Benign prostatic hyperplasia to HPIN to carcinoma; the impact of mast cells and VEGF

Abstract

Background

The process of transformation from Benign Prostatic Hyperplasia (BPH) to Prostate Cancer (PCa) is highly debatable; however, there is convincing information that it takes place in the background of inflammation. Multifocal high-grade prostatic intraepithelial neoplasia (HPIN), intraductal carcinoma (IDC-P), atypical intraductal proliferation (AIP), atypical adenomatous hyperplasia (AAH)/ adenosis, and proliferative inflammatory atrophy (PIA) are considered as precursor lesions during such transformation by Working Group 1 at the ISUP Cancer Precursors Meeting in September 2024. We specifically emphasize the need for further investigation of the inflammatory process, where stromal epithelial activity, angiogenesis, and mast cells are involved.

In this study, we investigated the role of vascular endothelial growth factor (VEGF), and MCs during its progression from BPH to PCa. Additionally, we investigated the relationship between MCs and Gleason Grade Group (GG), as well as other prognostic variables.

Methods

Retrospectively, 100 PCa patients whose pathology included areas of BPH and multifocal HPIN close to cancerous tissue. All the specimens were stained with CD117 as an MC marker and VEGF to evaluate for angiogenesis.

Results

While there was overall intense MC infiltration in all areas of BPH in general, it was relatively less intense in areas of carcinomatous tissue. Nevertheless, MC infiltration intensified with higher GG. Similarly, with increasing MC infiltration in cancerous and HPIN tissues, VEGF staining also intensified, specifically in glandular areas. In HPIN areas, the relationship between MCs and VEGF was intermediate between BPH and cancerous areas, representing a progression.

Conclusion

Stromal changes, stromal-epithelial interactions, and coexisting inflammatory changes are important elements in the development of BPH and PCa. We specifically investigated the intratumoral MC population and glandular VEGF staining intensity with increasing GG; thus, assessing the role of MCs and VEGF in tumor progression towards cancer development. MCs and VEGF staining characteristics in HPIN areas emphasize the efficient role these elements play in the transition of tissues from BPH to HPIN and eventually to PCa.

Introduction

Globally, prostate cancer is a common disorder [1, 2], and BPH is a physiologic process in men over 50. BPH is characterized by the expansion of epithelial and stromal elements [3] coexisting with inflammatory changes during this process [2, 4–7]. There are publications claiming that BPH and PCa have distinct genetic composition and molecular pathways, making BPH unlikely to be a precursor of PCa [8]. Nontheless, there are also publications that in the presence of inflammatory processes, BPH may progress to become PCa [2, 7, 9, 10]. Janiczek -Polewska et al., already demonstrated the existance of inflammatory pathways ‘’IL17A/IL-17RC/TRAF6/NF-kB and IL17A/IL-17RC/AKT1/NF-kB’’ that leads to its progression to PCa [11]. T lymphocytes and macrophages, MCs, eosinophils, and neutrophils constitute the principal elements of this inflammatory reaction [4–6].

One-third of men over 50 years old harbor preneoplastic changes in their prostate gland that lose their benign features in time and transform into a malignancy [1]. Working Group 1. at ISUP’s Cancer Precursors meeting (September 2024) evaluated 5 precursor lesions for Invasive Prostate Cancer, namely HPIN, intraductal carcinoma (IDC-P), atypical intraductal proliferation (AIP), atypical adenomatous hyperplasia (AAH) and proliferative inflammatory atrophy (PIA) [12]. Concensus was reached that the significance of unifocal HPIN was unclear; however, isolated multifocal HPIN was related to PCa [13]. The similar cytologic and genetic features [12] as well as its peripheral localization in the prostate gland and increasing incidence with age, make multifocal HPIN a precursor of PCa. Thus, HPIN is considered an evolving step between benign prostate epithelium and malignancy [14].

