MRI-US Fused Imaging in Prostate Cancer: Targeted Biopsy vs. Standard Biopsy vs. Template Biopsy

MRI-US Fused Imaging in Prostate Cancer: Targeted Biopsy vs. Standard Biopsy vs. Template Biopsy

This is a fantastic meeting. And I think that this morning was a nice setup for what we’ll discuss right now. The pitfalls do exist, but we hope to show that this is not an MRI you get in your office or with a script, but special MRIs that are being performed at certain centers but soon-to-be present everywhere. How soon, I’m not sure. I have a lab for which the government has a credo with a company called Phillips. I’ll show how we built over the past ten years a device that allows us to do MR guided biopsies in an office setting.

The recipients of the money that’s received from the selling of the device are the taxpayers of the U.S. and not me personally because I’m a federal employee. My wife is upset by that, but that’s how it works. I have patents, that’s good for my CV, but I have no financial gain on any of these devices.

And I hear, I am here as a member of a team for which radiology is probably the most important center for it, Peter – – . Giving urologists good information is paramount to success. Brad Woods, our IR doctor and partner in crime, as we like to say, we have a single pathologist who keeps us on our toes.

Yes, so the first part of the talk I was asked to give was where are we today with our standard of care biopsy? And we start with the six-core biopsy, a sextant biopsy, move on to a 12-core extended sextant biopsy, and detection rate improves slightly but not much.

And when you repeat that TRUS biopsy, whether it’s a six-core or a sextant, 12 cores extended sextant, the improvement is really minimal, and even when we use that ugly word, saturation, we find that we don’t really help our patients in any way because we’re not guiding ourselves to the area, and we’re doing this again just randomly through the rectum.

Yet, this is what we learned to be and have been told is the gold standard today, and that has to change. I would argue that as we look forward to template mapping biopsies, that probably gleans out the truth behind where are all the tumors setting because, and this is very good work here done by the host institution here, that they wanted to find out for sure if what we did in regards to a biopsy was actually what the pathologist could see on whole mount.

So they took about 25 men, had over 14 cores taken with that transperitoneal template biopsy system; each of these individuals went on to surgery; they carefully whole-mounted step-sectioned the prostate and reconstructed it with computer models, representations of each tumor volume. And they found 64 tumors in these specimens. And although at times, even a template biopsy can miss some of the tumors, here, 18 out of 64, only one was significant clinically and that’s important.

You heard this morning there was no perfect mousetrap. But what you want to try to avoid is missing the aggressive angry tumors, and this seemed to have done that.

Now, this work out of Tacoma, Washington and West Virginia looked at where I think there is the most utility, which is a failed TRUS biopsy, say, a failed 12-core biopsy, then you move on to a template guided mapping biopsy, about 500 patients. Each man at each session had about 60 cores removed. And what they found overall was a detection rate that approaches 50%. But more important, of that cohort, almost 90% of those tumors that were found were significant clinically. Again, we are finding out those plays that we have to avoid missing.

And so now, is template guided transperitoneal biopsy the new gold standard? Well, one could argue it really does the best for finding all those tumors. But how does it apply to office based urology?

We have to do these in surgical centers or office based practice. It’s really a hospital based procedure. It’s often used with anesthesia, whether it’s general or it’s sedation. And as you can see from the reports, it’s up to 50, 60, maybe 100 cores per session. And so, you have to wonder, can we improve upon this? Well, if so, what about directing biopsies to the tumor? That’s where my work has been for the past ten years.

What we can look at it and say is other solid tumors have this paradigm. I think prostate cancer may be one of the only solid tumors that we randomly biopsy for the hope of hitting the cancer.

But why is that the case? Well, we lacked imaging to find these tumors. We do that for breast cancer, we do that for kidney, and et cetera, but not for prostate.

Yet, many patients are told, I will use imaging in your biopsy. I have a TRUS probe in my hand, and that’s ultrasound, that’s imaging, and I don’t really look for your tumor. How often are we so busy that we just ensure the needle hits the cancer, I’m sorry, the needle hits the prostate but we’re not guiding it towards the cancer? And if we look to spend time in Gray-scale B-mode, we’re often misguided because a lot of those tumors that are hypoechoic lesions are not really cancer.

So as we heard this morning, MRI of the prostate, a high resolution image, this seems to be kind of the standard of care today for current technology. Old technology was really poor. And so, when people say to me, this afternoon, I was, during the break, our MRI doesn’t work, a lot of false positives, false negatives, the imaging quality is poor, the answer is yes, that’s what it was then; it’s different now.

