Evaluation of Loss of the Y Chromosome in Peripheral Blood as a Biomarker for Prostate Cancer

ABSTRACT

Background

Prostate cancer (PCa) is the most common malignant tumor in men, but the widely used prostate‐specific antigen (PSA) test has limited diagnostic accuracy. Research proposes that the loss of the Y chromosome (LOY) may affect the occurrence and development of prostate cancer, aiming to assess its potential as a diagnostic biomarker to improve the accuracy of prostate cancer detection and clinical management.

Methods

LOY levels were measured using droplet digital polymerase chain reaction (ddPCR) method, integrating relevant clinical indicators to evaluate the potential clinical significance of LOY in PCa. Logistic regression was employed to estimate risk prediction capability of LOY, and model performance was validated through bootstrapping.

Findings

131 men patients were enrolled in this study whose PSA level exceeding 4.0 ng/ml and underwent prostate biopsies. According to post‐biopsy pathological results, subjects were classified into normal and cancer groups. LOY levels were significantly higher in the cancer group than in the normal group (p < 0.001). LOY levels show a consistent increase with advancing AJCC clinical stage and escalating PI‐RADS scores. LOY is a promising biomarker for PCa (AUC = 0.898). Clinical decision curve analyzes demonstrated that incorporating LOY into each conventional model provided greater clinical benefit compared with the original model.

Interpretation

In biopsy patients, the LOY levels in the cancer group were higher than those in the normal group, and this was more pronounced in patients at advanced clinical stages. LOY levels have strong diagnostic efficacy for PCa, indicating that LOY may serve as an auxiliary biomarker for early PCa diagnosis.

Abbreviations

AJCC stage

American Joint Committee on Cancer stage

AMEL

amelogenin

AUC

area under the curve

BMI

body mass index

CI

confidence interval

cNRI

category‐free NRI

CNV

copy number variation

ddPCR

droplet digital polymerase chain reaction

FISH

fluorescent in situ hybridization

GS

Gleason score

GWAS

genome‐wide association studies

IDI

integrated discrimination improvement

LOY

loss of the Y chromosome

NGS

next‐generation sequencing

NRI

net reclassification index

OR

odds ratio

PCa

prostate cancer

PI‐RADS score

prostate Imaging‐Reporting and Data System score

PSA

prostate‐specific antigen

QF‐PCR

multiplex

quantitative fluorescence polymerase chain reaction

ROC curves

receiver operating characteristic curves

WGS

whole gene sequencing

1. Introduction

Prostate cancer (PCa) is the most common cancer diagnosed in men in the United States in 2024, accounting for 29% of all male incident cases and becoming the second leading cause of cancer‐related death [1]. PCa mortality rates were stable from 2013 through 2021 [2], which has declined by 53% since the peak in 1993 due to earlier detection through widespread screening with the serum prostate‐specific antigen (PSA) testing and treatment advances [3]. Although PSA remains as an initial screening indicator, the U.S. Preventive Services Task Force (USPSTF) has revised its recommendations over time. After advising against routine PSA screening for all men in 2012 due to concerns about overdiagnosis and overtreatment, the USPSTF updated its guidance in 2018 to recommend individualized, shared decision‐making for men aged 55–69 years. This reflects recognition that PSA testing can reduce PCa mortality, but it remains imperfect and may still lead to unnecessary biopsies and treatments [4]. According to the 2024 revision of the European Association of Urology (EAU) guidelines, widespread population‐based PSA testing is also not recommended since it remains challenging to accurately evaluate the balance of benefits and harms associated with early detection for individual men. Several PSA‐related biomarkers (%fPSA, p2PSA, %p2PSA) have been reported that may hold promise for improving prostate cancer screening [5, 6]. Finding new screening indicators and improving risk stratification are critical for diagnosing PCa, differentiating between low‐ or high‐ risk patients and guiding clinical decisions for active surveillance or treatment.

Loss of the Y chromosome (LOY) in hematopoietic cells has been recognized as a common age‐related phenomenon in elderly men [7]. Up to the present, LOY detected in peripheral blood cells represents one of the most prevalent somatic mutations [8], which is considered by the prevailing consensus as the incidence of LOY increases with age and is further exacerbated by environmental factors such as smoking [9, 10, 11]. Furthermore, accumulating evidence suggests that the Y chromosome plays roles beyond sex determination, contributing to broader physiological functions. When LOY occurs, it has been associated with increased susceptibility to a range of common age‐related diseases, in particular cancer, Alzheimer's disease, and cardiovascular disorders [12, 13, 14]. These findings indicate that LOY may serve as a potential biomarker and mechanistic contributor to systemic aging and disease vulnerability.

