Gynecologic Surgeries for Pain

Pain is the clinical indication for 40% of laparoscopies¹ and approximately 12% of hysterectomies.² Although pain has been attributed to virtually all forms of pelvic reproductive pathology, there is no gynecologic pathologic entity that causes pain in everyone who has it. The decision to operate therefore represents delicate clinical judgment in many instances. Because evaluation methods are subjective and imprecise, outcomes are variable.

The purpose of this chapter is to review procedures commonly done for pelvic pain from the perspectives of diagnosis, surgical procedure, and outcome. I discuss newer techniques such as pain mapping for evaluating the somatic and visceral contributions to pain. However, the discussion assumes that complete clinical assessment of contributions from other organ systems and evaluation of psychosocial factors has already taken place. This chapter reviews treatment of postoperative adhesions, retained ovaries, ovarian remnants, vaginal apex revision, presacral neurectomy, and uterovesical ganglion excision.

As described elsewhere in greater detail,^3,4 most problems of pelvic pain are diagnosed on the basis of a careful history. Attention to the details of the pain complaint, including a careful chronologic review of the accrual of new aspects of the pain, changes in its intensity, and the time course of the pain over the day and work week, often yield a relatively short list of diagnostic possibilities.

In terms of examination technique, a few general points are worth making. When evaluating a pelvic pain problem, the examiner should be willing to think flexibly about the sequence of the components of the examination. For example, routine insertion of the speculum before digital palpation sometimes serves only to evoke substantial discomfort, making the patient more reactive to subsequent digital examination and confusing the diagnostic interpretation. For example, if the history suggests that levator spasm may be present, I often palpitate the pelvic floor musculature before speculum insertion. It is wise to examine the areas in which you anticipate less pain before moving on to more sensitive areas. The accuracy of pain reports can be improved by encouraging the patient to focus on what she is experiencing at a given moment, rather than allowing her to look ahead with anxious anticipation to what will be painful in future parts of the examination.

It is often difficult to distinguish internal visceral or reproductive organ components of pain from more somatic or myofascial components emerging from the pelvic floor or the abdominal wall. To try to distinguish these sources, separately employ the vaginal and abdominal hand components of the examination before merging them in the standard bimanual approach. Transvaginal palpation of the uterus, broad ligaments, and adnexal areas can be accomplished without the addition of the abdominal hand for the purpose of identifying a general area from which pain signals emerge. The vaginal hand can then retreat, allowing a repeat examination of these same areas with the abdominal hand only. The examiner should ask which technique evokes discomfort most closely resembling the pain that she usually experiences. The two examining hands can then be joined in the standard bimanual examination and the diagnostic impression formed. When this technique fails to adequately clarify the diagnosis, conscious pain mapping using microlaparoscopic techniques under conscious sedation can be helpful.⁵

Performing hysterectomy for the treatment of chronic pelvic pain has always been a controversial subject. The uncertainties surrounding this procedure stem from several sources. Most cautionary reports are derived from referral center practices in academic medical centers.^6,7,8 Patients following this care path are more likely to have complications in terms of overlapping physical, multisystem, and psychosocial elements. Reports from these populations report usually “failure” rates of 25% to 50% for hysterectomy for chronic pelvic pain. In contrast, reports derived from nonreferred primary care practices are more optimistic.^9,10 In these studies, when reevaluated at 1 year after surgery, more than 90% of women who had hysterectomy for pelvic pain were satisfied. It appears that the primary care gynecologists in that study made accurate judgments concerning the general indications for hysterectomy. In contrast, academic centers derive a greater proportion of their practice volume from women referred because of the chronicity of their pain and medical or psychosocial complexity. Success rates would be expected to be lower for this group.

The traditional psychoanalytic interpretation of some pain complaints may be another factor that colors the approach to hysterectomy for chronic pelvic pain. If evaluation does not reveal concrete pathology with an obvious relationship to pain, the analytic interpretation may suggest that the pain is an outgrowth of internal conflict and hence is likely to reside in a different organ system after the hysterectomy has been performed. This phenomenon has been called symptom substitution. Many valid clinical examples of this phenomenon exist, but in the Maine Women's Health Study,⁹ only 8% of women undergoing hysterectomy for pelvic pain had experienced new unexplained symptoms during the subsequent year. Given the very high proportion of complaints seen in the general primary care office that are functional or stress related, this would seem to be a very reasonable figure.

