Correlations of PTEN and ERG Immunoexpression in Prostate Carcinoma and Lesions Related to Its Natural History: Clinical Perspectives

Purpose: The aim of our study was to observe the associations between the ETS-related gene (ERG) and the phosphatase and tensin homolog gene (PTEN) immunoexpression in prostate cancer and related lesions and highlight the clinical significance of these findings. Methods: We evaluated the immunohistochemical expression of ERG and PTEN in a series of 151 invasive prostate adenocarcinomas, including low-grade (Gleason grade pattern 3) and high-grade (Gleason grade patterns 4, 5) morphological patterns which corresponded to 45.5% and 54.4% of the cases, respectively. Additionally, we evaluated the immunoexpression of the two markers both in foci of high-grade prostatic intraepithelial neoplasia (HGPIN), as a precursor lesion of cancer, and in foci of intraductal carcinoma of the prostate (IDCP). Finally, to ensure the malignant nature of the prostate glands examined, we employed p63 and alpha-methylacyl-CoA racemase (AMACR) expression. Results: We found that PTEN loss was observed in 50.7%, and ERG positivity was detected in 41.8% of our cancerous samples. In HGPIN, PTEN loss appeared to be linked with a high-grade adjacent invasive carcinoma component which also displayed PTEN loss. As far as IDCP is concerned, ERG immunonegativity was correlated with adjacent high-grade invasive cancer, which was also ERG immunonegative. Conclusions: Our findings suggest that the clonal expansion of invasive cancer appears to be associated with distinct immunophenotypic cellular alterations of both early and late cancer-related histological lesions. Patients with PTEN loss in HGPIN in prostate biopsies should be closely monitored due to the increased likelihood of having an associated invasive high-grade carcinoma that may have not been sampled. Given the clinical significance that derives from PTEN expression in HGPIN lesions, we suggest the routine use of PTEN immunohistochemistry in prostate cancer biopsies in which HGPIN is the only finding.


