Increase in Angiogenesis and Vascularization in Patient-Derived Endometriosis Tissue: Insights from a 3D In Vivo Model

Abstract

Aim

Endometriosis is a gynecological disorder characterized by endometrial-like tissue outside the uterus. This study evaluates the vascularization and proliferation of human endometriosis and endometrium tissues engrafted onto the chorioallantoic membrane of chicken embryos using immunohistochemistry and laser speckle contrast analysis imaging. For the assessment of clinical relevance, a comparison between laboratory and clinical data was performed.

Material and Methods

Tissue samples from 10 patients categorized by #Enzian scores and undergoing endometriosis surgery were investigated in the chorioallantoic membrane model. Hematoxylin-eosin staining and immunohistochemical markers, including CD10, cytokeratin, Ki67, and Caspase-3, assessed cellular structures, proliferation, and apoptosis. Changes in blood perfusion, implemented as a surrogate marker for angiogenesis and vascularization, were analyzed over three days using laser speckle contrast analysis. The fertilized chicken eggs used for the chorioallantoic membrane model were stratified for their gender utilizing an in ovo sexing technique.

Results

Immunohistochemistry confirmed stromal and glandular cells in transplanted tissues. Ki67 indicated variable proliferation, while Caspase-3 identified apoptosis. Perfusion increased significantly in 75% of endometriosis samples. Endometrium from a patient with endometriosis showed increased perfusion, contrasting with stable perfusion in healthy endometrium. Higher #Enzian scores partly correlated with increased vascularization.

Summary

The chorioallantoic membrane model is a viable platform for studying endometriosis vascularization and angiogenesis. Endometriosis tissue showed enhanced vascularization influenced by lesion size and anatomical location, offering insights into disease progression and therapeutic strategies.

 

MissingLinkText

Zusammenfassung

Zielsetzung

Endometriose ist eine gynäkologische Erkrankung, die durch endometriumähnliches Gewebe außerhalb der Gebärmutter gekennzeichnet ist. Diese Studie untersucht die Vaskularisation und Proliferation von humanem Endometriose- und Endometriumgewebe, das auf die Chorioallantoismembran (CAM) von Hühnereiern transplantiert wurde, mittels Immunhistochemie und Laser-Speckle-Kontrastanalyse (LASCA). Zur Bewertung der klinischen Relevanz wurde ein Vergleich zwischen experimentellen und klinischen Daten durchgeführt.

Material und Methoden

Gewebeproben von 10 Patientinnen, klassifiziert nach #Enzian-Score und operiert aufgrund von Endometriose, wurden im CAM-Modell untersucht. Zelluläre Strukturen, Proliferation und Apoptose wurden mittels Hämatoxylin-Eosin-Färbung und immunhistochemischer Marker (CD10, Zytokeratin, Ki-67 und Caspase-3) beurteilt. Veränderungen der Durchblutung – als Surrogatmarker für Angiogenese und Vaskularisation – wurden über 3 Tage hinweg mittels LASCA analysiert. Die befruchteten Hühnereier im CAM-Modell wurden mithilfe eines In-ovo-Geschlechtsbestimmungstests hinsichtlich ihres Geschlechts stratifiziert.

Ergebnisse

Die Immunhistochemie bestätigte Stromazellen und Drüsenzellen in den transplantierten Geweben. Ki-67 zeigte eine variable Proliferation, während Caspase-3 Apoptose nachwies. In 75% der Endometrioseproben nahm die Durchblutung signifikant zu. Endometrium einer Patientin mit Endometriose zeigte eine erhöhte Perfusion im Gegensatz zu stabiler Durchblutung in gesundem Endometrium. Höhere #Enzian-Scores korrelierten teilweise mit erhöhter Vaskularisation (p < 0,001).

Schlussfolgerung

Das CAM-Modell stellt eine geeignete Plattform zur Untersuchung der Vaskularisation und Angiogenese bei Endometriose dar. Endometriosegewebe zeigte eine verstärkte Gefäßneubildung, die von der Größe und anatomischen Lokalisation der Läsionen beeinflusst wurde, und liefert somit wertvolle Erkenntnisse über den Krankheitsverlauf und potenzielle Therapieansätze.

 

MissingLinkText

Keywords

CAM model - endometriosis - chorioallantoic membrane - 3D in vivo model - human endometriosis tissue - human tissue

Schlüsselwörter

CAM-Modell - Endometriose - Chorioallantoismembran - 3D-in-vivo-Modell - humanes Endometriosegewebe - humanes Gewebe

Introduction

Endometriosis

Women of reproductive age often suffer from endometriosis, which is defined by endometrial tissue outside of the uterus. It affects around 10% of women worldwide and represents one of the most common benign diseases in gynecology. Some of its effects are pain and infertility and affect the women’s quality of life [1]. There are a few theories that try to explain the pathology of endometriosis but the definite underlying cause is still unknown [2] [3]. This makes its treatment and cure challenging. Endometriosis shows to be a tumor-like disease as it has some characteristics that are defined as the hallmarks of cancer: induction of angiogenesis, activation and invasion of metastasis, to name a few [4] [5]. One of the disease’s characteristics is its dependance on high estrogen levels [6]. The peritoneal fluid of patients with endometriosis consists of different hormones, cells, and proangiogenic factors. Those and their signaling products seem to enhance the spreading of new blood vessels, which lead to endometriotic implant survival [7] [8]. Also, the endometrium has an angiogenic potential and both, endometrium and peritoneal fluid influence lesion formation [9] [10]. In patients, endometriosis is classified using the #Enzian classification that categorizes endometriotic lesions into different groups, describing the depth of infiltration and affected organs, allowing for a more precise assessment of the severity and spread of the disease [11] [12].

