{"paper_id":"1067d2ea-4dc3-42f7-b4b9-5c2d6ab1c2b0","body_text":"Gynecol Surg (2006) 3: 265 –269\nDOI 10.1007/s10397-006-0219-8\nORIGINAL ARTICLE\nPentti K. Heinonen . Riikka Helin . Kari Nieminen\nLong-term impact and risk factors for hysterectomy\nafter hysteroscopic surgery for menorrhagia\nReceived: 21 March 2006 / Accepted: 15 June 2006 / Published online: 30 September 2006\n# Springer-V erlag Berlin / Heidelberg 2006\nAbstract The objective of this study was to assess the\nlong-term impact of management and establish the inci-\ndence of hysterectomy, and to identify factors predictive of\nfailure of the procedure among women who had undergone\nhysteroscopic endometrial resection with or without myo-\nmectomy for menorrhagia. Clinical history and data on\nadditional treatment and follow-up status were obtained by\nmedical record review and postal questionnaire for 279\nwomen who had undergone hysteroscopic surgery. Follow-\nup data were available for 259 (93%) cases, and the mean\nfollow-up was 6.0 years. Subsequent hysterectomy was the\nprimary endpoint, and its incidence was calculated by\nsurvival analysis. Univariant analysis and Cox regression\nmodel were used to identify predictors of failure. Myomas,\npolyps, adenomyosis, or endometrial hyperplasia were\nfound in 40.9% of hysteroscopic procedures. Perioperative\ncomplications occurred in 5.7% and late complications in\n7.7%. During the follow-up period, 97 (37.5%) of 259\nwomen underwent at least one gynecological procedure.\nThe incidence of hysterectomy was 23.6% (95% con-\nfidence interval: 18.8 –29.1%). Positive predictive factors\nfor hysterectomy were long uterine cavity ( ≥9 cm) and\ntubal ligation. Most (82.8%) of the 198 women who did not\nundergo hysterectomy had postoperative oligo- or amen-\norrhea. Hormone replacement therapy was common (67%)\namong postmenopausal women after endometrial resec-\ntion. Hysteroscopic resection of the endometrium and\nconcomitant hysteroscopic resection of fibroids for the\ntreatment of menorrhagia is a suitable alternative to\nhysterectomy and offers lasting results. A large uterine\ncavity indicating possible uterine pathology and tubal\nligation associated with hematometra increase the risk of\nhysterectomy.\nKeywords Endometrial resection . Myomectomy .\nHysterectomy . Menorrhagia\nIntroduction\nHysteroscopic surgery, including endometrial resection,\nmyomectomy, and polypectomy, is an accepted alternative\nto hysterectomy in the treatment of menorrhagia [ 1–3].\nSeveral transcervical endometrial ablation methods have\nbeen used for the purpose. The first-generation hystero-\nscopic procedure is performed using electrical energy for\nendometrial resection with loop or endometrial coagulation\nwith rollerball or the endometrium is destroyed with laser\nenergy [ 1, 2]. The second-generation methods are mostly\nnon-hysteroscopic techniques using different sources of\nenergy to destroy the endometrium [ 1, 2]. Hysteroscopic\nresection of the endometrium is one of the oldest, often\ncombined with myomectomy or polypectomy, which is a\ngreat advantage of this technique [ 1, 2, 4]. Operative\noutcomes, perioperative complications, and short-term\nfollow-up results are well documented [ 2–4]. However,\nlong-term results of endometrial resection are sometimes\nless striking than those in the immediate postoperative\nperiod. The long-term impact and factors prognostic for\nsubsequent hysterectomy are not well known. In addition,\nhormonal replacement therapy in women who have\nreached menopause after endometrial ablation has not\nbeen extensively studied.\nOur aim was to evaluate long-term outcomes and\nhysterectomy rates after endometrial resection for menor-\nrhagia and to identify factors predictive of hysterectomy in\nthese patients.