Uterine Artery Embolization for Symptomatic Uterine Myomas: Japanese Society of Interventional Radiology Procedural Guidelines 2021: English Version.

There were a total of seven RCTs comparing the efficacy of UAE with surgery for symptomatic uterine myomas, six of which were accepted into the Cochrane Library [125] in 2014, and the remaining one [126] was an additional 10-year report of one of them [107] . The breakdown was as follows: three [107 - 109] compared within 2 years with total hysterectomy, one [107] compared 5 years later, one [126] compared 10 years later, two [112 , 127] compared within 2 years with total hysterectomy or myomectomy, one [112] compared 5 years later, and one [111] compared within 2 years with myomectomy. Thus, patient satisfaction within 2 years after UAE was comparable to that of surgery (ranging from 41% lower to 48% higher for UAE, OR 0.94, 95% CI: 0.59-1.48, 6 RCTs, 640 cases, I 2 = 5%, certainty of evidence medium) [125] . Patient satisfaction at 5 years was similar to that of surgery (OR 0.90, 95% CI: 0.45-1.80, 2 RCTs, 295 patients, I 2 = 0%, moderate certainty of evidence) and at 10 years (OR 0.58, 95% CI: 0.27-1.28, 1 RCT, 156 patients) [125 , 126] . Quality improvement guidelines from the SIR in the US summarize that reduction in pressure symptoms is observed in 88% to 92% of patients, disappearance of abnormal genital bleeding is observed in >90%, and symptoms disappear in 75% of patients [20] . In terms of length of hospital stay, in all six RCTs described above, UAE was associated with a lower length of hospital stay than surgery (mean: 1.3-4.2 days for UAE vs. 3.5-7.6 days for surgery, 6 RCTs) and a faster return to normal activities after treatment (mean: 9-18.9 days for UAE vs. 21.7-39.8 days for surgery, 6 RCTs) [125] . Regarding re-treatment, in the seven RCTs mentioned above, UAE had a higher re-treatment rate than surgery (within 2 years: OR 3.72, 95% CI: 2.28-6.04, 6 RCTs, 732 cases, I 2 = 45%, certainty of evidence moderate; after 5 years: OR 5.79, 95% CI: 2.65-12.65, 2 RCTs, 289 cases, I 2 = 65%; after 10 years: OR 2.77, 95% CI: 1.30-5.92, 1 RCT, 156 cases) [125 , 126] . The 2-year re-treatment rate was 7% after total hysterectomy or myomectomy compared with 15%-32% after UAE. A literature review compared the 2-year recurrence rate with myomectomy [111] ; however, the confidence interval was very wide and therefore inconclusive (OR 1.32, 95% CI: 0.38-4.57, 1 RCT, 120 cases) [125] . According to the Fibroid Registry for Outcomes Data (FIBROID), a database of approximately 2000 cases in the US, 14.4% of patients underwent re-treatment (total hysterectomy, myomectomy, or re-UAE) 3 years after UAE, and this is considered to be due to the presence of poorly treated myomas or the development of new myomas [20 , 128] . In addition, the US FDA issued a safety notice in April 2014 that the use of motorized morcellators in laparoscopic uterine myomas surgery is not recommended because of the risk of intraperitoneal seeding of cancer tissue, particularly uterine sarcomas [11] . Approximately 1 in 350-500 patients undergoing total hysterectomy or myomectomy for uterine myomas have sarcoma [11 , 12] . In light of this, the aforementioned guidelines from the SIR state that although UAE is unlikely to cause tumor dissemination, informed consent is necessary to ensure that malignancy may be missed and treatment delayed by undergoing UAE [20] . In addition, at this time, although fertility preservation after UAE is debated, it is considered appropriate to perform UAE in patients who have no desire to become pregnant or give birth (see CQ6). Based on the aforementioned results, UAE for symptomatic uterine myomas is associated with a slightly higher re-treatment rate than surgery, but requires fewer days of hospitalization and allows for return to daily activities. UAE is considered to have similar levels of patient satisfaction. It is reasonable to consider UAE as a treatment option for symptomatic premenopausal women with symptomatic uterine myomas who do not wish to become pregnant in the future, after informed consent is given regarding the possibility of malignancy. There are no Cochrane reviews on the efficacy of UAE for uterine adenomyosis. The adopted literature was observational studies. A RCT is currently underway in the Netherlands [129] . There is limited information about the efficacy of UAE for adenomyosis and myomas associated with adenomyosis [130 - 142] . Uterine adenomyosis is often associated with myoma and is a challenge when determining the efficacy of UAE for adenomyosis alone. One review article evaluated the rate of symptom improvement in pure adenomyosis and myoma complicated by adenomyosis in the short and long term and found that 83.3% of patients with pure adenomyosis and 92.9% of patients with adenomyosis with myomas were treated in the short term. In the long term, there was a significant difference between the two: pure adenomyosis 64.9%, and adenomyosis with uterine myomas 82.4% [142] . UAE for myoma with adenomyosis and UAE for adenomyosis alone may be a future treatment option; however, the evidence is less reliable. There are no studies on the efficacy of UAE for endometriosis-associated adenomyosis. The indication for spherical embolic material in the package insert is “ arterial embolization of patients with hypervascular tumors or arteriovenous malformations, ” with no recommendations for uterine adenomyosis. For the UAE of symptomatic uterine myomas, spherical embolic substances such as trisacryl gelatin microspheres (Embosphere) [22 , 72 , 143 - 154] , polyzene F-coated hydrogel microspheres (Embozene) [27 , 134 , 155] , PVA, and acrylamido PVA (aa-PVA; Beadblock) [19 , 22 , 156] are used worldwide. Of these, Embosphere (Merit Medical Japan) was listed as an indication for hypervascular tumors or AV malformation in Japan in March 2013 and has been widely used since then. Although DC Bead (Boston Scientific Japan) had its indication extended to hypervascular tumors, including uterine myomas, uterine myomas were excluded from the indication in 2018. Embosphere is a hydrophilic, nonabsorbable, biocompatible spherical embolization material made of acrylic copolymer with porcine gelatin impregnation and coating. Because of its relatively high elasticity and hardness, and resistance to deformation, it is considered to have good depth distribution and is less likely to aggregate or redistribute. The 90%+ infarction rate of UAE using Embosphere is 82%-97% [22 , 143 , 144 , 146 , 147] . In contrast, gelatin sponges have been widely used in Japan since before spherical embolization materials were covered by insurance. Gelatin sponge is a porous structure made of gelatin extracted from bovine and porcine skin and ligaments [157] . As a temporary embolic agent, it has advantages such as an inexpensive price, the ability to cut into smaller strips, high visibility that makes it easy to evaluate the endpoint of embolization. Currently, there are four types of gelatin sponges commercially available in Japan: Sponzel (LTL