Clinical and morphometric characteristics of endometrial polyps and underlying mechanisms of abnormal uterine bleeding

The prevalence of endometrial polyps (EP) is generally estimated to be between 25% and 35%, which constitutes the greatest proportion of intrauterine pathologies. EP is an indication for every fourth hysteroscopy, due to their association with abnormal uterine bleeding (AUB), infertility, and the risk of malignancy [1-5]. The proportion of EP manifested by AUB, according to various sources, ranges from 38 to 81.5%, depending on the diagnostic methods and the patient age [1-7-7]. Some evidence suggests that EP causes AUB in 81.5% and 69% of patients in perimenopause and postmenopause, respectively [8]. According to other data, the incidence of AUB in patients with EP is significantly lower and constitute 38% and 46.5% among patients of reproductive age and in postmenopause, respectively [5, 6, 9]. EPs can manifest as prolonged and excessive menstrual bleeding (PEMB) or intermenstrual bleeding (IMB). However, the current literature is lacking sufficient coverage of the evidence regarding the incidence and possible causes of different types of AUB in patients with EP. The mechanisms leading to the clinical manifestation of the EP are still a subject of debate. The occurrence of AUB is often associated with chronic inflammation, pathological angiogenesis, thrombosis, capillary micro-erosion, and ischemic necrosis in EP [1, 2, 10]. It can be assumed that large and multiple EPs may cause AUB more often than small and solitary ones. However, most authors have not reported such a correlation [11, 12].

Clinical manifestation of the EP may also depend on their morphological structure, in particular, on the relative preponderance of glandular and stromal components, which underlies the categorization of EPs into glandular, fibrous and glandular-fibrous [13]. It should be noted that WHO experts (2014) do not support this classification due to its questionable clinical significance, including the association between the EP morphological structure and AUB [14].

Therefore, the incidence of various types of AUB in patients with EPs and their underlying mechanisms remain unclear, which causes difficulties in the diagnosis and the choice of optimal management strategy. To address these problems, this study aimed to investigate the histological and morphometric characteristics of EPs and their association with AUB.

Materials and methods

The study comprised 130 women aged 21 to 45 (mean 31.8 ± 8.6) years with histologically confirmed EPs. Their medical records were reviewed to retrospectively analyze the course of the disease, histological, and morphometric characteristics of EPs. The exclusion criteria were diseases that could be the cause of AUB, including submucous fibroids with centripetal growth, grade II-III adenomyosis, and coagulopathy. FIGO classification system (2018) was used to categorize the causes of AUB. All cases of AUB were categorized into intermenstrual bleeding (IMB), prolonged and excessive menstrual bleeding with a regular cycle (PEMB) or oligomenorrhoea (Ol + PEMB), and a combination of prolonged and excessive menstrual bleeding with intermenstrual bleeding (PEMB + IMB).

In addition to the morphological characteristics of the EP, the state of the neighboring endometrium was assessed. All samples were subdivided into three subgroups including EP with coexistent proliferative phase endometrium (EP + PPh, n = 58), EP with co-occurring chronic endometritis (EP + CE, n = 63), and EP with coexistent endometrial hyperplasia (EP + EH, n = 7). Chronic endometritis (CE) was subdivided into complete and incomplete forms. The criteria for the complete form included the presence of plasma cells, focal lymphoid infiltration, and fibrotic stroma. An incomplete form of CE was diagnosed in the presence of focal or diffuse lymphocyte infiltration, stromal fibrosis or spiral artery sclerosis with coexistent proliferative phase endometrium.

The morphometric analyses included 30 samples of EP from patients with (n = 15) and without (n = 15) AUB. The samples were stained with hematoxylin and eosin at an optical magnification of × 100. In each EP sample, two fields of view of the vascular pedicle were selected to measure the following parameters: the number of vessels, the thickness of the vascular wall, the area of ​​the vascular component, endometrial glandular area, stromal area, and the gland-to-stroma ratio. The measurements were performed with the Axio Vision 4.0 image analysis software (Karl Zeiss, Germany), and the results were presented in micrometers and square micrometers.

