#424 Wortman FINAL

Gynecology SURGICAL TECHNOLOGY INTERNATIONAL XXII

Ultrasound-guided Reoperative : Managing Endometrial Ablation Failures

MORRIS WORTMAN, MD, FACOG CLINICAL ASSOCIATE PROFESSOR OF GYNECOLOGY UNIVERSITY OF ROCHESTER MEDICAL CENTER DIRECTOR, CENTER FOR MENSTRUAL DISORDERS AND REPRODUCTIVE CHOICE ROCHESTER, NEW YORK

ABSTRACT ndometrial ablation and hysteroscopic myomectomy and polypectomy are having an increasing impact

on the care of women with abnormal uterine bleeding (AUB). The complications of these procedures Einclude the late onset of recurrent vaginal bleeding, cyclic lower abdominal pain, and the inability to adequately sample the in women with postmenopausal bleeding. According to the 2007 ACOG Practice Bulletin, approximately 24% of women treated with endometrial ablation will undergo within 4 years.1 By employing careful cervical dilation, a wide variety of gynecologic resectoscopes, and continuous sonographic guidance it is possible to explore the entire in order to locate areas of sequestered endometrium, , and occult hematometra. Sonographically guided reoperative hysteroscopy offers a minimally invasive technique to avoid hysterectomy in over 60% to 88% of women who experience endometrial ablation failures.2,3 The procedure is adaptable to an office-based setting and offers a very low incidence of operative complications and morbidity. In addition, the technique provides a histologic specimen, which is essential in adequately evaluating the endometrium in postmenopausal women or women at high risk for the development of adenocarcinoma of the endometrium.

- 1 - #424 Wortman FINAL

Ultrasound-guided Reoperative Hysteroscopy: Managing Endometrial Ablation Failures WORTMAN

INTRODUCTION It is well known that of women who Troublesome vaginal bleeding, may undergo EA a significant number will occur months or years following EA and eventually require a hysterectomy. has been attributed to inadequate Endometrial ablation [EA] has been Longinotti et al. analyzed the long-term endometrial destruction,11,12 unsuspect- an important addition to the gynecolog- results of 3,681 women undergoing ed deep adenomyosis,13 or the occur- ic armamentarium for the treatment of endometrial ablation at 30 Kaiser Per- rence of new pathology — myomata, abnormal uterine bleeding. The first manente Northern California facilities endometrial hyperplasia, or cancer. techniques utilized an operative hys- and noted 26% required hysterectomy Pelvic pain is generally cyclic and has teroscope in combination with either a during the 8-year follow-up period.9 been attributed to cornual and central neodymnium:YAG laser4 or a surgical The ACOG Practice Bulletin, May 2007 hematometra10 as well as the post-abla- electrode.5,6 Since 1995, five non-resec- edition, states that hysterectomy rates tion tubal sterilization syndrome toscopic endometrial ablation (NREA) within 4 years following both resecto- (PATSS).14 McCausland et al. note that devices have been introduced and scopic and NREA are at least 24%.1 the etiology of cyclic pelvic pain “fol- extensively employed to manage AUB in The late complications of EA are lowing both resectoscopic and nonresec- women who have completed their three-fold: persistent or recurrent vagi- toscopic endometrial ablations is due to childbearing.7 Oftentimes, EA is com- nal bleeding,8 cyclic pelvic pain,9 and the intrauterine scarring and contracture bined with hysteroscopic polypectomy the inability to adequately assess the that can occur following the or myomectomy to optimize the man- endometrium in women who later procedure.”10 Hopkins et al. performed agement of AUB.8 require sampling.10 hysterosalpingograms on 21 women at 3-, 6-, and 9-month intervals following radiofrequency global endometrial abla- tion and observed that intrauterine synechiae actually increased throughout the observation period suggesting that the continues to undergo remod- eling long after the original procedure.15 The inability to adequately assess the uterine cavity is an important and under-reported delayed complication following EA. Ahonkallio et al. demon- strated that endometrial biopsies failed Group 1. Recurrent Vaginal Bleeding in 23% of women with previous EA and were likely unreliable in many of the remaining patients given that endometrium is often trapped in the cornual region, which is frequently inaccessible. Any of these delayed com- plications are sufficient reasons for per- forming .16 Several authors, however, have reported “repeat” or “reoperative” endometrial ablation procedures. In 1992, Gimpelson reported a series of 16 women who underwent repeat EA uti- Group 2. Cyclic Pelvic Pain lizing either a hysteroscopic Nd:YAG or an electrosurgical technique; all were able to successfully avoid hysterectomy during the study period.17 In a series of 118 women who were offered reopera- tive hysteroscopic surgery (RHS), Istre et al.18 were successful in avoiding hys- terectomy in 72% of the subjects during a mean follow-up period of 22 months. In 2001, the author reported a series of 26 women who had undergone sono- graphically guided RHS following EA failures and noted that hysterectomy was avoided in 88.5% during a mean follow- Group 3. Endometrial sampling because 3 of abnormal perimenopausal or postmenopausal bleeding. up period of 23.2 + 22.7 months. The author believes that the use of a sonographically guided hysteroscopic Figure 1. Three groups of women requiring reoperative hysteroscopic surgery (RHS). resection technique provides two distinct

