Prostate cancer is a leading cause of cancer death in men, but many cases are potentially curable if caught early. Current detection relies on prostate-specific antigen (PSA) blood testing and digital rectal examination (DRE), but these methods have significant limitations in specificity -- elevated PSA can be caused by benign conditions such as benign prostatic hyperplasia (BPH) or prostatitis, not just cancer.
When PSA is elevated, patients typically undergo transrectal ultrasound (TRUS) guided biopsy, in which multiple tissue cores are sampled from the prostate. However, prostate tumors are often invisible on ultrasound, and the sampling is not targeted, leading to significant error rates where cancers are missed or biopsies are unnecessary.
Conventional T2-weighted MRI improves anatomical visualization of the prostate -- normal peripheral zone tissue appears bright, while cancerous tissue tends to appear dark -- but its sensitivity is only 60-75% in some zones, particularly the transition zone where BPH creates confusingly similar signal patterns.
There is a recognized need for imaging tools that go beyond anatomy to provide functional, cellular, and metabolic information about prostate tissue, enabling more accurate distinction between cancer and benign disease without invasive biopsy. This review focuses on two such advanced MRI techniques: diffusion-weighted imaging and magnetic resonance spectroscopy.
Diffusion-weighted imaging (DWI) measures how freely water molecules move within tissue. In healthy tissue, water diffuses relatively freely through the extracellular space, but in cancer, increased cell density and tightly packed cells restrict water movement. This restriction appears as bright signal on DWI images.
From DWI data, the apparent diffusion coefficient (ADC) is calculated pixel by pixel to produce a quantitative map. A low ADC value indicates restricted water diffusion -- a hallmark of densely packed cancer cells. Malignant prostate lesions consistently show ADC values approximately 20-40% lower than benign or normal tissue.
ADC maps provide a quantitative biomarker that is directly related to tissue cellularity. When a tumor grows, it increases cell density and reduces extracellular space, lowering ADC. Conversely, when cancer cells die in response to treatment, ADC rises as space opens up -- making DWI/ADC a potential tool for monitoring treatment response.
DWI sequences require careful technical optimization. Multiple b-values (the parameter controlling diffusion sensitivity) must be acquired to calculate ADC maps. Different pulse sequences including echo planar imaging (EPI), fast spin echo, and line scan diffusion offer different trade-offs in spatial resolution, signal-to-noise ratio, and susceptibility to artifacts. The choice of sequence can affect ADC values by up to 20%.
Magnetic resonance spectroscopy (MRS) measures the concentrations of chemical compounds within a tissue region, providing a metabolic fingerprint. In the prostate, the key metabolites are choline, creatine, and citrate. Healthy prostate epithelium secretes citrate at very high concentrations (greater than 60 mM), while cancer significantly reduces citrate production as cells lose their secretory function.
Cancer simultaneously drives up choline levels, which reflect increased membrane synthesis and cell proliferation. The ratio of (choline + creatine) to citrate is therefore a metabolic marker of malignancy: a high ratio indicates cancer, while a low ratio is consistent with normal or benign tissue. This ratio has been validated across numerous clinical studies.
Spectral thresholds for clinical interpretation have been established: peripheral zone tissue with a (choline + creatine)/citrate ratio more than two standard deviations above the mean of normal tissue (~0.22) is considered suspicious, and more than three standard deviations is considered highly suspicious for cancer. A standardized 5-point MRS scoring system has been developed based on these cutoffs.
Technical challenges limit MRS in practice. The citrate signal is complex because of strong magnetic coupling between adjacent protons, causing the signal to split into multiple peaks whose appearance depends on scanner settings, echo time, and field strength. Contamination from adjacent fat and the seminal vesicles can create false-positive results. Despite these challenges, MRS achieves sensitivity and specificity of around 82% and 88% respectively in meta-analysis of combined MRI/MRS studies.
Multiple studies summarized in this review demonstrate that adding DWI/ADC to conventional T2-weighted MRI increases sensitivity from roughly 60-75% to 54-98% and specificity from variable to 58-100%, depending on the study. The wide range reflects differences in imaging protocols, patient populations, and ADC thresholds used. Combining both modalities consistently outperforms either alone.
At higher magnetic field strengths (3 Tesla versus the older 1.5 Tesla standard), both signal-to-noise ratio and spectral resolution improve. At 3T, choline, creatine, and polyamine peaks that blend together at 1.5T become individually distinguishable, potentially improving diagnostic accuracy. Some studies have shown that a surface pelvic coil at 3T can match the performance of an endorectal coil at 1.5T, improving patient comfort.
Emerging 7 Tesla MRI offers dramatically improved spectral resolution, enabling visualization of polyamines such as spermine that are barely detectable at lower field strengths. These metabolites may provide additional cancer markers. While currently experimental, 7T prostate MRS is being explored as a platform for understanding prostate cancer biology at near-histological resolution.
A key strength of ADC mapping is its ability to serve as a treatment response biomarker. As cancer cells respond to therapy and die, the ADC rises -- detectable within days of treatment initiation, before anatomical tumor shrinkage is visible on conventional imaging. This early indicator could guide decisions about continuing, modifying, or stopping treatment.
Advanced prostate cancer frequently spreads to bones -- particularly the axial skeleton including vertebrae, pelvis, ribs, and femur -- in approximately 90% of cases of metastatic prostate cancer (MPC). The spread is typically osteoblastic, meaning tumors stimulate bone formation and replace normal bone marrow.
Traditional detection of metastases uses Tc99m bone scintigraphy, CT, and PET imaging. While PET offers near whole-body coverage, MRI is the most sensitive method for detecting bone lesions. However, conventional MRI is limited to focused regional scanning and cannot efficiently survey the entire skeleton in a single session.
New developments in MRI hardware including advanced gradient systems, radiofrequency coil arrays, and rolling-bed technology now allow whole-body MRI coverage in a single examination. This whole-body DWI (WB-DWI) approach can visualize metastatic lesions across the entire body simultaneously, combining the sensitivity of MRI with the comprehensive coverage of nuclear medicine.
WB-DWI also enables quantitative monitoring of metastatic disease through ADC mapping of individual lesions, potentially providing a more sensitive measure of treatment response than anatomical imaging or bone scanning alone. This approach could replace multiple separate imaging sessions and reduce radiation exposure from nuclear medicine scans.
No single MRI parameter provides sufficient diagnostic accuracy for prostate cancer on its own. The emerging paradigm is multiparametric MRI (mpMRI), which combines T2-weighted imaging, DWI/ADC, MRS, and dynamic contrast-enhanced MRI into a single comprehensive examination that captures anatomical, cellular, and metabolic dimensions of tissue simultaneously.
Each parameter contributes distinct and complementary information: T2 imaging provides structural anatomy, DWI reflects cell density and water diffusion, MRS reveals metabolic activity, and dynamic contrast imaging shows tumor vascularity and permeability. Together, these create a multiparametric dataset that is more powerful than any individual measure for detecting, localizing, and characterizing prostate cancer.
Clinical applications of this multiparametric approach include targeted biopsy guidance (directing needles to suspicious regions identified on imaging), local staging (determining whether cancer has spread beyond the prostate capsule), active surveillance (monitoring low-risk cancer without immediate treatment), and treatment monitoring (detecting early response or recurrence after surgery, radiation, or systemic therapy).
While technical challenges remain -- particularly for MRS -- continued advances in scanner hardware, acquisition sequences, and image analysis software are expected to make multiparametric prostate MRI a routine clinical tool. This could reduce unnecessary biopsies, improve treatment planning, and ultimately translate to better outcomes for patients with prostate cancer.