\nPatients and methods\nBetween November 1990 and May 1999, 279 consecutive\nwomen underwent hysteroscopic endometrial resection\nwith or without concomitant myomectomy at the Depart-\nment of Obstetrics and Gynecology in the University\nP . K. Heinonen ( *) . R. Helin . K. Nieminen\nDepartment of Obstetrics and Gynecology,\nTampere University Hospital, and Medical School,\nUniversity of Tampere,\nTampere, 33014, Finland\ne-mail: pentti.heinonen@uta.fi\nTel.: +358-3-31164650\nFax: +358-3-35516164\n\nHospital of Tampere, Finland. The mean age of the patients\nwas 42 years (range: 15 –68 years); 171 (61.3 %) were\nbetween 40 and 49 years, and 18 (6.5%) women were\npostmenopausal and menorrhagia was linked to hormone\nreplacement therapy. Mean body mass index (BMI) (SD)\nwas 25.9 (4.8) kg/m\n2. Mean parity was 2.0 (range: 0–5) and\n29 (10.4%) had undergone cesarean section. Tubal ligation\nfor sterilization had been performed in 141 (50.5%) cases.\nAll patients suffered from menorrhagia and were\ncandidates for hysterectomy. One hundred and eighty\n(64.5%) women had previously received medical therapy\nfor menorrhagia, most frequently oral cyclical or contin-\nuous progestogen (34.1%), tranexamic acid (28.0%), or a\nlevonorgestrel-releasing intrauterine system (17.9%).\nFifty-two (18.6%) patients were treated hysteroscopically\nfor a medical disorder constituting a significant risk at\nhysterectomy. Seventeen had heart disease, 8 were\nreceiving anticoagulant therapy, 12 were mentally retarded,\n8 had a neurological disease, 7 a nongynecological\nmalignancy, and 8 kidney or other disease. Five women\nhad severe obesity (BMI 40 –53 kg/m\n2).\nCervical smear and endometrial biopsy were taken\npreoperatively. Transvaginal ultrasonography was em-\nployed to determine the size and location of possible\nmyomas or endometrial polyps. Women with a uterus\nlarger than the size of a 12-week pregnancy were excluded\nfrom the study, but isolated submucous fibroids less than\n5 cm in diameter were not an exclusion criterion. Women\nwith endometrial hyperplasia, uterovaginal prolapse, un-\ntreated adnexal disease, or acute pelvic inflammatory\ndisease were excluded from resection. Women desiring\nfuture pregnancy were also excluded.\nPreoperative endometrial suppression with hormonal\ntherapy was used in 100 cases (35.8%). The treatment\nconsisted of danazol, gonadotropin-releasing hormone\n(GnRH) agonist (goserelin), or continuous progestogen\ntreatment with a levonorgestrel-releasing intrauterine\nsystem or continuous oral progestogen therapy (lynesterol).\nIn other cases surgery was performed on days 4 –9 of the\nmenstrual cycle.\nThe procedure was carried out under spinal anesthesia in\n136 cases (48.7%). The uterine cavity was distended with\n2.2% isotonic glycine (Baxter Health Care Ltd., Thetford,\nUK) and the endometrium was resected using a 26 French\nresectoscope as previously described [ 5]. Antibiotic pro-\nphylaxis was given when appropriate. Sixty-eight women\n(24.4%) had concomitant hysteroscopic myomectomy and\n42 (15.1%) laparoscopic tubal ligation.\nThe length of the uterine cavity was measured prior to\nthe hysteroscopic procedure using uterine sounding. The\nmean (SD) length of the uterine cavity was 8.7 (1.0) cm and\nranged from 6.5 cm to 13.0 cm. A cavity 9 cm or more in\nlength was regarded as large.\nHospital records of the patients were examined for\ndetails of preoperative history, operative procedures,\nfurther surgical treatment, and documented follow-up.\nPatients who had not undergone hysterectomy during the\nfollow-up period were sent a postal questionnaire to assess\nlong-term effects. Questions concerned the need for further\ntreatment, hormonal treatment, bleeding patterns, and\nsuccess of treatment. Respondents were asked whether\nthey had amenorrhea, slight menstruation for 1 –3 days,\nslight menstruation for 4 –6 days, or no improvement in\nmenstrual flow or increased menstrual flow.