One weakness of the Maine Women's Health Study is that the type and degree of pathology seen in patients undergoing hysterectomy for pain was not well documented. However, examination of pathologic or histologic findings have never been very informative in explaining the intensity of pelvic pain. The most common forms of pathology associated with pelvic pain, such as adhesions, endometriosis, pelvic congestion, adenomyosis, or alteration of pelvic support or uterine positioning, do not lend themselves to ready pathologic quantitation.

A further weakness in the general literature on hysterectomy for pelvic pain is the absence of structured assessment of the contributions of other organ systems to pelvic complaints. In women undergoing hysterectomy in which no pathology is found, a higher than expected proportion qualify for the diagnosis of irritable bowel syndrome.¹⁰ By the same token, much emphasis is being placed on the muscular components of pelvic pain, especially pelvic floor levator spasm.¹¹ In the “failures” seen in previous reports of hysterectomy, it is impossible to tell the degree to which these other components were evaluated.

Ovarian function can be completely suppressed with GnRH agonist medications. This would seem to eliminate reproductive tract function from the clinical environment and thereby provide a way to evaluate the contribution of the reproductive organs to the pain syndrome. However, gonadotropin-releasing hormone (GnRH) agonists may also alter gastrointestinal motility¹² and have even been investigated as possible therapeutic agents for severe irritable bowel syndrome. I have had clinical experience supporting this notion and therefore remain unconvinced that improvement of pain on GnRH agents proves that the pain is solely of gynecologic origin. Conversely, failure of pain to improve on GnRH agonists does not invariably prove that pain associated with, for example, known endometriosis is not of gynecologic origin. In the former case, the medication may be treating irritable bowel components. In the latter case, a more aggressive laparoscopic surgical approach may be necessary.

More complete descriptive reports are needed that encompass multisystem evaluation, clinical and surgical approaches, pathologic findings, and outcomes. These reports would allow generation and further testing of hypotheses regarding appropriate treatments for the different components of pelvic pain and the relative value of operative intervention such as hysterectomy.

Adhesions are more prevalent in women with chronic pain.^13,14 Contradictory reports have often used inappropriate control groups such as infertility patients, who may self-select for being pain free.¹⁵ Acute problems such as intestinal obstruction can occur in as many as 0.3% of benign adnexal gynecologic surgeries, 3% of hysterectomies, and 5% of radical hysterectomies.¹⁶ Autopsy studies reveal a 30% prevalence of pelvic adhesions in women who have never had pelvic surgery.¹⁷

The mechanism of pain production in women who do have pelvic adhesions is uncertain. Some adhesions seem to contain neuronal fibers, although the association of these fibers with pain is imprecise at best.¹⁸ More information concerning this association may appear in the future as a result of studies done using conscious pain mapping techniques and histologic assessment of adhesions.

The presence of adhesions is difficult to diagnose. By history, patients with significant adhesive disease may often describe that their pain is somewhat positional, such as being aggravated by assuming a sleeping position on one side or the other. The physical examination is inadequate to identify the presence and location of adhesions.¹⁹ Similarly, standard vaginal ultrasound, including observations of “slide-by” of adnexal structures compared with intestines, is not an accurate predictor of endoscopic findings.

Results of studies of the treatment of pelvic pain by adhesiolysis are somewhat encouraging. Only one controlled study has been performed. Peters and colleagues²⁰ performed diagnostic laparoscopy on 48 women and then randomly assigned one half of them to adhesiolysis by laparotomy and the other half to standard clinical management only. They found no difference in pain levels found between the two groups at 9 to 12 months postoperatively, except for improvement in those with severe bowel adhesions at the initial diagnostic laparoscopy. Unfortunately, a second-look laparoscopy to evaluate the regrowth of adhesions after the laparotomy was not performed.

