Genital prolapse surgery A study of methods, clinical outcome and impact of pelvic floor muscle function

(1)

1 Linköping University Medical Dissertations, No. 1103

Genital prolapse surgery

A study of methods, clinical outcome and impact of pelvic floor muscle function

Kristina Crafoord

(2)

2

Serie: Linköping University medical dissertations, No. 1103

©

Kristina Crafoord, MD, 2009 ISSN 0345-0082

ISBN: 978-91-7393-689-7

Printed by LiU-Tryck, Linköping, Sweden, 2009 Cover design: Dennis Netzell

(3)

3 To the women in the study who endured all investigations in my search of enlarging the knowledge of pelvic floor disorders...

(4)

4

不怕走慢，就怕不動。

Do not fear going forward slowly; fear only to stand still. (Chinese proverb)

(5)

5

List of Original Articles

This dissertation is based on the following original publications, which will be referred to in the text by their Roman numerals:

I Crafoord K, Sydsjö A, Nilsson K, Kjølhede P. Primary surgery of genital prolapse: a shift in treatment tradition. Acta Obstet Gynecol Scand 2006; 85(9): 1104-8.

II Crafoord K, Sydsjö A, Kjølhede P. Genital prolapse surgery after a shift in treatment tradition: an analysis of subsequent prolapse surgery. Acta Obstet Gynecol Scand 2008; 87(4): 445-56.

III Crafoord K, Sydsjö A, Johansson T, Brynhildsen J, Kjølhede P. Factors associated with symptoms of pelvic floor dysfunction six years after primary operation of genital prolapse. Acta Obstet Gynecol Scand 2008; 87(9): 910-5.

IV Crafoord K, Brynhildsen J, Lindehammar H, Kjølhede P. Pelvic floor neuropathy in relation to symptoms, anatomy and outcome of vaginal prolapse surgery: a

neurophysiologic study. In manuscript 2009.

V Crafoord K, Brynhildsen J, Hallböök O, Kjølhede P. The predictive value of anorectal manometry on subjective and objective findings and outcome of pelvic organ prolapse surgery. A prospective study. In manuscript 2009.

The published papers are reprinted with permission from the publisher.

The studies in this thesis are approved by the Ethics research committees of the Faculty of Health Sciences at Linköping University (paper I – V) and at the Regional Hospital in Örebro (paper I – III). The amendment of the prospective study (paper IV and V) is approved by The Regional Ethical Review Board in Linköping.

(8)

8

Abbreviations

ANCOVA analysis of covariance ANOVA analysis of variance ARM anorectal manometry ARPA anal resting pressure area ASPA anal squeeze pressure area BMI body mass index CI confidence interval EAS external anal sphincter muscle EMG electromyography

FI fecal incontinence IAS internal anal sphincter

ICS International Continence Society MARP maximal anal resting pressure MASP maximal anal squeeze pressure MLR multiple logistic regression MUP motor unit potential

OR odds ratio

PCM pubococcygeus muscle PFD pelvic floor dysfunction PFM pelvic floor muscle

PNTML pudendal nerve terminal motor latency POP pelvic organ prolapse

POP-Q pelvic organ prolapse quantification SD standard deviation

UI urinary incontinence UTI urinary tract infection

(9)

9

Introduction

In Sweden, genital prolapse surgery has traditionally been carried out with a complete vaginal repair encompassing of all three vaginal compartments concomitantly. This was done independent of the symptoms, size and extent of the prolapse. The Manchester operation was the most common prolapse operation (Danielson 1957). In the late 1980s this surgery with repair in all three compartments was debated and posterior colporrhaphy was especially questioned (Rud 1989). The risk of de novo urinary stress incontinence and dyspareunia was highlighted (Borstad et al. 1989; Brun-Poulsen 1990; Haase et al. 1988). POP surgery of only the vaginal compartments with symptomatic prolapse or very large prolapse was advocated by the gynecological profession and so called prophylactic surgery in compartments without obvious prolapse was abandoned (Rud 1989). The extent to which these recommendations were followed and the clinical consequences they had have not been evaluated.

Genital prolapse or vaginal prolapse is characterized by a portion of the vaginal canal protruding to or from the opening of the vagina. The term is synonymous with pelvic organ prolapse (POP). POP is defined by the International Continence Society (ICS) as the descent of a pelvic organ into or beyond the vagina, perineum or anal canal (Bump et al. 1996). On the basis of this definition, more than half of the adult female population may have POP (Swift et al. 2005). The size of a prolapse is not well correlated to the occurrence of symptoms but women with a prolapse that protrudes outside the vagina are more likely to have symptoms (Tan et al. 2005). Epidemiologic studies from Scandinavia have reported the prevalence of symptomatic POP to be 12 - 15% (Tegerstedt et al. 2004; Uustal Fornell et al. 2003). In clinical studies the prevalence of POP that protrudes outside the vaginal introitus is 2-6%, (Samuelsson et al. 1999; Versi et al. 2001).

The etiology of POP is complex and multi-factorial (Dietz 2008; Jelovsek et al. 2007; Maher et al. 2007). Several studies have demonstrated that changes in collagen metabolism are a factor in the development of POP (Edwall et al. 2008; Söderberg et al. 2004; Song et al. 2007). The pelvic floor muscles seem to play an important role in the support of the pelvic organs and for the development of POP (Athanasiou et al. 2007; DeLancey et al. 2007; Dietz 2007). Information in the literature about the pathophysiological impact of pelvic floor muscles on POP and symptoms of prolapse is scarce.

Surgical treatment of POP is a challenge to the gynecologist. The issues when, where and how to perform the surgery, preferably as a single procedure, giving the best outcome for the patient, are constantly in question. This attracted my curiosity and inspired me to start the

(10)

10

studies reported on here. In this thesis I will highlight some aspects of POP surgery and its consequences as well as the influence of pelvic floor muscle pathophysiology on pelvic organ prolapse.

Anatomy of the pelvic floor

Both the anatomy and physiology of the pelvic floor is complex. To understand the mechanism of POP and treatment it is important to have detailed knowledge about the anatomy.

The pelvic viscera are part of the pelvic floor and play an important role in the support through their connections with the cardinal and the uterosacral ligaments (Wei et al. 2004).

With normal anatomy in upright position, the distal vagina has a 45-degree angle from the vertical. In the mid-vagina the angle between vagina and vertical increases to 60 to 70 degrees giving the vagina a “banana-shape” (Rovner 2000).

The vaginal support can be described in three levels (Fig. 1) (DeLancey 1999):

Level I: The cardinal and uterosacral ligaments (parametrium) attach the uterine cervix to the pelvic walls and the sacrum, respectively and the uppermost portion of the paracolpium suspends the superior part of vagina to the pelvic walls.

Level II: In the mid portion of vagina the paracolpium attaches the vagina lateral to the pelvic wall, thereby stretching the vagina transversely between the bladder and the rectum. The vagina is, in this level, attached to the arcus tendineus fascia pelvis (“the white line”) and the superior fascia of the levator ani.

Level III: The distal vaginal walls are directly attached to the surrounding structures. Anteriorly, the vagina fuses with the urethra; posteriorly, with the perineal body, and laterally with the levator ani muscles. The lower part of the posterior wall is, in part,

Figure 1. The levels of support (DeLancey, with permission)

Figure 1. The three levels of vaginal support. (DeLancey, with permission.)

(11)

11 supported by the rectovaginal septum, which connects the perineal body to the pelvic wall. It is present in the caudal 2 to 4 cm of vagina. Above this, there is only adventitia between the vagina and the rectum.