Introduction
Prostate cancer represents the most common malignancy among men in the western world [1] and a highly biologically variable disease. In the USA, 268,490 new cases of prostate cancer were recorded during 2022, comprising 14% of all new cancer cases, and it is estimated that approximately 12.6 percent of men will be diagnosed with prostate cancer at some point during their lifetime [1]. Gleason scoring [1,2] of prostate cancer tissues remains the most widely applied and robust utility for assessing tumor progression and aggressiveness, as well as a trustworthy system upon which most decisions regarding patients' management are carried out. A considerable proportion of prostate cancers display a relatively indolent course without significant associated morbidity, as 5-year survival rates amount to 96.8%, according to the National Cancer Institute [1]. However, some cases follow a highly aggressive clinical course, leading to polymetastatic disease which eventually becomes treatment-resistant and ultimately lethal. Taking into account the considerable prevalence of prostate cancer in the male population as well, it becomes evident that, besides the Gleason grading pattern [2,3], there is an urgent need for the identification of novel predictive biomarkers to differentiate low-risk patients from those at increased risk for disease progression and metastasis. Despite the fact that our knowledge of genomic markers has greatly improved over the years, there is still a universal lack of predictive biomarkers that could be reliably used in clinical practice [4].
The ERG is located at chromosome 21q22.2 and it is a member of the erythroblast transformation-specific (ETS) family of transcriptions factors, a group of highly conserved molecules with an important role in multiple cellular events, including embryonic development, proliferation, and differentiation [2].
Under normal circumstances, ERG is not expressed in the epithelial cells of the prostate. However, increased expression is frequently present in patients with prostate cancer [3,4]. The over-expression of ERG is highly attributed to the TMPRSS2: ERG fusion, a recurrent chromosomal rearrangement that represents the most common molecular alteration of prostate cancer, approximately occurring in one-half of the cases [2,5]. TMRSS2: ERG fusion can occur due to genomic translocations or due to interstitial deletions of the intergenic region between the two genes. This genomic event results in the fusion of the ERG protooncogene with the androgen-driven promoter of the TMPRSS2, setting the transcription of the former under hormonal affection and resulting in its over-expression in cancer cells [5]. The exact molecular mechanisms underlying the pathogenic role of TMPRSS2: ERG fusion in prostate cancer are yet to be fully uncovered, but there is evidence suggesting its association with the clinical course of the disease [3,6].
The PTEN gene is located in chromosome 10, encoding an enzymatically active molecule that acts as a phosphatase, which impedes the activity of the PI3K/Akt pathway. PTEN has been shown to constitute the most frequently inactivated tumor suppressor gene in primary prostate cancer, and its loss has been associated with disease progression in both hormone-naive and castration-resistant prostate cancer [7]. Various methods, including immunohistochemistry, in situ and array comparative genomic hybridization, have been used to highlight the loss of PTEN function in prostate cancer [7,8]. Genomic deletion is the most commonly found mechanism of PTEN alteration, whereas point mutations are less frequent. A loss of PTEN expression in pathology specimens, as evidenced by the use of immunohistochemistry staining, has been correlated with low PTEN signaling at the molecular level, regarding DNA and mRNA [9,10]. Immunohistochemistry-targeting PTEN protein stains the nucleus and the cytoplasm of the basal and luminal prostatic cells, and loss of this staining pattern is frequently found in prostate cancer. Higher rates of PTEN loss have been associated with disease severity and progression along with poorer clinical outcomes [11].
The fusion of the androgen-regulated serine protease TMPRSS2 with the ETS family transcription factor ERG along with the inactivating rearrangements of the tumor suppressor gene phosphatase and tensin homolog (PTEN) are among the most common genetic alterations in prostate cancer [5]. PTEN loss has been linked to poor overall survival, adverse pathological findings and the development of castration-resistant and metastatic disease [6][7][8]. Meanwhile, TMPRSS2-ERG fusions represent an early event in prostate carcinogenesis and are present in approximately 36-78% of cases [9][10][11]. Although PTEN loss is less prevalent, it likely represents a more clinically relevant genetic rearrangement than TMPRSS2-ERG fusion in prostate cancer. The detection of the aforementioned alterations through immunohistochemistry has been proven as a reliable and useful technique in detecting the expression of ERG, a surrogate marker for TMPRSS2-ERG fusion, and of PTEN [9,[12][13][14][15].
Whereas the overexpression of ERG alone does not constitute a determinant for disease progression or overall survival [14,15], it could indicate the initiating trigger for the gene fusion upon other cancer cell regulatory pathways, such as a loss of PTEN, and thus reflect disease aggressiveness [16][17][18][19][20][21]. On the other hand, the loss of PTEN alone seems to add some value in predicting patient outcomes [22,23]. Nevertheless, there is more clinical significance when considering the combined effect of TMPRSS2-ERG fusion and PTEN deletion than either one alone [24,25]. Studies have shown that the ERG fusion-positive-PTEN-negative prostate cancers display unfavorable disease outcomes such as earlier biochemical recurrence [16,26] and cancer progression [27,28]. However, to date, there are few studies depicting disease-specific survival based on ERG and PTEN status.
In the present study, we aimed to observe the patterns of ERG immunoexpression and PTEN immunohistochemical loss in prostate cancer precursor lesions (HGPIN), invasive prostate carcinoma, as well as in IDCP in correlation with the immunohistochemical expression in the adjacent invasive carcinoma.

Patient Material Collection and Characterization
Following systematic research of the database of our department, formalin-fixed paraffin-embedded (FFPE) tissues from 151 radical prostatectomy specimens of Gleason scores 6-10 were retrieved from the pathology laboratory archives of the First Department of Pathology, School of Medicine, NKUA, "Laikon" General Hospital in Athens, Greece. Out of the 151 cases, 17 were excluded due to technical problems. Low-grade (Gleason grade pattern 3) prostate cancers represented 45.7% of our cases, whereas high-grade (Gleason grade patterns 4, 5) prostate cancers were 54.3% of the specimens (Table 1). HGPIN was detectable in 113 cases and IDCP in 35 cases. All tissue samples selected regarded retrospective cases of our department. Inclusion criteria were the presence of prostate acinar adenocarcinoma, HGPIN and IDCP.

