Read the Full Video Transcript

Michael Hofman: Thank you. Pleasure to be here. Thanks to the organizing committee for inviting me. I'm pleased to talk about the PRIMARY score and where it plays a role. There are no really relevant disclosures for this talk.

As a background, we've been doing PSMA PET in our center for a little over a decade. And over that time, we've seen this transformation in availability, now available at more than 140 PET/CT scanners around Australia. This is actually patient number two we did at our center in 2014, and it was a medical oncologist who had an equivocal MRI. There's a PI-RADS 3 lesion hiding somewhere there, but the PSMA PET showed a focal abnormality that turned out to be a clinically significant prostate cancer.

So it's been evident, really, from the very first days that this is a very good modality for looking at the prostate gland itself. Some time ago, in collaboration with Louise Emmett in Sydney, we did a two center study. The PRIMARY study published in 2021. Before the PRIMARY study, this is Declan Murphy, our head of GU oncology. He was pretty happy. And after the PRIMARY study, you can see he's developed gray hair and he's looking a lot older because suddenly he realized how many maybe significant prostate cancers he might have missed prior to the PSMA PET era.

So in this study, we took 296 men who had had a recent MRI. They'd had a biopsy and they underwent a PSMA PET CT as well. And what we found is that we really improved the diagnostic accuracy with the addition of PSMA PET. So without the PSMA PET, 17% of clinically significant prostate cancers were missed. I think this was alluded to in the prior talk. So we increase both negative predictive value and sensitivity as well.

We also looked at the SUVmax, the intensity of uptake, and there was a clear correlation between that and the ISUP grade group. Here is looking at this in a probabilistic way. So as your SUVmax increases on the x-axis, the likelihood of a clinically significant prostate cancer increases. If we deep dive on this, when our SUVmax was over 9 and you had a PI-RADS 4, 5 lesion, 100% of patients had a clinically significant prostate cancer. So begging the question, maybe you don’t even need a biopsy in this population.

If you had a PI-RADS 2 to 5 lesions and your SUV was over 12, you also had 100% likelihood of clinically significant prostate cancer. But there was this large area of overlap, an equivocal zone. Your SUVs, if they're between 2.5 and 7.5, SUV alone has insignificant, insufficient accuracy. And when we looked at the pattern, we realized that pattern was really important, much like the PI-RADS score from MRI. So we devised this five score based on pattern score. 1, 2 are likely benign. Score 2 is an interesting. You can have diffuse transitional zone activity. This can be significantly intense, but if it’s just diffuse, we call this score 2, it’s not prostate cancer. And the same for central zone activity. However, score 3, which is focal transitional zone or focal peripheral zone activity, score 4, we say likely suspicious. And score 5, uptake over 12 SUV, almost certainly you have clinically significant prostate cancer.

Now, this is another way of looking at that SUV relationship. And this has now been integrated into clinical guidelines. So the PROMISE version 2 recommends to use the PRIMARY score. And we have the spark score coming out as well. This is a recent meta analysis that we were involved with looking at PSMA PET for detection of prostate cancer. And you can see PSMA PET on the left. Smaller number of studies looking at PSMA in combination with MRI. And you can see when we combine it, we generally do better, improving both sensitivity and specificity.

But it's important to be cautious because we do have PSMA negative tumors. Here’s a PI-RADS 5 lesion that’s completely PSMA negative. Despite being a very aggressive prostate cancer, it’s not lighting up. It correlates with the histopathology. There was no significant cell surface expression of PSMA. However, this is rare. We’ve got a publication just out in JNM a few years ago, where we looked at the 300 patients in the ProPSMA study, how many of them were PSMA negative. And this was only 3.3%. This was defined by SUV. If we looked at the PRIMARY score, actually 5 of 10 of these patients had a PRIMARY score 4 or 5 or 3, so they had significant prostate cancer by the PRIMARY score.

In order to change global practice and really get PSMA PET/CT into practice guidelines for diagnosis, as opposed to staging or biochemical recurrence, we think the PRIMARY data is insufficient. So we've gone on to do the PRIMARY2 in Australia. The protocol has been published. You can have a look at this.

And we’re doing this trial because we think there are still men where MRI alone is insufficient, and too many men that undergo an MRI go on to have a biopsy, which maybe is not required if you have a normal PSMA PET scan. So what we’re aiming to do here is take men with equivocal MRI results and randomize them to undergo PSMA PET. We hope to reduce unnecessary prostate biopsies and thereby reduce overdiagnosis.

Now, this is the protocol. It’s a 660-patient study. Patients are eligible if they have a PI-RADS 2 lesion with a significant risk factor, such as a BRCA mutation or a strong family history, or PI-RADS 3 lesions where the urologist said they’re going to go on and do a biopsy as a standard of care because of their clinical risk. Half the population are having a standard of care, which is a template biopsy, and half the population are having a PSMA PET. If it’s positive, we have a targeted biopsy, both of MRI and PSMA-positive lesions. In the negative arm, they go on to surveillance.

We’re hoping to demonstrate non-inferiority. This will mean in both arms we should have the similar incidence of clinically significant prostate cancer. But in the PSMA arm, around one third will not require a biopsy. So they’ll go on to surveillance. We’re also looking at complications since men will have targeted biopsies and some won’t have biopsies at all. And we’re looking at measures of patient-reported outcomes, the stress of having a biopsy, having a diagnosis of a clinically insignificant prostate cancer, and the health economics.

So in summary, PSMA PET is evolving. I think in the next few years we’re going to see it enter the guidelines for diagnosis. We do see invisible tumors on PSMA PET that MRI misses. Potentially, we can skip histopathology in a group of patients who either have a low pretest likelihood and a negative scan, or a PI-RADS 4, 5 lesion and a positive lesion. The intensity matters. But the pattern of uptake, which the PRIMARY score defines, is very important for this equivocal group. And we really await the results of the PRIMARY2 study.

I can tell you, a few days ago, we enrolled randomized patient 560 of 660. So we’re marching towards the finishing line. Thank you very much.