Other studies have suggested that the laparoscopic approach to adhesiolysis results in less adhesion reformation than adhesiolysis by laparotomy.²¹ In general, the adhesion burden is reduced by approximately 50% after laparoscopic treatment.^22,23 Direct randomized comparison of laparoscopy with laparotomy for the treatment of pelvic adhesions associated with pain has never been performed.

Clinical series^24,25 suggest that 70% to 85% of women with chronic pelvic pain may obtain relief after laparoscopic adhesiolysis. The success rate may decline to less than 50% for those with complicating psychosocial factors described as a “chronic pain syndrome.”²⁵ Conscious pain mapping using microlaparoscopic techniques has been employed to obtain a better understanding of the visceral contributions to pelvic pain. Studies using this method may provide further information about the role of various types of adhesions in producing chronic pelvic pain. I expect to find that dense adhesions around the adnexal areas or dense adhesions of the small bowel to the pelvic floor or pelvic viscera may be most consistently associated with clinically important pain. A more complete discussion of this topic is provided elsewhere.²⁶

A generation ago, gynecologists almost uniformly recommended bilateral oophorectomy at the time of hysterectomy for any woman undergoing the procedure who was older than 40 years of age. Because of generally changed treatment philosophies and patient wishes and because of the development of operative laparoscopy, the age at which oophorectomy is recommended has increased, and most texts refrain from offering a purely numeric recommendation. The variables to be considered in this decision include the patient's age, obesity, past surgical history, risk of ovarian cancer, history of past pelvic infectious disease, presence of endometriosis, and her own personal preferences.

A major reason for recommending oophorectomy at the time of hysterectomy was that reoperation for ovarian problems occurred in approximately 1% to 5% of women who chose ovarian conservation at the time of abdominal hysterectomy.²⁷ When this problem required laparotomy, treatment entailed substantial operative and postoperative morbidity. Operative laparoscopy substantially reduces that morbidity, altering the decision parameters at the time of hysterectomy.

A “retained ovary” is defined as an ovary left behind after previous hysterectomy that has subsequently become symptomatic by virtue of intrinsic pathology, periovarian adhesions, or its location close to the vaginal apex. Associated symptoms typically include unilateral cyclic pelvic pain and deep dyspareunia. Variations occur, including more continuous discomforts and associated bowel dysfunction. In such cases, it is difficult to separate the contributions of general pelvic adhesive disease from those for which the ovaries are responsible.

Traditional surgical approaches to vaginal hysterectomy may contribute to this problem. Many versions of this procedure include suturing the utero-ovarian ligament into the vaginal cuff for purposes of augmenting vaginal apex support and extraperitonealizing any ligated or vascular pedicles.²⁸ This practice seems to have abated in recent years; it often resulted in the ovary being densely adhered to the vaginal apex, resulting in postoperative dyspareunia.

The traditional recommendation for women with endometriosis included bilateral oophorectomy at the time of hysterectomy when that procedure was performed. The literature suggests that the probability of requiring further abdominal surgery is approximately sixfold higher when one or both ovaries are left behind at the time of hysterectomy for endometriosis.²⁹ Although this is generally true, many surgeons are reluctant to perform bilateral oophorectomy when the degree of disease is minimal. A reasonable judgment may be to leave one ovary behind in the absence of overt disease in that ovary, provided complete surgical resection of visible disease is possible. Later operative intervention sometimes may be necessary, but it can almost always be accomplished laparoscopically. Other variables in this decision include the age of the patient, the pace of progression of endometriosis and her symptoms, and her wishes in terms in terms of ovarian conservation, hormonal replacement, and other issues. Because the behavior of endometriosis varies over time, it would be most informative to develop chronologic series in which ovarian conservation was practiced, with results stratified by degree of endometriosis present initially.

The diagnosis of the retained ovary syndrome is made primarily on the basis of clinical history and bimanual examination, sometimes supplemented by transvaginal ultrasound. In many cases, the final diagnosis cannot be made until laparoscopy is performed.

The typical history for retained ovary syndrome records a good level of postoperative comfort until approximately 4 to 12 months after hysterectomy. During this interval, the cyclicity of ovarian function may be disturbed, and the timing of exacerbations of pain may be somewhat variable. Words used to describe the pain typically include “dull” or “aching” with some sharp exacerbations. Especially when the left ovary is involved, alterations of bowel function can accompany the process, including increased pain with defecation. Constipation may aggravate the problem.

