There used to be an off-price clothing store called Syms, described by one journalist as “unlike any department store I’ve seen before or since.” If you were a savvy shopper, Syms was the place for you. Their slogan, “An educated consumer is our best customer,” was one of the greatest in retail history!

Today’s prostate cancer (PCa) patients are able to educate themselves thanks to the internet – though a lot of PCa information found continues to be disappointingly outdated. Still, at our Center we find that a patient who has done a lot of web-based reading asks excellent questions about his own case. In addition to educating our patients so they understand our findings, such questions help us address each individual’s personal concerns.

Not long ago, a patient asked me about a new study whose results suggested that multiparametric MRI (mpMRI) is not as good as other studies have found. In particular, the study results had to do with what is called negative predictive value.

What is meant by predictive value?

The term “predictive value” refers to probability, or to put it simply, chance. Predictive value is rated according to a range of zero to one, where zero is the outcome can never be predicted, and one means correct prediction 100% of the time. Everything in between is a decimal, or percent.

For example, let’s take a coin toss: what is the probability that it will come up heads or tails? If the coin is perfectly balanced, there’s a 50/50 chance of either outcome, so no one can predict the outcome. This is zero predictive value. On the other hand, if Psychic Sam can correctly predict the outcome every single toss, this would be an amazing predictive value of one, or 100%. Now, let’s say Psychic Sam predicts the correct outcome three times out of four (3/4), this would be a 75% predictive value. This is very good probability compared to zero, so if you were a gambler, it might be worth betting on Sam.

How mpMRI accuracy is calculated

When a man’s PSA is abnormally high, the next step (before biopsy) should be a multiparametric MRI done on powerful 3T (3 Tesla) magnet. Hundreds of published clinical studies have been done to test the probability that mpMRI can accurately identify the absence of clinically significant PCa. This type of study typically involves patients who will have their glands removed (prostatectomy) if they are found to have PCa based on mpMRI followed by biopsy. After surgery, the prostates are examined and compared with their pre-surgery MRI scans. “Accuracy” means how well does the scan match the specimen?

Why negative predictive value is important

Thus, negative predictive value means the probability that that a patient’s MRI correctly identified the areas where PCa was absent. Why is this important? According to a 2020 expert review of 42 such studies, “The reported negative predictive value (NPV) of mpMRI is used by some clinicians to aid in decision making about whether or not to proceed to biopsy.”^[i] In other words, an educated patient is likely to ask, “Doctor, my PSA has gone from 3.6 to 5.1 in a year but my MRI doesn’t show any clinically significant cancer. Can I count on the MRI results to reliably skip a biopsy?”

Clinical decision-making

Let’s face it, no one wants to have a needle biopsy of his prostate if it’s not necessary. The problem is, all the clinical studies have shown that the NPV is generally accurate BUT there is a range of reported values from average low end (75%) to average high end (97%) and everything in between. One paper reported that if patients skip a biopsy based only on their MRI results, “Some 7-10% of men, depending on the setting, will miss a diagnosis of clinically significant cancer if they do not proceed to biopsy.”^[ii] Betting on a coin toss is a gamble, but deciding on a biopsy shouldn’t be.

Thankfully, there’s a way to greatly improve negative predictive value! Studies are showing that combining negative MRI results with other factors such as patient age, prostate volume, free PSA, and PSA density pushes NPV to nearly 100%.^[iii] In such cases, patients can safely skip a biopsy.

As far as I can determine, no one has yet published a durable 100% NPV for mpMRI. I recognize this is as frustrating for our educated patients as it is for those of us who serve their clinical interests. We recommend that for personal care, patients turn to highly experienced mpMRI centers, and discuss with their doctors the integration of information from biomarkers like PSA density, free PSA, and genomic analysis before deciding whether or not to biopsy if MRI is negative. In addition, we encourage managing expectations by recognizing that we clinicians continue to improve as we learn from each other’s results.

NOTE: This content is solely for purposes of information and does not substitute for diagnostic or medical advice. Talk to your doctor if you are experiencing pelvic pain, or have any other health concerns or questions of a personal medical nature.

[i] Sathianathen NJ, Omer A, Harriss E, Davies L et al. Negative Predictive Value of Multiparametric Magnetic Resonance Imaging in the Detection of Clinically Significant Prostate Cancer in the Prostate Imaging Reporting and Data System Era: A Systematic Review and Meta-analysis. Eur Urol. 2020 Sep;78(3):402-414.
[ii] Ibid.
[iii] Haack M, Miksch V, Tian Z, Duwe G et al. Negative multiparametric magnetic resonance imaging for prostate cancer: further outcome and consequences. World J Urol. 2022 Dec;40(12):2947-2954.