Refining prostate cancer diagnosis: correlative analysis of Gleason scores and immunohistochemical markers

Abstract

Background

Distinguishing prostate cancer (PCa) from benign prostatic tissue can be challenging in equivocal cases. Gleason grading remains the gold standard, with immunohistochemical (IHC) markers such as androgen receptor (AR), prostate-specific antigen (PSA), and α-methylacyl-CoA racemase (P504S/AMACR) serving as established adjuncts. This study aimed to confirm their diagnostic utility in a retrospective Chinese cohort of PCa and benign controls (primarily benign prostatic hyperplasia [BPH]).

Methods

This retrospective analysis included 147 patients (88 PCa, 59 BPH) evaluated between 2020 and 2024 at two Chinese institutions. Gleason scores were extracted from pathology reports. IHC for AR, PSA, P504S (AMACR), and hematopoietic cell kinase (HCK) was performed routinely when clinically indicated. AR staining (when available) was assessed exclusively in tumor cell nuclei, with positivity defined qualitatively by any discernible nuclear staining (typically >1% of cells, including weak intensity). Due to the retrospective design and routine practice, no uniform cutoff or re-scoring was applied. AR and PSA results were available in most cases, whereas P504S was assessed in a subset (40/88 PCa; 19/59 benign), typically diagnostically ambiguous samples. Marker positivity was compared between PCa and benign tissue and correlated with Gleason score components and patient age.

Results

AR and PSA positivity were significantly higher in PCa than benign tissue (AR: 25.0% vs. 8.5%, p=0.011; PSA: 40.9% vs. 20.3%, p=0.009), consistent with prior reports. The lower AR rate in PCa (25.0%) likely reflects routine clinical reporting variability and thresholds. In assessed subsets, P504S was positive in 100% (40/40) of PCa cases versus 26.3% (5/19) of benign cases (p<0.0001), aligning with its established sensitivity. AR positivity peaked in intermediate Gleason grades but declined in high-grade tumors, while PSA showed a modest increasing trend with grade. HCK was uniformly negative in all assessed cases, supporting its limited diagnostic relevance. No significant age association was observed with Gleason categories.

Conclusion

This retrospective analysis corroborates the complementary roles of P504S (in assessed subsets), AR, and PSA as valuable IHC adjuncts to Gleason grading for evaluating prostate tissue, including differentiation from benign conditions such as BPH. HCK negativity further supports its deprioritization in diagnostic panels.

Introduction

Prostate diseases represent a major health burden among aging men, with prostate cancer (PCa) being the second most commonly diagnosed malignancy in men worldwide [1, 2]. Globally, PCa accounts for over 1.4 million new cases and approximately 375,000 deaths annually, with incidence rates highest in developed regions due to widespread prostate-specific antigen (PSA) screening [1]. Risk factors include advanced age (median diagnosis at 66 years), family history, genetic predispositions (e.g., BRCA2 mutations), and lifestyle elements such as diet and obesity [1, 2]. While localized PCa is often indolent, high-risk disease can progress to metastatic castration-resistant PCa (mCRPC), which remains incurable and carries a dismal prognosis.

While prostate cancer can be aggressive and life-threatening, benign prostatic hyperplasia (BPH) is a non-malignant enlargement of the prostate that often presents with overlapping clinical symptoms such as urinary obstruction [3, 4]. The challenge in distinguishing between these two entities—especially in early or equivocal cases—necessitates the use of reliable histopathological and molecular markers [5]. However, while classic nodular BPH is often straightforward to differentiate from cancer, the more typical diagnostic difficulty in needle biopsies involves separating adenocarcinoma from its histologic mimickers (e.g., adenosis, partial atrophy, basal cell hyperplasia). The benign cohort in this study consists primarily of classic BPH cases from surgical specimens. Basal cell markers (e.g., p63 and CK5/6) are routinely used to demonstrate loss of the basal layer in adenocarcinoma and are particularly valuable for excluding mimickers.