- 2 - #424 Wortman FINAL

Gynecology SURGICAL TECHNOLOGY INTERNATIONAL XXII

Figure 2. Transvaginal ultrasound demonstrating hematometra. Figure 3. Transvaginal ultrasound demonstrating endometrial regrowth. advantages over other methods of RHS. group of women who develop asympto- diagnosis is often delayed. In general, the First, sonographic guidance is an excel- matic hematometra as an incidental find- pain will resolve spontaneously within a lent tool for locating areas of hematome- ing on ultrasound or MRI generally few days only to recur the next cycle. tra, endometrial regrowth, and performed for another reason. This lat- A and transvaginal leiomyomas in a setting where standard ter group is usually amenorrheic but may ultrasound (TVUS) are best performed intrauterine landmarks are often absent. require endometrial sampling depending when the patient is symptomatic. The Second, hysteroscopic resection tech- on their age and risk factors for develop- former often reveals uterine tenderness niques provide ample tissue for histolog- ing endometrial cancer. without cervical motion or adnexal ten- ic analysis, an important requirement in derness. The latter often reveals the the management of women with a histo- Preoperative Evaluation presence of one or more hematometrae ry of abnormal peri-menopausal or post- The most important preoperative (Fig. 2). In some instances well-circum- menopausal bleeding and a prior EA. tool is a good history. Often the history scribed areas of endometrium may be This paper will summarize the author’s is straightforward; a woman with a pre- found in the cornua (Fig. 3) or scattered technique, which has evolved over the vious EA presents with gradually along the central uterine axis. Addition- past two decades and is now performed increasing vaginal bleeding accompanied ally, ultrasound may be useful in the in an office-based setting. by few if any cramps. In other instances, diagnosis of PATSS. Advanced studies the patient may present with severe such as sonohysterography or hysteros- Indications for Reoperative lower abdominal pain accompanied by alpingography are often painful and add Hysteroscopic Surgery (RHS) little or no vaginal bleeding. The pain little to the evaluation. The indications for RHS fall into the may be suprapubic or localized to one In determining whether or not pelvic three groups (Fig. 1) already mentioned: of the lower quadrants; it is often pain is related to a previous EA the use of those who experience recurrent vaginal described as suprapubic “sharp,” “stab- endometrial suppressive agents — oral bleeding, (Group 1) cyclic pelvic pain, bing,” “cramping,” or even “labor-like” contraceptives, leuprolide acetate, dana- (Group 2) or require endometrial sam- in quality. On occasion, the pain may be zol, and medroxyprogesterone — may pling because of abnormal peri- localized to the lower back or in one of be helpful in establishing the diagnosis. menopausal or postmenopausal bleeding the lower quadrants. When the pain is The improvement of pelvic pain with (Group 1). In addition, there is a fourth unaccompanied by vaginal bleeding the endometrial suppression, however, is not specific to the diagnosis of hematometra Table I or endometrial regrowth. However, the failure of symptoms to resolve with Candidacy for Reoperative Hysteroscopy: endometrial suppression strongly sug- Factors to Consider gests another etiology.

A symptom-free interval following the initial EA of at least 1 year Who are good candidates for Women who are 45 years of age or older RHS? Once the diagnosis has been estab- Uterine dimensions lished several factors must be weighed < 12 cms long, < 6 cms AP, and < 7 cms transverse in determining whether or not the Absence of multiple intramural leiomyomas patient is an appropriate candidate for RHS. Although there is little evidence- Absence of severe adenomyosis based data at this time the author Highly motivated women believes that appropriate candidates for Subjects are well informed of risks, consequences, and alternatives RHS include women who meet the cri- teria listed in Table I.