Accumulating evidence have begun to clarify that LOY in peripheral blood was linked to higher cancer‐specific mortality (HR = 3.62) and all‐cause mortality (HR = 1.91) in prospective cohorts, underscoring its prognostic significance [12]. A meta‐analysis including both East Asian and European populations revealed genetically predicted LOY has been associated with increased PCa risk, demonstrating an odds ratio of 1.09 (95% CI: 1.05–1.13, p = 1.2 × 10⁻⁵) [15]. Uppsala Longitudinal Study of Adult Men (ULSAM) subjects over a 6–14 year period revealed a marked increase in the proportion of cells exhibiting LOY, with PCa subjects demonstrating a rise in LOY‐positive cells ranging 14%−37% [12]. This progressive accumulation suggests a potential link between LOY detected in peripheral blood and increased tumor risk. Complementary to this, LOY has been detected in approximately 30% of primary prostate tumors and metastatic castration‐resistant prostate cancer (mCRPC), suggesting its early emergence during tumorigenesis rather than as a consequence of treatment resistance [16, 17]. Mechanistically, LOY may disrupt immune surveillance and genome maintenance by altering expression of Y‐linked and autosomal immune genes [8, 18]. Recent large‐scale cancer genomic studies further suggest that LOY may act as either a driver or passenger mutation depending on tumor context, affects tumor suppressor genes such as KDM5D, promoting tumor progression and chemoresistance [17, 19]. Despite these insights, the clinical utility of LOY as a diagnostic or predictive biomarker for PCa remains elusive. No prospective study has evaluated the additive predictive value of LOY over PSA or other established risk factors in early detection or biopsy decision‐making. While its correlation with worse survival outcomes and urologic cancers progression is increasingly evident [20], the absence of diagnostic model comparisons precludes conclusions regarding its superiority or complementarity to available clinical indicators.

Therefore, this study aimed to evaluate whether LOY in peripheral blood can serve as a complementary or superior predictor of PCa diagnosis when combined with traditional markers, and to assess its clinical relevance in PCa.

2. Methods

2.1. Clinical Information Collection

Clinical characteristics of the 131 participants with PSA > 4.0 ng/mL, all of whom were outpatients of the Department of Urology at Juntendo University Hospital and underwent a prostate needle biopsy, were included in this study. Based on pathological diagnostic results, individuals were classified into a PCa group (n = 79) and a normal group (n = 52). Recorded variables included age, body mass index (BMI), smoking history (ever or never), smoking status was also assessed by Brinkman Index, calculated as the number of cigarettes smoked per day multiplied by years of smoking and serum PSA levels. MRI findings were evaluated using the Prostate Imaging Reporting and Data System (PI‐RADS), which scores prostate lesions on a scale from 1 to 5. Pathological characteristics were classified as: grade group 1 (Gleason ≤ 6), grade group 2 (Gleason 3 + 4 = 7), grade group 3 (Gleason 4 + 3 = 7), grade group 4 (Gleason ≥ 8). No Grade Group 5 cases were present in the study cohort. Additionally, clinical staging was performed according to the American Joint Committee on Cancer (AJCC) 8th Edition criteria. Data regarding metastasis, including lymph node metastasis, bone metastasis, and organ metastasis, were collected from clinical records at the time of initial diagnosis.

2.2. DNA Extraction and LOY Detection by ddPCR

Peripheral blood samples were obtained from each patient prior to the biopsy procedure. DNA was subsequently extracted from nucleated blood cells using the Genomic DNA Purification kit (A1125, Promega) according to the manufacturer's instructions and stored at −80°C until use.

Quantification of LOY was performed on the QuantStudio 3D Digital PCR System (Thermo Fisher, A29154) using TaqMan primers and probes targeting on the homologous Amelogenin genes (AMEL) located on both the chromosome X and Y that differ by 6 bp in length, which could be amplified with the same primers, offering an unbiased detection without bias related to primer properties. This approach quantitatively assesses the absolute copy number of chromosome X and Y in DNA samples.

Primers and probes (#C_990000001_10, GTGTTGATTCTTTATCCCAGATG[‐/AAGTGG]TTTCTCAAGTGGTCCTGATTTT) were purchased from Thermo Fisher Scientific (MA, USA) [21]. 80 ng of each DNA sample was digested for 37°C, 1 h; 65°C, 20 min with HindIII (Thermo Fischer, MA USA). Subsequently, digested DNA samples were added in QuantStudio 3D Digital PCR Master Mix v2 (Thermo Fischer, A26359) together with primers and probes. Following manufacturer's instructions, loading per sample to chips was generated by QuantStudio 3D Digital PCR Chip Loader (Thermo Fischer MA, USA), then PCR was performed on the ProFlex 2 × Flat PCR System (Thermo Fischer MA, USA) following conditions: 96°C for 10 min, 39 cycles of 60°C for 2 min and 98°C for 30 s, ending with 60°C for 2 min and 10°C hold. The QuantStudio 3D Digital PCR Instrument (Thermo Fischer MA, USA) read each chip and calculated the concentration of AMELY and AMELX targeted by FAM and VIC dye‐labeled probes. We determined Y‑chromosome integrity (ChrY integrity) by quantifying the absolute copy numbers of AMELY and AMELX from positive and negative droplet counts then calculating their ratio (AMELY/AMELX). LOY was then calculated as 1−ChrY integrity, which is equivalent to (AMELX − AMELY)/AMELX.