Histologically, PCa is characterized by glandular architecture disruption, assessed via the Gleason scoring system, which remains the cornerstone for classification [6, 7]. It evaluates the architectural patterns of prostate tumor cells and assigns a primary and secondary score based on the two most predominant patterns. The resulting total Gleason score is strongly correlated with tumor aggressiveness and clinical outcome. The International Society of Urological Pathology (ISUP) grade groups further refine this into five prognostic categories [1]. Emerging histopathological variants, such as cribriform pattern 4 and intraductal carcinoma, are associated with worse outcomes and may influence risk stratification [3]. Novel histopathological biomarkers, including CD169 + macrophages (indicative of immune infiltration), neuropilin-1 (angiogenesis regulator), cofilin-1 (actin dynamics), interleukin-17 (inflammation), STAT3/LIMK1 (signaling pathways), CD15 (adhesion), AMACR (metabolism), PSMA (membrane antigen), Appl1/Sortilin/Syndecan-1 (endosomal trafficking), and p63 (basal cell integrity), show promise for enhancing prognostic accuracy [3].

Diagnostic approaches for PCa rely on serum PSA screening, digital rectal examination, and prostate biopsy, but concordance between biopsy and radical prostatectomy (RP) Gleason scores remains imperfect, with upgrading/downgrading rates affecting treatment decisions [4, 5]. Systematic transrectal ultrasound-guided biopsy (TRUS-Bx) has been the standard, but it misses up to 30% of clinically significant PCa (csPCa) due to sampling errors [4]. Targeted biopsies, guided by multiparametric MRI (mpMRI), improve detection of csPCa by focusing on suspicious lesions (PI-RADS ≥ 3), but may risk downgrading in high-grade cases (e.g., GG4) if systematic sampling is omitted [5, 6]. Comparative studies show targeted biopsy alone may underestimate Gleason scores compared to combined targeted-systematic approaches, particularly in intermediate-risk patients [5]. Transperineal biopsy offers reduced infection risk but similar accuracy [6].

Advanced imaging plays a pivotal role in guiding tissue acquisition and staging. mpMRI, incorporating T2-weighted, diffusion-weighted, and dynamic contrast-enhanced sequences, has revolutionized biopsy targeting, reducing unnecessary procedures by 25–30% and improving csPCa detection [6]. Prostate-specific membrane antigen (PSMA) PET/CT, using 68Ga-PSMA ligands, excels in nodal and metastatic staging, with 92% accuracy in high-risk PCa but lower (64%) in aggressive variants like ductal adenocarcinoma [7]. In active surveillance (AS) protocols, PSMA PET/CT aids reclassification, identifying occult csPCa with high sensitivity (SUVmax cutoff > 5.3) [8].

Novel molecular findings further inform PCa biology and diagnostics. Common alterations include TMPRSS2-ERG fusions (50% of cases), PTEN loss (driving aggressiveness), and lncRNA dysregulation, such as H19 overexpression promoting metastasis via cell adhesion molecule circuitry (e.g., E-cadherin suppression) [1, 2]. Epidrugs like GSK-J4 target H19 to inhibit epithelial-mesenchymal transition and metastatic progression [2]. PD-L1 expression correlates positively with higher Gleason grade groups, suggesting immune checkpoint roles in advanced disease [9].

However, Gleason scoring alone may not fully capture the biological behavior of prostate tumors, particularly in borderline or low-volume cancers. In this context, immunohistochemical (IHC) markers such as androgen receptor (AR) [8, 9], prostate-specific antigen (PSA) [10, 11], α-methylacyl-CoA racemase (P504S/AMACR), and others—serve as routine adjuncts to support diagnosis and differentiation from benign mimics such as BPH. Basal cell markers (e.g., p63 and CK5/6) are routinely used to demonstrate loss of the basal layer in adenocarcinoma and are particularly valuable for excluding mimickers.

This retrospective study from two Chinese tertiary centers aimed to confirm the diagnostic utility of these established IHC markers (AR, PSA, P504S, and HCK) in a real-world cohort, while exploring their correlations with Gleason score components and patient age.

Methods

Study design and setting

This was a retrospective cohort study conducted at Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, a tertiary care academic medical center. Given its retrospective design and use of anonymized archival pathology data, the requirement for informed consent was waived by the ethics committee. The study analyzed prostate tissue specimens collected between January 2020 and December 2024. The objective was to evaluate the histopathological and immunohistochemical characteristics of benign and malignant prostate conditions, focusing on the expression of androgen receptor (AR), HCK, and prostate-specific antigen (PSA), and their relationship with Gleason scores and patient demographics.