 

MissingLinkText

The chorioallantoic membrane model (CAM)

The chorioallantoic membrane (CAM) model, a 3D in vivo model using fertilized chicken eggs, has emerged as a promising platform for translational research in endometriosis. The CAM consists of three layers (ectoderm, mesoderm, and endoderm), with the mesoderm being highly vascularized, creating an ideal environment for tissue engraftment [13] [14]. The immunodeficient nature of the chicken embryo during early developmental stages further supports the growth and invasion of heterologous tissue.

Endometriosis exhibits tumor-like features such as angiogenesis, local invasion, and the formation of metastasis-like lesions, which can be effectively studied using the CAM model. This innovative platform allows for real-time monitoring of engrafted endometriosis tissue, providing a unique opportunity to analyze processes like inflammation, neuronal interactions, and angiogenesis. The CAM model’s versatility also extends to drug testing, offering a bridge between preclinical animal studies and clinical trials [15] [16] [17] [18].

A key advantage of the CAM model is its alignment with the 3R principles – replacement, refinement, and reduction of animal experiments [19]. By utilizing this model, researchers can minimize reliance on traditional animal models, reducing ethical concerns while maintaining robust experimental outcomes. Moreover, the CAM model’s ability to mimic the microenvironment of endometriotic lesions enables detailed investigations into angiogenesis but also biomarker identification.

Aim

This study aims to investigate the role of angiogenesis and vascularization in endometriosis using the CAM model in combination with immunohistochemistry and Laser Speckle Contrast Analysis (LASCA) imaging. We hypothesize that endometriotic lesions induce increased neovascularization, which can be visualized and quantified within the CAM model. Furthermore, we propose that these angiogenic responses correlate with clinical parameters derived from patient data, such as the #Enzian classification. The role of the chicken embryo’s sex on perfusion was analyzed utilizing an in ovo sexing technique. By integrating preclinical findings with patient-derived clinical information, this translational approach aims to deepen our understanding of the vascular component of endometriosis and support the development of targeted therapeutic strategies — all within the framework of ethical and reductionist animal research.

 

MissingLinkText

 

MissingLinkText

 

MissingLinkText

Methods

Patient cohort and data

Human endometrium and endometriosis tissue was obtained from patients who underwent laparoscopic endometriosis surgery at the EuroEndoCert certified Endometriosis Clinic at the St. Marien Hospital Amberg between May 2023 and February 2024. Only patients with intraoperatively and histologically confirmed endometriosis were ultimately included in the study. Initial screening was based on a high clinical suspicion of endometriosis, such as suggestive findings during ultrasound, a medical history consistent with endometriosis-related symptoms, or a previously confirmed diagnosis during earlier surgery. Screening was performed during visits to the endometriosis outpatient clinic. Exclusion criteria included postmenopausal patients and patients without histologically confirmed endometriosis. However, one healthy patient remained in the study to serve as a control, although no clinical or histological lesion was confirmed, as healthy endometrium could be extracted during an already planned hysterectomy. All patients signed an informed consent. The study was approved by the Ethics Committee of the University of Regensburg (23–3211–101, 22–2862–104), ensuring compliance with ethical standards for human tissue research and patient data processing. Patient data were pseudonymized. Data processing and analyses were conducted in accordance with the Declaration of Helsinki and the General Data Protection Regulation of the European Union ([Fig. 1] a, b).

Fig. 1 Methods – Overview. a, b Patient recruitment and tissue retrieval. c Assessment of tissue in the CAM model: sex determination of the chick, perfusion measurements via LASCA, microscope images of engrafted tissue on inoculation and exoculation day, respectively. d Immunohistochemical staining and histological analysis.

 

MissingLinkText

In ovo sex determination

The determination process is based on a patented in ovo sex determination method developed by Prof. Dr. Einspanier (University of Leipzig) [20]. On embryonic developmental day (EDD) 11, the allantoic cavity was accessed by puncturing the CAM, and allantoic fluid was collected ([Fig. 1] c). Then the allantois laboratory kit was used, which works with a 96-well plate (NUNC A/S, Roskilde, Denmark) coated with sheep anti-rabbit IgG (dilution 1 : 100 for E₁S, Technische Universität München, Institute of Physiology, München, Germany). The allantois samples were pipetted into the wells in duplicate. Also, two gender-specific control fluids were used to assign the gender of the individual samples later. The Elisa plates were evaluated by photometric determination at 450 nm wavelength in an Elisa reader (Multiscan FC ThermoScientific) and subsequent evaluation via Microsoft Excel was done. The device took advantage of the hormone level dependent color change that occurs in the presence of the female sex hormone estrone sulfate to determine the sex of the embryo.  The calibration curve and final analysis was performed by Prof. Dr. Einspanier and her group.

 

MissingLinkText

The endometriosis CAM Model

The CAM model was performed as previously described in the established protocol [21]. In brief, fertilized chicken eggs were incubated in a ProCon egg incubator (Grumbach, Asslar, Germany) at 37.8 °C and 63% humidity. To access the CAM, the eggshells were windowed to a size of approximately 1.5–2 cm and sealed with Leukosilk tape. Patient tissue was sliced into approximately 3 × 3 mm pieces and engrafted onto the CAM. LASCA measurements were conducted, and exoculation was performed ([Fig. 1] c).

