Diffusion is Brownian motion of free water molecules at the cellular level. Diffusion-weighted imaging can add valuable information about a tissue at the cellular level to the information from conventional T1-weighted and T2-weighted imaging [51]. Because diffusion-weighted imaging measures the Brownian motion of water molecules, it provides important information about the functional environment of water in tissue and reflects the cellular status of normal and pathologic tissue. Furthermore, diffusion-weighted imaging is sensitive to changes in the microdiffusion of water within the intracellular space and extracellular space and cytotoxic edema due to alterations in the adenosine triphosphate-dependent sodium–potassium pumps. Diffusion can be quantitated in terms of the ADC on diffusion-weighted MR images. The ADC quantifies the combined effects of both diffusion and capillary perfusion. A decreased ADC is interpreted as reduced motion of water molecules or diffusion, whereas in tissue with an increased ADC there is less restriction of water molecule motion. In prostate tumors, as in other solid tumors, the motion of water molecules is restricted because of their increased cellular density, with reduction of the extracellular space and restriction of the motion of a larger portion of water molecules to the intracellular space, resulting in decreased ADC. Therefore, diffusion-weighted imaging and ADC provide powerful indicators for characterization of prostate tissue, particularly in differentiation between benign and malignant lesions [52].

To optimize the ability of diffusion-weighted imaging to allow characterization of the diffusion properties of a tissue, imaging parameters should be optimized according to the imaging unit and magnetic field strength used. Selection of the correct b value is particularly important for prostate tissue with a prolonged T2 relaxation time. The b value specifies the sensitivity of diffusion. Use of high b values increases diffusion sensitivity by diminishing the hyperintensity of tissues with long T2 relaxation times (thus avoiding T2 shine-through). In general, a high b value (>700 s/mm² at 1.5 T and >1,000 s/mm² at 3 T) is recommended for prostate diffusion-weighted imaging, and a postprocessed ADC map is absolutely necessary for identifying and evaluating prostate tumors both objectively and subjectively. The higher the b value, the lower the signal-to-noise ratio; thus, an incremental increase in the number of excitations is required for optimization of image quality. Although studies have shown significant differences between cancerous and normal prostatic tissue in ADC [52, 53], this should be interpreted carefully in clinical patient care due to the current technical variability and lack of consensus. The advantages of diffusion-weighted imaging are a relatively short acquisition time and high contrast resolution between tumors and normal tissue. In general, regional ADC map values differ depending on location and tissue composition. Malignant lesions have lower ADC values (about 20–40 %) than benign or normal prostatic tissue; also, there are regional variations in the normal tissue values for different zones of the prostate (Fig. 8.8). Some authors have stratified ADC values into benign and malignant and demonstrated that diffusion-weighted imaging and ADC mapping can increase the sensitivity of MR imaging in detection of prostate cancer when diffusion-weighted imaging is used in conjunction with T2-weighted imaging [45] (Fig. 8.9); in particular, the addition of DWI imaging to T2-weighted MR imaging significantly improved sensitivity to 81 % (sensitivity for T2-weighted MR imaging alone, 54 %), whereas specificity was slightly lower for T2-weighted MR imaging combined with DWI (84 %) than for T2-weighted MR imaging alone (91 %). Also, in other prospective studies [54, 55], the addition of DW imaging to T2-weighted MR imaging improved prostate cancer localization performance. However, in a recent retrospective 3-T study in 51 patients, with prostatectomy specimens as reference standard [56], DW imaging did not add value to T2-weighted MR imaging for prostate cancer localization. Preliminary results suggest that diffusion-weighted imaging has the potential to support prediction of tumor aggressiveness [57].

The location of prostate cancer affects the sensitivity of diffusion-weighted imaging. Noncancerous peripheral zone tissue has been found to have higher average ADC (less overlap with cancerous tissue) than the transition zone and prostate base. Overlap limits the ability to differentiate prostate cancer from noncancerous tissue [58]. The high prevalence of BPH in elderly men significantly contributes to this difficulty. The stromal form of BPH in particular exhibits lower ADC and low T2 signal intensity, mimicking prostate cancer, whereas glandular BPH and prostatic intraepithelial neoplasia can be more readily distinguished because of their higher average ADC and higher T2 signal intensity [59] (Figs. 8.7 and 8.10). A higher Gleason score has repeatedly been shown to be associated with decreased ADC, likely due to the dedifferentiated infiltrative growth of these tumors, as opposed to the glandular organization of a more well-differentiated prostate cancer, which more closely resembles normal prostatic tissue [60, 61] (Fig. 8.11).

The principal pitfall of diffusion-weighted imaging includes a potential for false-positive findings of tumor related to postbiopsy hemorrhage: it has been reported that postbiopsy hemorrhage lowers the ADC of benign peripheral zone tissue and therefore limits the usefulness of diffusion-weighted imaging in this setting [60]. However, more recent data demonstrated excellent ability of the ADC in differentiation of prostate cancer from hemorrhage in the peripheral zone, and it was suggested that delayed imaging after biopsy may not be necessary [61] (Figs. 8.4 and 8.5).