Patient selection

Patients were identified through the hospital pathology database. Inclusion criteria were:

1. Male patients aged ≥ 18 years;

2. Histologically confirmed diagnosis of either prostate adenocarcinoma or benign prostatic hyperplasia (BPH);

3. Availability of complete pathology reports, including Gleason scoring (for cancer cases);

4. Immunohistochemical (IHC) staining results for AR, HCK, and PSA.

Exclusion criteria included incomplete pathology records, absence of IHC results, or ambiguous diagnostic classification. A total of 147 patients met the inclusion criteria, comprising 88 cases of prostate cancer (PCa) and 59 cases of BPH.

Histopathological analysis

Tissue specimens were formalin-fixed, paraffin-embedded, and stained with hematoxylin and eosin (H&E) for routine diagnostic evaluation. Prostate cancer specimens were graded using the Gleason grading system, following the 2014 International Society of Urological Pathology (ISUP) recommendations. For each case, the primary Gleason pattern (most predominant histological pattern), the secondary Gleason pattern (second most predominant), and the total Gleason score (sum of primary and secondary) were recorded from pathology reports. In cases where only the total score was provided, trained reviewers extracted primary and secondary components from the diagnostic text.

Immunohistochemistry (IHC)

IHC was performed on representative tissue sections using commercially available monoclonal antibodies for AR, HCK, P504S, and PSA. The IHC protocol included:

1. Antigen retrieval using citrate buffer (pH 6.0) or EDTA (pH 9.0);

2. Incubation with primary antibodies followed by HRP-conjugated secondary detection;

3. Visualization with DAB chromogen and counterstaining with hematoxylin.

IHC slides were independently reviewed by two pathologists blinded to clinical data. Marker expression was assessed qualitatively and classified as positive (+) or negative (–) based on nuclear staining for AR, and cytoplasmic staining for PSA and HCK. Cases with ambiguous or equivocal staining were discussed and resolved by consensus. The representative images of HE and IHC staining of tissue sections with Gleason score of 4 and 5 are shown in Fig. 1.

Fig. 1.

Representative histological and immunohistochemical features of prostate adenocarcinoma with Gleason patterns 4 and 5

Data extraction and variable definitions

Data were extracted manually from digital pathology reports and entered into a structured spreadsheet. Variables included patient age, diagnosis (PCa vs. BPH), AR/HCK/PSA status, and Gleason scores. Gleason scores were standardized and parsed into numeric primary, secondary, and total components using pattern-matching functions (e.g., from strings such as “7 (3 + 4)”). Marker status was binarized (1 = positive, 0 = negative) for statistical analysis. For reporting purposes, the term “positivity rate” was defined as the proportion of cases showing positive immunohistochemical staining for a given marker, calculated as the number of positive cases divided by the total number of evaluable cases in the specified group, expressed as a percentage.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 10.0. Summary statistics (mean, standard deviation, median, interquartile range) were computed for continuous variables such as age. Categorical variables, including AR, HCK, and PSA positivity, were summarized as counts and percentages. Comparisons between two independent groups (PCa vs. BPH) for categorical variables were conducted using the chi-square test or Fisher’s exact test, depending on expected cell counts. Comparisons of continuous variables between two groups were performed using the independent samples t-test for normally distributed data or the Mann–Whitney U test for non-normally distributed data.

Correlations between IHC marker positivity (AR, PSA) and Gleason score components (primary, secondary, and total) or patient age were evaluated using Spearman’s rank correlation coefficient.

Results

Demographic overview

The cohort consisted of a total of 88 patients diagnosed with prostate cancer (PCa) and 59 patients diagnosed with benign prostatic hyperplasia (BPH). The patients’ ages ranged from 58 to 82 years in the PCa group and from 52 to 85 years in the BPH group. The average age for prostate cancer patients was 71.3 years, while for BPH patients, the mean age was 70.3 years. The distribution of ages in both groups showed that the majority of patients were in their 60s to 70s. A summary of the demographic characteristics, including the age range, mean age, and percentiles, is presented in Table 1.

Table 1.

