One-Button PSMA PET Quantification Predicts Survival in Metastatic Prostate Cancer - Louise Emmett
June 20, 2025
Oliver Sartor hosts Louise Emmett to discuss her "one button push" automated PSMA PET quantification system. Dr. Emmett explains that while quantitative PSMA PET analysis provides valuable prognostic information, it typically requires 20 minutes of expert manual input, limiting clinical adoption. Working with MIM Software, they developed automated quantification that eliminates benign uptake without human intervention. Testing the system using a medical student with no anatomy knowledge versus expert analysis showed identical Kaplan-Meier survival curves, with the automated system actually performing slightly better for predicting overall survival. While imperfect, the system successfully predicts patient outcomes. Dr. Emmett's goal is making quantitative parameters like total tumor volume, SUV mean, disease sites, and intensity patterns easily accessible to busy clinicians through automated reporting.
Biographies:
Louise Emmett, MD, MBChB, FRACP, FAANMS, Professor, Director of Theranostics and Nuclear Medicine, St. Vincent’s Hospital Sydney, University of New South Wales, Sydney, Australia
Oliver Sartor, MD, Medical Oncologist, Director, Transformational Prostate Cancer Research Center, East Jefferson General Hospital Cancer Center, Tulane University Cancer Center, New Orleans, LA
Biographies:
Louise Emmett, MD, MBChB, FRACP, FAANMS, Professor, Director of Theranostics and Nuclear Medicine, St. Vincent’s Hospital Sydney, University of New South Wales, Sydney, Australia
Oliver Sartor, MD, Medical Oncologist, Director, Transformational Prostate Cancer Research Center, East Jefferson General Hospital Cancer Center, Tulane University Cancer Center, New Orleans, LA
Related Content:
ASCO 2025: ‘One Button Push’ Fully Automated PSMA PET Quantification: Correlation with Progression Free and Overall Survival in Patients Undergoing 177Lu PSMA Therapy for mCRPC
ASCO GU 2025: Overall Survival and Quality of Life with 177Lu-PSMA-617 plus Enzalutamide Versus Enzalutamide Alone in Poor-Risk mCRPC in ENZA-P (ANZUP 1901)
PSMA PET Scans Predict Enzalutamide Efficacy in Metastatic Prostate Cancer - Louise Emmett
Phase II ENZA-p Trial Explores Effect of Lutetium PSMA With Enzalutamide - Louise Emmett
ASCO 2025: ‘One Button Push’ Fully Automated PSMA PET Quantification: Correlation with Progression Free and Overall Survival in Patients Undergoing 177Lu PSMA Therapy for mCRPC
ASCO GU 2025: Overall Survival and Quality of Life with 177Lu-PSMA-617 plus Enzalutamide Versus Enzalutamide Alone in Poor-Risk mCRPC in ENZA-P (ANZUP 1901)
PSMA PET Scans Predict Enzalutamide Efficacy in Metastatic Prostate Cancer - Louise Emmett
Phase II ENZA-p Trial Explores Effect of Lutetium PSMA With Enzalutamide - Louise Emmett
Read the Full Video Transcript
Oliver Sartor: Hi, I'm Dr. Oliver Sartor. I'm here at ASCO 2025 with UroToday, talking about interesting topics of metastatic CRPC and have the honor to be able to welcome Professor Louise Emmett from Sydney, Australia. Welcome, Louise.
Louise Emmett: Thank you very much for having me.
Oliver Sartor: So when I was reading through the abstracts, I came across an odd title. It was something about one button push. And I said, oh my goodness, gracious. What does that mean? And why is she presenting this at ASCO? So Louise, what's this about pressing the button? What have you done?
Louise Emmett: I like simple things, right? One button push, nothing complex. So we've published so much recently on the value of quantitation in PSMA PET. But it's just for the elite. If you don't have 20 minutes to spend, deciding whether something is or is not tumor on these quantitative programs, you don't get an output.
And quite often, it can be very much impacted by the person who's doing the yes, no, yes, no, for the tumor deposits. And you can get wrong results as well. So we've been working for quite a long time now to try and push companies to automate this. Not AI, really, but just intelligent coding of the way that we do total body quantitation. And so there's a company called MIM Software, who we've been working with within the ENZA-p trial to try and improve this.
And we kind of got to a point where they can now define all the benign uptake and eliminate it. So the question is, is it good enough to actually be used clinically and to predict outcomes, predict overall survival and progression-free survival? So with one button push, what we did was, we took a prospective data set that we had at St. Vincent's in patients who had received lutetium PSMA 617 and I&T. We knew that outcome data. And we took their screening PSMA PETs.
And I actually gave it to a medical student who had done no anatomy. So I just got him to push them all through and then write down the numbers. So no input at all. And then we got one of our experts to do an automated quantitation. And we wrote down those numbers as well. And then we obviously put it through. And we got some Kaplan-Meier curves, which look absolutely identical.
