Prostate cancer screening has entered a new era with the integration of MR imaging and artificial intelligence (AI) into diagnostic workflows. This paradigm shift allows for the more accurate identification of clinically significant prostate cancer while reducing the overdiagnosis of low-grade disease. AI-driven image reconstruction and lesion detection platforms are improving diagnostic accuracy and access. This article synthesizes recent advances in MR imaging-based prostate cancer screening, focusing on the evidence for its utility, risk-adjusted screening strategies, emerging AI applications, and future directions aimed at increasing precision, efficiency, and equity in prostate cancer care.
Key points
-
•
Prostate-specific antigen (PSA)-based prostate cancer screening reduces cancer-related mortality but leads to overdiagnosis and overtreatment.
-
•
Prostate MR imaging improves detection of clinically significant cancers, reduces overdiagnosis of low-grade disease, and helps avoid unnecessary biopsies.
-
•
International guidelines are increasingly supporting MR imaging-first strategies, especially for high-risk individuals.
-
•
Emerging artificial intelligence platforms are enhancing the accuracy and consistency of identifying clinically significant disease, potentially reducing costs and clinician variability.
Abbreviations
| ACR | American College of Radiology |
| AI | artificial intelligence |
| AUA | American Urological Association |
| bpMRI | biparametric MR imaging |
| csPCa | clinically significant prostate cancer |
| DCE | dynamic contrast-enhanced |
| EANM | European Association of Nuclear Medicine |
| EAU | European Association of Urology |
| ESTRO | European Society for Radiotherapy and Oncology |
| ESUR | European Society of Urogenital Radiology |
| ISUP | International Society of Urologic Pathology |
| ISUP GG | International Society of Urological Pathology Grade Group |
| mpMRI | multiparametric MR imaging |
| MRI | MR imaging |
| NCCN | National Comprehensive Cancer Network |
| PCa | prostate cancer |
| PI-RADS | Prostate Imaging Reporting and Data System |
| PIVOT | Prostate Cancer Intervention versus Observation Trial |
| PROMIS | PROstate MR Imaging Study |
| ProtecT | Prostate Testing for Cancer and Treatment |
| PSA | prostate-specific antigen |
| SIOG | International Society of Geriatric Oncology |
| SOU | Society of Urologic Oncology |
| US | ultrasound |
Introduction
Advances in imaging and artificial intelligence (AI) have revolutionized the prostate cancer (PCa) screening landscape, shifting the paradigm toward precision diagnosis.
Epidemiology and Diagnosis
Worldwide PCa is the second most frequent cancer and fifth leading cause of cancer-related mortality. In the United States, PCa is the most commonly diagnosed malignancy in male individuals accounting for almost one-third of cancer diagnoses, affecting approximately 1 in 8 male individuals during their lifetime. , While incredibly common, the 5 year relative survival of PCa is high in the United States at approximately 97%, owing in large part to the detection of localized low-grade disease. PCa is a disease of older men, more common among non-Hispanic Blacks, and exhibits a relatively strong heritable component. , The disease is diagnosed by histopathology of prostate tissue samples typically acquired using a transrectal or transperineal ultrasound (US) core biopsy approach in men suspected of harboring PCa. The individual tissue cores are microscopically assessed for the presence of PCa using the Gleason pattern system and assigned a composite Gleason score, which is composed of the sum of the 2 most prominent patterns observed. The Gleason score assigned by pathologists directly corresponds to a grade group in the International Society of Urological Pathology Grade Group (ISUP GG) system, which is used globally ( Table 1 ).
Table 1
International Society of Urologic Pathology grading system
| ISUP Grade Group | Gleason Score |
|---|---|
| 1 | 3 + 3 |
| 2 | 3 + 4 |
| 3 | 4 + 3 |
| 4 |
4 + 4
3 + 5 5 + 3 |
| 5 |
4 + 5
5 + 4 5 + 5 |
Screening
Prostate-Specific Antigen-Based Screening
Prostate-specific antigen (PSA) is a serum protein biomarker with a well-established role in monitoring established disease progression, but its use in PCa screening is controversial. Several disease processes can lead to an elevated serum PSA level, including benign prostatic hyperplasia, prostatitis, prostatic trauma, and PCa. The European Randomized Study of Screening for Prostate Cancer trial evaluated the effectiveness of PSA-based screening in reducing PCa-related mortality, observing that while screening significantly reduced the number of deaths attributed to PCa, it was associated with a higher risk of overdiagnosis. ,, In contrast, the Prostate Cancer Intervention versus Observation Trial (PIVOT) evaluated the effectiveness of prostatectomy versus observation in men with localized PCa detected by PSA-based screening. The study reported no significant difference in all-cause or PCa mortality, between the surgery and observation groups at up to 20 years. , The Prostate Testing for Cancer and Treatment (ProtecT) trial also studied the outcomes of patients with PCa diagnosed by PSA-based screening who underwent monitoring or treatment. Similar to the PIVOT trial, the ProtecT investigators found no significant difference in the rates of all-cause and PCa-specific mortality between patients who underwent active monitoring and those treated with prostatectomy or definitive radiation therapy at 10 years. The Cluster Randomised Trial of PSA Testing for Prostate Cancer (CAP) evaluated the effect of PSA screening on PCa mortality at 10 years in men aged 50 to 69 years compared to men not undergoing PSA screening. CAP investigators reported no significant difference in PCa mortality between groups, but in the PSA screening cohort, more men were diagnosed with low grade (ISUP GG1) PCa. These studies, and others, suggest that treatment may be unnecessary for a considerable proportion of men harboring low-grade disease as it may remain indolent during their lifetime. In fact, the PIVOT study reported higher rates of urinary incontinence and sexual dysfunction in treated patients, indicating that intervention may cause more harm.
