Dr. Abu-Judeh was a Cardiovascular Interventional Radiology fellow at New York Presbyterian Hospital, Cornell Medical Center, New York, NY. He is currently an Assistant Professor at the University of Dentistry of New Jersey­New Jersey Medical School. Dr. Khilnani and Dr. Min are Professors and Dr. Kandarpa is the Chief of the Cardiovascular Interventional Radiology Section at New York Presbyterian Hospital, Cornell Medical Center, New York, NY.

Recently, uterine artery embolization (UAE) has received much attention in both the scientific literature and the lay press as a minimally invasive procedure for treating symptomatic fibroid disease. Increasing numbers of women report satisfaction with the effectiveness and efficiency of the procedure over the conventional options. The purpose of this article is to review the literature regarding UAE and its development, techniques, indications, results, and complications. The results to date indicate that UAE is a highly successful, uterine-sparing alternative to hysterectomy for women with symptomatic fibroid disease.

A MEDLINE literature search using the words "uterine fibroid embolization" (UFE) or "uterine artery embolization" (UAE) revealed 196 articles as of May 1, 2002. Those articles include review articles, case series, case reports of adverse effects, and technical notes; addressing anatomical vascular considerations, fertility and pregnancy issues, animal experiments, embolic material choices, quality of life, and cost effectiveness.

Many women with fibroids choose to seek treatment only for symptomatic relief, with the main motivation often being quality-of-life issues. Conservative measures and traditional treatments may not be an option for many women. Surgical interventions are an undesirable option for some women since many consider the uterus an important aspect of their femininity. ¹ The absence of an effective complication-free treatment for fibroid disease has led to a plethora of treatment modalities. Uterine artery embolization is a relatively new procedure that produces impressive results in the treatment of symptomatic uterine fibroids. It carries a very low rate of complications and preserves the uterus.

Short history of UAE

Transarterial embolization for active pelvic bleeding has been considered the therapy of choice for decades. Trauma surgeons have long used pelvic embolization to control acute traumatic pelvic hemorrhage. The first reported case of pelvic embolization for the treatment of obstetrical hemorrhage was in 1979. ² Since then, embolization has gained more recognition and has been used to treat a wide variety of obstetric and gynecologic diseases associated with bleeding, including postpartum hemorrhage and ectopic pregnancy. ³ Uterine artery embolization has also been used successfully as presurgical gynecologic prophylaxis to decrease the amount of blood loss. In 1994, Ravina et al ⁴ were the first to describe pelvic embolization or UAE as an adjunct to surgery in the treatment of fibroids. A year later, he described the procedure as the primary treatment of uterine fibroids. ^5,6

Since 1994, there has been an exponential rise in the number of UAEs performed annually. Worldwide, approximately 15,000 women, including more than 10,000 in the United States, have been treated with UAE for symptomatic uterine fibroid disease. Numerous published reports involving a total of more than 2000 patients have demonstrated the effectiveness and safety of this procedure. Symptomatic uterine leiomyoma are a significant source of morbidity to many women and place a substantial burden on our healthcare system. Uterine artery embolization has emerged as a minimally invasive procedure for treatment of symptomatic uterine fibroids and has gained acceptance as a primary treatment for symptomatic fibroids.

Pathophysiology of fibroids

Uterine fibroids are the most common benign tumors in women with an estimated incidence of 20% to 40% in women over 35 years of age. Fibroid tumors affect women of all ages and races; however, there is a higher incidence among African-American women. ^7,8

Leiomyoma is a benign smooth-muscle tumor that most commonly affects the body of the uterus but may also be found in the cervix, broad ligament, and, rarely, the ovary. The precise etiology of leiomyomas is unknown, although it is clear that the effects of hormones are pivotal as they typically occur during the reproductive period when hormonal influences are at their maximum. Studies have shown that leiomyomas have abnormal gene expression that maintains high levels of sensitivity to estrogen during the estrogen-dominated proliferative phase of the menstrual cycle. ⁹ Fibroids may vary in size from a few millimeters to 20 cm in size and may become symptomatic depending on their location and size.