There were cell phones in your pocket with more technology than a desktop that you had five years ago, and this is true for MRI. There’s no difference. This, look, the quality of the imaging has improved; the details improved; and therefore, we have the opportunity now to see tumors in the prostate.
But as we heard before this morning, it’s multiparametric. It’s different sequences. It’s almost four MRIs in one session. And as we have to learn the terminology for SNPs, – – profiling, and the new biomarker CTC, we have to learn the terminology for this. This will go right past us, and radiology will be more than happy to see a tumor, biopsy it themselves, and maybe treat it focally. And I think Dr. Berentz has advocated for that.

So we as urologists need to understand how to speak the language and understand the terminology. I’ll break it up into two categories. The first is signal-to-noise ratio. This year, whether the magnet is at 1.5 or 3T strength, whether we use a coil in the rectum on the body or both, all we’re trying to do is make the image quality clear. That’s the first step.

The second step is the different parameters for which the terms are unknown to us but not difficult to understand. T2: we’re familiar [phonetic] with this, very easily, it’s an anatomic view of the prostate, and the peripheral zone often is bright, central gland could be heterogeneous or dark, but tumors that stand out are usually darker in those areas. For example, this young man who had a negative 12-core TRUS biopsy, and you can ask me, well, this is a pretty big tumor, how could this be a negative biopsy, because we are taught to direct our needles laterally. In the sagittal plane, we often do extreme lateral medial, and it’s a possibility this was missed.

Well, that drive-by [phonetic] signal can be confirmed with the second parameter called diffusion imaging. And if you get one thing from this lecture, it’s this parameter. This really has the most potential to guide this to where the tumors may be hiding in the peripheral zone, which is the kind of bright spot here, and it can even see within the central gland, which is the anterior lesions where we often miss those tumors.

Well, how does this work? It looks at the movement of water through tissue. And where these cells become tight, overgrown, the tumors are forming, that restricted motion is picked up on the MRI. And they continue to improve this sequence, we call that B-mode, to a point now where they can discern tumors and often Gleason grade.

And so, – – radiologist and Pete – – we took our MR guided biopsies and I asked them to give a Gleason grade, not a six, seven, eight, or nine, but a low grade versus high grade. And they were able to discern, based on this one modality, low from high. And this work was also published and corroborated by the group – – by Halle Baron [phonetic], so this is not just single institutions’ results.

And when you have a second lesion, it may be hard to see, but it’s a dark spot in two areas; you have maybe a confirmation that this first area of suspicion maybe now becomes a tumor.

Contrast enhancement: the third parameter. It’s just gadolinium. It’s a blush of contrast. It’s like a tumor in a kidney showing increased cellularity, I mean, increased vascularity, increased vessels having kind of the growth needed for tumor formation. And again, you see a blush of contrast amongst other parts of normally perfused tissue.

And – – spectroscopy which we may not be performing in the future because it’s hard to reproduce between different MR systems at different institutions, but when performed, can be very powerful. In essence, it’s like the old NMR days: it looks at the peak of citrate and choline, that ratio within the tumor. Citrate levels are low, choline is high when you have a tumor or a suspicious tumor. And so, they were called voxels, little areas that can be interrogating and you can look at the peaks and discern where these tumors may be hiding.

Put all together, this multiparametric image is where the excitement is for MR today. Having consistently determined tumor location, levels is a very highly suspicious area. This is a grading system we use at NIH. There are something called PIRADS [phonetic] or other institutional kind of unique grading systems, doesn’t make a difference, as long as you are consistent with the radiologist and you understand what he or she is telling you, if everything is clean, it’s negative, and of course, this is where all the detail is.

So how do we know MRI works? Back in 2004, we ran a protocol similar to what Dr. Crawford’s done. We actually took all the MR specimens that we have for surgery and we whole-mount sectioned them. We made a mould of that individual person’s prostate. We took out the prostate after surgery, put it in the mould. We designed the slits of the mould to have our pathologist slice the gland where the MR actual images were, so we knew we had the stringent criteria for direct correlation. And work prior to this had been close but not perfect.

And this came out in 2011, and in essence, when we have a single modality, we fall short. When you combine all modalities, the PPV approach is one and that’s very helpful, but people ask me, what about the NPV? I want to be able to rule out a tumor, the person’s had a TRUS biopsy, he’s had multiple TRUS biopsies that are negative, can the MRI tell me to leave him alone? And in these cases also, the answer is yes; it performs very well. So these parameters have been present for a couple years now, but they are new and they are specific to your institution.

