Increased Angiogenesis and Lymphangiogenesis in Adenomyosis Visualized by Multiplex Immunohistochemistry

Abstract

There is evidence for increased angiogenesis in the (ectopic) endometrium of adenomyosis patients under the influence of vascular endothelial growth factor (VEGF). VEGF stimulates both angiogenesis and lymph-angiogenesis. However, information on lymph vessels in the (ectopic) endometrium of adenomyosis patients is lacking. In this retrospective matched case-control study, multiplex immunohistochemistry was performed on thirty-eight paraffin embedded specimens from premenopausal women who had undergone a hysterectomy at the Amsterdam UMC between 2001 and 2018 to investigate the evidence for (lymph) angiogenesis in the (ectopic) endometrium or myometrium of patients with adenomyosis versus controls with unrelated pathologies. Baseline characteristics of both groups were comparable. In the proliferative phase, the blood and lymph vessel densities were, respectively, higher in the ectopic and eutopic endometrium of patients with adenomyosis than in the endometrium of controls. The relative number of blood vessels without α-smooth muscle actinin (α SMA) was higher in the eutopic and ectopic endometrium of adenomyosis patients versus controls. The level of VEGF staining intensity was highest in the myometrium but did not differ between patients with adenomyosis or controls. The results indicate increased angiogenesis and lymphangiogenesis in the (ectopic) endometrium affected by adenomyosis. The clinical relevance of our findings should be confirmed in prospective clinical studies.

Keywords: adenomyosis; angiogenesis; ectopic endometrium; histology; lymphangiogenesis.

Figure 1

Multiplex images viewed in PerkinElmer’s inForm^® advanced image analysis software (A,C,E) and analysis of multiplex images in NIS Elements using binary layers for an automated analysis (B,D,F) of control endometrium (A,B), eutopic endometrium (C,D), and ectopic endometrium (E,F). All magnifications were ×200 with each field representing 0.9 × 0.7 mm. Blood vessel density: blood vessels (count red areas)/mm² endometrium stroma (surface light blue area). Lymph vessel density: lymph vessels (count light green areas)/mm² endometrium stroma (surface light blue area).

Figure 2

Multiplex images viewed in PerkinElmer’s inForm^® advanced image analysis software (A,C) and analysis of multiplex images in NIS Elements using binary layers for an automated analysis (B,D) of control myometrium (A,B), and adenomyosis myometrium (C,D). All magnifications were ×200 with each field representing 0.9 × 0.7 mm.

Figure 3

Representative images of the individual IHC markers using the pathology view in PerkinElmer’s inForm^® advanced image analysis software, with the primary antibody staining brown (columns from left to right of CD31, a-SMA, podoplanin, and vascular endothelial growth factor (VEGF)) and cell nuclei blue in samples of control proliferative eutopic endometrium (top row), adenomyosis proliferative eutopic endometrium (middle row), and adenomyosis ectopic endometrium (bottom row).

Figure 4

Boxplots of blood vessel density determined by the count of CD31 stained areas (vessels) per mm² CD10 stained area (A) and lymph vessel density determined by the count of podoplanin stained areas per mm² CD10 stained area (B) in the eutopic endometrium in the secretory (n = 20, missing n = 4) and proliferative (n = 11, missing n = 3) menstrual cycle phase. The blood vessel density was significantly higher in the secretory than in the proliferative cycle phase (p < 0.05). There was no significant difference in lymph vessel density between the secretory and proliferative cycle phase. Data are shown as median ± range for all patient data pooled. NS: not significant; * significance p < 0.05.

Figure 5

Boxplots of blood vessel density in the eutopic endometrium determined by the count of CD31 stained areas (vessels) per mm² CD10 stained area (A) and lymph vessel density determined by the count of podoplanin stained areas per mm² CD10 stained area (B) in the eutopic endometrium per medication use, with ‘no’ is no/other medication, and yes is NSAIDs/tranexamix acid. There were no significant differences in blood or lymph vessel density between patients who reported to use medication with an effect on angiogenesis. Data are shown as median ± range for all patient data pooled. NS: not significant.

Figure 6

Top row: Boxplots of blood vessel density (BVD) determined by the count of CD31 stained area (vessels) per mm² in the secretory phase (A) and proliferative phase (B) of the menstrual cycle per tissue type. In the secretory phase, there were no significant differences in BVD between tissue types. In the proliferative phase, the BVD was significantly higher in the ectopic endometrium than in the control endometrium (p < 0.05). Bottom row: Boxplots of lymph vessel density (LVD) determined by the count of podoplanin stained area (vessels) per mm² in the secretory phase (C) and proliferative phase (D) of the menstrual cycle per tissue type. In the secretory phase, there were no significant differences in LVD between tissue types. In the proliferative phase, the LVD was significantly higher in the eutopic endometrium than in the control endometrium (p < 0.05). Data shown as median ± range. NS: not significant; * significance p < 0.05.

Figure 7

Boxplots of the ratio of the relative number of immature blood vessels, determined by the ratio of mature blood vessels that presented double immunostaining of CD31 and α-SMA to the total number blood vessels in the endometrium in the stroma in the secretory (A) and proliferative phase (B) of the menstrual cycle. In the secretory phase, the ratio of aSMA/CD31 vessels to CD31 vessels was lower in the ectopic endometrium than in the control endometrium (p < 0.01). In the proliferative phase, the ratio of aSMA/CD31 vessels to CD31 vessels was lower compared to the control endometrium in both the eutopic (p < 0.01) as well as the ectopic endometrium (p < 0.05). Data shown as median ± range. NS: not significant; * significance p < 0.05; ** significance p < 0.01.

Figure 8

Boxplots of immunohistochemical score (IHS) of VEGF staining per tissue type in the secretory (A) and proliferative phase (B) in adenomyosis eutopic and ectopic endometrium stroma and glands (n = 16) versus control endometrium stroma and glands (n = 18), and in adenomyosis myometrium (n = 19) versus control myometrium (n = 18) (right). There were no significant differences in VEGF IHS in each tissue type between the adenomyosis and control patients. Data shown as median ± range. NS: not significant.

Figure 9

Examples of Phenochart™ whole slide image viewer and the selection of multispectral image (MSI) areas of interest (green rectangles) of eutopic endometrium (red rectangle on panel (A)) and ectopic endometrium (red rectangle on panel (B)) after multicolor IHC staining of an adenomyosis sample.