 

MissingLinkText

Angiogenesis measurements

Changes in blood flow in the CAM were measured as a surrogate for angiogenesis using laser speckle contrast analysis (LASCA) with the PeriCAM perfusion speckle imager (PSI) system, high-resolution (HR) model (PERIMED, Järfälla, Sweden), and Pimsoft software version 1.5.4.8078 (64-bit) by Perimed, as previously described [21] [22] [23]. The LASCA technology enabled real-time visualization of tissue blood perfusion through a small opening in the eggshell, providing insights into angiogenesis in chicken eggs ([Fig. 1] c). Measurements were performed on days 1, 2, and 3, with measurement day 1 conducted after 2 days, measurement day 2 after 4 days, and measurement day 3 after 6 days following tissue engraftment. For accurate results, LASCA was conducted during a low-movement period of the embryo lasting at least 20 seconds, referred to as the time of interest (TOI). A region of interest (ROI) was defined, including only the tissue surrounding the CAM, with a 0.2 mm wide zone around the tissue samples to ensure precise readings. A 550 perfusion units (PU) cutoff filter was applied to exclude areas with minimal or no blood perfusion. LASCA-derived perfusion values are expressed in perfusion units (PU), which are arbitrary units derived from optical speckle contrast analysis. As the method provides relative, not absolute, measurements of blood flow, there is no direct conversion to physiological units such as ml/min. Consequently, PU is the intrinsic and appropriate unit for representing perfusion in LASCA-based assessments.

 

MissingLinkText

Immunohistochemistry

All excised human endometriosis and endometrium samples, along with surrounding CAM tissue, were immediately fixed in 4% buffered formalin for 24 hours and subsequently stored in sodium azide until paraffin embedding. The tissue was then cut into 6 µm sections using a microtome. For histological and immunohistochemical analysis, sections were stained with hematoxylin (Gill No. 3, Sigma-Aldrich, St. Louis, MO, USA) and eosin (Chroma, Waldeck GmbH & Co. KG, Münster, Germany) (H&E staining) or subjected to histochemical antibody staining. Endometrial glands were detected using an anti-cytokeratin (CK) antibody (AE1/AE3 clone anti-mouse from DAKO, Santa Clara, USA), and endometrial stromal cells were identified using a CD10 antibody (CD10–270-L-CE anti-mouse from Leica Biosystems, Nußloch, Germany). For proliferation analysis, Ki67 staining was performed with the ZytoChem Plus (HRP) Anti-Rabbit Kit (Biozol, Eching, Germany) using the rabbit monoclonal Anti-Ki67 antibody [Sp6] (ab16667, Abcam, Cambridge, UK). For the detection of apoptosis, cleaved Caspase-3 (Asp175) antibody #9661 from Cell Signaling Technology, Inc. (Danvers, Massachusetts, USA) was used ([Fig. 1] d). Slides were scanned with the Precipoint M8 and Precipoint Fritz scanner at various magnifications, enabling standardized digital microscopy using Viewpoint Light (PreciPoint, Munich, Germany).

 

MissingLinkText

Statistics

The statistical analysis was performed using International Business Machines Corporation (IBM) SPSS Statistics, version 29.0.0.0 (241) (Armonk, New York, USA), and GraphPad Prism 8, version 8.0.2 (263), by GraphPad Software Inc. (La Jolla, California, USA). As sample sizes for each patient consisted of several CAM models, different statistical tests were conducted depending on data distribution and scale level. For normally distributed, paired data, paired t-tests were used. When normality could not be assumed, the non-parametric Wilcoxon signed-rank test was applied instead. For comparing more than two related samples, the Kruskal-Wallis test for related samples was used. In cases of two independent groups with non-normally distributed data, the Mann-Whitney U test was performed. All tests were two-sided, and p-values lower than 0.05 were considered statistically significant. Furthermore, we evaluated and analyzed the collected data from PeriCam using Microsoft Excel, version 2402, developed by Microsoft Corporation (Redmond, Washington, USA) for chart creation, and GraphPad Prism for diagram generation.

 

MissingLinkText

 

MissingLinkText

Results

Description of patient cohort

Tissue samples were taken from different origins: pelvic wall, ovarian cyst, uterus, peritoneum, mesosalpinx, or endometrium ([Fig. 1] b, [Table 1]). A total number of 10 patients were included in the final analysis, of which nine showed a histologically confirmed endometriosis. One healthy patient served as a control: as surgery was performed due to dysmenorrhea and hypermenorrhea and included a hysterectomy, endometrium tissue of this patient could be extracted. Patients were excluded if endometriosis could not be confirmed histologically. CAM samples were excluded if the tissue failed to adhere (n = 2) or if the movements of the chicks distorted the angiogenesis measurements, resulting in falsely elevated values (n = 4).

Table 1 Patient characteristics (ID 1 to ID 10). Note: For the bivalent categories of the #Enzian classification (such as B, O, T), only a single value (the highest value) was recorded to simplify the analysis. n. s. = not specified.

ID	Tissue origin	#Enzian classification	Hormone therapy	Age at surgery
1	peritoneum, pelvic wall, ovary and ovarian cyst	P3 O2 T2 A1 B3 C0 FA	ethinyl estradiol 0.03 mg, levonorgestrel 0.15 mg	27
2	peritoneum, pelvic wall	P2 O3 T2 A0 B3 C1 FA	n. s.	24
3	peritoneum, pelvic wall, ovary, ovarian cyst and mesosalpinx	P3 O0 T2 A0 B2 C2 FA	no	38
4	ovary, ovarian cyst	P2 O3 T1 A1 B2 C1 FA	n. s.	31
5	peritoneum, pelvic wall	P3 O1 T0 A0 B1 C0 FA	chlormadinone acetate 2 mg	38
6	peritoneum, pelvic wall	P1 O0 T0 A1 B3 C1 FA	no	39
7	ovary, ovarian cyst	P0 O1 T0 A3 B2 C1	no	31
8	peritoneum, pelvic wall	P1 O0 T1 A0 B2 C0	dienogest 2 mg	25
9	Endometrium (healthy)	P0 O0 T0 A0 B0 C0	no	39
10	Endometrium (endometriosis)	P1 O1 T0 A0 B1 C0 FA	n. s.	24