Menorrhagia is the most common presenting symptom of fibroids. ¹⁰ The bleeding can be severe enough to cause microcytic anemia that iron supplementation alone is not enough to correct. Submucosal fibroids cause most bleeding in women with fibroids. Because of their location on the endometrium, submucosal fibroids place pressure on the uterine lining that builds with each menstrual cycle. This, in turn, can cause abnormal bleeding. ¹¹ Another common fibroid-related symptom is pelvic pain or pressure and may even, depending on their location (ie, subserosal), affect the gastrointestinal or genitourinary systems. Pelvic pressure from the increasing size of a fibroid, regardless of location, can also produce menorrhagia. Fibroids may transmit transmural pressure on the bladder, causing urinary frequency. If large enough, fibroids may obstruct urinary flow through pressure on the ureters. Similarly, a fibroid may put pressure upon the bowel, causing constipation, painful bowel movements, and hemorrhoids. Fibroids may even cause neuropathic pain often involving the lower back and legs. Frequent complaints from patients include increased abdominal girth and distention, producing a disturbing cosmetic effect. During pregnancy, fibroids can cause recurrent miscarriage and premature birth. ^10,11

Therapeutic options

Medical treatment for uterine fibroids may include the use of nonsteroidal anti-inflammatory agents, oral contraceptive pills, or progesterone analogs. Gonadotropin-releasing hormone agonist (GnRH) is reserved for more severe cases and is mainly limited to presurgical management to decrease blood loss during subsequent myomectomy or hysterectomy. ¹⁰ This treatment may result in improvement of symptoms and reduction in fibroid size of an average of 40% in the first 3 months of therapy. ¹² GnRH is associated with postmenopausal symptomatology and cannot be used for periods longer than 6 to 9 months. ¹² The fibroids often regrow to their original size within a few months of discontinuing the treatment. ^10,12

Operative management of uterine fibroids includes laparoscopic, hysteroscopic, vaginal, or abdominal hysterectomies. These procedures may be associated with significant surgical morbidity (adhesions, infection, blood loss, thromboembolic events) and infertility. ¹⁰ Fibroid disease is the single most common condition (30%), and gynecologists in the United States perform more than 150,000 hysterectomies each year to relieve symptoms of uterine fibroids. ¹³ The rate of hysterectomy, however, has shown to vary with geographic region, the woman's age, level of education, and insurance coverage. ¹³ The main advantage of hysterectomy is that the risk of further fibroids is nonexistent. It is, however, a major surgical procedure with an extended recovery period of up to 6 weeks and with significant postoperative morbidity and mortality. Any hysterectomy means the end of childbearing, and in addition to the direct surgical risks, there may be long-term physical, emotional, and psychological effects, potentially including depression and loss of sexual pleasure. ^14,15 If a salpingo-oophorectmy is performed along with the hysterectomy prior to menopause, it triggers premature onset of menopause with loss of ovarian-derived estrogen and androgens and may increase the risk of early bone loss and the onset of menopausal symptoms. ¹⁶

Surgical alternatives to hysterectomy have evolved, resulting in a 20% decline in the prevalence of the procedure in the last quarter century. Each year 35,000 myomectomies (transvaginal or laparoscopic) are performed to relieve the symptoms of uterine fibroids. Laparoscopic myomectomies are limited to a maximum fibroid size of 10 cm, or not more than four leiomyomas. ¹⁷ There is an average of 81% resolution of menorrhagia with myomectomies. ¹⁷ However, this procedure has its associated increased risk of hemorrhage, adhesions, lengthy operating time, postoperative morbidity (including pain), and longer hospital stays than with hysterectomy. ¹⁶ In addition to these shortcomings, 20% to 25% of women who undergo a myomectomy will ultimately require an additional procedure to further treat the fibroid. ¹⁶

Cost

Two recent cost studies from a single institution found that UAE compares favorably with hysterectomy ¹⁸ and myomectomy. ¹⁹ The estimated hospital cost for UAE was $3080. The cost for hysterectomy ranged from $3100 to $4900, depending on the type of procedure performed, and the cost for abdominal myomectomy was $5597. Although not included in these studies, there is an indirect cost of the lost work days during recovery, which is significantly lower in patients receiving UAE than in those undergoing hysterectomies or myomectomies.

