Gynecologic Myomectomy



Uterine leiomyomas are among the most common problems encountered by the obstetrician/gynecologist. They are a frequent cause of pain and abnormal bleeding and are thought to be involved in infertility. Uterine leiomyomas are the most frequent indication for hysterectomy in the US. Many patients suffer from myomas but desire to retain the option for future childbearing or simply want to preserve their uterus. For these women, myomectomy, the removal of the myomas with reconstruction and preservation of the uterus, is an important option.

History of the Procedure: Successful abdominal myomectomy was reported as early as 1845 by brothers Washington and John Atlee in the American Journal of Medical Science. Washington, the older of the two, eventually published his experience with 14 abdominal myomectomies, winning the annual prize essay award of the American Medical Association despite the death of 5 of the patients (Speert, 1980, p. 180). The operation was slow to gain widespread use. In 1875, W.H. Byford gave the Chairman’s address to the American Medical Association Section on Obstetrics and Gynecology and said abdominal myomectomy was “so dangerous and difficult as not to be thought of except in desperate conditions?(Speert, 1980, p. 39). At the turn of the century, abdominal myomectomy had a mortality rate of 40% compared to 6-7% for abdominal hysterectomy. Victor Bonney is credited for advocating and popularizing the procedure in the 1920s (Hutchins, 1995).

Problem: Uterine leiomyomata, or fibroids as they are more frequently but less correctly known, affect upwards of 20% of women of reproductive age. Patients can have a single or numerous myomas. They are benign growths in the wall of the uterus (see Picture 1). Most are small, asymptomatic, and are noted as incidental findings on routine pelvic examination or pelvic imaging studies. When they enlarge, they can cause mass effect causing pelvic pressure or pain or distort the uterine wall or endometrial cavity, causing abnormal uterine bleeding. At times they can cause other problems; for instance, prolapsing through the cervix or being confused for an ovarian mass. They are one of the problems most frequently encountered by the obstetrician/gynecologist. While they also can cause problems in pregnancy, myomas are thought in some patients to be linked to infertility.


Frequency: Uterine leiomyomata occur in 20-25% of women of reproductive age (Reiter, 1994). Some autopsy studies detected the presence of at least one small myoma in 50% of women (Thompson, 1997). Myomas grow in response to estrogen stimulation and regress after menopause. Thus, they are most frequently found in women in their fifth decade of life and are quite rare before age 20 years. Racial differences also have been noted, with myomas diagnosed more frequently in women of color than white women (Kjerulff, 1993). Although quite common, there also is some evidence for a familial predisposition, with one study noting a 2.2-fold increased risk in families with 2 or more affected individuals (Vikhlyaeva, 1995).

Diet and body fat have been implicated as well in myoma risk. Marshall et al in 1998 noted an increased risk of myoma formation with increased body mass index (BMI); however, they did not note any relation to cigarette smoking. Sato (1998) studied Japanese women and noted that women with occult obesity as defined by a BMI greater than 24 and a body fat percentage greater than 30% were at elevated risk of surgical treatment for myomas, as were women with increased upper-body fat distribution. In a case-control study in Italy, Chiaffarino and coworkers (1999) studied diet and myoma formation and noted increased risk with increased red meat or ham consumption. Increased consumption of green vegetables seemed to be protective.

Most myomas are asymptomatic and of no consequence. Nonetheless, many women have significant symptoms from myomas, and myomas are the most frequent indication for hysterectomy in the US. This indication constituted approximately 33% of all hysterectomies (223,000) in 1993 (Lepine, 1997). Figures for myomectomy are older, but in 1984 18,000 myomectomies were performed compared with 112,000 hysterectomies, suggesting about 1 myomectomy per 6 hysterectomies among women aged 15-44 (Greenberg, 1995).

Etiology: Leiomyomas are smooth muscle tumors that form in the uterine wall. The precise etiology is not known, although many cytogenetic and genetic studies have been performed. Some 40-50% of myomas have karyotypic abnormalities, particularly involving chromosomes 6, 7, 12, and 14. Within a myoma, all cells are identical, and a monoclonal origin has been confirmed. In patients with multiple myomas, different karyotypes are noted suggesting that each myoma arises as an individual event (Ligon, 2000). Other changes noted include increased expression of estrogen, progesterone, and insulin-like growth factor (IGF) I and II receptors, as well as abnormalities in the myometrium adjacent to the myoma (Tiltman, 1997).


Pathophysiology: Although myomas are common, relatively few actually cause symptoms. Whether symptoms are present depends largely on a combination of size, number, and location of the myomas. In general, myoma growth is a result of the stimulation of estrogen, which is present until the menopause. Over time, previously asymptomatic myomas may grow and become symptomatic. Conversely, many myomas begin to shrink as menopause removes the estrogen stimulation, and many myoma-related symptoms resolve spontaneously shortly after the menopause.

Myomas generally are categorized by location.

Pelvic pressure and pain symptoms usually are the result of mass effect. This can occur either from a single large myoma or from a combination of multiple smaller myomas. A fibroid uterus can grow to be quite large, at times reaching the size of a term gravid uterus. Interestingly, perhaps due to the slow growth and accommodation by the patient, some extremely large uteri are well tolerated by the patient and do not require intervention. Some large myomas that impinge on the ureters can cause hydronephrosis and, very rarely, ureteral obstruction.