- 3 - #424 Wortman FINAL

Ultrasound-guided Reoperative Hysteroscopy: Managing Endometrial Ablation Failures WORTMAN

Figure 4. Laminaria japonica placed under ultrasound guidance. Figure 5. Operating room setup.

Patient Preparation Equipment, Setup, and Personnel Whenever bipolar electrosurgery is Once the decision has been made to for RHS used, a saline-C-cut, effect 5, setting is perform RHS, it is best to schedule The equipment and personnel uti- employed. During unipolar electro- surgery in the presence of symptoms or lized for RHS are identical to what the surgery, glycine 1.5% is used for disten- clear ultrasound findings. One should author has reported for sonographically tion; normal saline is used during refrain from using endometrial suppres- guided hysteroscopic endomyometrial bipolar electrosurgery. sive agents during the cycle prior to RHS. resection.19 Low-viscosity fluids are Most cases are performed with We insist that patients undergoing delivered through one of several fluid- either a 22 Fr or a 26 Fr continuous RHS undergo cervical dilation and lami- management systems at initial pump flow resectoscope (CFR), the former naria placement the day prior to settings varying from 100 to 180 mm being preferred during earlier portions surgery. This is done under minimal to Hg. Fluids are allowed to egress by of the case. A variety of ultrasound moderate sedation. Patients are fore- gravity alone during the resection phase machines have been used over the past warned that the evening prior to RHS of the procedure. Active suction may be decades. We presently use a Siemens can be quite uncomfortable as the lami- supplied during the coagulation phase in Acuson X150 (Siemans Corp., New naria often expands against a fibrotic order to remove as many water vapor York, NY) equipped with a variable fre- lower uterine segment. bubbles as possible. quency abdominal transducer in order Laminaria placement is preceded by Our operating room (Fig. 5) is to provide continuous sonographic careful dilation under sonographic equipped with an Autocon II 400 (Karl monitoring. Our operating room is out- guidance with Hegar dilators. General- Storz Endoscopy, Culver City, CA) fitted with two side-by-side monitors ly, a 3- or 4-mm laminaria is sufficient enabling us to utilize both unipolar and that facilitate real-time observation for this purpose and should be passed bipolar electrosurgery. Unipolar elec- (Fig. 6). All procedures are digitally well beyond the internal cervical os trosurgery is generally performed at recorded using a MediCapture USB 200 (Fig. 4). Patients are given prescrip- 140 watts of C-Cut, effect 4, during the (MediCapture, Inc., Philadelphia, PA). tions for NSAIDs or opiates and asked resection phase and 120 watts of forced Most cases are presently performed to not eat or drink anything prior to coagulation current, effect 4, for the in an office setting and require a mini- their procedure. ablation portion of the procedure. mum of 4 assistants. The first assistant

Figure 6. Operating room with monitors side by side. Figure 7. New Igor drawing of ultrasound-guided reoperative hysteroscopic surgery.

- 4 - #424 Wortman FINAL

Gynecology SURGICAL TECHNOLOGY INTERNATIONAL XXII stands to the operator’s left while the sonographer stands to the right. A fluid management technician is responsible for all functions related to the hys- teroscopy pump and reports the rate of fluid absorption as well as the net fluid deficit. The fourth member of the team is an appropriately trained and creden- tialed registered nurse who administers midazolam and fentanyl while monitor- ing the patient for her level of conscious- ness.

