Introduction:

PSA-based screening has improved the early detection of prostate cancer, resulting in many more men being diagnosed and treated.  Prospective cohort studies have shown that active surveillance is a safe management option for men with low and favorable-intermediate risk prostate cancer. Tools such as prostate biomarkers have been incorporated into the management of prostate cancer risk stratification, which in turn reassures the patient and clinician regarding appropriateness of active surveillance. Another such clinical tool is the multi-parametric prostate MRI. 

The NCCN guidelines’ active surveillance protocol previously recommended consideration of prostate MRI in response to PSA increase if an anterior lesion was suspected. These recommendations have recently changed in favor of expanded utilization of mpMRI. 

An unchanged mpMRI has been associated with an 80% negative predictive value for biopsy upgrading during AS (Henderson et al). However, prior studies have also demonstrated biopsy upgrading (> Gleason 6) in 27% of men with a negative mpMRI, suggesting that mpMRI alone cannot be used to monitor men on AS. The American Urological Association (AUA) Multiparametric Prostate MRI Consensus Panel deems current data to be insufficient regarding repeat MRI without a prostate biopsy for monitoring men on AS. 

Thus, given conflicting data and society stances, it is the goal of this study to evaluate the utility of prostate mpMRI for clinical decision making in patients on active surveillance. 

Methods:

The study was deemed exempt by the Institutional Review Board of University Hospitals, Cleveland Medical Center. Low and favorable-intermediate risk categories were defined according to National Comprehensive Cancer Network (NCCN) guidelines. A retrospective chart review was conducted to identify patients on active surveillance for low or favorable-intermediate risk prostate cancer receiving at least two prostate MRIs at University Hospitals Cleveland Medical Center from 2009-2016.

69 patients on active surveillance with at least two prostate MRI and two prostate biopsies were identified. All MRIs were read and scored by board-certified radiologists using the Prostate Imaging Reporting and Data System (PIRADS) version 2. If the patient had multiple suspicious lesions, the lesion with the highest PIRADS score was included in the analysis and this index lesion was tracked throuhgout subsequent MRIs.  

The utility of prostate MRI was determined by calculating the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the change from the first to the second/subsequent MRI in relation to changes in biopsy pathology. MRI progression was defined as increase in PIRADS score of the previously existing index lesion or the appearance of a new high grade lesion. 

Noting that some patients had an initial mpMRI with a high risk lesion (PIRADS >3) that did not change, we then separately calculated the relative risk of biopsy upgrade among these patients with a high grade lesion at any time during AS compared to those without a high grade lesion. 

Results:

19 out of 69 patients (28%) had progression of PIRADS lesion when followed with ≥2 mpMRI. 50 patients had no progression of MRI lesion (Figure 2). Of those with progression on MRI, 16 were found to have upgrade on prostate biopsy pathology, resulting in a positive predictive value of 84%. An upgrade in MRI lesion was noted to be 91% specific for biopsy progression (95% CI 76% to 98%). All patients who had an upgrade on pathology were noted to upgrade to intermediate risk disease (Gleason 3+4=7 or Gleason 4+3=7). Of the patients who had pathologic upgrade predicted by mpMRI, treatment change was noted in ~50% (data not shown). 

A total of 35 patients had biopsy progression on final pathology. Of these patients, 16 had upgrade of MRI lesion. Thus, the sensitivity of a change in MRI lesion was calculated at 46%. 

A total of 44 patients had a high grade PIRADS lesion detected on mpMRI during AS. Amongst those patients who had a high grade lesion on at any time, the relative risk of biopsy upgrade was 3.3 times greater than those without a high grade PIRADS lesion (95% CI 1.46-7.42; P<0.01).

Conclusion:

When following patients on active surveillance with multiple prostate mpMRIs, most MRIs (~3/4) will not upgrade. A patient with a high risk lesion on mpMRI identified at any time while on surveillance has a 3.3-fold increased risk of progression on subsequent biopsy compared to those without a high risk leasion (PIRADS>3). If initial mpMRI demonstrates a high risk lesion, the majority of the benefit of imaging has been obtained. Clinicians should have a high suspicion for missed lesion and low threshold for re-biopsy. If a patient has upgrading of PIRADS lesion on MRI, there is a high likelihood of biopsy progression, often predicting treatment change (exit from AS). Thus the utility of mpMRI appears to be greatest when determining candidacy for active surveillance or when engaging in shared decision making on active surveillance to obtain additional information on the risk of pathologic upgrade prior to repeat biopsy. 

Funding: N/A