PSA, encoded by the KLK3 gene, is a glycoprotein predominantly secreted by the epithelial cells of the prostate gland (16). While its primary physiological role is the liquefaction of semen, its high organ specificity has established it as a cornerstone biomarker in the screening and diagnosis of prostate diseases (17). In the serum, PSA exists primarily in three molecular forms: Free PSA (f-PSA), which constitutes 10-30% of the total PSA (t-PSA); PSA complexed with α1-antichymotrypsin and PSA complexed with α2-macroglobulin. The latter two are collectively termed complexed PSA (18).

f/t PSA provides key diagnostic information beyond t-PSA levels alone (19). In patients with PCa, the proportion of f-PSA is markedly lower compared with those with benign conditions such as benign prostatic hyperplasia. The f/t-PSA ratio is calculated as follows (20): f-PSA/t-PSA=free PSA concentration (ng/ml)/total PSA concentration (ng/ml).

An f/t PSA ratio <0.15 is associated with an increased risk of malignancy and is a key factor in recommending prostate biopsy, thereby improving diagnostic specificity (21,22). This metric is particularly valuable within the diagnostic 'gray zone' of t-PSA (4-10 ng/ml) (23). A ratio >0.25 typically suggests a benign etiology, whereas a ratio <0.15 warrants further invasive assessment (24,25). Notably, when t-PSA >10 ng/ml, biopsy is generally indicated regardless of the f/t PSA ratio (26). A large multicenter study demonstrated that incorporating the f/t PSA ratio into clinical decision-making reduced misclassification by ~43%, underscoring its utility in enhancing diagnostic accuracy (27).

Emerging strategies seek to integrate the f/t PSA ratio with other clinical parameters, including patient age, family history and racial background, to develop multivariate predictive models (28,29). These refined algorithms promise to further personalize PCa risk stratification and support early intervention strategies (30).

PSAD, defined as the ratio of serum PSA concentration to prostate volume (PSAD=t-PSA/prostate volume), refines diagnostic precision by accounting for gland size (31). Prostate volume is typically quantified via transrectal ultrasound (TRUS) or, with increasing frequency, MRI. A PSAD value >0.15 ng/ml/cm³ is consistently associated with a heightened risk of PCa (32,33). This normalization mitigates the confounding effect of benign gland enlargement, thereby reducing false-positive results (34).

Longitudinal cohort studies have confirmed that an elevated PSAD is a robust predictor of PCa incidence and progression (35,36). Recent evidence indicates that MRI-based PSAD calculations, benefiting from superior soft-tissue resolution, outperform TRUS-derived values (37). Consequently, MRI-calculated PSAD exhibits enhanced diagnostic sensitivity and specificity in discriminating malignant from benign lesions, supporting its integration into modern diagnostic pathways (38).

PSAV measures the dynamic change in PSA levels over time, calculated as the absolute change in PSA per year [PSAV=(PSA₂-PSA₁)/ΔTime in years] (39). A velocity >0.75 ng/ml per year is considered clinically considerable and suspicious for underlying malignancy (40). PSAV is particularly useful for monitoring individuals with borderline PSA levels and for assessing disease progression in patients' post-treatment (41). A rapid rise in PSA is often associated with aggressive tumor biology and a worse prognosis (42).

The clinical utility of PSAV is being augmented by AI and machine learning (43). Emerging models integrate PSAV with a multitude of clinical, imaging and genomic parameters to achieve more accurate predictions of disease trajectory, thereby aiding in personalized surveillance and therapeutic decision-making (44,45).

PHI is an advanced blood test that combines three distinct forms of PSA: f-PSA, t-PSA and a truncated form of kallikrein (KLK)-related protein 4 (k-PSA, also known as hK4). It is calculated using the following formula (46): PHI=(f-PSA)/(t-PSA) x t-PSA √(t-PSA) x k-PSA.

