Detection of chronic endometritis at fluid hysteroscopy

Size: px

Start display at page:

Download "Detection of chronic endometritis at fluid hysteroscopy"

Aubrey Baldwin Lyons
5 years ago
Views:

1 Journal of Minimally Invasive Gynecology (2005) 12, Detection of chronic endometritis at fluid hysteroscopy Ettore Cicinelli, MD, Leonardo Resta, MD, Roberto Nicoletti, MD, Massimo Tartagni, MD, Marco Marinaccio, MD, Carlo Bulletti, MD, and Giuseppe Colafiglio, MD From the 1st Department of Obstetrics and Gynaecology (Drs. Cicinelli, Nicoletti, Tartagini, Marinaccio, and Colafiglio), and Institute of Pathology (Dr. Resta), Faculty of Medicine, University of Bari, Bari; and the Department of Obstetrics and Gynecology, Hospital of Rimini (Dr. Bulletti), Rimini, Italy. KEY WORDS: Hysteroscopy; Chronic endometritis; Micropolyps; Saline distention Abstract STUDY OBJECTIVE: Chronic endometritis is a subtle condition that is difficult to detect; however, it may cause abnormal uterine bleeding and infertility. Few data exist about the appearance of chronic endometritis at fluid hysteroscopy and about the value of diagnostic fluid hysteroscopy in the detection of this condition. In our experience, at fluid hysteroscopy chronic endometritis is characterized by consistent association of stromal edema and either focal or diffuse hyperemia; in some cases, this finding is associated with endometrial micropolyps (less than 1 mm in size). This study attempted to describe diagnostic criteria for chronic endometritis at fluid hysteroscopy and assess the diagnostic accuracy of fluid hysteroscopy in the detection of this condition. DESIGN: Retrospective Study (Canadian Task Force classification II-2). SETTING: University hospital. PATIENTS: Nine hundred-ten women in whom hysteroscopy was indicated. INTERVENTIONS: Fluid hysteroscopy followed by endometrial biopsy. MEASUREMENTS AND MAIN RESULTS: Sensitivity, specificity, positive and negative predictive values, and accuracy of fluid hysteroscopy in the detection of chronic endometritis, based on the association of edema; hyperemia; and, if present, micropolyps were calculated. Based on the presence of hyperemia and edema, chronic endometritis was diagnosed in 158 patients (17.4%); in 61 patients (6.7%), micropolyps also were present. Histology confirmed the diagnosis in 101 patients (63.9% of positive cases at hysteroscopy) and was positive in 9 additional cases not detected by hysteroscopy. Chronic endometritis at histology was found in approximately 30% of infertile women and 35% of cases related to abnormal uterine bleeding. The sensitivity, specificity, and positive and negative predictive values of hysteroscopy for chronic endometritis based on detection of only hyperemia and edema were 91.8%, 92.9%, 63.9%, and 98.8%, respectively; the diagnostic accuracy was 92.7 %. The combination of hyperemia, edema, and micropolyps had sensitivity, specificity, and positive and negative predictive values of 55.4%, 99.9%, 98.4%, 94.5%, respectively, with a diagnostic accuracy of 93.4%. CONCLUSIONS: Fluid hysteroscopy is very reliable in diagnosing no inflammation, while detection of micropolyps is a very reliable sign of inflammation. When performing hysteroscopy for abnormal uterine bleeding or infertility, signs of chronic endometritis should always be sought AAGL. All rights reserved /$ -see front matter 2005 AAGL. All rights reserved. doi: /j.jmig