Bimanual examination often localizes pain to the area of either ovary. When the ovary is surrounded by adhesions, it may become progressively more difficult to outline and move. During the examination, transvaginal palpation of the ovary without the abdominal hand may replicate the patient's dyspareunia.

Transvaginal ultrasound is sometimes used to estimate the proximity of the ovary to the vaginal apex. However, I have observed multiple false-positive results using this approach, probably because the transvaginal ultrasound probe can be manipulated over an extensive range, giving the false impression of proximity to the ovary.

When the symptoms associated with the ovary are caused by intrinsic ovarian function or dysfunction, suppression of the ovary may be useful diagnostically and therapeutically. Cyclic or continuous birth control pills or GnRH agonists may be employed. When GnRH agonists are used, they may reduce symptoms arising from irritable bowel disorder, confusing the diagnostic impression. In some cases, when medical therapy can be successfully employed for approximately 1 year after surgery, ovarian cyclicity may return and the degree of discomfort diminish. During this time, some couples are successful in adjusting coital technique to avoid the dyspareunia.³⁰

The surgical approach to the retained ovary is usually salpingo-oophorectomy. In the past, when laparotomy was needed to accomplish the procedure, the standard was removal of the innocent and the symptomatic ovary. The availability of laparoscopic techniques may allow exception to this rule. The soundness of leaving one ovary in place has not been systematically investigated. I make this decision with the patient based on her general level of concern about ovarian hormone replacement, the potential need for future surgery, and the difficulty of her past surgeries.

It is sometimes tempting to consider ovaripexy—mobilizing the ovary from the pelvic floor or the vaginal apex and suturing it to the pelvic sidewall or to the more lateral portion of the round ligament. The literature on this approach is scanty in terms of assessing postoperative comfort levels, and my experience with this technique is most disappointing.

A residual ovary is a normal ovary intentionally left in place. An ovarian remnant is diagnosed when both ovaries have ostensibly been removed at the time of previous surgery, but true ovarian tissue reemerges over time. Once felt to be uncommon, recent series suggested that this problem was more common than previously thought.^31,32 In most situations, it arises when the initial hysterectomy and oophorectomy surgery was difficult because of severe pelvic adhesive disease or endometriosis. A small fragment of ovarian tissue may be left behind in a bed of dense scar tissue. As peripheral levels of ovarian hormone subsequently decline, follicle-stimulating hormone (FSH) levels rise, stimulating the fragment to enlarge over time. Typically, a mass becomes diagnosable and symptomatic 1 to 3 years after the purported oophorectomy.

In its purest form, the ovarian remnant manifests as a cyclically painful, unilateral adnexal mass that is palpable on bimanual pelvic examination and can be visualized by transvaginal ultrasound, magnetic resonance imaging (MRI), or computed tomography (CT) scan. Premenopausal levels of FSH and estrogen demonstrate continued ovarian function.

Unfortunately, many ovarian remnants manifest in a more subtle fashion. I have operated on five histologically proven ovarian remnants that were not hormonally functional. Peripheral estrogen levels were at the upper end of the menopausal range, and FSH levels were between 80 and 90 mIU/ml. A number of other ovarian remnants were found in the presence of FSH levels that were only minimally elevated into the menopausal range.

Ultrasound and MRI studies of ovarian remnants are fallible; ultrasound images especially may be corrupted by the presence of postoperative adhesions involving adjacent loops of bowel or sterile inflammatory cysts. As a rule of thumb, many gynecologists have learned that the diagnosis of an ovarian remnant or other pelvic mass is uncertain when it rests only on the pelvic ultrasound results.

In many clinical situations, it is difficult to determine the relative contributions of the functional ovarian tissue compared with pain emerging from dense pelvic adhesive disease. In this setting, total ovarian suppression by a GnRH agonist may be informative. In cases not involving endometriosis, I use “add-back” estrogen replacement to reduce clinical symptoms. When these manipulations completely eliminate the pelvic pain, the ovarian remnant can be held responsible for the problem. In patients who are very poor operative candidates or who are close to the age of anticipated menopause, medical management in this manner can be continued until the remnant stops functioning. In my experience, monthly injections of GnRH agonists have not always been necessary; I have a number of patients managed successfully on GnRH agonist injections three to six times yearly.