Preprocedure evaluation

Patient selection and preprocedure evaluation are extremely important: the correct diagnosis of symptomatic fibroid disease must be reached and other causes of uterine bleeding must be excluded. The patient should be interviewed, and a consultation with a gynecologist is advocated. In our institution, the patient is given reading materials to further familiarize her with all aspects of the treatment. Currently, premenopausal women who have uterine fibroids and symptoms that correlate reasonably well with the size and location of the fibroids are candidates for UAE. A complete medical history should be obtained. A physical examination, including a pelvic examination, should be performed within 6 months of the procedure. A normal PAP smear within 12 months is also required. The possibility of pregnancy should be excluded. If the patient has history of pelvic infection, cultures for gonorrhea and chlamydia should be obtained. In our institution, we obtain follicle-stimulating hormone (FSH) levels prior to UAE. Endometrial biopsy may be obtained for patients with abnormal bleeding, particularly if they are older than age 40. ²⁰ If the patient is taking GnRH, the medication should be discontinued 8 to 12 weeks prior to procedure.

Contraindications

Contraindications to UAE include pregnancy, acute or chronic pelvic infection, arteriovenous malformation, renal insufficiency, active vasculitis, or pelvic irradiation. Although severe allergy to contrast is a contraindication to perform UAE, there is a case report in the literature describing technical success using CO ₂ . ²¹

Other considerations

The possibility of a leiomyosarcoma should always be considered part of the differential diagnosis of uterine fibroids. It is extremely difficult, however, to prospectively differentiate a growing fibroid from a leiomyosarcoma with the currently employed imaging modalities, including magnetic resonance imaging (MRI). Theoretically, a positron-emission tomography (PET) scan might be of value.

This lack of adequate imaging options has led many physicians to perform deep fibroid biopsies, although this procedure has a low reliability rate (specificity and sensitivity), even in cases with known leiomyosarcoma. The incidence of leiomyosarcoma is rare and routine biopsies are not advocated, particularly given the low positive yield. ²⁰

Postprocedure follow-up imaging is necessary; both ultrasound and MRI are effective. The physician should be suspicious of cancerous lesions if the fibroids do not decrease in size after embolization or, particularly, if the fibroids continue to enlarge.

Women with extremely large fibroids (>20 cm) may not experience significant benefit from UAE due to bulk-related symptoms. McLucas et al ²² identified, with a strong statistical significance ( P = 0.05), that leiomyomata with a diameter >8.5 cm is a predictor of failure. Although Spies et al ²³ could not duplicate the statistical strength of McLucas's report, their conclusion was that smaller size leiomyoma and submucosal location were indicators for a positive outcome. There are no strong data to suggest that large fibroids are at increased risk for post-UAE infection. ²⁰

The initial reports of UAE suggested a link between UAE failure and the presence of adenomyosis. ^24,25 These early studies, which were conducted retrospectively to find an explanation for UAE failures, discovered islands of adenomyosis on the pathological examination. In a more recent retrospective series of 15 patients, the authors concluded that UAE can be beneficial in patients with menorrhagia and adenomyosis. ²⁶ We also recently reported our experience in patients with uterine fibroids and co-existing adenomyosis. ²⁷ Although the number of patients was small (10 patients), there was a suggestion that women with focal adenomyosis may be more responsive to UAE than those with diffuse adenomyosis.

Many interventional radiologists say they would consider offering UAE to patients with adenomyosis, since the only other successful alternative is hysterectomy, and would perform the procedure after counseling the patient and explaining the potential for failure (G. Siskin, informal oral communication, 2002 SCVIR meeting).

Menorrhagia in postmenopausal women can be severe enough that some women may seek relief. In our experience, postmenopausal women who refuse surgery and who are still suffering from fibroids can experience relief following UAE. Careful gynecologic clearance must be performed to exclude malignancy. This procedure can be very challenging technically, particularly in patients not receiving hormone replacement therapy. ²⁸ The uterine artery is typically small and can be very difficult to access.