The pelvis has two fascias: the parietal fascia covers the muscles, and the visceral fascia surrounds the pelvic organs. The vagina is attached by fibrous connective tissue to the parietal fascia, usually named the endopelvic fascia. The white line is a band of connective tissue reaching from the pubic symphysis to the ischial spine on both sides of the pelvis. The perineal body is the central fusion between the halves of the perineal membrane in the lower 1/3 of vagina. The connection extends cranially to approximately 2-3 centimeters above the hymeneal ring. The densest part is the distal perineal body. Cranially it is becomes progressively thinner. The posterior wall has a U-shaped contour in level III (DeLancey 1999). The levator ani muscles are attached to the perineal body and directly connected to the upper surface of the perineal membrane (DeLancey 1999).

The perineal membrane (also called “the urogenital diaphragm”) in the anterior part of the pelvis, below the levator ani muscles and at the level of the hymen, is a three-dimensional structure composed of a dorsal and a ventral portion (Stein et al. 2008). The dorsal portion has bilateral fibrous sheets, which attach the lateral walls of the lower one third of the vagina and the perineal body to the ischiopubic ramus (level III) (Fig. 2).

The ventral portion is part of a solid tissue mass involving the compressor urethra and the urethrovaginal sphincter of the distal urethra. Here the perineal membrane is continuous Figure 2. The perineal membrane and

attachments (A). Separation of the fibers in this area (B) leaves the rectum unsupported and results in a low rectocele.

(12)

12

with the insertion of the arcus tendineus fascia pelvis. The levator ani muscles are connected with the cranial surface of the perineal membrane (Stein et al. 2008) (Fig. 3).

The levator ani muscles consist mainly of striated muscles and lie below the endopelvic fascia. The levator ani muscles are traditionally subdivided into an anterior part, the pubococcygeus muscle, and a posterior part, the ileococcygeus muscle. Together with the coccygeus muscles the levator ani muscles form the bottom of the pelvis (Fig. 4).

Figure 3. A) Position of the perineal membrane. B) Levator ani muscles seen from below the edge of the perineal membrane. (DeLancey, with permission.)

A B

(13)

13 The levator ani muscles extend from the pubic bone anteriorly to the ischial spine posteriorly and are attached to the white line. In the midline the levator ani muscles fuse but leave a hiatus (the urogenital hiatus) through which, from ventrally to dorsally the urethra, vagina and rectum pass. The hiatus is posteriorly supported by the perineal body and the external anal sphincter, and anteriorly is bounded by the pubic bone. The iliococcygeus, pubococcygeus and coccygeus muscles form a transverse muscular shelf (the levator plate) from the pelvic sidewalls, above the ischioanal fossa. Medially, the muscles slope down sharply to form a vertical portion. The external anal sphincter muscles and the puborectalis muscle are situated outside and the internal sphincter inside this vertical portion.

The puborectalis muscle has previously been considered to be a part of the levator ani muscle but is now considered to be below the transverse portion of the pubococcygeus, iliococcygeus and coccygeus muscles and outside the vertical portion of the levator muscle (Shafik 1979) (Fig. 5). The puborectalis muscle binds the vertical portion of the levator ani muscle to the pubic symphysis, and encloses the intrahiatal structures (i.e. the urethra, vagina and rectum).

Figure 5. The anorectal angle. (Reprinted with permission.)

EAS IAS Rectum M. puborectalis Anorectal angle Anal canal Septum rectovaginale Pubic symphysis M. pubococcygeus Cervix

(14)

14

The anal canal is surrounded by the internal and external anal sphincter muscles. The external anal sphincter (EAS) forms a ring of muscle around the anal canal and is continuous with the puborectalis muscle The EAS is a striated muscle under voluntary control but is continuously active with a resting tonus. It is approximately 4 cm in length and surrounds the internal anal sphincter. The internal anal sphincter (IAS) is a smooth muscle continuous with the circular muscles of the rectum. It exhibits a continuous state of tonic contraction. A high-pressure zone is found in the anal canal during the resting state, the peak high-pressures are found 2 cm from the anal verge. Both sphincters contribute and predominantly the IAS with 70-85% of the resting tone.

The somatic innervations of the levator ani muscles is by direct branches from the sacral plexus (S3-5) and the pudendal nerve ((S2-4) both autonomic and somatic) is responsible for innervations of the external anal sphincter and the puborectalis muscles (Barber et al. 2002) (Guaderrama et al. 2005) (Fig. 6). The autonomous sympathetic (L2-4) and parasympathetic (S2-4) innervations is supplied by the inferior hypogastric plexus, which has an important function in the smooth muscle relaxation and contraction of the pelvic floor, bladder, vagina and rectum.

(15)

15

Development of pelvic organ prolapse

The interaction between the pelvic floor muscles and the supportive connective tissue is essential for the support of the pelvic organs. When the pelvic floor muscles work properly the urogenital hiatus is closed. The pelvic ligaments and fascias are normally under minimal tension due to the support of the levator ani muscles. If the pelvic muscles are damaged, the urogenital hiatus opens and the stress on pelvic ligaments and fascias thereby created can cause breaks or stretching in these structures thus promoting development of prolapse (Wei et al. 2004). This has been described as “the boat in the dock”-mechanism by DeLancey (DeLancey 2002) (Fig. 7).

The muscle may be injured directly or indirectly by damage to the nerves that innervate the muscles. The development of POP is progressive when the muscles are no longer able to maintain the support (Chen et al. 2006).

Lacerations and stretching of the pelvic ligaments and fascias caused by direct trauma or continuing stress can lead to the formation of prolapse even without obvious muscle damage. Inherited disorders of collagen or muscle tissue causing weakness may also promote development of POP.

Figure 7. The boat corresponds to the uterus/vagina, the ropes to the ligaments/fascias and the water to the supportive pelvic floor muscle. The ropes act to hold the ship in the center of its berth. - The support system will not function if the water level falls; the ropes holding the boat will be stressed and eventually break.

(16)

16

Symptoms associated with pelvic organ prolapse

Pelvic floor symptoms may be due to dysfunction of an adjacent organ system because of loss of vaginal support (for instance urinary stress incontinence) or muscular or neurogenic damage to the pelvic floor (i.e. fecal incontinence). The source of a symptom (i.e. constipation or obstructed defecation and chronic straining) can also contribute to the development of POP. Pelvic floor dysfunction (PFD) is the joint designation of a group of clinical conditions that includes urinary incontinence, fecal incontinence, pelvic organ prolapse, sensory and emptying abnormalities of the lower urinary tract, defecatory dysfunction, sexual dysfunction and several chronic pain syndromes. Symptoms that usually are associated with POP and their prevalence are shown in Table 1.

Table 1. Common symptoms of pelvic organ prolapse and reported prevalence of the symptoms.

Type of symptom Symptom Prevalence

Local symptoms: Vaginal protrusion or bulge 10-100%

Pelvic pressure or heaviness 10-49%

Urinary symptoms: Manual reduction to start or complete voiding 9-44%

Feeling of incomplete emptying 30-50%

Urinary stress incontinence 13-83%

Urgency/ urge incontinence 21-73%

Bowel symptoms: Difficulty in defecation 20-23%

Digitation or splinting of vagina, perineum or anus to complete

defecation 6-35%

Incontinence of flatus, liquid stool or solid stool 10-31%

Sexual symptoms: Dyspareunia 8-69%

The POP symptoms are not consistently associated with the size or site of POP (Ellerkmann et al. 2001) as demonstrated in Fig. 8. The prolapse often needs to extend beyond the level of the hymen if mechanical symptoms are to occur (Swift et al. 2003; Gutman et al. 2008; Mouritsen et al. 2003). Reports of a bulge are very common in association with occurrence of POP (Swift et al. 2003; Tegerstedt et al. 2005). The symptoms of needing to manually reduce the bulge to urinate or defecate are also associated with prolapse (Tan et al. 2005).