Immunohistochemistry Procedure and Evaluation
Immunohistochemistry was carried out using standard procedures in all tested specimens. Firstly, the sections were stained with the following antibodies on a Dako system, according to the manufacturer's protocol. The sections were stained based on the double-staining protocol of Leica Biosystems in the Bond-III fully automated stainer, with antibodies against p63 (clone 4A4 Biocare/at dilution 1:100, Pacheco, CA 94553, USA), AMACR (monoclonal rabbit, clone 13H4, Dako/at dilution 1:50, Santa Clara, CA 95051, USA), ERG (clone EP111, ready to use, Dako, Santa Clara, CA 95051, USA) and PTEN (monoclonal mouse, clone 6H2.1, Dako/at dilution 1:200, Santa Clara, CA 95051, USA). Antigen retrieval was performed at pH 6. The Envision (Dako) visualization system was used. DAB (3,3-diaminobenzidine) was used as a chromogen and hematoxylin as counterstain. A double staining protocol was used both for PTEN (red chromogen)-ERG (brown chromogen) and for AMACR (red chromogen)-p63 (brown chromogen). The evaluation of the slides was carried out by two experienced pathologists (D.G. and A.C.L) independently who were initially blinded to the clinicopathological data of the patients as well as to each other's results. After a literature review [29], a tissue specimen was considered to have PTEN protein loss if the intensity of cytoplasmic and nuclear staining was noticeably reduced or completely negative across >10% of tumor cells in relation to the surrounding benign glands or stroma, which were used as internal positive controls. In cases where PTEN was lost in >10 and <100% of the tumor cells, the specimen was considered to have heterogeneous PTEN loss (score 1). Otherwise, when the specimen showed PTEN loss in 100% of the tumor glands, it was considered to have homogeneous PTEN loss (score 0).
PTEN preservation across the entirety of the specimen was considered as score 2. ERG expression was considered to be positive when any tumor cells displayed nuclear ERG expression and negative if there was a complete absence of ERG expression. Endothelial cells were used as internal positive controls [29]. Each immunomarker was separately assessed in each Gleason pattern. Double staining with p63 (brown chromogen) and AMACR (red chromogen) was performed to confirm the diagnosis of prostate adenocarcinoma. AMACR was considered positive in the carcinoma when it displayed a strong granular cytoplasmic pattern in the luminal cells of the neoplastic prostate glands, whereas the absence of p63 in the same glands was considered diagnostic. The nuclear staining of the basal cell layer of the adjacent normal prostate glands was used as an internal control. Our group then proceeded to correlate the staining patterns of the aforementioned markers with a multitude of histopathological characteristics, including the Gleason score of prostate carcinoma specimens and the presence of HGPIN and IDCP. As HGPIN we define the proliferation of atypical secretory cells within prostatic glands. It is thought that HGPIN is the earliest histologically recognizable precursor of invasive adenocarcinoma of the prostate. On the other hand, IDCP is a neoplastic epithelial proliferation involving pre-existing, generally expanded, duct-acinar structures and characterized by architectural and cytological atypia beyond what is acceptable for HGPIN 2 .

Statistical Analysis
Statistical analysis was performed with IBM-SPSS v26. The level of statistical significance was set at 5% (α = 0.05). Pearson's χ 2 test (Pearson's chi-square test for categorical data) with continuity correction for 2 × 2 tables was performed in order to check for a statistically significant relation between studied markers.

