Morphological changes in hysterectomies after endometrial ablation

Francisco Tresserra1,3, Pablo Grases2, Alicia Ubeda2, Maria Angela Pascual2, Pedro J. Grases1 and Ramon Labastida2

1 Department of Pathology and 2 Department of Obstetrics and Gynecology, Instituto Universitario Dexeus, Paseo de la Bonanova 69, 08017 Barcelona, Spain


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Electrosurgical ablation of the endometrium is a therapeutic choice for those patients having abnormal uterine bleeding. When ablation is followed by a hysterectomy, tissue damage due to thermal effect can be seen. From a total of 350 women with endometrial ablation, 12 required subsequent hysterectomy. The histological features found in these specimens are described and related to the elapsed time between the two surgical procedures. The mean elapsed time between ablation and hysterectomy was 19 ± 17.3 months. Scarring with formation of additional endometrial cavities was seen in five cases, and endocervical stenosis in two cases. In seven patients, endometrial regrowth was seen at hysterectomy. Necrosis, granulomatous and foreign-body giant cell reaction, eosinophilic infiltrate and pigment-containing macrophages in the myometrium were seen in the long-term post-ablation hysterectomies. Necrosis was seen in short period post-ablation hysterectomies. Six of the seven patients with endometrial regeneration had adenomyosis in the hysterectomy specimen. Endometrial ablation induces thermal effects in the endometrium and granulomatous reaction with foreign-body giant cell reaction, fibrosis and deposition of pigment within macrophages in the myometrium. Adenomyosis is a possible explanation for endometrial regeneration in cases of ablation failure.

Key words: endometrial ablation/histology/hysterectomy/morphological changes/uterus


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Electrosurgical ablation of the endometrium represents an alternative procedure to hysterectomy for those patients having abnormal uterine bleeding refractory to hormonal therapy (Ke and Taylor, 1991Go; Fernandez et al., 1997Go; Davis et al., 1998Go). It consists of an electro-loop resection of the whole endometrium including the basal layer and the first 2–3 mm of the myometrium under direct visualization of the endometrial cavity through the hysteroscope. It can be accomplished with electrosurgical resection, roller-ball cautery, or both (Magos et al., 1991Go; Dexeus et al., 1993Go). This is a minimally invasive technique that also allows the resection of endometrial polyps and submucous myomas (Dexeus et al., 1993Go). Occasionally after ablation there is persistence of abnormal uterine bleeding; in resected specimens, pathological alterations can be found. Under these circumstances, post-ablation requirements are needed, for example repeated ablation or hysterectomy (Davis et al., 1998Go). In post-ablation resection specimens, some pathological alterations attributable to thermal effects have been described, mainly consisting of necrotizing granulomatous endometritis (Ferryman et al., 1984; Goldfarb, 1990Go; Ashworth et al., 1991Go; Thurrell et al., 1991Go; Letterie et al., 1993Go; Davis et al., 1998Go). These changes are most likely a healing response and expressed by distinct endometrial regenerative patterns that should be recognized and not be confused with lesions due to other aetiologies. The presence of adenomyosis in some hysterectomy specimens can explain the endometrial regeneration occurring in some instances. The morphological endometrial and myometrial changes in 12 post-ablation hysterectomies are presented.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
From January 1991 to August 1998, a total of 1066 hysterectomies were performed at our institution. During the same period 350 endometrial ablations were done in the same number of patients and 12 of them (3%) underwent post-ablation hysterectomy. The procedure used for ablation was a combined loop and roller-ball diathermy with an electrosurgical high frequency monopolar 27 F resectoscope (Karl Storz, GmbH & Co., Tuttlingen, Germany) connected to a generator of 100 W mean power (Karl Storz). When the endometrial ablation was concluded, electrocoagulation was applied using a roller-ball electrode at 100 W mean power (Karl Stortz) that burned the myometrial surface. A sorbitol–mannitol solution was used as distension medium. The surgical technique lasted from 15 to 30 min.

Ultrasonographic studies were performed in eight of these 12 patients before endometrial ablation and in six before hysterectomy. The mean time elapsed between ultrasonography and ablation was 4 ± 3.8 months (range 1–12 months) and between ultrasonography and hysterectomy 0.8 ± 0.7 months (0.1–2 months). The ultrasonographic studies were performed with a Toshiba SSA 270 A ultrasonograph (Toshiba Co., Tokyo, Japan) using a 6 mHz real time vaginal probe at a 150° angle. The endometrial thickness was determined measuring the larger distance between both endometrium–myometrium interfaces in longitudinal sections.

