Standardized method to develop an endometriosis model in Sprague–Dawley rats: A step-by-step description

Article information

Korean J Fertil Steril. 2026;.cerm.2025.08690

Publication date (electronic) : 2026 August 4

doi : https://doi.org/10.5653/cerm.2025.08690

Herbert Situmorang

Department of Obstetrics and Gynecology, Dr. Cipto Mangunkusumo National Referral Hospital, Faculty of Medicine, University of Indonesia, Jakarta, Indonesia

Corresponding author: Herbert Situmorang Department of Obstetrics and Gynecology, Dr. Cipto Mangunkusumo National Referral Hospital, Faculty of Medicine, University of Indonesia, Jl. Salemba Raya No. 6, Central Jakarta, Jakarta Capital Special Region, Indonesia Tel: +62-813-8310-0567 E-mail: datadrherbert@gmail.com

Received 2025 September 18; Revised 2025 December 1; Accepted 2026 February 23.

Abstract

Introduction

Endometriosis is a chronic, estrogen-dependent disorder characterized by the presence of functional endometrial glands and stroma outside the uterine cavity. Ectopic lesions most commonly occur on the ovaries, pelvic peritoneum, and uterosacral ligaments, although extrapelvic manifestations—such as involvement of the diaphragm or abdominal wall—have also been reported [1,2]. The condition affects an estimated 10% of women of reproductive age and is particularly prevalent among those with infertility, in whom rates may reach 30% to 50%, and among those with chronic pelvic pain, where prevalence may be as high as 70% to 80% [1,3]. Beyond its clinical manifestations, endometriosis imposes substantial psychological and socioeconomic burdens, underscoring its importance as a major public health concern.

Despite decades of investigation, the pathogenesis of endometriosis remains incompletely understood. Several theories have been proposed to explain the ectopic implantation and persistence of endometrial tissue. Sampson’s theory of retrograde menstruation remains the most widely cited; however, alternative hypotheses—including coelomic metaplasia and Müllerian remnant theories—also retain relevance [4,5]. These mechanisms are increasingly viewed not as mutually exclusive but as interconnected processes involving immune dysfunction, hormonal dysregulation, angiogenesis, and genetic susceptibility that collectively drive lesion establishment and persistence [4,5]. This multifactorial complexity has hindered the development of definitive diagnostic strategies and targeted therapies.

Clinically, endometriosis presents heterogeneously. Although some women remain asymptomatic, many experience cyclic pelvic pain, dysmenorrhea, dyspareunia, dyschezia, or urinary pain. Notably, symptom severity does not consistently correlate with disease stage or depth of invasion [1]. Infertility, often resulting from pelvic adhesions and distorted pelvic anatomy, further highlights the profound reproductive consequences of the disease [2]. Laparoscopy with histopathological confirmation remains the diagnostic gold standard, and current treatment strategies include hormonal suppression, surgical excision, or a combination of both. However, none of these approaches guarantees long-term remission, and recurrence remains common.

The limitations of current diagnostic and therapeutic options underscore the need for reliable experimental models of endometriosis to elucidate disease mechanisms and facilitate therapeutic development. Because spontaneous endometriosis occurs almost exclusively in nonhuman primates—models limited by ethical and financial constraints—rodents have become indispensable in preclinical research. Among these species, the Sprague–Dawley rat is particularly advantageous because of its predictable estrous cycles, ease of handling, and ability to generate reproducible lesions through surgical induction [6]. However, unlike menstruating species, rats do not undergo cyclic endometrial shedding, limiting replication of Sampson’s retrograde menstruation theory. Consequently, surgically induced homologous models remain the primary experimental approach in rats [7].