ERG and PTEN Associations in Prostate Carcinoma
Statistical analysis of PTEN and ERG expression in prostate carcinoma provided statistically significant results (p < 0.0001) (Figure 1). More specifically, when ERG was negative in prostate carcinoma, PTEN displayed a heterogeneous loss (score 1). Meanwhile, the positive expression of ERG proved to be significantly associated with a homogeneous loss (score 0) of PTEN in prostate cancer ( Figure 2). negative in prostate carcinoma, PTEN displayed a heterogeneous loss (score 1). Meanwhile, the positive expression of ERG proved to be significantly associated with a homogeneous loss (score 0) of PTEN in prostate cancer ( Figure 2).   negative in prostate carcinoma, PTEN displayed a heterogeneous loss (score 1). Meanwhile, the positive expression of ERG proved to be significantly associated with a homogeneous loss (score 0) of PTEN in prostate cancer ( Figure 2).

Association of ERG Expression with Gleason Grade Pattern in Prostate Cancer
The Gleason grade pattern displayed a statistically significant association with the expression of ERG (p < 0.0001) ( Figure 3). Characteristically, high-grade carcinomas of Gleason grade patterns 4 and 5 were ERG-negative, whereas low-grade prostate carcinomas of Gleason grade pattern 3 were ERG-positive (Figure 4).

Association of ERG Expression with Gleason Grade Pattern in Prostate Cancer
The Gleason grade pattern displayed a statistically significant association with the expression of ERG (p < 0.0001) (Figure 3). Characteristically, high-grade carcinomas of Gleason grade patterns 4 and 5 were ERG-negative, whereas low-grade prostate carcinomas of Gleason grade pattern 3 were ERG-positive (Figure 4).

Association of ERG Expression with Gleason Grade Pattern in Prostate Cancer
The Gleason grade pattern displayed a statistically significant association with expression of ERG (p < 0.0001) (Figure 3). Characteristically, high-grade carcinomas Gleason grade patterns 4 and 5 were ERG-negative, whereas low-grade prostate carci mas of Gleason grade pattern 3 were ERG-positive (Figure 4).

PTEN Expression in HGPIN and Its Association with the Adjacent Invasive Prostate Cancer
In cases with HGPIN, the Gleason grade pattern of the co-existent adjacent prostate carcinoma demonstrated statistically significant differences when compared to PTEN expression in the HGPIN foci (p = 0.05) ( Figure 5). Cases showing PTEN homogeneous loss in HGPIN (score 0) were associated with high-grade adjacent invasive carcinoma (Gleason grade pattern 4 or 5) and a PTEN score of 0 ( Figure 6). In cases where PTEN was preserved in the HGPIN (score 2), the adjacent prostate carcinoma had low-grade features (Gleason grade pattern 3) and retained its PTEN expression (Figure 7). When the PTEN was heterogeneously lost in HGPIN (score 1), then 51% of adjacent invasive prostate cancer was low-grade and had a PTEN score of 2.

PTEN Expression in HGPIN and Its Association with the Adjacent Invasive Prostate Cancer
In cases with HGPIN, the Gleason grade pattern of the co-existent adjacent prostate carcinoma demonstrated statistically significant differences when compared to PTEN expression in the HGPIN foci (p = 0.05) ( Figure 5). Cases showing PTEN homogeneous loss in HGPIN (score 0) were associated with high-grade adjacent invasive carcinoma (Gleason grade pattern 4 or 5) and a PTEN score of 0 ( Figure 6). In cases where PTEN was preserved in the HGPIN (score 2), the adjacent prostate carcinoma had low-grade features (Gleason grade pattern 3) and retained its PTEN expression (Figure 7). When the PTEN was heterogeneously lost in HGPIN (score 1), then 51% of adjacent invasive prostate cancer was low-grade and had a PTEN score of 2.