Preprocedure imaging

Prior to UAE, imaging should be performed to confirm the presence of fibroids, to exclude other causes of bleeding, and to serve as a baseline for future follow-up. Both ultrasound and MRI are capable of diagnosing fibroids, showing their location, and obtaining volumes to determine both fibroid and uterine size. Although ultrasound is significantly less costly than MRI, there is a trend toward obtaining an MRI rather than an ultrasound. An MRI is less operator-dependent and is reproducible. It has been shown to be more specific and sensitive than ultrasound for imaging of fibroids. ²⁰ It is better at defining pedunculated fibroids and concomitant diseases, such as adenomyosis, which may affect management. Patients with subserosal pedunculated fibroids have a theoretical risk of devascularization of the attachment leading to torsion or even detachment of the fibroid; in these cases, a necrotic mass will be left in the peritoneal cavity. ²⁹ Patients may benefit from laparoscopic myomectomy with or without premyomectomy UAE to decrease intraoperative bleeding. Patients with submucosal fibroids may benefit from hysteroscopic myomectomy since there is a risk of degenerating fibroid falling into the uterine cavity, a phenomenon noted even with nonpedunculated fibroids.

There is a recent trend to perform perfusion MRI with gadolinium. In one study, it was found that there is a difference in perfusion between the myometrium and leiomyoma following bilateral UAE. ³⁰ At 1 month, there was recovery of myometrial perfusion but perfusion to the leiomyoma remained depressed. The reduction of leiomyoma perfusion correlated with the improvement in clinical score at 1 month. Leiomyomas high in signal intensity on the initial T2-weighted images showed greater volume reduction than those with low signal intensity. The authors concluded that dynamic MRI may be used to predict clinical response, while signal intensity on T2-weighted images predicts volume reduction. Three-dimensional color Doppler sonography can depict fibroid vascularity and, in some cases, can reveal collateral flow not depicted by uterine artery (UA) arteriography ³¹ ; however, post-UAE sonographic peak systolic velocity did not show predictive value. ³²

UAE procedure

Infection control and pain management

On the day of the UAE, informed consent is obtained from the patient. Preprocedure prophylactic antibiotic administration is advocated. Most interventional radiologists use cefazoline (Ancef, SmithKline Bechum, Pittsburg, PA), others may use antibiotic with more gram-negative coverage such as ceftriaxone (Rocephin, Roche, Nutley, NJ). A Foley catheter is inserted to eliminate impairment of visualization of pelvic vasculature due to a distended bladder.

A protocol for pain management should be developed in consultation with the anesthesia service. Different pain management schemes have been developed, mostly dependent on each institution's protocol. This protocol should include preprocedure, intra-procedure, and immediate postprocedure management, as well as in-hospital and outpatient pain management.

Most centers administer a combination of nonsteroidal anti-inflammatory drugs and narcotics prior to initiation of the procedure. Most UAE procedures are performed under conscious sedation, achieved with an intravenous (IV) combination of narcotics and benzodiazepines, typically fentanyl, morphine sulphate, and midazolam. Subcutaneous 1% lidocaine solution is the local anesthetic of choice. Unlike tumor embolization in other organs, lidocaine injection into the UA has led to spasm and constriction of the UA, resulting in incomplete embolization and no significant pain relief. ³³

There are several published strategies for pain management following UAE. Oral, IV, or epidural pain medications have been employed. Patient-controlled analgesia is frequently employed post-UAE and the set-up must be in place as soon as the procedure is over because frequently the pain may start immediately after UAE or even near the end of the procedure.