2.3. Statistical and Analysis

Statistical analyzes and data visualization were conducted using SPSS (version 29.0, IBM Corp., Armonk, NY, USA), R 4.1 and GraphPad Prism (version 10.0, San Diego, CA, USA). Continuous variables were described as mean ± standard deviation (SD) or median with interquartile range (IQR), and categorical variables were presented as frequencies and percentages.

Student's t test or Mann‐Whitney U test were used for continuous variables and Chi‐square test or Fisher's exact test for categorical variables, as appropriate. The correlation between LOY levels and age was assessed using Spearman's rank correlation analysis. Receiver operating characteristic (ROC) curves, based on logistic regression conducted in SPSS, were used to assess the diagnostic performance of LOY. AUC, Brier score and log‐loss were used to estimate risk prediction performance of models after containing LOY. Model reclassification performance for estimating the improvement between the prediction models was also evaluated using net reclassification improvement (NRI), category‐free NRI (cNRI), and integrated discrimination improvement (IDI). To evaluate the clinical utility of LOY, decision curve analysis (DCA) was conducted to calculate the net benefit. A two‐sided p value of less than 0.05 was considered statistically significant for all analyzes.

2.4. Results

2.4.1. Characteristics of the Subjects

The study included 131 male subjects (mean age = 68.34) with PSA above 4.0 ng/mL who underwent prostate biopsy due to suspected PCa. All blood samples were collected between September 2020 and June 2021, with a subsequent median follow‐up duration of 43 months (range, 3–53 months). Based on the pathological findings, subjects were classified into a PCa group (n = 79) and a normal group (n = 52). A significant difference in age and PSA was observed between two groups (p < 0.001, Table 1), necessitating their adjustment as covariates in the LOY analyzes to control for potential confounding effect. And a positive correlation between age and LOY was also observed in both overall group and PCa group (Supporting Information Figure S3A,B), however, no such trend was identified in the normal group (Supporting Information Figure S3C). The mean LOY level in the entire cohort was 0.10 ± 0.17, while the LOY levels were significantly higher in the PCa group compared to the normal group (p < 0.001, Tabel 1). Notably, no statistically significant difference in smoking history or Brinkman Index were observed between groups (p = 0.499, p = 0.521, Table 1).

Table 1.

Clinical characteristics for total 131 patients.

	All (n = 131)	PCa group (n = 79)	Normal group (n = 52)	p value
Age (years)	68.34 ± 8.61	70.54 ± 8.17	64.98 ± 8.24	< 0.001
< 65	41 (31.30%)	16 (20.25%)	25 (48.08%)	< 0.001
≥ 65	90 (68.70%)	63 (79.75%)	27 (51.92%)	< 0.001
BMI (kg/m²)	24.14 ± 2.94	24.16 ± 3.2	24.11 ± 2.57	0.925
Smoking history	83 (63.36%)	51 (64.56%)	32 (61.54%)	0.499
Brinkman index	200 (0, 660)	200 (0, 720)	150 (0, 750)	0.521
Gleason score
1 ( ≤ 6)	—	18 (22.8%)	—
2 (3 + 4 = 7)	—	14 (17.7%)	—
3 (4 + 3 = 7)	—	16 (20.3%)	—
4 ( ≥ 8)	—	31 (39.2%)	—
AJCC stage
I	—	6 (7.6%)	—
II	—	42 (53.2%)	—
III	—	21 (26.6%)	—
IV	—	10 (12.7%)	—
Metastasis	—	10 (12.7%)	—
Lymph node	—	2 (2.5%)	—
Bone	—	8 (10.1%)	—
Organ	—	3 (3.8%)	—
PI‐RADS score
1	9 (6.87%)	5 (6.33%)	4 (7.69%)
2	12 (9.16%)	2 (2.53%)	10 (19.23%)
3	45 (34.35%)	19 (24.05%)	26 (50.00%)
4	37 (28.24%)	30 (37.97%)	7 (13.46%)
5	28 (21.37%)	23 (29.11%)	5 (9.62%)
PSA (ng/ml)	15.37 ± 23.76	20.80 ± 29.35	7.22 ± 3.87	< 0.001
LOY	0.10 ± 0.17	0.17 ± 0.17	0.00 ± 0.10	< 0.001

Note: Values are given as n (%), median (25th percentile, 75th percentile) or mean ± standard deviation.

Abbreviations: AJCC stage, American Joint Committee on Cancer stage; BMI, body mass index; LOY, loss of Y chromosome; PCa, prostate cancer; PI‐RADS score, Prostate Imaging‐Reporting and Data System score; PSA, prostate‐specific antigen.