VEGF is a potent angiogenic factor and was originally recognized as a mitogen specific for CD 34 + endothelial cells. It is a member of a large growth factor family that includes VEGF-A, VEGF-B, VEGF-C, VEGT-D, and placental growth factor [6, 14, 15]. Each member of the family has a distinct biological activity and targets a specific receptor. VEGF is also released by macrophages, thrombocytes, renal mesenchymal cells, keratinocytes, and importantly neoplastic cells. A large array of growth factors and cytokines induce VEGF production and release through mRNA. Furthermore, VEGF is also involved in normal physiologic processes such as wound healing, skeletal growth, hematopoiesis, and tumor angiogenesis. Elevated VEGF levels are linked to the development of cancer cells, increased neoangiogenesis, tumor growth, and increased metastatic potential [16]. Already, numerous studies indicate the presence of elevated levels of VEGF in prostate cancer [6, 17]. Moreover, agents that suppress VEGF receptors through cancer immunosuppression may be an effective treatment tool.

Lately, investigators focused on MC infiltration and their role in angiogenesis [1, 6, 18]. MCs play a significant role in the emergence of malignancies, including prostate cancer [4, 19, 20]. There are 3 different variants of MCs [1, 4, 19, 20]. Depending on the MC variant, fibroblast growth factor, VEGF, TGF-B, and other similar growth factors are produced [18]. MCs also play a crucial role in the infiltration of malignant cells into stroma by producing trıptase and matrix metalloproteases [18, 20, 21].

İn this study, we investigated the importance of MCs and VEGF in the transformation of BPH to HPIN and eventually to PCa in inflammatory background. Likewise, the relationship of MCs and VEGF during such transformation. We also investigated the prognostic implication of MC concentration. It may be possible to influence the development and progression of PCa at the very onset by medications supposedly inhibiting the MC and VEGF activity.

Materials and methods

We investigated all the patients with a diagnosis of PCa who underwent prostatectomy between 2011 and 2022 at Baskent University Zubeyde Hanım Hospital in İzmir, Turkey. All the slides and archival tissues in paraffin blocks were reviewed. The study included 100 patients between ages 46 and 75, who had histological evidence of carcinoma, BPH, and multifocal HPIN, all coexisting in the same microscopic field. The patients who did not contain all three elements in the same microscopic field were excluded from the study.

In all patients, the paraffin blocks that included all three histologies, including PCa near BPH and multifocal HPIN, were included in the analysis. Two slides, each containing 4 micron-thick cuts, were prepared from tissue blocks. One of the slides was stained with H.E., the other with Toluidin Blue to identify MCs. Furthermore, additional 4-micron cuts were also made from each tissue block to prepare 2 sets of positively charged slides. Each slide was processed in DAKO AUTOSTAİNER, an immunohistochemical stainer, one for VEGF and the other for CD117. Two pathologists were double-blinded to interpret the slides under light microscopy, independently.

VEGF staining was evaluated in glandular areas as well as stroma in peri glandular areas surrounding the glands in all three circumstances including carcinoma, BPH, and HPIN. Cytoplasmic as well as membranous staining was considered positive. Scoring of VEGF was done according to Stefano et al. [22]. In an area where staining is most intense within a single large magnified field: 0 to 25% staining was scored weak (score1), 26% and over staining of the cells was scored as strong (score 2).

İnterpretation of CD117 staining was correlated with staining of MCs with Toluidin Blue. MC count was limited to intraglandular and stromal areas that were within 2 microns of glands. Henceforth, if CD117 stained MCs were 1 to 5, it was scored as 1; 6 to 10 as 2, and 11 and above as 3.

The cancerous area of each case has been classified according to the 2014 ISUP, Grade Group (GG), and Gleason Score system, after careful microscopic review. We also evaluated each case with respect to extraprostatic extension and lymph node invasion.

Statistical analysis

The statistical analysis of data in this study was evaluated in accordance with IBM SPSS Statistics Standard Concurrent User V 26 (IBM Corp., Armonk, New York, USA) statistical package program. The representing numbers (n) and percentages (%) were expressed in mean and +/- standard deviation values. The normal distribution of data belonging to numerical variables was evaluated with the Shapiro-Wilk normality test. When 2 groups were compared, an independent samplest test was employed, in case of normal distribution of data. Mann-Whitney test was utilized if the data was not normally distributed. For variables with more than two categories, One Way Analysis Of Variance (ANONA) was used in the case of normal distribution of data, and if the data was not normally distributed, Kruskal Wallis H was used. If the results of the Variance Analysis were significant, Dunn-Bonferroni was used as multiple comparison tests. Relationships between categorical variables were assessed using the Pearson chi-square and Fisher’s exact test. Fisher’s exact test was used if the frequency values in the cells were less than 5. A p < 0.05 value was considered statistically significant.