 

MissingLinkText

Immunohistochemical staining

All tissue samples were routinely stained with H&E to enhance the visualization of cellular structures. This staining served as a prescreening method for further immunohistochemical analyses. Various markers were used to specifically identify tissue components, all of which yielded positive results: the CD10 marker successfully identified stromal cells, which were typically localized around the glands in endometrial tissue. Anti-CK primarily stained the epithelial cells of endometrial glands, facilitating the differentiation between epithelial and stromal regions. Ki67, a marker of cellular proliferation, was used to assess cell growth activity. While Ki67 showed little to no proliferation in endometrium, some proliferative activity was observed in endometriosis tissue before and after engraftment on the CAM. Additionally, Caspase-3 staining showed some positive results, revealing areas of necrosis within the endometriosis and endometrium tissue after exoculation. The typical histological differences between endometriosis and endometrium tissue were also seen in the tissue samples after exoculation: numerous glands embedded within abundant stromal tissue for endometrium versus only sporadic glands, with significantly less stromal tissue for endometriosis ([Fig. 2]).

Fig. 2 Immunohistochemical staining for human endometriosis and endometrium samples. Engrafted tissue of endometriosis in the CAM model. a–f endometriotic tissue from patient ID 8 egg 7. g–l endometrium tissue from patient ID 10 egg 25 in the CAM model. a, g Overview image H&E staining. b, h H&E staining. c, i CD10 staining. d, j Caspase-3 staining. The tissue sections analyzed were from the same patient, respectively, but not from the same egg, as the paraffin block was unfortunately completely depleted before this staining method could be established. e, k Ki67 proliferation marker. f, l CK staining. 1 = stromal cells; 2 = gland; * = CAM.

 

MissingLinkText

Increase in angiogenesis over time for endometriosis patients

Depending on data distribution, paired t-tests for ID 1, 2, 5 and 10 or Wilcoxon tests for ID 3, 4, 6 and 7 were applied (Supp. Table S1, Supplementary Material, Online). No tests were performed for ID 8 and 9 as the sample sizes were too small. Across most endometriosis patients (75% of cases), an increase in blood perfusion, hence angiogenesis, over time was observed from day 1 to day 3 ([Fig. 3] a–g1). For ID 1, a significant increase for PU was observed between day 1 and day 3 (p < 0.001). ID 2 showed a significant increase between day 2 and day 3 (p = 0.016). In contrast, no significant differences were detected for ID 3 across any of the measurement days. For ID 4, a highly significant increase was observed between all measurement days (p < 0.001 for all comparisons). ID 5 showed a significant increase between day 1 and day 2 (p = 0.015). ID 6 exhibited significant increases between day 1 and day 3 (p = 0.026) and between day 2 and day 3 (p = 0.008). Finally, ID 7 demonstrated significant increases between day 1 and day 2 (p = 0.036) and between day 1 and day 3 (p = 0.025). ID 8 showed no visible consistent differences between all measurement days.

Fig. 3 Angiogenesis measurements for endometriosis samples. a–g1: Representation of blood perfusion values (PU) on measurement days 1 to 3 for various patients (ID 1 to ID 7). g2: Individual blood perfusion trends for 2 out of 10 eggs of ID 7 over days 1 to 3, represented by colored points (red for egg 9, blue for egg 21). h: Summary of the average blood perfusion values (PU) for all measured eggs inoculated with endometriosis tissue (n = 109, ID 1 to ID 8) on days 1 to 3. g3 and g4 (top row): LASCA images showing blood perfusion for egg 10 (red, Fig. g3) and egg 11 (blue, Fig. g4) on days 1 to 3. The perfusion images visualize regional differences in perfusion. g3 and g4 (bottom row): The same images as the top row, with a PU filter of 550 nm applied to cancel out areas with low to no perfusion. The black circle in the image represents the ROI. Each bar represents the mean PU value with standard error of the mean (SEM). n = number of eggs measured for each patient. * p < 0.05, ** p < 0.01 and *** p < 0.001. PU = perfusion unit.

The overall average perfusion for all endometriosis IDs and all measured eggs (n = 109, ID 1 to ID 8) over the three days increased ([Fig. 3] h). The mean perfusion values increased progressively across the days, with an average of 623.8 perfusion units (PU) on day 1, 644.2 PU on day 2, and 668.3 PU on day 3 (day 1 vs. day 2 p < 0.001, day 2 vs. day 3 p < 0.001, day 1 vs. day 3 p < 0.001).

When analyzed separately for each patient, the trends in perfusion values of the individual eggs varied ([Fig. 3] g2–4).

 

MissingLinkText

Increase in blood flow over time for endometrium of endometriosis patients

To investigate differences in blood perfusion of endometrium tissue, LASCA measurements were performed on days 1 to 3 for a patient with and without endometriosis, respectively. For endometrium of a healthy patient (ID 9) no visible consistent changes were observed in perfusion levels in the CAM model over time. In contrast, a significant increase in perfusion was observed for the endometrium sample of an endometriosis patient (ID 10): between day 1 and day 2 (p = 0.004) ([Fig. 4], Supp. Table S1). The paired t-test was used to compare measurements across the three days for IDs 9 and 10.