Bleeding abnormalities usually are the result of distortion of the endometrial cavity by myomas. Unlike pain, which usually requires large or multiple myomas, some patients have significant intermenstrual bleeding or menorrhagia from a single small strategically placed myoma. A submucosal myoma sometimes can prolapse through the cervix, which can occur asymptomatically or cause significant bleeding.

Acute pain resulting from myomas is uncommon and usually stems from 1 of 2 possibilities. Some pedunculated myomas can torse, causing the same severe pain as a torsed ovary. Large myomas also can outgrow their blood supply, leading to infarction (degenerating myoma), which can be extremely painful.

Although general agreement is lacking on the mechanism, myomas also are thought to be related to infertility and early pregnancy wastage as well as preterm labor (Wallach, 1995). Possibilities include distortion of the endometrial cavity and abnormal endometrium overlying the myoma.

Very rarely, myomas can be associated with erythrocytosis. This triad of myomatous uterus, erythrocytosis, and restoration and maintenance of normal hematologic values after hysterectomy is called the myomatous erythrocytosis syndrome (LevGur, 1995). A number of etiologies have been hypothesized, but alterations in erythropoietin levels seem likely.

Clinical: Most leiomyomata are small and are asymptomatic. Many are found as incidental findings on obstetrical or gynecologic ultrasound (see Picture 3) or on routine pelvic examination.

However, myomas can cause a number of symptoms. They can cause menstrual irregularities, particularly intermenstrual bleeding or menorrhagia. This bleeding usually begins gradually and progressively worsens as the responsible myoma enlarges. A regular menstrual pattern should be discernible within the extra bleeding. If no regular pattern is noted, an etiology such as chronic anovulation more likely is present. Bleeding from chronic anovulation can be severely compounded if concomitant myomas are present.

Some patients present with progressive pelvic pressure, pelvic pain, or low back pain. Many other possible etiologies for these problems exist. However, if these problems are noted in someone with a medium or large sized uterus (greater than 14 or 15 weeks?size), the myomas likely are contributing. Some fibroid uteri can grow out of the pelvis and into the abdomen, where they can be palpated by the patient. This can be disturbing even when relatively asymptomatic. Some will be visible, distorting the abdominal wall and at times making the patient appear pregnant.

Most myomas grow slowly, and some remain relatively unchanged over prolonged periods. In the past, rapid growth was considered worrisome for leiomyosarcoma. Parker, Berek, and Fu (1994) showed that sarcoma was quite rare (0.27%) even in rapidly growing uteri and that this risk was identical to stable myomas (0.21%). However, most of the patients were premenopausal. Rapid growth in postmenopausal women should be treated with greater suspicion.

In a patient with known fibroids and acute onset of pelvic pain, the diagnosis of degenerating myoma should be considered. This pain generally is acute onset and can be quite severe. The patient also can develop fever and an elevated white blood cell count (without a left shift) that can be confused with infection. On exam, tenderness usually is quite specific and localized to the exact region of the degenerating myoma.

Infertility evaluations usually include determining the presence of myomas, specifically submucosal myomas. Ultrasound, hysterosalpingogram (HSG), sonohysterography, or hysteroscopy frequently are utilized, as submucosal myomas may not be detectable on pelvic exam.

Several unusual presentations also can occur. A prolapsed myoma (see Picture 4) sometimes is found on routine speculum exam. A pedunculated myoma is a possible cause for an adnexal mass.


The American College of Obstetricians and Gynecologists (ACOG) criteria for hysterectomy for leiomyomata (Reiter, 1994) are reasonable for myomectomy, as well. They are as follows:


All of these criteria are directed at relieving symptoms or improving quality of life by decreasing the patient’s concerns. There is no indication for the removal of asymptomatic fibroids. Most of these patients will never develop symptoms, and evidence does not support that earlier intervention improves long-term outcomes. An older argument proposed as an indication for hysterectomy, that the removal of uteri greater than 12 weeks?size improved the ability to detect adnexal masses, also is fallacious or obviated by ultrasound. No evidence exists to show that routine pelvic examination improves outcome in the early detection of ovarian cancer. Similarly, there is no reason to think that removal of large uteri to make adnexa easier to examine would alter long-term ovarian cancer risk, detection, or outcome.

Additional criteria include the patient's interest in preserving her fertility or a personal preference of retaining her uterus. A definite risk exists of myoma recurrence after myomectomy and with it the need for a repeat surgical procedure in the future. If the patient no longer desires to retain her fertility or her uterus, hysterectomy is the usual procedure of choice. Interestingly, a number of women who have completed childbearing still request myomectomy for management of symptomatic myoma. This decision usually is motivated by patient preference and the desire to retain organs. As the short-term risks of myomectomy compare favorably with hysterectomy (Iverson, 1996), and despite the recurrence risk, most patients do not require future surgery, a myomectomy is not unreasonable for appropriately counseled patients.

Myomectomy also is performed frequently in patients with infertility with the presence of fibroids (Vercillini, 1998). Studies supporting myomectomy as a fertility-enhancing procedure are uncontrolled and do not use life table analysis; nonetheless, a number of authors recommend offering myomectomy to infertile women after other causes of infertility have been eliminated (Hutchins, 1995b; Nachtigal, 1989; Verkauf, 1996).