The resulting score stratifies patients according to their risk of harboring clinically considerable PCa, with higher values indicating greater risk (47). The established clinical thresholds guide management as follows: i) A PHI <15 suggests a low likelihood of malignancy, ii) values between 15 and 35 fall into an indeterminate range where adjunctive evaluation with multiparametric MRI (mpMRI) is recommended, iii) a PHI >35 signifies high risk and typically warrants biopsy (48,49). PHI demonstrates superior diagnostic accuracy compared with t-PSA or f/t PSA alone, particularly within the t-PSA gray zone and has been shown to reduce unnecessary biopsies by ~20% (50,51). Prospective validation studies confirm that PHI effectively identifies aggressive PCa, improving patient selection for invasive procedures (52,53).

The integration of PHI with mpMRI represents a powerful diagnostic synergy. This combined model enhances detection confidence for clinically considerable cancer while minimizing overtreatment (54). Furthermore, ongoing research into ethnicity-specific PHI thresholds highlights the importance of personalized diagnostic standards to optimize accuracy across diverse populations (55,56).

KLK family comprises 15 secreted serine proteases (KLK1-KLK15) clustered on chromosome 19q13.4. These enzymes share structural similarity, characterized by a conserved catalytic triad of histidine, aspartic acid and serine residues (57-59). KLKs participate in diverse physiological processes, including skin desquamation, neural plasticity and inflammation (60-62). For instance, KLK1 cleaves kininogen to generate bradykinin, thereby modulating vascular tone, permeability and inflammatory signaling (63). KLK5 and KLK7 regulate keratinocyte differentiation and desquamation, with aberrant activity implicated in psoriasis and atopic dermatitis (64,65).

Among all KLKs, KLK3 (a PSA) is the best-established biomarker for PCa (66). Nevertheless, increasing evidence highlights that other family members, particularly KLK2, act synergistically with KLK3 to promote prostate tumorigenesis and reshape the tumor microenvironment (67,68). Dysregulation of the KLK proteolytic network facilitates cancer progression by enhancing tumor invasion, angiogenesis and metastasis (69).

Transcriptomic analyses demonstrate distinct expression patterns of multiple KLKs in prostate carcinoma compared with benign tissue, underscoring their combined diagnostic and prognostic potential (70,71). KLK2, sharing ~78% sequence identity with KLK3, is also AR-regulated and activates pro-PSA by cleaving its pro-peptide (72). Due to its highly restricted expression profile, KLK2 has emerged as an attractive therapeutic target (73). Several innovative platforms are currently in development, including: i) Radioligand therapies delivering cytotoxic payloads selectively to KLK2-expressing cells (74), ii) bispecific T-cell engagers redirecting cytotoxic lymphocytes toward KLK2-positive tumor cells (75); and iii) chimeric antigen receptor T-cell therapies engineered to recognize KLK2 on the cancer cell surface (76). Promising clinical candidates from companies such as Johnson & Johnson and EpimAb Biotherapeutics have advanced into Phase I trials, representing an encouraging step toward precision immunotherapy in PCa (77,78).

Citric acid and polyamines have attracted increasing attention as metabolic biomarkers for PCa diagnosis and management (173). Citric acid, abundantly present in prostatic fluid, blood and tissues, serves as an indicator of prostate metabolism (174). Reduced citric acid levels are typically associated with malignant transformation, offering potential diagnostic utility in differentiating benign from malignant lesions (175). The use of magnetic resonance spectroscopy (MRS) enables non-invasive quantification of citric acid, providing valuable metabolic information for clinical decision-making (176). However, its diagnostic sensitivity remains limited when used alone and combination with other biomarkers such as PSA improves accuracy and specificity (177).

Polyamines, such as putrescine, spermidine and spermine-also show association with tumor aggressiveness and metastatic potential (178,179). Elevated polyamine concentrations in serum or urine may indicate enhanced tumor invasiveness and poor prognosis (180). Monitoring polyamine profiles could therefore assist in predicting disease progression and evaluating treatment responses. Nevertheless, benign conditions such as inflammation can modestly elevate polyamine levels, limiting diagnostic specificity (181).