Giving medications with different mechanisms of action, such as nonsteroidal anti-inflammatory drugs and opioids, has proven to provide adequate pain control in both the inpatient and outpatient settings. ²⁰ In most centers, patients will stay overnight for pain control; however, some centers have had success performing the procedure in an outpatient setting. ³⁴ Almost all patients will experience postembolization syndrome, consisting of a variable constellation of pelvic pain, cramping, nausea, vomiting, fever, and lethargy. The pain may be most severe immediately after the UAE. Although pain may indicate devascularization of the fibroid or fibroid ischemia, pain has not been found to be a reliable indicator of procedure success. ³⁵ Lack of postprocedure pain, however, correlates well with procedure failure. It is important to have a team approach and an agreed-upon protocol among all services and individuals who may encounter the patient during or after this procedure.

Technique

It should be kept in mind that UAE is a vascular procedure and that careful attention to technique is essential for successful results. For the most part, most interventionalists are well versed in the different catheters and wires required to access the uterine arteries and are fairly skilled at doing so. However, there are certain issues that have been discussed in the literature that deserve attention.

Most interventional radiologists use the single-access, right common femoral artery approach using a 5F sheath. ³⁶ There are some who advocated using the double-access approach, right and left common femoral arteries, citing the lower radiation risk as well as shortened procedure time. ²⁰ This approach has not been used widely; however, due to the need for two interventional radiologists and the doubling of the risk of access-site­related complications. Moreover, there is a minor additional cost of a selective catheter, and a significant increase in cost if there is a need for two microcatheters. ¹⁸

Most interventional radiologists will place the tip of a 5F or 4F catheter at the aortic bifurcation and perform a pelvic arteriogram to illustrate the origin of the internal iliac arteries, confirm the presence of dilated uterine arteries, and demonstrate the presence and location of a dominant UA. This step may be substituted with a pelvic magnetic resonance arteriogram (MRA). ³⁷

Once in the internal iliac artery, most interventionalists use an oblique posterior projection and perform a road map to assist in selecting the UA. Typically, a contralateral oblique (20š to 30š) is sufficient. Occasionally, a steep oblique (40š to 50š) is needed, and infrequently, an ipsilateral oblique is required, particularly in the case of upper origin of the UA from the main trunk of the internal iliac artery. ³⁸

For complete and successful embolization of the UA, the interventional radiologist should be aware of the anatomical variations of the artery. It is typically the first branch of the anterior division of the internal iliac artery, and it may arise from the medial or the lateral aspect of the anterior division. Other variations of the origin of the UA relate to its take-off in relation to the division of the internal iliac artery. Typically, there is a shared origin with the cystic or vaginal arteries. The UA has a tight curve in its proximal segment, followed by a descending, transverse, and then ascending segment. ^20,38 The cervicovaginal branches, which some believe play a role in post-UAE sexual dysfunction, ³⁹ typically originate from the transverse segment of the UA. If possible, the tip of the catheter should be placed distal to those branches. Arcuate (main intramural), radial, and peripheral uterine arteries are further subdivisions of the UA that supply the body of the uterus. In a hypervascular uterus "parasitized vascularity" from ovarian or other pelvic vessels may occur. ^40,41

In initial reports, UA catheterization was performed with a 4F or 5F catheter; however, given that the UA is fairly tortuous artery, introducing a 4Fr or 5Fr catheter may lead to straightening the vessel, introduce artificial kinking, and reduce the flow around the catheter. Multiple authors have described spasms and constrictions preventing complete embolization with the use of such catheters. There is a strong current trend toward use of a 3F microcatheter to cannulate the UA. In a hand poll conducted at the 2002 Annual Meeting of the Society of Cardiovascular & Interventional Radiology (SCVIR), close to 40% of interventional radiologists said that they routinely used 3F microcatheters in every case, as do we in our practice. Furthermore, nearly 100% of those polled said they had used microcatheters at least in some cases.