ERG Expression in IDCP and Its Association with the Adjacent Invasive Prostate Carcinoma
In IDCP, the Gleason grade pattern of the adjacent invasive component showed statistically significant differences when analyzed in relation to the ERG expression (p = 0.002) (Figure 8). More specifically, when IDCP expressed ERG, the adjacent invasive prostate cancer was also ERG-positive with low-grade features (Figure 9). On the contrary, when IDCP was ERG-negative, the adjacent prostate cancer was also ERG-negative and displayed high-grade morphology. As such, we observed that the pattern of ERG staining in IDCP is identical to that of the adjacent invasive cancer.

AMACR Expression in Prostate Cancer and Its Association with PTEN
AMACR expression in prostate cancer was correlated with PTEN expression in prostate cancer (p = 0.039). More precisely, when AMACR was negative, the PTEN displayed a heterogeneous loss (score 1) expression pattern, whereas when AMACR was positive, PTEN demonstrated a homogeneous loss (score 0).   In IDCP, the Gleason grade pattern of the adjacent invasive component showed statistically significant differences when analyzed in relation to the ERG expression (p = 0.002) ( Figure  8). More specifically, when IDCP expressed ERG, the adjacent invasive prostate cancer was also ERG-positive with low-grade features (Figure 9). On the contrary, when IDCP was ERG-negative, the adjacent prostate cancer was also ERG-negative and displayed high-grade morphology. As such, we observed that the pattern of ERG staining in IDCP is identical to that of the adjacent invasive cancer.    Figure 9. Simultaneous ERG nuclear immunopositivity in intraductal carcinoma and in adjacent low-grade invasive carcinoma (×200).

AMACR Expression in Prostate Cancer and Its Association with PTEN
AMACR expression in prostate cancer was correlated with PTEN expression in prostate cancer (p = 0.039). More precisely, when AMACR was negative, the PTEN displayed a heterogeneous loss (score 1) expression pattern, whereas when AMACR was positive, PTEN demonstrated a homogeneous loss (score 0).

