Our study demonstrates that ADC values can effectively differentiate between different invasive lesions, which is beneficial for the preoperative evaluation of PCa.

This research elucidates that with the elevation of serum PSA levels, the risk of PCa increases progressively [8]. Currently, clinical guidelines advocate for early serum PSA testing among individuals at high risk of PCa, including men aged over 50, those over 45 with a familial history of PCa, men over 40 with PSA > 1 ng/mL, and those over 40 carrying BRCA2 gene mutations [9]. Although conventional norms designate serum PSA levels under 4 ng/mL as within the normal range, it’s important to recognize that PSA levels within this interval do not comprehensively rule out the potential presence of PCa. Accurate diagnosis necessitates a comprehensive approach involving digital rectal examinations, biopsies, and MRI scans, aligning with insights from various studies and clinical guidelines [9, 10]. For PSA > 10 ng/mL, prompt prostate biopsy is recommended. Conversely, in the PSA “gray zone” spanning 4 to 10 ng/mL, an intricate assessment involving the free/total PSA ratio or prostate-specific antigen density is deemed prudent [11].

Presently, the definitive preoperative diagnosis of PCa is achieved through the acquisition of pathological results via needle biopsy—a benchmark procedure. Yet, this invasive method’s detection rate stands at a modest 62.2% [4]. The Gleason scoring system remains the cornerstone for assessing PCa’s malignancy degree [12,13,14]. Its utility extends to gauging biological activity, invasiveness, guiding treatment strategies, and prognostic evaluations. Furthermore, it emerges as a robust predictor of postoperative progression and survival [15]. The GS encompasses the summation of primary and secondary architectural patterns observed in hematoxylin and eosin-stained slides. A higher GS value denotes increased malignancy and invasiveness. Typically, GS ≤ 6 signifies a favorable prognosis and low risk, prompting active surveillance and vigilant monitoring instead of immediate intervention. GS = 7 represents an intermediate risk, often managed with monotherapy. Meanwhile, GS > 7 signifies high risk, warranting aggressive combined therapeutic strategies anchored in radical resection.

Notably, the distribution of Gleason patterns 3 and 4 exerts a bearing on patient prognoses. Consequently, GS = 7 PCa is parsed into two strata: GS 3 + 4 and GS 4 + 3. Evidence indicates diminished overall and specific survival rates for GS 4 + 3 in comparison to GS 3 + 4 [16]. Under specific circumstances like stable PSA levels and limited biopsy tumor volumes, guidelines even advocate for active monitoring among GS 3 + 4 patients [17]. Responding to this, a novel grading paradigm was unveiled in 2014, wherein GS = 7 is no longer treated as a single entity, but bifurcated into GS 3 + 4 (prognostic grade group II) and GS 4 + 3 (prognostic grade group III) [13]. To prevent undue biopsies and treatment in prostate patients, this study chiefly explores the feasibility of employing radiological imaging techniques to distinguish between benign, GS 3 + 3, GS 3 + 4, and GS 4 + 3 lesions.

It should be noted that previous studies often combined biopsy results with resection results; however, some studies [18, 19] have highlighted limitations in biopsies and tumor heterogeneity, leading to potential upgrading of pathological scores after PCa resection. To address this issue comprehensively, our study exclusively relied on radical prostatectomy pathology for all PCa lesions.

PCa lesions generally exhibit indistinct boundaries and irregular shapes, presenting as low signals in the PZ or TZ on T2WI. However, these visual attributes lack specificity, as there exists considerable overlap with images depicting prostate hyperplasia and inflammation. Leveraging the principles of DWI, which captures water diffusion dynamics, yields insights into tissue functionality at the cellular level. The resulting ADC map functions as a model delineating signal decay. Computed from multiple DWI b-values, this metric proves efficacious in unveiling lesions that elude detection through T2WI and standard DWI. Consequently, it demonstrates enhanced accuracy in localizing and detecting PCa [20]. Given that tumor tissues exhibit an augmented cellular composition, water molecule diffusion is impeded, consequently resulting in a distinct reduction in ADC values. The quantification of these ADC values serves as an objective measure of the degree of restricted water diffusion within the tissue [21].

This study’s outcomes manifest a significant diminution in both the median ADCmin and ADCmean values for GS 4 + 3 when contrasted with GS 3 + 4. Additionally, the findings affirm heightened sensitivity and specificity when deploying ADCmin and ADCmean for differentiating various low to intermediate-risk PCa grades. The inverse correlation observed between GS values and ADCmin/ADCmean substantiates the assertion that as the malignancy of prostate lesions intensifies, ADC values correspondingly decline, potentially portending an unfavorable PCa prognosis. Remarkably, these conclusions are aligned with the investigations of Saito et al [22] and Bajgiran et al [23]. Within the context of lesions spanning diverse low to intermediate-risk PCa grades, both ADCmin and ADCmean exhibit diagnostic value, facilitating the nuanced prediction of risk stratification pertinent to prognosis. This capacity extends further, enabling the differentiation of higher-invasiveness PCa from an imaging perspective.

Upon juxtaposing the diagnostic performance of ADCmin and ADCmean, it becomes evident that ADCmean consistently outperforms ADCmin in terms of sensitivity, specificity, the Youden index, and the AUC for low to intermediate-risk PCa diagnosis. This emphasis underscores the heightened discriminating power of ADCmean in delineating PCa instances, corroborating the insights drawn from the work of Yang et al [24]. This phenomenon is potentially grounded in the heterogeneous composition of PCa tissue [25]. While ADCmin emphasizes regions characterized by the most prolific cellular proliferation and the densest distribution within tumor tissue—portraying the most potent constituents—ADCmean reflects the holistic composition of the lesion. Moreover, considering that the GS amalgamates primary and secondary architectural patterns within cancerous tissue, it is postulated that ADCmean encapsulates the lesion’s comprehensive nature most effectively, warranting a recommendation for radiologists to prioritize the integration of ADCmean in their clinical practice.

This study is subject to several limitations. First, this study is a retrospective clinical study, which may entail some selection bias. Second, pathological results for benign patients were obtained through needle biopsy, which might involve a certain risk of missed diagnosis. Furthermore, the size of the selected ROI could affect the measurement of ADC values, thereby introducing some degree of error. Finally, this study is a single-center retrospective study, and whether the derived ADC threshold is applicable to other models requires further multicenter research.

In conclusion, a clear negative correlation between ADC values and PC’s invasiveness is evident. This inverse relationship suggests that as ADC values decline, the probability of malignancy in prostate lesions rises. ADC values exhibit remarkable sensitivity and specificity in diagnosing low to intermediate-risk PCa. Notably, ADCmean proves more valuable than ADCmin in diagnosing low to intermediate-risk PCa. The diagnostic potential of ADC values is notably heightened for patients with G = 7, offering a robust tool for accurate preoperative assessment of tumor invasiveness. This, in turn, aids clinicians in devising tailored treatment strategies and making informed prognosis assessments. Furthermore, it serves as a means to prevent unnecessary biopsies and treatments. Additionally, radiologists are encouraged to prioritize the integration of ADCmean for enhanced diagnostic precision. However, it’s essential to acknowledge the aforementioned limitations when interpreting the study’s outcomes.