The Combination of PSMA + MRI vs to MRI Alone in the Diagnostic Performance for Detecting Clinically Significant Prostate Cancer (PRIMARY) - Christopher Wallis & Zachary Klaassen

September 27, 2022

Christopher Wallis and Zachary Klaassen discuss the publication titled, "The Additive Diagnostic Value of Prostate-specific Membrane Antigen Positron Emission Tomography Computed Tomography to Multiparametric Magnetic Resonance Imaging Triage in the Diagnosis of Prostate Cancer (PRIMARY): A Prospective Multicentre Study." This trial aimed to determine whether the combination of PSMA + MRI was superior to MRI in diagnostic performance for detecting clinically significant prostate cancer.

Biographies:

Christopher J.D. Wallis, MD, Ph.D., Assistant Professor in the Division of Urology at the University of Toronto.

Zachary Klaassen, MD, MSc, Urologic Oncologist, Assistant Professor Surgery/Urology at the Medical College of Georgia at Augusta University, Georgia Cancer Center


Read the Full Video Transcript

Christopher Wallis: Hello, and thank you for joining us for this UroToday Journal Club. Today we are discussing the recently published PRIMARY trial assessing the Additive Diagnostic Value of Prostate-specific Membrane Antigen Positron Emission Tomography, Computed Tomography to Multiparametric Magnetic Resonance Imaging Triage in the Diagnosis of Prostate Cancer. This is a prospective multi-center study. I'm Chris Wallis, an Assistant Professor in the Division of Urology at the University of Toronto, and joining me today is Zach Klaassen, Assistant Professor in the division of Urology at the Medical College of Georgia.

This is the citation for this recent publication led by Dr. Emmett and published in European Urology.

Prostate cancer, as most viewers of UroToday know, is common, affecting one in every six to eight men with a further one in five men undergoing prostate biopsy as a result of elevated PSA. In the last five years, there has been transformative evidence with the realization that MRI-based triage for men with elevated PSA can improve diagnostic accuracy and avoid biopsy for a significant proportion of men with elevated PSA. However, the positive predictive value of MRI is relatively low and its negative predictive value does not exceed 90%. Thus, there remains an unmet need to improve decision-making for these patients.

As we know, prostate imaging forms a critical component of prostate cancer care, and recently Novel PET radiotracers predominantly led by PSMA-based radiotracers have revolutionized the care of patients with prostate cancer with increased sensitivity for prostate cancer diagnosis. The ProPSMA trial looked at the use of PSMA as a staging study in men diagnosed with prostate cancer prior to their initial treatment with surgical resection or radiotherapy.  And as we can see here, PSMA-PET/CT had substantially higher accuracy than conventional imaging for detecting the evidence of distant metastasis.

When we looked at this in a variety of ways, whether we looked at any metastatic disease, pelvic nodal disease, distant metastatic disease, and whether we consider the AUC, the specificity or the sensitivity, a PSMA-PET/CT was superior. The question, therefore arises, does the use of PSMA-PET/CT earlier in the diagnostic algorithm have a benefit, and so can we improve the diagnosis of clinically significant prostate cancer by using PSMA-PET/CT?

To assess this, the authors performed a prospective multicenter phase two imaging trial at three academic centers. They included patients who had clinical suspicion of prostate cancer on the basis of an elevated PSA, though, it must have remained below a maximum of 20 nanograms per milliliter or an abnormal DRE.  As a result of these findings, patients had to be planned for prostate biopsy. Patients were excluded if they had a prior prostate cancer diagnosis or had previously undergone a prostate biopsy or MRI.

Patients underwent a multiparametric MRI that was performed and reported locally according to the preferred subspecialist prostate MRI radiologist. Predominantly 3.0 T MRI was used. Although, some patients received 1.5 T scans. These are interpreted according to the PI-RADS version 2 criteria and a PI-RADS lesion score of three to five was considered positive, and these results were used to inform targeted biopsy.

Additionally, patients underwent a PSMA-PET with pelvic only imaging applied in a low dose non-contrast CT protocol utilized. Again, as with MRI, these images were used to inform if targeted biopsy were relevant. The authors segmented the prostate into four segment regions and the SUV max was recorded in each of these quadrants. All PSMA-PET/CTS were read centrally by two reviewers and an SUVmax exceeding four was considered a positive study. The authors then considered combining these imaging modalities, and so they interpreted each of them independently and then considered the overall positive imaging finding to be either a positive PSMA-PET or a PI-RADS four to five lesion.

The authors then undertook a biopsy, and so systematic transperineal prostate biopsy was used. A standard minimum of 18 cores was recommended with additional cores taken in a targeted fashion for abnormalities seen on either the MRI or the PSMA-PET. These were assessed by a uropathologist and interpreted according to ISUP criteria assessing the number of cores as well as the location of cancer, the maximum cancer length, and the proportion of high-grade cancer. Clinically significant prostate cancer is considered to be ISUP grade group two or greater. The authors considered the magnitude of change and the negative predictive value of MRI versus MRI plus PSMA-PET for the diagnosis of clinically significant prostate cancer, and secondarily, they considered the proportion of patients who could avoid biopsy, the diagnostic parameters for MRI and PSA imaging alone and in combination, as well as the proportion of clinically significant prostate cancer that was detected by PSMA, by MRI, or by biopsy. Finally, within the PSMA data set, they assessed the association between SUVmax for the identification of clinically significant prostate cancer.

