both sides. ing the vagina to swing caudally. The one woman whose arcus continued to be attached The nature of these defects suggests hypotheses con- posteriorly had atypical anatomy on that side. cerning how support system damage may influence ure- Volume 187, Number 1 DeLancey 97 Am J Obstet Gynecol

Table VI. The proportion of the distance between the Table VII. Status of the pubococcygeal muscle pubic bone and the ischial spine in which the ATFP was Left Right visible Muscle No. % No. % Left Right Normal 32 45.1 28 39.4 ATFP seen on General atrophy 20 28.2 18 25.7 pelvic wall No. % No. % Localized atrophy 16 22.5 21 30 Not evaluable 3 4.2 3 4.3 None 27 38.0 26 36.6 Other 0 0.0 1 1.4 One fourth distance 6 8.5 7 9.9 Total 71 100 71 100 One half distance 16 22.5 14 19.7 Three fourths distance 2 2.8 3 4.2 To spine 12 16.9 13 18.3 No defect 2 2.8 1 1.4 Not evaluable 6 8.5 7 9.9 Total 71 100.0 71 100.0 tachments. In fact, earlier studies have proved that some women lose the voluntary ability to elevate their urethral support and stress continence. Detachment of the thras,10 substantiating the fact that muscle defects can arcus from the spine could loosen the tissues of the fascia happen. Several authors have called attention to the pres- that span the distance between the two arcus tendineae, ence of pelvic muscle damage,11 and work is ongoing in compromising the stiffness of the suburethral tissues. In our unit to define the magnetic resonance imaging ap- addition, it would also alter the relationship between the pearance of these lesions so that case-controlled observa- urethral supports and the angle of action of the levator tions can be carried out on asymptomatic women. The ani muscles. In normal circumstances, these act perpen- relationship between these observations and the growing dicular to the longitudinal axis of the urethra. As the ure- data concerning neuromuscular injury and pelvic floor thra drops, however, it may no longer be in a position dysfunction12-14 remains to be fully elucidated. where the muscles can exert a force that counterbalances This study predominantly concerns women with ure- abdominal pressure. thral hypermobility and stress urinary incontinence and The presence of occasional atypical defects, however, does not include women with stress incontinence and was seen. The nulliparous woman with a cystourethrocele normal support (intrinsic sphincter deficiency). There- and cotton swab straining angle of 90 degrees despite an fore, these observations relate to women with abnormal intact arcus is, perhaps, the exception that proves the urethral support and do not provide a full picture of the rule. This variation reveals that each woman can have an anatomic situation in all women with the symptom of individual injury to one of several parts of the supportive stress urinary incontinence. system. The idea that there is a single cause of hypermo- This is an observational study and, as such, has limita- bility is not born out by this study. tions. It has provided detailed anatomic information The findings of this study expand and add detail to the about the status of the tissues involved in urethral sup- original description of the paravaginal defect by Richard- port. The author could not be blinded to the patient’s son et al.1 It confirms the frequent nature of this defect in clinical status, so these observations will need to be tested women with stress urinary incontinence and cys- with objective measures that permit blinded evaluation of tourethrocele and adds specificity to the location of struc- both continent and incontinent women. Such a study can tural detachments that occur. be carried out with use of magnetic resonance imaging The pubococcygeal portion of the levator ani muscles where examiners, blinded to patient symptoms and ante- also shows damage in these women. Because activity of rior vaginal wall support, can review individual scans and the muscles influences the stiffness of the urethral sup- make objective assessments. Although it is possible to see ports,6 loss of this muscle may influence stress conti- certain changes present in women with the paravaginal nence. In this study we found that the medial portion of defect,15 demonstrating the location of the detachments the levator ani muscle involved in urethral support is visi- that give rise to this appearance will need further work. bly abnormal in half these women. This is consistent with We believe that understanding the precise nature of the findings of others7,8 that reveal this portion of the the anatomic changes associated with symptoms and muscle to be abnormal in women with pelvic floor dys- physical findings in women with pelvic floor disorders will function. This type of abnormality is associated with vagi- allow us to sharpen our knowledge of causation. Hope- nal delivery,9 as is stress incontinence.7 fully, this will lead to better treatment selection and pre- Defects in the muscles involved in urethral support vention. could lead to increased urethral mobility as a result of loss of the upward forces provided by these muscle at- Submitted in memory of A. Cullen Richardson, MD. 98 DeLancey July 2002 Am J Obstet Gynecol

REFERENCES 9. Dimpfl T, Jaegar C, Mueller-Felber W, Anthuber C, Hirsch A, 1. Richardson AC, Edmonds PB, Williams NL. Treatment of stress Brandmaier R, et al. Myogenic changes of the levator ani muscle urinary incontinence due to paravaginal fascial defect. Obstet in premenopausal women: the impact of vaginal delivery and Gynecol 1981;57:357-62. age. Neurourol Urodynam 1998;17:197-206. 2. DeLancey JOL. Structural aspects of the extrinsic continence 10. Peschers U, Schaer GN, Anthuber C, DeLancey JOL, Schuessler B. mechanism. Obstet Gynecol 1988;72:296-301. Changes in vesical neck mobility following vaginal delivery. Ob- 3. DeLancey JOL. Structural support of the urethra as it relates to stet Gynecol 1996;88:1001-6. stress urinary incontinence: the hammock hypothesis. Am J Ob- 11. Kirschner-Hermanns R, Wein B, Niehaus S, Schaefer W, Jakse G. stet Gynecol 1994;170:1713-20. The contribution of magnetic resonance imaging of the pelvic 4. Baden WF, Walker T. Surgical repair of vaginal defects. In: Fun- floor to the understanding of urinary incontinence. Br J Urol damentals, symptoms and classification. Philadelphia: Lippin- 1993;72:715-8. cott-Raven; 1992. p. 9-23. 12. Snooks SJ, Badenoch DF, Tiptaft RC, Swash M. Perineal nerve 5. DeLancey JOL. Correlative study of paraurethral anatomy. Ob- damage in genuine stress urinary incontinence: an electrophysi- stet Gynecol 1986;68:91-7. ological study. Br J Urol 1985;57:422-6. 6. Howard D, Miller JM, DeLancey JOL, Ashton-Miller JA. Differ- 13. Smith ARB, Hosker GL, Warrell DW. The role of pudendal nerve ential effects of cough, valsalva, and continence status on vesical damage in the aetiology of genuine stress incontinence in neck movement. Obstet Gynecol 2000;95:535-40. women. Br J Obstet Gynaecol 1989;96:29-32. 7. Koelbl H, Strassegger H, Riss PA, Gruber H. Morphologic and 14. Smith ARB, Hosker GL, Warrell DW. The role of partial dener- functional aspects of pelvic floor muscles in patients with pelvic vation of the pelvic floor in the aetiology of genitourinary pro- relaxation and genuine stress incontinence. Obstet Gynecol lapse and stress incontinence of urine: a neurophysiological 1989;74:789-95. study. Br J Obstet Gynecol, 1989;96:24-8. 8. Hanzal E, Berger E, Koelbl H. Levator ani muscle morphology 15. Huddleston HT, Dunnihoo DR, Huddleston PM, Meyers PC. Mag- and recurrent genuine stress incontinence. Obstet Gynecol netic resonance imaging of defects in DeLancey’s vaginal support 1993;81:426-9. levels I, II, and III. Am J Obstet Gynecol 1995;172:1778-84.