\nFollow-up data were received from 259 patients\n(92.8%). Five had died and 15 were lost to follow-up.\nThe mean follow-up period was 6.0 years (range: 0.3 –\n11.0 years).\nStatistical analysis was made using the SPSS for\nWindows, version SPSS 11.5 (SPSS Inc., Chicago, IL,\nUSA). Subsequent hysterectomy was the primary endpoint\nand its incidence was calculated by survival analysis\n(Kaplan-Meier). The association between hysterectomy\nand possible factors was tested by univariant analysis. A\nmultivariant analysis was then made using the Cox\nregression model with backward stepwise method to\nidentify the subset of variables most accurately predictive\nof the risk of hysterectomy. The level of significance was\nset at p<0.05.\nResults\nOperative complications were recorded in 16 (5.7%) of 279\noperations. Uterine perforation occurred in two women\n(0.7%). The first had cervical perforation during the\ncervical dilatation at commencement of surgery and\nendometrial resection was not undertaken. This patient\nhad previously received intracavitary radiation therapy for\nmenorrhagia. The second sustained uterine perforation and\ndisruption of the uterine artery in connection with\nmyomectomy, requiring emergency hysterectomy. Excess\nbleeding was controlled by tamponade using a Foley\ncatheter placed in the endometrial cavity in six cases\n(2.2%). Eight women (2.9%) with postoperative endome-\ntritis recovered completely after antibiotic treatment.\nGlycine deficit during the operation was over 1 l (range:\n1.0–1.8 l) in only eight patients (2.9%), but none of these\ndeveloped hyponatremia. Of 136 women with spinal\nanesthesia, 13 (9.6%) suffered from postspinal headache,\nwhich was successfully treated with a blood patch.\nAt least one abnormal finding in endometrial specimens\nwas reported in 114 (40.9%) cases. Concomitant hystero-\nscopic resection of fibroids was performed in 68 (24.4%)\nand polypectomy in 25 (9.0%). The mean size of the\nfibroids was 2.0 cm (range: 0.5–5.0 cm). Adenomyosis was\ndetected in 21 specimens (7.5%). Despite earlier endome-\ntrial biopsy seven (2.7%) women were found to have\nendometrial hyperplasia without atypia.\nLate complications occurred in 20 (7.7%) of 259\nwomen with follow-up. Hematometra developed in 18\n(6.9%) cases and 14 of these had undergone tubal ligation.\nTwo pregnancies (0.8%) occurred after endometrial\nresection, one having induced abortion during the first\ntrimester 6 months postoperatively and the other a\nspontaneous miscarriage with placenta accreta with\nsubsequent hysterectomy.\n266\n\nDuring the follow-up, 97 (37.5%) of 259 women\nunderwent at least one gynecological procedure\n(Table 1). Endometrial ablation was repeated because of\nhematometra, spotting bleeding, or request for amenorrhea.\nHysterectomy was performed in 61 cases [23.6%, 95%\nconfidence interval (CI): 18.8 –29.1%]. The main indica-\ntions for subsequent hysterectomy applied to 21 cases with\nmyomas, 13 adenomyosis, 11 persistent menorrhagia or\nendometrial hyperplasia, 7 peri- or postoperative compli-\ncations (uterine perforation, pregnancy, hematometra), 8\nendometriosis or genital prolapse, and 1 case with\nunknown reason. Malignancy was not found in any case.\nSurvival analysis shows a relationship between the prob-\nability of not undergoing subsequent hysterectomy and the\ntime since the endometrial ablation procedure (Fig. 1).\nMost (83.6%) of the hysterectomies were performed during\nthe first 5 years after the hysteroscopic surgery.\nTable 2 shows the results of univariant analysis and the\nCox regression model.