Chronic pelvic pain (CPP) and dysmenorrhea are the most common health problems that affect women of childbearing age [1]. There is a variety of causes for CPP which include endometriosis, adenomyosis, chronic infection, vulvodynia, irritable bowel syndrome, and bladder pain syndrome. Endometriosis is the most common cause of CPP [2], accounting for 24–40% of all CPP diagnoses [3]. For the management of endometriosis, surgery is a frequent choice since the efficacy of medical treatment alone is either poorly documented or of limited efficacy. However, owing to the unclear etiology of endometriosis, its recurrence rate following surgery remains high. Reoperation occurs in 51% of patients with endometriosis, often resulting in damage to ovarian reserve [4]. The risk for reoperation, coupled with uncertainty for results and the presence of continued pelvic pain, makes endometriosis a chronic disease.

Reya et al. [5] proposed a cancer stem cell (CSC) theory that some rare cell populations that exist in cancer tissues, had the capacity for self-renewal, multipotential differentiation, tumorigenesis, metastasis, relapse and treatment-resistance. Evidence exists that endometrial mesenchymal stem cells (eMSCs) are located within the endometrium [6]. eMSCs are believed to contribute to cyclical changes of human endometrium, including proliferation, differentiation, tissue breakdown and shedding under the influence of estrogen and progesterone during the menstrual cycle [7]. The migration of eMSCs is similar to that of CSCs, with endometriosis possessing biological behaviors of local aggressiveness, distant metastasis and high disease recurrence. eMSCs may play a key role in the development and relapse of endometriosis. A suggested hypothesis is that eMSCs are abnormally shed during menses, present in amniotic fluid and blood. They are capable of gaining access to the peritoneal cavity or an abdominal wall scar, where they establish ectopic implants in those women who develop endometriosis.

This work aimed to identify eMSCs in retrograde menstruation, various endometriosis lesions, and normal control tissues in order to investigate the association between eMSCs and endometriosis. The effectiveness of gonadotropin-releasing hormone agonists (GnRH-a) on endometriosis was also compared using percent of eMSCs between the groups.

In the period from 1 January 2015 to 31 December 2019, a total of 508 patients, which included 270 cases of ovarian endometriosis, 129 cases of adenomyosis, 37 cases of AWSE, and 72 cases of DE, were enrolled. The age (mean $\pm$ SD) of the study population was 34.8 $\pm$ 0.3 years and BMI was 20.9 $\pm$ 0.1 kg/m${}^2$. Two hundred and sixty-four patients (52.0%) were nulliparous. A total of 242 open surgeries and 266 laparoscopies were performed. The 2-year recurrence rate was 6.9% (26/379), which included 20 cases of ovarian endometriosis, 1 case of AWSE and 5 cases of DE. Baseline data for all patients is shown in Table 1.

For evaluation of GnRH-a, a treatment group consisted of 50 adenomyosis patients and 16 DE patients, with a similar control group. There was no significant difference from baseline between the treatment group and the control group (p $>$ 0.05, Table 2).

The operation was carried out after menstruation. Twenty-three cases of retrograde menstruation were found during surgery and gathered for analysis. We detected eMSCs in retrograde menstruation material. As shown in Fig. 1, the eMSCs expressing PDGFR-$\beta$${}^{+}$ and CD146${}^{+}$, showed red and green light, while the nuclei, stained with DAPI, showed blue light.

Fig. 1.

The eMSCs in retrograde menstruation ($\mathrm{\times }$400).

eMSCs were detected in various endometriotic lesions. However, eMSCs could not be identified in normal ovary, myometrium, adipose tissue or peritoneal membrane. Fig. 2 shows the comparison between endometriosis and normal tissues. The percents (mean $\pm$ SD) of eMSCs were 0.88% $\pm$ 0.01% of ovarian endometriosis (0.88% $\pm$ 0.01% of relapse-free patients vs 0.88% $\pm$ 0.03% of relapse patients; p = 0.88), 1.01% $\pm$ 0.01% of adenomyosis, 0.76% $\pm$ 0.02% of AWSE (0.76% $\pm$ 0.02% of relapse-free patients vs 0.68% of 1 relapse patient; p = 0.40), and 1.06% $\pm$ 0.01% of DE (1.06% $\pm$ 0.01% of relapse-free patients vs 1.04% $\pm$ 0.05% of relapse patients; p = 0.65) (Fig. 3). Statistical differences were not found between relapse-free patients and those experiencing a relapse.