Clinical data were focused on hormonal therapy, previous pregnancy, the cause of endometrial ablation and the indication for hysterectomy in each patient.

In ablation specimens the variables recorded were weight, size and the greatest endometrial thickness (measured from the surface epithelium to the endometrial–myometrium junction of perpendicularly oriented specimens in areas without pathological changes). The number of histological slides examined for each case ranged from three to six.

Endometrial thickness in hysterectomy specimens was also measured, and all the specimens were microscopically examined for tissue destruction and pathological associated disorders. The number of endometrial representative histological slides reviewed ranged from one to three.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The patients' mean age at ablation time was 49 ± 11.6 years (range 36–78 years). Clinical indication for endometrial ablation was abnormal uterine bleeding in all patients, associated or not with a benign organic condition (Table IGo). Nine women were parous and eight of them had had an abortion. The endometrium was in the proliferative phase in five cases, in the secretory phase in one case and atrophic in six cases. Four patients had endometrial hyperplasia (two atypical, one of them complex and two non-atypical, one of them complex), six had adenomyosis, three had myomas, four had endometrial polyps and one had an endometrioid adenocarcinoma grade I. The endometrial thickness in ablation specimens ranged from 0.5 to 2.2 mm (mean 1.3 ± 0.5).


View this table:
[in this window]
[in a new window]
 
Table I. Relevant features in endometrial ablations and hysterectomies
 
There was an ablation failure in eight patients (persistence of abnormal uterine bleeding). In the four remaining, the indication for hysterectomy was atypical endometrial hyperplasia in two patients, endometrial adenocarcinoma in one and ovarian carcinoma in the other. The mean elapsed time between endometrial ablation and hysterectomy was 19 ± 17.3 months (range 1–47). The mean age of patients at hysterectomy was of 50 ± 11.1 years (range 37–79). In five case, gross scarring and stenosis of the uterine cavity were seen with formation of secondary cavities, mainly located symmetrically in the cornual areas (Figure 1Go). These cavities showed microscopically an endometrial lining. In two cases stenosis was seen in the isthmus, also narrowing the endocervical canal (Figure 2Go).



View larger version (168K):
[in this window]
[in a new window]
 
Figure 1. Gross appearance of the uterus in case 9 with a narrowed endometrial cavity and the presence of two symmetrically cornual cavities. The specimen measured 8.8 x 5.5 x 3.8 cm.

 


View larger version (156K):
[in this window]
[in a new window]
 
Figure 2. Gross appearance of the uterus in case 2 with stenosis of the endocervical canal. Scale is marked in mm.

 
In five patients, the endometrium was absent in the hysterectomy specimen. Seven cases had adenomyosis and one had non-atypical simple hyperplasia of the endometrium. The case with cancer at ablation did not show any residual tumour, and only one of the four cases with endometrial hyperplasia at ablation showed a residual lesion.

The mean endometrial thickness in the seven cases in which the endometrium was present was 1.7 ± 1.2 mm (range 0.5–4) (Table IIGo). Ultrasonography did not reveal an endometrial line in the four cases in which no endometrial lining was seen at hysterectomy. Coagulative necrosis was seen in three cases, in which the endometrium and superficial myometrium showed only ghost remnants with variable amounts of inflammatory cells. In these cases numerous coagulated and thrombosed small vessels with fibrinoid necrosis of their walls were also seen. Fibrosis with hyalinosis and scar formation was observed in four cases. In one of these cases (no. 1) electrosurgical myometrial damage such as nuclear streaming, hyperchromasia and cytoplasmatic eosinophilia were seen. Four specimens showed granulomatous reaction, consisting of a central core of amorphous, eosinophilic necrotic material surrounded by a rim of palisade histiocytes (Figure 3Go). In these cases, multinucleated foreign body giant cells were also seen, most of them containing golden-brown intracytoplasmatic pigment. This pigment was also seen in the centre of granulomas and mixed with necrotic material in the inner uterine surface. In two of the four cases with granulomatous reaction, a prominent inflammatory eosinophilic cell infiltrate was detected and in another case the eosinophilic response was diffusely seen throughout the endometrium. In three specimens, scattered macrophages containing golden-brown intracytoplasmic pigment were seen located deep in the myometrium (Figure 4Go). These cells were not accompanied by granulomatous reaction and no inflammatory response could be elicited.


View this table:
[in this window]
[in a new window]
 
Table II. Morphologic findings in uteri related to ablation effect
 


View larger version (189K):
[in this window]
[in a new window]
 
Figure 3. Granulomatous reaction in case 5 with a necrotic core with golden-brown material and multinucleated giant cells (haematoxylin–eosin). Original magnification x700.