From a therapeutic perspective, abnormal metabolism of citric acid and polyamines represents a promising target for intervention (182). Strategies aiming to restore mitochondrial metabolism or inhibit ornithine decarboxylase, a key enzyme in polyamine synthesis, have demonstrated potential in enhancing endocrine therapy and overcoming drug resistance (182,183). Recent metabolomics-based studies have identified urinary derivatives such as N¹-acetylputrescine, which, when analyzed together with citric acid metabolites, notably improve early screening sensitivity (184,185). Integration of MRS-derived citric acid signals and serum polyamine levels with machine learning algorithms has further enhanced diagnostic precision, particularly for small lesions (186,187). Cross-ethnic studies have revealed distinct metabolic profiles of citric acid and polyamines, suggesting that population-specific diagnostic thresholds may improve biomarker accuracy (188,189). Moreover, emerging evidence indicates that gut microbial metabolites can modulate prostate metabolism by influencing citric acid and polyamine pathways, offering new insights into prostate carcinogenesis and potential microbiota-targeted therapies (189).

Basal cell markers p63 and 34βE12 are indispensable tools in PCa histopathology (190). In normal glands, basal cells form a continuous layer expressing nuclear p63 and cytoplasmic 34βE12 (191-193). Their absence is a hallmark of adenocarcinoma, reflecting the replacement of normal basal epithelium by malignant cells (194,195).

Compared with p63, which is nuclear and occasionally affected by tissue fixation, 34βE12 provides stable cytoplasmic staining, allowing reliable detection even in suboptimally preserved specimens (196,197). During prostate biopsy evaluation, the concurrent loss of both markers indicates carcinoma, with diagnostic sensitivity >95% and specificity >90% (198-200). When basal cells appear partially positive, additional markers such as PSA or α-methylacyl-CoA racemase are used to confirm malignancy (201,202).

Recent advances combine p63 and 34βE12 immunostaining with AI-assisted pathology, markedly improving interpretive accuracy and reducing observer variability (203,204). Single-cell sequencing and organoid models have further elucidated the dynamic expression of these markers across tumor grades, associating their loss with malignancy progression, recurrence risk and treatment outcomes (205-207).

DNA methylation represents one of the earliest and most consistent molecular alterations in PCa. Among the numerous genes identified, GSTP1 promoter methylation is the most extensively validated (208). GSTP1 encodes glutathione S-transferase, an enzyme involved in detoxification and oxidative defense. Hypermethylation-induced silencing of GSTP1 occurs in >90% of PCa cases but rarely in benign tissues, making it a highly specific diagnostic indicator (209). Combined detection of GSTP1 methylation with urinary PCA3 enhances diagnostic specificity >90% (210).

Other genes such as APC, RASSF1A, PTEN and CDH1 also demonstrate marked clinical relevance (211). APC promoter methylation is observed in 50-70% of cases and is associated with tumor progression and lymph node metastasis (212). RASSF1A methylation frequently co-occurs with GSTP1 and helps detect high-grade tumors with greater accuracy (213). PTEN methylation activates the PI3K/AKT pathway and contributes to castration resistance (214), while CDH1 and SFRP2 methylation associate with increased invasiveness and disease progression (Table III) (215).

Genome-wide methylation profiling has recently identified novel PCa-related epigenetic signatures with potential diagnostic and prognostic applications (216). Combining methylation markers with genetic and RNA-based assays, particularly in liquid biopsy formats, offers a more comprehensive and non-invasive diagnostic strategy (217). Studies also indicate that non-coding RNAs may regulate methylation marker expression, adding another layer of complexity to prostate tumorigenesis (218,219). These findings highlight the promise of methylation-based diagnostics and their integration into precision oncology (Fig. 4) (219).