Introduction
Arteriovenous malformations (AVM) of skin and soft tissues are rare congenital vascular lesions, arising from dysmorphogenesis during embryogenesis. They usually have a slowly progressive growth commensurate with the growth of the patient, and do not regress overtime [
1‐
3]. Compared to other types of simple vascular malformations, AVM are the most problematic lesions in terms of clinical behaviour due to their high flow characteristics [
2]. Despite an overall quiescent type of growth, AVM can enlarge disproportionally over time and may become symptomatic due to trauma, inflammation, thrombosis, hormonal changes, or complications related to pressure on surrounding tissues [
4,
5].
Episodic growth of AVM has also been attributed to occurrence of areas of microvascular proliferations (MVP) [
4‐
6]. Meijer-Jorna et al. have reported presence of MVP areas in 30% of 107 resection and amputation specimens of AVM obtained from patients who were treated for symptomatic disease [
4,
5]. In 2009, Duyka et al. reported that a disproportional growth of the AVM mass occurs during episodes of hormonal disbalance such as adolescence, puberty, pregnancy or the use of hormonal contraception [
7‐
11]. Such observations raise the question whether hormonal stimulation or dysregulation in patients with AVM could be involved in stimulating or aggravating a process of angiogenesis. Indeed, the highest risk of progression to a higher Schobinger Stage occurs during adolescence, suggesting that circulating hormones might contribute to AVM proliferation [
12,
13].
Several studies have demonstrated that specific hormone receptors (HR) such as progesterone receptor (PGR), growth hormone receptor (GHR), and follicle-stimulating hormone receptor (FSHR) were present in vascular tumors and vascular malformations [
5,
7,
14]. In these studies, expression of GHR, FSHR, and PGR was reported mostly in endothelium of vascular malformations. GHR was found in the endothelial/perivascular area of vascular malformations, and also in the stroma of control tissues [
5]. FSHR was found mostly in endothelial cells of vascular anomalies, but not in normal control specimens [
14]. PGR was found in lesional endothelial cells (EC) and smooth muscle cells (SMC) [
7]. Estrogen receptors (ER) could not be detected in the vessel walls of vascular malformations [
5,
7], but interestingly were reported to be present in infantile hemangiomas (IH), which are characterized by an initial stage of growth due to microvascular (capillary) proliferations [
15].
Hormones such as estrogen and follicle-stimulating hormone (FSH) have been described to be involved in angiogenesis. Several studies have reported a positive correlation between ER expression, angiogenic activity, and an invasive growth pattern of breast tumors [
16‐
18]. These findings support the angiogenic effect of estrogen mediated by ER. Moreover, a study reported that FSHR was found in pericytes of IH, but it remains unclear whether this concerns the proliferation phase of the these lesions [
19]. Other studies also found that FSHR expressed on the EC of the blood vessels of tumors like lung, breast, colon, kidney, and metastatic prostate carcinoma [
20,
21].
Expression of several HR on the vascular walls of several types of vascular malformations including AVM has been described only in mature vessels of vascular malformations, but was not investigated specifically in areas of MVP [
5,
22]. Hypothetically, an increase in expression of HR in vasoproliferative areas could explain the expansive growth of AVM types of vascular malformations during episodes of hormonal excess.
Therefore, the present study was designed to immunohistochemically investigate the expression of several types of HR in which a vasoproliferative response had occurred. Archived paraffin blocks of large AVM resection specimens were selected which contained histologically proven vasoproliferative areas amidst the vessels of the pre-existent arteries and veins of the malformation. Immunostaining of ER, PGR, GHR and FSHR was applied, followed by comparing the relative expression of the respective HR in both areas of interest (MVP versus pre-existent mature vessel areas). Since several cell types can be involved in angiogenesis, particularly EC, SMC, and mast cells (MC), we also investigated the co-localization of HR expression in these cell types. Potentially this could improve our knowledge on angiogenic growth in AVM lesion, in addition to the previously known insight of angiogenic factors released by mast cells [
23‐
25].
Discussion
Expansive growth related to onset of MVP has been reported in a substantial fraction of arteriovenous malformations (AVM) [
6]. The present study reports the expression pattern of several types of HR in relation to vasoproliferative activity. We found that ER, PGR, GHR, and FSHR expression could be detected immunohistochemically in all lesions, both in histological areas of closely packed immature microvessels representing proliferative growth, and in the thick-walled arteries and veins of the malformations. This was further confirmed by RT-qPCR analysis in 3 lesions of which frozen tissue was available. On the other hand, expression of HR appeared to be absent in vessels of samples of normal skin tissue, which indicates that HR expression could be a specific feature of the malformations, albeit not solely for the proliferative component.
HR expression has been reported previously in tissue specimens of vascular malformations, especially in AVM [
5,
7,
14,
32]. Kulungowski et al. reported that GHR was primarily present in the endothelium/perivasculature of malformations [
5], Maclellan et al. reported that FSHR was found in the endothelium [
14], and Duyka et al. reported that AVM, VM, and LM specimens showed positive staining for PGR within the nuclei of EC and SMC of the malformed vessels [
7]. However, occurrence of MVP areas was not noticed or specifically described in these reports, so relative expression of HR in proliferative areas has not been published thus far.