Discussion
PTEN deletion and ERG rearrangement exhibit a significant role in the pathogenesis and clinical course of prostate cancer, and both represent frequent genetic alterations [8,28,[30][31][32]. Among others, the contribution of PTEN deletion to prostate cancer tumorigenesis and progression was underlined in a large study that incorporated more than 4.5 thousand specimens and correlated its presence to advanced tumor stages, higher Gleason score grades, positive lymph nodes and androgen-independent disease, linking it to an overall aggressive tumor phenotype [12]. The synergistic action of ERG rearrangement in carcinogenesis has also been explored, with researchers demonstrating the frequency of coexistence of the two phenomena in malignant tissues. Specifically, TMPRSS2-ERG fusions, which constitute by far the most frequent aberration (>97%), were shown to strongly associate with PTEN deletions [7,13,14].
In the current study, we demonstrated that homogeneous PTEN loss was associated with immunohistochemical ERG positivity; the latter being mostly linked to low-grade prostate carcinomas. On the contrary, PTEN heterogeneous loss was statistically linked with ERG negativity, which was in turn related to high-grade prostate carcinomas.
Genomic PTEN loss has been associated, as previously mentioned, with prostate cancer progression, aggressiveness and poor prognosis [33][34][35]. Among others, researchers have linked this molecular feature with increased tumor grade and even increased lethality in patients who underwent radical prostatectomy [33,34], with the risk being more pronounced when tumors exhibited a concurrent loss of ERG immunoexpression [35]. In our study, we observed that PTEN homogeneous loss was observed in 50.7% of cases which has been reported to correspond to PTEN gene homozygous deletion [18]. Meanwhile, PTEN heterogeneous loss was present in 41.8% of cancer specimens, which most likely corresponds to a heterogeneous or subclonal PTEN gene deletion [18]. Nevertheless, we did not manage to establish an association of PTEN loss (when evaluating PTEN loss alone) with prostate carcinoma aggressiveness, perhaps due to the relatively small number of our specimens. On the contrary, when assessing the ERG-PTEN association, we observed that PTEN homogeneous loss was linked to ERG positivity which was mostly found in low-grade prostate carcinomas, and PTEN heterogeneous loss was linked to a negative ERG expression which constitutes a key finding in high-grade carcinomas. Therefore, we can support that immunohistochemical homogeneous PTEN loss is predominantly linked to ERG fusions (ERG-positive immunostaining), a mechanism mostly seen in prostate adenocarcinomas of low grade. It has been claimed that PTEN alterations typically develop subsequent to ERG fusions [10]. More specifically, Krohn et al. in the aforementioned study performed a tissue microarray analysis for PTEN alterations (including deletions and breaks) and compared PTEN and ERG status in the same tumor areas [10]. The data from their study showed the substantial heterogeneity in PTEN aberrations in prostate cancer and advocates that ERG activation is a key driver of such abnormalities. On the contrary, PTEN alterations, as a primary lesion in cancer, did not display an increased risk for developing TMPRSS2: ERG fusions [10].
PTEN and ERG immunoexpression has been investigated with the aim to explore the utility of the aforementioned biomarkers in the challenging task of distinguishing between HGPIN and IDCP in the setting of prostate biopsies. In our study, a PTEN immunostaining pattern in HGPIN was identical to that of the adjacent invasive carcinoma tissues when the PTEN pattern was either score 0 or 2 in high-grade or low-grade carcinomas, respectively. Therefore, we demonstrated, for the first time to our knowledge, that HGPIN cases showing homogeneous PTEN loss (score 0) are associated with the presence of an adjacent high-grade invasive carcinoma focus, whereas the preservation of PTEN in the HGPIN component was linked to an adjacent low-grade invasive prostate carcinoma. This is partly explained as HGPIN represents a precursor lesion, so it is speculated that in these cases, the same neoplastic clone gives rise to and progresses to invasive prostate carcinoma with similar genomic characteristics. The same applies for the cases of PTEN preservation in the HGPIN. However, what is of particular interest is the fact that 51% of HGPIN cases that displayed heterogeneous PTEN loss were linked to an adjacent low-grade invasive prostate adenocarcinoma. In these HGPIN cases with heterogeneous PTEN loss, the adjacent carcinoma displayed PTEN expression preservation (score 2) only in its low-grade form. So, complementary to what we observed in the former cases, in these PTEN score 1 HGPIN cases, we see that probably the HGPIN clone that gives rise to the adjacent low-grade invasive carcinoma is the one that preserves PTEN expression. These findings further underpin the model of clonal progression in prostate cancer [36]. With regard to the clone of low-grade adjacent cancer, cancerous cells are likely to arise from those cells that preserve PTEN immunoexpression within HGPIN; on the contrary, high-grade cancer cells might derive from those cells of HGPIN which lack PTEN immunoexpression. The individual cells within a precursor lesion acquire potentially additional genomic abnormalities, resulting in subclonal tumor cell populations and invasion [36].