2.5. LOY Detected in PCa Subjects

Significantly elevated levels of LOY in PCa group were demonstrated compared with normal group across the entire age cohort (p < 0.001, Figure 1A). To evaluate age‐related effects, the cohort was further stratified into two age subgroups (categorized as < 65 and ≥ 65 years, which based on clinical relevance and previous studies) [22, 23]. Same trends were observed in both subgroups, where LOY levels remaining significantly higher in PCa group than in normal controls (both p < 0.001, Figure 1B,C). These findings suggest that increased LOY is consistently associated with PCa regardless of age stratification, underscoring its potential utility as a robust biomarker in clinical diagnosis.

Figure 1.

Comparison of LOY by age stratification. (A) LOY levels between prostate cancer group (PCa; blue) and normal group (Nor; red) then all samples were stratified by age: (B) <  65 years, (C) ≥ 65 years. The y‐axis displays LOY, representing the estimated proportion of nucleated blood cells retaining loss of Y chromosome. Y chromosome integrity (ChrY integrity) was determined by quantifying the absolute copy numbers of AMELY and AMELX from positive and negative droplet counts in ddPCR and calculating their ratio (AMELY/AMELX). LOY was then derived as 1−ChrY integrity, equivalent to (AMELX − AMELY)/AMELX. Thus, higher LOY values reflect greater loss of the Y chromosome. Asterisks above each comparison indicate statistically significant differences (*p < 0.05; **p < 0.01; ***p < 0.001). [Color figure can be viewed at wileyonlinelibrary.com]

2.6. LOY Associated With Clinical Stage of PCa

To further explore the potential relationship between LOY and PCa clinical stage, according to AJCC Cancer Staging Manual (8th Edition), individuals were classified into early stage (AJCC grades I and II) and advanced stage (AJCC grades III and IV). LOY levels were significantly higher in advanced stage group compared to the early stage group (p < 0.01, Figure 2A). AJCC stages I to IVB were converted into numerical scores ranging from 1 to 9, creating a continuous variable reflective of escalating disease severity. Correlation analysis demonstrated a positive association between LOY levels and clinical stage (p < 0.05, Figure S4A). Subjects were also stratified by age, using 65 years as the cutoff. LOY levels did not differ significantly between early and advanced stage group among subjects younger than 65 years (Figure S4B). Conversely, subjects older than 65 years exhibited significantly higher LOY levels in the advanced stage group compared to the early stage group (p < 0.05, Figure S4C), indicating that elevated LOY levels may predict advanced disease stage predominantly in older individuals.

Figure 2.

Relationship between LOY and clinical diagnosis. (A) LOY levels between early stage (AJCC grades I–II, blue) and advanced stage (AJCC grades III–IV, red). (B) LOY levels between subjects with PI‐RADS scores 1–3 (blue) and those with scores 4–5 (red). [Color figure can be viewed at wileyonlinelibrary.com]

Consistent with the findings from clinical stage comparisons, LOY levels were markedly elevated in the MRI‐positive subgroup (PI‐RADS scores 4–5) compared to the MRI‐negative subgroup (PI‐RADS scores 1–3) (p < 0.01, Figure 2B). LOY levels exhibited a significant positive correlation with PI‐RADS scores, progressively increasing with higher PI‐RADS categories (p < 0.01, Supporting Information Figure S4D). Additionally, within the MRI‐negative subgroup, LOY levels were significantly elevated in PCa group compared to normal controls (p < 0.001, Fig S4E), highlighting potential of LOY to identify PCa in patients with negative MRI findings and thus reducing false‐negative diagnosis.

2.7. The Potential of LOY as an Adjuvant Diagnostic Biomarker for PCa

To further evaluate the clinical utility of LOY as a diagnostic biomarker in PCa, ROC curve analyzes were performed comparing LOY with established clinical indicators, including PSA, PI‐RADS scores, and age. The analysis demonstrated that LOY alone exhibited robust diagnostic capability, with an area under the curve (AUC) of 0.898 (95% CI 0.836–0.960), substantially outperforming traditional indicators such as PSA (AUC 0.708; 95% CI 0.620–0.795) and PI‐RADS (AUC 0.727; 95% CI 0.637–0.816) (Figure 3A). ROC‐based Youden index analysis identified an optimal LOY cut‐off of 0.138. When applied to this cohort, this threshold provided positive predictive values (PPV) of 91.1% and Negative Predictive Values (NPV) of 86.5%, consistent with the strong diagnostic performance of LOY (Supporting Information Tables S1). Importantly, when LOY was combined with PSA, the resulting predictive model attained the highest diagnostic efficacy observed, with an AUC of 0.906 (95% CI 0.850–0.961) (Figure 3B).

Figure 3.