Pharma), Gelfoam (Pfizer), Serescue (Astellas), and Gelpart (Nippon Kayaku); however, as of 2024, they have not been approved for the treatment of symptomatic uterine myomas. To date, only a few RCTs have provided a high level of evidence regarding the use of embolic substances in UAE for uterine myomas. Five RCTs and one systematic review have been reported for Embosphere and PVA (Embosphere vs. nonspherical-PVA 1 [143] , Embosphere vs. spherical-PVA 4 [72 , 144 - 146] , and Embosphere vs. spherical-PVA systematic review [158] ); however, the radiological evaluation, clinical evaluation, and evaluation period used have not been consistent. Most previous reports on the outcomes of gelatin sponge therapy [69 , 71 , 159 - 161] originated in Japan. The medium- to long-term efficacy and safety of gelatin sponge were first demonstrated by Katsumori and colleagues [159 - 161] . Subsequently, Sone et al. [71] reported the results of a multicenter phase I/II prospective single-arm study. Based on these outcomes, the JSIR guidelines on the use of gelatin sponge particles in embolotherapy [157] recommended the use of gelatin sponges in UAE for uterine myomas. Subsequently, prospective single-arm [162] and retrospective studies [69 , 163] comparing porous gelatin sponge granules (Gelpart) with gelatin sponge shredded using conventional cutting methods were reported, and the outcomes were similar. Reports comparing gelatin sponge with other embolic materials remain scarce. Single-center, prospective, nonrandomized studies [164] , a small number of RCTs [165] , and a single-center, retrospective study [166] with postoperative pain as an outcome have been reported comparing gelatin sponge and Embosphere. A single-center, prospective, nonrandomized study [166] using infarct rate as the outcome showed significant differences in the percentage of patients with ≥90% infarct rate on contrast-enhanced MRI at 1 week (94.4% [17/18] in the Embosphere group and 93.6% [44/47] in the gelatin sponge group), and improvement in symptoms and quality of life at 4 months and adverse events at 4 months were not significantly different between the Embosphere and gelatin sponge groups. In contrast, a randomized controlled study [165] comparing the Embosphere group (10 patients) with the gelatin sponge group (10 patients) showed no significant differences in maximum myoma diameter, UFS-Qol symptom severity score, UFS-Qol HRQL score, laboratory data including inflammation and hormone levels, mean fluoroscopy duration, and radiation exposure. However, the Embosphere group had a significantly smaller uterine volume at 6 months than that of the gelatin sponge group (p = 0.01). In a single-center retrospective study [166] comparing the gelatin sponge group (49 patients) and the Embosphere group (52 patients) for postoperative pain (mean and maximum visual analog scale score within 24 hours), Embosphere was associated with less pain than gelatin sponge. For treating large or hypervascular uterine myomas that require a large amount of embolization material, embolization with a certain amount of Embosphere may be followed by additional embolization using gelatin sponges. This helps reduce medical costs. In a retrospective comparison of 106 patients embolized with Embosphere alone and 123 patients embolized with Embosphere followed by additional embolization with gelatin sponge [167] , there was no significant difference in treatment efficacy, but the Embosphere group used an average of 2.7 vials less Embosphere than the group embolized with Embosphere alone (p = 0.0001), resulting in a $708.90 (42%) reduction in medical costs associated with embolization materials. This report did not find a significant increase in pain at 24 hours post-UAE when embolization was performed using Embosphere, followed by additional embolization with gelatin sponge. Another prospective, nonrandomized, controlled trial [168] compared nine patients embolized with Embosphere, followed by additional embolization with gelatin sponge and eight patients embolized with gelatin sponge alone. This study found that embolization using Embosphere, followed by additional embolization with a gelatin sponge may cause more pain than embolization using a gelatin sponge alone. As described above, there are no reports with a high level of evidence regarding the use of gelatin sponges for additional embolization after embolization with spherical embolic material. The warning document on pulmonary embolization via AV shunt or AV fistula in uterine myomas embolization (Nippon Kayaku Embosphere Proper Use Information) [76] recommends that when myoma staining does not decrease on angiography even after a certain amount of Embosphere is administered, the presence of a shunt should be suspected and measures such as changing to a larger embolic material, increasing the particle size, or discontinuing administration should be considered. In this case, the “ larger embolic material ” could be a large piece of gelatin sponge or a metallic coil that does not easily pass through AV shunt. Pain, the main symptom of postembolization syndrome, is an important clinical issue because it is almost inevitable after UAE and requires intense analgesia in most cases. Pain lasts approximately 24 hours immediately after treatment, peaking at approximately 7 hours [169 - 171] . Various types of analgesia have been reported, including intravenous and subcutaneous infusions of opioids (narcotic analgesics such as morphine and fentanyl and narcotic antagonist analgesics such as pentazocine), epidural anesthesia, nerve block, intra-arterial lidocaine infusion, and general anesthesia, with various drug and dosage combinations. Although there is no standard drug or dosage regimen, there is a consensus that adequate pain management is necessary during and after treatment. Intravenous and subcutaneous injection of opioids using patient-controlled analgesia (PCA) is widely used as an analgesic method in the intraoperative and early postoperative period for the UAE, and numerous reports are available. Morphine and fentanyl are the most common opioids used for PCA. Non-RCTs have shown that morphine is more analgesic than fentanyl [169] , and RCTs have shown that short-acting remifentanil and morphine provide comparable analgesia, with remifentanil providing better pain relief within the first 4 hours [172] . However, no convincing evidence supports the recommendation of a specific drug. In addition, in these studies, opioids were not used alone for analgesia but were used in combination with other analgesics. The authors concluded that there was no significant difference in analgesic efficacy between “ opioids, NSAIDs, and acetaminophen ” alone and other combination therapies [173] . Regarding side effects, although nausea and vomiting, which are the primary side effects of opioids, are present, none of them are severe. However, RCT results showing improved analgesic efficacy and decreased side effects with the addition of dexmedetomidine [174] and dexamethasone [175] have been reported and may be