The shortcomings include susceptibility-induced distortions, which are caused by susceptibility effects from an air-filled rectum or endorectal coil balloon, bone–tissue interfaces, poor local magnetic field homogeneity, and chemical shift artifacts caused by periprostatic fat.

Dynamic contrast-enhanced MR imaging was developed to help achieve a higher accuracy in prostate cancer localization and staging than that obtained with conventional T2-weighted MR imaging. DCE MR imaging is an advanced, fast, dynamic imaging technique (e.g., temporal resolution <5–10 s) that allows derivation of parameters that are closely related to microvascular properties and angiogenesis in tissues. As a response to the tumor hypoxia, there is a strong expression of angiogenesis-inducing factors, such as vascular endothelial growth factor and fibroblast growth factor receptor in their endothelial cells [62], with budding of new blood vessels from existing blood vessels (angiogenesis) or de novo formation of blood vessels (vasculogenesis). Tumor neovessels are in general more permeable than normal vessels, more heterogeneous in size and branching pattern, and more disorganized, making prostate tumors highly vascular and manifesting early hyperenhancement (higher and earlier peak enhancement than in normal tissue) and rapid washout of contrast material from the tumor, in comparison with normal prostate tissue. Microvascular alterations and neovascularity are in general most severe in prostate cancer, in comparison with other processes in the prostate such as BPH or prostatic intraepithelial neoplasia [63].

Numerous contrast enhancement parameters can be used to differentiate cancerous from benign tissue, including onset time, time to peak enhancement, peak enhancement, relative peak enhancement, and washout time; results of numerous studies suggest that the peak enhancement of cancer relative to that of surrounding benign tissue is the most accurate parameter for cancer localization [64] (Fig. 8.12). Typically, prostate cancer appears as an early and rapidly enhancing lesion with fast washout of contrast material in comparison with that of normal prostate tissue. The presence of rapid washout is highly indicative of prostate cancer [65], even in the absence of low T2 signal intensity. On contrast-enhanced MR images, the peripheral zone enhances more than the transition or central zone. In prostate cancer, onset time and time to peak enhancement are lower and peak enhancement is higher than in high-grade prostatic intraepithelial neoplasia and chronic inflammation, and abnormalities are more distinct in high-grade prostate cancer [66]. All DCE MR imaging parameters can be converted into pseudocolor parametric maps and overlaid on the anatomic T1- and T2-weighted images for interpretation. The contrast resolution is similar to that seen on T2-weighted images.

Results of studies indicated that dynamic contrast-enhanced MR imaging can help improve the accuracy of prostate cancer staging. Localization accuracy with dynamic contrast-enhanced MR imaging increased to 72–91 %, as compared with 69–72 % for anatomic T2-weighted MR imaging only [49, 67–69]. DCE MR imaging in combination with MR spectroscopy allows detection of prostate cancer in 46 % of patients with prior negative transrectal US-guided biopsy results and a persistently elevated PSA level (4–10 ng/mL) versus a prostate cancer detection rate of 24 % with repeat transrectal US-guided biopsy in these patients [66] (Fig. 8.13). Although much progress has been made, challenges for DCE MR imaging remain. Benign prostate tissue (e.g., some BPH nodules) is also highly vascular, limiting the overall accuracy of DCE MR imaging in tumor detection, in particular the differentiation of prostatitis from cancer in the peripheral zone and BPH from transition zone tumors [63].

Prostatic intraepithelial neoplasia and well-differentiated prostate cancer demonstrate less neovascularity and permeability changes than higher-grade prostate cancer, therefore representing a source of false-negative results at DCE MR imaging. Differentiation of chronic prostatitis from low-grade prostate cancer may not be possible, while differences were found in DCE MR imaging parameters between high-grade prostate cancer and chronic prostatitis [70]. In addition, low tumor volumes and infiltrative prostate cancer are affected by partial volume effects and are more difficult to detect. However, continuous evolution of rapid imaging techniques has allowed well-designed DCE MR imaging to improve the accuracy of intraprostatic detection of prostate cancer [47–49].

Criteria for extracapsular extension or seminal vesicle invasion at DCE MR imaging include abnormally high or asymmetric peak enhancement, contrast agent washout, and short onset time and time to peak enhancement. In the case of extracapsular extension, these findings are detected near the neurovascular bundle or rectoprostatic angle, broadly abutting the capsule, or in an extracapsular location. In seminal vesicle invasion, these abnormalities are found in the lumen of the ejaculatory ducts or in the seminal vesicles themselves and are associated with wall thickening of the ejaculatory ducts (Fig. 8.14). DCE MR imaging can improve the staging accuracy of less-experienced readers for detection of capsular penetration and seminal vesicle invasion and can allow detection of cancers that are not apparent on T2-weighted images. In addition, in multifocal cancers, DCE MR imaging may demonstrate additional foci of higher stage than the known lesions, leading to adequate upstaging of tumors [71].