Results

100 patients (mean age 64) who underwent radical prostatectomy were included in this study. No lymph node metastasis was discovered in 95% of the patients. Only 5% of the patients had metastatic lymph node involvement, whereas 41% had extraprostatic extension. The percentage of patients according to their GG were as follows: GG1: 15%, GG2: 34%, GG3: 37%, GG4: 4%, and GG5: 10%.

When data was analyzed according to MC infiltration, except for 4 patients whose tumors were GG3, all patients had MC infiltration above 6 in BPH areas, surrounding the tumor, and the mast cell count was above 10 in most patients (Fig. 1). Conversely, when MC infiltration within the tumor was analyzed, it was low in tumors with low GG; whereas, its number increased with rising GG. Nonetheless, in contrast to BPH, a few patients (5 patients) had MC infiltration of more than 10 (Fig. 2) in their tumors, and 3 of these 5 patients had GG3.

Fig. 1.

CD 117 positive mast cells in areas of BPH (CD 117 × 200)

Fig. 2.

Massive CD 117 positive mast cell infiltration in areas of tumour (CD 117 × 200)

MC infiltration in areas of HPIN (Fig. 3) correlated with tumors, in the same manner according to their GG. In tumors with GG1 and GG2, as well as in their HPIN areas, the MC infiltration was correspondingly low. Whereas with higher GG, increased MC infiltration was observed.

Fig. 3.

CD 117 positive mast cells in areas of HPIN (CD 117 × 200)

When HPIN areas were reviewed in prostatectomy materials, none except 2 patients with GG3 had increased MC infiltration over 10. MC infiltration ranged between 1 and 10, in general.

No meaningful correlation was observed between the degree of MC infiltration and lymph node involvement; albeit, the few number of patients with lymph node involvement in our study may be the reason for this.

VEGF was analyzed separately with respect to glandular and stromal staining characteristics in areas of tumor (Fig. 4), BPH (Fig. 5), and HPIN (Fig. 6). Glandular staining in tumors was more intense with increasing GG. Whereas no statistically meaningful difference was discovered in stromal staining. When areas of BPH were analyzed, the glandular as well as stromal stainings were weak, independent of GG, in general. In areas of HPIN, staining, especially in glandular areas, was weak compared to the tumor; nonetheless, it was more intense than BPH, somewhere between BPH and tumor at 35% (Table 1).

Fig. 4.

Intense glandular VEGF staining in areas of tumour (VEGF x 100)

Fig. 5.

Positive stromal VEGF staining (red arrow), absence of glandular staining (yellow arrows) in BPH. Blue arrow indicates staining of vascular endothelial cells (VEGF x 100)

Fig. 6.

İntense glandular VEGF staining in areas of HPIN (VEGF x 100)

Table 1.

VEGF staining characteristics

VEGF staining	TUMOR	HPIN	BPH
Glandularly strong	62 (62%)	35 (35%)	22 (22%)
Glandularly weak	38 (38%)	65 (65%)	78 (78%)
Total (n/ %)	100 (100%)	100 (100%)	100(100%)
Stromal strong	34 (34%)	5 (5%)	11 (11%)
Stromal weak	66 (66%)	95 (95%)	89 (89%)
Total (n/ %)	100 (100%)	100 (100%)	100 (100%)

İn 80% of patients with lymph node involvement, glandular VEGF staining was strong in tumors; however, there was no meaningful difference with stromal staining.

When MC infiltration was compared with VEGF staining properties, in BPH with increasing MC infiltration (6–10 and > 10), there was a corresponding decrease in VEGF staining in stromal areas of BPH and HPIN. This was statistically significant (p < 0.05). Furthermore, in all BPH patients with low MC count, VEGF staining in glandular areas of the tumor and HPIN was intense with statistical significance (p < 0.05).

With increasing MC count in tumors, increased glandular VEGF was stained with statistical significance (p < 0.05). Similarly, with an increased MC population in areas of HPIN, the intensity of glandular VEGF staining increased correspondingly, though this was not statistically significant (Table 2). There was no correlation between MC infiltration in BPH with MC infiltration and VEGF staining in tumors.

Table 2.