Several papers suggest that patients with fibroids who are undergoing assisted reproductive technology procedures may have lower success rates than patients without fibroids. Stovall (1998) noted decreased fertility rates in patients with any myomas undergoing in vitro fertilization (IVF) or zygote intrafallopian transfer (ZIFT). Eldar-Geva and coworkers (1998) noted decreased success in patients with intramural and submucosal myomas but not in subserosal myomas. Ramzy (1998) noted no change in fertility, but that study specifically excluded anyone with large, submucous, or intramural myomas that distorted the endometrial cavity. It is important to note that no randomized studies document that removal of these myomas improves success rates.


Relevant Anatomy: Leiomyomata usually are confined to the myometrium but can occur in the lower uterine segment, the cervix or project out into the broad ligament. These myomas frequently are the most difficult to remove and are problematic at hysterectomy as well as myomectomy.

Within the uterus, the myomas can be at any level within the uterine wall. Those mostly confined to the wall are termed intramural. Myomas projecting into the endometrial cavity are termed submucosal (see Picture 2), whereas myomas projecting outward from the uterus are subserosal (see Picture 1).

Some myomas are attached by pedicles, which can extend outward into the abdomen and may be confused with adnexal masses or can project into the endometrial cavity and can cause bleeding. Some of these will prolapse through the cervix (see Picture 4).

Contraindications: A number of important contraindications to myomectomy exist. Myomectomy is not reasonable in the management of symptomatic leiomyomata in patients who no longer desire fertility or uterine preservation. It should not be performed when there is a possibility of endometrial cancer or uterine sarcoma. It should be avoided in pregnancy. With the possible exception of otherwise unexplained infertility, it should not be performed in asymptomatic patients. There is no evidence whatsoever to support prophylactic myomectomy of asymptomatic myomas in decreasing the risk of any adverse outcome later in life.

Relative contraindications include strong suspicion that a functional uterus could not be reconstructed after excision of the myomas. For myomectomy to be successful, it must be possible to reconstruct the uterus with patent tubes. If the leiomyomata are suspected in the region of the uterine vessels or broad ligament, it sometimes is difficult to remove them without a hysterectomy to control hemorrhage. If the patient has numerous small myomas, it may be impossible to remove them and reconstruct the uterus in a way that will support pregnancy. Excision of very large leiomyomata that constitute the entire anterior or posterior wall of the uterus may leave defects so large that closure is prohibited. In addition, if adenomyosis is present, it is not amenable to resection and cannot be treated by myomectomy.


Lab Studies:

Imaging Studies:

Diagnostic Procedures:


Medical therapy: Management of symptomatic uterine fibroids includes a number of nonsurgical approaches. Of note, treatment usually is strictly for patient comfort, and withholding treatment is reasonable in the patient with no symptoms or with mild, well-tolerated symptoms. While medical treatment currently does not allow a permanent cure for fibroids, a number of studies are examining nonsteroidal anti-inflammatory drugs, oral contraceptive pills, progestins, and androgens, as well as gonadotropin-releasing hormone (GnRH) analogs for treatment (Davis, 1995).

Of note, the Agency for Health Care Quality Research (AHRQ) commissioned a systematic review on the Management of Uterine Fibroids (Myers, 2001). They comment, "with the exception of trials of GnRH agonist therapy as an adjunct to surgery, there is a remarkable lack of randomized trial data demonstrating the effectiveness of medical therapies (non-steroidals, progestins, or oral contraceptives) in the management of women with symptomatic fibroids." Most importantly, no randomized trials compare medical management to surgery.

In general, surgery is reserved for people who have failed medical management. Despite the lack of good randomized evidence for the use of nonsteroidal anti-inflammatory drugs and oral contraceptive pills, these seem to be appropriate options for properly selected women without contraindications. Many women with fibroids, particularly those who have fibroids that are compounding dysfunctional bleeding, can be managed successfully with a combination of nonsteroidal anti-inflammatory drugs, birth control pills, or cyclic progestins. A short course is reasonable for patients with fibroids before committing to surgery, as some patients will be successfully managed medically. Most studies of medical management are short, from 3 months to 1 year, and long-term success is unclear.

Patients who are being managed expectantly usually are examined more frequently than once a year. If the myomas are very large and extend laterally, consideration can be given to periodic ultrasound to monitor for the development of hydronephrosis or the rare occurrence of ureteral obstruction.

Surgical therapy: A number of surgical therapies are available for the management of myomas. The most commonly performed procedure is total abdominal hysterectomy, and many thousands are performed every year in this country. In this section, we will focus on conservative surgery for leiomyoma. The traditional procedure performed is abdominal myomectomy, which is covered in detail below.

Myomectomy can be performed laparoscopically under certain conditions. Development of this procedure was driven by the desire to create a minimally invasive approach that would obviate a major abdominal laparotomy. Numerous published series document the feasibility of laparoscopic myomectomy (Seinara, 1997; Hasson, 1992; Darai, 1997; Dubuisson, 1997 are representative).

This is a newer procedure and has many unresolved questions including selection of appropriate candidates, the optimal way for removing the myomas, and closure of the defects as well as the safety of pregnancy and labor when compared to an abdominal procedure. The most recent ACOG Practice Bulletin (Stewart, 2000) suggests that it "may be a safe and effective option for women with a small number of moderately sized uterine leiomyomas who do not desire future fertility. Further studies are necessary to evaluate the safety of this procedure for women planning pregnancy."

A third technique, hysteroscopic resection, also can be used selectively for myomas impinging on the endometrial cavity that are thought to contribute to abnormal bleeding or infertility.