(17)

17 0 50 100 -3 -2 -1 0 1 2 >=3 Pr ev ale nc e % .

Maximum extent of prolapse

0 50 1 00 - 3 -2 - 1 0 1 2 > = 3 .0 Pr ev ale nc e % . M a x im um e x ten t of p ro lap s e

Figure 8. Relationship between symptoms and maximal extent of prolapse. Maximal extent of prolapse measured in 1-cm increments with the hymen = 0. Prolapse in which maximal extent is proximal to the hymen is in negative numbers and beyond the hymen positive numbers.

(Modified from Tan et al. 2005). Bulging sensation Pelvic pressure Prev alenc e % Prev alenc e %

(18)

18

Clinical evaluation of pelvic organ prolapse

Evaluation of local/mechanical symptoms as well as functional symptoms from the lower urinary tract, the bowels, and the effects on sexual life and on quality of life should be made for all women with POP. An evaluation form is presented in Table 2.

Table 2. Evaluation form for assessing POP, based on clinical experience. History

 Complete medical (including heredity, hypermobility), surgical, obstetric and gynecological history

 Pelvic symptoms (as vaginal protrusion/ pelvic pressure, worsening by lifting)  Urinary symptoms (difficulty starting voiding/splinting, emptying, incontinence)  Bowel patterns (straining, frequency, splinting to defecate, fecal incontinence)  Sexual function problems

 Patient self-reported pelvic symptom questionnaire/s that assess bowel, urinary and sexual function.

Physical examination  Abdominal palpation

 Vaginal examination (start with visual inspection at maximal straining; in lithotomy and upright position; staging level of protrusion in each vaginal compartment related to the hymenal plane; if cystocele central/paravaginal/both)  Cough provocation test with and without reposition of the prolapse

 Digital pelvic examination and assessment of pelvic floor muscle strength  Bimanual gynecological/anorectal examination

 Vaginal ultrasound examination (residual urine, ovarian tumor, myoma)  Neurological examination to assess sensory and motor function

(physical disability, mental status, perineal sensation)

POP-Q

The Pelvic Organ Prolapse Quantification (POP-Q) system as the standardization of grading of prolapse was approved by the International Continence Society, the American Urogynecologic Society, and the Society of Gynecologic Surgeons in 1996 (Bump et al. 1996). Measurements of the descent of the anterior, posterior wall and the anterior lip of the cervix/vaginal cuff in cm above or below the plane of the hymeneal remnants (negative above and positive values below) are measured with the patient straining (Fig. 9). The length of the genital hiatus and the perineal body are measured when the patient strains. The length of

(19)

19 vagina is measured in a resting state. Depending on the measurements the pelvic organ status is divided into stages 0-4.

Quantitative POP

Grade 1

Grade 2

Vaginal Profile (Baden)

Grade 3 Grade 4 Stage 1 Stage 2 Stage 3 Stage 4 Stage 0 (-)1 cm (+)1 cm Midplane of vagina Hymeneal ring Introitus Complete eversion

Vestibule Labia minora Labia majora Lower half vagina Upper half vagina

Points:

Aa. Located in the midline of the anterior wall 3 cm proximal to the external urethral meatus.

Ba. The most distal part of the upper vaginal wall.

C. The most distal edge of the cervix or leading edge of the vaginal cuff.

D. Point representing the location of posterior fornix (not measured if cervix is abscent) Bp. The most distal part of the posterior vaginal wall.

Ap. In the midline of the posterior wall 3 cm proximal to the hymen.

Figure 9. Six points (Aa, Ba, C, D, Bp and Ap), genital hiatus (gh), perineal body (pb), total vaginal length (tvl) used for pelvic organ support quantification. (Reprinted from Bump et al 1996, with permission.)

(20)

20

In Fig.10 the different levels of staging according to the Baden-Walker halfway system and the POP-Q are shown (Baden et al. 1972; Bump et al. 1996). Vaginal prolapse in the different vaginal compartments are shown in Fig. 11.

Surgical treatment of pelvic organ prolapse

The aims of surgical treatment are to restore vaginal support and function and relieve symptoms in women with symptomatic POP. The indications for POP surgery are most often relative, depending on subjective and objective findings and the patient’s preferences.

History

In the 19th century amputation of the cervix and perineorrhaphy were used in surgical prolapse treatment for repairing damages of childbirth and restoring vaginal closure. The aim of these vaginal operations was to enable a pessary to be used for the prolapse (Pearce 2004).

Archibald Donald (1860-1937) of Manchester, England started the surgical treatment of genital prolapse by combining correction of all three vaginal compartments. In this way no pessary was required postoperatively.

A

D

Figure 11. Cystocele (A), Rectocele (B), Descent of uterus (C) and Descent of the vaginal vault (D). C

(21)

21 Donald performed his first “colporrhaphy” in 1888. He emphasized “What one must recognize, if one is to have any success in the operations, is that the raw surface, anterior or posterior, or in exceptional cases, both, must be high, wide, and deep, and that the deep tissues, triangular ligament, or levator ani muscles must be brought together by deep buried stitches”. His colleague William Edward Fothergill (1865-1926) further developed the method by extending the incision in the anterior vaginal wall to join the incision of the cervix, and plicate the parametria anterior to the cervix (Pearce 2004). This has later been known as the Manchester-Fothergill operation (Bemis 1974).

In 1957 Carl-Olof Danielson described the Manchester operation to be the most common vaginal prolapse operation in the Scandinavian countries during the 1940-50s (Danielson 1957). He emphasizes, in his thesis, the aim of a “complete” operation and especially the amputation of the cervix to be of importance to avoid recurrence of the prolapse.

In 1959 Jeffcoat reported a high prevalence of dyspareunia in women operated with combined anterior and posterior colpoperineorrhaphy (Jeffcoate 1959). He recommended that the posterior colpoperineorrhaphy be practiced only when a significant degree of rectocele or perineal deficiency existed, to be determined by examination preoperatively, before the patient’s muscles were relaxed by anesthesia.

Dyspareunia and urinary incontinence were highlighted as postoperative problems in reports from Scandinavia (Borstad et al. 1989; Brun-Poulsen 1990; Haase et al. 1988). The recommendations given to the Swedish gynecologists by the Reference Group in Urogynecology within the Swedish Society of Obstetrics and Gynecology in 1989 were intended to avoid posterior repairs in the women with asymptomatic rectocele. Advice was also given to avoid “overcorrection” of the cystocele (Rud 1989). Levator ani plication is nowadays not recommended in sexually active women (Weber et al.

2005)(Fig. 12). Figure 12. Colpoperineorrhaphy with plication of

the levator muscles (from Shaw´s textbook of Operative Gynaecology 1960 2nd_{ed. With} permission)

(22)

22

Several modifications of the methods have been described in order to make it possible to more selectively repair specific support defects. Paravaginal repair was first described by White 1909 as a vaginal procedure. The repair corrected the paravaginal defects by reattaching the anterior lateral vaginal sulci to the obturator internus muscles and fascia at the level of arcus tendineus fascia pelvis (“white line”) transvaginally or abdominally.

In 1993 Richardson described rectocele repair by suturing specific breaks in the rectovaginal fascia (Richardson 1993). These breaks are detected peroperatively after incision of the vaginal mucosa and are sutured specifically.