MR Imaging-Guided Cancer Detection
Recognizing the problem of overdiagnosis and overtreatment, significant investigative efforts were undertaken toward refining screening approaches that led to studies evaluating the usefulness of prostate MR imaging in diagnosis. The intention of incorporating MR imaging was to triage patients to enrich detection of clinically significant disease (most often considered as ISUP GG ≥2), while avoiding detection of low-grade disease (ISUP GG1), as well as to localize the anatomic location of the disease within the prostate to facilitate MR imaging-US targeted biopsy and ultimately pathologic confirmation. For the purposes of this article, clinically significant prostate cancer (csPCa) will be referred to as ISUP GG 2 or greater, unless otherwise stated.
One of the earliest high-quality studies illustrating the utility of MR imaging in diagnosing csPCa was the PROstate MR Imaging Study (PROMIS), published in the Lancet in 2017. PROMIS was a multicenter paired-cohort confirmatory study conducted in the United Kingdom of biopsy-naïve men suspected of harboring csPCa, which they defined as ISUP GG 3 or greater. The authors reported superior detection of csPCa by multiparametric MR imaging (mpMRI) over transrectal US-guided biopsy alone as well as a reduction in the detection of clinically insignificant disease. The following year, the PRostate Evaluation for Clinically Important Disease: Sampling Using Image-guidance Or Not? (PRECISION) study was published showing that MR imaging-US guided targeted biopsy was superior for detecting csPCa and its utilization could reduce overdiagnosis of low-grade disease. In 2019, the 4M and MR imaging-FIRST trials were published, both providing strong evidence for incorporation of MR imaging into diagnostic workflows. Indeed, the 4M trial supported the conclusions of the preceding PROMIS and PRECISION studies in establishing the usefulness of MR imaging in detecting csPCa. However, the MR imaging-FIRST study placed nuance on the prior results illustrating a continued role for systematic biopsy in helping capture csPCa that may be missed by MR imaging. , The subsequently published TRIO study further illustrated that a combined approach, MR imaging-US targeted biopsy with systematic biopsy, gives the most representative picture of disease residing in the prostate. While an MR imaging-US targeted approach with systematic biopsy appears to better capture most cases of csPCa, it does so through more biopsies and detection of more ISUP-GG1 cancers. The STHLM3-MRI study, a randomized noninferiority trial, compared MR imaging-targeted plus systematic biopsy to systematic biopsy alone. The study reported a higher detection rate for csPCa and a lower detection rate for ISUP GG1 disease concluding the MR imaging-targeted plus systematic approach was noninferior.
Since PSA-based screening and systematic biopsy of at-risk patients leads to overdiagnosis and unnecessary biopsies, and MR imaging clearly adds value in diagnosing disease, more recent investigations have evaluated the effectiveness and safety of MR imaging-guided screening efforts to minimize overdiagnosis. The GÖTEBORG-2 study, a large prospective randomized clinical trial conducted in Sweden, compared MR imaging with targeted biopsy to MR imaging-targeted biopsy plus systematic biopsy for PCa detection in men with a PSA 3 ng/mL or greater. GÖTEBORG-2 demonstrated that MR imaging-targeted biopsy alone detects less ISUP GG1 disease and resulted in fewer biopsies. However, a criticism of the GÖTEBORG-2 study was that it did not account for the patients that may have been missed in the MR imaging-targeted only group. That is, is it safe and effective to use an MR imaging only approach? To address this question, the GÖTEBORG-2 investigators followed patients over multiple rounds of screening with the assumption that csPCa would reveal itself at imaging. In 2024, the results for 4 years of follow-up were published demonstrating that MR imaging-targeted biopsy detected approximately 84% of csPCa cases detected by systematic biopsy and the risk of missing aggressive cancers was low, less than 0.5%. These results suggest that it may be safe to forego biopsy in patients with a negative MR imaging.