Preoperative details:

Abdominal myomectomy

GnRH analogues have been studied extensively in patients undergoing abdominal myomectomy. Potential advantages include decreased blood loss, smaller uterine defects, decreased operating time, decreased surgical difficulty, and additional time prior to surgery to allow the patient to donate autologous blood for potential transfusion. Several authors have done randomized trials looking at blood loss (see Table 1, below) and operating room time (see Table 2, below). Blood loss in all studies was less, although not necessarily enough to decrease transfusion risk. Similarly, operating room (OR) time was approximately 10 minutes shorter. Although there are theoretical advantages, no published evidence suggests that decreasing the overall size of the defect improves outcome or decreases risk of uterine rupture during pregnancy.

Table 1. Estimated Blood Loss (mL) in Randomized Blinded Trials Comparing GnRH to Controls


Author Year Approach Analog Control
Friedman 1989 Abd 213 320*
Fedele 1990 Abd 235 275
Golan 1993 Abd 476 320*
Zullo 1998 L/S 172 232



Table 2. Operating Time in Randomized Blinded Trials Comparing GnRH to Controls (minutes)


Author Year Approach Analog Control
Friedman 1989 Abd 87 99
Fedele 1990 Abd 92 98
Golan 1993 Abd 80 96
Zullo 1998 L/S 99 113


Several potential disadvantages exist. GnRH analog treatment is quite expensive. It is possible that GnRH agonist treatment may make it more difficult to remove fibroids.

Deligdisch et al (1997) made a histologic study of the cleavage planes between the myoma and the myometrium in hysterectomy specimens in patients treated with GnRH analogues and a group of untreated controls. They noted that there was a loss of the histologic cleavage plane in 91% of patients pretreated with GnRH analogue and 50% of controls. Theoretically, by decreasing the size of fibroids, some fibroids may decrease in size to the point where they are not detectable at the time of surgery, thus increasing the recurrence rate. Friedman et al (1992) followed 20 patients for a total ranging 27-38 months. They studied the occurrence of myomas greater than 1 cm detected by ultrasound. They noted a 70% recurrence rate in the group treated with GnRH and a 50% in the placebo treated group.

Fedele (1990) similarly studied 24 patients for 6 months and measured as an outcome any myoma recurring on ultrasound. It was noted that 63% in the analog group and 13% in the placebo group had recurrences.

Preoperative autologous blood donation is offered frequently to patients undergoing myomectomy but likely is overutilized. Transfusion-related infection has become tremendously rare, and by making patients more anemic preoperatively autologous donation actually may increase the transfusion rate.

A single study in a gynecologic population (Kanter, 1999) suggests that using patient and provider education to decrease autologous blood donation in hysterectomy patients decreased by 50% (from 53% to 26%) the number of units donated without altering the allogeneic transfusion rate. An individualized approach where autologous blood donation is reserved for patients at particularly high risk of transfusion (very large or numerous myomas) is appropriate. If autologous blood is to be donated, GnRH analog, usually a 3-month course, is extremely useful. It also is tremendously useful for allowing anemic patients to attain normal hemoglobin levels prior to surgery.

Preoperative bowel preparation is not necessary for these patients unless unusual adhesions are expected. No studies are available on prophylactic antibiotic use. Because this procedure usually is performed on patients desiring to retain their fertility and the consequences of infection on tubal function are severe, many providers do give prophylactic antibiotics empirically. A vaginal prep also is a sensible precaution, as 1-2% of these procedures will of necessity be converted into abdominal hysterectomies.

Laparoscopic myomectomy

A number of authors have attempted to define who is a candidate for a laparoscopic myomectomy. Although several suggestions have been made, none has been validated. Nezhat (1991) laparoscopically removed myomas up to 15 cm in diameter. Other authors are more conservative. Dubuisson (1996b) suggested not removing any myoma greater than 8 cm and not performing laparoscopic myomectomy if more than 2 myomas are present. Parker (1994b) suggested limiting the procedure to patients with uteri less than 14-week size, fibroids less than 8 cm, at least 50% of the myoma being subserosal, and the myoma not being located in any critical location.

GnRH analog use can slightly decrease operating time and blood loss during laparoscopic myomectomy as well (see Table 1, Table 2). All patients undergoing laparoscopic myomectomy should also give consent for laparotomy as it will be necessary to convert up to 8% of the procedures to a laparotomy intraoperatively (Dubuisson, 1997).

Hysteroscopic myoma resection

GnRH analog also has been studied for patients undergoing hysteroscopic myomectomy (Römer, 1998). In patients where the myoma did not impinge on the endometrial cavity, no difference was seen in operative time, fluid deficit, or hemoglobin drop. A slight improvement was seen in eumenorrhea in the patients who were not pretreated. However, when patients with submucous myomas extending over 50% into the cavity were studied, operative time was significantly shorter and the likelihood of postoperative eumenorrhea was significantly greater for those who had been pretreated with GnRH analogs.

Intraoperative details:

Abdominal myomectomy

Patients undergoing abdominal myomectomy require anesthesia adequate for a laparotomy, usually general endotracheal (GET). An incision is chosen that allows maximal exposure. Many myomectomies can be performed through a Pfannenstiel incision, but vertical incisions can be used when necessary. There are instances when better exposure potentially could be the difference between the ability to stop hemorrhage and preserve the uterus and the need to proceed with hysterectomy to control bleeding. Several excellent atlases are available for full details of surgery (Wheeless 1988, Te Linde 1997).