Estrogens are considered to be the most-involved hormones in puberty of both male and female [
7]. Kulungowski et al. [
5] reported the expression of ER in a series of malformations, which included also a case of AVM, but expression was less abundant compared to GHR expression in their samples. Duyka et al. [
7] and Ventejou et al. [
22] reported negative expression for ER in vascular malformations. In our series, ER expression was found in all types of vessels involved, including arteries, veins and microvessels, albeit in variable proportions. In general, estrogen may promote angiogenesis in vivo and in vitro through several mechanisms, including activation of ER and consequently enhance the pathophysiological process of angiogenesis in EC [
5,
33,
34]. Estrogen also stimulates vascular endothelial growth factor, which is involved in neovascularization regulating angiogenesis [
24,
35]. Additionally, our finding of concomitant expression of ER and GHR in lesions could imply that estrogen affects GHR functionally [
5,
24,
36]. Estrogen stimulates growth hormone (GH) and its receptor in SMC, and proliferation and migration of EC [
37,
38]. We found indeed ER expression both in EC and in SMC (Fig.
4A, B). Generally, the staining pattern of ER was mostly nuclear, although we also found additional cytoplasmic staining. This cytoplasmic pattern meets the requirements of ‘positive assessment of cytoplasmic ER’ as described in a previous publication [
39].
GH also tends to have a role during puberty, when levels of GH increase rapidly [
22]. Increased levels of GH are associated with altered circulating levels of angiogenic factor affecting endothelial cell function [
40]. Kulungowski et al. showed an elevated expression of GHR, but no expression of PGR in vascular malformations compared to control samples [
5]. Their finding of GHR expression is in line with our findings in the present study, supporting the view that GH may have a role in the expansion of vascular malformations.
Hormonal influence through progesterone in AVM lesions was supported by Duyka et al. who found that 83% of AVM samples in their study group showed diffusely positive PGR expression in EC and SMC compared to controls [
7]. However, a possible contribution of progesterone to vasoproliferations in AVM is still unclear, as progesterone is a less potent endothelial mitogen during neovascularization compared to estrogen [
3]. Indeed, in our series, PGR showed no significant differential expression between proliferative and non-proliferative areas.
Maclellan et al. have reported the expression of FSHR in a diverse series of vascular anomalies, which included different types of vascular malformations including AVM, and also benign vascular tumors, whereas vascular control tissues were negative. Among these, the highest levels of expression were found in the proliferative stages of IH, which is characterized by abundant microvascular proliferations [
14]. The authors argued that the secretion pattern of the hormone, FSH, correlates with the growth cycle in IH, increasing after birth when IH proliferates rapidly and decreasing later when the growth of IH slows. And of note, FSH surges again during adolescence, when AVM are most prone to worsen clinically [
14]. Our observation of FSHR expression in EC of MVP areas, might support the view that FSHR plays an important role in AVM expansion.
Taken together, previous publications and this study have provided sample evidence for expression of hormone receptors by vascular wall cells of AVM lesions. At the same time, vascular malformations are likely to grow under hormonal influences as has been noticed clinically during adolescence and pregnancy [
7]. Moreover, growth of vascular lesions under hormonal stimulation can be explained by the absence or scarcity of these receptors in normal vascular tissues as shown in this study and others [
5,
7,
14,
19]. HR receptors are not only expressed on vascular wall cells of mature AVM vessels but also on vascular wall cells of proliferating microvessels, resulting in high tissue densities of especially positive EC in MVP areas (Fig.
3), which is a major finding of this study, and we postulate that they could provide a target area for rapid expansive growth under hormonal influence. Although we only focused on ER, PGR, GHR, and FSHR expression, other studies have reported on androgen receptor expression [
5,
22].
It is assumed that the vasoproliferative response represents a reactive phenomenon, which can be induced by external factors such as inflammation, trauma, and especially circumstances of tissue hypoxia as has been published previously [
41‐
48]. The high EC: SMC ratios that we found in proliferative areas opposed to the mature areas of malformations suggests that EC are the most important target for hormonal influences, stimulating the process of angiogenesis. We investigated also the expression in interstitial MC, which occur in high numbers in proliferative IH and in MVP areas of AVM, and are considered to exert important angiogenic effects through production of various angiogenic peptides [
4,
49]. We found hardly no expression of HR in MC (see Fig.
3). Expression of ER and PGR has been reported previously in MC by other studies [
50,
51], but not specifically in lesions of AVM. Moreover, Hou et al. showed ER expression in MC of IH samples, but they reported that these MC might be involved in the regressive stage, and not in proliferating stage of IH [
15]. This needs further research as literatures on this subject of HR in MC is very limited.
Conclusion
The present study advanced our knowledge on angiogenesis and hormonal involvement in the progression of AVM, since ER, PGR, GHR, and FSHR were expressed not only in the mature malformed vessels of AVM but also in the vasoproliferative areas of these lesions. Based on these findings, we propose that the growth of vascular malformations in situations of hormonal dysbalance could be explained by the expression of hormone receptors on vessels of the malformations, in which MVP components may participate. These MVP areas, characterized by high vascular densities, may serve as a target area for episodic rapid growth under such hormonal dysbalanced situations. Unfortunately, due to the anonymized status, no information was available on the hormonal status of the patients involved in this study, but these findings could be useful for ongoing research on a molecular level to determine the definitive association between hormones and angiogenesis in AVM, and development of targeted therapies for recurrence AVM by reducing the use of hormonal therapy and preventing the use of hormonal drugs or contraception.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.