The diagnosis of endometriosis traditionally needs invasive procedures which include laparoscopy and biopsy of suspected lesions, [1] however, this can delay diagnosis for many years [2-5] and can carry surgical risks [6]. Many studies have assessed the role of endometrial nerve fibers detection as a diagnostic modality for endometriosis which is surly simpler and less invasive than laparoscopy and initial results were reassuring [7-9].Conversely, many previous studies found that nerve fibers are found in relation to pain regardless the gynecological disease and declined its role in diagnosis of endometriosis [10,11].

The aim of this study is to assess the presence of endometrial nerve fibers in different gynecological pathologies compared to endometriosis and associate their density to the type and degree of pelvic pain if present.

This cross sectional study was conducted at Ain Shams University department of obstetrics and gynecology between December 2010 and February 2014, including 325 women 20-60 years old, undergoing laparoscopy or laparotomy for different indications namely, infertility, polycystic ovarian disease (PCOD), ovarian cysts, fibroid uterus, adenomyosis, endometrial hyperplasia, endometrial carcinoma and suspected endometriosis. Women who received hormonal treatment on the last 6 months or having combined pathologies were excluded in addition to cases with unsatisfactory biopsy specimens. Approval was obtained from the ethical committee of the department of Obstetrics and Gynecology, Ain-Shams University. A written informed consent was obtained from each participant.

Pain intensity was documented before surgery using a 100-mm visual analogue scale (VAS) bounded by “no pain” and “the worst pain”. Endometrial biopsies were obtained by an endometrial curette for women undergoing surgeries other than hysterectomy in which a strip of the endo-myometrial interface was sectioned. Endometrial biopsy specimens were fixed in 10% neutral buffered formalin immediately after collection for at least 24 hours, processed, paraffin embedded and stored in room temperature until further use. Paraffin blocks were sectioned at 4 mm thickness on a Leica microtome (type 2055 Autocut, Nussloch, Germany). Serial sections for histological and immunohistochemical examinations were made. Endometrial biopsy was morphologically assessed prior to immunohistochemical staining where unsatisfactory endometrium and uterine polyps were excluded.

Immunohistochemistry was performed using polyclonal rabbit anti-PGP 9.5, a highly specific pan neuronal marker, which recognizes all types of nerve fibers. We used normal small intestine as positive control containing rich supply of submucous nerve plexus with each run of slides prepared in order to ensure proper staining of the run.

Figures 1-3 shows the presence of endometrial nerve fibers while Figure 4 shows their absence.

Figure 1: A tiny brown small nerve fibers pointed by arrows in the functional layer of endometrium in a patient presenting with pain; immunohistochemical staining with anti-protein gene product 9.5(PGP9.5) and × 400 magnification.

Figure 2: Multiple endometrial nerve fibers shown as woven light brown bundles parallel and in close proximity to endometrial glands pain; immunohistochemical staining with anti-protein gene product 9.5(PGP9.5) and × 400 magnification.

Figure 3: Tiny brown nerve fibers indicated by the black arrows in the functional layer of endometrium in a woman with endometriosis (immunohistochemical staining with pan neuronal marker PGP 9.5 and x400 magnifications).

Figure 4: Functional layer endometrium showing no nerve fibers in the glands or the adjacent stroma; immunohistochemical staining with antiprotein gene product 9.5(PGP9.5) and × 400 magnification.

Data was statistically analyzed using IBM SPSS version 22 (SPSS Inc., Chicago, IL). Chi-square test (Fisher’s exact test) was used to examine the relation between qualitative variables. For quantitative data, comparison between two groups was done using independent sample t-test or Mann-Whitney test. Comparison between 3 groups was done using Kruskal-Wallis test. Spearman-rho method was used to test correlation between numerical variables. Binary logistic regression was used for multivariate of analysis of factors associated with the presence of endometrial nerve fibers. A p-value<0.05 was considered significant (Figures 1-4).

Three hundred and sixty-nine women were assessed for eligibility and 44 has been excluded after histopathological assessment due combined pathology (Figure 1). Demographic and clinical characteristics of the studied subjects. Pain was the main presentation in patients with endometriosis (88%) followed by adenomyosis (43%), while no patients with PCOS had pain. The proportion of patients with pain was significantly associated with disease type (p<0.001). In those suffering from pain, pain scores were significantly lower in patients with adenomyosis relative to the other pathological entities (p=0.033).

Nerve fibers in the functional layer of the endometrium were detected in 65 patients (20%). All patients with endometriosis (n=25) had nerve fibers in addition to 41 patients representing 13.7% of nonendometriotic pathologies. Apart from endometriosis, the highest proportion of positive nerve fibers (17.1%) was found in cases of adenomyosis and fibroid uterus. Thus, endometriosis had significantly higher proportion of nerve fiber detection (p<0.001). No endometrial nerve fibers were present in patients with PCOS. However, density of nerve fibers did not differ significantly between the studied pathological groups (p=0.108) (Figure 5).