Comparative receiver operating characteristic analyzes for models integrating LOY. (A) ROC curves for PSA, PI‐RADS scores, and their combinations in the detection of PCa. (B) ROC curves for LOY, LOY combined with PSA, PI‐RADS, and age, and their various combinations. AUC values and 95% confidence intervals for each model are summarized in the corresponding tables below each panel. The diagonal dashed line indicates the reference line (AUC = 0.5). [Color figure can be viewed at wileyonlinelibrary.com]

Meanwhile, clinical decision curve analysis (DCA) was performed to assess the potential clinical value of various predictive models across a spectrum of threshold probabilities (Figure 4A). The resulting curves indicated that the LOY‐integrated detection strategy provided a superior net benefit over a broad range of threshold probabilities compared to strategies without LOY (Figure 4B).

Figure 4.

Decision curve analysis comparing clinical net benefit of models integrating LOY. (A) Decision curve analysis (DCA) for PSA, PI‐RADS scores, and their combinations in prostate cancer detection. (B) DCA curves for LOY, LOY combined with PSA, PI‐RADS, and age, and their various combinations. The solid black lines represent the strategies of treating all individuals (“All”) or treating none (“None”). [Color figure can be viewed at wileyonlinelibrary.com]

2.8. Performance of PCa Predictive Model Containing LOY

The predictive performance of the logistic regression model integrating LOY and PSA was robust, with an AUC of 0.906 (95% CI: 0.850–0.961, p < 0.001) significantly exceeding the performance of PSA alone (Table 2, Figure 3A). Incorporating PI‐RADS into the LOY model yielded an AUC of 0.888 (95% CI: 0.825–0.952, p < 0.001), while addition of age to LOY produced an AUC of 0.896 (95% CI: 0.833–0.960, p < 0.001). Models integrating LOY with multiple clinical indicators (PSA, PI‐RADS, age) demonstrated consistently elevated AUC values, ranging from 0.887 to 0.905 (Table 2). The addition of LOY significantly improved reclassification metrics compared to conventional models, reflected by NRI values ranging from 0.284 (95% CI: 0.135–0.423) to 0.547 (95% CI: 0.395–0.701). Correspondingly, IDI ranged from 0.114 (95% CI: 0.053–0.177) to 0.219 (95% CI: 0.136–0.307). cNRI demonstrated substantial enhancement across models, with values spanning from 0.888 (95% CI: 0.570–1.161) to 1.340 (95% CI: 1.091–1.560). Overall, incorporating LOY substantially enhanced predictive accuracy and improved reclassification power for PCa diagnosis beyond conventional clinical parameters.

Table 2.

Risk prediction performance of logistic regression for PCa diagnosis.

Performance	LOY + PSA	LOY + PI‐RADS	LOY + Age	LOY + PSA + PI‐RADS	LOY + PSA + Age	LOY + PI‐RADS+Age	LOY + PSA + PI‐RADS + Age
1) Discrimination
Brier score	0.121 (0.093–0.155)	0.129 (0.100–0.159)	0.128 (0.101–0.159)	0.123 (0.095–0.155)	0.121 (0.092–0.157)	0.128 (0.097–0.162)	0.122 (0.092–0.157)
Log‐loss	0.423 (0.308–0.568)	0.454 (0.337–0.620)	0.459 (0.346–0.630)	0.420 (0.307–0.566)	0.425 (0.310–0.584)	0.454 (0.335–0.606)	0.419 (0.306–0.569)
AUC	0.906 (0.850–0.961)	0.888 (0.825–0.952)	0.896 (0.833–0.960)	0.905 (0.849–0.961)	0.904 (0.847–0.960)	0.887 (0.823–0.950)	0.902 (0.846–0.959)
p for AUC	< 0.001 (vs PSA)	< 0.001(vs PI‐RADS)	< 0.001 (vs Age)	< 0.001(vs PSA + PI‐RADS)	< 0.001(vs PSA + Age)	< 0.001(vs PI ‐RADS + Age)	< 0.001(vs PSA + PI‐RADS + Age)

2) Reclassification	vs PSA	vs PI‐RADS	vs Age	vs PSA + PI‐RADS	vs PSA + Age	vs PI‐RADS + Age	vs PSA + PI‐RADS + Age
NRI	0.361 (0.195–0.523)	0.547 (0.395–0.701)	0.521 (0.354–0.691)	0.383 (0.240–0.531)	0.295 (0.135–0.458)	0.313 (0.166–0.462)	0.284 (0.135–0.423)
cNRI	1.168 (0.909–1.412)	1.340 (1.091–1.560)	1.196 (0.944–1.444)	1.051 (0.760–1.321)	1.004 (0.701–1.249)	1.024 (0.715–1.298)	0.888 (0.570–1.161)
IDI	0.195 (0.112–0.281)	0.163 (0.083–0.241)	0.219 (0.136–0.307)	0.137 (0.068–0.203)	0.139 (0.068–0.210)	0.123 (0.057–0.191)	0.114 (0.053–0.177)

Abbreviations: AUC, area under curve; cNRI, category‐free NRI; IDI, integrated discrimination improvement (p represents the p‐value at a significance level within the bootstrapped t‐test for AUC between prediction models); NRI, net reclassification index.