The earliest systematic rat model of endometriosis was described by Jones [8], who introduced a homologous autotransplantation technique in Sprague–Dawley rats. Their method involved excising a uterine horn segment, separating the myometrium from the endometrium, and transplanting a standardized 5×5 mm endometrial fragment onto the peritoneal wall with the epithelial surface facing the cavity. The graft was secured at four corners, and viability was assessed 3 weeks later [8]. Building on this approach, Vernon and Wilson [9] compared three induction strategies—sutured uterine squares, intraperitoneal luminal lavage, and dispersed endometrial scrapings—and demonstrated that only sutured autotransplants yielded viable, cystic lesions, thereby establishing surgical implantation as the most reliable method. do Amaral et al. [10] subsequently adapted the Jones procedure in Wistar rats using a full-thickness uterine flap sutured to the abdominal wall, improving graft handling and lesion reproducibility. Collectively, these foundational models established the key methodological principles—autologous transplantation, stable tissue fixation, and hormonal responsiveness—that continue to underpin contemporary induction techniques. Nevertheless, variability persists in implantation success, tissue orientation, suture configuration, infection control, estrous-cycle timing, and postoperative outcomes, with reported success rates ranging from 60% to 80%. Inconsistent graft adherence remains a significant limitation, often leading to epithelial loss or incomplete lesion establishment.

In the absence of a rigorously standardized rodent model with consistently reproducible results, a methodological gap persists that limits comparability across studies. The present study addresses these shortcomings by refining the Vivian technique through strict estrous-phase synchronization, verified endometrial orientation, four-point graft fixation, and prophylactic antibiotic administration. These modifications are intended to reduce interoperator variability, enhance reproducibility, and provide a clearly defined standardized protocol suitable for mechanistic and translational research.

Methods

2. Estrous-cycle determination

Estrous cyclicity was assessed to ensure hormonal synchronization before surgery. Vaginal smears were obtained by gently inserting either a plastic pipette containing 10 μL of sterile 0.9% sodium chloride or a saline-moistened cotton swab into the vaginal canal. Secretions were transferred to glass slides, fixed in 90% ethanol, and stained with Giemsa. Slides were examined using a standard light microscope. Estrous phases were classified as proestrus, estrus, metestrus, or diestrus based on the relative proportions of epithelial cells and leukocytes (Figure 1) [11]. Cycles were considered normal when they repeated every 4–7 days. Only animals in proestrus or estrus underwent surgical induction to maximize implantation success.

Figure 1.

Estrous cycle in rats. (A) Diestrus: Predominance of leukocytes with few or no epithelial cells. (B) Proestrus: Decreased leukocytes with increasing epithelial cells. (C) Estrus: Field composed entirely of epithelial cells, with no leukocytes observed. (D) Metestrus: Postestrus phase characterized by declining epithelial cells with a concomitant increase in leukocytes. Original micrographs reproduced from Zenclussen et al. [11].

3. Induction of endometriosis

Anesthesia was induced with intramuscular ketamine hydrochloride (90 mg/kg) and xylazine (10 mg/kg). Prior to surgery, animals were weighed to calculate anesthetic dosage. After induction, the abdominal wall hair was removed using clippers, and rats were positioned supine on a surgical board with limbs abducted. Routine asepsis and antisepsis were performed.

A 4-cm midline incision was made below the umbilicus to expose the peritoneal cavity. The bicornuate uterus was identified (Figure 2), and a 1-cm segment of uterine horn was excised after securing hemostasis by proximal and distal ligation with 3–0 silk sutures. The excised segment was opened longitudinally to expose the endometrial surface (Figure 3). The flap of uterine tissue was prepared and sutured to the anterior peritoneal wall of the right lower quadrant, approximately 2 cm from the abdominal incision, with the endometrial side facing the peritoneum. Fixation was achieved using four interrupted 6–0 nylon sutures (Figure 4).

Figure 2.

Identification of the uterus.

Figure 3.

Excision of a uterine horn segment.

Figure 4.

Suturing of the uterine tissue to the peritoneal wall.

After confirming hemostasis, the abdominal wall was closed with 4–0 chromic catgut in one layer, and the skin was closed with 4–0 silk sutures. Rats were returned to their cages after recovery from anesthesia. Prophylactic antibiotic treatment was administered as intraperitoneal ceftriaxone at a dose of 10 mg per rat on postoperative days 0, 1, and 2.

4. Lesion assessment

One month later, a second laparotomy was performed under the same anesthesia and asepsis protocol. A 4-cm midline incision was made, and the peritoneal cavity was inspected. Macroscopic features of endometriotic lesions were documented and photographed using an iPhone 7 Plus camera (Apple Inc.) with a millimeter scale for calibration (Figure 5).

Figure 5.