considered in cases of substantial side effects. Regarding intraoperative, intra-arterial lidocaine infusion, an RCT compared three groups, “ mixed infusion with embolic material, ” “ infusion after embolization, ” and “ control ” [176] . No difference was found between the two groups for lidocaine intravenous infusion, with significantly lower pain scores at 4 hours and no significant difference in pain scores at 7-24 hours compared with the control group, but with lower analgesic doses. Two other RCTs have shown the efficacy of lidocaine infusion [177 , 178] , and although it is combined with various analgesic and sedative agents, the bias was reduced by randomization, and evidence for the analgesic effect of lidocaine infusion was “ moderate. ” PVA was used as the embolization material in these three RCTs, the use of trisacryl gelatin microspheres (Embosphere, Nippon Kayaku), covered by insurance in Japan, was not included in this guideline. Two studies have been published after the systematic review for this guideline was conducted. In 2019, Katsumori et al. [179] reported the results of a nonrandomized, retrospective, comparative trial. They found that the pain scores at 3 hours after embolization were lower in the group receiving intravenous lidocaine (80 mg in 20 mL of saline) than in the control group. However, no significant difference in pain scores or analgesic use occurred up to 24 hours after embolization. In 2020, Duvnjak et al. [180] reported an RCT in which pain scores at 2 hours were significantly lower in the “ embolization followed by lidocaine infusion (200 mg/20 mL) ” group, and morphine use was also lower. However, there was no significant difference in pain scores between the two groups at 4-24 hours [180] . In both studies, there were no serious adverse events, and the pain at 2-3 hours was adequately controlled, but no consensus has been reached on dosage or dilution. In Japan, as of 2024, lidocaine products are not covered by insurance for intra-arterial infusion. Several reports have demonstrated the efficacy of performing superior epigastric nerve block in conjunction with UAE. Yoon et al. [181] , in a double-blind RCT in 2018, showed that the superior epigastric nerve block group required significantly fewer analgesic narcotics and antiemetics than those by the sham procedure group. Epidural anesthesia is also used: however, evidence is scarce. Given potential severe adverse events caused by nerve blocks and epidural anesthesia, evidence of their safety and efficacy compared with conventional pharmacotherapy is needed to make them standard supportive care for the UAE. The certainty of evidence recommending prophylactic antimicrobial administration is low; no studies have assessed the need for prophylactic antimicrobials in the UAE. An overseas review of prophylactic antimicrobials during IR procedures [182] noted that the need for prophylactic antimicrobials in UAE is unclear and stated that prophylactic antimicrobials are routinely administered in many cases reported in studies. Statements in a Cochrane review [125] regarding the effects of UAE on pregnancy and childbirth can be summarized as follows: In addition to pain, postembolization syndrome, and infection, complications of UAE include ovarian dysfunction, amenorrhea secondary to endometrial atrophy and luminal adhesions, and unknown effects on fertilization and pregnancy [183] . Potential effects on the uterine adnexa after UAE include ovarian dysfunction due to decreased ovarian blood flow, fallopian tube damage due to infection and resulting infertility, multiple reports of pregnancies and births after UAE, and reports of uterine rupture at birth. There have also been reports of increased miscarriage rates [184] and placental abnormalities, such as placental adhesions [185] ; thus, UAE may have various effects after pregnancy is established. The Cochrane review examined two RCTs on the fertility of UAE (Mara et al. [111] and REST2011 [127] ), and the literature other than the studies included in the Cochrane review included zero RCTs, zero non-RCTs, and three observational studies [121 , 186 , 187] (all single-arm). Mara et al. [111] was the only study to mention childbirth, in which UAE was correlated with a lower birth rate than with myomectomy (17 pregnancies, five deliveries, and nine miscarriages out of 26 in the UAE group; 33 pregnancies, 19 deliveries, and six miscarriages out of 40 myomectomies). However, this is low-quality evidence from a subgroup of patients (those who wanted to have a baby) and should be treated with great caution. Ovarian dysfunction manifests as transient or permanent amenorrhea due to unintentional embolization of the ovarian artery. Menopause (permanent amenorrhea) after UAE occurs in only 2%-3% of patients younger than 45 years and in approximately 8% of those older than 45 years [188 , 189] . Within 3 years after UAE, 1.6% of patients younger than 40 years had amenorrhea, although the association with UAE is unknown [128] . An RCT comparing postoperative anti-Müllerian duct hormone levels 24 months after UAE and after total hysterectomy also found that they were lower in the UAE group [190] . Ovarian function declines spontaneously with longer postoperative observation periods, making it difficult to assess whether the decline is due to aging or complications. In a Cochrane review of ovarian dysfunction, three studies (Mara et al. [111] , EMMY2010 [107] , and REST2011 [127] ) evaluated changes in postoperative follicle-stimulating hormone (FSH) levels in comparison with non-UAE treatment groups. Among them, two reports (EMMY2010 and REST2011) evaluated the rate of decrease in FSH levels above 40 IU/L over a 2-year follow-up period and found no significant difference between the UAE and non-UAE treatment groups (OR 1.01; 95% CI 0.53-1.94, 297 patients). Furthermore, reports on the rate of decline >10 IU/L at 6 months postoperatively [111] did not show sufficient results to conclude that there was a significant difference (OR 4.80; 95% CI: 0.97-23.64, 120 patients). Future RCTs with refined designs that clarify the differences in pregnancy and delivery rates among the treatment modalities for uterine myomas are warranted. Further data are also awaited on the association between UAE and pregnancy-related complications, such as preterm labor, intrauterine growth restriction, and postpartum hemorrhage. Currently, it is appropriate to obtain informed consent carefully and appropriately and to use UAE in patients who do not desire to become pregnant or give birth. However, even if there is a desire for pregnancy and childbirth, in cases with a high risk of systemic management of surgery due to significant comorbidities or medical history, or in cases in which surgical treatment such as enucleation is diagnosed as high risk (leiomyomatosis, multiple myoma, cervical myoma, etc.), it may be necessary to consider the following factors to determine fertility after UAE. After informed consent is obtained, UAE should be considered as an alternative to surgical treatment.