To clarify the relative impact of individual variables on PCa risk, odds ratio (OR) was calculated per one standard deviation (SD) increase (Table 3). LOY showed a significant association with PCa, exhibiting an OR of 8.132 (95% CI: 3.826–17.318, p < 0.001). PSA also demonstrated a substantial contribution to PCa risk (OR = 6.024, 95% CI: 1.993–64.691, p = 0.049), although with broader CI. In contrast, neither age (OR = 1.188, 95% CI: 0.902–1.564, p = 0.512) nor PI‐RADS score (OR = 1.237, 95% CI: 0.782–1.957, p = 0.420) showed significant associations with PCa.

Table 3.

LOY as an independent risk factor from logistic regression.

	OR (95%CI)	p
LOY	8.132 (3.826–17.318)	< 0.001
PSA	6.024 (1.993–64.691)	0.049
PIRADs	1.237 (0.782–1.957)	0.420
Age	1.188 (0.902–1.564)	0.512

3. Discussion

LOY in hematopoietic cells was first reported nearly six decades ago by Jacobs et al, [24] and is recognized as one of the most prevalent somatic genomic alterations in elderly men, closely associated with aging processes [9, 10, 13]. Recent studies highlighted that LOY in blood cells correlates with an elevated risk for overall mortality [8, 12]. Additionally, a large‐scale study using data from the UK Biobank revealed that LOY in leukocytes is moderately correlated with an increased incidence of solid tumors [25], with PCa is certainly on the list of LOY‐related malignant tumor [26, 27].

Commonly utilized methods for detecting LOY include genome‐wide association studies (GWAS) [28], copy number variation (CNV) analyzes for large‐scale mosaic autosomal abnormalities [29], and whole genome sequencing (WGS) to identify mosaic mutations at the base‐pair level [30]. P. Noveski et al. reported that LOY is a stronger predictor of cancer risk than chronological age alone, suggesting that age by itself might not fully account for the increased prevalence of detectable LOY in cancer cohorts, whose conclusion was based on employing multiplex quantitative fluorescence polymerase chain reaction (QF‐PCR) [31]. Furthermore, GWAS identified 18 out of 156 LOY‐associated variants in leukocyte overlap with known susceptibility variants for various non‐hematological malignancies, including PCa [8].

Recently, a novel PCR method known as droplet digital PCR (ddPCR) has emerged, offering advantages over traditional real‐time PCR by enabling absolute quantification of target genes without the need for a reference sample or standard curve [32]. This technique has been applied in the detection of tumor‐related LOY [33], however, its application in PCa remains limited. Given the simplicity, sensitivity, and efficiency of ddPCR compared to techniques such as DNA sequencing, further exploration of its utility in assessing the role of LOY in prostate cancer is of significant value. In this study, we employed ddPCR to detect LOY in patients with suspected PCa to evaluate its potential as a diagnostic and predictive biomarker of PCa. To validate accuracy of the LOY results from ddPCR, we simultaneously analyzed 198 DNA samples from patients diagnosed with PCa in our hospital using both low coverage WGS and ddPCR. Our findings demonstrated that the LOY results obtained via ddPCR were highly consistent with those from low coverage WGS (p < 0.001, r = 0.4738, Supporting Information Figure S2). Importantly, compared to low coverage WGS, ddPCR substantially reduced both measurement time and cost, marking a major step toward its future clinical implementation. Our data further substantiate that LOY in peripheral blood cells significantly elevates Pca risk, aligning with previous studies that have established a link between LOY and various non‐hematological cancers. It is noteworthy that there is indeed an overlap in the distribution of LOY values between men with and without PCa. Biologically, LOY in peripheral blood represents a somatic and mosaic alteration that accumulates with age and is influenced by multiple factors such as smoking and clonal hematopoiesis—many of which are unrelated to prostate carcinogenesis. As a result, some “normal” individuals may exhibit relatively high LOY values, whereas certain patients with PCa may retain a comparatively preserved Y chromosome. Importantly, despite this overlap, LOY still demonstrates strong discriminative performance at the population level, as evidenced by its high AUC (0.898) and its independent association with PCa in multivariable analyzes adjusted for age. LOY should therefore not be interpreted as a perfectly separating diagnostic test at the individual level, but rather as a robust risk‐stratification biomarker that can complement PSA and other clinical parameters to support improved clinical decision‐making.