Comparison of mast cell count with glandular VEGF positivity in tumor and HPIN, as well as stromal VEGF positivity in BPH

VEGF STAINING DENSİTY		Mast Cell Count			Test/p
		1–5	6–10	>10	P ≤ 0,05
TUMOR	Glandulary weak	32 (44,4%)	6 (26,1%)	0 (0,0%)
	Glandulary strong	40 (55,6%)	17 (73,9%)	5 (100%)
HPIN	Glandulary weak	37 (63,8%)	27 (67,5%)	1(50,0%)	p > 0,05
	Glandulary strong	21 (36,2%)	13 (32,5%)	1 (50,0%)
BPH	Stromal weak	2 (50,0%)	28 (93,3%)	59 (89,4%)	P < 0,05
	stromal strong	2 (50,0%)	2 (6,7%)	7 (10,6%)

There was a corresponding increase in VEGF staining intensity of glandular areas in both tumor and HPIN, with increased MC infiltration in HPIN.

No statistically significant correlation was found between extraprostatic extension and MC infiltration.

While VEGF glandular staining was strong in HPIN of extraprostatic extensive tumors, it was weak in tumors confined within the prostate and this was statistically significant (p < 0.05). Similarly, glandular VEGF staining was intense in patients with tumors extending outside the prostate.

There was no relationship between VEGF staining in BPH and extracapsular extension.

No relationship was found between patient age and VEGF staining characteristics, as well as MC infiltration.

Discussion

Thus far, the results of observational studies, meta-analysis, and subgroup-analysis indicate that the difference between BPH and PCa incidence has risen by 2.9 fold, worldwide [12]. Although the difference may be related to ethnicity, hormonal and inflammatory changes taking place in the prostate gland may play an important role [ 5, 7, 10, 19, 23]. Specifically, the imbalance between cell proliferation and cell death in BPH patients may explain the increased risk of PCa development [23, 24].

Changes in stroma and inflammation in the prostate gland play an important role in the development of BPH. The metabolic stimulation of the stromal microenvironment, epithelial-stromal interaction, and inflammation have been reported by numerous studies to play an important role in the development and progression of cancer [1, 10, 25]. Thus, it is important to understand the different roles played by the cells in the stroma in order to comprehend the relationship between benign and malignant lesions in the prostate gland. Reactive stroma is composed of fibroblasts, endothelial cells, myofibroblasts, MCs, and other inflammatory cells. The role of MCs is controversial and debated, without a fully understood role [5, 26]. The increased number of MCs in tumors has been a subject of debate. Some authors claim that they stimulate the growth of malignant cells, while others claim the opposite [5, 9]. Hempell et al. reported increased recurrence of PCa when intratumoral MC count is low [5]. Conversely, there are studies that increased release of VEGF by MCs generates tumor angiogenesis [6, 18, 27].

Aydın et al. reported increased MC infiltration in BPH areas in contrast to intratumoral areas. Furthermore, they reported no relationship between GG and MC infiltration [4]. Similarly, we also demonstrated increased MC infiltration in areas of BPH, while MC infiltration was low in tumorous areas. Contrary to Aydın et al.’s observation, we discovered that low MC count in tumors correlates with low GG and high MCs with correspondingly higher GG.

Similar MC infiltration in areas of HPIN and tumor may indicate the importance of these cells, playing an important role in the progression from relatively benign histology to a malignancy.

MCs are known to release mediators such as VEGF, FGF-2, and TGF during degranulation that lead to angiogenesis [18]. VEGF is one of the important agents that induce angiogenesis. Angiogenesis plays an important role in tumor progression, invasion, and metastasis [6, 18, 28, 29]. In PCa patients that have elevated blood VEGF levels correlate with poor prognosis [6, 16, 28].

MCs are activated by Ig E antibodies, antigens, mediators, and hormones with resultant degranulation [5, 6]. Shaik et al. reported that Substance P and IL-33 induce VEGF release by MCs, supporting tumor progression. A strong relationship has been reported between VEGF, MCs, and inflammation [6]. In our study, we also reported the strong relationship between VEGF staining intensity and MC infiltration, specifically ın tumors. Similar findings are also reported in HPIN areas as well; however, no relationship existed between VEGF staining and MC infiltration in BPH. Nonetheless, similar to our findings and studies by Globa et al. this may be related to different subtypes of MCs infiltrating areas of BPH than the MCs infiltrating tumors [1].

