Matthew Cooperberg: 

Wonderful. Thanks so much. We are now going to go over to our questions. We have a few pre-set questions that our advocates are going to lead with.

Leszek Izdebski: 

Hi, my name is Leszek Izdebski. I'm one of patient advocates here at UCSF. Really, thank you so much for this presentation. I think it's really important for the patients. I would add to some constraints, even in places where there are quite a few MRI machines, patients sometimes wait for MRI many, many weeks. Availability is definitely an issue, so this really solves the problem. But because of that, for patients who decide to move forward with the micro-ultrasound versus MRI, what are the risks? In particular, I'm thinking about a few different cases. One, patient with significant BPH, because I would imagine there's some attenuation of the ultrasound signal, especially for patients who have anterior cancer. Similarly, does this detect cancer in lymph nodes? The last thing that came to my mind was, what about fusion biopsy? Because if you go only with ultrasound, micro-ultrasound, you don't have, I'm assuming, 3D image to guide in the fusion biopsy.

Adam Kinnaird: 

Okay. Those are great points, okay, so thank you for those questions. In terms of the PRI-MUS scoring system, it is only for the peripheral zone. There is a secondary scoring system for the transition zone and the anterior portion of the prostate. But I will say as a qualified expert user, I do find it very difficult to call lesions in the transition zone and, somewhat, in the anterior portion of the prostate. In fact, when we looked at, in Dr. Sonn's trial, when we had six independent investigators look at patients with known prostate cancer and determine where the lesions were blinded to the results, we found that actually we were much worse at calling lesions in the anterior portion of the prostate than we were in the peripheral zone of the prostate. So that is a limitation, but there is a scoring system.

Now, I will say that something that MRI has the luxury of is it's over 20 years old for the prostate, and it's been through five iterations of the PI-RADS scoring system. We are only on the first iteration of the PRI-MUS scoring system, and so certainly we will need to go through several iterations to improve it, and I'm thinking that's where AI may be helpful. The bigger the prostate, the more challenging it is to do a micro-ultrasound biopsy, but the software does allow for a depth of six centimeters of imaging, so most prostates can fit on that.

I think your final question was the fusion platform. The device does come with an MRI micro-ultrasound fusion platform. It is simple to use. I would say that it is not perfect in terms of the fusion as with other technologies that I use with MRI, but it does give you a general idea as to where in the prostate you should be aiming. Now then, that said, the nice thing is with micro-ultrasound, theoretically, you can actually see the lesion in real-time. Whereas with MRI ultrasound imaging, you are completely relying on the fusion technology. With fusion technology, there's something called registration error where you think the lesion is over here, but actually the lesion's over here. When you fuse things together, it gets misaligned, and so you biopsy exclusively from over here, missing the actual lesion. Whereas with MRI micro-ultrasound, you figure out the general area, and then you focus your interpretation on that area and hit that hotspot.

I don't think that that is a major limitation to micro-ultrasound and may pan out to actually being a theoretical benefit to micro-ultrasound. I think you also mentioned the wait time. I'm Canadian. In the Canadian healthcare system, the wait time for MRI is massive. It is about six months to get a prostate MRI in Canada.

Leszek Izdebski:  

Even if you fit that-

Geoffrey Sonn: 

The other thing that he asked about-

Matthew Cooperberg: 

... it can be about two months. Of course, you said $1,000 is a lot, Canadian. In the States, in US dollars, it's several times that, two, three or more thousand is not uncommon for MRI here. In many parts of the civilized world, it's about 250, in the UK and Australia, but we do pay quite a bit for imaging.

Let's go to a related question. Stan, I think this is yours. Just talking a little more about... Not this one, sorry. This one. You're muted, Stan. You need to unmute, Stan.

Stan Rosenfeld: 

So sorry. Thank you so much. Geoff and Adam, my question had to do with comparing PI-RADS to PRI-MUS, and you couldn't have described it better, so we'll move on.

Matthew Cooperberg:  

Well, on this thing, how about reproducibility and accuracy? I mean, Geoff, you've done a lot of work on reproducibility and inter-reader variability in terms of PI-RADS. Any sense of whether the same problems face the PRI-MUS scoring system? If two urologists take a look at the same scan or the same urologist day-to-day, how much consistency is there? What's the learning curve look like? How many do you need to do before you're confident and stable as a urologist?