Aging, widely recognized as an independent risk factor for LOY, has been extensively documented [34, 35, 36]. Additionally, smoking is identified as a major contributor to LOY in blood cells, with evidence indicating that current smokers experience up to a four‐fold elevated risk [11]. Building on these observations, a critical question of our study is whether LOY observed in PCa patients reflects a direct consequence of cancer biology or merely reflects age‐ or smoking‐related phenomenon. To address these potential confounders, we accounted for age and smoking status at the time of sampling by including them as co‐covariates in the statistical models assessing the relationship between LOY and PCa. Our results revealed that, even after adjusting for age and smoking status, LOY remains an independent risk factor for PCa. In subsequent analyzes, we conducted an age‐based cutoff to minimize the confounding effect of sampling age. Notably, the association between LOY and increased risk of PCa persisted consistently across age subgroups. The elevated prevalence of LOY observed in the cancer cohort implies that chromosomal instability may play a critical role in the initiation and progression of PCa. This finding is consistent with previous studies that have implicated chromosomal alterations as key drivers of oncogenesis across a range of malignancies [17, 33, 37, 38]. The frequent detection of LOY in PCa patients further suggests the presence of an underlying pathogenic mechanism that may contribute to the malignant transformation of normal prostate epithelial cells. Collectively, these findings support the consideration of LOY not only as a biomarker for PCa but also as a potential driver of adverse clinical outcomes.

Although PSA testing remains one of the most widely employed screening methods for PCa [39], numerous studies have underscored its limitations as a reliable indicator [40, 41, 42]. Consequently, growing research efforts have increasingly focused on identifying more effective and complementary approaches, including advanced computational algorithms [43], liquid biopsy techniques [44], and novel molecular biomarkers [45, 46]. This study highlights a significant enhancement in PCa diagnostic performance by incorporating LOY as a novel biomarker alongside PSA testing. Notably, ROC curve analysis demonstrated LOY alone achieved an AUC of 0.898, surpassing the 0.708 observed for PSA alone. Based on ROC analysis using the Youden index, we identified an optimal LOY cut‐off (0.138) (Supporting Information Table S1). When applied to our cohort, the corresponding PPV and NPV were 91.1% and 86.5%, reflecting both the strong intrinsic performance of LOY and the elevated disease prevalence in this selected population. While this threshold aids clinical interpretation, LOY is inherently a continuous variable, and the optimal cut‐point may differ across populations. Therefore, external validation in broader and more diverse cohorts—particularly including men with PSA < 4.0 ng/mL—is warranted before definitive clinical recommendations regarding LOY‐based decision‐making can be established. This study cohort consisted exclusively of men with PSA ≥ 4.0 ng/mL, reflecting the inclusion criteria of our prospective protocol. As a result, the diagnostic performance of LOY reported here pertains to a PSA‐elevated population and may not fully represent its utility in men with lower PSA levels. Whether LOY can discriminate PCa in individuals with PSA < 4.0 ng/mL remains unknown and warrants further investigation. Future population‐based or low‐PSA cohort studies will be essential to determine the broader applicability of LOY as an early detection biomarker. These findings indicate that LOY may serve as a complementary and potentially superior biomarker to PSA testing, addressing limitations of PSA in false‐positives issues. Moreover, integrating LOY into the screening process could facilitate more targeted PCa detection strategies, offering a less invasive and more efficient preliminary screening approach. This may markedly reduce the number of unnecessary biopsies and associated complications, with enhancing patient care and improving clinical outcomes. During follow‐up of the biopsy‐negative cohort, two individuals were subsequently diagnosed with PCa at 18 and 19 months, respectively, with peak PSA levels of 18.372 ng/mL and 12.732 ng/mL during the observation period. Since our study did not conduct longitudinal assessment of LOY over time, we were unable to confirm the fluctuations in LOY for these two newly identified samples during the follow‐up period. This observation highlights an important direction for future work. Longitudinal evaluation of LOY trajectories may provide additional insights into whether temporal fluctuations in Y‐chromosome loss can predict progression from an initially negative biopsy to clinically significant disease.

From a clinical perspective, this approach aligns with the principles of personalized medicine by facilitating tailored diagnosis pathways based on individualized biomarker profiles. Incorporating LOY into routine clinical practice may be able to revolutionize PCa diagnostics by enhancing both precision and efficiency. Additionally, identifying patients with elevated LOY levels may support the development of individualized treatment strategies, thereby enhance therapeutic efficacy and potentially reduce the risk of recurrence. The observed positive correlation between LOY and PCa clinical diagnostic grading, according to the American Joint Committee on Cancer (AJCC 8th edition), indicates that the clinical relevance of LOY extends beyond initial screening and contribute to tumor staging and risk stratification. Patients with elevated levels of LOY may exhibit more aggressive disease progression, underscoring the need for intensified therapeutic strategies. We further examined the relationship between LOY and Gleason Grade. Unfortunately, our cohort revealed no statistically significant correlation between LOY and Gleason Grade (1–4) of the diagnosed PCa. A potential reason may be the absence of Gleason grade 5 samples in this cohort, coupled with the limited number of grade 1–4 samples, resulting in the absence of statistically significant differences in the final correlation analysis. This indicates that while LOY serves as a strong biomarker associated with the presence of PCa in men with elevated PSA, its level in blood does not appear to reflect the histopathological aggressiveness of the tumor within the grade spectrum observed in this study. Therefore, its primary utility in the tested population (PSA > 4 ng/mL) may lie in augmenting the positive predictive value of PSA, helping to identify men who are most likely to have a positive biopsy and thus benefit most from the procedure, rather than in replacing pathological grading. It remains possible that a relationship exists with very high‐grade or metastatic disease not represented here. Future studies with larger, more diverse cohorts encompassing the full spectrum of disease aggressiveness, including high grade tumor, are necessary to conclusively determine if LOY has any role in aggressiveness stratification. Nevertheless, the clear association with PCa diagnosis itself establishes LOY as a compelling candidate biomarker worthy of further investigation in broader clinical settings. This insight may aid in guiding clinical decision‐making and hold promise for improving patient survival outcomes. However, given the strong diagnostic performance observed in this study, it is important to acknowledge that these findings require further validation. This study should be replicated in independent external cohorts and confirmed through prospective studies, which to avert the potential reduced effectiveness when evaluated across broader populations. Therefore, while LOY appears to be a highly informative biomarker in our dataset, rigorous validation across diverse clinical settings will be essential before its incorporation into routine diagnostic workflows.