Statistical analysis was performed using the Statistica 10.0 software. Categorical variables were compared by chi-square test. The differences in morphometric variables between groups were assessed by Mann-Whitney U-test. Continuous data are presented as median (Me) and quartiles (Q1; Q3). Differences were considered statistically significant at p <0.05.

Results

Histological findings showed that in two women (1.5%) aged 40 and 45 years EP had the signs of atypical hyperplasia (AH). The remaining EPs were characterized as benign glandular-fibrous (n = 106; 82.8%), glandular (n = 13; 10.2%) or fibrous (n = 9; 7.0%). In 60 (46.1%) cases, EPs coexisted with proliferative phase endometrium (PPh), and the remaining 53.9% had concurrent uterine mucosa pathologies. Most frequently, EPs coexisted with CE (n = 63, 48.5%) and EH without atypia (n = 7, 5.4%). The complete and incomplete forms of CE were found in almost equal proportions - 24.6% and 23.9%, respectively.

Of the whole study cohort, 79 (60.8%) patients with EP experienced AUB, while the remaining study participants had asymptomatic Eps, which were incidentally found during ultrasound examinations. In 37 (46.8%), 24 (30.4%), and 12 (15.2%) patients AUB manifested as PEMB, IMB, and PEMB + IMB, respectively. PEMB with concomitant oligomenorrhea was observed in 6 (7.6%) patients; in all of these cases, EPs coexisted with EH. EPs with foci of AH also manifested as Ol + PEMB. Because AUB in the above cases could be attributed to both EH and EP, they were excluded from further analysis. There were no statistically significant differences in rates and types of AUB between age groups of patients with EP.

The relationship of AUB types with a histological diagnosis is presented in Fig.1. As can be seen, AUB was slightly more often (by 15.1%) associated with the combination of EP with CE than with PPh. After dividing CE into complete and incomplete forms, it turned out that the manifestation of EP was significantly more often observed in patients with the complete form of CE than with PPh (OR = 2.56; 95% CI [1.01; 6.49]; p = 0.05 ). In the group with incomplete CE, no significant differences from PPh were found. Regarding the structure of different types of AUB, the prevalence of PEMB was observed in all subgroups irrespective of the neighboring endometrium. Their proportion in the complete form of CE was slightly higher than in PPh. However, no statistically significant differences were found.

To assess the relationship of clinical manifestation of EPs with their diameter, all samples were categorized into small (<0.6 cm), medium (0.6-1.0 cm), and large (>1.0 cm) polyps. Proportions of the small, medium, and large EPs were 31.6%, 46.9%, and 21.5%, respectively. The diameter of the EP was found not to be associated with the patients’ age or the presence of CE.

As shown in fig. 2, the rates of AUB in patients with EPs of different diameters did not differ significantly. However, an increase in the EP size was in parallel with an increase in the proportion of PEMB from 9.7% in the patients with small to 38.1% with large EPs (p <0.001). Significant differences in the PEMB rates were also observed between groups of medium and large EPs (p = 0.001), while no significant difference was found between small and medium ones. In patients with small and medium-sized EPs, the rates of IMB were approximately equal, but they were six times lower in patients with large ones. It was noted that the rates of PEMB + IMB were inversely associated with the EP diameter. One hundred and nine (83.9%) and 21 (16.1%) patients had single and multiple Eps, respectively. At the same time, the incidence of AUB was similar - 57.1 and 60.0%, respectively. Compared to patients with single EPs, in patients with multiple EPs, PEMB was observed 1.9 times more often and IMB 2.2 times less frequently, although no statistical differences were found.

The results of the morphometric study, conducted along with the histological analysis, are summarized in Table. 1. The number of vessels and the area of ​​the vascular component did not differ significantly between EP with and without AUB. However, in AUB samples showed microvessels congestion and expansion, which was reflected in their lumen increase (p = 0.03). Depending on their diameter, the vessels were divided into small (less than 40 microns), medium (40-80 microns) and large (more than 80 microns). In AUB samples, the vessels had mainly medium and large lumen diameters, and only 13% had a small diameter. Patients with the vessel diameter of more than 80 μm were 2.3 times more likely to have symptomatic EPs than asymptomatic ones (p = 0.05) (see Fig. 3).

