Here, a digital immunohistochemistry image analysis platform was developed and validated to quantitatively analyze the endometrial immune cells of patients with recurrent miscarriages in the window of implantation.
To evaluate the endometrial immune microenvironment of patients with recurrent miscarriage (RM), a digital immunohistochemistry image analysis platform was developed and validated to quantitatively analyze endometrial immune cells during the mid-luteal phase. All endometrium samples were collected during the mid-luteal phase of the menstrual cycle. Paraffin-embedded endometrial tissues were sectioned into 4 µm thick slides, and immunohistochemistry (IHC)staining was carried out for detecting endometrial immune cells, including CD56+ uNK cells, Foxp3+ Tregs, CD163+ M2 macrophages, CD1a+ DCs, and CD8+ T cells. The panoramic slides were scanned using a digital slide scanner and a commercial image analysis system was used for quantitative analysis. The percentage of endometrial immune cells was calculated by dividing the number of immune cells in the total endometrial cells. Using the commercial image analysis system, quantitative evaluation of endometrial immune cells, which are difficult or impossible to analyze with conventional image analysis, could be easily, and accurately analyzed. This methodology can be applied to quantitatively characterize the endometrium microenvironment, including interaction between immune cells, and its heterogeneity for different reproductive failure patients. The platform for quantitative evaluation of endometrial immune cells may be of important clinical significance for the diagnosis and treatment of RM patients.
Recurrent miscarriage (RM) is the loss of two or more consecutive pregnancies and is a complex disease drawing attention from clinicians in recent years.The incidence rate of RM in women of childbearing age is 1%-5% 1. Results of previous studies show that immune factors are closely associated with the pathogenesis of RM2,3,4,5. Maintaining immune homeostasis at the maternal-fetal interface is required for embryo implantation and development. Endometrial immune cells perform several regulatory roles to maintain this homeostasis, such as promoting trophoblast invasion, remodeling spiral arteries, and contributing to placenta development6,7,8,9.
Aberrant endometrial immune cells in women with RM have previously been reported. Results show a close association between the high density of uterine natural killer cells (uNKs) and the occurrence of RM10,11,12. An increased number of macrophages has been reported in the endometrium of women with RM, compared with those who had a live birth13. Regulatory T cells (Treg) play a role in maternal immune tolerance toward the embryo, and their level and function are decreased in the decidua of RM patients14. Cytotoxicity T cells (CTL) and dendritic cells (DCs) also play a role in the immune regulation of pregnancy15,16. Therefore, a comprehensive quantitative analysis of local endometrial immune cells during the mid-luteal phase could help to better understand the pathogenesis of RM. Some current methods for quantitative analysis of endometrial immune cells use flow cytometry which can accurately label immune cells with multiple markers17,18. However, clinical application of flow cytometry is limited because it can only be performed on fresh tissue. Obtaining fresh tissue is only feasible when a large volume of excess tumor is available, a rare occurrence for endometrium. Immunohistochemistry can observe tissue morphology well in situ and can also label various immune cells, while traditional immunohistochemical techniques cannot perform quantitative analysis of immune cells.
Compared to conventional immunohistochemistry experiments, quantitative immunohistochemical analysis of immune cells in the endometrium has important clinical significance. IHC intensity scoring is usually ranked on a four-point scale or strong and weak in pathological diagnostics and research19,20,21. However, this semi-quantitative technique is subjective, highly inaccurate, and demonstrates significant intra-observer and inter-observer variability22. One possible solution is the application of machine learning, which is valuable indigital image analysis23,24. By providing quantitative measurements, this approach enables a more precise assessment of immune cell infiltration, distribution, and density within the uterine tissue. This quantitative information can help to elucidate the dynamic changes in immune cell populations during the menstrual cycle and in various pathological conditions. Overall, the ability to quantitatively analyze immune cells in the endometrium through immunohistochemistry offers valuable insights into the immune microenvironment of the uterus.