Demographic overview

Group	Prostate Cancer (PCa)	Benign Prostatic Hyperplasia (BPH)
Number of Patients	88	59
Mean Age (Years)	71.32	70.27
Age Range (Years)	58–82	52–85
25th Percentile Age (Years)	67	67
Median Age (Years)	71	70
75th Percentile Age (Years)	75.25	74.5

Correlation between AR and Gleason scores

The expression of androgen receptor (AR) was evaluated across Gleason categories. For primary Gleason scores, AR positivity was highest in tumors with score 3 (26.8%), decreased to 15.4% in score 4, and was 25.0% in score 5. For secondary Gleason scores, AR positivity was 22.2% for score 3, increased to 27.6% for score 4, and was 16.7% for score 5. When considering total Gleason scores, AR positivity was 18.2% in patients with score 6, rose to 26.7% in those with score 7, and reached 50.0% in the limited subset of score 8 cases, while no AR positivity was observed in tumors with a total score of 9. These findings are summarized in Table 2, which details AR positivity rates stratified by primary, secondary, and total Gleason scores.

Table 2.

AR positivity rates by Gleason score

Gleason Score	Primary	Secondary	Total
3	26.8% (19/71)	22.2% (4/18)	–
4	15.4% (2/13)	27.6% (16/58)	–
5	25.0% (1/4)	16.7% (2/12)	–
6	–	–	18.2% (2/11)
7	–	–	26.7% (16/60)
8	–	–	50.0% (4/8)
9	–	–	0.0% (0/9)

Correlation between PSA and Gleason scores

Prostate-specific antigen (PSA) expression was analyzed across different Gleason categories. For primary Gleason scores, PSA positivity was 42.3% in tumors with score 3, decreased to 30.8% in score 4, and rose to 50.0% in those with score 5. For secondary Gleason scores, PSA positivity was 33.3% in score 3, increased to 46.6% in score 4, and was 25.0% in score 5. When considering total Gleason scores, PSA positivity was 18.2% in tumors with a score of 6, markedly higher in those with score 7 (48.3%), intermediate in score 8 (37.5%), and declined again in score 9 (22.2%). These results are summarized in Table 3.

Table 3.

PSA positivity rates by Gleason score

Gleason Score	Primary	Secondary	Total
3	42.3% (30/71)	33.3% (6/18)	–
4	30.8% (4/13)	46.6% (27/58)	–
5	50.0% (2/4)	25.0% (3/12)	–
6	–	–	18.2% (2/11)
7	–	–	48.3% (29/60)
8	–	–	37.5% (3/8)
9	–	–	22.2% (2/9)

Correlation between P504S and Gleason scores

P504S (AMACR) expression was analyzed across different Gleason categories. For primary Gleason scores, positivity was 47.9% (34/71) in tumors with score 3, 30.8% (4/13) in score 4, and 50.0% (2/4) in score 5. For secondary scores, P504S positivity was 27.8% (5/18) in score 3, 51.7% (30/58) in score 4, and 41.7% (5/12) in score 5. When considering total Gleason scores, positivity was 18.2% (2/11) in tumors with score 6, 51.7% (31/60) in score 7, 50.0% (4/8) in score 8, and 33.3% (3/9) in score 9.

Age distribution by Gleason score in prostate cancer

We analyzed age distributions across primary, secondary, and total Gleason scores. For primary Gleason scores, patients with score 3 (n = 71) had a mean age of 71.1 ± 5.7 years, those with score 4 (n = 13) averaged 72.5 ± 4.5 years, and those with score 5 (n = 4) averaged 70.8 ± 8.3 years.For secondary Gleason scores, patients with score 3 (n = 18) had a mean age of 72.4 ± 5.9 years, score 4 (n = 58) averaged 70.7 ± 5.5 years, and score 5 (n = 12) averaged 72.7 ± 5.9 years.For total Gleason scores, patients with score 6 (n = 11) had a mean age of 71.1 ± 6.1 years, those with score 7 (n = 60) averaged 71.2 ± 5.4 years, score 8 cases (n = 8) averaged 73.0 ± 7.4 years, and score 9 cases (n = 9) averaged 70.9 ± 5.3 years (See Fig. 2).

Fig. 2.