Among the patients with AUB, the thickness of the vascular wall in the EP was thinner than in the group without AUB (16.5 μm vs. 24.8 μm, p = 0.02). When the samples were categorized into vessels with thin (less than 20 microns), medium (20-30 microns) and thick (more than 30 microns) walls, the formers were found to account for 60% of all AUB cases, and only 13% of AUB samples had thick walls. The visual differences of EPs in asymptomatic patients and patients with AUB are shown in fig. 4.

The morphometric evaluation of EPs in patients with different AUB types showed no statistically significant differences between PEMB and IMB. The presence of CE was associated with pathological changes in microvessels. In the presence of CE, the vessel diameter in the EP was greater (p = 0.01), and the wall thickness was thinner (p = 0.03). This observation correlates with data on the higher detection rate of AUB in EP with CE (see Fig. 1). Comparison of morphometric parameters of EP coexisting with CE and PPh with and without AUB showed that regardless of the state of the neighboring endometrium, AUB was associated with a tendency for thinner walls and greater diameter of microvessels. However, in the presence of CE, these changes were more pronounced than in the case of endometrial PPh. Thus, among symptomatic EPs, EP + CE had significantly greater vessel diameter than in EP + PPh (p = 0.04), but the thickness of the vascular walls was similar. The diameter and number of vessels and the thickness of the vascular wall had no significant relationship with the size of EPs. The area of the vascular component gradually increased from 72922.7 (24082.6; 106660.3) to 209176.9 (108197.6; 446452) and 308760.6 (112070.7; 356753.6) μm2 in small, medium, and large Eps, respectively. Therefore, the area of the vascular component in large EPs was 4-fold greater than in small ones (p = 0.007).

Of the 30 EPs selected for the morphometric study, 16, 8, and 6 were found to be glandular-fibrous, glandular, and fibrous, respectively. According to morphometry, the gland-to-stroma ratio ranged from 0.1 to 0.3 and had no significant differences between these groups. The glandular and stromal areas and the gland-to-stroma ratio did not correlate with the presence of AUB, concomitant endometrial pathology, and the EP diameter.

Discussion

EPs are the most common intrauterine pathology and the cause of AUB in both reproductive and postmenopausal patients. EPs also have a high recurrence rate resulting in multiple intrauterine interventions that not only reduce women’ quality of life but also inflict endometrial injury, increasing the risk of intrauterine synechiae and infertility. EP is also associated with the risk of malignancy, which occurs in 0.5–3.4% and 12.9% of EPs in the reproductive and postmenopausal age, respectively [2, 4, 12]. In this study, no polyp turned malignant; however, 1.5% of them had AH, which is generally consistent with the literature. In the overwhelming majority of cases, EPs were glandular-fibrous, less commonly, glandular and fibrous.

The results of a more objective morphometric study revealed no significant differences in endometrial glandular area, stromal area, and the gland-to-stroma ratio in samples with the above histological diagnoses. It is worth noting that the gland-to-stroma ratio did not correlate with any of the clinical, histological, and morphometric parameters. This observation reflects a low clinical significance of the EP classification, based on the endometrial gland to stroma ratio, and confirms the feasibility of using the WHO classification (2014), which divides the EP into only two categories - benign and with AH.

The main clinical manifestation of EPs is AUB. In our cohort, they were found in 60.8% of patients. Almost every second patient had AUB characterized by PEMB, every third had IMB, and much fewer of them had their combination. PEMB with oligomenorrhea in all cases was associated with EH, including AH in EPs. These results indicate that EP can manifest as different types of AUB, but they usually arise against the background of a regular menstrual cycle, whereas AUB alternating with oligomenorrhea is characteristic of EH. Thus, the nature of menstrual irregularities may indirectly indicate the presence of a particular endometrial pathology even at the prehospital stage.