Figure 5: Flowchart showing the distribution of study population among different groups. AM: Adenomyosis; F: Fibroid; EH: Endometrial Hyperplasia; EC: Endometrial Carcinoma; PCO: Polycystic Ovary; BOT: Benign Ovarian Tumors; MOT: Malignant Ovarian Tumors; ENDO: Endometriosis; P: Presenting By Pain; NP: Not Presenting By Pain; NF: Nerve Fibers Present.

The presence of nerve fibers was positively associated with the presence of pain (p<0.001) (Figure 6). The density of nerve fibers was significantly higher in the presence of pain; 6.0 (1.3-30.0) NF/mm³ compared to absence of pain; 3.3 (2.2-8.0) NF/mm3 (p=0.004). In the 52 patients with pain and positive nerve fibers, pain scores had a strong correlation with nerve fiber density (r=0.479, p<0.001).

Figure 6: Relation between the presence of pain and the presence of nerve fibers in the functional layer of the endometrium.

Patients with endometrial nerve fibers were significantly older (p=0.003), but with comparable parity (p=0.742) and BMI (p=0.449) with those without endometrial nerve fibers. Using multivariate analysis the presence of pain was the only independent factor associated with the presence of endometrial nerve fibers with an Odds ratio of 14.5 (95% CI: 6.7-31.2).

This study demonstrates that nerve fibers in the functional layer of the endometrium are detected in 13.7% of non-endometriotic gynecological pathologies compared to 100% of endometriotic patients (p<0.001). The presence of nerve fibers is positively associated with the presence of pain (p<0.001). Pain is the only independent factor associated with the presence of endometrial nerve fibers (OR 14.5; 95%CI: 6.7-31.2). The nerve fiber density is strongly correlated with pain scores (r=0.479, p<0.001). Adenomyosis has the lowest pain scores relative to the other pathological entities (p=0.033).

The presence of endometrial nerve fibers in cases of endometriosis has been evaluated in many previous studies [8,12,13] and significant association has been confirmed. One study confirmed the validity of immunohistochemical detection of endometrial nerve fibers for diagnosis of endometriosis with high sensitivity and specificity (98% and 83%, respectively) which is supposed to reduce the more invasive laparoscopy for diagnosis of endometriosis. No endometrial nerve fibers were found in women with adenomyosis, endometrial hyperplasia, chronic endometritis, leiomyomas and endometrial polyps [8]. Another study showed a sensitivity of 92% and specificity of 80% and the addition of serum IL-6 markedly improved the test performance [14].

Other studies confirmed that the presence of nerve fibers demonstrated by immunohistochemistry was restricted to patients with endometriosis [7,15]. Additionaly, determining the density of nerve fibers can be a reliable marker for diagnosis of endometriosis [13].

The concept of high specificity of endometrial nerve fibers to endometriosis has been challenged in other studies with growing evidence of its relation to pelvic pain rather than pathology. PGP9.5- positive nerve fibers have been found in endometrium in women with pain, whether they have endometriosis, uterine fibroids or adenomyosis, and to correlate with the severity of pain symptoms in these three disorders [10,11].

Endometrial tissue from 74 hysterectomy specimens of cases with adenomyosis and uterine fibroids was examined and nerve fibers were only detected in the functional layer of endometrium only in patients with pain with a frequency higher than that in the current study; 64% in adenomyosis and 67% in uterine fibroids [16]. Remarkably, they found immunoreactive nerve fibers in the basal layer of endometrium in patients with and without pain.

Another study [17] included 31 patients with endometriosis and 29 patients as a control group with other diagnoses and the results showed that the detection of nerve fibers is not statistically different between the 2 groups. Leslie et al. [18] found endometrial nerve fibers in conventional biopsy specimens of 19% of patients with and 29% of patients without endometriosis. In a group of 44 women with pelvic pain, endometrial fine nerve fibers were present in patients with and without endometriosis. Poor sensitivity and specificity for diagnosis of endometriosis were found (31.8 and 45.5% respectively) [19].

The findings of the current study indicate that endometrial nerve fibersare linked to pain rather than a specific pathological condition. Detecting endometrial nerve fibers more frequently in cases of endometriosis is explained by the higher proportion of endometriotic women complaining of pain.

The presence of nerve fibers in the functional layer of the endometrium was suggested to contribute to pain generation during office hysteroscopy in women with endometriosis and adenomyosis [20]. A positive relationship between density of neural fibers and pain intensity was found in peritoneal endometriotic lesions. Neural fibers were significantly higher in deep endometriotic lesions compared to normal peritoneum [10].

We can conclude that the presence of nerve fibers in the functional layer of the endometrium is fairly common in endometriotic as well as non-endometriotic gynecological pathologies principally in association with pain. Thus, detection of these nerve fibers cannot be safely trusted for the diagnosis of endometriosis. Future studies may be directed toward certain molecular products that interact with and/or influence these nerve fibers within eutopic endometrium or endometriotic plaques. To further investigate a definite mechanism of endometriosisassociated pain, further research is needed to reveal the detailed interplay between functional molecules and identified nerve fibers, and their tandem interactions aiding in the progression of symptoms.

All the authors declare no conflict of interest.