Clinical Characteristics of Women with Surgical Signs of Superficial Peritoneal Endometriosis but a Negative Histology: A Nested Case-Control Study

Endometriosis remains a controversial entity regarding its conceptual and working definitions. While defined by most experts as the presence of endometrium-like tissue outside the uterus [1, 2], many disagree about what is an “endometrium-like tissue” for diagnostic and management purposes [3]. Even the localization of lesions “outside the uterus” is sometimes replaced by “outside the uterine cavity” [4, 5], which stricto sensu means considering adenomyosis a type of endometriosis. Current guidelines include diagnostic criteria that range from symptoms to noninvasive imaging to surgical visualization of lesions to histological evidence of glands and/or stroma [2, 6‒8], which leaves some patients with a partial diagnosis of endometriosis that is neither ruled out nor fully confirmed. This is particularly valid for superficial peritoneal endometriosis (SUP) due to the high proportion of biopsy specimens that fail to be corroborated by a histological diagnosis [9].

We have questioned what clinical features distinguish this unique group of symptomatic patients with surgical visualization of SUP lesions but no histological confirmation of the disease. Our hypothesis was that by having only nonspecific peritoneal lesions they would have a milder peritoneal inflammation and be less symptomatic than those with true, histologically proven endometriotic lesions. Thus, the aim of this study was to investigate how endometriosis symptoms affect women with SUP diagnosed by laparoscopy and not confirmed by histology, in comparison with a group with histologically proven SUP and a control group without endometriosis.

This was a single-center, hospital-based, nested case-control study. We retrospectively identified all non-pregnant patients aged 18–47 years submitted to a laparoscopy or laparotomy for a benign gynecologic condition at our institution from October 2004 to February 2023. Indications for surgery have been detailed elsewhere [10] and included pelvic pain, infertility, pelvic mass (e.g., uterine fibroids, benign adnexal cysts) and abnormal uterine bleeding. Exclusion criteria were pregnancy, malignancy, and refusal or impossibility to answer to the interviewer’s questions.

We then classified the selected participants (n = 2,989) based on the surgical report into a group with (n = 1,552) and a group without endometriosis (n = 1,437, online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000543910). The group with endometriosis was filtered to retain patients with only SUP (n = 390) and exclude cases with the presence of any other endometriosis phenotype or without a histological report. The same number of participants (n = 390) was randomly selected from the group without endometriosis to form a control group. To avoid selection bias, the list of 1,437 potential controls was ordered by date of intervention and the first 3 of each sequence of 11 were included. The SUP group was subdivided into a subgroup with histological confirmation of endometriosis (histology positive, n = 245) and a subgroup whose biopsy did not confirm the presence of endometriosis (histology negative, n = 145, online suppl. Fig. 1).

For each patient, face-to-face interviews had been conducted by the surgeon during the month preceding surgery. These interviews used a structured questionnaire that queried about clinical history, detailed menstrual and reproductive history, past and current use of hormonal contraception, and painful symptoms (dysmenorrhea, deep dyspareunia, and non-cyclic chronic pelvic pain). Pain severity was assessed preoperatively through a 10 cm visual analog scale (VAS) ranging from no pain to the worst possible pain. Dysmenorrhea and dyspareunia were considered mild when VAS was 1–3, moderate when VAS was 4–6, and severe when VAS was 7–10 [11]. The participants from both groups had a complete visual evaluation of the abdominopelvic cavity by the same surgeons following a standard procedure [10, 12]. The histological examination of the excised specimens was based on hematoxylin-eosin staining and performed by pathologists who were specialized and experienced in the diagnosis of endometriosis [13, 14].

Data were analyzed with the software IBM SPSS 22.0 (IBM Corporation, Armonk, NY, USA). Data distribution was tested using the Shapiro-Wilk test. Normally distributed data are summarized as means ± standard deviation and the groups were compared by unpaired t test or one-way analysis of variance (ANOVA) followed by Tukey post hoc test. Categorical data were compared by chi-square tests with Bonferroni correction. The study sample was sufficient to detect differences of at least 15% in the prevalence of symptoms or 0.3 standard deviation in scale variables between the groups, with power (1-β) = 0.80 and significance level (α) = 0.05 [15].

As shown in Table 1, the two subgroups with SUP and the control group were comparable as they had similar age (group means ranging from 31 to 32 years), body mass index (22–23 kg/m²), smoking prevalence (27%–36%) serum anti-müllerian hormone levels (3.8–4.2 ng/mL) and menstrual cycle patterns (73%–78% with regular menstrual cycles). In contrast, when clinical characteristics related to endometriosis were compared (Table 2), both SUP subgroups (histology-positive and histology-negative) differed significantly from the control group. Remarkably, the SUP/histology-negative subgroup was more likely to have a familial history of endometriosis (14% vs. 1%) or a personal history of primary infertility (29% vs. 19%) or primary dysmenorrhea (50% vs. 33%) compared to the control group (all p values <0.01, chi-square test).