Geoffrey Sonn:  

I'll defer some of that to Adam then too, because he's certainly much more familiar with the initial development and validation of PRI-MUS. I think it's hard to imagine that that would not be the case. This is just how humans are. Some of us are better at certain things than others, so I think that that definitely will be something we have to overcome.

In the study that I mentioned earlier that Adam was a part of as well, the AI part, we also was a multi-reader study where we're trying to look. In that case, we weren't asking people to designate PRI-MUS. It was just like, "Do you see a lesion or not?" And there was quite a lot of variability, and there wasn't a clear cut. One guy was really, really good, and the rest were not so good. There were a lot of people that are about the same. So I think that the variability is going to be something that will need to be overcome, and maybe AI is going to be that technology that helps us or maybe more training. Like I said, Exact has really good resources that are available, but we're all busy, and so it's tough to take the time to learn a new technology and do as much as we'd like.

Adam Kinnaird: 

The learning curve certainly is real. It has been shown that if you go from 40 cases that are proctored to 90 cases that are proctored, your detection of clinically significant prostate cancer goes up by about 16%. That was previously done. In terms of the scoring system, these Likert scales of 1 to 5 are really bothersome because 3, equivocal, what do you do with that? So it would be nice if it were binary, if it were, "Don't biopsy," or, "Biopsy." That would be much preferred. But whenever you see a scale of 1 to 5, it means that you're not exactly sure what is prostate cancer or not. It just is... You're sort of guessing at it.

As a user, I would say that even with the scoring system that we have right now, starry sky appearance is a PRI-MUS four. In my hands, that's mostly prostate inflammation. I think that's what inflammation looks like on micro-ultrasound. So I think that in the second iteration or PRI-MUS 1.1 or PRI-MUS 2.0, we will move starry sky appearance to be maybe a 3 or a 2. This work has to be done and, in fact, it already is being done as we speak.

Matthew Cooperberg: 

Are there data? I know in the MRI literature, there are data that as an MR radiologist gets more experienced, they call fewer 3s and more 1, 2 versus 4, 5. Is the same true of urologists, as they get more experienced with PRI-MUS, they make fewer equivocal calls?

Adam Kinnaird: 

I can't answer that with actual data, but I can say that, as my user, I'm still calling 3s, and I've done about a thousand.

Matthew Cooperberg: 

Okay, okay.

Adam Kinnaird: 

Yeah.

Matthew Cooperberg: 

Yeah, okay. Amazing.

Geoffrey Sonn: 

I mean, the figure you showed earlier though too, the number of people with PRI-MUS 1 was like nine or something like that, so there's not very many, and that's been my experience too. There's not very many that are just stone cold normal, and it's tough to know what to do with those intermediate cases.

Adam Kinnaird: 

That's a good point.

Matthew Cooperberg: 

The AI story is kind of fascinating. There are literally now over 1,000 FDA-approved radiology AI systems and things out there, a number of which are relevant to prostate MR specifically. One of my favorites is PI-CAI initiative, which is an open source, multinational initiative published a couple years ago, and they had a follow-up study last year that PI-CAI was able to help a human radiologist do a better job in assigning the PI-RAD score. But it turned out when you dropped the human and just let the AI read the scan, accuracy was still better. The robots are coming for us very, very hard and fast in radiology. I'm just curious, Adam, those are some beautiful images from the initial reviews paper. How soon do you think this is going to be primetime use where we're really going to have these sort of heat maps to point at?

Adam Kinnaird: 

 

Yeah, so this is exciting. What you need for AI, and I'm not an AI expert, but I think I'm a millennial, so I understand a little bit, what you need is you need large data sets for machine learning processes to work. What we're doing is we currently have a multi-center initiative where we are trying to create a large imaging repository, micro-ultrasound imaging repository. Certainly, as part of the OPTIMUM trial, and then my trials, MUSIC-Screen and MUSIC-AS, all of these images are being uploaded to a central imaging repository. Hopefully, over the course of the next year or two, we're able to get across the world several thousand of these images stored so that imaging experts at places like Stanford and University of Florida and Queens University can develop these models for us to then roll out.