Age distribution by Gleason score components in prostate cancer patients. A Age distribution stratified by primary Gleason score. Median age was similar across scores, with modest variation observed particularly in patients with higher scores. B Age distribution by secondary Gleason score. Patients with score 4 were the most prevalent, and no clear trend in age differences was observed across scores. C Age distribution across total Gleason scores. No significant differences in age were observed across Gleason groups

Comparison of AR, PSA, and P504S positivity between prostate cancer and benign prostatic hyperplasia

We compared immunohistochemical expression of androgen receptor (AR), prostate-specific antigen (PSA), and P504S between the prostate cancer (PCa) and benign prostatic hyperplasia (BPH) cohorts. AR was positive in 22/88 PCa cases (25.0%) versus 5/59 BPH cases (8.5%) (Fig. 3A). This difference was statistically significant (χ²=6.43, p = 0.011), corresponding to an estimated odds ratio (OR) of 3.60 (95% CI 1.28–10.14), indicating AR expression was ~ 3.6× more likely in malignant than benign specimens. Additionally, PSA was positive in 36/88 PCa cases (40.9%) and 12/59 BPH cases (20.3%) (Fig. 3B). The difference was also statistically significant (χ²=6.80, p = 0.009), with an OR of 2.71 (95% CI 1.26–5.82), suggesting PSA expression was roughly 2.7× more likely in PCa than BPH. Furthermore, P504S positivity was observed in 40/40 evaluable PCa cases (100%), compared to only 5/19 (26.3%) in BPH (Fig. 3C). This striking difference underscores the superior diagnostic relevance of P504S in distinguishing malignant from benign prostate lesions, supporting its inclusion alongside AR and PSA in immunohistochemical diagnostic panels .

Fig. 3.

Comparison of AR, PSA, and P504S positivity rates between prostate cancer and benign prostatic hyperplasia. A Bar graph showing the percentage of AR-positive cases in prostate cancer (PCa) and benign prostatic hyperplasia (BPH). AR expression was significantly higher in PCa (25.0%) compared to BPH (8.5%) (p = 0.011). B Bar graph showing the percentage of PSA-positive cases in PCa and BPH. PSA expression was also significantly higher in PCa (40.9%) than in BPH (20.3%) (p = 0.009). Data represent proportions of positive cases; statistical significance assessed using chi-square test. C Bar graph showing the percentage and case counts of P504S-positive specimens in PCa and BPH. P504S was universally positive in PCa (100%, 40/40) but only detected in a minority of BPH cases (26.3%, 5/19), highlighting its strong discriminatory power between malignant and benign lesions

Discussion

This retrospective analysis of 147 patients (88 with prostate cancer [PCa] and 59 with benign prostatic hyperplasia [BPH]) from two Chinese tertiary centers confirms the well-established diagnostic utility of α-methylacyl-CoA racemase (P504S/AMACR), androgen receptor (AR), and prostate-specific antigen (PSA) immunohistochemistry as adjuncts to Gleason grading in the evaluation of prostate cancer relative to benign prostatic tissue (primarily classic nodular BPH).

The observed near-universal positivity of P504S in PCa (100%, 40/40 assessed cases) and markedly lower expression in BPH (26.3%, 5/19) aligns closely with extensive prior literature that has established P504S as a highly sensitive and relatively specific marker for prostatic adenocarcinoma [12]. Similarly, the significantly higher rates of AR and PSA positivity in malignant compared with benign tissue (predominantly BPH) corroborate long-standing reports on the supportive role of these markers in diagnostically challenging cases [13].

In contrast, PSA showed a modest but positive correlation with increasing Gleason scores, with positivity rates rising from 18.2% in tumors with a Gleason score of 7 to 48.3% in those with a score of 7. PSA expression was also significantly more common in PCa than in BPH (40.9% vs. 20.3%). These observations are consistent with previous reports that PSA levels and tissue expression can increase with tumor burden and grade [14]. However, the moderate strength of this association emphasizes the limitations of PSA as a stand-alone diagnostic or prognostic marker, especially given its presence in benign conditions and variability in expression across cancer grades.

Interestingly, HCK was uniformly negative across all cases and Gleason categories, suggesting it lacks diagnostic relevance in the prostate pathology spectrum. While HCK has been implicated in hematopoietic and inflammatory signaling, its absence in both benign and malignant prostate tissues indicates that it may not be suitable as a marker for epithelial neoplasia in this context [15].