One important issue with myomectomy is controlling blood loss from the raw myoma beds after they have been excised. Several techniques have been studied. A randomized trial comparing vasopressin and saline injected into the serosa prior to the uterine incision (Frederick, 1994) showed that vasopressin is extremely effective in decreasing blood loss. In their study, patients receiving saline had a 50% transfusion rate while in the vasopressin group no patients were transfused. There was a difference of a 13% decrease in hematocrit compared to a 5% decrease. Many providers place tourniquets to control bleeding (DeLancey, 1992). This usually is done by making a window in the broad ligament at the level of the internal cervical os bilaterally and passing a Foley catheter or red rubber catheter through the windows and around the cervix and then tightening it with a clamp to constrict the uterine vessels. In combination with this, vascular clamps generally are placed on the utero-ovarian ligaments.

Two randomized trials compared vasopressin and tourniquet use. Fletcher (1996) showed that vasopressin was associated with lower blood loss and less risk of either transfusion or blood loss greater than 1 liter. Ginsberg (1993) noted no statistically significant difference between the groups, although her study was much smaller. No studies are available comparing tourniquet to no tourniquet. Studies very clearly suggest that vasopressin (usually 20 units in 50-100 cc normal saline) should be injected routinely prior to making the incision in the wall of the uterus. It is unclear whether additional use of a tourniquet further decreases blood loss.

After dilute vasopressin has been injected, an incision is made through the wall of the uterus into the myoma. Once the plane between the myometrium and myoma has been defined, it is dissected bluntly and sharply until the entire fibroid is removed. As many fibroids as possible are removed through a single incision. Once the fibroids have been removed, the defect is closed in layers with delayed absorbable suture.

Placement of the incision frequently is overlooked and is important, as well. Tulandi (1993) studied 26 women with uteri greater than 6-8 weeks?size. He performed abdominal myomectomies on these patients and did a second-look laparoscopy 6 weeks later. He noted patients with incisions in the posterior wall of the uterus had a much higher likelihood of significant adhesions as measured by percentage with adhesions or American Fertility Society (AFS) adhesion score compared to patients with incisions in the fundus or anterior wall of the uterus.

Laparoscopic myomectomy

Unfortunately, a number of important unresolved technical issues remain regarding laparoscopic myomectomy. One difficulty is in the removal of the fibroids, which has been done by morcellation (Carter, 1997), mini-laparotomy (Nezhat, 1994), or colpotomy. An alternative to this has been to destroy the fibroids in place with cryotherapy (Zreik, 1998), bipolar cautery (Goldfarb, 1995), or laser (Goldfarb, 1992). No trials have compared these techniques to determine which is the safest or the most effective.

The second critical technical issue is closing the uterine defect left by the destruction or removal of the fibroid. This is much more difficult than during an open procedure. Nezhat et al (1991, 1996) performed laparoscopic myomectomy and then did a second-look laparoscopy on 28 women 6 weeks later. He had removed 37 myomas from these women and closed the defects with laparoscopically placed sutures. At the time of second look, indentations were visible at all of the sites where the myomas had been excised. In addition, 6 uterine fistulas were visible on postoperative HSG.

Only a single randomized trial (Mais, 1996) compares abdominal and laparoscopic myomectomy in a randomized fashion. The authors randomized 40 patients to abdominal myomectomy or laparoscopic myomectomy with myoma morcellation. Complications and short-term outcomes were then compared. The study was restricted to patients with fairly small and few myomas, and the authors excluded patients with submucosal myomas. Each patient had a mean of 2.5 myomas with an average size of 4.4 cm. Surgical outcomes were very similar. The estimated blood loss was 200 cc laparoscopically versus 230 cc abdominally. OR time was very similar; approximately 100 minutes in both groups. There were no intraoperative complications. The patients receiving abdominal myomectomy had longer hospital stays and, as might be expected, had lower pain scores.

In summarizing the published laparoscopic myomectomy literature, the numerous small series are difficult to interpret because of their mixed populations, variable and short-term follow-up, and the lack of controls. In general, no fatalities were noted in any of these series. The risk of conversion to laparotomy was approximately 2-8%. Most patients were hospitalized postoperatively. The role of this procedure remains unclear because none of the published studies gives clear insight to the appropriate patients on which to perform the procedure, the benefits and risks of the abdominal procedures versus no procedure, and the long-term outcomes.

Hysteroscopic resection of myomas

A thorough discussion of Hysteroscopy can be found in a separate article. Patients undergoing diagnostic hysteroscopy also should give consent for resection of myomas projecting into the endometrial cavity. Myoma resection usually is performed with a loop electrode by shaving the visible portion of the myoma into small pieces (see Picture 5) (Corson, 1991, 1995; Gimpelson, 2000). Sometimes myomas deeply embedded in the myometrium cannot be completely excised. Other techniques for removing the myoma hysteroscopically include using a neodymium:YAG laser fiber (Donnez, 1990) or electric myoma vaporizer (Brooks, 1995). In patients who do not desire future fertility and whose main problem is excessive bleeding, concomitant endometrial ablation may help.

Postoperative details: After abdominal myomectomy, patients are treated as are any other patients who have had a laparotomy. Patients should be ambulated early, and the diet advanced at the surgeon's discretion. The patients tend to be young and healthy, so the recovery usually is fairly rapid. Vaginal bleeding is very common postoperatively. Fever also is quite common, particularly in the first 48 hours, but it does not appear to be from infection (Iverson, 1999). No studies exist determining the optimal amount of time patients should wait prior to attempting conception. It would appear prudent to allow patients to heal for at least several months prior to any attempts. Patients should not use tampons, douche, engage in sex, or place anything in the vagina for at least 4-6 weeks postoperatively.