Fig. 4 Perfusion measurements for endometrium samples: a PU values measured over three days of an endometrium sample of a healthy patient (ID 9) without endometriosis. b PU values measured over three days in the endometrium of a patient (ID 10) with endometriosis, classified with an #Enzian Score of P1 O1 T0 A0 B1 C0 FA. Each bar represents the mean PU value with SEM. n = number of eggs measured for each patient. * p < 0.05, ** p < 0.01 and *** p < 0.001. PU = perfusion unit.

 

MissingLinkText

The role of the chicken embryo’s sex

To evaluate whether the sex of the chicken embryos influenced angiogenesis measurements, sex determination for the eggs of ID 5, 6 and 7 was performed. The significance levels of blood perfusion measurements for day 1, day 2, and day 3 were analyzed. Across all three experimental runs, a total of n = 38 eggs were investigated, of which 16 were male and the remaining were female. The Mann-Whitney U-test revealed no significant differences between sexes for blood perfusion measurements on day 1 (p = 0.866), day 2 (p = 0.651), and day 3 (p = 0.988).

 

MissingLinkText

The role of the #Enzian classification on angiogenesis in the CAM model

The analysis focused on the relationship between angiogenesis measurements via LASCA over three days (day 1–3) and the #Enzian classification categories of the individual patients (ID 1 to ID 8 and ID 10), particularly the categories A (vagina, rectocervical and rectovaginal), O (ovary), P (peritoneum), and T (tubal/ ovarian), B (uterosacral ligament, cardinal ligaments, and pelvic sidewall) and C (rectum) and FA (adenomyosis uteri) ([Fig. 5]).

Fig. 5 Angiogenesis measurements results for #Enzian categories. The blood perfusion from days 1 to 3 (top to bottom) is shown for the following #Enzian classification categories: A in panel a, O in panel b, P in panel c, and T in panel d. y-axis = PU; x-axis = extension of disease; box plots with mean (thick black line) and standard deviation. *** p < 0.001, ** p < 0.01, * p < 0.05. n = 123.

The relationships were evaluated using Kruskal-Wallis tests for comparisons between more than two groups. In category A (vagina and cervix), an inverse relationship was observed, where higher #Enzian scores correlated with lower perfusion values on days 1 to 3. For example, lesions larger than 3 cm (A3) were associated with a lower perfusion compared to cases without lesions (A0). The statistical significance of this relationship was as follows: day 1: p < 0.001, day 2: p = 0.021 and day 3: p = 0.017. In category O (ovary), no consistent effects on angiogenesis measurements were found. Significant effects on angiogenesis measurements were identified in this category as follows: day 1: p < 0.001 and day 2: p = 0.033. In contrast, higher #Enzian scores in category P (peritoneum) significantly correlated with increased perfusion values on all three days, indicating greater perfusion as lesion size increased (day 1: p < 0.001, day 2: p = 0.025, day 3: p = 0.020), demonstrating strong variation in perfusion across the peritoneal extension categories. For category T (tube), the presence of adhesions involving the pelvic sidewall and uterus (T2) was associated with significantly higher perfusion values on day 1 to 3 compared to cases without adhesions (T0): day 1: p < 0.001, day 2: p < 0.001 and day 3: p = 0.034. In category C, statistical significance was observed on day 1 with p < 0.001, while on day 2 and 3, there was no significance. Still, no clear pattern was seen. In category B, no statistical significance was detected across all days. For category FA, significant results were found on day 1 (p = 0.010) and day 3 (p = 0.040), showing no clear pattern. In summary, for #Enzian categories B, C, O, and FA, the perfusion values fluctuated between the different classes, showing neither a correlation nor an inverse trend compared to categories A (inverse correlation), P, and T (positive correlation), respectively (Supp. Table S2, Supplementary Material, Online).

 

MissingLinkText

 

MissingLinkText

Discussion

This study is among the first to investigate human endometriosis tissue in the CAM model – to our knowledge this has been performed by only four other groups. Liu et al. focused on isolated endometrial stem cells from human endometriosis lesions in the CAM model, measuring angiogenesis by evaluating the vascular area and branch count [24]. Pluchino et al. investigated endometriosis tissue derived from human ovarian endometriomas in the CAM model, applying local treatments with testosterone or anastrozole and measuring lesion growth [25]. Wang et al. and Ria et al. also studied tissue from human ovarian endometriomas, with Wang et al. treating lesions locally with estradiol or puerarin and assessing angiogenesis by manually counting vessels and calculating an angiogenic index, while Ria et al. analyzed angiogenesis using microscopic vessel counting after immunohistochemical staining [26] [27]. In contrast to these studies, our research analyzed tissue from a variety of endometriosis lesions. Additionally, our method of measuring angiogenesis using LASCA provides the advantage of dynamic, real-time measurements. This approach also allowed for multiple assessments over the course of three separate days. In other prior endometriosis studies, the CAM model has primarily been used to examine endometrial tissue, which was subsequently classified as endometriosis tissue following engraftment onto the CAM [28] [29] [30] [31] [32] [33] [34] [35] [36].

Immunohistochemical staining

The analysis of human endometriosis tissue poses a challenge, as endometrial glands are often only scarcely present compared to endometrium tissue and surrounded by scarring tissue. Nevertheless, in the above study, the endometriosis tissue in the CAM model demonstrated vitality, as evidenced by positive Ki67 staining, confirming cellular activity, and only partly positive Caspase-3 staining, a marker for necrosis. We observed more Ki67 positive cells in endometriosis tissue than in endometrium, although no clear changes were seen after inoculation in the CAM model. This was also observed by Nap et al., who found that Ki67 does not accurately reflect the ability to form lesions in the CAM model [37]. The successful implantation of viable endometriosis tissue (glands and endometrial stroma) onto the CAM was confirmed via CK and CD10 staining.