Macroscopic appearance of the lesion and area measurement. (A) Representative macroscopic image of the lesion. (B) Representative macroscopic image after area measurement using SketchAndCalc software (iCalc Inc.).

Results

All 60 rats (100%) developed macroscopically visible endometriotic lesions, indicating that each implanted endometrial fragment successfully formed a lesion using the modified Vivian technique. At the second laparotomy, performed 1 month after implantation, cystic lesions were consistently identified on the peritoneal surface. Lesions appeared as fluid-filled, cystic masses with smooth surfaces and a hyperemic coloration compared with the surrounding peritoneal tissue. Prominent blood vessels converged toward the lesions, and adhesions to adjacent bowel or omentum were observed in several animals (Figure 6). The surface area of each lesion was documented by digital photography with a millimeter scale and subsequently measured using SketchAndCalc software (iCalc Inc.).

Figure 6.

(A) Endometriotic lesion and (B) endometriotic lesion with intestinal adhesion.

Quantitative measurements demonstrated expected biological variability, with lesion areas ranging from 0.83 to 131.50 mm² (mean, 45.27 mm²; median, 37.34 mm²) across the cohort (Table 1). This distribution aligns with prior reports of surgically induced endometriosis, in which lesion heterogeneity is typical despite standardized implantation techniques (Figure 7). The uniformly successful lesion formation observed in the present study compares favorably with earlier models, which often report 60% to 80% viability because of inconsistent graft orientation, limited suture fixation, or suboptimal estrous timing at implantation.

Table 1.

Lesion area distribution after surgical induction of endometriosis in rats

Figure 7.

Distribution of endometriotic lesion area in the rat model. Scatter plot showing the distribution of lesion area measurements.

Histological analysis of excised lesions further confirmed the establishment of endometriotic tissue. H&E staining demonstrated cystic structures lined by a single layer of cuboidal epithelial cells, with surrounding endometrial-like glands and stromal components. Stromal cellularity was most pronounced adjacent to the cyst wall and gradually decreased toward the periphery, consistent with organized tissue architecture within the implants (Figures 8 and 9).

Figure 8.

Histologic appearance of rat endometriotic lesions (hematoxylin and eosin stain). (A) Tissue section from the endometriosis model (40×). Scale bar: 0.1 mm. (B) Cystic lesion containing endometrial cells and leukocytes; the cyst wall is lined by a single layer of cuboidal epithelium (yellow arrow) (40×). Cystic lesion surrounded by endometrial-like stroma and gland-like structures (pink arrow) (40×). Scale bar: 0.1 mm.

Figure 9.

(A) High-power view of the cyst wall and surrounding endometrial-like tissue (hematoxylin and eosin stain). Cystic lesion containing endometrial cells and leukocytes; the cyst wall is lined by a single layer of cuboidal epithelium (yellow arrow) (400×). (B) Cystic lesion surrounded by endometrial-like stroma and gland-like structures (pink arrow) (400×). Scale bar: 0.1 mm.

Discussion

This study established a standardized rat model of endometriosis with a 100% lesion induction rate, exceeding the success typically reported in classical models such as Vernon and Wilson [9] and the original Vivian technique. The high reproducibility observed here likely reflects several purposeful refinements. First, implantation was performed exclusively during the proestrus or estrus phase, when endogenous estrogen levels peak, thereby providing a hormonally favorable environment that supports endometrial cell survival, attachment, and subsequent growth [9,10,12]. Second, the uterine implant was fixed using four interrupted 6–0 nylon sutures, ensuring uniform apposition of the endometrial surface to the peritoneum and reducing the risk of graft detachment, a recognized limitation of earlier two-suture techniques. This more comprehensive anchoring likely contributed substantially to consistent lesion establishment across animals. Third, perioperative prophylactic antibiotic administration may have reduced postoperative infection, thereby limiting inflammatory complications that can compromise graft viability. Collectively, these modifications improved procedural stability, addressed key limitations of prior rodent endometriosis models, and enabled reliable lesion formation.