During the recent decade meshes/implants have been introduced in prolapse surgery. Various mesh materials, biologic and synthetic, have been suggested and different surgical approaches have been suggested. So far scientific information about outcomes of operations with meshes and the indications for use of these is limited.

There are numerous surgical methods described to treat pelvic organ prolapse. Often these are modifications of the traditional methods or they are modifications adapted to individual circumstances of the prolapse and the patient. The methods may also differ in use of suture material, surgical technique and conditions related to anesthesia and perioperative care. A summary of the most commonly reported procedures and their application in treating each vaginal compartment is given in Table 3.

Table 3. Summary of procedures described for treatment of pelvic organ prolapse. Vaginal

compartment Vaginal procedures Abdominal procedures Anterior Anterior colporrhaphy

Paravaginal repair

Anterior prolapse repair with mesh

Colposuspension Paravaginal repair Anterior repair with mesh Apical Vaginal hysterectomy

McCall culdoplasty Amputation of the cervix Plication of cardinal ligaments

Vault attachment to iliococcygeus fascia Vault attachment to sacrouterine ligaments Vault repair with/without mesh

Sacrospinosus ligament suspension Levator myorrhaphy

Sacrocolpopexy Paravaginal repair Levator myorrhaphy

Posterior Posterior colporrhaphy Posterior colpoperineorrhaphy Levator ani plication

Site-specific repair

Rectovaginal fascia reattachment

Rectovaginal fascial repair with/without mesh Posterior repair with mesh

Laparoscopic

sacrocolpo-perineopexy (combination of apical and posterior defects)

(23)

23 Anterior colporrhaphy is shown Fig. 13.

Figure 13. Traditional anterior colporrhaphy (from Urogynecologic Surgery, ed. W G Hurt, with permission)

(24)

24

Background for the study

It was a clinical impression that a change in mode of surgical treatment of genital prolapse occurred in late 1980s. Information about the long-term effects of POP surgery is incomplete and no Swedish data have been presented. In order to investigate these matters we formulated a series of research questions:

Research questions

 Did any change in the extent of POP surgery, concerning the number of compartments treated, occur between the early 1980s and 1990s?

 If a shift de facto was observed in treatment tradition, what were the consequences regarding subsequent POP surgery?

 What was the subjective outcome in terms of symptoms of pelvic floor dysfunction six years after primary prolapse surgery?

 Which risk factors for adverse outcome regarding symptoms of pelvic floor dysfunction at long-term follow-up could be identified?

The normal supporting function of the pelvic floor muscles depends on the anatomical positions of the muscles, on the activity of the pelvic floor muscles at rest (active support) and on the integrity of the fascia (passive support). Several studies have indicated that impaired function and constitution of the pelvic floor muscles are associated with development of POP, and urinary- and fecal incontinence (Snooks et al. 1984; Allen et al. 1990; Hanzal et al. 1993; Parks et al. 1977; Weidner et al. 2000b; Zhu et al. 2005). The outcome of surgery for pelvic floor dysfunction has also been associated with neuromuscular damage (Benson et al. 1993; Kjølhede et al. 1997; Morley et al. 1996; Welgoss et al. 1999), but there are only few studies in this field. The associations between symptoms of pelvic floor dysfunction and the neuromuscular constitution of the pelvic floor muscles in POP have not previously been investigated. The results of POP surgery vary and are not always satisfactory (Maher et al. 2007). Symptoms may be cured by POP surgery, persist or occur de novo after POP surgery. Even the anatomical outcome varies in a similar way. So far, no study has investigated the neuromuscular function of pelvic floor muscles as a prognostic factor for outcome of POP surgery. Analyses of the function of pelvic floor muscles can be established either by direct measurement of muscle fiber function by means of electromyography (Weidner et al. 2000a) or by means of anorectal manometry, which measure the summary function of the internal and external anal sphincter muscles and the puborectalis part of the levator ani muscles (Read et al. 1992). Neurophysiological measurements of pelvic floor muscles with electromyography

(25)

25 (EMG) and nerve conduction studies are methods that can characterize the integrity of the muscle, nerve and neuromuscular junction and can localize an injury and quantify the severity of the problem (Olsen et al. 2001). Anorectal manometry (ARM) is a method that assesses the anorectal function and provides insight into the pathophysiology of anorectal disorders (Read et al. 1992).

The impact on the neuromuscular function of pelvic floor muscles on symptoms and extent of POP as well as on the outcome of POP surgery is so far not well investigated. In addition to and as a consequence of our previous studies of POP surgery, I found it of great interest to investigate the neuromuscular function of pelvic floor muscles in women who were scheduled for POP surgery.

The following hypotheses and research questions were elaborated:

Hypotheses Research questions

 Neuromuscular damage of pelvic floor muscles identified by neurophysiologic investigations can be related to symptoms of pelvic floor dysfunction and degrees of pelvic organ prolapse and further may be predictive for the outcome of the POP surgery.

 Are neurophysiologic findings of the pubococcygeus muscles and external anal sphincter muscle associated with symptoms of pelvic floor dysfunction and size and extent of POP in women who are undergoing primary surgery for POP?

 Can the neurophysiologic findings predict the outcome of surgery with respect to symptoms and anatomy?  Pelvic organ prolapse and bowel

emptying symptoms, and degree and extent of POP as well as outcome of vaginal POP surgery, in particular posterior colporrhaphy, can be predicted by anorectal manometric findings preoperatively.

 How are the long-term results concerning symptoms of pelvic floor dysfunction and anatomy after primary POP surgery?

 Are anorectal manometric findings associated with symptoms of pelvic floor dysfunction and size and extent of POP in women who are undergoing primary surgery for POP?

 Can the anorectal manometric findings and changes in these predict long-term outcomes in terms of symptoms and anatomy after primary POP surgery? Based on these research questions a series of studies were conducted and the specific aims of this thesis were prepared.

(26)

26

Aims of the study

The aims of the study were:

i. to evaluate whether a shift in the extent of genital prolapse surgery occurred between 1983 and 1993 regarding prevalence of complete repair versus selective repair.

ii. if a shift occurred, to analyze the consequences of such a shift regarding the need for subsequent prolapse surgery and to analyze the frequency of complementary and recurrent surgery with respect to size of the prolapse and extent of the primary POP surgery.

iii. to determine the prevalence of symptoms of pelvic floor dysfunction in women six years after vaginal prolapse surgery and to analyze predictive factors for these symptoms.

iv. to analyze subjective and objective outcomes of POP surgery concerning change in PDF-symptoms and anatomical status of the vaginal compartments at long term follow up.

v. to evaluate whether neurophysiologic findings of the levator ani muscles and external anal sphincter muscle and anorectal manometry pressure measurements, respectively were associated with preoperative symptoms of pelvic floor dysfunction and the extent and degree of pelvic organ prolapse.

vi. to investigate whether the neurophysiologic findings and anorectal manometry pressure measurements respectively could predict the anatomical and subjective outcomes of the prolapse surgery.

(27)

27

Material

Study populations

The retrospective material (Paper I-III)

The cohorts of women operated for genital prolapse at the departments of Obstetrics and Gynecology at Linköping University Hospital, Örebro Regional Hospital and the County Hospital of Värnamo in 1983 (Period I) and 1993 (Period II) were identified from the local County patient and operation registers. In order to achieve samples of approximately similar size in the hospitals the samples from the County Hospital in Värnamo were expanded to also include all patients operated on in 1982 and 1992, respectively.

The patient records were retrieved and reviewed. The women without previous prolapse surgery were identified and these women constitute the study group. The prolapse operation was the first POP operation in these women, i.e. it was primary surgery.