It should also be remembered that patients who are on GnRH therapy might experience spasm and constriction of the UA, leading to difficult and sometimes incomplete embolization. ⁴²

There are important collaterals between the right and left UA, between the ovarian artery and the UA, and between the round ligament artery and the UA. In most cases, the fibroid receives its blood supply from the UA alone. In some cases, the ovarian artery may provide vascular supply to the fibroid. If reflux into the ovarian artery is noted during embolization, most authors recommend continuing the embolization as long as the embolic material is not allowed to go beyond a short segment into the ovarian artery so that it immediately washes back into the uterus. ²⁰

The presence of ovarian collaterals to the fibroids is a recognized risk for UAE failure. ^43-45 It has been suggested that routine pre-embolization search (aortography and selected ovarian arteriography) for ovarian supply of fibroids is not warranted because of the assumed rarity of such collateral supplies and the radiation dose to the patient. ⁴³ Given the impact of the discovery of ovarian collaterals on patient management, there is a current trend--including in our practice--to perform aortography at the conclusion of the embolization, with the tip of the catheter at the level of the renal arteries in search for ovarian collaterals. ^44,45 If dilated ovarian arteries are discovered, the option of distal particulate embolization or proximal embolization with a temporary agent should be considered. ⁴⁴ If no intervention was made, the mere recognition of a major ovarian collateral may expedite the surgical treatment or bring the patient back for a second procedure with the intent to embolize the ovarian artery. Although not reported in the literature, an abdominal/pelvic MRA may be useful in discovering dilated ovarian arteries.

Embolic material

The choice of embolic material should be based on the size of the vessels. The blood supply of uterine leiomyoma is provided by the uterine arteries and is characterized by two vascular networks of different sizes. The peripheral blood supply comes from dilated and tortuous arteries, and the central blood supply is composed of few fine arterioles derived from the peripheral network. ⁴⁰ According to research by Pelage et al, ³⁸ the target vessels for UAE are the prefibroid vascular plexuses, which are typically 500 to 600 µm in diameter.

There are currently three embolic materials on the market that are used in UAE ⁴⁶ : gelatin sponge pledgets (Gelfoam, Pharmacia & Upjohn Co., Kalamazoo, MI), polyvinyl alcohol particles (PVA) (Contour, Interventional Therapeutic, Fermont, CA), and tri-acryl gelatin microspheres (Embospheres, BioSphere Medical, Rockland, MA). Although each agent has advantages and disadvantages, there is no consensus at this time as to which embolic material is best. Thus, the choice of embolic material is an individual choice for the interventional radiologist.

Gelfoam is significantly less expensive than any other embolic material. In the past, it was used as the embolic material of choice for obstetrical hemorrhage since it preserved fertility in such cases. ⁴⁷ It may be used in UAE for fibroid disease as a sole embolic material ⁴⁸ or it can be used to supplement the embolization or to "cap" a vessel that has been embolized with a different agent. ²⁰

The majority of UAE procedures reported in the literature have been performed with PVA as the embolic agent. The recommended size is 300 to 500 µm or 500 to 700 µm. ⁴⁹ There are disadvantages, however, to using PVA. It tends to occlude catheters, particularly the 3F size. It also tends to aggregate in and lodge in proximal vessels that are larger than those vessels believed ideal for UAE.

Calibrated microspheres have been introduced into the market recently. ⁵⁰ They are appealing because, unlike PVA, they are less likely to occlude a microcatheter and they are deformable. Their homogenous flow-dependent distribution makes them well suited for occluding the target vessels thought to be ideal for embolization. Recommended sizes are 500 to 700 µm or 700 to 900 µm depending on the use of microcatheters. ⁵⁰ In a late-breaking abstract at the 2002 Annual SCVIR meeting, the authors compared the new generation of Embospheres, EmboGold, with the bland Embospheres and found a surprisingly high rate (44% or 4 of 9 patients) of endometritis with the EmboGold. ⁵¹ We have used EmboGold in more than 20 patients, and our endometritis rate for both bland Embospheres and EmboGold is 0%.

Although Gelfoam is considered a temporary agent because it is reabsorbed relatively rapidly, it stimulates the most inflammatory response of the embolic agents used.

In women who undergo UAE with PVA alone, there appears to be restoration of flow in the main UA within 16 months after UAE. ²⁰ Sterling et al ²⁰ found that by using both PVA and Gelfoam together, the vessels remain occluded 3 to 12 months postprocedure. They do not, however, advocate the use of this method for patients desiring fertility.