Notably, patient age did not show a significant association with Gleason scores, supporting the notion that tumor grade is more likely driven by intrinsic tumor biology rather than age-related physiological changes. This finding indicates that differences in tumor grade in our cohort are likely driven by intrinsic tumor biology rather than patient age. Including age analysis not only confirmed its independence from grade but also ruled out its potential as a confounding variable in interpreting AR and PSA expression patterns. Also, this finding is consistent with previous studies that emphasize the independence of age and tumor aggressiveness in prostate cancer prognosis [16, 17]. Notably, HCK was uniformly negative in both PCa and benign prostatic tissue across all cases where it was assessed. This consistent lack of expression corroborates prior suggestions of limited HCK utility in prostate pathology [18–20] and supports its deprioritization or exclusion from routine diagnostic IHC panels, potentially optimizing resource allocation without compromising diagnostic accuracy.

From a clinical perspective, these findings reinforce the established practice of using P504S, AR, and PSA as standard immunohistochemical adjuncts in equivocal prostate biopsies, particularly when histologic features are ambiguous or tissue is limited. While the study does not introduce novel diagnostic markers, it provides confirmatory evidence of their reliability in a contemporary Chinese patient cohort evaluated in real-world pathology practice.

However, this study has several limitations. First, its retrospective design from a single tertiary center may limit generalizability, and the sample size, although adequate for initial analysis, restricts the ability to detect smaller effect sizes or marker interactions. Second, the assessment of IHC positivity was qualitative; future studies incorporating quantitative image analysis could provide more nuanced insights into expression patterns. The uniform negativity of HCK, while informative as evidence supporting its exclusion from standard prostate IHC panels, warrants confirmation in larger cohorts to fully establish its lack of diagnostic relevance. Furthermore, the benign cohort consisted predominantly of classic nodular BPH from surgical specimens rather than needle biopsy cases featuring cancer mimickers (e.g., adenosis, partial atrophy, or basal cell hyperplasia), which represent the more frequent diagnostic pitfalls in routine prostate biopsy practice. Inclusion of such mimickers would have provided a more direct evaluation of the markers’ utility in ambiguous cases. Additionally, P504S data were available only in a subset of cases (40 PCa, 19 benign), reflecting its selective use in routine practice for morphologically challenging lesions. This introduces potential selection bias, as assessed cases may overrepresent diagnostically ambiguous samples. Accordingly, the observed positivity in PCa is consistent with prior literature on P504S sensitivity but is limited to these subsets and should not be extrapolated universally. The incomplete P504S dataset also precludes exploration of correlations with AR or PSA expression. Nevertheless, the relatively low AR positivity rate (25% in PCa) compared with many published studies likely stems from routine clinical IHC variability (e.g., antibody sensitivity, thresholds including weak staining) and selective staining; interpretations of AR patterns, particularly regarding potential androgen independence, are thus limited by these factors. Finally, the follow-up period was insufficient to correlate marker expression with long-term clinical outcomes such as biochemical recurrence or survival, which should be addressed in prospective studies.

In summary, our results corroborate the complementary role of these well-established immunohistochemical markers (P504S/AMACR, AR, and PSA) alongside Gleason grading and basal cell immunohistochemistry (e.g., p63 and high-molecular-weight cytokeratins) for accurate differentiation of PCa from benign prostatic tissue. Basal cell markers remain essential for confirming the absence of a basal layer in suspicious glands, particularly when distinguishing adenocarcinoma from mimickers. The uniform HCK negativity further supports its deprioritization in diagnostic panels. These practice-based data from a Chinese cohort support continued reliance on existing multimarker diagnostic algorithms in prostate pathology.

Authors’ contributions

H.F.A.: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Supervision, Writing - original draft. Q.L.: Conceptualization, Project administration, Validation, Writing - review & editing.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Data availability

The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

This was a retrospective cohort study conducted at Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, a tertiary care academic medical center (No. Not assigned, our institutional ethics committee has formally approved this study and confirmed that our institution does not issue dedicated ethics approval numbers for retrospective, observational studies of this nature (which uses de-identified archived clinical samples with no direct patient intervention).All research conducted on humans and human data/material in this study adhered to the Helsinki Declaration. Given its retrospective design and use of anonymized archival pathology data, the requirement for informed consent was waived by the ethics committee.

Competing interests

The authors declare no competing interests.