Dr. Hiroki Higashihara, a member of the Editorial Board of Interventional Radiology, is an author of this paper. This author was not involved in the peer-review or editorial decision-making process.

CQs and Recommendations are presented in Table 4 . CQs and Recommendations. CQ: clinical question; UAE: uterine artery embolization

This guideline is based on its Japanese version, which was published in 2022 on the website of JSIR.

The indications for UAE for uterine myomas are symptomatic uterine myomas confirmed by ultrasound, MRI, or other imaging studies [17] ; symptoms that are difficult to control with drug therapy; premenopausal age [18] ; and no desire for surgery or no indication for surgery because of comorbidity or disease history. Symptoms of uterine myomas that may be treated by UAE include excessive menstrual periods, anemia, dysmenorrhea, and mass effects from the myomas themselves, such as pain (including pain during intercourse), abdominal distention, urinary urgency, urinary frequency, urinary retention, hydronephrosis, and constipation [19 , 20] . Absolute contraindications for UAE of uterine myomas are pregnancy, active pelvic inflammation, and malignancy or suspected malignancy of the uterus or adnexa [20 - 23] . Particular attention should be paid to malignant tumors such as sarcomas because malignancy treatment delay due to inappropriate UAE must worsen the prognosis [20 , 24] . Relative contraindications include abnormal coagulation disorder, severe iodine contrast medium allergy, and renal dysfunction [17 , 20 , 22 , 25 , 26] . Other contraindications include the presence of an intrauterine device, history of pelvic radiation therapy, and recent history of pseudomenopause therapy (see 3.3 Preoperative Hormonal Therapy for details), which may decrease the efficacy of treatment [27] . Some authors believe that large myomas (e.g., >8 cm in diameter) should be avoided because of the increased risk of complications, whereas others have reported no relationship between myoma size and complications [25 , 28] . The incidence of UAE-associated myoma infarction decreases with smaller myoma diameter [29] . The Society of Interventional Radiology (SIR) guidelines do not have size restrictions [20] ; however, caution should be exercised when performing UAE for large or small myomas. For stalked serous submucosal myomas (stem diameter <50% of myoma diameter), early case reports have shown that stalk torsion, ischemic necrosis of the myoma, or separation of the myoma from the uterus can occur [30 , 31] . However, numerous case reports have revealed no differences in efficacy or complications [32 - 35] . The SIR guidelines state that it should not be considered off-label [20] . There have been subsequent reports that the tumor necrosis rate is low for stalked subserosal myomas [35] , whereas others have reported that the infarction rate of stalked subserosal myomas is similar to that of myomas located in other sites [29] , and it is still under evaluation. Therefore, careful and appropriate informed consent should be obtained when performing UAE for stem subserosal myoma, considering the aforementioned circumstances. Some specific locations of the tumors, including the uterine cervix, lower body, and anterior wall, have also been reported as potential factors for low rates of infarction after UAE [29] . Before performing UAE for uterine myomas, a medical interview, physical examinations, general preoperative workup, and imaging studies should be conducted. During the interview, the patient's menstrual cycle, blood volume, and accompanying signs and myoma-related symptoms (e.g., hypermenorrhea, dysmenorrhea, bulk symptoms, etc.) should be checked. In addition, comorbidities, medical history, and past treatments (pelvic surgery, hormone therapy, etc.), and history of allergies (particularly to iodine contrast media and local anesthetics) should be confirmed. As a rule, patients who wish to become pregnant in the future are not eligible for the therapy, so it should be confirmed whether they expect a baby or not. Patients can be informed about other treatment options, including surgery (myomectomy, total hysterectomy, etc.) and hormonal therapy. Furthermore, the patient must be adequately evaluated by the gynecologist for the absence of cervical or body cancer with cytological diagnosis. Other general preoperative tests, such as blood samples (blood cell count, platelets, blood biochemistry tests including coagulation capacity, inflammatory response, renal function, and D-dimer, etc., and infectious diseases), electrocardiogram, and simple chest X-ray should be performed, as in other vascular interventions. Imaging studies for uterine myomas include ultrasound and MRI. MRI should be performed because ultrasound alone may not accurately identify the myoma's primary location [36] , and because it is difficult to distinguish it from sarcomas and to evaluate other diseases. It is important to know the location of the myoma in advance because submucosal myomas have the potential for delivery [37] , and there are reports of a low infarction rate for stalked subserosal myomas [35] , as well as for cervical myomas, myomas of the lower uterine body, myomas of the anterior body wall and small myomas [29] . A detailed MRI evaluation for possible malignancy is necessary because an incorrect UAE for malignancy, such as uterine leiomyosarcoma, may delay treatment and worsen the prognosis [20 , 24] . MRI findings suggesting malignant tumors include a sole mass, high signal on T2-weighted images, T1-weighted images, diffusion-weighted images, low apparent diffusion coefficient values, and gradual and heterogenous enhancement, endometrial thickening, ascites, etc., have been reported as additional findings [38] . Other findings, such as large size, irregular margins, and presence of metastases, should also be noted, but differentiation is not always easy, and the possibility of malignancy should be kept in mind in cases of poor response to treatment after UAE or residual contrast effect [24] . Contrast-enhanced MRI using a gadolinium contrast agent allows the evaluation of blood flow in uterine myomas, which is beneficial for estimating treatment efficacy and determining treatment efficacy during follow-up (the necrosis rate of the largest myoma [poor contrast zone] is proportional to the symptom improvement rate) [39 , 40] . Moreover, additional contrast-enhanced MRA and non-contrast MRA imaging can provide an understanding of pelvic arterial anatomy (the common iliac artery bifurcation, the internal and external iliac artery bifurcation, the uterine artery origin, etc.) [40] . In particular, since there are many anatomical variations in the origin of the uterine artery, it has been reported that imaging the origin contributes to a reduction in the number of angiographic scans, radiation dose, procedure time, and contrast medium used [40 - 43] . In addition, hypoplasia of the uterine artery and predominant blood flow in the ovarian artery can be evaluated, contributing to preoperative prediction of treatment efficacy and determination of treatment strategy [43] . Contrast-enhanced CT can also be used to evaluate blood flow and to understand vascular anatomy, but it is usually not performed due to radiation exposure issues for nonelderly