Recent presentations at the SCVIR meeting described extravascular location of Embospheres on pathologic examination of hepatic tumors resected postembolization. ⁵² The size of embo-spheres used in hepatic tumor embol-ization is significantly smaller (40 to 120 µm) than those used in UAE, and it remains to be seen if larger particles behave in the same manner as the smaller particles.

The end point

The initial goal of UAE using PVA was to maintain embolization until there was complete stasis in the UA. This approach is being abandoned by most interventional radiologists who now look for other signs to end the embolization. This has become more important since Embospheres do not accomplish the complete stasis seen with PVA. Most interventional radiologists now look for signs such as the "pruned-tree" appearance of the ascending segment of the UA (Figure 1), or the appearance of new collaterals or filling of previously undemonstrated ovarian artery. Other signs include staining of the lower uterine segment and reflux of contrast around the catheter in the UA. ²⁰ More recently fluoroscopic visualization of 5 pulsations of contrast in the UA before it dissipates has been used as another sign to end the embolization (J. Spies, informal oral communication, SCVIR 2002 Annual Meeting).

Outcome

There are currently more than 20 large (>50) case series published worldwide. Table 1 summarizes the results of 11 of them. ^{24,28,33,53-60} The reported technical success rate of bilateral UAE is between 95% and 100%, and may approach 100% in cases in which the risk of dissecting the UA would be significantly reduced with the use of a 3F microcatheter. Symptomatic relief of menorrhagia is between 85% and 96%, and relief from bulk symptoms is between 69% and 92%. There is a 30% to 50% decrease in uterus size and 40% to 100% decrease in fibroid size. Most patients (90% to 100%) had a 1-night hospital stay.

Pathologically, fibroids undergo hyaline degeneration following UAE. Grossly, they are soft in consistency and pale on pathologic sectioning. The hyalinized areas are translucent and may be either small and localized or extensive. Microscopically, typical hyaline changes are characterized by marked cellularity and loss of the usual myomatous architecture. ⁶¹ After UAE, the majority of leiomyomata will decrease in volume and show changes on MRI, characterized by an increased signal intensity on T1-weighted images and a homogenous decreased signal intensity on T2-weighted images (Figure 2). ⁶² Fibroids may have a hypoechoic appearance on sonography, which correlates well with the hyaline degeneration seen in fibroids following UAE. ⁶³ Unlike the typical dense, coarse, and central calcification of spontaneously degenerating fibroids, fibroids degenerating post-UAE have peripheral calcification that correlates with the peripheral aggregation of PVA particles in peripheral fibroid arteries. ⁶³

UAE failure

There is an average failure rate for UAE of approximately 10%. ²⁰ Suspected causes include incomplete or unilateral embolization, extremely large fibroids, presence of uterine leiomyosarcoma, coexisting adenomyosis, and presence of collateral blood supply, such as ovarian artery and round ligament artery (a branch of the inferior epigastric artery).

In some cases, incomplete embol-ization is due to spasm or constriction of the UA. This can be due to the use of 4F or 5F catheters and has also been associated with theuse of GnRh. Many interventional radiologists now use 3Fr microcatheters routinely to select and embolize the UA. In a recent study, 8 of 12 patients who underwent unilateral UAE had unsuccessful outcomes due to technical failure. ⁶⁴ One failure was in a patient with a previously ligated internal iliac artery and another was in a patient who had unilateral UAE for technical reasons and did not undergo a second UAE. The authors concluded that patients who undergo unilateral UAE for technical reasons (inability to access the UA, eg, spasm) should be offered a second UAE to embolize the failed site shortly (within 3 to 4 weeks) after the initial procedure. Patients with congenitally absent UA may respond with similar success to those who undergo bilateral embolization.