\nA history of tubal ligation and length of uterine cavity\n9 cm or more were associated with an increased risk of\nhysterectomy in both univariant and multivariant analysis.\nPerioperative complication was a significant factor in\nunivariant but not in multivariant analysis (Table 2).\nLong-term follow-up data were available for 198 women\nwho did not require hysterectomy. One hundred women\n(50.5%) reported amenorrhea, 64 (32.3%) had slight\nbleeding for 1 –3 days, and 24 (12.1%) had slight bleeding\nfor 4 –6 days. Ten women (5.1%) estimated that their\nbleeding had become scantier than before the hysteroscopic\nprocedure.\nOf 198 women, 112 (56.6%) had reached menopause\nduring the follow-up period; 75 of them (67.0%) used\nhormonal replacement therapy, 30 used a continuous\ncombined estrogen and progestogen regimen or tibolone,\n38 a cyclical combined regimen, 4 reported using only\nestrogen, and 3 had only a cyclical progestogen regimen.\nOf 38 women on a cyclical combined estrogen and\nprogestogen regimen, 19 (52.6%) reported that they were\namenorrheic.\nDiscussion\nHysterectomy was the primary endpoint in this study, as a\nhysteroscopic procedure was undertaken mostly instead of\nhysterectomy in women suffering from menorrhagia. The\nrate of hysterectomy after endometrial resection was 23.6%\nin the present study and 15 –24% in previous studies with a\nfollow-up period of 4 years or more [ 6–8]. The indication\nfor uterine removal was in most cases uterine fibroids or\nadenomyosis. Our study also included cases undergoing\nhysterectomy with no relationship to endometrial resection,\nfor example endometriosis or uterine prolapse. It is evident\nthat a long follow-up period makes it possible to develop\nmany diseases of the uterus which are not related to the\ntreatment of menorrhagia. These patients therefore need\nfollow-up years after hysteroscopic treatment.\nHysterectomies were performed mostly during the first\n5 years after hysteroscopic surgery. Those women who had\namenorrhea or hypomenorrhea after hysteroscopic surgery\nalso maintained this response years after treatment. This\ntechnique did not merely postpone hysterectomy as three of\nfour women had a permanent result.\nBoth univariant and multivariant analysis showed that a\nlength of the uterine cavity of 9 cm or more was a\nsignificant risk factor for hysterectomy. A long uterine\ncavity may be associated with a large uterine cavity, when\nthe endometrial surface to resect is larger and liable to\nincomplete resection [ 2, 9]. In addition, the uterine cavity\nmay be insufficiently expanded preventing complete\nresection. Furthermore, menorrhagia and a long uterine\ncavity may be associated with other uterine pathology such\nas adenomyosis or uterine fibroids. Neis and Brandner [ 10]\nreported that women with dysmenorrhea and a uterine\ncavity over 10 cm show a high incidence of adenomyosis.\nThey run an increased risk of failure and should be\nexcluded from endometrial ablation [ 10]. Submucous\nFig. 1 Probability of not having a hysterectomy (Kaplan-Meier\ncurve). The upper figure line indicates the number of cases in\nfollow-up and the lower line the cumulative number of cases with\nhysterectomy\nTable 1 Further treatment after hysteroscopic management of\nmenorrhagia in 259 patients during the follow-up\nTreatment Number of patients (%)\nAny further treatment 97 (37.5)\nHysterectomy 61 (23.6)\nEndometrial reablation 20 (7.7)\nHysteroscopy, curettage 18 (6.9)\nCervical dilatation and drainage 9 (3.5)\nSalpingo-oophorectomy 6 (2.2)\nContinuous progestin 3 (1.2)\n267\n\nfibroids were the most common pathology needing\nadditional treatment with endometrial resection. Only this\nkind of myoma was possible to resect, intramural myomas\nbeing impossible to treat although they may enlarge the\nuterus [ 9].