The study material in Paper I-III is illustrated in the flow-chart (Fig. 14).

The prospective material (Paper IV and V)

Women admitted to the department of Obstetrics and Gynecology at the University Hospital in Linköping for surgical treatment of genital prolapse, comprising at least a posterior colporrhaphy, between November 1999 and Mars 2004 were eligible for the study. Exclusion criteria in the study were occurrence of previous surgery for POP, urinary or fecal incontinence, or total hysterectomy, significant physical, neurological or psychic disability; or complete vaginal eversion (stage 4 – prolapse).

One single investigator (the Ph.D. student) was to perform all preoperative examinations and participate in the laboratory investigations. Given this constraint – that all women in the study were to be examined by a single physician - only those patients scheduled for preoperative evaluation on the days when this physician was in residence were given written and verbal information about the study. No selections were otherwise done deliberately among the potentially eligible women.

(28)

28

The retrospective studies (Papers I – III)

1982/83

n = 293 n = 317

Previous POP surgery

n = 32 Previous POP surgeryn = 36

No subsequent POP surgery n = 254 No subsequent POP surgery n = 259

1992/93

Six y ea rs pos to pe ra tiv e fo llow-up n = 4 n = 3 n = 5 Complete repair n=180 Selective repair n=81 Complete repair n=101 Primary Surgery: n = 281 Selective repair n=180 Primary Surgery: n= 261 Answered the questionnaire n = 188 n = 9 n = 3 n = 5

Paper III

Paper I

: Missing records n = 4 Not answered n=52 Period I Period II

Paper II

Deceased n =19 Po st al q uest io nn ai re

Complementary POP surgery

Complementary and recurrent POP surgery

Recurrent POP surgery

(29)

29 The prospective study (Paper IV and V)

Figure 15. Flow chart showing the study population in the prospective material. One woman deceased before long-term follow-up and two other women did not want to participate in the clinical examination at long-term follow up. Follow-up visits with clinical examination were conducted approximately two years postoperatively in these three women. No questionnaires were obtained from the deceased woman or from one of the two women who declined to participate in the clinical assessment at follow-up.

(30)

30

Methods

Review of patient records

In the retrospective study the patient records were systematically reviewed in 1999 by three experienced gynecologists. Demographic, clinical, and surgical data were extracted and collected in a standardized form. The categorization of the size of the prolapse noted in the patient record was done according to the Baden-Walker half-way system (Baden et al. 1972). Information about subsequent POP surgery was noted and categorized as a) recurrent POP surgery, i.e. the subsequent surgery was done in a vaginal compartment that previously was operated upon at the primary surgery; b) complementary POP surgery, i.e. the subsequent surgery was carried out in a previously not operated vaginal compartment; or c) combination of a) and b).

Questionnaires

Follow-up study of the retrospective material (Paper III)

A postal questionnaire with 68 questions concerning symptoms of pelvic floor dysfunction was developed. The questionnaire consisted of questions regarding demographic and clinical data and detailed questions concerning symptoms of prolapse, bowel and urinary function and sexual issues. The questions concerning the symptoms of pelvic floor dysfunction used in the study were excerpted from validated questionnaires (Oliveira et al. 1996; Uustal Fornell et al. 2003).

The questions were constructed as simple sentences and the answers were given by placing a cross or checkmark in boxes next to the written alternatives. The number of boxes was limited to between two and ten. The alternative answers to the questions concerning symptoms were mostly constructed to facilitate reporting of the frequency of occurrence of the symptom. The descriptive and demographic continuous data were given by specification of a number. The questionnaire was sent by mail to the patient with a letter providing information and a stamped envelope. One reminder was sent to those who did not respond within four weeks.

(31)

31

The prospective study questionnaires (Paper IV and V)

A questionnaire consisting of 81 questions was prepared. The questionnaire was in construction and content essentially similar to the questionnaire used in the retrospective study as described above. More detailed questions about certain aspects of demographic and clinical issues were added to this questionnaire. The Swedish version of the questionnaire is presented as Appendix B. The questionnaire was completed by the women in the prospective study before the POP surgery and at the follow-up visits two-three years postoperatively. At the long-term follow-up visit four to eight years after surgery the questionnaire was extended by the addition of questions concerning the degree of satisfaction with the result of the surgery and quality-of-life aspects (excerpted from UDI 6 and IIQ7 (Uebersax et al. 1995)). The questionnaire at long-term follow-up is shown as Appendix C.

Clinical assessment

(Paper IV and V)

Pelvic examination was done with the patient in dorsal lithotomy position. The pelvic organ prolapse quantification system (POP-Q) (Bump et al. 1996) was used to objectively characterize the vaginal profile. Urethral competence was assessed by means of a cough provocation test with and without repositioning of the prolapse and by a modified one-hour pad test (Abrams et al. 1990). All patients were preoperatively examined by one urogynecologist (KC). The postoperative assessments of POP-Q were done by three investigators.

Methods of assessing the pelvic floor muscle function

Neurophysiological investigations

Concentric needle EMG

Electromyography (EMG) involves measurement of electrical activity in striated muscle tissue and the activity relates to the neuromuscular condition. Motor unit potentials (MUPs) are recorded (Fig. 16). The analysis of recorded EMG (“motor unit” EMG) helps in the differentiation of normal, denervated/reinnervated and myopathic muscle. Surface or intramuscular electrodes can be used. In the context of this study the concentric needle electrode has a central insulated platinum wire (Fig. 16). Electomyographic activity can be measured mostly to a distance of 2.5 mm from the electrode (Sehapayak 1985). The signals are amplified and recorded. Analysis of the recordings is computerized. The number and

(32)

32

characteristics of MUPs recorded depend upon the local arrangements of muscle fibers within the motor unit and the extent of muscle contraction.

All neurophysiologic tests were performed with MedelecSynergy™ EMG-system (Viasys™ Healthcare, UK). A ground electrode was placed on the right thigh and the patient was placed in dorsal lithotomy position.

The pubococcygeus part of the levator ani muscles were located on both sides by transvaginal palpation during slight contraction and reached transvaginally with the concentric needle electrode guided though a needle holder placed on the examiners index finger. Several positions within the muscles were examined. Local anesthetic was not used. Several sites were sampled from the muscle by the concentric needle electrode. The EAS was examined with the EMG needle inserted through the skin on each sides of the anal orifice at “3 and 9 o´clock” with a finger inside the anus for guidance. All muscles were examined during slight and maximal contraction. The audio output from the electromyographic system´s loudspeaker allowed assessment of the quality of recordings. The EMG-electrodes used were stainless steel disposable concentric needle electrodes (Medtronic, Denmark) with length 50 mm, diameter 0.46 mm, and recording area 0.07 mm². Filter setting was 10 Hz – 10 kHz. Analysis of the EMG signal was done when sharply contoured motor unit potentials (MUPs) were seen and the audio output of the recorder revealed a crisp sound. Only MUPs without noise and with good baseline were analyzed.

The EMG activity of the muscles was analyzed by two methods that measure the size and configuration of the MUPs: The first method was the Multi-MUP analysis. Ten to 20

Figure 16. Concentric needle electrode in needle guide and EMG curve. EMG from the external anal sphincter muscle

(33)

33 unique MUPs from each muscle were recorded from different sites in the muscles during rest and moderate contraction. The following MUP parameters were analyzed: amplitude, duration, number of phases, number of turns, area and the percentage polyphasic MUPs (more than four phases). The mean value for each MUP parameter in each muscle was used for analysis.