 


View larger version (173K):
[in this window]
[in a new window]
 
Figure 4. Cluster of pigment containing macropahges within the myometrium (case 6) (haematoxylin–eosin). Original magnification x700.

 
For all patients in which the endometrium was absent in the uterus, the elapsed time between ablation and hysterectomy was <1 year (except case 3), but greater endometrial thickness did not correlate with longer elapsed time between both procedures. Necrosis and granulomatous reaction were always seen in hysterectomies performed <1 year postablation. Fibrosis was present after 7 months post-ablation and the multinucleated foreign-body giant cells reaction was seen at any time, but mainly within the first year post-ablation, in contrast to pigment-containing macrophages in the myometrium apparently representing a late post-ablation finding.

Six of the seven patients with endometrial regeneration had adenomyosis in the hysterectomy specimen and four of them also had this finding in the ablated specimens. Among those five patients without endometrial regeneration, only two had adenomyosis in the ablation and only one showed this alteration at hysterectomy. Only one case without endometrial regeneration showed adenomyosis at hysterectomy. The case with electrosurgical burn changes (no. 1) was the only case with an elapsed time of <1 month. In cases with endometrial re-growth, the endometrium was thinner in hysterectomy specimens than ablation tissue in two cases, in contrast to a thicker mucosa in the five remaining cases.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Endometrial ablation is an electrosurgical procedure used for endometrial resection in those patients having abnormal uterine bleeding, endometrial polyps and submucous leiomyomas (Magos et al., 1991Go; Dexeus et al., 1993Go). The success rate of the method is variable (Ke and Taylor, 1991Go; Gimpelson and Kaigh, 1992Go; Dexeus et al., 1993Go), and sometimes a post-ablation hysterectomy is needed. The principal causes of this ulterior procedure are persistent metrorrhagia, mainly due to adenomyosis, and other pathological conditions found during the histological examination of ablation specimens such as atypical hyperplasia and carcinoma (Gimpelson and Kaigh, 1992Go). In these post-ablation specimens it is possible to find some pathological alterations attributable to thermal effects, similar to those described following transurethral resection of the prostate (TURP) (Hedeling et al., 1981Go; Lee and Shepherd, 1983Go; Evans et al., 1984Go), that is, a surgical technique quite similar to endometrial ablation.

In gross specimens, stenosis and adhesions are the most common features (Davis et al., 1998Go) resulting in irregularities and distortion of the uterine cavity that can be a hindering factor for post-ablation pregnancies. It has also been speculated that these, along with retained endometrial islets, can subsequently undergo malignant degeneration (Horowitz et al., 1995Go).

In-vitro studies in fresh hysterectomies have demonstrated that using a mean power resectoscope of 50–100 W, almost all endometrial glands are destroyed, the degree of destruction being related to the mean power used. However a small portion of endometrial glands survive beneath the zone of destruction regardless of the power setting. This fact could explain the recurrence of abnormal uterine bleeding after ablation occurring in some patients and the presence of regrowing endometrium in post-ablation specimens (Letterie et al., 1993Go). In our series, seven of 12 post-ablation hysterectomies showed endometrial regeneration, the elapsed time between ablation and hysterectomy in these patients being larger than in patients without regeneration of the endometrium. The finding of adenomyosis in six of these seven patients favours the hypothesis of endometrial regeneration from endometrial islands located deep in the myometrium, but not destroyed by ablation as has been previously suggested (Davis et al., 1998Go). Other studies have demonstrated that angiogenesis can play an important role in endometrial regeneration (Kooy et al., 1996Go) because in histologically normal post-ablation endometrium there is an increase in endothelial cell number but a low endothelial cell proliferation, suggesting a low turn-over of endothelial cells in these specimens.

Necrotizing granulomas similar to those seen in rheumatoid arthritis are one of that most frequent microscopical findings in post-ablation specimens (Ashworth et al., 1991Go; Thurrell et al., 1991Go; Ferryman et al., 1992Go; Davis et al., 1998Go). They consist of a central core of necrosis with a rim of palisade histiocytes and a multinucleated foreign body giant cell reaction. Sometimes within the giant cells and in the necrotic centre, a golden-brown material can be seen (Clarke and Simpson, 1990Go; Ashworth et al., 1991Go; Thurrell et al., 1991Go). Electron diffraction analysis of this material has proved it contains carbon and aluminium oxalate (Thurrell et al., 1991Go). The origin of these granulomas in post-TURP specimens has been attributed to an altered collagen reaction as a result of previous surgery or to the metal deposition from the instruments used (Henry et al., 1993Go). Eosinophilic reaction in these granulomas should not be attributed to allergy. Its presence is more evident in those cases with a short-interoperative period (Lee and Shepherd, 1983Go). We have not been able to find previous reports dealing with the presence of deeply located pigment-containing macrophages in the myometrium. The pigment, most likely carbon particles due to thermal effects during the ablation, has similar optical characteristics to that found in the granulomas and in the inner surface of the uterine wall.