patients and patients with nonmalignant tumors, and the difficulty in differentiating sarcomas from myomas. Although there have been no reports of health hazards associated with MRI gadolinium contrast agents in recent years, it has been reported that the linear form is more likely to be deposited in the brain than the cyclic form [44] . In response, the FDA has issued a warning, and the European Medicines Agency has temporarily suspended the marketing of linear contrast media (except liver contrast media). The Japanese Ministry of Health, Labour and Welfare revised the “ Precautions for Use ” in November 2017, stating that the need for using with gadolinium contrast media should be carefully determined, the cyclic form should be the first choice, and the linear form should be administered when the cyclic form is not appropriate for use, so care should be taken in selecting contrast media. One of the hormonal therapies for uterine myomas is GnRH agonist therapy (pseudomenopause, which temporarily suspends ovarian function). A Letter to the Editor reported that the uterine arteries were smaller in diameter and more tortuous in the group that had undergone this therapy within 8 weeks before UAE, which may increase technical difficulties and the risk of complications, than in the group that had not [45] . This may be due to an increase in vascular endothelial growth factor caused by GnRH agonist [46] . However, others have reported that UAE safety, technical success rates [47] , and patient satisfaction [48] did not differ between patients who received GnRH agonist therapy from those who did not receive GnRH agonist therapy. Patients who have received GnRH agonist therapy are not excluded from UAE. Because it may affect the procedure and therapeutic effects, it is advisable to wait 8-12 weeks after the last GnRH agonist therapy before performing UAE. Cases in which low-dose estrogen-progestin combination (LEP) drugs are prescribed for dysmenorrhea are routinely encountered. In this case, the accompanying documentation states that its use is contraindicated for 4 weeks before surgery because of the increased risk of thrombosis (see also “ 3.6 Complications ” ). Although the UAE for uterine myomas is one of the common interventional procedures, there is no standardization regarding the choice of access vessels, myoma feeding vessels requiring angiography and embolization, type of catheters, and embolization techniques, which are discussed below. Embolic material, pain management, and prophylactic antimicrobial therapy are summarized in a separate CQ. In Japan, UAE is generally performed under local anesthesia. The left or right common femoral artery is often selected as the access vessel, and a 4- to 5-Fr short sheath is placed after puncture using the Seldinger technique [25 , 49] . The bilateral femoral arteries are occasionally approached to allow simultaneous contrast and embolization, which reportedly results in shorter fluoroscopy duration and lower radiation dose than those with unilateral puncture [50 , 51] . Recently, in Europe and the United States, the UAE via the upper extremity, particularly the left radial artery approach, as well as other interventional procedures, has been performed [52 - 54] . The radial artery approach does not require postural retention of the trunk after UAE, may relieve patient distress, and reduce the risk of deep vein thrombosis. Studies comparing the radial and femoral artery approaches reported no difference in fluoroscopy duration [54] and no difference in technical success rates or clinical outcomes [53] . The upper extremity approach has potential risks of hand ischemia, cerebral infarction due to passage through the aortic arch. Serious complications, including cerebral infarction, associated with transradial UAE procedure, except for transit radial artery occlusion, have been reported [53] . However, informed consent for these specific complications should be obtained when an upper extremity approach is chosen. As mentioned above, the unilateral femoral artery approach is often chosen for UAE; however, the choice should be based on preoperative MRA, vascular anatomy, and the advantages and disadvantages of each approach. The uterine artery is the main vessel that supplies blood to the uterine myomas, followed by the ovarian artery. The uterine artery is a branch of the internal iliac artery system. However, its branching patterns vary, with 45% reported as the first branch of the inferior gluteal artery, 6% as the second or third branch of the inferior gluteal artery, 43% originating at the bifurcation of the superior and inferior gluteal arteries, and 6% as the first or unknown branch of the internal iliac artery [55] . The ovarian artery branches ventrally from the abdominal aorta near the level of the second lumbar vertebra, just below the renal artery bifurcation, on both sides. However, because it sometimes branches from the renal artery [56] , aortography should be performed at the level of origin of the renal arteries [57] . The ovarian artery is known to anastomose with the uterine artery [57 - 59] . Razavi et al. [57] found that 49 (32.2%) of 152 bilateral ovarian arteries anastomosed with the uterine artery on aortography. Among these, 21.7% (type I) flowed from the ovarian artery to the uterine artery, 3.9% (type II) directly fed the myoma, and 6.6% (type III) exhibited a large amount of blood flow from the uterine artery to the ovarian artery. Additionally, the UAE was not effective in 3 of 6 patients with type II anastomosis [57] . A small case series of ovarian artery embolization for myoma with type I and type II anastomosis showed ovarian artery embolization did not increase the risk of permanent amenorrhea, and higher symptom recurrence rates were observed when ovarian artery embolization was not performed for those cases [60 , 61] . Sheikh et al. [59] reported that by carefully observing the washout of contrast medium during uterine arteriography with an injector, anastomoses between the uterine and ovarian arteries were observed in 178 (96.7%) of the 184 bilateral uterine arteries. Of these, when the anastomosis and ovarian vascular bed are visualized and prolonged (type III of the modified Razavi classification, 24%) or when the anastomosis is contrasted and washed out retrograde in the direction of the uterus (type Ib, 45%), it is a risk factor for ovarian dysfunction (7% of all cases) [59] . They recommend coil embolization of the uterine-ovarian artery anastomosis before UAE for those cases if available [59] . However, a systematic review [62] reported that the UAE has no significant effects on ovarian reserve, and the need for pre-UAE embolization of the uterine-ovarian artery anastomosis is controversial [63] . Although there is no consensus on the need for aortography or ovarian angiography, when the uterine arteries do not supply blood to the entire uterus or myoma, involvement of the ovarian arteries should be evaluated before or after embolization. Other vessels that supply blood to the uterine myomas include, although rare, the uterine circumflex artery (which forms an anastomosis between the inferior abdominal wall artery and the uterine artery branching from the external iliac artery) [64] , superior and inferior mesenteric arteries [65] , and bladder artery, a