In terms of statistical models, they performed an interim analysis, which showed that the negative predictive value of MRI alone for clinically significant disease was approximately 70% in this cohort. So, using the relative predictive value method, they recalculated a sample size of 283 patients necessary to detect an increase in this negative predictive value to at least 85% with a combined approach, utilizing an alpha of 0.05 and a beta of 0.2 for a power of 80%. Evaluation of the difference in NPV was assessed using relative predictive values. The authors further assessed test characteristics including sensitivity and specificity and compared these using McNemar's test. They assessed the inter-rater agreement on imaging interpretation using Cohen's kappa and the Kruskal-Wallis test to assess the association between SUVmax and PI-RADS score with the ISUP grade group.

At this point in time, I'm going to hand it over to Zach to walk us through the results.

Zachary Klaassen: Thanks, Chris. So, this is the flow diagram for this trial. You can see the inclusion criteria here on the left with patients having undergone an MRI and clinical suspicion of prostate cancer either with an abnormal DRE or an elevated PSA. The key exclusion criteria for this study were a PSA greater than 20 and previously diagnosed prostate cancer or prostate biopsy. Ultimately, there were 303 patients that were assessed for eligibility, 296 patients enrolled, and 291 patients that underwent a PSMA PET, MRI, and a prostate biopsy.

This is table one patient characteristics. We can see that the median age was 64 years. The median PSA was 5.6. The most common clinical T-stage was T1-C at 68%, the median time from MRI biopsy was 35 days, and the median time from PSMA PET to biopsy was eight days. In terms of the most common PI-RADS score, this was PI-RADs two at 33% and the second most common was PI-RADs four at 31%. The median SUVmax was 5.4 and 73% of PSMA PETs were positive. In terms of positive MRI plus PSMA combined, 81% of cases were positive. The median number of biopsy cores obtained was 33, and the most common ISUP grade group score was ISUP grade group two at 35%.

This table looks at the diagnostic test parameters, looking at PSMA PET, MRI, and PSMA plus MRI, and I've highlighted on the right, the box that shows the combined PSMA plus MRI and this is the sensitivity specificity positive predictive value, negative predictive value as compared to MRI alone. So, looking at the sensitivity for the combined PSMA plus MRI, this was 97%, which was significantly higher than MRI alone with a ratio of 1.16. If we dropped down to the specificity, this was only 40% for the combined approach, which was significantly lower than MRI alone with a ratio of 0.75. Looking at the positive predictive value, this was 67%, which was comparable to MRI alone, and at the bottom, we see the negative predictive value of 91%, which was significantly higher than MRI alone with a ratio of 1.27.

This is a patient classified type of false-negative finding on PSMA plus MRI.  We can see here on the left, this yellow arrow shows an area of difficult interpretation secondary to high central renal excretion, and the red arrow above this masking a low-grade PSMA uptake in the right apex of the prostate. Looking at the MRI, this was reported as a PI-RADs two, however, on PSMA PET targeted biopsy, this patient had a true positive biopsy with an ISUP grade three prostate cancer.

This table looks at the characteristics of patients with clinically significant cancer in total and by MRI and/or PSMA PET positivity. It's a bit of a busy table so I've sort of summarized the key findings over here on the right, and this is basically looking at the false-negative rates.  And if we look at clinically significant prostate cancer, 17% of patients with PI-RADs two on MRI end up being 37% ISUP equal to or greater than three.  And in terms of PSMA, 9.9% with a negative PSMA were associated with 25% of patients having an ISIP equal to or greater than three prostate cancer.

This slide looks at two case presentations.  On the left, we have case number one which shows a PSMA-avid transition zone lesion with the red arrow, as you can see here. This is reported as a PI-RADs two only on MRI and was confirmed to be ISUP grade three on prostate biopsy. Case two is on the right here. This is a PSMA PET with no appreciable activity in the prostate with a PI-RADs five lesion as highlighted by the red arrow and transperineal biopsy confirming an ISUP grade four prostate cancer.

This figure looks at the decision curves for the cohort and for PI-RADs only patients. Among all patients, combined imaging outperforms individual imaging modalities, though its superiority to biopsy all patients, is only for high rationale threshold probabilities.  On the right, which is specific to just PI-RADs two patients, the addition of PSMA PET for PI-RADs two patients demonstrated a benefit for all threshold probabilities greater than 5%.

This is a box and whisker plot between PI-RADs or grade group on histopathology and PSMA SUVmax, and the summary for both of these figures is that PSMA intensity was associated with both PI-RADs and biopsy grade.

So, several discussion points from this PRIMARY trial. This is the first prospective multicenter trial evaluating the potential of PSMA PET for the diagnosis of intraprostatic malignancy in men with MRI PI-RADs two to five lesions. This study found that PSMA plus MRI improved sensitivity as well as negative predictive value for clinically significant prostate cancer compared to MRI alone. It also identified a further proportion of men who could safely avoid biopsy over and above an MRI triaged approach.

PSMA and MRI displayed similar sensitivity and specificity for clinically significant prostate cancer. Although, as I mentioned previously, false-negative rates were high. This included 17% of clinically significant prostate cancer being negative on MRI and 10% of clinically significant prostate cancer being negative on PSMA PET.  However, what's important is that clinically significant prostate cancer was identified by the alternate modality in all but only two patients.

So, to summarize the study and some of the data, the compelling advantage of PSMA is in men with a negative or equivocal MRI.  Because on biopsy, 28% of men with a PI-RADs two lesion and 47% of men with a PI-RADS three lesion had clinically significant prostate cancer and 90% of these were identified by PSMA. So, in conclusion, PSMA plus MRI improved the negative predictive value and sensitivity for clinically significant prostate cancer in an MRI triaged population. However, further randomized studies will determine whether biopsy can safely be omitted in men with a high clinical suspicion of clinically significant prostate cancer but negative combined imaging.

Thank you very much for your attention, and we hope you enjoyed this UroToday Journal Club discussion.