 

MissingLinkText

Angiogenesis measurements

A key observation was the increase in blood perfusion in endometriosis tissue throughout the CAM experiments, indicative of active angiogenesis. When analyzed individually, 75% of all patients showed an increase in blood perfusion, underscoring significant interindividual differences. These variations could be influenced by factors such as patient age, disease severity, the age of the endometriotic lesions, phase of the menstrual cycle, different tissue origin or differences in proangiogenic factors within the individual tissues. However, no definitive explanation for these differences could be identified, highlighting the need for further investigation and consideration of the above-mentioned factors in future experiments. The comparison between healthy and endometriotic endometrium revealed significant differences. While healthy endometrium showed no visible consistent changes in perfusion, the diseased endometrium exhibited a notable increase. However, these observations are limited by the small sample size and should be interpreted with caution. On the contrary, Gescher et al. observed no differences in vessel density index analysis or vascular epithelial growth factor (VEGF) mRNA expression between the endometrium of endometriosis patients and that of non-endometriosis patients when incubated on the CAM for varying durations. This suggests that factors other than the eutopic endometrium may contribute to the angiogenesis of endometriosis lesions [30].

 

MissingLinkText

The role of the chicken embryo’s sex

Another focus of the study was the potential influence of the chicken embryo’s sex on blood perfusion, as endometriosis is highly hormone-dependent and influenced by differences in estrogen receptors. Female chicken embryos exhibit higher estrone sulfate levels, leading to the hypothesis that blood perfusion might be gender dependent. However, the results showed no significant differences between male and female embryos, suggesting that perfusion in the CAM model is not primarily hormone-dependent and that the sex of the chicken embryos does not affect the degree of blood perfusion. Hence, a gender stratification does not seem mandatory for future angiogenesis investigations in the endometriosis CAM model. Still, in ovo sexing could be relevant for analyzing endometriosis tissue growth in the CAM model, as Pluchino et al. observed gender-specific differences in the growth of endometriosis lesions [25].

 

MissingLinkText

The role of the #Enzian classification on angiogenesis in the CAM model

Noteworthy was the analysis of perfusion across different categories of the #Enzian classification. A positive correlation was seen in perfusion and the #Enzian categories P (peritoneum) and T (tube): higher classifications in both categories resulted in higher perfusion values for all measurement days, indicating robust angiogenesis. The higher the extent of endometriosis at the anatomical sites of the peritoneum or the peritubarian region, the higher the angiogenesis. Paradoxically, category A, when correlated with the angiogenesis measurements, demonstrated an inverse effect. The lower the extent of endometriosis in the rectovaginal space and adjacent areas, the higher the induced angiogenesis in the CAM model. This effect might be possible due to specific tissue characteristics or local factors. Categories O (ovary), C (rectum), and FA (adenomyosis uteri) showed inconsistent results without clear trends, while category B (uterine ligaments) displayed no significant differences. These findings should be interpreted cautiously and warrant further investigation as not all subgroups of the #Enzian classification (such as A2, B0 and C3) were represented. The observed correlation between the clinical extent of endometriosis and perfusion changes in the CAM model can be explained by the presence of proangiogenic factors. In categories P and T, the larger volume of affected tissue likely is accompanied by higher local levels of VEGF and other angiogenic mediators, promoting increased perfusion. In contrast, the inverse relationship observed in category A may result from specific tissue properties, which could suppress angiogenesis. The lack of clear trends in other categories might be due to variability in tissue composition or local microenvironmental factors.

 

MissingLinkText

Angiogenesis in the CAM model

Endometriosis induced angiogenesis in the CAM model might be explained by the tissue accompanying human angiogenic factors such as VEGF or angiopoietin 2 found in the microenvironment of the tissue. Nap et al. demonstrated that human anti-VEGF effectively inhibited neovascularization in the CAM model, but only when angiogenesis was stimulated by the engraftment of endometrial tissue. They hypothesized that this effect was driven by the presence of accompanying human VEGF. In contrast, no impact on vessel growth was observed when the human anti-VEGF antibody was applied to control CAMs without human tissue [34]. It has also been suggested that tissue engraftment onto the CAM initially places the tissue in a hypoxic state. Since hypoxia is a key driver of angiogenic factor expression, this condition stimulates the tissue to produce higher levels of angiogenic factors. This, in turn, promotes increased neovascularization and facilitates the integration of the engrafted tissue, enabling it to connect with the chick’s vascular system [31].

Angiogenesis plays a crucial role in endometriosis and is driven by proangiogenic factors. These mechanisms explain the increase in perfusion and support the hypothesis that endometriosis and tumors share similar angiogenic processes as angiogenesis has been identified as a hallmark of cancer [7] [8] [38]. The CAM model is a versatile 3D-in-vivo-model, widely acknowledged for its ability to assess tissue growth, metastasis, invasion, and angiogenesis, as well as to test potential drugs and substances [39]. Its unique structure and accessibility make it particularly suited for studying angiogenesis dynamics [38]. Notably, the CAM model facilitates precise evaluations of angiogenesis and can be used to test the effects of angiogenesis-modulating substances, a method that has also been employed by others [40].

 

MissingLinkText

Limitations and strengths of the study

Although angiogenesis was evaluated functionally via perfusion measurements, the molecular mechanisms driving neovascularization – such as VEGF expression or hypoxia-induced pathways – were not directly quantified in this study. No molecular assays were performed to directly assess their involvement in the CAM model. Future studies should integrate molecular analyses to better elucidate the mechanistic basis of angiogenesis in endometriosis.