branch of the internal iliac artery. These approaches are considered when uterine or ovarian arteriography does not fully visualize the myoma or when preoperative MRA raises concerns about their involvement. The unilateral femoral artery approach requires some techniques in selecting the ipsilateral internal iliac artery on the puncture side, and various catheters are used [63] . After embolization of the contralateral side is completed, the catheter is pushed up to form a loop (Waltman loop) in the aorta and pulled back to the ipsilateral side to select the internal iliac artery [25 , 66] . In addition, a combination of a catheter with a small U-shaped tip called the Rosch inferior mesenteric catheter and a hydrophilic guidewire has been reported to be useful [67] . In Japan, Mohri-type (Mohri) catheters, which have a long inverted portion, are used because they are easy to form loops and have a tip that faces ventrally in three dimensions, making them suitable for uterine artery selection [68 , 69] . As mentioned above, the branching morphology of the uterine artery origin varies, and selection can be difficult; thus, the origin should be identified by preoperative MRA or intraoperative angiography in an oblique position (e.g., 30° contralateral) [40 , 70] . In the case of the radial artery approach, a 125-cm-long angled catheter is used because of the long distance and downward bifurcation of the iliac and uterine arteries [52 - 54] . When inserting a catheter into the uterine artery, crude manipulation and unnecessary guidewire insertion should be avoided because induction of vascular injury and spasm at the uterine artery origin can reduce the procedural success rate and therapeutic efficacy [25 , 66] . Although the uterine artery can be selected using a parent catheter, the microcatheter should be inserted without moving (or using) the guidewire as much as possible to prevent spasm [25] . In such cases, a high-flow microcatheter with a large lumen is recommended to ensure good passage of the embolic material because the particle size of the spherical embolization material used for UAE is generally ≥500 μm [69 , 71] . The following describes the general use of trisacryl gelatin microspheres (Embospheres), a spherical embolic material currently used for UAE of uterine fibroids in Japan. First, the Embospheres are mixed with the contrast agent (usually 1:1) and transferred into a syringe for injection. The syringe is rotated several times to distribute the particles evenly and injected slowly under fluoroscopy to avoid nontarget embolization. Embolization failure or recanalization may occur if particles aggregate in the proximal vessel. Spherical embolic material is generally diluted as needed to penetrate the peripheral tumor vasculature. However, in the case of UAE, the blood flow is abundant, and the required volume of embolic materials may be high, and excessive dilution is not essential because it leads to increased contrast and radiation dose. Regarding size, small particles carry the risk of uterine necrosis. The Embosphere package insert states that particles ≥500 μm should be used for uterine fibroids, and a formulation ≥500 to 700 μm should be used. Some studies have suggested that ≥700 μm is better for preventing influx into the ovarian artery [25 , 61] . The endpoint of embolization is considered good when the contrast agent injected under fluoroscopic guidance stagnates in the ascending branch of the uterine artery for about five heartbeats [72] . Because blood flow may reopen after embolization, angiography should be performed approximately 5 minutes after the end of the injection to ensure that staining of the uterine fibroids disappears and the ascending uterine artery is preserved [73] . Unlike gelatin sponges, spherical embolic material may take longer to achieve stagnation of blood flow, making it difficult to determine the endpoint; however, it is not necessary to embolize to the extent that the uterus no longer shows up on the imaging, as with gelatin sponges. Care should be taken to avoid excessive injection. In contrast, gelatin sponge can be used in two ways: (1) cutting using a scalpel and scissors, or (2) pumping using a syringe and a three-way stopcock. While the latter method is quick to prepare, it contains a relatively large number of microfragments and results in variations in fragment diameter [74] . The sponge is shredded into 1-mm pieces, mixed with contrast medium, and injected to embolize the ascending branch of the uterine artery to an extent that it becomes stagnant [71] . The uterine artery is rich in blood flow and rarely forms AV fistulas. There have been reports of fatal pulmonary embolization resulting in death during UAE [75] . Always be mindful of the presence of an AV shunt and check for early venous return on uterine arteriography (particularly in cases of large myomas or abundant blood flow, as the shunt may be difficult to see) [76] . Intermittent fluoroscopy during embolization is recommended to assess decreased blood flow. If blood flow is not decreased, angiography should be performed as appropriate to assess for decreased myoma staining and the presence of AV shunt. If myoma staining does not decrease, the presence of a shunt should be suspected, and consideration should be given to increasing the particle size, changing to a larger embolization material, or discontinuing administration. Furthermore, the patient's condition should be monitored during treatment, and changes in arterial blood oxygen saturation must be carefully monitored [76] . There are no reports of radiation-related disorders due to UAE for uterine myomas [20] . Therefore, it is necessary to understand the radiation dose and the risk of radiation injury and to reduce the radiation dose [20] . Since 2000, there have been many reports on radiation exposure in UAE from Europe and the United States. In these reports, the fluoroscopy duration ranged from 9 to 27 minutes, the dose area product ranged from 30.6 to 155.1 Gy/cm 2 , and the average estimated ovarian exposure ranged from 7 to 223 mGy [41 , 50 , 51 , 77 - 86] . In the early reports, there were many cases in which pulsed fluoroscopy was not used, and over time, the radiation dose tended to decrease because of advances in equipment, the establishment of methods to reduce radiation exposure, and increased awareness [78 , 87 , 88] . Therefore, although stochastic effects are unknown, the possibility of local skin damage or infertility, which are definitive effects, occurring with a single UAE exposure dose is considered minimal [83 , 87] . To reduce radiation exposure during UAE, a smaller number of shots, avoiding magnified fluoroscopy, and limiting oblique angiography and aortography are recommended [20 , 88] . In particular, because aortography accounts for >20% of the total radiation dose, selective contrast (e.g., ovarian artery) should be used as much as possible [20 , 88] . There are also reports that cone-beam CT can be used instead of aortography to confirm the embolization area while reducing radiation exposure [89] , and the bilateral approach using bilateral femoral artery puncture can reduce radiation exposure without compromising treatment results [51] . In addition to general methods to reduce radiation exposure, such as pulsed fluoroscopy, narrowing