Risk-Adjusted Patient Selection
The risk of developing PCa is highly influenced by heritable genetic factors, which is estimated to at 58%. , The minority of genetic variants contributing high risk of developing PCa is monoallelic mutations in DNA-repair pathway genes and may lead to more aggressive disease ( Table 2 ). This has led to MR imaging-based screening efforts for pathogenic germline carriers with promising early results reported for the Prostate Cancer Genetic Risk Evaluation Screening Study. However, the majority of genetic risk is conferred by inheritance of commonly occurring low-risk variants referred to as single nucleotide polymorphisms that have been established through genome-wide association studies. In total, over 450 variants have been attributed to an elevated risk of PCa, which have been used to derive genetic risk scores. Using a polygenic risk score adapted PCa screening approach, the recently published BARCODE1 study in the United Kingdom found that by inviting patients with risk scores in the 90th percentile or higher to undergo MR imaging and targeted MR imaging-US transperineal biopsy, they identified a higher percentage of patients with csPCa than would have been identified using PSA or MR imaging alone. Indeed, the study reported a cancer detection rate of 40% with 55.1% of patients harboring intermediate or higher risk disease. Integration of plasma biomarkers, genetic markers, and clinical variables has also been shown to better stratify patient risk before MR imaging. Prioritization of higher risk patients may be helpful in improving MR imaging access, particularly as MR imaging utilization is expected to rise with growing new PCa cases.
Table 2
Most common monogenic high-risk variants
| Gene | Nonprostate Cancer Predisposition |
|---|---|
| ATM | Breast and pancreatic |
| BRCA1 | Breast and ovary |
| BRCA2 | Breast, ovary, and pancreas |
| CHEK2 | Breast |
| HOXB13 | – |
| MSH2 /MSH6/MLH1/PMS2 | Colorectal, endometrial, and ovarian |
| NBN | Breast endometrial, and ovarian |
| PALB2 | Breast, ovary, and pancreas |
| RAD51C /D | Breast and ovary |
Current Screening Guidelines
The European Association of Urology (EAU), combined with other stakeholder organizations, updated their guidelines in 2024, emphasizing a risk-based screening approach that incorporates MR imaging in the diagnostic pathway. The EAU-European Association of Nuclear Medicine (EANM)-European Society for Radiotherapy and Oncology (ESTRO)-European Society of Urogenital Radiology (ESUR)-International Society of Urologic Pathology (ISUP)-International Society of Geriatric Oncology (SIOG) guidelines promotes a patient centered approach for screening, typically beginning at age 50 years, and recommends MR imaging-based screening protocols to enhance detection of clinically significant PCa and avoid unnecessary biopsies. The American Urological Association (AUA) and Society of Urologic Oncology (SUO) published updated guidelines in 2023 for early detection of PCa, which acknowledge the growing role of MR imaging in refining screening strategies. , These guidelines emphasize shared decision-making and risk stratification in screening protocols. The National Comprehensive Cancer Network (NCCN) also strongly recommends that multiparametric MR imaging precede tissue sampling in at-risk patients undergoing biopsy toward utilizing image-guided lesion-targeting technologies. For imaging-based PCa detection recommendations, the current American College of Radiology (ACR) Appropriateness Criteria lists MR imaging as “Usually Appropriate” for the initial detection of biopsy-naïve patients suspected of harboring PCa. These guidelines are summarized in Table 3 .
Table 3
Societal recommendations regarding MR imaging utilization in the initial diagnosis of patients at risk for prostate cancer
| Society | Recommendation |
|---|---|
| AUA and SOU |
|
| EAU-EANM-ESTRO-ESUR- ISUP-SIOG | MR imaging is recommended for helping better stratify intermediate-risk patients for determining the need for biopsy and prior to biopsy for lesion targeting purposes |
| NCCN | For patients undergoing biopsy, it is strongly recommended that multiparametric MR imaging precede biopsy so that targeted MR imaging-US may be employed |
| ACR Appropriateness Criteria | Prostate MR imaging is usually appropriate as an initial imaging modality for biopsy-naïve patients suspected of harboring PCa |
Prostate MR imaging technique, interpretation, and quality
Technical and Reporting Standards
To promote uniform, high-quality image acquisition, imaging centers should adhere to the technical parameters outlined in the Prostate Imaging Reporting and Data System (PI-RADS) version 2.1 manual. A PI-RADS v2.1 compliant MR imaging is composed of multiple sequences including triplane T 2 -weighted images, high b -value ( b ≥1400 s/mm 2) diffusion-weighted imaging, and T 1 -weighted dynamic contrast-enhanced (DCE) images. The combination of these sequences comprises a mpMRI, which provides comprehensive tissue characterization and improved diagnostic accuracy. Whenever possible, higher field strength machines (3 T) should be used to achieve images with higher signal-to-noise. Similar to promoting uniformity in high-quality image acquisition, reporting should also adhere to PI-RADS v2.1 reporting standards. Examples of patients with clinically significant cancers in the peripheral and transition zones are illustrated in Figs. 1 and 2 .