In fact, I think that the next trial that I'm planning right now is actually called MUSIC-ScreenAI. So it's going to be exactly the trial that I showed you guys, but it's going to be AI assisting the urologist to make the biopsy decision as to whether you need a biopsy or not.

Matthew Cooperberg:  

Fantastic. Let's take a couple of related ones from the Q&A here. There's a question, "After hip surgery, that titanium, the hip implant can throw off an MRI, for example. Are there situations in which micro-ultrasound cannot be done due to anything that's been done to the patient before?" I think the related question, there's a question about before and after aquablation, "So if there have been BPH procedures performed, how well does PRI-MUS hold up?"

Adam Kinnaird: 

Sure. I will say this, one disadvantage to micro-ultrasound versus MRI is that micro-ultrasound requires a transrectal probe, whereas MRI can be done non-invasively. So if you have anal stenosis or fissures or something that is painful, then you would find a micro-ultrasound done under local anesthetic uncomfortable, but it's just like having a biopsy. There are no other contraindications to micro-ultrasound. This just reminded me actually, a question that was asked was, "Does it image the lymph nodes?" It does not image the lymph nodes. It images only the prostate.

Matthew Cooperberg: 

What about the prior BPH procedures? Patients had a TURP or an aquablation or Rezum or whatever?

Adam Kinnaird: 

No problem. Yeah, no problem. It's actually really nice because micro-ultrasound focuses on the peripheral zone, which does not get touched by BPH procedures. So it actually makes things easier.

Matthew Cooperberg: 

Super. All right. Let's go to another one from our advocates. Whose is this one? Nathan, I think. Nathan, I think this is your question.

Nathan Roundy: 

Is that for Nathan?

Matthew Cooperberg:  

I think it's yours, yes.

Nathan Roundy: 

Okay, hi. Yeah. Is there a protocol where a man can undergo all the testing he needs to do before deciding to get a biopsy and have all that information available and then do the micro-ultrasound examination with the idea that, if it's negative, the biopsy would already be decided not to do, or, if it's positive, to do the biopsy right away at the same time?

Adam Kinnaird: 

Yeah, so I'll answer this in two ways. First of all, you can still undergo the exact same pathway as you would with an MRI-guided biopsy where you can get an MRI first and then make your biopsy decision based on the MRI as to whether you undergo a micro-ultrasound or not. And if the MRI does show something, then you can use the micro-ultrasound MRI fusion technology.

The other way of answering that is if you exclude the MRI and you're just looking at micro-ultrasound alone to make your biopsy decision, that is exactly the point of the MUSIC-Screen trial that we are currently conducting. Hopefully, in the next two to three years, we will be able to answer your question, Nathan, as to whether you can forget about MRI and just undergo a micro-ultrasound. But for right now, we don't have any evidence to tell us one way or another. So I would recommend to patients that if you are looking to avoid a prostate biopsy, you should definitely get an MRI first at this point in time.

Nathan Roundy: 

My question is, can you do the decision of making the biopsy and then doing the biopsy at the same session, or do you have to make two appointments?

Adam Kinnaird:  

It can be done in the same session. It can be done in real-time. So if you do it transperineally, you don't have to take any antibiotics before. You can just prepare the patient that they may or may not get a biopsy, but that is still future state because we don't know whether the imaging itself can tell you to avoid a biopsy or not. We need the study, yeah.

Matthew Cooperberg: 

We have a question in the Q&A, I think following up your last comment, Adam, about the peripheral zone. "So since microbiome ultrasound focuses on the peripheral zone, should patients with lesions in the transition zone stick with MRI?"

Adam Kinnaird: 

Well, the answer is that we don't know who has lesions in the transition zone until you've had an MRI. So the answer is yes, that if you have known, like if you're on active surveillance and you have known to have a cancer in the transition zone, then you should probably continue with MRI and MRI-guided biopsy follow-up.

Now, that said, there is a scoring system. I feel like I'm not as good at it yet as maybe some others. I do prefer MRI in the transition zone. I mean, that said, MRI in the transition zone has been an issue as well. It's not as good as MRI in the peripheral zone. The actual architecture of the prostate gland is different, the cells are different, the amount of water that they contain is different, and that's what you're imaging with MRI in the transition zone versus the peripheral zone.