the irradiation field, and reducing the frame rate, other methods include roadmap imaging [88] and last image hold [77 , 88] . DSA should be avoided as much as possible. In contrast, in UAE, the irradiation field is the pelvis, and in the femoral artery approach, the irradiation field is close to the operator. The operator's radiation exposure should be reduced by moving away from the irradiation field as much as possible, keeping hands out of the irradiation field, avoiding manual DSA, wearing protective eyewear and protective clothing, and using shields and protective curtains. Several review and commentary articles have been published on the complications of UAE for uterine myomas [20 , 90] . In addition to arterial puncture complications, postembolization syndrome, deep vein thrombosis, and pulmonary artery thromboembolism, which are common with arterial embolization, pain (pelvic pain), chronic banding [91] , myoma delivery [92 - 94] , uterine and adnexal infection, and ovarian dysfunction are known to be characteristic of UAE. Complications such as myoma abscess [90 , 95] , tubo-ovarian abscess [69] , uterine necrosis [96 , 97] , sepsis [98 , 99] , partial uterine wall defect [100] , bowel perforation [101] , and bowel obstruction [102] have also been reported. Serious complications resulting in death included two cases of sepsis and multiple organ failure [98 , 99] , one case of pulmonary embolism [103] , and one case of nontarget embolization due to AV fistula and foramen ovale opening within the myoma [75] . Although both of these serious complications are infrequent, the operator should be fully aware of their risks and explain them to the patient being treated preoperatively. After UAE, patients may present with subjective symptoms such as pain (pelvic pain), fever, nausea, vomiting, decreased appetite, discomfort, or abnormal laboratory values, such as elevated levels of C-reactive protein (CRP) or increased white blood cell count. These are called postembolization syndromes, and they usually improve within a week or so after UAE. If subjective symptoms or laboratory abnormalities do not improve or recur, serious complications, such as ischemic necrosis, infection, and associated sepsis, should be considered and managed appropriately. Pain is often manifested as pelvic pain and is almost always present. The degree of pain varies from mild to severe. However, depending on the symptoms, pain management is necessary during the intraoperative period and several days after the procedure. Fever is also frequently observed. In some cases, there is excretion of a blood-tinged bandage [20] , which may subside within 2 weeks after UAE but may persist for several months or longer and require transvaginal treatment [91] . Submucosal myomas and intramuscular myomas in contact with the endometrium may protrude into the uterine cavity and be excreted into the cervix or vagina during follow-up after UAE, which is known as myoma delivery [20] . Although they are often excreted spontaneously without serious symptoms, if they become lodged in the cervix or remain vaginally, they are associated with labor-like lower abdominal pain and foul odor and require gynecological procedures to drain them because of the increased risk of conversion to myoma infection via vaginal ascending infection [92 - 94] . When infection or inflammation in the uterine and adnexal regions is suspected, antibiotics (intravenous, oral, or vaginal) should be administered; if this is not effective, surgical drainage, transvaginal myomectomy, or other procedures may be performed, and ultimately a total hysterectomy may be necessary. Prompt simple abdominal total hysterectomy should be performed to save lives, particularly if the patient has transitioned or is at high risk of transitioning from uterine infection to sepsis. Nontarget embolization is caused by the unintentional influx of embolic material into vascular-dominated areas outside the target region, which may affect the ovaries, bladder, bowel, muscles, nerves, lower extremities, etc. Nontarget embolization results in the infarction of various organs and associated pain. Transient amenorrhea is often observed; however, embolization of the ovarian artery can cause permanent ovarian dysfunction [59] . The SIR guidelines summarize that amenorrhea occurs in 0%-3% of patients younger than 45 years and in 20%-40% of patients older than 45 years [20] . In addition, based on overseas reports of deaths from nontarget embolization (pulmonary embolization and cerebral infarction) due to AV fistula within the myoma and patency of the foramen ovale [75] , a warning document (Nippon Kayaku Embosphere Proper Use Information) has been published regarding pulmonary embolization via AV shunt and AV fistula during uterine fibroids embolization [76] . Because coagulation is transiently elevated after UAE and rest is necessary, prophylactic measures against deep vein thrombosis and associated pulmonary thromboembolism (wearing elastic stockings, avoiding unnecessarily prolonged rest, encouraging ankle exercises, etc.) should be taken during the perioperative period. Low-dose oral contraceptives (OCs), LEP, etc., are risk factors for deep vein thrombosis. Uterine effects after UAE include endometrial atrophy, uterine necrosis, partial uterine wall defects, adhesions, and Asherman's syndrome [20] . The results of a meta-analysis of seven randomized controlled trials (RCTs) [104 - 110] involving a total of 793 patients comparing the efficacy and safety of UAE with non-UAE treatments for symptomatic uterine myomas were published in the Cochrane Database of Systematic Reviews (UAE for symptomatic uterine fibroids) [111] . The control groups (non-UAE treatment groups) in the RCTs included in the analysis were total hysterectomy in three trials [104 - 106] , myomectomy in two trials [107 , 108] , and total hysterectomy or myomectomy in two trials [109 , 110] . In total, 53 patients were randomized to total hysterectomy and 62 patients to myomectomy. The incidence of serious complications in the control group for total hysterectomy was approximately 3% [112] , whereas that for myomectomy, although unknown, was considered to be lower. The meta-analysis concluded that the risk of complications was not higher in the UAE and control groups for any of the several safety endpoints, and there were no significant differences between the two groups. In addition, a meta-analysis [113] of eight RCTs comparing complications between surgical treatment and UAE did not find a significant difference between surgical treatment and UAE in terms of complication risk. However, UAE has been shown to have a relatively higher incidence of minor complications and a higher rate of requiring additional treatment within 2 and 5 years. Specific analysis results presented by the aforementioned meta-analysis are as follows: The OR for intraprocedural complications was 0.91 (95% CI: 0.42-1.97, 4 trials, 452 cases, I 2 = 40%, low-quality evidence), the OR for major complications within 1 year was 0.65 (95% CI: 0.33-1.26, 5 trials, 611 cases, I 2 = 4%, moderate quality of evidence), and the OR for serious complications within 5 years was 0.56 (95% CI: 0.27-1.18, 2 trials, 268 cases). The OR for minor complications within 1 year was 1.99 (95% CI: 1.41-2.81, 6 