Rodent models of endometriosis are broadly categorized as homologous or heterologous. Homologous models, such as ours, use autologous or syngeneic uterine tissue transplanted into the peritoneal cavity, whereas heterologous models use human tissue in immunocompromised hosts [13]. Despite phylogenetic limitations, rat models remain widely used because they balance affordability, surgical feasibility, and preservation of immunological integrity. Our findings support the use of standardized homologous models as reliable preclinical platforms and complement menstrual-mimicry approaches, such as those proposed by Persoons et al. [14].

A defining feature of this model was the consistent development of neovascularization and adhesions, both of which are hallmarks of clinically significant endometriotic disease. Angiogenesis is essential for the survival and expansion of ectopic endometrial implants and is largely mediated by vascular endothelial growth factor (VEGF), which has been repeatedly implicated as a central driver of lesion vascularization in rodent and human studies [15,16]. Although VEGF expression was not quantified, the pronounced macroscopic neovascularization observed aligns with established angiogenic mechanisms and supports the biological validity of the induced lesions. Similarly, the frequent formation of adhesions—often underreported in rodent models despite their contribution to infertility, dyspareunia, and chronic pelvic pain—enhances the translational relevance of this approach [13].

The inflammatory microenvironment also underpins these pathological processes. Activated macrophages and pro-inflammatory cytokines, particularly tumor necrosis factor alpha (TNF-α), promote implantation, persistence, and progression of ectopic endometrium, while emerging mediators such as interleukin 37 correlate with disease severity and may serve as biomarkers of lesion activity [17,18]. Although cytokine profiles were not measured in this study, the macroscopic and histological findings—cystic morphology, prominent vascularization, and well-preserved glandular and stromal architecture—are consistent with inflammatory cascades described in contemporary rodent endometriosis models. H&E staining provided morphological confirmation; however, current evidence suggests that optimal validation of experimentally induced lesions includes immunohistochemical (IHC) markers to substantiate tissue identity, distinguish endometriotic implants from nonspecific peritoneal reactions, and assess angiogenic or inflammatory activity [19]. The absence of molecular and IHC analyses therefore represents a methodological limitation, particularly because current standards emphasize evaluation of mediators such as TNF-α, VEGF, and oxidative stress markers (e.g., malondialdehyde) to strengthen mechanistic interpretation and translational relevance. Future applications of this standardized model should incorporate targeted molecular and IHC assays to enhance lesion characterization, support reproducibility, and improve comparability with human disease.

Oxidative stress also plays a central role in endometriosis pathophysiology. Reactive oxygen species and byproducts such as malondialdehyde contribute to peritoneal inflammation and cellular stress, which may promote lesion persistence and impair reproductive function [6]. The interaction between oxidative stress and inflammation has been linked to infertility in both clinical and experimental studies, underscoring the value of incorporating oxidative stress biomarkers in future applications of this model.

These findings are also consistent with evidence showing reproductive consequences of endometriosis in rats, including impaired ovarian reserve and reduced fertility [6]. By reliably inducing lesions with adhesion formation, this model provides a robust framework for investigating how endometriosis disrupts reproductive function and for evaluating targeted interventions. Pharmacological studies have shown that agents such as rosiglitazone can reduce lesion progression in surgically induced rat models, further illustrating the utility of standardized systems for preclinical drug testing [20].

Several limitations should be acknowledged. First, rodents do not menstruate, which limits the extent to which retrograde menstruation can be modeled as a mechanism of disease establishment [7,14]. Second, pain assessment—an essential translational endpoint—was not included and should be prioritized in future studies [5]. Third, molecular and IHC evaluation of angiogenic, inflammatory, and oxidative stress pathways would complement the morphological findings reported here. In addition, the present method specifically generates peritoneal wall endometriosis, whereas human disease encompasses multiple subtypes, including superficial peritoneal lesions, deep infiltrating endometriosis, and ovarian endometriomas. Developing parallel protocols to model these phenotypes, and clearly subclassifying experimental models accordingly, would improve translational applicability and better reflect clinical heterogeneity [21].

In conclusion, this modified rat model of endometriosis reliably recapitulates essential disease features, including vascularization, cyst formation, and adhesion development. The reproducibility achieved in the present study supports its use as a platform for mechanistic studies and preclinical evaluation of novel therapies. Future research should extend this framework by incorporating molecular validation and pain phenotyping to better align experimental outcomes with the clinical complexity of endometriosis.

Article information Continued