In our study, the incidence of AUB did not depend on the diameter or the number of EPs, which is consistent with the literature [11, 12]. However, the type of AUB was associated with the size of EPs. Patients with larger EPs were more likely to have PEMB than those with smaller EPs. Some authors suggest that the torsion movements of large EPs may result in the injury to the capillary endothelium resulting in thrombosis, and apical ischemic necrosis of EPs [15]. This causes an extensive area of ​​damage to the microvascular network, which leads to an increase in blood loss at the onset of menstruation. Small-sized EPs were more often characterized by IMB. There is an opinion that IMB can occur against the background of leukocyte infiltration of the endometrium or with accompanying CE because proteolytic enzymes of leukocytes destruct intercellular bonds and cell membranes resulting in the release of prostaglandins, vascular thrombosis, ischemic necrosis and vessels’ autolysis. It was shown that the affected area of arterioles, in this case, represents micro-erosions, which are manifested by scant bleeding in the middle of the menstrual cycle [15]. Even though patients with multiple EPs were almost twice more likely to have PEMB than IMB, no statistically significant data were obtained. This may be due to the relatively small number of patients with multiple EPs in our study.

The mechanism of occurrence of AUB in EP remains poorly understood, and therefore a morphometric analysis of symptomatic and asymptomatic EPs was conducted. The study findings showed no differences in area and number of vessels in the EP between the study groups. However, in patients with AUB, the diameter of the vessels was significantly larger, and the vessel walls were thinner than in those with asymptomatic EPs. No similar studies on the morphometric characteristics of EP vessels and their association with AUB were found in the available literature. However, it has been shown that the large, dilated, thin-walled vessels are characteristic of AUB resulting from local and systemic use of gestagens [16, 17].

It should be noted that, unlike this work, the authors of these studies did not use computer morphometry, and the pathological changes of the vessels were determined only during the histological evaluation of endometrial specimens. Pathologically dilated endometrial vessels are characterized by a decrease in the number of pericytes (cells bearing a supporting function in the capillaries) and vascular smooth muscle cells, which leads to a significant thinning and increased fragility of vessel walls. The expansion of capillaries can be associated with blood stasis, increased pressure, and mild traumatization of the thin vessel walls [16-18].

Vascular pedicle of EP contains a tangle of large blood vessels that are thought to be a result of pathological angiogenesis, which may be one of the possible causes of structural changes in the vessels, leading to AUB [10, 13, 16]. One of the most important inducers of neo-angiogenesis is vascular endothelial growth factor (VEGF), which expression may be increased in EP [19, 20]. Increased VEGF expression contributes to the formation of dilated thin-walled capillaries of the functional endometrial layer [16], which may underlie structural vessel changes and the occurrence of EP-associated AUB. Another important link in the regulation of angiogenesis is the NOTCH signaling pathway, which plays a key role in the formation of blood vessels in both normal endometrium and EP [21]. The NOTCH-3 gene mutations are associated with the pathological changes in vascular smooth muscle cells and the development of arteriovenous malformations, which are one of the causes of bleeding [22]. It is possible that a similar mechanism underlying the impaired angiogenesis also plays a role in the occurrence of AUB in patients with EP.

As has been previously reported in several studies, the formation of vascular pedicle of EP may be the result of progressive vasculopathy associated with CE [10, 23]. Carvalho et al. [23] found a significant association between CE and endometrial vascular changes in 85.7% of infertile patients. EPs are known to co-occur with CE in 19.7–27.5% [4, 23]. In the study, they were detected against the background of inflammatory changes in the endometrium in every second (48.5%) patients. This high incidence of concomitant CE can be explained by the presence of both the complete (24.6%) and incomplete form of CE (23.9%), in which plasma cells and a high level of proinflammatory cytokines in the endometrium were also detected by the IHC analysis [24].

It is important to note that in EPs co-occurring with CE, the thinning of the walls and dilation of microvessels’ diameter was more pronounced compared to EPs associated with PPh endometrium. This observation may indicate that the long-term inflammatory process aggravates the capillary pathological changes leading to a 2.5-fold higher incidence of AUB when the EP is combined with the complete form of CE.

Conclusion

Considering the results of the study, it can be concluded that the classification of EP into glandular, fibrous, and glandular-fibrous types is not sufficiently validated for clinical application. In our cohort, 60.8% of patients with EPs and the regular menstrual cycle had AUB, every second patient had AUB characterized by PEMB, every third had IMB, and much fewer of them had their combination. One of the possible mechanisms underlying AUB in patients with EP may be structural changes in microvessels resulting in the formation of dilated thin-walled capillaries, which probably occur due to impaired angiogenesis. The presence of CE in the neighboring endometrium aggravates the vascular morphological abnormalities and increases the likelihood of AUB.