Hysteroscopic myoma resection: Some authors (Blackwell, 1992; Ubaldi, 1995) advocate the postoperative administration of oral estrogen to decrease the chance of intrauterine synechiae formation, although there are no supporting studies.

Follow-up care: Patients who have had myomectomies should be followed for recurrence of myomas. Patients normally are seen for a routine postoperative exam 2-6 weeks postoperatively. A pelvic exam at 3 months, 6 months, and 1 year for myoma recurrence seems reasonable, although no studies have been done to support this. If there is no recurrence in 1 year, annual exams likely are adequate.

Vavala et al (1997) studied the use of GnRH analog for preventing the recurrence of myomas after myomectomy. The 65 patients who were studied were given leuprolide acetate (Lupron) depot 3 months a year for 3 years. At the end of this time, the patients showed significantly reduced uterine volume and myoma recurrence rate compared to untreated patients. Although this suggests that GnRH analog might be useful, they did not present data on symptoms or the need for repeat surgery. Further study clearly is needed.



Abdominal myomectomy

Both short-term and long-term problems are associated with abdominal myomectomy. Short-term complications include all of the usual complications of gynecologic laparotomy including bleeding, infection, visceral damage, and thromboembolism.

Iverson, Chelmow and co-workers at Tufts (1996) reviewed the relative morbidity of abdominal myomectomy as compared to abdominal hysterectomy. In patients undergoing myomectomy the average blood loss was 464 mL, and the risk of transfusion was approximately 28%. However, nearly three fourths of these transfused units were autologous blood replacement, and many of these patients likely would not have received blood if only random donor blood had been available. Approximately 13% of patients had temperatures of greater than or equal to 38.5°C after 48 hours postoperatively and were started on antibiotics for presumed infection. When compared to hysterectomy, the operative time was nearly identical. The blood loss, traditionally thought to be higher than with hysterectomy, actually was quite a bit lower, a difference that was maintained even with multivariate analysis.

No intraoperative visceral injuries occurred in patients undergoing myomectomy, although a number occurred in patients undergoing hysterectomy. Lamorte (1993) noted similar complication rates in an uncontrolled series of patients at Yale, and Sawin and coworkers (2000) noted similar results in a more recent work comparing hysterectomy and myomectomy.

Patients undergoing myomectomy have an unusually high incidence of fever occurring in the first 48 hours postoperatively, a phenomenon that appears to be unique to this procedure. This also was studied by Iverson, Chelmow, and coworkers (1999). They noted a baseline risk of approximately 33% for fever greater than or equal to 38.5°C within the first 48 hours. When compared with hysterectomy and using multivariate analysis to control for age, parity, estimated blood loss, and type of physician performing the surgery, they noticed a 3.9-fold increased risk. This "myomectomy fever" may be from release of unknown pyrogenic factors during the myoma dissection or from hematomas forming in defects left by the removed myomas.

The most significant short-term risk is the potential for the need to convert a myomectomy to a hysterectomy intraoperatively. This occurs largely for 2 reasons. First, at times it may not be possible to reconstruct the uterus because of the many defects left by the removal of multiple small fibroids or a single large fibroid. Second, a hysterectomy may be necessary intraoperatively to control bleeding. In the Tufts series (Iverson, 1996), it became necessary to convert 2 of 103 myomectomies to hysterectomies intraoperatively. In the Yale series (Lamorte, 1993), a single patient out of 128 required a hysterectomy. All patients undergoing myomectomies should be apprised of this possibility as part of the consent process.

Laparoscopic myomectomy

Laparoscopic myomectomy has all of the usual risks of laparoscopy, predominantly those related to trocar placement and the need to convert to a laparotomy. Rates of conversion vary from very low to 8-10%, largely depending on the complexity of the cases. The suboptimal defect closures are very concerning for uterine rupture in labor. A number of case reports have been published describing uterine rupture after laparoscopic myomectomy (Nkemayim, 2000; Hockstein, 2000; Pelosi, 1997; Arcangeli, 1997 to cite a few of the most recent). It is difficult to estimate incidence from case reports, but the number of case reports seems to be unusual for such a rarely performed procedure. In addition, a review of published series of abdominal myomectomies finds no reports of uterine rupture in pregnancy whereas the largest published series of laparoscopic myomectomy (Dubuisson, 1997) reported one uterine rupture out of 213 patients.

Hysteroscopic myoma resection

Complications of hysteroscopic myoma resection include hemorrhage, uterine perforation, damage to the cervix, and excessive absorption of the distention media (usually glycine) into the vascular system, which can cause metabolic disturbance (Cooper and Brady, 2000). Corson and Brooks (1991) noted one case of heavy bleeding requiring transfusion and 3 uterine perforations out of 92 patients undergoing hysteroscopic myoma resection. Indman (1993) noted distension media complications in 2 of 51 women. Intrauterine synechiae also can occur after hysteroscopic myoma resection (Hallez, 1995).


Despite the long use of myomectomy and extensive literature on this procedure, the data actually are poor for 2 important issues related to outcome. In particular, both recurrence rate and impact on fertility have been poorly studied.