Because MR spectroscopy provides metabolic information about prostatic tissue by displaying the relative concentrations of chemical compounds within contiguous small volumes of interest (voxels), it is increasingly being used as a biomarker for detection of cancers, including prostate cancer [72]. MR spectroscopy is also known as chemical shift imaging. Chemical shift is the physical phenomenon in which the electron cloud surrounding the imaged nucleus (in prostate MR spectroscopy, hydrogen nuclei = protons) shields the nucleus partially from the external field and is highly dependent on the chemical environment (chemical bond, adjacent molecules, or atoms) and the molecule the proton is located in. Some chemical compounds have simple spectra (a singlet or single spectral peak) while others possess highly coupled spectra with multiple spectral main peaks, each of which may be composed of multiplets (doublet, triplet, etc.). The analyzed metabolites are choline, citrate, creatine, and polyamine.

Choline is represented by its distinct methyl proton resonance, which forms a composite peak of phospholipid cell membrane components (e.g., phosphocholine, glycerophosphocholine, and free choline) at 3.2 ppm. Owing to elevated cell membrane turnover (phospholipid synthesis and degradation) and increased cell surface compared with cell volume in cellular tumors, the choline level is elevated in the proliferating malignant tissue. Increased choline signal or concentration is considered the spectroscopic hallmark of cancer [73]; however, it has also been found in benign conditions of the prostate such as prostatitis [74].

Citrate is a normal secretion of the healthy prostatic glandular epithelium, and normal prostate tissue contains high levels of citrate (higher in the peripheral zone than in the central and transition zones) [75]; glandular hyperplastic nodules, however, can demonstrate citrate levels as high as those observed in the peripheral zone. The normal prostate has an MR spectrum with a prominent citrate peak at 2.6 ppm, which is usually seen as a doublet with occasional visualization of small additional side peaks. In the presence of prostate cancer, the citrate level is diminished or undetectable because of a conversion from citrate-producing to citrate-oxidating metabolism, but reduced levels can be found also in prostatitis and hemorrhage.

Creatine resonates at 3.0 ppm and is related to energy metabolism. Normal prostatic tissue contains high polyamine levels, of which spermine is a predominant component. Polyamines are represented by a relatively broad spectral peak between creatine and choline (3.1 ppm) that shows significant overlap with creatine and choline. Polyamine levels are reduced in prostate cancer because of higher metabolism.

Lipids cause signal in a broad range at the lower end of the citrate peak (typically 1.3 ppm). Adequate placement of saturation bands around the prostate is important to avoid contamination of the spectra by inclusion of extraprostatic fat, which causes large lipid signals.

Typically, the method for depicting tumors is based on an increased choline/citrate ratio. Because the creatine peak is very close to the choline peak in the spectral trace, the two may be inseparable; therefore, for practical purposes, the (choline + creatine)/citrate ratio is used as a biomarker in prostate tumor detection. An increase in the choline-to-citrate ratio or the (choline + creatine)/citrate ratio is often used as a marker of malignancy in prostate cancer and increases the specificity of diagnosis; however, it is most reliable in the peripheral zone (Figs. 8.12 and 8.15).

There is no consensus about spectral interpretation. The classification system described by Kurhanewicz et al. [76] is often used. In that system, a voxel is classified as normal, suspicious for cancer, or very suspicious for cancer. Furthermore, a voxel may contain nondiagnostic levels of metabolites or artifacts that obscure the metabolite frequency range. Voxels are considered suspicious for cancer if the ratio of choline and creatine to citrate is at least 2 standard deviations (SDs) higher than the average ratio for the normal peripheral zone. Voxels are considered very suspicious for cancer if the ratio of choline and creatine to citrate is higher than 3 SDs above the average ratio [77]. Voxels considered nondiagnostic contain no metabolites with signal-to-noise ratios greater than 5. In voxels in which only one metabolite is detectable, the other metabolites are assigned a value equivalent to the SD of noise.

Although MR spectroscopy shows promise as a problem-solving modality with high specificity, it is limited by low sensitivity. Partial volume effects may obscure the presence of prostate cancer, especially small or infiltrative lesions, which are obscured by strong signals from glandular BPH or surrounding normal tissue present in the MR spectroscopic voxels. By shifting the MR spectroscopic voxels to a position centered on the area of suspected abnormality, a voxel containing the highest possible partition of abnormal tissue can be generated; with this method, some of the lost information can occasionally be recovered.

The combined use of MR imaging and MR spectroscopy improves detection of tumors within the peripheral zone [44, 77, 78]; it has also been shown to increase the specificity in the localization of prostate cancer in the peripheral zone. MR spectroscopic imaging has shown higher specificity (68–99 %) and lower sensitivity (25–80 %) for prostate cancer localization, when compared with anatomic T2-weighted MR imaging (specificity, 61–90 %; sensitivity, 68–87 %) in prospective studies with prostatectomy specimens as reference standard [44, 46, 49, 79]; however, a multicenter trial that included 110 patients, with prostatectomy findings as reference standard, did not show any benefit for the addition of 1.5-T MR spectroscopic imaging to T2-weighted MR imaging in prostate cancer localization [80].