Now, the MRI is one part of this. It’s easy to just say the MRI sees the cancer, I will biopsy it in the gantry. Well, that’s not urologically based, is it? That would be a radiology procedure. So our feeling was to get this back in the office would be the fuser with ultrasound. It makes it truly office based, the patient’s comfort level is increased, being in the MR gantry for hours for a biopsy is difficult, and lastly, it delegates authority to those who know what they do best. Radiology will read the films, we don’t need to read MRs, but they’ll give us the information that we can decide: is the MRI helpful? Should we do a biopsy, repeat it next year? It’s our decision.

I will speak today about Urine-F. This was the system that we built in my lab over the past few years, but I will say that there are excellent products by different companies even today available. And I will argue that this would be like cystoscopy the next ten years. You can have a good Storz, Olympus, ACMI, Wolf, you pick your scope. They’re all good scopes. They work well; you may have multiple scopes and multiple vendors. This will be the same thing here: over time, you’ll see this technology exist, and their technology will be present.

This was the fusion platform in ’06, hardly office based. We have engineers who sit back here. We miss a lot of targets; spatial accuracy was poor. This was within actually my cysto suite; we had to use that for the procedure. Currently now, we have a platform that is what we call ultrasound machine agnostic. This works on your ultrasound machine, whether you have a BNK or a GE or a Hitachi, whatever you have, it’ll be an add-on feature.

The biopsy guidance is tracked. This little box here is an electromagnetic tracking system; it’s an EM generator. And this technology allows us to look in both ultrasound and MRI. The software is looking at the contours of both organs. The MRI is built in and sent in previously. The ultrasound is real time in your hand. This is the ultrasound probe in your hand, not table fixed. You’re holding it and moving it in real time. And as you go back and forth, the software is picking up little details and then fusing the images together.

Once it’s aligned itself, the images move in concert. So here I am taking the ultrasound probe, moving from base to apex, and the MRI moves in concert with it. Those little blue sides, little dots are the targets or the MR lesions that we’re going to sample. I can move to a sagittal view; here is the bladder up here. I can move out lateral. And at the same time, it travels with me. So again, the handheld probe where the MRI becomes a mirror image of the ultrasound.

Here are the targets being present. This little line is kind of a guided sight line, allows us to do a sample core. This is the individual where the tumor is hard to see, but it’s a small tumor right in this bright peripheral zone, and the ultrasound here would never see it. It’s right there; it’s very small. We can navigate to that spot. We have residents in DC who travel for rotations with us, and these are first year residents doing this. Learning curve is relatively quick: four or five cases, they’re up and running. So not a lot of difficulty in acquiring the tissue. It’s almost like playing video games, which a lot of them are used to, but you can get that core from there.

And this is the person: he had that lesion right there. It was a 3+4=7. Because of his age, he went on to surgery. I would argue now if I had seen that patient today, he’d be part of our focal therapy protocol. We have a protocol since 2011 where we treat individuals just like this, and we use our image guidance system to destroy that lesion. And we’ll see some day whether that’s available or not.

We’ll hear a good lecture series tomorrow about this in more detail, but again, this imaging, the ability to see is giving us more options to treat. Whether that’s good or not, we’ll find out.

People often ask me, I can’t spend a half an hour doing this; it sounds like it’s complicated. It’s a 15-minute procedure from start to finish. And so, the publications that came from this were initially questions. I had many, many rejections. I was told to come back with more larger numbers. But eventually, it got in.

And this is our first 100 patients where we showed when you use MRI on top of, in addition to a standard sextant biopsy, extended sextant 12-core biopsy, we have an improvement in detection rate. Each patient is his internal control. They get a 12-core biopsy by one physician, Brad Wood; he’s blinded to the MR information and does an extended sextant 12-core biopsy. I’ll come in with the MR information and turn on the machine and guide the needle to those areas.

In areas on MRI that were deemed suspicious, highly suspicious for cancer, we can see that the standard random biopsy can miss almost half of those tumors, not surprising. But more importantly, what happens when you go ahead, here it is, what happens when you have a negative biopsy? That’s really where I think the utility is. So you don’t need this over time, but you have a negative TRUS biopsy: can we find those tumors?

And now, focus here for the purpose of time and just say that when the MRI has a level of suspicion that is relatively high, again, we approach 70%. Remember that template biopsy paper where they were finding tumors after failed biopsies 85% of the time? I’m sorry, 40% of the time where almost all of them were going to be high grade cancers. Same thing here: some patients have failed one biopsy, two, three, or four or more, and yet, the lesions can be found.

Why is that? Well, they’re often anterior lesions, tumors that we missed because the routine biopsy is too posterior.