PROCEDUREPROCEDURE

The procedure is begun by removing the previously placed laminaria and prepping the and with a bactericidal solution. The cervix is Figure 8. Two-tenaculae technique. grasped at 12 o’clock with a tenaculum. Next, a vasopressin solution containing almost entirely with ultrasound guid- highly variable, and the particular path 2.5 units in 20 mL saline is injected ance (Fig. 7). In most cases, the removal of dissection is determined by both the deep into the cervical stroma at 3 and 9 of this tissue strip provides sufficient intrauterine and sonographic findings, o’clock using a 21-gauge x 1½-inch room within the uterine cavity so that the two complementing each other. needle. The cervix is carefully dilated continuous flow and visualization are Initial inspection of the cavity may under sonographic guidance to either 8 facilitated. Once this central cavity is reveal active endometrial elements. In mm (for a 22 Fr CFR) or 10 mm (for a established the resection margin is some cases, islands of endometrium are 26 Fr CFR). Glycine 1.5% is delivered widened in all quadrants with care to hemosiderin-stained, especially if the at a pump infusion pressure of 120 to leave at least 5 mm to 10 mm of tissue area has been contained within a 180 mm Hg. We carefully monitor fluid from the central cavity to the uterine sequestered hematometra. The extent absorption and adjust the infusion pres- serosa at any given point. It is important of intrauterine adhesions (IUA) can be sure accordingly. In the presence of a to keep the uterine cavity well distend- quite variable. In some instances there small or tubular cavity the initial resec- ed allowing an adequate “sonohystero- are few, if any, IUA — this is especially tion is carried out on the thickest uter- gram” during the procedure. This is true when RHS is offered to women ine wall, which is often the posterior or done by maintaining sufficient pump who have received little relief of their anterior wall. Even in the presence of pressure and minimizing extravasation menstrual symptoms and in whom new- poor hysteroscopic visualization this of fluid from the cervix; the latter can onset cyclic pelvic pain is not a major first critical strip is removed by extend- be accomplished by placing additional presenting problem. In other clinical ing the loop approximately 7 mm and tenaculae at the 3 or 9 o’clock positions scenarios, initial inspection reveals vir- removing a continuous ribbon of tissue (Fig. 8). Unlike endomyometrial resec- tually no endometrial elements; this is from the upper reaches of the cavity to tion (EMR),20 there is no predeter- especially true if there is a sequestered the internal os, a maneuver performed mined order of resection. Each case is area of endometrium or a hematometra.

Figure 9. Igor drawing of centrally located hematometra. Figure 10. Reconfigured loop electrode.

- 5 - #424 Wortman FINAL

Ultrasound-guided Reoperative Hysteroscopy: Managing Endometrial Ablation Failures WORTMAN

a b Figures 11a and 11b. Dissection of cornual hematometra.

Oftentimes, ultrasound guidance is able fied and excised, the freshly exposed 26%, the majority of which were done to determine the relationship between is then deeply coagulated within 3 years of the original proce- the resectoscope loop and the with a ball-electrode, usually at 120 dure.9 The failure rate appears to be hematometra. In this scenario it is watts of coagulation current. The final unrelated to the type of procedure or important that only thin strips of tissue result generally reveals a cavity, which is the existence of leiomyomas. Age at the be removed (1 to 3 mm in thickness). quite larger than originally noted (Fig. time of the original procedure appears The management of centrally located 12); a transabdominal scan confirms the to play an important role in predicting hematometra (Fig. 9) is fairly straight- extent of dissection and the disappear- hysterectomy risk. Women under the forward, and RHS can be managed ance of any preoperative evidence of a age of 40 have a 40% probability of much like a primary EMR.20 In most hematometra. hysterectomy during the 8-year follow- cases of RHS the tubal ostia are not up period, while the risk drops to 20% clearly identified. Should one fail to for women in the 45 to 50 year age identify the tubal ostia it is important to CONCLUSIONSCONCLUSIONS cohort. The results were similar for all carefully explore the cornua. This can types of EA procedures and indepen- be done with a reconfigured loop elec- dent of leiomyomata. trode (Fig. 10) with care to perform Although the precise number of all Longinotti et al. also note that the both blunt and electrosurgical dissection types of endometrial ablation and most common indications were vaginal under scrupulous sonographic guidance. resection performed in the United bleeding (51.6%), pain (22%), and Cautious dissection often allows the States is difficult to uncover, one esti- vaginal bleeding with pain (20.3%).9 operator entry into a cornual mate suggests that in 2008 alone, Although there are relatively few series hematometra — often signaled by the approximately 312,000 global endome- of RHS for EA failures the available data egress of chocolaty material — beyond trial ablations were performed in this suggest that between 60% to 88.5% of which active endometrial elements are country.21 Longinotti et al. noted that hysterectomies can be avoided in clearly seen (Figs. 11a & 11b). Once the in a cohort of 3,681 women with a his- women followed up to 84 months.2,3,18 surgeon is confident that all remaining tory of EA the probability of subse- The author’s experience suggests areas of endometrium have been identi- quent hysterectomy at 8 years was that the incorporation of careful patient