The mechanisms underlying the association between LOY and PCa remain incompletely elucidated. Emerging evidence suggests that LOY influences the transcriptome and modulates downstream biological processes, including cell cycle regulation, apoptosis, and immune response [47]. Among these, immune dysregulation appears to be most prominent [48, 49, 50], with impaired immune surveillance contributing to tumorigenesis through increasing genomic instability and decreasing immune‐mediated clearance of aberrant cells. Further comprehensive studies are warranted to clarify the functional role of LOY in PCa pathogenesis.

Several novel biomarkers and multi‐parametric diagnostic tests, such as MPS2, are currently being developed to improve PCa risk stratification. While MPS2 integrates multiple urinary molecular markers and has shown promising diagnostic performance, it requires multi‐analyte measurement and may be affected by pre‐analytical variability inherent to urine‐based assays. In contrast, LOY is derived from a single, stable genomic measurement obtained from peripheral blood, offering practical advantages in terms of assay simplicity, standardization, and scalability. These methodological distinctions underscore the potential complementary role of LOY within the evolving landscape of PCa biomarkers. Future studies comparing LOY directly with tests such as MPS2 will be important to determine their relative strengths and optimal clinical integration.

3.1. Limitations

This study was conducted within a specific and limited population. Future research should aim to validate these findings in larger and more diverse cohorts, integrate animal models, and perform fundamental experiments to investigate the molecular mechanisms underlying LOY. Such efforts will not only advance the optimization of PCa screening strategies but also support the development of more effective diagnostic and therapeutic interventions for clinical management.

4. Conclusion

This study identified LOY as an independent risk factor for PCa among individuals with PSA levels > 4 ng/mL undergoing prostate biopsy. ROC analysis demonstrated that LOY exhibited a high AUC for PCa detection, supporting its potential utility as a reliable biomarker to enhance the accuracy of current screening strategies. Furthermore, the association between LOY levels and clinical staging highlights its relevance not only for diagnostic purposes but also for guiding therapeutic decision‐making in PCa management. Future investigations involving larger, more diverse cohorts and comprehensive molecular biology experiments are warranted to elucidate the interaction mechanisms between LOY and PCa tumorigenesis.

Author Contributions

Jun Lu: experiments, formal analysis, investigation, visualization, writing – original draft, writing – review and editing. Yan Lu: methodology, validation, writing – review and editing. Takuroi Kobayashi, Tsuyoshi Hachiya, and Yoshihiro Ikehata: software, visualization, writing – review and editing. Haruhiko Wakita, and Yasunari Tanaka: resources, data curation, writing – review and editing. Masayoshi Nagata: validation, writing – review and editing. Yiming Jin: experiments, writing – original draft. Hisamitsu Ide: conceptualization, methodology, validation, writing – review and editing. Shigeo Horie: conceptualization, project administration, supervision, writing – review and editing.

Ethics Statement

This study was obtained with approval of the Institutional Review Board (IRB) of Juntendo University Hospital (IRB #M18‐0234; #M19‐0158) and clinical data has been prospectively collected in accordance with a standardized protocol. All study participants provided written informed consent.

Consent

Written informed consent was obtained from all participants prior to their inclusion in this study.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Fluorescence distribution plot of FAM (blue) vs. VIC (red) in a ddPCR result. Figure S2: Correlation analysis between whole gene sequencing (WGS) data and ddPCR data (n=198). Figure S3: Correlation analysis between age and LOY. Figure S4: Association of LOY with clinical stage and MRI findings. Table S1: Diagnostic performance of LOY at the optimal Cut‐off.

Data Availability Statement

Data underlying the findings of the Juntendo dataset are available from the corresponding author upon reasonable request.