The surgical approach to an ovarian remnant can be made by laparoscopy or laparotomy. The literature suggests a list of 10% to 15% of recurrence of ovarian remnant tissue even after the most careful surgical excision done by a laparotomy.³³ Comparable figures after laparoscopic removal of remnants have not been published. My experience is that the approach can be carried out by laparoscopy successfully when the anatomy is less complicated. After performing a thorough bowel preparation, it is almost always reasonable to start the procedure laparoscopically, with the proviso that laparotomy be carried out if necessary.

The timing of the surgery is important. If GnRH agonist suppression has been used as a diagnostic manipulation, it may succeed in making the remnant ovarian tissue so small as to be very difficult to find at the time of surgery. It is therefore prudent to allow some time to elapse after the last GnRH agonist injection and to reevaluate to be certain that the residual mass is again palpable if not imagable. The remnant may be provoked to enlarge by administering clomiphene (100 mg each day for 5 days), starting 10 days to 2 weeks before the procedure.

Regardless of the surgical approach, ureteral visualization is imperative. When performed laparo scopically, visualization may be aided by placing ureteral stents by means of cystoscopy at the beginning of the procedure. Whether performed laparoscopically or by laparotomy, dissection should be fundamentally the same, beginning with retroperitoneal entry lateral to the remnant, isolation of its vascular supplies (which usually involve the infundibular pelvic circulation), and separation of these structures from the ureter. After the vascular supply has been divided, further sharp and blunt dissection with mobilization of the ureter usually frees the mass from the surrounding tissue. Especially when this surgery involves endometriosis, it can be one of the more difficult benign gynecologic surgeries performed.

A number of techniques may more adequately identify the boundaries of surrounding structures, prevent injury to them, or evaluate them for the possibility of injury. In terms of the urinary tract, these techniques may include an intentional cystotomy over the dome of the bladder with retrograde stent placement, if stents have not been placed preoperatively, and intravenous methylene blue injection to evaluate the possibility of ureteral injury. The dye can be visualized cystoscopically or through the intentional cystotomy. At times, performing the intentional cystotomy may aid the precise definition of the bladder and the adjacent ureter while dissecting off the last portions of ovarian tissue.

Definition of the vaginal apex can be problematic at times. Placement of a stent in the vagina can help. I often place a patient in modified lithotomy position, allowing access to the vagina and the rectum during this type of procedure. This allows standard rectovaginal examination during the surgery, assisting definition of the vaginal apex and the rectum adjacent to remnant ovarian tissue.

Preoperative assessment usually determines the anticipated degree of difficulty of the surgery with reasonable accuracy. When an especially difficult case is anticipated, it should be tackled by the most experienced surgical hands available.

Uncommonly, deep dyspareunia is reported despite an appropriately performed and successfully completed hysterectomy. In the context of a referral pelvic pain practice, I have seen approximately 40 women during the past several years who appear to have pain that originates from the suture line of the vaginal cuff itself. As physicians have become more aware of this phenomenon, the diagnosis is made more frequently, and its true prevalence is open to speculation.

The clinical diagnosis is made by altering the usual technique of the pelvic examination. For this problem, commonly suspected on the basis of clinical history, I usually perform a speculum examination first. Total visualization of the vaginal apex is accomplished without directly touching it with the speculum. Touching the suture line with a cotton-tipped applicator allows mapping of focally tender areas. Typically, the tenderness is localized in one vaginal fornix or the other, although occasionally it can be present along the entire suture line. Observing that complete pain syndrome can be evoked by gentle palpation, it would appear that intraperitoneal pathology, even when present, may not be the culprit.

This diagnostic impression can be confirmed by painting the vaginal apex with a 2% Xylocaine gel and rechecking the same examination several minutes later. If this yields an inconclusive impression, the area can be injected with a 1% Xylocaine solution, buffered with 1.0 ml of 0.9% sodium bicarbonate per 10 ml of 1% Xylocaine. When technically feasible, to diminish the pain of injection itself, I use a 10-ml syringe equipped with a 27-gauge needle.