Therefore, the protocol aimed to developed and validated a digital immunohistochemistry image analysis platform to quantitatively analyze endometrial immune cells including uNK cells, Tregs, macrophages, DCs and cytotoxic T cells during the mid-luteal phase in RM patients.
The research content and protocol has been ethically reviewed and approved by the research ethics committee of Shenzhen Zhongshan Urology Hospital. All women (20- 40 years old) involved in the study provided informed consent for sample collection and usage.
1. Acquisition of pathological tissue
2. Tissue dehydration
3. Tissue embedding
4. Tissue sections
5. Immunohistochemical staining
6. Dehydration and sealing of the slide
7. Scanning slide
8. Analysis of images
In order to evaluate endometrial immune cells quantitatively and reduce the instability caused by man-made operational mistakes, we established a digital quantitative analysis platform for endometrial immune cells by using automatic immunohistochemical detection and digital quantitative evaluation system. Immunohistochemistry image analysis platform was established to quantitatively analyze endometrial immune cells of patients with recurrent miscarriage (RM) in the window of implantation. All endometrium tissues were collected during the mid-luteal phase of the menstrual cycle. Paraffin-embedded endometrial tissues were sectioned into 4 µm thick slides, and IHC staining was carried out for detecting endometrial immune cells, including CD56+uNK cells, Foxp3+Tregs, CD163+M2 macrophages, CD1a+ DCs, and CD8+ T cells. The panoramic slides were scanned using the digital slide scanner, and quantitative analysis was performed using a digital image analysis system. For calculating the percentage of endometrial immune cells, divide the number of immune cells by the total number of endometrial cells. (Figure 1 and Figure 2). The proportion of different endometrial immune cells in women with RM (N=30) is shown in Table 2.
Figure 1: Schematic workflow for the analysis of mid-luteal endometrial immune cells. Collection of samples was done during the mid-luteal phase of the menstrual cycle, at 7-9 days after theluteinizing hormone (LH) surge. Fixation was done in 10% neutral buffered formalin for 4-6 h at room temperature, and then embedded in paraffin wax. IHC staining was performed for detection of immune cells in the endometrium during the mid-luteal phase, including CD56+uNK cells, Foxp3+Tregs, CD68+ macrophages, CD163+M2 macrophages, CD1a+DCs and CD8+T cells25,26. Tissues were scanned using a commercial scanner and quantitative analysis was performed using the immunohistochemistry image analysis system. Please click here to view a larger version of this figure.
Figure 2: Immunostaining of endometrial cells for identification of positive immune cells. Immunostaining of CD56+uNK cells, Foxp3+Tregs, CD68+ macrophages, CD163+M2 macrophages, CD1a+DCs and CD8+T cells in endometrium from RM patients. Brown represents positive immune cells and blue represents the nucleus. Please click here to view a larger version of this figure.
Primary antibody | Clone | Dilution |
CD56 | 123C3 | 1/800 |
Foxp3 | 236A/E7 | 1/100 |
CD163 | 10D6 | 1/1200 |
CD1a | 10 | 1/200 |
CD8 | 4B11 | 1/300 |
Table 1: Primary antibodies used during immunohistochemical staining. The table shows the clone and dilution of the primary antibodies.
Immune markers | Median (%) | Minimum (%) | Maximum (%) | Mean (%) | 5th percentile (%) | 95th percentile (%) |
CD56 | 4.83 | 1.8 | 16.76 | 6.03 | 2.04 | 13.63 |
Foxp3 | 0.05 | 0.02 | 0.12 | 0.06 | 0.02 | 0.11 |
CD68 | 0.78 | 0.37 | 3.62 | 0.99 | 0.44 | 2.67 |
CD163 | 0.84 | 0.35 | 2.38 | 0.93 | 0.38 | 2.13 |
CD1a | 0.04 | 0.01 | 0.11 | 0.05 | 0.01 | 0.1 |
CD8 | 1.69 | 0.76 | 4.1 | 1.85 | 0.81 | 3.72 |
Table 2: Percentage of various endometrial immune cells in women with RM. The percentage of CD56+uNK cells, Foxp3+Tregs, CD68+ macrophages, CD163+M2 macrophages, CD1a+DCs and CD8+T cells in RM patients was calculated by immunohistochemistry image analysis platform.