Fig. 2.

The eMSCs in different endometriosis lesions and control normal tissues ($\mathrm{\times }$100). A-1,2,3 ovarian endometriosis, A-0 ovary tissue. B-1,2,3 adenomyosis, B-0 myometrium of uterine. C-1,2,3 abdominal wall scar endometriosis, C-0 abdominal adipose tissue. D-1,2,3 deep endometriosis, D-0 peritoneal membrane.

Fig. 3.

Percent of eMSCs in different endometriosis lesions. (A) Ovarian endometriosis relapse-free. (a) Ovarian endometriosis with relapse. (B) Adenomyosis. (C) Abdominal wall scar endometriosis relapse-free. (c) Abdominal wall scar endometriosis with relapse. (D) Deep endometriosis relapse-free. (d) Deep endometriosis with relapse.

Specimens obtained from the GnRH-a treatment and control groups were tested to quantitatively compare eMSCs in both groups. As shown in Fig. 4, eMSCs were found in both adenomyosis and DE in the two groups. There was no statistical difference in percents of eMSCs between the treatment group and the control group for adenomyosis patients (1.00% $\pm$ 0.03% vs 1.05% $\pm$ 0.03%, p = 0.27, Fig. 5). Similar to patients with DE, no statistical difference was found for eMSCs between the groups (1.13% $\pm$ 0.06% vs 1.15% $\pm$ 0.05%, p = 0.82, Fig. 5). Evaluation of the 2-year recurrence rate demonstrated no statistical difference (p = 0.64, Table 2).

Fig. 4.

The eMSCs in the GnRH-a treatment group and the control group ($\mathrm{\times }$100). AM, adenomyosis; DE, deep endometriosis.

Fig. 5.

Percents of eMSCs in the GnRH-a treatment group and the control group. (A) Treatment group of adenomyosis. (B) Control group of adenomyosis. (C) Treatment group of deep endometriosis. (D) Control group of deep endometriosis.

CPP and dysmenorrhea are common health problems and affect women of reproductive age. Women with CPP, regardless of a diagnosis of endometriosis, experience significant negative impact across a range of life issues including education, work, social, and sexual relationships [8]. Although endometriosis is one of the most common gynecological disorders, the etiology of endometriosis still remains unclear, resulting in unsatisfactory management and high risk of recurrence after surgery. Theories include implantation theory, metaplasia theory of coelomic epithelium and induction theory, none of which clearly explain the etiology of endometriosis resulting in non-ideal treatments [9, 10, 11]. To improve therapeutic effect, the etiology of endometriosis needs to be clarified. Previous studies have suggested that the recurring endometriotic lesions arise from lesions or cells not completely removed during the primary surgery [12]. Thus, the recurrence may be unavoidable until the etiology is clarified and targeted therapy developed. The CSCs theory proposes that both tumor development and progression are driven by undifferentiated stem cells capable of self-renewal and tumor-initiation. Considering the migration of eMSCs being similar to that of CSCs, and with endometriosis possessing biological behaviors of local aggressiveness, distant metastasis and high disease recurrence, similar to those of ovarian cancer, eMSCs may play a key role in development and relapse of endometriosis.

In 2007, Schwab et al. [6] harvested CD146${}^{+}$/PDGFR-$\beta$${}^{+}$ cells from endometrial biopsy samples, which resided in both basalis and functional layers of the endometrium, indicating that eMSCs are likely to be shed in menstrual blood [13, 14]. CD146${}^{+}$/PDGFR-$\beta$${}^{+}$ cells, which play a key role in regeneration of the endometrium after menstruation, participated in endometrial cell proliferation, angiogenesis, and implantation [15]. Masuda et al. [7] transplanted these cells onto the kidney capsule of NOD/SCID mice whose ovaries were removed, and induced these mice with estrogen and progestin. The cells developed endometrial like structure containing gland and stroma, changing cyclically by estrogen and progestin, indicating these cells owned characteristics of stem cells. Retrograde menstruation, a common phenomenon in women of reproductive age, is a widely accepted risk factor for developing endometriosis. Approximately 76–90% of patients experience retrograde menstruation, particularly those with stenosis or atresia of the cervix [16, 17]. In addition, typical endometriosis lesions developed in animal models when menstrual blood was injected [18]. Membrane surface markers, PDGFR-$\beta$ and CD146, as well as nuclei marker DAPI, were chosen to identify eMSCs in retrograde menstruation [19]. We hypothesized that eMSCs might be shed by menstrual blood into the peritoneal cavity and being induced by sex hormones and microenvironment to evolve into endometriosis, which is similar to the implantation theory. We analyzed various specimens using immunofluorescence and found eMSCs existing in endometriosis lesions. The mean percents of eMSCs varied between a low of 0.76% to a high of 1.06%, slightly below 1.50% in normal endometrium as reported by Schwan et al. [13] We did not identify eMSCs in control normal tissue from the same case, indicating that eMSCs found in endometriosis may arise from other tissues or pathways. We also detected eMSCs in AWSE, indicating that eMSCs might shed with amniotic fluid when patients underwent a cesarean section, differentiate by being exposed to sex hormones and microenvironment, resulting in the development of endometriosis.