Abdominal myomectomy

Vercellini et al (1998) extensively reviewed abdominal myomectomy as a fertility enhancing procedure. They noted that although numerous papers report on fertility outcomes after myomectomy, they all share the same serious flaws. In particular, not a single study included controls or used randomization. Only a few of the studies used life table analysis. All used differing definitions of infertility and included heterogeneous uses of other infertility treatments. Nonetheless, the studies were fairly consistent, with approximately two thirds of patients with myomas and otherwise unexplained infertility conceiving after myomectomy.

Similar results were noted in both prospective and retrospective studies. However, results were very inconsistent when subgroups were analyzed. Myomectomy continues to be offered routinely to patients with uterine fibroids and infertility, but until controlled studies with expectantly managed controls are performed, the benefit of this procedure to patients remains unclear.

The risk of myoma recurrence similarly is poorly studied. Patients undergoing myomectomy should be counseled that they are at risk for fibroid recurrence and the potential for additional surgery in the future. Unfortunately, proper studies to determine recurrence risk are not available. Most studies are limited because they include heterogeneous study groups composed of a mixture of symptomatic patients who are managed for treatment of fibroids and asymptomatic patients who are managed for treatment of infertility. Follow-up in all of these studies is poor with many patients lost to follow-up and most with very short follow-up periods. As fibroids may recur slowly over a long period of time, studies with short follow-up do not give the necessary information.

In addition, the studies use different definitions of recurrence, some limited only to symptomatic recurrence and some including patients with asymptomatic fibroids detected on ultrasound or pelvic exam. Most do not use life table analysis.

It is likely that many women will have recurrence of myomas after myomectomy. Fedele (1995) used ultrasound to diagnose recurrences and noted a cumulative recurrence rate of 51% over 5 years. Asymptomatic recurrence is, however, not really a relevant outcome. Limiting work to studies that look at patients requiring reintervention after myomectomy and appreciating that most of these studies have short, less than 5-year follow-up, recurrence rates of 8-27% are noted (see Table 3, below). Most of the higher rates are noted in older studies when hysterectomy was done for much more liberal indications and was performed much more frequently. Looking at these studies, there probably is a 5-10% medium-term risk of the need for repeat surgery after myomectomy.

Several studies established particular risk factors for reoperation. Malone (1969) noted that removal of multiple myomas was a strong risk factor for need for reoperation. Future repeat surgery was required for 26% of patients with multiple myomas as compared to 11% of the patients with single myomas. Also, pregnancy after myomectomy appears to be protective. Candiani (1991) noted that over 10 years following myomectomy, 15% of patients achieving pregnancy and 30% of patients not achieving pregnancy required repeat surgery.

Table 3. Summary of Studies Reporting Need For Future Surgery For Myomas After Myomectomy


Study Year Follow-up Re-operation
Finn & Muller 1950 24-120+ 13%
Brown 1956 >72 17%
Malone 1969 >60 27%
Berkeley 1983 17 8%
Garcia & Tureck 1984 >10 6%
Rosenfeld 1986 >12 4%
Smith & Uhlir 1990 NR 5%
Verkauf 1992 42 6%
Gehlbach 1993 >12 12%
Acien & Querada 1996 4-144 8%


One other long-term complication that must be considered is the risk of uterine rupture during pregnancy. This, fortunately, is quite rare and is most likely to occur in labor. Spontaneous rupture in patients undergoing abdominal myomectomy is unusual, although there are a number of case reports (Nkemayim, 2000; Hockstein, 2000; Pelosi, 1997; Arcangeli, 1997) in patients who became pregnant after laparoscopic myomectomy.

Although data are limited to support it, the usual recommendation is to offer cesarean delivery to patients who had myomectomies where large defects in the active segment of the uterus are created by removal of the fibroids. Some recommend cesarean delivery any time the endometrial cavity is entered during the procedure, but it seems more likely that it is the total extent of the defect and not the entry into the endometrium that presents the risk to the patient. Recommendations must be individualized for each patient. These recommendations are best made by the physician performing the myomectomy and should be clearly documented in the chart and conveyed to the patient so that they are clear in the event of future pregnancy.

Laparoscopic myomectomy

Fertility rates after laparoscopic myomectomy have been reported in several studies. These studies have the same limitations as the published work on abdominal myomectomy but are even more limited in number. They seldom use life table analysis and have short and incomplete follow-up. Pregnancy rates of 22-62% have been reported (Dubuisson, 1996a, 2000; Darai, 1997; Miller, 1996).

A single study looked at recurrence after laparoscopic myomectomy (Nezhat, 1998). The study followed 114 women for a mean of 37 months and defined recurrence as the return of any myoma. The cumulative recurrence risk was 10.6% at 1 year, 31.7% at 3 years, and 51.4% at 5 years. Eight patients underwent repeat laparoscopic myomectomies. One patient underwent 2 laparoscopic myomectomies. One patient had a myomectomy and then a total abdominal hysterectomy, and 6 had total abdominal hysterectomies. Of the patients, 14% required repeat surgery.

Goldfarb (1992) presented data on patients undergoing myolysis with the neodymium:YAG laser. He studied 75 patients and reported 50% shrinkage of the myomas at 6 months. This series, like most others on minimally invasive techniques, reported no medium-term or long-term data on pregnancy or need for future procedures. Chapman (1998) performed a similar procedure on 293 patients. Of these, 6 patients required hysterectomy and 30 had future pregnancies. Goldfarb (1995) performed myolysis with bipolar cautery and reported 83% shrinkage over 6 months. One series reported 2 out of 3 patients with uterine rupture after they became pregnant within 3 months of the procedure (Vilos, 1998).