A major strength of this study is the application of real-time, dynamic perfusion measurement using LASCA over multiple time points, which offers a functional perspective on angiogenesis in viable human endometriosis tissue. Furthermore, the comparison across a range of lesion localizations using the #Enzian classification adds clinical relevance. However, the limited sample size and the exclusion of direct molecular profiling represent methodological constraints. Additionally, variability in tissue origin and quality may introduce heterogeneity that is difficult to control. A key limitation of this study is the absence of data on the menstrual cycle phase of the included patients. To enhance experimental standardization in future research, it is recommended to include only patients in a defined phase of the menstrual cycle – such as the proliferative phase – and to exclude those using hormonal contraceptives, as their use may represent a potential confounding factor. Another limitation is that, for the bivalent categories of the #Enzian classification (such as B, O, T), only a single value was recorded, which may reduce the resolution of subgroup analyses.

 

MissingLinkText

 

MissingLinkText

Conclusion

This study demonstrates that higher #Enzian scores correlate with increased blood perfusion in categories P and T, while an inverse relationship was observed for category A. The gender of the chicken embryo had no impact on outcomes, but blood flow increased over time in 75% of endometriosis cases, contrasting with stable perfusion in healthy tissue. This study is limited by its small sample size, which restricts the generalizability of the findings. Furthermore, the lack of long-term follow-up data prevents conclusions regarding sustained angiogenic activity or tissue viability beyond the observation window. The absence of molecular analyses also limits mechanistic insight into the angiogenic processes observed. An additional limitation is that for bivalent categories of the #Enzian classification, only a single value per case was recorded, which may reduce accuracy in reflecting multifocal lesion distribution. These limitations underscore the need for further studies with larger cohorts, and the inclusion of molecular endpoints. By successfully analyzing angiogenesis in human endometriotic tissue using the CAM model, this work lays the foundation for future studies with larger sample sizes to confirm these findings, explore further clinical correlations, and extend investigations, for example, with angiogenesis-modulating substances.

 

MissingLinkText

Supplementary Material

• Supplementary Table S1: Angiogenesis measurements results for all IDs. D1–3 = Measurement Day 1–3, σ = standard error, PU = perfusion units.

• Supplementary Table S2: Angiogenesis measurements results for #Enzian classification categories. D1–3 = Measurement Day 1–3, σ = standard error, PU = perfusion units.

 

MissingLinkText

 

MissingLinkText

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgement

We would like to thank the BayWISS (Bayerisches Wissenschaftsforum) for their financial support. Our sincere gratitude goes to Prof. Dr. Einspanier from the University of Leipzig, Germany, for assisting with the sex determination analyses. We also extend our heartfelt thanks to Dr. Donutiu and Dr. Tsaousidis for their invaluable efforts in providing endometriosis tissue samples from the EuroEndoCert certified clinic at the St. Marien Hospital Amberg. Special recognition is given to Lucia Hoffmann for her exceptional technical support. Furthermore, we express our gratitude to Dr. Weber from the Department of Pathology in Regensburg for his contributions to the Caspase-3 staining. We would also like to thank Dr. Gerken from Tumorzentrum Regensburg, Germany, for reviewing the statistical analysis and for his valuable corrections, suggestions, and statistical expertise.