\nA significant number (13%) of women after endometrial\nresection have patent fallopian tubes [ 11]. Contraception is\nnecessary after endometrial resection, since pregnancy is\npossible, as two cases in the present study demonstrated.\nTubal ligation, which was also a risk factor for hysterec-\ntomy, was the most common method of contraception. An\nassociation between hematometra, endometrial resection,\nand tubal ligation has previously been reported [ 12]. Patent\nfallopian tubes may allow egress of blood and prevent the\nformation of hematometra. Most cases with painful\nhematometra can be treated by transcervical drainage.\nThe definitive treatment is resectoscopic diathermy to\nresidual areas of the endometrium [ 12]. However, this\ntreatment is not always successful and hysterectomy may\nbe necessary.\nAbout half of the women without hysterectomy had\nreached menopause during a long follow-up period, and\nmany of them (67%) were on hormone replacement\ntherapy (HRT). Combined treatment with estrogen and\nprogestogen was the most popular as it does not cause\nendometrial proliferation. Almost half of the postmeno-\npausal women on cyclic estrogen and progestogen regimen\nhad amenorrhea, showing good results after endometrial\nresection. Unopposed estrogen treatment is not recom-\nmended after endometrial ablation even where a woman\nhas amenorrhea after endometrial resection. Endometrial\ncells are found in endometrial samples in women with\namenorrhea after endometrial resection [ 13]. Istre and\nassociates reported endometrial hyperplasia in women on\nan unopposed estrogen regimen [ 14]. Those receiving a\ncontinuous combined estrogen and progestogen regimen\nhad no endometrial hyperplasia. Women who have under-\ngone endometrial resection must follow the same guide-\nlines during HRT for endometrial protection as women\nwith an intact uterus [ 14]. Postmenopausal women after\nendometrial resection and receiving HRT should undergo\nmeasurement of endometrial thickness using transvaginal\nultrasound at follow-up examinations. Those with post-\nmenopausal bleeding should be checked by endometrial\nbiopsy, as endometrial carcinoma has been reported after\nendometrial resection [ 15].\nNowadays endometrial resection has been replaced\nby second- and third-generation endometrial ablation\ntechniques [ 16]. However, submucous and pedunculated\nintracavitary fibroids and e ndometrial polyps are treated\nby the resection technique [ 2, 3]. Submucous myomas\nand polyps were the most common findings in this as in\nprevious studies [ 17]. The combination of resection of\nfibroids or polyps prior to endometrial ablative technol-\nogy may be one possible way to treat these patients.\nHormonal treatment with a levonorgestrel-releasing\nintrauterine system has been reported as an alternative to\nhysterectomy in women with heavy menstrual bleeding\n[18]. Studies comparing intrauterine hormonal treatment\nand endometrial resection have given similar results for the\n12-month follow-up period [19, 20]. In the present study 50\n(17.9%) women had used a levonorgestrel-releasing intra-\nuterine system prior to endometrial resection. Hystero-\nscopic surgery may also be an effective mode of therapy in\nwomen who have not been satisfied using a levonorgestrel-\nreleasing intrauterine system to treat menorrhagia instead\nof hysterectomy.