Endometrial regeneration, scarring and fibrosis seem to be long-term post-ablation findings. Necrosis, granulomatous and foreign-body giant cell reaction, eosinophilic infiltrate and pigment containing macrophages in the myometrium are histological features found in post-ablation hysterectomies with a short intersurgical period. All these findings in post-ablation hysterectomies should not be confused with lesions due to other agents, allergic or infectious (Ashworth et al., 1991Go).


    Notes
 
3 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Ashworth, M.T., Moss, C.I. and Kenyon, W.E. (1991) Granulomatous endometritis following hysteroscopic resection of the endometrium. Histopathology, 18, 185–187.[ISI][Medline]

Clarke, T.J. and Simpson, R.H. (1990) Necrotizing granulomas of peritoneum following diathermy ablation of endometriosis. Histopathology, 16, 400–402.[ISI][Medline]

Davis, J.R., Maynard, K.K., Brainard, C.P. et al. (1998) Effects of thermal endometrial ablation. Clinicopathologic correlations. Am. J. Clin. Pathol., 109, 96–100.[ISI][Medline]

Dexeus, S., Ubeda, A., Labastida, R. et al. (1993) Endometrial ablation in dysfunctional uterine bleeding. J. S. O. G. C., 15, 957–962.

Evans, C.S., Goldman, R.L., Kleinm, H.Z. et al. (1984) Necrobiotic granulomas of the uterine cervix. A probable postoperative reaction. Am. J. Surg. Pathol., 8, 841–844.[ISI][Medline]

Fernandez, H., Capella, S. and Audibert, F. (1997) Uterine thermal balloon therapy under local anaesthesia for the treatment of menorrhagia: a pilot study. Hum. Reprod., 12, 2511–2514.[Abstract]

Ferryman, S.R., Stephens, M. and Gough, D. (1992) Necrotising granulomatous endometritis following endometrial ablation therapy. Br. J. Obstet. Gynaecol., 99, 928–930.[ISI][Medline]

Gimpelson, R.J. and Kaigh, J. (1992) Endometrial ablation repeat procedures. Case studies. J. Reprod. Med., 37, 629–635.[ISI][Medline]

Goldfarb, H.A. (1990) A review of 35 endometrial ablations using the Nd:YAG laser for recurrent menometrorrhagia. Obstet. Gynecol., 76, 833–835.[Abstract]

Hedeling, H., Johansson, S. and Nilsson, S. (1981) Focal prostatic granulomas. A sequel to transurethral resection. Scand. J. Urol., Nephrol., 15, 193–196.[ISI][Medline]

Henry, L., Wagner, B., Faulkner, M.K. et al. (1993) Metal deposition in post-surgical granulomas of the urinary tract. Histopathology, 22, 457–465.[ISI][Medline]

Horowitz, I.R., Copas, P.R., Aaronoff, M. et al. (1995) Endometrial adenocarcinoma following endometrial ablation for postmenopausal bleeding. Gynecol. Oncol., 56, 460–463.[ISI][Medline]

Ke, R.W. and Taylor, P.J. (1991) Endometrial ablation to control excessive uterine bleeding. Hum. Reprod., 6, 574–580.[Abstract]

Kooy, J., Taylo, N.H., Healy, D.L. et al. (1996) Endothelial cell proliferation in the endometrium of women with menorrhagia and in women following endometrial ablation. Hum. Reprod., 11, 1067–1072.[Abstract]

Lee, G. and Shepherd, N. (1983) Necrotising granulomata in prostatic resection specimens: a sequel to previous operation. J. Clin. Pathol., 36, 1067–1070.[Abstract]

Letterie, G.S., Hibbert, M.L. and Britton, B.A. (1993) Endometrial histology after electrocoagulation using different power settings. Fertil. Steril., 60, 647–651.[ISI][Medline]

Magos, A.L., Baumann, R., Lockwood, G.M. et al. (1991) Experience with the first 250 endometrial resections for menorrhagia. Lancet, 337, 1074–1078.[ISI][Medline]

Thurrell, W., Reid, P., Kennedy, A. et al. (1991) Necrotizing granulomas of the peritoneum (letter). Histopathology, 18, 190.[Medline]

Submitted on November 23, 1998; accepted on February 17, 1999.