The second method was turns/amplitude analysis. Turns per second, mean amplitude per turn and consequently the ratio were analyzed. Signals from five to ten different sites were recorded at different levels of muscle contraction.

The needle positioning in the muscles was performed by the urogynecologist (KC) and the analysis of the recordings by the neurophysiologist (HL).

Pudendal nerve terminal motor latency (PNTML)

Pudendal nerve terminal motor latency (PNTML) involves measurement of the time, the latency, from stimulation of the pudendal nerve at the level of ischial spine to the responding contraction of the EAS. It is a functional evaluation for a motor nerve. A special electrode (St Mark’s electrode) (Fig. 17) has been developed (Kiff et al. 1984). The electrode is placed on the examiner’s gloved index finger and the pudendal nerve is stimulated at the ischial spine using a transrectal approach. At the base of the finger the electrode records the muscular contraction response of the anal sphincter muscle. Stimulation is applied of square wave pulses of 0.2 ms duration and increasing intensity. By adjustment of the fingertip’s position until maximal response is received, the sphincter muscle responses are recorded. The stimulation is repeated until two reproducible responses are recorded from each side. The latency from stimulation to onset of the response is measured.

Figure 17. St Mark´s electrode for recording of pudendal nerve latency. Test of the pudendal nerve latency and the amplitude.

(34)

34

The amplitude of the recorded compound muscle action potential is measured from baseline to first negative peak (Fig. 17). The examination is classified as abnormal if both sides have a latency of more than 2.5 ms (Österberg et al. 2000; Ricciardi et al. 2006).

The pudendal nerve terminal motor latency (PNTML) was measured using a disposable St Mark’s electrode (Dantec-Medtronics, Minneapolis, MN) (Kiff et al. 1984). Stimulation of square wave pulses of 0.2 ms duration and increasing intensity was applied. Small adjustment of the stimulator position was done until the maximal response was recorded from the sphincter muscle. The stimulation was repeated until two reproducible responses were recorded from each side. The latency from stimulation to onset of the response was measured. The amplitude of the recorded compound muscle action potential was measured from baseline to first negative peak. The pudendal nerve latency tests and analysis of the recorded signals were performed by the neurophysiologist (HL).

Ano rectal manometry (ARM)

Anorectal manometry gives information about anal sphincter pressure at rest and pressure generated at voluntary contraction. The sphincter pressures are measured at different levels of the anal canal and give an indication of the strength of the sphincters (Fig. 18).

The resting pressure is mainly due to the function of the internal anal sphincter and normally 50 mmHg. The maximum squeeze pressure, normally > 100 mmHg, is mostly due to the external anal sphincter muscle function. The anal sphincter function is characterized by the closure pressure generated by the sphincter muscles and by the length of the pressure zone in Figure 18. Measurements of anal sphincter pressures at rest and squeeze at different levels in the anal canal. The resting and maximum squeeze pressures are plotted at the different levels from the anal verge. The areas under the curves at the distance 0 - 5 cm from the anal verge are calculated by computer.

(35)

35 the anal canal. Anal continence is dependent on sufficient closure pressure generated by the sphincter muscles in conjunction with a sufficient length of the pressure zone, i.e. the functional length of the anal canal. The area under the curve of the pressure-distance plot curve is considered to be a more adequate measure of anal sphincter function than pressure measurement alone (Fig. 18)(Hallböök et al. 1995).

In the present study the ARM measurements were conducted according to the method described by Sundblad et al (Sundblad et al. 1993) using the stationary pull-through technique. The equipment used was ABB Goerz. Metrawatt SE 120, with microtip transducer (MTC ®; MMS ups 2020, Enschede, NL) and the computer program software MMS Version 8.3; Windows ™ (version 2004 (8.0.4) ©4D SA, 1985-2006; 4D Sweden AB, Isafjordsgatan 36, 164 40 Kista).

Before the manometry the patient had used an enema to clean the rectum. The patient was positioned in the left lateral position, and the catheter was gently inserted 8 cm in the rectum. Following equilibration, the pressures were measured at rest, while the patient performed a single maximal squeeze effort, followed by a period of rest. The measurements were repeated at 6 cm and at five subsequent stations by 1 cm intervals, as the catheter was progressively moved in caudal direction. The maximal anal resting pressure (MARP) and squeeze pressure (MASP) were determined and registered. From the graphic presentation of the pressure-distance plot the area under the curve at the distance zero to five centimeters was automatically calculated by the software Kalkylations 4th Dimension, 4-D Runtime Interpreted™ according to the method described by Hallböök et al (Hallbook et al. 1995). Consequently values of the anal resting pressure area (ARPA) and anal squeeze pressure ASPA) area were derived. All ARM examinations and readings were done by one specially trained assistant nurse.

Prolapse surgery

(Paper IV and V)

The surgery was performed under general or spinal anesthesia. The POP surgery should include a posterior repair. Surgery in the anterior and apical vaginal compartments was performed concomitantly if considered necessary by the surgeon depending on the pre- and peroperative status and symptoms. The colporrhaphies, posterior and anterior, were carried out with the two-layer suturing technique; one of the endopelvic fascia and recto vaginal septum using interrupted sutures in the midline plication, and the second by suturing the

(36)

36

vaginal epithelium. Plication of the levator ani muscles and perineorrhaphy was performed if the attachments to the perineal body were considered by the surgeon to be disrupted and/or if the perineal body was deficient, respectively. Only resorbable sutures were used in the colporrhaphies. Meshes and implants were not used.

Statistical methods

Statistical evaluations were accomplished by means of non-parametric tests. For comparisons of unpaired data on ordinal scale odds ratio (OR) with 95 % confidence interval (CI), Yates corrected _{-test or Fishers exact test were calculated when appropriate. For continuous data} Mann-Whitney U-tests were used. For comparison of paired ordinal data the McNemar test was used and for continuous data the Wilcoxon Signed Rank test was applied.

Analysis of variance (ANOVA) or analysis of covariance (ANCOVA) tests were used to determine associations between effect parameters and independent variables. Subsequent post hoc testing wasdone with Fisher's PLSD test.

Multiple logistic regression (MLR) analyses were used to identify predictive factors. Adjustments were carried out for known and potential confounders when appropriate. Results of MLR are presented as OR and 95%CI.

A difference was considered significant when p < 0.05. As a consequence of multiple testing, the Bonferroni adjustment can be applied ad lib.

(37)

37

Results and comments

Shift in surgical treatment

(Paper I)

During the two time periods, 542 women were treated with primary POP surgery, 261 in the period I (1983) and 281 in period II (1993). The demographic and clinical preoperative data did not differ significantly between the two periods.

The proportion of POP surgery with complete repair was significantly lower in period II compared with period I (36% vs. 69%). It was especially evident that the use of posterior colporrhaphy had significantly changed. Eighty-six per cent of the women in period I had a posterior colporrhaphy and 58% in period II.

The results indicate that a shift from complete repairs to selective repairs occurred between the time periods and that it was the use of posterior colporrhaphy in particular that was reduced. Thus it seems that Swedish gynecologists complied with the recommendations generated from the reports published in the 1980s expressing concern about the troublesome side effects of complete repairs and posterior repairs.

It seems that this has continued to evolve. Data from the National Board of Health and Welfare in Sweden show a continuing reduction of the use of complete repairs, especially the Manchester repair, from 1998 and onwards as presented in Fig. 19.