Hysteroscopic myoma resection

Ubaldi (1995) reviewed fertility rates after hysteroscopic myomectomy. He noted pregnancy rates similar to those after abdominal myomectomy, averaging approximately 60%. Again, none of these studies had expectantly managed control groups. Reoperation after hysteroscopic myomectomy also has been studied. As usual, these studies are compounded by short follow-up periods. Ubaldi reviewed older studies, which had reoperation rates of 5-25% on follow up of up to 3 years. In more recent studies, Vercellini (1999) studied 108 women who had hysteroscopic resection of submucous pedunculated, sessile, or intramural leiomyomas. After a mean follow-up of 41 months, 27 patients had recurrence of myoma on ultrasound, with a 3-year cumulative recurrence rate of 34%. Twenty women had recurrent menorrhagia with a 3-year rate of 30%.

Emanuel (1999) studied 285 women who had submucous myomas treated with hysteroscopic myoma resection without endometrial ablation. Several patients required multiple procedures. Patients were followed for a median of 46 months. Forty-one patients (14.5%) required repeat surgery. Patients requiring repeat surgery were more likely to have larger uteri and higher numbers of submucous myomas. Hysterectomy was required in 20 of the 41 patients who required repeat surgery. Most (90.3%) patients with normal size uteri and 2 or less myomas did not require future surgery at 5 years.

A second series from Britain (Hart, 1999) studied 122 women for a mean of 2.3 years. Twenty-one percent of women required repeat surgery by 4 years and 0% thereafter. Their regression analysis suggested that outcome was better in older women where the uterus was less than or equal to 6 weeks?size gestation or the fibroid was less than or equal to 3 cm and mainly intracavitary. Donnez (1994) studied the recurrence of menorrhagia by the site of myoma. He noted that women who had multiple submucosal myomas were much more likely to have recurrent symptomatic menorrhagia than people who had only 1-2 myomas. Having the greatest diameter inside the uterine cavity and the largest portion of the uterine wall were less accurate predictors.

Vaginal removal of a prolapsed myoma

Management of the prolapsed vaginal myoma also can be problematic. A single study (Ben-Baruch, 1988) noted that removal of prolapsed myomas represented 2.5% of all procedures for myomas. Some 93.5% of these procedures were successful with transvaginal removal, and 6.5% of patients needed a total abdominal hysterectomy. Of the failures, only 1 had very serious complications. After the initial vaginal myomectomy, 34 patients were followed for a median of 5.5 years. In these patients, 79% had no further symptoms from their fibroids. Of those remaining, 21% developed other symptoms of which 6% required a hysterectomy, 6% had a single repeat prolapsed myoma, and 3% of patients (1 patient) had multiple repeat procedures.


Many questions remain regarding the natural course of untreated fibroids, the efficacy of medical management, as well as the unanswered questions regarding surgery discussed above. The Agency for Health Care Research (AHCR) sponsored an evidence-based review (Myers, forthcoming) that provides a superb review of these topics.

Current controversies include the role of minimally invasive procedures. In particular, laparoscopic myomectomy has many theoretical advantages including lower cost and avoidance of prolonged hospitalizations. However, it is not at all clear that the repair of the defect is as effective as that performed with abdominal myomectomy, and it may have an increased risk of uterine rupture during pregnancy.

Medical therapy also is being explored. At present, data for medical management, particularly new treatments such as Tibolone or older treatments such as controlling symptoms with birth control pills, are very scant. The role of GnRH analogs also requires further clarification. The advantages in terms of making the fibroids smaller must be balanced against the high cost and the subsequent inability to find fibroids reduced by treatment.

A number of new treatments are being explored at present. Uterine artery embolization in particular may be an especially promising minimally invasive approach to fibroids (Siskin, 2000; Worthington-Kirsch, 1998). In this procedure, angiographic catheters are introduced in the groin and passed to the uterine artery under fluoroscopic guidance. Microspheres are then injected, which lodge in the blood supply to the myomas and cause them to infarct. Significant pain from the acute infarction of the myomas usually requires hospital admission for pain control.

A number of complications have been reported including prolapsing myomas, infection and hematoma at the catheter placement site in the groin, and bleeding and infection requiring hysterectomy. However, short-term results for relief of heavy bleeding and pelvic pain and pressure have been good. As this is a new procedure, there is only a small amount of follow-up past 2 years to estimate recurrence rates and extremely limited information for patients who plan future fertility.

It also is conceivable that as more is learned about the genetics of leiomyomas it may be possible to target the specific genetic defect to prevent or treat myomas.


Caption: Picture 1. Small uterine myomas visualized at laparoscopy.
Click to see larger picture
Picture Type: Photo
Caption: Picture 2. Submucosal myoma visualized at hysteroscopy.
Click to see larger picture
Picture Type: Photo
Caption: Picture 3. Uterine leiomyoma as seen on ultrasound imaging (image courtesy of Rick Doherty, MD).
Click to see larger picture
Picture Type: Photo
Caption: Picture 4. Prolapsed uterine myoma visualized through a speculum (image provided by Kris Strohbehn, MD).
Click to see larger picture
Picture Type: Photo
Caption: Picture 5. Submucosal myoma being resected hysteroscopically with a loop resectoscope.
Click to see larger picture
Picture Type: CT