In fact, there's slightly better prediction of overall survival for the totally automated, no human touch, total body quantitation for tumor volume and for SUV mean, than for the one that was adjusted. I think it's important. We can get lots of information on how great total tumor volume is on PSMA PET and on FDG PET.
But radiologists don't have time to spend 10 or 15 minutes creating that. It has to flick up on their screen for them to put it in the report. And so I'm really excited about one button push, because I know that the guys that I work with in my department will write those numbers down in their reports. And I think it's pretty close to good enough.
Oliver Sartor: Well, your critical work has shown us the value of things like SUV mean, total tumor volume. We have clinical implications for these findings, not only for lutetium treated patients, but also for hormonally-treated patients.
And now, you can press a button and get that output in a relatively reliable way. That is wonderful! Now, what about the rest of the world? You can do it at St. Vincent's. But I want to do it in New Orleans. Can I do it in New Orleans?
Louise Emmett: So I think that one data set showing that it's predictive of overall survival and progression-free survival is not enough. We need to test this much more broadly. I'd love to take all your data sets from Novartis. I'd like to get a medical student to shove them through and see how it performs.
And then we probably need to adjust a little bit. But it's so close. Nuc med people like to be perfect when they report. Every single image, it has to be perfect. It can't have any mistakes. But the fact that it's predicting survival means that it's very close to being usable.
Oliver Sartor: That is useful. When you predict survival, you predict outcomes. You do a test today. And it tells you what happens tomorrow.
Louise Emmett: Yep.
Oliver Sartor: That's useful. Well, let me ask a, I don't want to say, technical question, but a question that I really do have. So I looked at a number of PET scans myself. And men do things like dribble the urine and get stuff on their leg, OK? Now, how do you know in the one button push that that dribbled urine is urine and not tumor metastatic deposits. And you can also see extravasation at the injection site.
Louise Emmett: Yes.
Oliver Sartor: So how do you deal with those kind of weird things?
Louise Emmett: So the answer is, at the moment, not perfectly. So it predicted overall survival. But there were some patients who had incontinence pads that were still full of urine. And that was included in the volume. So it's not perfect. But I guess it's predictive, not perfect, is the answer. But what they're doing at the moment is, they're currently coding in for me, skin.
So they're going to exclude continence pads. They're going to exclude catheter bags. And it's about identifying these small problems that, of course, happen all the time.
Oliver Sartor: Of course.
Louise Emmett: And coding them in such a way that they're excluded.
Oliver Sartor: I like it. The truth is that you've been able to demonstrate, as stated, that measurements on the PET have implications for our patients, for prognostic and predictive value. And now you're telling me that it's getting easier to be able to measure those parameters in a pragmatic way that will actually come into the clinic and influence care. I think that's very important.
Louise Emmett: I want you as an oncologist to have available to you the total tumor volume, the SUV mean, the sites of disease, whether it's visceral, non-visceral, bone or lymph node, high intensity or low intensity, and heterogeneous. We can put all of that together. And that can be quantifiable. That helps the reporter. It doesn't take away from a clinical report. But it's such a huge adjunct to how we can use those scans.
Oliver Sartor: I'm going to do a little bit of corrective action. Yes, good for the reporter. But it's really good for the clinician.
Louise Emmett: Yeah.
Oliver Sartor: I'm the one that has to make those decisions about what to do with the patient. And this is a tool, a new tool that helps me be a better doctor.
Louise Emmett: And the guys I work with, they're fantastic. But they're busy. They're so busy. So we need to make it as simple as possible for people to put that into the report. And I think we're on the cusp of that. I'm hoping that all this data that we've published in the last few years on the value of PET quantitation will become available to everybody in the next two years.
Oliver Sartor: I think that's fantastic. So Professor Louise Emmett, always a pleasure to be able to have you here on UroToday. And thank you for explaining one button push.
Louise Emmett: Thank you.
Oliver Sartor: Hi, I'm Dr. Oliver Sartor. I'm here at ASCO 2025 with UroToday, talking about interesting topics of metastatic CRPC and have the honor to be able to welcome Professor Louise Emmett from Sydney, Australia. Welcome, Louise.
Louise Emmett: Thank you very much for having me.
Oliver Sartor: So when I was reading through the abstracts, I came across an odd title. It was something about one button push. And I said, oh my goodness, gracious. What does that mean? And why is she presenting this at ASCO? So Louise, what's this about pressing the button? What have you done?
Louise Emmett: I like simple things, right? One button push, nothing complex. So we've published so much recently on the value of quantitation in PSMA PET. But it's just for the elite. If you don't have 20 minutes to spend, deciding whether something is or is not tumor on these quantitative programs, you don't get an output.