trials, 735 cases, I 2 = 0%, moderate quality of evidence). In the 350 patients randomized to UAE (mean age 41.90 years, mean follow-up 1022.63 days), the most frequent complications were bilateral UAE failure (4.00%), leukorrhea and fever (4.00%), postembolization syndrome (2.857%), and deep vein thrombosis (0.286%). The less frequent complications were severe vasovagal reflex (0.286%) and retained uterine hematoma (0.286%). There were no deaths. In the 346 patients randomized to surgery, the most frequent complications were stress urinary incontinence (3.757%), compression symptoms (2.890%), and excessive menstrual periods (2.601%). The study also showed that UAE tended to require reintervention at a higher rate than non-UAE treatment (control group) (OR 3.72; 95% CI: 2.28-6.04, 6 trials, 732 cases, I 2 = 45%, moderate quality of evidence; the re-treatment rate within 5 years was OR 5.79; 95% CI: 2.65-12.65, 2 trials, 289 cases, I 2 = 65%). If 7% of patients require additional treatment within 2 years after total hysterectomy and myomectomy, 15%-32% of patients will require additional treatment within 2 years after UAE. The study also showed a relatively low rate of blood transfusion requirements in the UAE group (OR 0.07; 95% CI: 0.01-0.52, 2 trials, 277 cases, I 2 = 0%). In the Embosphere package insert [114] , the risk of nontarget embolization due to AV fistula or regurgitation of embolic material particles is listed as a potential serious adverse event that may occur during UAE for uterine myomas. Other gynecologic adverse events listed include uterine bleeding, early menopause, amenorrhea, infection of the pelvic region, uterine/ovarian necrosis, vaginal discharge (band excretion), myoma shedding/myoma delivery, necrotic tissue removal, and hysterectomy, many of which are listed as having an unknown frequency. In the US clinical trial of Embosphere for uterine myomas, 132 patients were included in the safety analysis, and treatment-related adverse events occurred in 34 of these patients (25.8%). These included allergic reaction/rash in 6.1%, puncture site injury in 4.5%, myoma delivery or removal of myoma tissue in 3.8%, vaginal infection/vaginitis in 3.8%, pain in 3.0%, urinary tract infection/cystitis in 2.3%, vaginal irritation/burn/vaginal discharge (band discharge) in 1.5%, and hysterectomy after embolization in 0.8% (1/132 patients). The safety profile in the domestic clinical trials was 0.8% (1/132 patients). However, only three patients were subjected to safety analysis in the domestic clinical study. All three patients had subjective symptoms, including pain in 100% (3/3), hypertension in 66.7% (2/3), fever in 33% (1/3), and decreased appetite in 33.3% (1/3). Abnormal laboratory values were also observed in all three patients, mainly increased CRP and lactate dehydrogenase. The guidelines of CIRSE, the European SIR [25] , recommend the following follow-up: Most complications after UAE are observed within the first week after treatment, and postembolization syndromes such as pain can be treated with analgesics and sedatives. Infection is a rare but fatal complication, and a lack of communication between the physician and patient can lead to delayed treatment and fatal sepsis [98] ; thus, a system that allows for urgent consultation is required. The treatment effect is maximal and stable after 6 months; therefore, follow-up at that time is recommended. Patients should also be informed that myoma symptoms may not change for several months after treatment, and that patients who underwent UAE can return to normal activities about 2 weeks after treatment [115 , 116] . Follow-up after UAE often includes blood tests and radiological studies. Radiological studies mainly include ultrasound and MRI to evaluate changes in myoma size. Contrast-enhanced MRI can evaluate the percentage of contrast-enhanced area of the myoma in addition to the size change of the myoma, and re-treatment is often unnecessary, particularly when the contrast-enhanced area of the myoma is reduced by 90% to 100% [117] . Other reports have evaluated the results by hysteroscopy [118 , 119] . There are various reports on the timing of radiological studies, ranging from 48 to 72 hours after UAE [117] to 3 months to 1 year [119 - 124] ; the CIRSE guidelines [25] state that the maximum and stable effect is achieved after 6 months of UAE, and evaluation around that time may be appropriate. However, because the re-treatment rate of UAE is higher than that of surgery [125] (see CQ1 for details), longer follow-up and informed consent regarding the possibility of symptom relapse are necessary.

UAE has become a treatment option for symptomatic uterine myomas. To perform UAE procedures effectively and safely, radiologists should be aware of the basic techniques and the complications and limitations of each UAE procedure.

Uterine artery embolization (UAE) is an interventional radiology (IR) procedure to embolize the uterine artery or other vessels via the transcatheter in patients with gynecological disorders. UAE was originally performed for obstetric hemorrhage. However, because Ravina [1] reported it as a treatment for uterine myomas in France in 1995, it has been widely performed worldwide as a less invasive alternative to total hysterectomy, primarily as a treatment for symptomatic uterine myomas. In Japan, UAE has been performed since 1997 and was not approved by insurance for a long time. In 2013, the embolization material trisacryl gelatin microsphere (Embosphere) was covered by insurance for the treatment of hypervascular tumors or arteriovenous (AV) malformations, making UAE possible under certain conditions (Please refer to the “ Requirements for Systems etc. on the Proper Use of Embosphere ” from the website of the Japanese Society of Interventional Radiology (JSIR): https://www.jsir.or.jp/ ). The number of UAE procedures has been increasing since insurance approval in Japan, as indicated by the JSIR web registry until 2018. In this article, we summarize clinical questions (CQs) regarding the evidence on the effectiveness of UAE for symptomatic uterine myomas to perform appropriate UAE. The target readers of this guideline are physicians, nurses, and radiologists who perform UAE for symptomatic uterine myomas. The target patients include those scheduled for UAE for symptomatic uterine myomas and those who have undergone UAE. The certainty of evidence and classification of the strength of the recommendation were determined using the Grading of Recommendation, Assessment, Development, and Evaluation (GRADE) system ( Table 1 and 2 ) [2 - 15] . Recommendations were classified into five types as presented in Table 3 . Furthermore, when a recommendation is labeled as “ weak, ” it can be mistaken for evidential weakness. Therefore, this term was replaced with “ with conditions, ” “ optimal, ” and “ limited. ” In this guideline, we use the term “ with conditions, ” or in other words, “ recommendation subject to the setting, available resources, and patient's sense of values ” [2 - 15] . The guideline is based on its Japanese version, which was published in 2022 on the JSIR website [16] , and the publication of the current English version has been approved by JSIR. Graded Scale of Outcome Importance. Classification of Evidence Certainty. Recommendation Type.

None