Second thing that I am always asked about is, well, the high grade cancers, can you miss high grade cancers? And so, we heard this morning of a case that Dr. Crawford had where the MRI had missed a high grade cancer. Of course; nothing is perfect.

But that’s a rare exception. The same way that it’s rare for template guided biopsy to miss a tumor of that size, it’s rare for MRI to miss it, too. And that’s the training set: go back to your radiologist, show him what you’ve seen, and say, listen, you missed this. Have him or her learn to read those films better.

In this case, 21 patients had a high grade cancer in this cohort. The MR targeted biopsy never missed those areas. But the standard 12-core biopsy did miss them half the time, and that’s where you want to avoid that. Here, we don’t mind missing those low grade cancers, but not the high grade ones.

And yes, multiple papers, this is not just our work: Correllus, Eigen, Biopsy UCL, this is work in the United States, work overseas, all showing that these systems are here to stay and will become more involved in our practice as the years go on.

I’ll end with just a few cases to highlight this work. One in particular is very common: 68-year-old male, 12-core biopsy, every year, his urologist gets a biopsy; on the seventh year, he gets that horrible saturation biopsy, it wasn’t template guided, it was just a put-you-in-the-hospital, 36 cores through the rectum.
And of course, they came back negative because the tumor was sitting here. This looks very similar to the image we saw this morning. This was again deemed highly suspicious. We were able to get up to that spot, as you heard from Dr. Partin. This is similar to work they’re doing at Hopkins. And yes, sure enough, there was the high grade cancer.

This one seems to be even more troublesome to me. This is a person in DC. He came to me for surveillance. He would like an MRI before surveillance. We said sure, let’s go ahead and do that. But based on the information we have today, he is eligible for surveillance. If he wants surgery because he’s young, he gets nerve sparing; no doubt about it, nerve sparing. And he gets seed implantation, maybe brachy, maybe that’s about it.

Well, the MRI showed this lesion. It was highly suspicious. This person has a very busy schedule. He works very hard. He travels internationally quite often. It was hard to get it back in, but we were able to do a biopsy, and he said, well, you know what? Assume it’s a cancer; let’s just treat it that way and re-discuss my options. We wanted to get the tissue first as part of the research protocol. So this came back Gleason eight.

Well, now the options are changed. What we first had was surveillance, nerve sparing on both sides, or seed implantation; this is all we had.

This gave us more information. Now with that, we can re-discuss it with more detail and say, I feel you’re not T1C, three, six, 5%; you’re eight, 60% of that one core, and you have tumor evasion, they felt confident in the nerve. Surveillance is off the table. If you get surgery, we don’t spare that nerve. Or he gets hormone treatments with his radiation based – – trial. Seeds is off the table, completely different picture based on the one MRI.

Yes, these are – – cases, but these just highlights what you can’t do in certain cases. So here, we took out that nerve and we found tumor in that part of the nerve through the EC. So as you heard from Dr. Partin, at Hopkins, everybody gets MRI as a part of their protocol. I’m not so sure if we’re there yet, but we may need to go to that at some point.

I threw this in sitting in the audience today in response to this morning’s lecture. And what we did for 74 patients was we took the Hopkins criteria for surveillance, they came to me for second opinions, they had to meet that criteria of what we call very low prostate, progressive [phonetic] prostate cancer, very low grade, and what we found is that about 41% of those patients would have been reclassified, and possibly disadvantaged by going on that.

Now, use those terms very carefully because as we know, the MRI the past ten years have shown that surveillance works very well. But we can give confidence not just to the patient but to ourselves. Urologists don’t want to have these individuals progress on their watch. We can rule out or pull out those high grade players.

We have just pulled our last ten years of data. We have about 300 patients in that, and maybe next year, we’ll have that available for this conference or for other conferences. So we’ll see.

So in summary, at centers of excellence that can perform these, what we call multi-parametric MRIs, they performed very well. But to the critics of this morning, they are specific to certain centers: Sloan Kettering as we’ve heard, our institution here in Denver, folks over in Hopkins and other institution. That’s today; that is not tomorrow. The radiologists will get onboard with this and will give this to their physicians in the community, and we’ll see this as part of our care.

I’m very happy to say that at least the devices we have built and others have built allows us to control this as urologists and use it to our advantage for our patients. We keep the biopsies in our possession. And as we go forward, you know what? We’ll look and see how to best apply it.

Will we get to a point where the MRI is a screening tool? Will it become the mammography for breast cancer? I’m not so sure. But we’ll look at that ourselves. Will we get to a point where the MRI is our avenue for focal therapy? I believe the answer is going to be yes. But again, should we be doing that? More to work on: more in the lab, more for research.