As far as ERG expression is concerned, we also observed an association with IDCP, a late event in prostate cancer progression. The incidence of IDCP amounts to an estimated 20% of radical prostatectomies [15]. The presence of IDCP constitutes a very important finding in prostate carcinoma, as it heavily influences patients' prognosis and management. A multitude of studies has underlined its association with poor outcomes [15][16][17][18] as well as with many adverse clinicopathological characteristics, including positive lymph nodes [19,20], distant metastasis [16,21] and castration-resistant disease [22]. Researchers have attempted to shed light on the link between PTEN loss and ERG rearrangements with IDPC, as all three phenomena represent important factors in the progression of prostate adenocarcinoma. More specifically, two studies reported the coexistence of IDPC with ERG rearrangements in 94% [23] and 75% [24] of cases, while PTEN loss has been observed in 84% [25] and 72% [26] of IDCP specimens. Characteristically, in our study, when IDCP expressed ERG, the adjacent invasive carcinoma was also ERG-positive and paradoxically low-grade, whereas when IDCP lacked ERG expression, the abutting invasive component was ERG-negative and high-grade. According to the most widely accepted IDCP theory, IDCP is by definition a high-grade lesion which represents an end-stage spread of highgrade invasive carcinoma into ducts and acini [2,37]. Therefore, it is reasonable that the ducts colonized by an elsewhere-developed invasive component will show the same clonal neoplastic traits. According to our findings, the ERG immunopositivity of the IDCP-adjacent low-grade invasive carcinoma cannot be a fundamental immunophenotypic characteristic related to the cancerous clone within IDCP.
Moreover, although we were not able to obtain a statistically significant result, probably due to the small number of our cohort, we observed that 43.8% of our IDCP cases displayed a homogeneous PTEN loss. This is in line with findings by Haffner et al., according to which, IDCP occurs subsequent to invasive carcinoma development [36].
Finally, AMACR was positive in 95.5% of our prostate cancer samples and negative in the remaining 4.5%. Interestingly, AMACR-negative expression in prostate cancer was associated with a heterogeneous PTEN loss, while its positive expression was linked to PTEN homogeneous loss. There have been a few studies discussing the potential oncogenic role of AMACR in prostate adenocarcinomas [38,39], so in conjunction with its strong association with PTEN homogeneous loss, it could prove to be an important candidate biomarker for prostate cancer progression and prognosis apart from its well-established diagnostic significance. This is in keeping with other groups of researchers that have also linked increased AMACR expression to higher-grade tumors [27,28] as well as to patients' age [27]. Moreover, AMACR-positive HGPIN lesions in biopsies that did not harbor prostate carcinoma have exhibited a five-fold increased risk of cancer diagnosis in subsequent biopsies, compared to patients with AMACR-negative HGPIN [29]. It is evident, however, that further studies are needed in order to establish AMACR as a strong predictor of clinical outcome.
All in all, our study confirmed the strong relationship between ERG rearrangement and PTEN deletion reinforcing the theory of an interactive or cooperative role of the two phenomena in the biology of prostate cancer. Furthermore, we displayed a possible clonal relationship between the neoplastic cells of precursor lesions (HGPIN) and the adjacent invasive carcinoma based on PTEN expression patterns, and among IDCP and the abutting invasive cancer, based on ERG expression. According to our findings and the existing knowledge regarding the biological role of PTEN loss and ERG rearrangements, we suggest that patients with HGPIN in prostate biopsies displaying homogeneous PTEN loss would benefit from a more intensive clinical follow-up. In those patients, should a coexistent invasive cancer be found, it would most likely be of high grade and thus clinically significant. Our findings underpin our understanding of the sequence of molecular and morphological events and of complex clonal relationships that arise during prostate cancer progression. Moreover, we provide evidence that would help optimize the histological risk stratification of prostate cancer patients based on the immunohistochemical evaluation of precursor lesions in prostate biopsies. Among the limitations of our study is the lack of a prolonged follow-up of more than 5 years, since the 5-year survival rate of prostate cancer is nearly 100% [40]. Furthermore, the fact that we did not use in situ molecular techniques (hybridization) in order to establish our immunohistochemical results represents another limitation of our study.

Conclusions
Overall, it is evident that disorders of PTEN gene expression seem to be associated with an early stage of prostate carcinoma progression as their significance appears to be focused on HGPIN lesions of our samples. On the contrary, ERG immunoexpression appears to correlate with the tumor grade of invasive carcinoma with its positivity linked to low-grade tumors. What appears, however, to be of importance and clinical significance derives from PTEN expression in HGPIN lesions. Therefore, we suggest the routine use of PTEN immunohistochemistry in prostate cancer biopsies when HGPIN lesions are encountered as the only finding. Informed Consent Statement: Patient consent was waived due to the retrospective nature of the study involving archival tissue specimens.

Data Availability Statement:
No new data were generated.

Conflicts of Interest:
The authors declare no conflict of interest.