2 Cicinelli et al Fluid hysteroscopy and chronic endometritis 515 Corresponding author: Ettore Cicinelli, MD, Department of Obstetrics and Gynaecology, University of Bari, Policlinico, Piazza Giulio Cesare, Bari, Italy. Submitted March 9, Accepted for publication July 14, Chronic endometritis is a subtle condition that may cause abnormal uterine bleeding (AUB) and infertility. 1,2 Clinically, chronic endometritis in most cases is asymptomatic or accompanied by mild disturbances like spotting, mild and undefined pelvic pain, and leukorrhea. 1,2 Vaginal bleeding is reported to be the major presenting symptom in up to 94% of patients with chronic endometritis. 2 Chronic endometritis may have severe reproductive consequences on fertility in spontaneous as well as in vitro fertilization (IVF) cycles. 1 From a diagnostic point of view, chronic endometritis is a condition that it is difficult to identify. It is not detectable on transvaginal sonography, and it can be only suspected in patients with complications like adhesions, pyometra, or hydrometra. Recently, the presence of vaginal polymorphonuclear leukocytes has been reported to have a high sensitivity and negative predictive value for the diagnosis of upper genital tract infection. 3 Hysteroscopy allows the inspection of the endometrial surface and the detection of signs of inflammation. At conventional CO 2 hysteroscopy, chronic endometritis is characterized by the presence of areas of red endometrium flushed with a white central point, localized or scattered throughout the cavity, presenting a typical aspect called strawberry aspect. 1 The reddish color with a starry sky white configuration is similar in appearance to diffuse colpitis seen with a colposcope. It may also appear as whitish plaques that are friable and bleed easily. 4 However, areas of red endometrium may be caused by vascular wall breaking due to the employment of CO 2, hormone effects, 5 hypertension, immunologic diseases, and other causes. Moreover, each of the above-described hysteroscopic signs may be mild so that even at hysteroscopy the diagnosis may be doubtful or missed. This explains why no general agreement exists in the literature about diagnostic usefulness of hysteroscopy in the detection of chronic endometritis. 1,6 In recent years, the use of saline as a distention medium of the uterine cavity at hysteroscopy has probably become more popular than standard CO 2. 7,8 Saline has no effect on endometrial micro-circulation, provides smoother distention and continuous washing of the uterine cavity, and allows floating of endometrial ingrowths. This implies findings at fluid hysteroscopy may be different from the traditional diagnostic images described with standard CO 2 hysteroscopy. Accordingly, we have recently demonstrated the presence of very small (less that 1 mm of size) pedunculated endometrial polyps, lesions never described at CO 2 hysteroscopy, is significantly associated with chronic endometritis (Figures 1A, 1B, and 1C). 9 Histologically, micropolyps are small, vascularized ingrowths covered by endometrium and characterized by an accumulation of inflammatory cells (lymphocytes, plasma cells, or eosinophilic granulocytes) (Figure 1D). 9 Whenever micropolyps were detected, homogeneous or, more frequently, nonhomogenous endometrial thickening and focal or diffuse periglandular hyperemia also were present; conversely, in about 45% of cases of chronic endometritis confirmed at histological examination, micropolyps were not detectable. 9 Another sign in our experience with fluid hysteroscopy could be related to endometrial inflammation is stromal edema in the proliferative phase that causes the endometrial mucosa to appear irregularly thick and pale. This study attempted to describe diagnostic criteria for chronic endometritis at fluid hysteroscopy and assess the diagnostic value of fluid hysteroscopy in the detection of this condition. For this purpose, we compared histologic and hysteroscopic findings from women who underwent hysteroscopy for different indications. Materials and methods From January 2004 through January 2005, we examined 910 outpatient women in whom diagnostic minihysteroscopy was indicated (Table 1). The study was approved by our institutional review board, and all women gave their informed consent. Minihysteroscopies were performed using a lens-based 2.7-mm outer diameter (OD) mini-telescope, 105-degree angle of visual field equipped with a 3.5 mm OD single-flow diagnostic sheath (Slim-line Hysteroscope, ACMI, Southborough, MA). All hysteroscopies were performed in the follicular phase. Heavy bleeding, severe cardio-vascular disease, and suspected pregnancy were considered contraindications. Saline was used to distend the uterine cavity at a pressure generated by simple drop from a bag suspended 1 m above the patient. A 300 W light source with a xenon bulb (mv 9087; ACMI, Southborough, MA), a 3 CCD digital camera (Micro-Digital IIIe, ACMI, Southborough, MA), and a 21- inch video color screen (Sony Trinitron, PVM-20M2MDE, Sinigawa-Ku, Tokyo, Japan) were used. The exploration of the uterine cavity consisted of a panoramic view of the cavity as a first step followed by the examination of both cornua, tubal ostia, and anterior and posterior walls; after that, a thorough evaluation of the endometrial mucosa was performed as previously described. 9 Briefly, the tip of the hysteroscope approached both uterine walls in order to get a view paralleling the surface; in this way, any irregularity of the mucosa can be identified easily. All hysteroscopies were performed by two of the authors (EC and RN). Cases of suspected endometrial neoplasia were not included in the study. After hysteroscopy, all patients underwent endometrial biopsy by means of a 3-mm Novak s curette connected to a 20-mL syringe. All proce-

516 Journal of Minimally Invasive Gynecology, Vol 12, No 6, November/December 2005 Figure 1 Different aspects of chronic endometritis at fluid hysteroscopy.