Matthew Cooperberg: 

All right. Next up. I think, Bruce, this one is yours. Okay.

Bruce Zweig: 

Okay. Hi.

Matthew Cooperberg: 

Oh, sorry.

Bruce Zweig:

Can you hear me?

Matthew Cooperberg:  

Go ahead. Yeah, go ahead. Go ahead.

Bruce Zweig: 

It looks from the data that there's some advantage to having both the MRI and the micro-ultrasound. Is that something that we should assume going forward if you can get them both, or, I don't know, how should a patient decide?

Adam Kinnaird: 

Yes. So that's a very intriguing question. In arm 3, the combined arm where you had MRI and a micro-ultrasound, there was a numerically larger amount of clinically significant prostate cancer that was found. It was not statistically different, but it was about 1% higher detection. In targeted biopsy only, it was 6% higher than MRI alone and 2% higher than micro-ultrasound alone. So I think that if you are a private payer and you can afford to have both, then it would make sense to have both. If you are a publicly-funded healthcare system, at this point you don't have enough evidence to say, "Do both."

Bruce Zweig:

Can you sort of tell from the micro-ultrasound that maybe you need an MRI? Does that ever happen if you're MRI only... you're right, sorry, mUS only? Yeah.

Adam Kinnaird:  

Yeah. So I would say that if a patient has a really high PSA... What I use is the PSA density, and so that is your PSA divided by your prostate volume. If your PSA density is too high and you don't see anything on micro-ultrasound, your pretest probability is high enough that you might want to investigate this further, and so you may say, "Okay, I didn't see anything. I took biopsies," if the biopsies are negative, because technically right now you're not supposed to get just a micro-ultrasound image. It should be micro-ultrasound with biopsies, then you would probably get an MRI to see if you missed something anteriorly. You can also just take a couple anterior samples right off the hop to try to diagnose that.

Bruce Zweig: 

Right. When you said you have the micro-ultrasound, then you'll know the volume so you can calculate the density, so that's just sort of a plus there.

Adam Kinnaird: 

Exactly, yeah.

Bruce Zweig: 

And then one other question, on all other uses, do you think this will be used in HIFU going forward. You won't have a registration error. It seems to have some advantages in other focal therapies.

Adam Kinnaird: 

Yeah. A two-part answer for this one. Two things with focal therapy that are important, one is imaging during the procedure, and then the second one is imaging after the procedure and follow-up.

In terms of imaging during the procedure, there have been two publications on this, one using imaging for cryoablation, and the other using imaging during focal laser ablation of the prostate. In both cases, you can image either the iceball or the heated area live on micro-ultrasound, which is a good advantage for laser to actually see the area that's being heated. But under standard ultrasound, you can see the iceball, so it's not an advantage there. But again, these are just case series and it's been done on just several patients. It hasn't been really rigorously analyzed. The other technical limitation with micro-ultrasound is that it only lets you image in the sagittal plane. When I do focal therapy, I like imaging in both the sagittal and the axial plane to see exactly where you are. So that's one technical limitation.

In terms of follow-up imaging, with MRI, there are certain scoring systems like PI-FAB that has been out there that is a scale of 1 to 3 to say whether it looks like there's something suspicious or not. That has not yet been done with micro-ultrasound. I have done micro-ultrasound biopsies on some patients' post-focal therapy. I don't have anything to guide me, but I would just say that it looks like a melted prostate gland, and there's nothing that really sticks out to me. Everything looks hypoechoic, everything looks suspicious, and so I don't know, at least using the human eye, that we will have a scoring system for micro-ultrasound post-focal therapy.

Matthew Cooperberg: 

Thank you. A couple more. Jim, I believe this is yours.

Jim White: 

Yeah, thank you. Adam and Geoff, thank you very much. Your presentation gave me a better understanding of the paper, so I appreciate that. I also apologize. I'll try to be fast here. My internet connection is unstable this morning. I know the paper was about initial diagnosis, but is there potential benefit from micro-ultrasound for men with biochemical recurrence? I'm asking partly because I've got a recurrence. It was in the prostate bed. It was not detectable on my PSMA scan. It was found with a traditional ultrasound. I'm wondering whether there are potential benefits here.