Fertility, pregnancy, and sexuality

Leiomyomas are a well-known cause of infertility, and there are serious risks in pregnancy in a leiomatous uterus. The effect of UAE on fertility is not clear for several reasons. To date, most treated patients are in their 40s and 50s and do not desire future pregnancies. However, numerous pregnancies have been reported after UAE. ^65,66 These positive reports are difficult to interpret since the exact number of women attempting to get pregnant is not available. In a series of 12 patients with obstetric hemorrhage who underwent Gelfoam embolization, 11 resumed normal menses and all 3 patients who desired pregnancy achieved full-term, healthy newborns. ⁴⁷ Currently, most interventional radiologists, for ethical reasons, counsel women desiring future pregnancy to consider other options. McLucas et al ⁶⁷ reported a 33% term pregnancy rate in women younger than 40 years old, a rate comparable to that of women who underwent myomectomy. The authors conclude that pregnancy is a viable option for women undergoing UAE and that their preliminary data suggest that fertility is enhanced in women undergoing UAE.

A UAE experiment was performed in sheep with no fibroids and was found to be associated with a decrease in the number of pregnancies. ⁶⁸ We wonder if it would not be the case in uteri with fibroids, but unfortunately there are no animal models with fibroids. Additional studies are needed to further evaluate the effect of UAE on fertility in leiomyomatous uteri.

The effect of UAE on menses is also difficult to assess for the same reasons. Most series report a rate of 2% to 5% cessation of menses. ^65,66 In a recent series of 66 women, the reported ovarian failure rate was 14% ⁶⁹ ; however, when the authors stratified their data, it was noted that no women under 45 years of age developed ovarian failure. Moreover, surgical options do not appear to be protective from early hormonal menopause even in women undergoing ovarian-conserving hysterectomies. ⁷⁰ In one study, there was no statistically significant change ( P = 0.16) in ovarian function measured by basal FSH after UAE for the overall group. ⁷¹ However, for patients aged 45 or older, there was approximately a 15% chance of an increase in basal FSH into the perimenopausal range. ⁷¹

Impaired sexual function after hysterectomy is well documented in the literature. ¹⁴ There are also substantial, well-described emotional and psychological problems that may occur in a woman after hysterectomy. ¹⁵ A woman may feel "desexed," or may consider herself "half a woman." There is a single case report in the literature describing sexual dysfunction after UAE. ³⁹ The authors propose that occlusion of the cervicovaginal branches may have caused cervical ischemia and, in turn, loss of orgasm; however, the patient eventually recovered sexual function. Recent abstracts reported that UAE was found not to cause sexual dysfunction and may even improve sexual function in patients presenting with bulk symptoms. ^72,73

Complications

Arteriographic complications (such as hematoma, bleeding, and puncture-site arterial thrombosis) and complications related to the use of closure devices ²⁰ have been reported. Most complications are minor, rarely requiring aggressive thrombolytic therapy. One case, however, did require a femoral-femoral bypass. ⁴⁹ There are also reports of uterine infections, most minor, but a few required hysterectomies, and one case resulted in fatal septicemia. ²⁰ There are reports of uterine ischemia and perforation. There have been cases of deep vein thrombosis and two reported cases of pulmonary embolism, one resulting in death. ²⁰ There are a total of four known cases of death following UAE, the other two deaths being from sepsis. ²⁰

Radiation exposure

Radiation exposure during UAE is of particular concern and deserves special attention. In a 20-patient series published by Nikolic et al, ⁷⁴ the mean estimated ovarian dose was 22.34 cGy. This dose is equivalent to one to three barium enema examinations, and is unlikely to result in acute or long-term radiation injury or genetic risk to the patient. However, the dose range reported in the same series was 4 to 65 cGY. ⁷⁴ It is important to note this variation in radiation dose, which was attributed to inconsistent fluoroscopic times. With more experience, the same group of physicians was able to decrease the radiation exposure by approximately 50% using pulsed fluoroscopy, avoiding magnification and oblique fluoroscopy, and by simultaneous bilateral embolization and arteriography. ⁷⁵

Conclusion

Uterine artery embolization is a very promising and minimally invasive procedure with extremely favorable results. SCVIR (which recently changed its name to the Society of Interventional Radiology) has a comprehensive research strategy to validate UAE as a primary treatment for symptomatic uterine fibroids and has created the UFE registry. The interventional community as a whole is committed to research and information gathering so that patients may benefit from a state-of-the-art procedure that provides better care with fewer adverse effects.