Just to show how important this was, this is Francis Collins, NIH Director; this is the engineering team, pathologist – – Marino – – Turpay; this received the NIH Director’s Award last year. Dr. Collins felt that this work was a paradigm shift in how we can look at prostate cancer and how we can go forward with or without genetic [phonetic] profiling, surgical interventions, radiation, focal therapy, but just the fact that we can bring MRI into the picture is a new level of understanding for this disease process.

We have a campus, not as nice as this campus here, and you’re always welcome to come and visit any time. And this is our branch, and Morris Lenahan recruited me many years ago, and without him, I wouldn’t be here. Thank you.

Bittner N, Merrick GS, Butler WM, et al. Incidence and pathological features of prostate cancer detected on transperineal template guided mapping biopsy after negative transrectal ultrasound guided biopsy. J Urol. 2013 Aug;190(2):509-14. http://www.ncbi.nlm.nih.gov/pubmed/ 23416641

Crawford ED, Rove KO, Barqawi AB, et al Clinical-pathologic correlation between transperineal mapping biopsies of the prostate and three-dimensional reconstruction of prostatectomy specimens. Prostate. 2013 May;73(7):778-87. http://www.ncbi.nlm.nih.gov/pubmed/ 23169245

Delongchamps NB, Peyromaure M, Schull A, et al. Prebiopsy magnetic resonance imaging and prostate cancer detection: comparison of random and targeted biopsies.
J Urol. 2013 Feb;189(2):493-9. http://www.ncbi.nlm.nih.gov/pubmed/22982424

Gondo T, Hricak H, Sala E, et al. Multiparametric 3T MRI for the prediction of pathological downgrading after radical prostatectomy in patients with biopsy-proven Gleason score 3?+?4 prostate cancer. Eur Radiol. 2014 Dec;24(12):3161-70.

Hadaschik BA, Kuru TH, Tulea C, et al. A novel stereotactic prostate biopsy system integrating pre-interventional magnetic resonance imaging and live ultrasound fusion. J Urol. 2011 Dec;186(6):2214-20 http://www.ncbi.nlm.nih.gov/pubmed/22014798

Hu Y, Ahmed HU, Taylor Z, et al .MR to ultrasound registration for image-guided prostate interventions. Med Image Anal. 2012 Apr;16(3):687-703. http://www.ncbi.nlm.nih.gov/pubmed/21216180

Loch T, Eppelmann U, Lehmann J, et al. Transrectal ultrasound guided biopsy of the prostate: random sextant versus biopsies of sono-morphologically suspicious lesions. World J Urol. 2004 Nov;22(5):357-60. http://www.ncbi.nlm.nih.gov/pubmed/15672278

Pinto PA, Chung PH, Rastinehad AR, et al. Magnetic resonance imaging/ultrasound fusion guided prostate biopsy improves cancerdetection following transrectal ultrasound biopsy and correlates with multiparametric magnetic resonance imaging. J Urol. 2011 Oct;186(4):1281-5. http://www.ncbi.nlm.nih.gov/pubmed/21849184

Sonn GA, Natarajan S, Margolis DJ, et al. Targeted biopsy in the detection of prostate cancer using an office based magnetic resonance ultrasound fusion device. J Urol. 2013 Jan;189(1):86-91. http://www.ncbi.nlm.nih.gov/pubmed/23158413

Turkbey B, Xu S, Kruecker J, et al. Documenting the location of prostate biopsies with image fusion.
BJU Int. 2011 Jan;107(1):53-7. http://www.ncbi.nlm.nih.gov/pubmed/20590543

Turkbey B, Shah VP, Pang Y, et al Is apparent diffusion coefficient associated with clinical risk scores for prostate cancers that are visible on 3-T MR images? Radiology. 2011 Feb;258(2):488-95. http://www.ncbi.nlm.nih.gov/pubmed/21177390

Ukimura O, Desai MM, Palmer S, et al. 3-Dimensional elastic registration system of prostate biopsy location by real-time 3-dimensionaltransrectal ultrasound guidance with magnetic resonance/transrectal ultrasound image fusion. J Urol. 2012 Mar;187(3):1080-6. http://www.ncbi.nlm.nih.gov/pubmed/22266005

Vourganti S, Rastinehad A, Yerram NK, et al. Multiparametric magnetic resonance imaging and ultrasound fusion biopsy detect prostate cancer in patients with prior negative transrectal ultrasound biopsies. J Urol. 2012 Dec;188(6):2152-7. http://www.ncbi.nlm.nih.gov/pubmed/23083875