This diagnostic method may also be therapeutic. This type of injection can sometimes afford days or weeks of relief of the symptom, making the sensitive vaginal apex appear more like a trigger point. Occasionally, serial injections of this type have been successful in relieving the dyspareunia when coupled with suggestions for altering coital technique to limit deep penetration. In a few instances, the precoital application of a small amount of 2% Xylocaine gel through a transvaginal cream applicator has provided sufficient comfort to allow more comfortable intercourse. However, in most cases, obtaining sufficient relief requires enough volume of Xylocaine gel that the male partner is then gifted with temporary genital anesthesia, an unhappy outcome for both partners.

For most patients who do not succeed with these medical treatment methods, I usually offer revision of the vaginal apex. Although I have seen this performed by open laparotomy, it can almost always be accomplished laparoscopically with far less trauma to the patient. A complete description of this procedure is described elsewhere.³⁴ The surgeon places a vaginal stent in the vagina, fills the bladder to identify its margin, and incises the peritoneum adjacent to the bladder to allow blunt and sharp dissection to separate the bladder from the vaginal apex. After a reasonable surgical margin is obtained, a unipolar needle or scissors can be used to make an elliptical incision encompassing the vaginal apex suture line with an adjacent margin of approximately 0.5 cm. Once colpotomy is obtained, the remainder of the incision can be completed with the unipolar needle or with cautery scissors. The specimen is then removed through laparoscopic trocar and the vaginal apex reclosed laparoscopically or vaginally, depending on the particular anatomy. I usually completed the closure laparoscopically, because I think this ensures precise approximation of the vaginal edges. Interrupted or figure-of-eight sutures of 0 or 00 delayed absorbable suture may be employed using a regular needle driver or laparoscopic suturing device.

With follow-up ranging between 1 and 3 years, postoperative success has been obtained in all but 2 of 20 cases operated. The 2 failures have had recurrence of pain at the same level as that experienced preoperatively. Pathologic findings have included scarring in most, with about one half of the specimens containing additional findings such as sterile abscesses, residual endometriosis, and fluid-filled inclusion cysts.

Presacral neurectomy is defined as the surgical division of the sympathetic nerve supply traversing the superior hypogastric plexus. This plexus arises from the lower thoracic sympathetic chain, crosses anterior to the common iliac vessels on both sides, and forms a highly variable network in front of the fifth lumbar and first sacral vertebrae. It continues inferiorly to the inferior hypogastric plexus and thence to supply the uterus and adjacent tissues, including a portion of the bladder and the rectum. In this course, some of the fibers also traverse Frankenhauser's plexus. Elant described a variation of this anatomy in a cadaver study published in the early 1930s.³⁵ In addition to anatomic variability over the sacral promontory, it appears that the ultimate destination of these fibers is also variable. The failures of interruption of the nerve supply to cure all forms of central pelvic pain is often attributed to redundant nerve supply emerging from other sympathetic fibers not traveling with this plexus and to parasympathetic fibers emerging from S2 through S4 nerve segments and arriving at the central pelvis through Frankenhauser's plexus.³⁶

Originally described by Jaboulay in 1899,³⁷ presacral neurectomy was popularized by Cotte beginning in the 1920s. The latter investigator ultimately reported more than 1500 cases of successful treatment of central pelvic pain and dysmenorrhea with this procedure.³⁸ These early reports were extremely optimistic, reporting 98% success rates with almost no complications. Later reports by such gynecologic authorities as Pfanneuf³⁹ and Meigs⁴⁰ for more selectively performed procedures found an approximately 5% to 15% incidence of clinically significant constipation and urinary retention.

The procedure was used most frequently for the treatment of dysmenorrhea until the early 1960s. After the introduction of the oral contraceptive pill in the mid-1960s and the development of nonsteroidal inflammatory drugs in the early 1970s, the frequency of this procedure declined further.