Indicators of the severity of dysmenorrhea such as the VAS score (mean 5.7 vs. 4.1), absenteeism from school during menstruation (34% vs. 15%), and faint during menstruation (10% vs. 5%) were all severer in the SUP/histology-negative subgroup than in the control group. The same pattern was observed for the intensity of other painful symptoms like deep dyspareunia and non-cyclic chronic pelvic pain, which was higher in both SUP subgroups than in the control group (p < 0.05, ANOVA and Tukey’s test, Table 2). Only gastrointestinal symptoms were significantly less intense in the SUP/histology-negative than in the SUP/histology-positive subgroup, but in both subgroups the scores indicated mild intensity pain (VAS = 1.2 ± 2.4 vs. 2.2 ± 3.1). Serum CA-125 levels did not differ between groups (Table 2).

The straightforward definition of endometriosis as the presence of endometrial-like tissue outside the uterus does not capture the huge complexity of this disease, from early pathogenesis to clinical manifestations and response to treatments [6]. In more than a century of attempted classification, there is still no consensus as if endometriosis is a single disease with various phenotypes or several diseases sharing some genetic, epigenetic, and clinical features [1, 16, 17].

The most recent ESHRE guideline recommends that “laparoscopic identification of endometriotic lesions is confirmed by histology although negative histology does not entirely rule out the disease” [2]. Another expert panel mentioned a “stage 0” disease, “which could mean strongly suspected endometriosis based on combinations of symptoms and examination findings strongly predictive of endometriosis, […] visualized but not histologically confirmed” [6]. If we assume that histology is not a perfect and definite criterion to exclude the presence of endometriosis in symptomatic people, we should consider why some patients could have a negative histology. In a large multicenter cohort study performed in Australia, 25% of SUP biopsy specimens resulted negative [18]. These cases were labeled false positives, considering histology as the gold standard. But could at least some of them be actually true SUP? Accordingly, the authors agree we the possibility that thermal damages during lesion excision may cause histological artifacts that hamper the proper identification of endometriosis structures [18]. Another explanation could be that many pathologists request the unequivocal identification of both glands and stroma to confirm the presence of endometriosis [9] despite the great morphological diversity of SUP lesions and the underappreciated role of fibrosis [19, 20]. In addition, the detection of stromal endometriosis can be enhanced by immunohistochemistry for CD10 antigen, but this is not used routinely [21].

We have previously observed that the presence of SUP is sufficient to increase the risk of infertility, even in the absence of ovarian endometrioma or anatomic distortion of the pelvis [22]. In the present study, the prevalence of primary infertility was significantly higher in both groups with suspected SUP compared to the control group (29% vs. 19%). Thus, the subgroup with a negative histology was equally prone to infertility as the subgroup with a positive histology, and this cannot be attributed to confounding bias of age, body mass index, or menstrual cycle pattern. Interestingly, serum anti-müllerian hormone levels were similar in all groups. In previous studies, we have noted that oligo-anovulation is as common in patients with endometriosis (regardless of the phenotype) than in a matched control group without endometriosis [12], and the present results indicate that the SUP phenotype associates with infertility by mechanisms other than ovarian reserve.

Pelvic pain was also more frequent and intense in the SUP groups than in the control group, regardless of SUP histological confirmation. This was valid for dysmenorrhea, dyspareunia, bowel, and vesical pain related to menstruation, as well as noncyclic chronic pelvic pain. We, therefore, rise several hypotheses for the painful symptoms in the SUP/histology-negative subgroup. First, the pain could be caused by classical endometriotic lesions that were not seen by the pathologist due to some artifact. Second, the lesions might have evolved from the classical morphology with stroma and/or glands to a more fibrotic pattern, and this was overlooked or dismissed. Third, the biopsied tissues did not contain active endometriotic lesions because they were atrophied by medical treatment, but the underlying inflammatory response was still present. Fourth, their symptoms were unrelated to endometriosis and the cause was not identified.

A main strength of this study is the large number of cases of SUP without the confounding effects of deep endometriosis or ovarian endometrioma. Another strength is the comprehensive assessment of symptoms through face-to-face interviews carried out within 1 month before surgery, using a standard questionnaire and a semiquantitative VAS to measure the intensity of pelvic pain symptoms. Furthermore, the endometriosis and control groups answered to the same questionnaire, which was applied for the same medical team. Finally, all the participants from both groups had a complete visual evaluation of the pelvic cavity performed by the same team of surgeons at the same hospital. One limitation of the present study is its setting being a referral surgical center that receives patients with more advanced disease compared to typical outpatient settings. Although our data collection was standardized and curated, this study was retrospective and therefore some source of performance bias cannot be ruled out. Further limitations are the lack of detailed records of the laparoscopic appearance of lesions and of the microscopic features described by the pathologists.

Our initial hypothesis was refuted by the finding that patients with suspected SUP lesions and a negative histology had clinical characteristics similar to those with histological confirmation of endometriosis. Therefore, the former condition should be further investigated using molecular tools and biomarkers, to clarify why these women are so symptomatic in the absence of histological evidence of endometriosis.

We used a database started after ethical approval by the Institutional Review Board of Hospital Cochin (protocol 05-2006), and all participants had provided written informed consent to have their data anonymously used in future research.

The authors report no conflict of interest.

This study received no funding. FMR was recipient of a scholarship from Conselho Nacional de Desenvolvimento Científico e Tecnológico (grant # 201876/2017-5).

Design and manuscript drafting: F.M.R., C.C.; data curation: P.S., L.M., M.B., C.M., and C.C.; data analysis: F.M.R. and P.S.; manuscript editing: P.S. and C.C.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author (C.C.) upon reasonable request.