Woodland et al., investigated the differences in VEGF variants expressed by PCa; specifically, VEGF-A and VEGFR-1 have been expressed more abundantly in tumors than in BPH areas. Whereas, no difference was reported in stromal VEGF staining, between tumors and benign areas [30]. Similarly, in our study glandular VEGF staining in tumor and HPIN were more pronounced than BPH. Furthermore, in patients with elevated GG indicating, more aggressive tumor behavior, the VEGF staining became pronounced in glandular areas. İn studies involving bladder cancer, ovarian cancer, and lung cancer there was a direct correlation between increased glandular VEGF -A expression and poor prognosis. These observations support our findings. Furthermore, we have discovered that in patients where extraprostatic extension and lymph node metastasis are demonstrated, glandular VEGF expression is profound; thus supporting the observation that strong glandular expression of VEGF correlates with aggressive behavior of the malignancy. This is supported by the other investigators as well [17, 30].

İn our study, we demonstrated that with elevated GG, the glandular VEGF expression and the number of MCs infiltrating the tumor are increased; thus, intratumoral MCs are responsible for increased VEGF release, resulting in tumor progression and increased aggressiveness. This observation was not supported by Hempell et al.; nonetheless, a similar observation was discovered by Johansson et al. They demonstrated that in PCa, MCs stimulate tumor angiogenesis; whereas, inhibition of MC activity by imatinib interrupts tumor growth and angiogenesis. İnvestigators failed to demonstrate any correlation between MC count and microvessel density in normal prostate tissue. Additionally, low intratumoral MC count was correlated with a good prognosis [6, 19], similar to our observations.

Contrary to our discovery, Pittoni et al. demonstrated increased MC infiltration in normal, HPIN, and well-differentiated carcinoma in experimental animal models. Specifically, they reported decreased MC infiltration in poorly differentiated areas. They claimed that this might have been related to the poor ability of the tumor to produce CSF as a ligand to CD117 released from MCs [19, 21]. Whereas, in our study, we demonstrated increased MC infiltration in BPH areas, contrary to tumors and HPIN. This probably is related to different subtypes of MCs in normal, HPIN, and tumorous areas, as well as decreased degranulation of MCs in cancerous glandular fields, leading to poor identification under light microscopy and IHC staining.

Conclusion

Changes in stroma and stromal epithelial interactions are important in the development of BPH and PCa. Inflammation is an essential element in the BPH and PCa microenvironment. Specifically, MCs, among other inflammatory cells, depending on their subtype and location in the prostate gland, play either a stimulatory or an inhibitory role through the release of cytokines, such as VEGF and other angiogenic factors. We discovered the important role played by the increased number of intratumoral MCs and strong glandular VEGF expression, in tumor progression and subsequent increase in GG.

Furthermore, we related the decreased VEGF staining intensity in BPH to the presence of different MC subtypes in these locations with a lack of stimulation to release VEGF.

Presumably, the release of other mediators and c-kit (CD117) expression directly from PCa cells plays a chemotactic role for some MC subtypes that subsequently lead to VEGF expression. We plan further studies to investigate this relationship.

In this study, we discovered that the MC number and VEGF expression intensity in HPIN are at an intermediate level between BPH and carcinoma. İn a stepwise progression from BPH to HPIN and eventually to carcinoma, MCs, and VEGF play an important role. Further comprehensive studies involving other preneoplastıc lesions such as intraductal carcinoma (IDC-P), atypical adenomatous hyperplasia (AAH)/adenosis, and proliferative inflammatory atrophy (PIA) will be valuable for PCa evaluation, recommendations regarding early diagnosis and discovering targeting agents for treatment.

Authors’ contributions

ON Sivrikoz designed the study, made the initial interpretation of all the slides, and special stains. She is the principal contributor to the authorship. A Eroglu and B Ozturk are urologists, they were involved in obtaining tissue samples. Oztop is a second pathologist who independently reviewed all the slides. SM Sanal, a senior author, contributed to the design and interpretation of data, and did major auditing.

Funding

The authors received no financial support for the research.

Data availability

Data will be available from the authors upon reasonable request.

Declarations

Ethics approval and consent to participate

Our study is a retrospective observational study. According to Article 2, subparagraph (2) of the Clinical Research Legislation of the Ministry of Health of the Republic of Turkey, it is outside the scope of ethical approval. Written informed consent forms were obtained from all patients.

Competing interests

The authors declare no competing interests.