3 516 Journal of Minimally Invasive Gynecology, Vol 12, No 6, November/December 2005 Figure 1 Different aspects of chronic endometritis at fluid hysteroscopy. (A) Endometrial mucosa appears thick, edematous, and focally hyperemic. (B) Endometrial mucosa shows focal hyperemia, adhesions, nonhomogeneous edema, and sporadic micropolyps. (C) Endometrial surface is completely covered by micropolyps. (D) Histologic preparation stained with hematoxylin & eosin (200 magnification) showing two micropolyps. The surface of these new growths is covered by endometrium while the stroma is characterized by an accumulation of inflammatory cells (lymphocytes, plasma cells or eosinophilic granulocytes) intermingled with normal stromal cells, small vessels, and glandular structures. dures were performed without any kind of anesthesia. A single oral dose of antibiotic (800 mg pefloxacin) was given at the end of examinations. Endometrial samples were fixed in neutral formalin and embedded in paraffin according to routine histologic procedure. Five microsections were stained with hematoxylin & eosin. A single operator (LR), who was unaware of hysteroscopic findings, performed histologic examination following diagnostic criteria described in the literature. 4 Superficial stromal edema, increased stromal density, pleomorphic Table 1 group Indications for hysteroscopy and prevalence of chronic endometritis at hysteroscopy and histology for each indication Indication Total No. CE (Hys) No. (%) Micropolyps (Hys) CE (His) No. TP% GP% No. TP% GP% Premenopausal AUB* (20.9) 21 (9.3) (44.7) 26 (11.6) (55.3) Postmenopausal bleeding (2.0) 0 2 (1.4) (66.7) Infertility (40.7) 23 (15.3) (37.7) 45 (30.0) (73.8) Polyp at TVE (14.8) 9 (5.3) (36) 14 (8.3) (56.0) Cervical polyp 92 9 (9.8) 3 (3.3) (33.3) 6 (6.5) (66.7) Submucous myoma 56 4 (7.1) 0 2 (3.6) (50.0) Malformation 72 9 (12.5) 5 (6.9) (55.5) 6 (8.3) (66.7) Total (17.4) 61 (6.7) (38.6) 101 (11.1) (63.9) AUB abnormal uterine bleeding; CE chronic endometritis; GP group population; His histology; Hys hysteroscopy; TP total population; TVE transvaginal echography. *Indicates women without evidence of any abnormality at transvaginal echography.

4 Cicinelli et al Fluid hysteroscopy and chronic endometritis 517 stromal inflammatory infiltrate dominated by lymphocytes, and plasma cells were considered as signs of inflammation. Data are reported as mean SD. Sensitivity, specificity, and positive and negative predictive values of hysteroscopy for chronic endometritis based on the detection of hyperemia and stromal edema, alone and in combination with micropolyps, in comparison with histology were calculated by means of a 2 2 table 10 ; the accuracy of hysteroscopic diagnosis of chronic endometritis (true positive true negative/total population) was also calculated. Results The prevalence of chronic endometritis at fluid hysteroscopy and at histology for each indication group is displayed in Table 1. Only six (0.7%) patients had undergone endometrial biopsy or any other intrauterine procedure within 6 months of hysteroscopy, three in the infertile group, two in the premenopausal AUB group, and one in the postmenopausal AUB group. At fluid hysteroscopy, based on the presence of hyperemia and edema, chronic endometritis was diagnosed in 158 patients (17.4%); in 61 patients, micropolyps were also present (6.7% of general population and 38.6% of all endometritis at hysteroscopy). Histology confirmed the diagnosis in 101 cases (63.9% of positive cases at hysteroscopy) and was positive in 9 additional cases not detected by hysteroscopy. The sensitivity, specificity, and positive and negative predictive values of hysteroscopy based on detection of only hyperemia and edema were 91.8%, 92.9%, 63.9% and 98.8%, respectively; the diagnostic accuracy was 92.7%. When considering the presence of hyperemia, edema, and micropolyps, the sensitivity, specificity, and positive and negative predictive values were 55.4%, 99.9%, 98.4%, 94.5%, respectively, with a diagnostic accuracy of 93.4%. Discussion The results of the present study demonstrate that diagnosis of no inflammation at fluid hysteroscopy is very reliable, while the detection of micropolyps strongly suggests the existence of endometrial inflammation. In fact, the absence of endometrial hyperemia and edema at fluid hysteroscopy has a very high negative predictive value (98.8%); in other words, when at fluid hysteroscopy we verify the absence of these signs, we can exclude the presence of inflammation. On the other hand, the presence of micropolyps that is always associated with hyperemia and stromal edema is a very reliable sign of inflammation as demonstrated by the very high positive predictive value of the presence of all these signs (98.4%). Micropolyps are subtle lesions that we have recently demonstrated to be related to inflammation. 9 In our experience, they have great diagnostic value; in fact, notwithstanding their small size, detection of micropolyps is easy when saline is used for distending the uterine cavity as saline allows the micropolyps to float. To the contrary, when using CO 2 these endometrial ingrowths flatten against the endometrial surface. Moreover, using saline the continuous washing of the uterine cavity ensures a clear vision in the presence of mucus or bleeding that is frequently found in cases of endometrial inflammation. The discovery of endometrial micropolyps and other subtle alterations of the endometrial surface with saline distention is in accordance with findings on the ovarian and tubal surface at transvaginal hydrolaparoscopy; also in this case hydroflotation and magnification are considered the key factors for detecting subtle lesions that are easily missed with conventional CO 2 laparoscopy. 11 The study also demonstrated that chronic endometritis was present in about one-third of women undergoing hysteroscopy due to reproductive or bleeding problems: histology was positive in 31.3% of women undergoing hysteroscopy due to unexplained infertility (up to 39.6% of patients if we also consider those with uterine malformation) and in 36.2% of women with indications related to AUB (premenopausal AUB, suspected endometrial polyp, endocervical polyp, intracavitary myoma). Endometrial biopsy or any other intrauterine procedure within 6 months of hysteroscopy was reported by only 0.7% of patients, so that in our population these procedures may account for a few cases of chronic endometritis. This is in accordance with a study that found some pathologic finding at diagnostic hysteroscopy before IVF in 45% of patients; most of these abnormalities were endometritis. 12 Others have reported that the inflammatory aspects of the endometrium (when observed by hysteroscopy) accompany the failures of IVF and that there is a positive correlation between endometrial congestion and positive spermculture. 13 Our results disagree with those of other researchers who concluded that hysteroscopy is not useful for screening chronic endometritis in asymptomatic infertile women. In fact, by estimating the incidence of chronic endometritis in their population as great as 10%, they concluded that a negative predictive value of about 89.1% did not add any useful information for risk assessment. However, these researchers used CO 2 and not saline as distention medium; moreover, an incidence of approximately 10% of chronic endometritis in infertile women is significantly lower than that reported by most authors. 12,13 Micropolyps were detected in premenopausal women but not in postmenopausal women, suggesting that micropolyps, characterized by the accumulation of inflammatory cells and edema in the stroma, are probably expression of an active and strong endometrial reaction and massive release of interleukins and local growth factors. Micropolyps were present only in approximately 40% of patients with chronic endometritis. Even considering that in

5 518 Journal of Minimally Invasive Gynecology, Vol 12, No 6, November/December 2005 some cases we could miss the lesions, it is likely that either the intensity of endometrial biochemical response or the different nature of infectious agents may determine the micropolyps formation. Unfortunately, infectious investigations were not performed, and therefore we cannot make any correlation between etiologic agents and presence of micropolyps. Stromal edema, if combined with hyperemia, is another sign of inflammation that could be easily detected using saline as distention medium at low pressure (70 mm Hg). The endometrial mucosa outside the areas of hyperemia, although in the proliferative phase, appears pale, whitish, and nonhomogeneously thick; in some cases it may fold, simulating a polyp. Stromal edema and endometrial thickening may be suspected at transvaginal echography as these alterations change the normal endometrial pattern. One study found that a nonhomogeneous hyperechogenic sonographic endometrial echo pattern predicts lower fertility potential in women who are not receiving follicle-maturing drugs. 17 It can be argued that stromal edema may be determined by the accumulation of saline into the mucosa during the examination; in fact, it is known that the longer hysteroscopy takes, the more edema accumulates into the mucosa. However, in our hands diagnostic hysteroscopy usually takes a very short time (approximately 2 3 min) so that it is unlikely that edema could accumulate. Moreover, we do not perform any surgical procedure, and we try to avoid any indentation of the mucosa. At histology, signs of chronic endometritis may be focal; and as inflammatory cells normally are present in the endometrial mucosa, the histologic diagnosis of chronic endometritis is not always easy. In order to facilitate the histologic diagnosis, we tried to obtain endometrial samples as large as possible. We therefore preferred to perform oriented endometrial biopsies by using a 3-mm Novak s curette instead of guided biopsies using endoscopic forceps. Moreover, as described above, all histologic examinations were performed by the same experienced histopathologist (LR). Conclusion Fluid hysteroscopy is a reliable and useful examination for investigating chronic endometritis. It could be useful for screening asymptomatic infertile patients and for evaluating women with AUB. Bearing in mind that chronic endometritis is a hidden condition that is difficult to detect with noninvasive examination, we suggest that fluid hysteroscopy should be always performed in the diagnostic work-up of women with unexplained infertility, especially before starting assisted reproduction procedures, or in women complaining of AUB. References 1. Cravello L, Porcu G, D Ercole C, et al. Identification and treatment of endometritis. Contracept Fertil Sex. 1997;25: Greenwood SM, Moran JJ. Chronic endometritis: morphologic and clinical observations. Obstet Gynecol. 1981;58: Yudin MH, Hillier SL, Wiesenfeld HC, et al. Vaginal polymorphonuclear leukocytes and bacterial vaginosis as markers for histologic endometritis among women without symptoms of pelvic inflammatory disease. Am J Obstet Gynecol. 2003;188: Dotto JE, Lerna B, Dotto JR Jr, et al. Classification of microhysteroscopic images and their correlation with histologic findings. JAm Assoc Gynecol Laparosc. 2003;10: Fraser IS, Hickey M. Endometrial vascular changes and bleeding disturbances with long-acting progestins. Steroids. 2000;65: Polisseni F, Bambirra EA, Camargos AF. Detection of chronic endometritis by diagnostic hysteroscopy in asymptomatic infertile patients. Gynecol Obstet Invest. 2003;55: Indman PD. Instrumentation and distention media for the hysteroscopic treatment of abnormal uterine bleeding. Obstet Gynecol Clin North Am. 2000;27: Cicinelli E, Parisi C, Galantino P, et al. Reliability, feasibility, and safety of minihysteroscopy with a vaginoscopic approach: experience with 6,000 cases. Fertil Steril. 2003;80: Cicinelli E, Resta L, Nicoletti R, et al. Endometrial micropolyps at fluid hysteroscopy suggest the existence of chronic endometritis. Hum Reprod. 2005;20: Stempel LE. Eenie, meenie, minie, mo... What do the data really show? Am J Obstet Gynecol. 1992;144: Gordts S, Campo R, Puttemans P, et al. Investigation of the infertile couple: a one-stop outpatient endoscopy-based approach. Hum Reprod. 2002;17: Feghali J, Bakar J, Mayenga JM, et al. Systematic hysteroscopy prior to in vitro fertilization. Gynecol Obstet Fertil. 2003;31: Taylor S, Frydman R. Hysteroscopy and sperm infection. Contracept Fertil Sex. 1996;24: Dechaud H, Maudelonde T, Daures JP, et al. Evaluation of endometrial inflammation by quantification of macrophages, T lymphocytes, and interleukin 1 and 6 in human endometrium. J Assist Reprod Genet. 1998;15: Richter HE, Holley RL, Andrews WW, et al. The association of interleukin 6 with clinical and laboratory parameters of acute pelvic inflammatory disease. Am J Obstet Gynecol. 1999;181: Romero R, Espinoza J, Mazor M. Can endometrial infection/inflammation explain implantation failure, spontaneous abortion, and preterm birth after in vitro fertilization? Fertil Steril. 2004;82: Check JH, Gandica R, Dietterich C, Lurie D. Evaluation of a nonhomogeneous endometrial echo pattern in the midluteal phase as a potential factor associated with unexplained infertility. Fertil Steril. 2003;79:

The major causes of female infertility include ovulatory dysfunction, tubal and peritoneal

The major causes of female infertility include ovulatory dysfunction, tubal and peritoneal Focused Issue of This Month YoungMin Choi, MD Department of Obstetrics and Gynecology, Seoul National University College of Medicine Email : ymchoi@snu.ac.kr J Korean Med Assoc 2007; 50(5): 400-405 Abstract