Adam Kinnaird:  

Yeah.

Matthew Cooperberg: 

Maybe comment on post-radiation as well as post-surgery, yeah.

Adam Kinnaird: 

Sure, yeah. Basically, I've biopsied people post-radiation with biochemical recurrence. Obviously, the number one thing to think about with micro-ultrasound, the number one thing to think about with biochemical recurrence is that it could be anywhere in the body, not just where the micro-ultrasound can image. So that's where a PSMA PET scan really is important to check out the lymph nodes, bones, and visceral organs in terms of a local recurrence. That's the only thing that micro-ultrasound could potentially be used for.

I've done biopsies post-brachytherapy, low dose-rate brachytherapy, high dose-rate brachytherapy, and external beam radiotherapy. I would say that the prostate gland just looks globally hypoechoic. It looks globally dark after radiation. I can't say that there are areas that look more suspicious than others on micro-ultrasound, but the whole prostate post-radiation just sort of looks bad on micro-ultrasound. I would not say that I would preferentially use micro-ultrasound over MRI for recurrence. I would still stick with PSMA PET.

In terms of post-radical prostatectomy, if you have a nodule still, like a positive surgical margin that has now become something, theoretically, you may be able to image this, but I have never done that. I don't have any experience. Geoff, do you have any experience in doing that? Of course, you would never have any experience post-radical prostatectomy biochemical recurrence, but maybe somebody referred to you.

Geoffrey Sonn: 

Yeah. Other people may have some failures then too. No, I have not used any micro-ultrasound for biochemical recurrence.

Matthew Cooperberg: 

I think the last one, there's a number of questions both from our advocates and in the Q&A today about availability and why are more centers not doing this, how does the patient find a center which is doing micro-ultrasound, and I'll throw this out to both of you. I can comment on the ECF experience too, but what does the rollout look like, and what are the barriers to adoption for a given urology practice?

Adam Kinnaird: 

Sure. I can go first. There is a Find a Provider on the Exact Imaging website. They can direct you, or you can email Exact Imaging and they can tell you a provider near them. They're a very responsive company. They're a Canadian company, so they're nice. I do know that, prior to the OPTIMUM trial coming out, there were 250 devices around the world. Sales went up by 800% after the OPTIMUM trial, so I think that more people are going to be using this technology and it will become available to you.

Matthew Cooperberg: 

Geoff, any thoughts?

Geoffrey Sonn:  

Yeah. We have a micro-ultrasound system available too. I think some of the barriers, they are significantly more expensive than a conventional ultrasound machine. Far less than buying an MRI scanner, but more expensive than your conventional ultrasound. So I think that that, one, is just the finances of buying it. And then two is just the training. Is someone going to spend the time and effort to learn how to read these scans? When we think about conventional ultrasound, most people still, I think, just use it as a way to find the prostate and are not specifically looking for lesions that are there, so it really is some additional training and it's a change of mindset to think, "Okay, now I'm using this ultrasound to look for something in the prostate and biopsy it, and not just find the gland."

Matthew Cooperberg: 

Yeah, I would just add to that. We've had at ECSF for a while. Our volume is restricted by a unique, local, crazy policy that the probe needs to be sent down to sterile processing instead of just cleaned in the clinic like every other sane institution does, and this is just unique vagaries of our bureaucracy at UC, so we are still ramping up our experience. But barring that sort of issue, the throughput and the cost and the ability to do everything in one visit as opposed to multiple visits, the consistency for center, for academic centers, with a large catchment, like a large referral area, in a way, it's much more appealing to have the patient come down and do a micro-ultrasound with us rather than have a local MRI at a local center where the quality may be quite variable.

My expectation is this, we will continue to see more of this. I think that the cost is going to favor it, and there's a lot of reasons that it's favorable in terms of urologist workflows, easier on the patients to have fewer visits. If everything is equal or at least equal, maybe even a little bit better if you believe the hints in the statistics from the OPTIMUM trial, I think this is the technology that hopefully we're going to be seeing used more, both for diagnosis and I think maybe especially in the active surveillance setting as we try to follow more patients without doing serial biopsies.