Type of POP surgery in Sweden

0 500 1000 1500 2000 2500 3000 3500 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 n Anterior repair Posterior repair Manchester repair Complete repair + vaginal hysterectomy

Figure 19. Number of patients with POP surgery in Sweden 1998-2007 (data from the National Board of Health and Welfare in Sweden (http://www.socialstyrelsen.se/Statistik/statistikdatabas/index.htm)).

(38)

38

Subsequent POP surgery (retrospective study)

(Paper II)

In all, subsequent POP surgery was performed on 29 of the 542 women (5.4%) within the follow-up period of six years. The prevalence of subsequent POP surgery after primary POP surgery in the two periods differed significantly. Of the women operated on in period I, 2.7% (7/261) had subsequent POP surgery and of those operated on in period II, 7.8% (22/281) had subsequent POP surgery. This increase was mainly due to an increase in the use of complementary procedures. None of the women from period I had a complementary procedure whereas 5% (9/180) of the women in period II had a complementary procedures. No significant difference was observed in occurrence of recurrent or combined procedures between the two periods.

Subsequent POP operations were performed significantly more often in women who had selective repairs as the primary surgery than in women who had complete repairs (7.7% (20/261) vs. 3.2% (9/281)).

Women treated with a posterior colporrhaphy either alone or combined with an apical repair were also more likely to have subsequent POP surgery than those who had neither of these procedures.

Neither the size of the prolapse, nor the specific vaginal compartment that dominated the prolapse were found to be associated with the occurrence of subsequent POP surgery.

The associations between the de facto surgery of the specific vaginal compartment and occurrence of subsequent recurrent or complementary POP surgery are shown in Table 4. No significant differences were seen in occurrence of recurrent or complementary POP surgery between the various vaginal compartments.

Table 4. Frequency and type of subsequent POP surgery in relation to vaginal compartment. Vaginal compartment

Type of subsequent POP surgery Anterior Apical Posterior Recurrent 6/488 (1.2%) 11/404 (2.7%) 11/385 (2.9%) Complementary 5/54 (9.3%) 5/138 (3.6%) 9/157 (5.7%) The denominator in “Recurrent” expresses the number who had an operation in the compartment at the time of primary surgery. In “Complementary”, the denominator corresponds to the number who was not operated on in the specific compartment at the time of primary surgery. The numerator reflects the number of women with subsequent surgery in the specified compartment.

(39)

39 The prevalence of subsequent POP surgery found in this study (5.4%) is low compared with findings from other studies where the prevalence is found to be up to 33% (Benson et al. 1996; Moghimi et al. 2005; Tegerstedt et al. 2004). The rate of subsequent POP surgery increases with follow-up time. The time interval between the first and the second operation was on average 12.5 years in the study by Clark (Clark et al. 2003). In my study, the follow-up time was six years. This may be a part of an explanation for the relative low prevalence. As seen in Fig. 20 the cumulative relative incidence of subsequent surgery in the retrospective material of this thesis continues to increase after the sixth postoperative year.

Subsequent POP surgery

0 2 4 6 8 10 12 14 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Years after primary surgery

Complete repair period I Selective repair period I Complete repair period II Selective repair period II

Figure 20. The cumulative relative incidence (%) of subsequent POP surgery in patients with complete and selective repair from period I and II beyond the 6-year follow-up in period I.

Complete repairs were encumbered with significantly higher complication rates than selective repairs according to information from the patient records. In particular POP surgery including posterior repair seemed to be more prone to higher complication rates than POP surgery without posterior repair as seen in Table 5.

Table 5. Associations between the extent of POP surgery and complications

Complete repair vs. selective repair posterior repair vs. no posterior repair POP surgery including

No. of women with

Complete repair (n = 281) Selective repair (n = 261) and (95% CI) OR Posterior repair (n = 385) Without posterior repair (n = 157) and (95% CI) OR Any complication 37% 20% 2.17 (1.48 – 3.18) 36% 12% 4.15 (2.46 – 7.00) Bleeding/hematom 10% 4% 2.67 (1.27 – 5.63) 9% 1% 8.49 (2.02 – 35.7) Wound infection 6% 3% 1.69 (0.73 – 3.90) 6% 0.6% 10.5 (1.40 – 78.1) UTI 21% 12% 1.82 (1.40 – 2.90) 20% 9% 2.35 (1.31 – 4.24). Blood transfusion 4.6% 0.8% 6.28 (1.40 – 28.1) 3.6% 0.6% 5.94 (0.77 – 45.6) UTI = Lower urinary tract infection

(40)

40

Anatomical outcome of POP surgery

(Paper V)

The POP surgery performed upon the 42 women in the prospective study is shown in Table 6. Table 6. Type of POP surgery in 42 women.

Type of POP surgery Number and (%) Complete repair: Manchester repair (12)

Including vaginal hysterectomy (7) 19 (45%) Anterior and posterior colporrhaphy 8 (19%)

Posterior colporrhaphy 12 (29%)

Posterior colporrhaphy and apical repair 3 (7%)

The long-term anatomical outcome median 6.4 years after the primary surgery is presented in Table 7. Overall, in 71% (30/42) of the cases, the surgery had improved the preoperative stage of prolapse in at least one stage. Still, 67% (28/42) had stage 2 or 3 prolapse at long-term follow-up. Subdivided according to the vaginal compartment, the outcome for the anterior compartment was worst with 50% (21/42) having stage 2 or 3 cystocele at follow-up. This was mainly attributed to the group who had had anterior colporrhaphy. That group had cystocele stage 2 – 3 in 58% (15/26). In the group without anterior colporrhaphy the corresponding figure was 37.5% (6/16). The apical compartment was anatomically well supported in all patients at follow-up with no stage 2 or 3 prolapse. The outcome of the surgery in the posterior compartment showed stage 2 and 3 prolapse in 26% (11/42). These results are in accordance with previously reported papers (Fialkow et al. 2008; Jeon et al. 2008; Kahn et al. 1997; Tegerstedt et al. 2004) showing that the surgery improves the vaginal anatomy but emphasize the “fragility” of the anterior vaginal compartment after POP surgery.

(41)

41 Table 7. Distribution of POP-Q stage preoperatively and at follow-up median 6.4 years later. Sub-classification according to whether or not surgery was performed of the anterior and apical vaginal compartments is included.

Preop

(no. with) Follow-up (no. with)

POP-Q staging (N = 42) Stage 0-1 Stage 2 Stage 3

Overall POP-Q stage

Stage 0-1 0 --- --- ---

Stage 2 15 6 7 2

Stage 3 27 8 16 3

Anterior compartment POP-Q stage

Stage 0-1 Anterior colporrhaphy 1 1 0 0

No anterior colporrhaphy 15 10 4 1

Stage 2 Anterior colporrhaphy 10 5 4 1 No anterior colporrhaphy 1 0 1 0 Stage 3 Anterior colporrhaphy 15 5 7 3 No anterior colporrhaphy --- --- --- ---

Apical compartment POP-Q stage

Stage 0-1 Apical repair 8 8 0 0

No apical repair 20 20 0 0

Stage 2 Apical repair 9 9 0 0

No apical repair --- --- --- ---

Stage 3 Apical repair 5 5 0 0

No apical repair --- --- --- ---

Posterior compartment POP-Q stage

Stage 0-1 11 8 3 0

Stage 2 21 14 6 1

(42)

42

Symptoms of pelvic floor dysfunction

Paper III and IV

The prevalence of PFD symptoms preoperatively and the changes after primary POP surgery are shown in Table 9.