In the 1970s and 1980s, presacral neurectomy remained in the surgical armamentarium primarily as a supplement to conservative resection for stage III or IV endometriosis. A number of open clinical trials suggested that the adding presacral neurectomy to conservative resection improved postoperative pain relief^41,42 Two randomized trials of the procedure have been done, both performed with women undergoing conservative resection of stage III or IV endometriosis. The trial reported by Tjaden and coworkers⁴³ described conservative resection by laparotomy in eight women, four of whom also underwent presacral neurectomy. Those who underwent neurectomy obtained complete relief, but none of the women with conservative resection only reported relief of pain. At that point, statistical significance was reached, and the study was terminated by the institutional review board. In a subsequent clinical series of presacral neurectomy with conservative resection, 15 of 17 patients reported pain relief. The second trial is that performed by Candiani and associates,⁴⁴ who reported essentially the same study in 71 patients. In this study, both groups obtained significant postoperative pain relief in terms of dysmenorrhea, pelvic pain, and deep dyspareunia. The addition of the presacral neurectomy apparently provided incremental improvement in central dysmenorrhea only, beyond that already attributed to conservative resection of endometriosis. One year postoperatively, both groups had similar rates of continued pain relief (80% versus 75%).

The added benefit of presacral neurectomy to thorough surgical resection of endometriosis remains in question. I think it should be carefully and selectively applied in the surgical management of chronic pelvic pain. Further work is needed to quantify the impact of this procedure on large bowel motility and bladder function. Sexual dysfunction after presacral neurectomy has not been investigated. If an argument can reasonably be made for preservation of the cervix in the interest of preserving sexual function, the physician must examine the presacral neurectomy patient for the possibility of impaired sexual response.

The uterovaginal ganglion incision is described briefly, if only as a cautionary note. The uterovaginal ganglion is a neurologic structure containing sympathetic and parasympathetic fibers and present near the ascending vaginal artery, ureter, and pelvic sidewall. It is unclear whether this is distinct from or is a part of Frankenhauser's plexus.³⁶ Laparoscopic surgical interruption of this ganglion has been recommended by one investigator.⁴⁵ In her series of 175 patients who underwent bilateral uterovaginal ganglion ablation, two thirds reported significant relief and one third reported improvement, with only a 3% failure rate. However, as part of the diagnostic evaluation, she employed a pudendal nerve latency test that has not been validated and most likely represents evaluation of motor function, rather than sensory. In her report, success was highly correlated with return of this test result to normal values, which is most unusual in nerve conduction studies. Approximately 60% of the subjects had other procedures such as resection of endometriosis performed at the same time as the neuroablative procedure, making it difficult to assess the true outcome. No postoperative urodynamic studies, voiding diaries, or other attempts were made to look for complications. Some patients were reported as having urinary retention postoperatively, requiring permanent catheterization. A study of similar size reported this same procedure being applied for the treatment of “lateral pelvic pain,” otherwise undefined. Similar outcomes were described, although many of the patients undergoing this procedure for relief of bladder complaints also had excision of endometriosis. Potential overlap between these series was not described.

Although this type of selective denervation approach may prove fruitful for some pelvic pain conditions, the potential for complications is significant, and the procedures require careful development under structured research protocol conditions.

Extant organic pathology by no means establishes a connection between the pathology and the clinical pain reported. Thorough history, careful pelvic examination by techniques somewhat different from those usually employed, and laparoscopic pain mapping may in some cases provide important supportive evidence in favor of such a connection or may point to other sources of pain. These additional contributing factors need to be evaluated and treated to a degree possible before embarking on a surgical approach. A woman undergoing an operation for pelvic pain deserves complete discussion of the pros and cons of the procedure, including the potential for failure and the potential for inducing postoperative adhesive disease that may simply replace one painful condition with another. Often, this discussion should include the notion that recovery from pain, especially if present for a long time, may require a rehabilitation approach, rather than the hope that the surgery can eliminate the pain in short order.

The surgical approach chosen should fit the clinical and anatomic situation discovered and the experience and skills of the surgeon. In particular, the degree to which laparoscopic procedures are successful is more often caused by the experience and skill of the surgeon than it is to inherent limitations of the technique itself. The value of the best surgery lies in incorporating the procedure into an overall treatment plan that carefully evaluates and treats as many of the components of a chronic pain problem as possible.