\nReferences\n1. Oehler MK, Rees MCP (2003) Menorrhagia: an update. Acta\nObstet Gynecol Scand 82:405 –422\n2. Abbott JA, Garry R (2002) The surgical management of\nmenorrhagia. Hum Reprod Update 8:68 –78\n3. Lethaby A, Shepperd S, Cooke I, Farquhar C (2004) Endome-\ntrial resection and ablation versus hysterectomy for heavy\nmenstrual bleeding (Cochrane Review). In: The Cochrane\nLibrary, Issue 2. Wiley, Chichester, UK, pp 1 –25\n4. Overton C, Hargreaves J, Maresh M (1997) A national survey\nof the complications of endometrial destruction for menstrual\ndisorders: the MISTLETOE study. Br J Obstet Gynaecol\n104:1351–1359\n5. Tapper A-M, Heinonen PK (1995) Hysteroscopic endomyo-\nmetrial resection for the treatment of menorrhagia —follow-up\nof 86 cases. Eur J Obstet Gynecol Reprod Biol 62:75 –79\n6. Aberdeen Endometrial Ablation Trials Group (1999) A\nrandomised trial of endometrial ablation versus hysterectomy\nfor the treatment of dysfunctional bleeding: outcome at four\nyears. Br J Obstet Gynaecol 106:360 –366\n7. Cooper KG, Jack SA, Parkin DE, Grant AM (2001) Five-year\nfollow up of women randomised to medical management or\ntranscervical resection of the endometrium for heavy menstrual\nloss: clinical and quality of life outcomes. Br J Obstet Gynaecol\n108:1222–1228\n8. Boujida VH, Philipsen T, Pelle J, Joergensen JC (2002) Five-\nyear follow-up of endometrial ablation: endometrial coagula-\ntion versus endometrial resection. Obstet Gynecol 99:988 –992\n9. Shamonki MI, Ziegler WF, Badger GJ, Sites CK (2000)\nPrediction of endometrial ablation success according to peri-\noperative findings. Am J Obstet Gynecol 182:1005 –1007\nTable 2 Predictors of hysterectomy in patients with hysteroscopic surgery for menorrhagia\nPredictora Univariant analysis Cox regression model\nHazard rate ratio 95% CI p Hazard rate ratio 95% CI p\nTubal ligation 4.15 2.36 –7.27 0.020 1.954 1.048 –3.643 0.035\nPerioperative complication 2.06 1.10 –3.,88 0.049 2.088 0.854 –5.106 0.106\nLength of uterine cavity ≥9 cm 1.87 1.19 –2.94 0.005 1.993 1.187 –3.347 0.009\naThe following items were not statistically significant: age, medical disease, cesarean delivery, endometrial thinning, use of levonorgestrel-\nreleasing intrauterine system, late complication, abnormal histology, adenomyosis, myomectomy\n268\n\n10. Neis KJ, Brandner P (2000) Adenomyosis and endometrial\nablation. Gynaecol Endosc 9:141 –145\n11. Istre O, Daleng W, Forman A (1996) The incidence of fallopian\ntube patency after transcervical resection of the endometrium\nincluding rollerball diathermy to the tubal ostia. Fertil Steril\n65:198–200\n12. Gannon MJ, Johnson N, Watters JK, Lilford RJ (1997)\nHematometra—endometrial resection —sterilization syndrome.\nGynaecol Endosc 6:45 –46\n13. Magos AL, Baumann R, Lockwood GM, Turnbull AC (1991)\nExperience with the first 250 endometrial resections for\nmenorrhagia. Lancet 337:1074 –1078\n14. Istre O, Holm-Nielsen P , Bourne T, Forman A (1996) Hormone\nreplacement therapy after transcervical resection of the endo-\nmetrium. Obstet Gynecol 88:767 –770\n15. V alle RF, Baggish MS (1998) Endometrial carcinoma after\nendometrial ablation: high-risk factors predicting its occur-\nrence. Am J Obstet Gynecol 179:569 –572\n16. Roy KH, Mattox JH (2002) Advances in endometrial ablation.\nObstet Gynecol Surv 57:789 –802\n17. Nagele F, O ’Connor H, Davies A, Badawy A, Mohamed H,\nMagos A (1996) 2500 outpatient diagnostic hysteroscopies.\nObstet Gynecol 88:87 –92\n18. Marjoribanks J, Lethaby A, Farquhar C (2004) Surgery versus\nmedical therapy for heavy menstrual bleeding (Cochrane\nReview). In: The Cochrane Library, Issue 2. Wiley, Chichester,\nUK, pp 1 –31\n19. Crosignani PG, V ercellini P , Mosconi P , Oldani A, Cortesi I, De\nGiorgi O (1997) Levonorgestrel-releasing intrauterine device\nversus hysteroscopic endometrial resection in the treatment of\ndysfunctional uterine bleeding. Obstet Gynecol 90:257 –263\n20. Istre O, Trolle B (2001) Treatment of menorrhagia with the\nlevonorgestrel intrauterine system versus endometrial resection.\nFertil Steril 76:304 –309\n269","source_license":"CC0","license_restricted":false}