GnRH-a, a man-made gonadotropin-releasing hormone, can suppress adenohypophysis and reduce estradiol. GnRH-a could be an adjuvant treatment for endometriosis after surgery [20]. However, the effect of GnRH-a is reversible and recurrence may occur if treatment ceases [21]. A multicenter randomized controlled trial demonstrated that compared with surgery alone, postoperative GnRH-a could prolong the period of recurrence but not effect fertility or reduce recurrence rate (p = 0.08) [22]. In our research, percents of eMSCs showed no statistical difference between the GnRH-a treatment group and the control group, whether adenomyosis (p = 0.27) or DE (p = 0.82) was present. We postulated that GnRH-a may only affect mature endometrial cells except for eMSCs. When treatment is completed or interrupted, the patients’ hormone concentrations return and eMSCs are activated, resulting in recurrence. This suggests that the effect of GnRH-a may be limited for endometriosis.

CSCs also have the capacity to affect treatment-resistant tumors [5]. The traditional therapy might not affect CSCs, since residual cells developed homologous ovarian cancer lesions near the primary site [23, 24]. Evidence has revealed that CSCs are not only responsible for primary tumor growth, metastasis and relapse of disease, but also for the development of chemoresistance [25]. Considering that the migration of eMSCs is similar to that of CSCs and that endometriosis possesses biological behaviors of local aggressiveness, distant metastasis and high disease recurrence similar to those of ovarian cancer, eMSCs may also play a leading role in the limited efficacy of various drug treatments. Effects of drugs and conservative surgery have a limited role for endometriosis [4]. More than half of patients who underwent conservative surgery required a second surgical procedure. Repeat surgery may damage the patient’s ovary harming fertility as well as endocrine function [26]. Recurrent dysmenorrhea and infertility also reduce the patient’s quality of life. Our results have demonstrated that peritoneal endometriosis (ovarian endometriosis and DE) might be associated with eMSCs in retrograde menstruation while AWSE is associated with their presence in amniotic fluid at the time of caesarean section. Regardless of whether patients received postoperative GnRH-a, the recurrence rate of AWSE is far below that of peritoneal endometriosis following complete removal of lesions [27]. This may suggest that retrograde menstruation is closely related to the inevitable recurrence of endometriosis within the pelvic cavity. Hence, we postulate that even though surgery could remove every lesion, recurrence cannot be prevented. We suggest that gynecologists regard endometriosis as a chronic disease and develop an individual lifetime management plan for patients based on age, clinical manifestations, fertility desire and quality of life.

Our study demonstrated that eMSCs played a critical role in the development and recurrence of endometriosis and that GnRH-a did not affect eMSCs. Gynecologists may regard endometriosis as a chronic disease resulting in lifetime management, especially for patients with CPP.

JZ—protocol development, data collection, data analysis, manuscript writing. XL—data analysis, manuscript editing. TY—protocol development, manuscript editing. AT—data collection, data analysis, manuscript writing. RP—data collection, data analysis, manuscript writing. GZ—data analysis, manuscript writing. SL—data analysis, manuscript writing. XZ—protocol development, manuscript editing. CB—protocol development, manuscript editing. GS—protocol development, manuscript editing. All authors read and approved the final manuscript.

This study was approved by the ethics committee on human research at West China Second Hospital, Sichuan University (No. 2020074). All subjects gave their informed consent for inclusion before they participated in the study.

Not applicable.

This work was supported by the Key Research Projects of Sichuan Province (grant numbers S15059).

The authors declare no conflict of interest.