Table 9. Preoperative symptoms in 42 patients and changes of these at postoperative follow-up

Preop Outcome of symptom at follow-up No. of women

Symptom No.* Resolved completely or improved# Persistent or worse# _{De novo}# p- values* Prolapse symptoms Vaginal protrusion 23/42 18/22 4/22 1/18 0.0002 Pelvic pressure or heaviness 29/42 19/28 9/28 2/12 0.0005

Bowel symptoms

Problems with emptying the bowel 17/37 13/16 3/16 2/19 0.0098

Splinting 11/37 9/11 2/11 2/24 0.0704

Feeling of incomplete bowel emptying 16/37 9/14 5/14 1/21 0.0269 Fecal incontinence

Gas 18/36 9/17 8/17 3/17 0.1489

Liquid 5/36 1/4 3/4 1/30 0.4795

Solid 1/36 0/1 1/1 1/33 1.000

Urinary symptoms

Difficulty starting voiding 6/35 6/6 0/6 0/27 0.0412 Feeling of incomplete bladder emptying 12/37 8/12 4/12 2/23 0.1138 Urinary incontinence 8/36 3/8 5/8 3/26 0.6831 All questions were not answered by all women. The number of affirmatives for each question is given in the denominator.

# Persistent means that the symptom is still present but has either improved, but not significantly, is unchanged or has become worse. De novo means that the symptom was not reported preoperatively but developed during follow-up period and was reported as occurring at the follow-up. Improvement means that the symptom still exists but has changed from occurring weekly or more often to occurring less than once a week.

* McNemar test with continuity correction; df = 1.

There were significant improvements in the symptoms of prolapse, problems with bowel emptying, feeling of incomplete bowel emptying and difficulty starting voiding but not in splinting, feeling of incomplete bladder emptying and urinary - and fecal incontinence. These results agree to a great extent with those of other published studies (Kahn et al. 1997; Maher et al. 2004; Mellgren et al. 1995; Sloots et al. 2003).

(43)

43 The prevalence of the subjective symptoms of PDF six years after primary POP surgery in the retrospective study and the prospective study are listed in Table 10. For most of the symptoms there were striking similarities in the figures. Age, BMI and follow-up period but not parity were comparable in the two studies.

Table 10. The prevalence of PFD symptoms reported in the questionnaire at follow-up 6 years postoperatively in the retrospective and prospective studies.

PFD symptom Retrospective study (N = 188) Prospective study (N = 42)

Prolapse symptoms

Vaginal protrusion 18% 12%

Pelvic pressure/heaviness 36% 26%

Bowel symptoms

Problems with emptying the bowel 19% 13.5%

Splinting 10.5% 11%

Feeling of incomplete bowel emptying 24% 16%

Fecal incontinence Gas 38% 36%

Liquid 13% 11%

Solid 4% 5.5%

Urinary symptoms

Difficulty starting voiding 7% 0%

Feeling of incomplete bladder emptying 28% 16% Urinary incontinence episodes ≥ weekly 40% 22% PFD = pelvic floor dysfunction

(44)

44

The distribution and combination of three main symptoms of PFD – vaginal bulge, urinary- and fecal incontinence- is illustrated in Fig. 21. Only a very small proportion of the women presented with the three main symptoms concomitantly.

Figure 21. Distribution of three main symptoms of PFD in women six years after primary POP surgery. (Reprinted with permission from Acta Obstetricia et Gynecologica Scandinavica.)

Factors associated with symptoms of PFD, POP and outcome of surgery

(Paper III, IV and V)

The statistically significant associations between the anorectal manometry findings and the prolapse and bowel symptoms are shown in Table 11. The mean value ± SD is given for each group, i.e. the group of women who has the symptom and those who do not, and the p- value of the subsequent comparison of the group means by the Fishers´ PLSD test is specified. No significant associations were otherwise observed between the anorectal manometry measurements and any of the other investigated symptoms of PFD or the size and extent of the POP.

(45)

45 Table 11. Statistically significant associations between the symptoms vaginal protrusion and feeling of incomplete bowel emptying, respectively and anorectal manometric findings

Anorectal manometry Group with the symptom Group without the symptom Analysis of covariance* Fishers´ PLSD test Symptom measurement Mean ± 1 SD Mean ± 1 SD p-value p-value

Vaginal protrusion MARP 42 ± 14 67 ± 31 0.0033 0.0013 MASP 84 ± 27 125 ± 55 0.0039 0.0037 ARPA 125 ± 46 193 ± 90 0.0074 0.0023 ASPA 250 ± 78 356 ± 156 0.0047 0.0073

Feeling of incomplete bowel MASP 122 ± 59 82 ± 26 0.0107 0.0102

emptying ASPA 332 ± 166 247 ± 76 0.0371 0.0489 *Adjusted for age and BMI.

No previous studies dealing with associations between anal pressures and symptoms or extent of POP have been reported. This study showed that in particular the symptom “vaginal protrusion” was significantly associated with a weaker anal sphincter muscle. In contrast to this, the symptom “feeling of incomplete bowel emptying” was associated with a stronger anal sphincter muscle. This may indicate that this symptom reflects an obstructed defecation and not an impaired muscle function as suggested by Klingele et al. (Klingele et al. 2005).

At long-term follow-up none of the PFD symptoms were significantly associated with the postoperative POP-Q stages, overall or compartmental. In contrast, urinary incontinence and feeling of incomplete bladder emptying were significantly associated with parity, BMI, and age (data not shown).

The anorectal manometry measurements deteriorated significantly after POP surgery as shown in Table 12.

Table 12. Anorectal manometry measurements preoperatively and at repeated measurement, median 3.6 years after.

Preoperative Follow-up

Measurement Median and (range) Median and (range) p – value*

MAPS (mmHg) 42 21 - 122 43 14 - 75 0.0114

MAPS (mmHg) 95.5 40 - 260 87 29 - 166 0.0125

ARPA (mmHg * cm) 131 61.5 - 370 111.5 42 - 256 0.0135 ASPA (mmHg * cm) 276.5 123 – 717.5 233 94 – 408.5 0.0019 *Wilcoxon Signed Rank test.

(46)

46

The anal sphincter pressures are age dependent (Åkervall et al. 1990). The deterioration in pressures observed in this study was more pronounced than expected for the influence of the ageing factor alone. This might imply that the prolapse surgery may influence the muscle function of the pelvic floor or that pelvic floor muscle function due to a pre existing damage is more prone to deteriorate in women with prolapse than in normal women. The results of anorectal manometry studies in rectocele repairs are contradictory (Maeda et al. 2003; Mellgren et al. 1995; Yamana et al. 2006). The results in this study are in accordance with those of Ho et al. and van Dam et al.(Ho et al. 1998; van Dam et al. 2000) Mellgren et al. in a Swedish study of 25 women found an increase in anal resting pressure (Mellgren et al. 1995). Contrary, Nieminen et al. found no changes in anal pressures (Nieminen et al. 2004). The surgical approach of the rectocele repair differs in these studies, which may explain the diversity of the results.

Statistically significant associations were found between the neurophysiologic findings and the symptoms of PFD and stage of the POP. However, the post hoc testing with Fisher's PLSD test revealed no statistically significant differences between group means in any of the significant associations. These results indicate that there are associations between neuromuscular damage of the pelvic floor muscles and symptoms of PFD and extent of POP. The associations were heterogeneously distributed in the muscles and it was not possible to determine discriminatory values of individual neurophysiologic characteristics to predict the symptom or stage of the prolapse. This may support the theory about the multifactorial pathophysiology of POP (Dietz 2008; Swift et al. 2001).

Predictive clinical factors for symptoms of PFD at long-term follow-up after primary POP surgery are shown in Table 13.

Genital prolapse surgery A study of methods, clinical outcome and impact of pelvic floor muscle function