By: Dan Sperling, MD

With the rapidly changing prostate cancer (PCa) treatment landscape, men diagnosed with early stage, low risk PCa may be candidates for either focal treatment (targeting just the tumor using ablation) or conservative management (active surveillance in order to delay treatment). Clinical ethics requires that patients be carefully qualified to determine the safety of pursuing either strategy in order to neither leave active cancer behind (focal treatment) nor miss a treatment window due to cancer progression (active surveillance). The problem is that conventional ultrasound-guided systematic 12-14 core biopsies are inadequate to give a true picture of the amount of PCa, the location(s), and even the highest grade of aggressiveness. While incomplete information may be excusable if a patient is to undergo a radical treatment, it is arguably irresponsible to base any other treatment decision on less than a thorough map of an individual’s disease.

A team of researchers out of Tokyo, Japan studied the imaging, biopsy and post-prostatectomy pathology results of 53 consecutive patients to determine the most efficacious way to map the presence of tumors.[i] They conducted their research retrospectively, meaning that all imaging, needle biopsies and prostatectomies had been completed before they began their analysis of the data. The pathology reports from the laboratory analysis of the removed gland specimens would be used to determine the actual size, site and tumor grade of each patient. However, the post-surgery pathology results were not revealed until experienced readers had examined the images and the biopsy results for the presence or absence of PCa.

The Japanese team set out to compare two specific MRI parameters to see if one performed better than the other: Diffusion Weighted Imaging (DWI) and MR Spectroscopy (MR-S).

Diffusion Weighted Imaging (DWI) is a method of magnetic resonance imaging that reflects thermal molecular motion of water molecules in tissues. Water molecules exhibit free random motion if unrestricted. Tissues with high cellular density restrict molecular motion (less diffusion), and changes in the magnetic resonance signal reflect this. An experienced reader who is familiar with the diffusion characteristics of healthy prostate tissue in its zonal anatomy can use DWI to determine if abnormalities represent benign tissue changes, inflammation, or malignancy.

MR Spectroscopy provides biochemical information about tissues. Different chemical nuclei such as carbon, sodium, fluorine and phosphorus give off specific signals that can be picked up by MR-S. In human tissue, specific biochemicals called metabolites can be detected at levels that are indicative of tumors. These metabolites include choline, creatine, and inositol, among others. MR-S is gaining wider use in the detection of brain, breast and prostate cancer. It can help rule in or out the need for a biopsy. As with DWI, the experience level of the reader is important.

The purpose of the Japanese study was to explore how well DWI and MR-S could diagnose the precise distribution and characteristics of all malignant prostate tumors, including small lesions. Their larger objective was to contribute toward an accurate means of identifying patients who can safely opt for a focal therapy or active surveillance.

To structure the comparison between the pre-treatment imaging/biopsy results with the actual specimens, the team divided the prostate into a 12-section model or scheme. First, each region was examined on the basis of DWI, MR-S and prostate biopsy results (in that order). Then, these results were correlated and compared with the respective regions in the prostatectomy specimens to assess positive predictive value (PPV, or correct prediction of the presence of cancer), negative predictive value (NPV, or correct prediction of the absence of cancer), and sensitivity (proportion of positives that are accurately predicted) of each diagnostic modality. Based on the imaging and biopsy, the team noted 636 “evaluable lesions” of which they identified 242 as cancer. They then explored the accuracy (compared with the specimens) of DWI, MR-S and biopsy. The results are shown below:

The authors concluded that DWI is more efficient than MR-S and prostate biopsy in detecting the distribution of prostate cancer within the gland.

It is noteworthy that on balance, imaging performed better than biopsy alone. In addition, the use of more than one parameter in detecting tumors increases confidence that the suspicious lesion is, in fact, cancer. Such noninvasive acquisition of information by imaging will increase the safety of choosing a targeted treatment or conservative management.