a b Figure 12ab Cavity before RHS. Figure 12b. Cavity after RHS

- 6 - #424 Wortman FINAL

Gynecology SURGICAL TECHNOLOGY INTERNATIONAL XXII selection, meticulous cervical prepara- 2008;15:704–6. Obstet Gynecol 1997;104:934–8. tion, and sonographic guidance are the 3. Wortman M, Daggett A. Reoperative hys- 12. Lisa JR, Gio JD, Rubin IC. Observations keys to performing safe reoperative hys- teroscopic surgery in the management of on the interstitial portion of the fallopian patients who fail endometrial ablation and tube. Surg Gynecol Obstet 1954;92:159–69. teroscopy with a high degree of patient resection. J Am Assoc Gynecol Laparosc 13. McCausland AM, McCausland VM. satisfaction and success. The assembly of 2001;8:272–7. Depth of endometrial penetration in adeno- a dedicated and stable operating room 4. Loffer FD. Hysteroscopic endometrial myosis helps determine outcome of rollerball team is a key component for this proce- ablation with the Nd:YAG laser using a non- ablation. Am J Obstet Gynecol 1996;174: dure’s success. Future work is needed touch technique. Obstet Gynecol 1987; 1786–93. to better delineate those women who 679–82. 14. Townsend DE, McCausland VM, McCaus- 5. Magos AL, Baumann R, Lockwood GM, et land AM. Post-ablation-tubal sterilization syn- are the best candidates for reoperative al. Experience with the first 250 endometrial drome. Obstet Gynecol 1993;82: 422–4. hysteroscopic surgery. STI resections for menorrhagia. Lancet 1991; 15. Hopkins MR, Creedon DJ. Radiofre- 337:1074–8. quency global endometrial ablation followed 6. Vancaillie TG. Electrocoagulation of the by hysteroscopic sterilization. J Minim Inva- AUTHOR’SAUTHOR’S DISCLOSURES DISCLOSURES endometrium with the ball-end resectoscope. sive Gynecol 2007;14:494–501. Obstet Gynecol 1989;74:425–7. 16. Ahonkallio SJ, Liakka AK, Martikainen 7. Zarek S, Sharp H. Global endometrial HK, et al. Feasibility of endometrial assess- ablation devices. Clin Obstet Gynecol ment after thermal ablation. Eur J Obstet The author is a consultant for Holog- 2008;51: 167–75. Gynecol Reprod Biol 2009;147:69–71. ic, Inc. However, they will not be fund- 8. Loffer FD. Improving results of hystero- 17. Gimpelson RJ, Kaigh J. Endometrial abla- ing any of this work. There are no scopic submucosal myomectomy for menor- tion repeat procedures case studies. J Repro pertinent conflicts of interest. rhagia by concomitant endometrial ablation. J Med 1992;37:629–34. Minim Invasive Gynecol 2005;3:254–60. 18. Istre O, Langebrekke A. Repeat hystero- 9. Longinotti MK, Jacobson G, Hung Y, et scopic surgery reduces the hysterectomy rate al. Probability of hysterectomy after endome- after endometrial and myoma resection. J Am REFERENCESREFERENCES trial ablation. Obstet Gynecol 2008;112: Assoc Gynecol Laparosc 2003;10:247–51. 1214–20. 19. Wortman M. Sonographically guided hys- 10. McCausland AM, McCausland VM. teroscopic endomyometrial resection. Surg 1. Munro MG. ACOG Practice Bulletin: Long-term complications of minimally inva- Tech Int 2011;XXI:163–9. endometrial ablation. Obstet Gynecol sive endometrial ablation devices. J Gynecol 20. Wortman M, Daggett A. Hysteroscopic 2007;109:1233–47. Surg 2010;26:133–49. endomyometrial resection. JSLS 2000;4: 2. Hansen BB, Dreisler E, Sorensen SS. Out- 11. Turnbull LW, Jumaa A, Bowsley SJ, et 197–207. come of repeated hysteroscopic resection of al. A magnetic resonance imaging of the 21. http://investors.hologic.com/ index.php? the endometrium. J Minim Invasive Gynecol uterus after endometrial resection. British J s=43&item=420. Queried August 5, 2012.

- 7 -