This protocol established a digital immunohistochemistry image analysis platform to quantitatively analyze endometrial immune cells of RM patients. Here, six endometrial immune markers were detected to evaluate the endometrial immune microenvironment in RM patients.
A receptive endometrium during the mid-luteal phase is key for successful implantation and pregnancy27,28. Therefore, evaluation of percent endometrial immune cells plays an important role in estimating endometrial receptivity. Endometrial immune cells analysis, by conventional pathologic methods, is predictive so it does not help in clinical application. At present, there is no standardized method for the measurement of immune cell percentage which poses a barrier to understanding the role of these cells in pregnancy. Conventional IHC analysis is based on selected visual fields and manual counting. Precise tissue segmentation and localization of immune cells cannot be evaluated with conventional IHC because manual analysis is subjective leading to observer bias29.
Compared to analysis of the distribution of immune cells in the selected field, panoramic analysis of tissues is more accurate for analyzing the distribution of endometrial immune cells. Digital pathology approaches that utilize machine-based learning algorithms have been tested in multiple tumors to evaluate large tissue areas and complex cell phenotypes30,31. In the present study, a commercial system was introduced for obtaining a panoramic image of endometrium samples. The percentage of endometrial immune cells was subsequently determined using an automated quantitative analysis based on the immunohistochemistry image analysis system.
The immunohistochemistry image analysis system used here is an image analysis platform specialized for pathological tissues, which enables tissue segmentation by using artificial intelligence. Many studies using various modules with this system have been reported32,33,34. The system was used to quantify various histopathological changes and findings that were difficult to analyze using conventional image processing software. For example, the number of CD56+NK cells account for about 5% of the total cells in the endometrium during the mid-luteal phase. Traditional immunohistochemical analysis is difficult to accurately calculate the number of CD56+ NK cells, and it may take at least 30 min to calculate the total number of positive cells on the whole section, but it takes 2-3 min with the system used here. Therefore, using the immunohistochemistry image analysis system used here, endometrial immune cells could be easily, and accurately analyzed. Quantification of pathological findings by image analysis can contribute to improving objectivity, precision, and persuasiveness of endometrial immune environment evaluation.
However, there is a limitation of the described method. Endometrium consists of various immunocytes that have an association with pregnancy outcome. Therefore, defining only one or two immune markers might be insufficient. Hence, multiparametric approaches are needed to comprehensively assess immune profiling of cells.
In summary, it can be concluded that this protocol has successfully applied digital image analysis in endometrial sections during mid-luteal phase of RM patients for the quantification of immune cells and panoramic analysis was conducted in the present study to determine the distribution of endometrial CD56+ uNKs, Foxp3+Tregs, CD68+ macrophages, CD163+ M2 macrophages, CD1a+ iDCs and CD8+T cells during the mid-luteal phase. More evidence is required to support the association between endometrial immune cells with pregnancy outcomes of patients with RM and future studies will focus on this. This was the first study that investigated endometrial immune environment in RM patients using digital pathology.
The authors have nothing to disclose.
The authors are grateful to all women who consented and donated samples for this study.
Automated coverslipper | Sakuraus | DRS-Prisma-P-JCS&Film-JC2 | |
CD163 | GrowGn Biotechnology | NCL-L-CD163 | |
CD1a | Gene Tech | GM357129 | |
CD56 | Gene Tech | GT200529 | |
CD8 | Novocastra | NCL-L-CD8-4B11 | |
Dehydrator | Thermo Fisher | Excelsior ES | |
Digital pathology and | Indica labs | HALO | |
Foxp3 | YILIFANG biological | 14-477-82 | |
IHC stainer | Leica | BOND III | |
Image analysis platform | Indica labs | HALO | |
Slide Scanner | Olympus life science | VS200 |