Geoff, any last comments? And then I'll give Adam the last word.

Geoffrey Sonn: 

I'll say we have the same problem that you do, it's with micro-ultrasound image and then they want to take the probe for three hours before we get it back. So we have the same workflow limitations that you do.

Matthew Cooperberg: 

Yeah, yeah. Adam, any last thought?

Stan Rosenfeld: 

I have a question.

Matthew Cooperberg: 

Future directions?

Oh, yeah. Sorry, go ahead, Stan.

Stan Rosenfeld: 

Oh, I just wondered, why is that probe cleaning different than a regular ultrasound probe cleaning?

Matthew Cooperberg: 

Well, at UCSF, it's not. They just have this crazy policy about cleaning all the probes. At most institutions, including other University of California campuses, there is a process for high-level wiping and cleaning in the clinic, and some infection control departments just make arbitrary decisions and they do not have adequate clinical oversight over their decisions.

Stan Rosenfeld: 

I know, but Geoff says it's got to go out for three hours in his location also.

Matthew Cooperberg: 

Yeah, go ahead.

Geoffrey Sonn: 

Ours is the same too. All probes have the same thing, and we are going through the same thing that Matt's describing to try to get a wiping system to get it clean because, half the time, these things come back broken too and there's all sorts of problems with that. We'll work on overcoming that obstacle.

Adam Kinnaird: 

Our clinic has its own, what's called MDRD, which is where they clean the probes. For us, we have about a 20-minute turnaround time. We do 12 biopsies a day, and we have two probes. While you're doing one biopsy, one is being cleaned. But still sometimes the patient is waiting, so we're buying another system and more probes. But each probe costs about $70,000, Canadian, so it is pretty expensive.

Matthew Cooperberg:  

That sort of throughput is what you need to get your experience up. If you can only do one or two a day as a urologist, you're not going to get familiar enough with the PRI-MUS system and get your volume experience. These are purely medical center bureaucratic issues that we are working through at a number of centers. I think it's a little easier in some of the urology clinical practices. There are some of these large urology groups around the country that are embracing this technology as well. So I do think we're going to see more and more of it.

Adam, any last thoughts on where the puck is going?

Adam Kinnaird: 

I love it. Love that. I would say that my take-home message for this is that if you need a prostate biopsy and you definitely are going to be getting a prostate biopsy, micro-ultrasound is not worse than MRI. But for right now, while the clinical trials are being completed, I still think that MRI has a significant role in prostate cancer diagnosis. But hopefully in two to three years, I can retract that statement and say it's just micro ultrasound all the way.

Thanks for having me. This has been a lot of fun. The patient advocates, your guys' questions have been great. I feel like your urologists yourself with the level of questions that you're asking.

Matthew Cooperberg: 

Thanks so much for joining us. Thanks, Adam. Thanks, Geoff. Thanks to all of our advocates, and thanks to everybody who has joined us. We look forward to seeing you soon for another installment of the UroToday at Journal Club for Patients. Thanks so much.

Jim White: 

Thank you. Thank you.

Bruce Zweig: 

Bye-bye.

---

Biographies:

Stan Rosenfeld, Prostate Cancer Patient Advocate, University of California San Francisco, San Francisco, CA

Nathan Roundy, Prostate Cancer Patient Advocate, University of California San Francisco, San Francisco, CA

Leszek Izdebski, Prostate Cancer Patient Advocate, University of California San Francisco, San Francisco, CA

Bruce Zweig, Co-Chair of the UCSF Patient Services Committee, University of California San Francisco, San Francisco, CA

Jim White, Prostate Cancer Patient Advocate, University of California San Francisco, San Francisco, CA

Geoffrey Sonn, MD, Cortland T. Hill Memorial Scholar, Associate Professor of Urology, Stanford University, Stanford, CA

Adam Kinnaird, MD, PhD, FRCSC, Surgeon, Scientist, Chair of APCaRI, Frank and Carla Sojonky Chair in Prostate Cancer Research, Division of Urology, Department of Surgery, University of Alberta, Alberta, Canada

Matthew R. Cooperberg, MD, MPH, Professor of Urology; Epidemiology & Biostatistics, Helen Diller Family Chair in Urology, UCSF Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA