The latest in prostate cancer imaging

Professor Anwar Padhani – Mount Vernon Cancer Institute, UK

Professor Padhani, Paul Strickland, Mount Vernon, guru of imaging in the oncology field in the UK, if not further afield, thank you for giving us a couple of minutes.

A pleasure.

I know you’re busy at this prostate cancer meeting. I’m convinced of the speed that imaging is moving at but a lot of other people don’t even know whether things are static or mobile. They’re mobile, aren’t they?

Yes. Imaging is changing at such a fast pace, even in the last five years we’ve seen a tremendous acceleration in imaging. And there are certain trends that are coming through: firstly there’s the quantitative aspect of imaging - we are moving as a community from a qualitative assessment of images to a more quantitative approach and that’s a very strong theme. The second thing that’s coming through is the emphasis on functional imaging and that is in the area of PET and MRI scans. We now see the emergence of machines that will do both PET and MRI at the same time, so PET MRI scans are now commercially available for whole body applications. So we see the technology moving, we see the quantification and this whole process is accelerated by improvements in biology. So as we understand cancer biology more, we can then begin to identify targets, receptors, chemicals that we can then label and then image. So the advantage of imaging in this particular sense is that you can image heterogeneity, so heterogeneity spatially in different parts of the body; heterogeneity within a tumour and heterogeneity across time. So we see a number of developments occurring both in the hardware, the biology and the software as well to be able to image these different processes at baseline and in response to therapies. And this is a very exciting time.

And of course cancer is the heterogeneous disease of all time.

Exactly.

We’re talking about prostate cancer at this meeting and where it’s going and where it’s been.

Yes.

Where it’s been in imaging has been bone scanning, especially in advanced prostate cancer, and you came out very clearly and said, ‘Hi guys, that’s for the birds. We’ve got to move on.’

Correct. So the interesting thing is that the multifunctional approach in primary prostate cancer has been around for a long time. The metastatic setting has been largely neglected and radiologists aren’t really that interested, until today, in the metastatic setting. And part of that issue is that there hasn’t been the impetus from clinicians to say ‘We need better ways of monitoring the AR receptor, the spatial distribution. We need a better bone scan,’ we just haven’t had that. But now that we’ve got new drugs that work…

We’ve got drugs that work.

Particularly in the metastatic setting, previously you had PSA and PSA seemed to work fine so why bother? But now when we’re moving into this castration resistant era, then imaging is becoming important and so you see that radiologists are now responding to that pressure and I think that you will see that that will change over time.

In your excellent talk, differentiating between bony disease, lytic and sclerotic, and soft tissue disease, horses for courses, so what’s the recipe for bony disease first of all, lytic and sclerotic?

Well, this is the important thing.

Because we’ve got rid of bone scans.

Up until now we’ve stuck to bone scans because we were looking at bone and it’s very, very clear to those of us who are in this game that actually it’s the cells in between the bony trabeculae, that’s where the importance is. And in terms of looking at the lytic disease or the fleshy tumour within the trabeculae, then clearly we need to either look at the cells or the cell density and that’s why I think that diffusion imaging is exploding. In my institution we do lots and lots of diffusion imaging and that is now beginning to catch on in the general community.

And that’s called DWI and it’s done with an MR image?

Correct, yes. It’s a fairly standard technique, instead of doing one body part with diffusion imaging we just do four body parts and then stitch them together in software which is fairly simple.

And quicker?

It’s quick, it’s about twenty minutes to do the whole body. So it’s very fast so you can do lots of patients. And it’s very reproducible, that’s the nice thing. The reproducibility is probably, in terms of the measurements, it’s probably around 10-15%. So, in other words, if you repeat the same experiment then you can spot the change of 10-15%. I guess the important thing is how much of a change do you need for it to be clinically important and that’s a different question.

So in terms of looking at the cells and if you’re interested in cell kill, yes, then diffusion is the way to go. If you looking for something else, a pharmacodynamics biomarker, which is something very different, it’s about drug development whether we modulate a target, that sort of thing, then you need something that’s very much more sensitive. And for that the PET scan, obviously, does best, and for something that targets the androgen receptor, the FDHT is a nice example of that.