• References

• 1 Zondervan KT, Becker CM, Koga K. et al. Endometriosis. Nat Rev Dis Primers 2018; 4: 9
• 2 D’Hooghe TM. Invisible microscopic endometriosis: how wrong is the sampson hypothesis of retrograde menstruation to explain the pathogenesis of endometriosis?. Gynecol Obstet Invest 2003; 55: 61-62
• 3 Zondervan KT, Becker CM, Missmer SA. Endometriosis. N Engl J Med 2020; 382: 1244-1256
• 4 Thomas EJ, Campbell IG. Molecular genetic defects in endometriosis. Gynecol Obstet Invest 2000; 50 (Suppl. 1) 44-50
• 5 Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011; 144: 646-674
• 6 Chantalat E, Valera M-C, Vaysse C. et al. Estrogen Receptors and Endometriosis. Int J Mol Sci 2020; 21: 2815
• 7 Rocha ALL, Reis FM, Taylor RN. Angiogenesis and endometriosis. Obstet Gynecol Int 2013; 2013: 859619
• 8 May K, Becker CM. Endometriosis and angiogenesis. Minerva Ginecol 2008; 60: 245-254
• 9 Taylor RN, Lebovic DI, Mueller MD. Angiogenic factors in endometriosis. Ann N Y Acad Sci 2002; 955: 89-100 discussion 118, 396–406
• 10 Taylor RN, Lundeen SG, Giudice LC. Emerging role of genomics in endometriosis research. Fertil Steril 2002; 78: 694-698
• 11 Dirou C, Fondin M, Le Pabic E. et al. Association of preoperative Enzian score with postoperative fertility in patients with deep pelvic endometriosis. J Gynecol Obstet Hum Reprod 2022; 51: 102408
• 12 Tuttlies F, Keckstein J, Ulrich U. et al. ENZIAN-score, a classification of deep infiltrating endometriosis. Zentralbl Gynakol 2005; 127: 275-281
• 13 Ribatti D, Tamma R, Annese T. Chorioallantoic membrane vascularization. A meta-analysis. Exp Cell Res 2021; 405: 112716
• 14 Schneider-Stock R, Ribatti D. The CAM Assay as an Alternative In Vivo Model for Drug Testing. Handb Exp Pharmacol 2021; 265: 303-323
• 15 Kue CS, Tan KY, Lam ML. et al. Chick embryo chorioallantoic membrane (CAM): an alternative predictive model in acute toxicological studies for anti-cancer drugs. Exp Anim 2015; 64: 129-138
• 16 Phumat P, Chaichit S, Potprommanee S. et al. Influence of Benincasa hispida Peel Extracts on Antioxidant and Anti-Aging Activities, including Molecular Docking Simulation. Foods 2023; 12: 3555
• 17 Rao Q, Yu H, Li R. et al. Dihydroartemisinin inhibits angiogenesis in breast cancer via regulating VEGF and MMP-2/-9. Fundam Clin Pharmacol 2024; 38: 113-125
• 18 Thomas EJ, Campbell IG. Evidence that endometriosis behaves in a malignant manner. Gynecol Obstet Invest 2000; 50 (Suppl. 1) 2-10
• 19 Sneddon LU, Halsey LG, Bury NR. Considering aspects of the 3Rs principles within experimental animal biology. J Exp Biol 2017; 220: 3007-3016
• 20 Weissmann A, Reitemeier S, Hahn A. et al. Sexing domestic chicken before hatch: a new method for in ovo gender identification. Theriogenology 2013; 80: 199-205
• 21 Kuri PM, Pion E, Mahl L. et al. Deep Learning-Based Image Analysis for the Quantification of Tumor-Induced Angiogenesis in the 3D In Vivo Tumor Model-Establishment and Addition to Laser Speckle Contrast Imaging (LSCI). Cells 2022; 11: 2321
• 22 Pion E, Haerteis S, Aung T. Application of Laser Speckle Contrast Imaging (LSCI) for the Angiogenesis Measurement of Tumors in the Chorioallantoic Membrane (CAM) Model. Methods Mol Biol 2023; 2572: 141-153
• 23 Pion E, Asam C, Feder A-L. et al. Laser speckle contrast analysis (LASCA) technology for the semiquantitative measurement of angiogenesis in in-ovo-tumor-model. Microvasc Res 2021; 133: 104072
• 24 Liu Y, Liang S, Yang F. et al. Biological characteristics of endometriotic mesenchymal stem cells isolated from ectopic lesions of patients with endometriosis. Stem Cell Res Ther 2020; 11: 346
• 25 Pluchino N, Poppi G, Yart L. et al. Effect of local aromatase inhibition in endometriosis using a new chick embryo chorioallantoic membrane model. J Cell Mol Med 2019; 23: 5808-5812
• 26 Ria R, Loverro G, Vacca A. et al. Angiogenesis extent and expression of matrix metalloproteinase-2 and -9 agree with progression of ovarian endometriomas. Eur J Clin Invest 2002; 32: 199-206
• 27 Wang D, Liu Y, Han J. et al. Puerarin suppresses invasion and vascularization of endometriosis tissue stimulated by 17β-estradiol. PLoS One 2011; 6: e25011
• 28 Juhasz-Böss I, Hofele A, Lattrich C. et al. Matrix metalloproteinase messenger RNA expression in human endometriosis grafts cultured on a chicken chorioallantoic membrane. Fertil Steril 2010; 94: 40-45
• 29 Nap AW, Dunselman GAJ, de Goeij AF. et al. Inhibiting MMP activity prevents the development of endometriosis in the chicken chorioallantoic membrane model. Hum Reprod 2004; 19: 2180-2187
• 30 Gescher DM, Siggelkow W, Meyhoefer-Malik A. et al. A priori implantation potential does not differ in eutopic endometrium of patients with and without endometriosis. Arch Gynecol Obstet 2005; 272: 117-123
• 31 Drenkhahn M, Gescher DM, Wolber EM. et al. Expression of angiopoietin 1 and 2 in ectopic endometrium on the chicken chorioallantoic membrane. Fertil Steril 2004; 81: 869-875
• 32 Kressin P, Wolber EM, Wodrich H. et al. Vascular endothelial growth factor mRNA in eutopic and ectopic endometrium. Fertil Steril 2001; 76: 1220-1224
• 33 Ren Q, Qian Z, Jia S. et al. Vascular endothelial growth factor expression up-regulated by endometrial ischemia in secretory phase plays an important role in endometriosis. Fertil Steril 2011; 95: 2687-2689
• 34 Nap AW, Dunselman GAJ, Griffioen AW. et al. Angiostatic agents prevent the development of endometriosis-like lesions in the chicken chorioallantoic membrane. Fertil Steril 2005; 83: 793-795
• 35 Maas JW, Calhaz-Jorge C, ter Riet G. et al. Tumor necrosis factor-α but not interleukin-1β or interleukin-8 concentrations correlate with angiogenic activity of peritoneal fluid from patients with minimal to mild endometriosis. Fertil Steril 2001; 75: 180-185
• 36 Oosterlynck DJ, Meuleman C, Sobis H. et al. Angiogenic activity of peritoneal fluid from women with endometriosis. Fertil Steril 1993; 59: 778-782
• 37 Nap AW, Groothuis PG, Demir AY. et al. Tissue integrity is essential for ectopic implantation of human endometrium in the chicken chorioallantoic membrane. Hum Reprod 2003; 18: 30-34
• 38 Storgard C, Mikolon D, Stupack DG. Angiogenesis assays in the chick CAM. Methods Mol Biol 2005; 294: 123-136
• 39 Ribatti D. The chick embryo chorioallantoic membrane (CAM) assay. Reprod Toxicol 2017; 70: 97-101
• 40 Ribatti D. The chick embryo chorioallantoic membrane in the study of tumor angiogenesis. Rom J Morphol Embryol 2008; 49: 131-135

Publication History

Received: 19 January 2025

Accepted after revision: 02 July 2025

Article published online:
22 July 2025