And quite often, it can be very much impacted by the person who's doing the yes, no, yes, no, for the tumor deposits. And you can get wrong results as well. So we've been working for quite a long time now to try and push companies to automate this. Not AI, really, but just intelligent coding of the way that we do total body quantitation. And so there's a company called MIM Software, who we've been working with within the ENZA-p trial to try and improve this.
And we kind of got to a point where they can now define all the benign uptake and eliminate it. So the question is, is it good enough to actually be used clinically and to predict outcomes, predict overall survival and progression-free survival? So with one button push, what we did was, we took a prospective data set that we had at St. Vincent's in patients who had received lutetium PSMA 617 and I&T. We knew that outcome data. And we took their screening PSMA PETs.
And I actually gave it to a medical student who had done no anatomy. So I just got him to push them all through and then write down the numbers. So no input at all. And then we got one of our experts to do an automated quantitation. And we wrote down those numbers as well. And then we obviously put it through. And we got some Kaplan-Meier curves, which look absolutely identical.
In fact, there's slightly better prediction of overall survival for the totally automated, no human touch, total body quantitation for tumor volume and for SUV mean, than for the one that was adjusted. I think it's important. We can get lots of information on how great total tumor volume is on PSMA PET and on FDG PET.
But radiologists don't have time to spend 10 or 15 minutes creating that. It has to flick up on their screen for them to put it in the report. And so I'm really excited about one button push, because I know that the guys that I work with in my department will write those numbers down in their reports. And I think it's pretty close to good enough.
Oliver Sartor: Well, your critical work has shown us the value of things like SUV mean, total tumor volume. We have clinical implications for these findings, not only for lutetium treated patients, but also for hormonally-treated patients.
And now, you can press a button and get that output in a relatively reliable way. That is wonderful! Now, what about the rest of the world? You can do it at St. Vincent's. But I want to do it in New Orleans. Can I do it in New Orleans?
Louise Emmett: So I think that one data set showing that it's predictive of overall survival and progression-free survival is not enough. We need to test this much more broadly. I'd love to take all your data sets from Novartis. I'd like to get a medical student to shove them through and see how it performs.
And then we probably need to adjust a little bit. But it's so close. Nuc med people like to be perfect when they report. Every single image, it has to be perfect. It can't have any mistakes. But the fact that it's predicting survival means that it's very close to being usable.
Oliver Sartor: That is useful. When you predict survival, you predict outcomes. You do a test today. And it tells you what happens tomorrow.
Louise Emmett: Yep.
Oliver Sartor: That's useful. Well, let me ask a, I don't want to say, technical question, but a question that I really do have. So I looked at a number of PET scans myself. And men do things like dribble the urine and get stuff on their leg, OK? Now, how do you know in the one button push that that dribbled urine is urine and not tumor metastatic deposits. And you can also see extravasation at the injection site.
Louise Emmett: Yes.
Oliver Sartor: So how do you deal with those kind of weird things?
Louise Emmett: So the answer is, at the moment, not perfectly. So it predicted overall survival. But there were some patients who had incontinence pads that were still full of urine. And that was included in the volume. So it's not perfect. But I guess it's predictive, not perfect, is the answer. But what they're doing at the moment is, they're currently coding in for me, skin.
So they're going to exclude continence pads. They're going to exclude catheter bags. And it's about identifying these small problems that, of course, happen all the time.
Oliver Sartor: Of course.
Louise Emmett: And coding them in such a way that they're excluded.
Oliver Sartor: I like it. The truth is that you've been able to demonstrate, as stated, that measurements on the PET have implications for our patients, for prognostic and predictive value. And now you're telling me that it's getting easier to be able to measure those parameters in a pragmatic way that will actually come into the clinic and influence care. I think that's very important.
Louise Emmett: I want you as an oncologist to have available to you the total tumor volume, the SUV mean, the sites of disease, whether it's visceral, non-visceral, bone or lymph node, high intensity or low intensity, and heterogeneous. We can put all of that together. And that can be quantifiable. That helps the reporter. It doesn't take away from a clinical report. But it's such a huge adjunct to how we can use those scans.
Oliver Sartor: I'm going to do a little bit of corrective action. Yes, good for the reporter. But it's really good for the clinician.
Louise Emmett: Yeah.
Oliver Sartor: I'm the one that has to make those decisions about what to do with the patient. And this is a tool, a new tool that helps me be a better doctor.
Louise Emmett: And the guys I work with, they're fantastic. But they're busy. They're so busy. So we need to make it as simple as possible for people to put that into the report. And I think we're on the cusp of that. I'm hoping that all this data that we've published in the last few years on the value of PET quantitation will become available to everybody in the next two years.
Oliver Sartor: I think that's fantastic. So Professor Louise Emmett, always a pleasure to be able to have you here on UroToday. And thank you for explaining one button push.
Louise Emmett: Thank you.