And what is that? Fluorine?

You basically use the fluorine, you tag it on to dihydrotestosterone and essentially what you look at is binding of that testosterone onto the androgen receptors, it’s as simple as that. So we can see the spatial distribution of that.

You can see soft tissues?

In soft tissue as well as in lytic bony disease. Now the important thing here is that the lytic disease is the one that’s dangerous, that’s what kills patients and that’s what you want to attack with chemotherapy or any other treatment because that’s what’s going to kill patients.

And the sclerotic bone metastases are interesting, they’re there, they’re not measurable, we’ve never even tried to measure them and, as you said in your talk, they don’t kill people because it’s pretty slow.

That’s right, and if you go over a certain threshold, and we don’t know what this threshold is, they are classified as inactive. So the question is, what is that threshold? Some people set it at 600 units and some people set it at higher levels but we know that if you have dense bone, some patients are symptomatic from really dense bone and they are more likely to fracture etc, but in general they’re considered inactive.

And it’s the lytic disease that are going to fracture and they’re going to cause the spinal compression etc.

That’s right, absolutely.

So that’s five years of phenomenal progress which has gone parallel with the development of new second or third generation androgen treatments etc, and we were talking about abiraterone today.

I don’t think it’s five years of development, I think we’re just catching the up-slope, I think we’re just catching the up-slope.

So the next five years, come on.

So the next five years, I don’t think there’s any doubt about the fact that in mainstream we will start to monitor castration resistant prostate cancer primarily with diffusion imaging, that’s my prediction. And I see that in my practice already. So I think that that’s what will happen, clearly a number of things need to happen for this to happen but, as a technique, that, I think, will develop because it’s quantitative, you can look at it visually, you can actually measure cell density indirectly with that. As far as the PET techniques are concerned, clearly the FDHT will be important for the drugs that target the androgen receptor, the colene is much more non-specific and I think will directly compete against diffusion imaging but diffusion imaging is so much more cheaper. So in general practice that’s what will get used and I suspect that colene will be used for problem-solving, detecting small volumes.

And to accompany new drug development?

The new drug development, I think it will depend on what targets are being… it depends on what is the question. The question is about modulating targets, then we will have to develop a compound in that pathway that is being modulated and that has to be PET, I don’t think there’s any doubt about that. But if it’s about the effects of those drugs in terms of response, then you’re really interested in cell kill. If you’re interested in cell kill then you need to have something that looks at cell killing, and that could be a PET tracer like annexin or it could be something much more cheaper like diffusion imaging.

OK. So clearly we’re multi-modality imaging in advanced disease?

Absolutely, yes.

We’re already targeting da Vinci scopes and CyberKnives with PET and MR in the pelvis for earlier disease, what about before that? What about pre-malignant disease? We’re going to talk about inflammation leading to prostate cancer, have you got something up your sleeve for picking up early inflammatory changes which might be pre-malignant? Screening tool for 2020?

Well I’m not sure about that. I think one can look at inflammation, and there are PET agents that will look at information, but MRI does not have the sensitivity, I don’t think, at this stage to look at the early inflammatory change or the stromal responses. That may precede the invasive phenotype but MRI is actually quite good at looking at cellularity. Sometimes you get the androgenic switch that comes on, so you can look at the androgenic switch that comes on, you can look at diffusion imaging to look at increasing cell packing, to give you some idea of where there is a transition going on between something like PIN and the invasive phenotype. But I don’t think that we can be very, very specific in that area yet.

Early days, but no question that molecular imaging is moving just as fast as molecular micro-pathology or experimental pathology and the field of biomarkers is exploding.

That’s absolutely right, yes.

But those biomarkers are also contained within your clever machines.

Yes.

Anwar, thank you very much indeed, I really appreciate you getting us up to date. Thank you.

My pleasure, thank you.