Sergio Calvo, general manager of theranostics at GE HealthCare
Photo courtesy of GE HealthCare
MobiHealthNews’ Emerging Technologies Series spotlights organizations developing, scaling and investing in innovative healthcare technologies. What follows is Part 4 of a seven-part series:
Sergio Calvo, general manager of theranostics at GE HealthCare, sat down with MobiHealthNews to discuss how theranostics uses molecular imaging to personalize treatment and help clinicians predict whether a patient is likely to respond to therapeutics.
MobiHealthNews: Can you tell our readers about theranostics? What's special about this technology?
Sergio Calvo: Theranostics combine targeted therapies with molecular imaging.
Some people say, for simplicity, it's therapy and diagnostics at the same time. It's not quite. It's actually a sequential process, and it's not about diagnosing. It's a therapy approach driven by molecular imaging.
So, you image the same biological target. That's the key, by the way – the same biological target that you're going to treat, you image before and you image the therapy after. That's what enables personalization of the targeted therapy.
MHN: How does information from medical imaging of, say, a tumor, helps determine the most effective therapy?
Calvo: In the last decades, we discovered certain proteins, especially receptors on the cancer membrane, not only but especially those, we discovered some of these proteins that are more present in the cancer cells than in normal cells. One example is the PSMA protein, which is more present in prostate cancer cells than in normal cells. But that's not necessarily true for all cells and all cancers. It's true for a lot, 80% or 90%, but you need to know.
We know the patient has prostate cancer, and we know it's metastasized, and at that very moment, we need to know if PSMA radioligand therapy is appropriate for this patient. The first step is to do a PSMA PET scan. So if you see that image – and yes, the tumors from this patient, they are expressing PSMA – the target is available. And that's when the doctor decides we will go for the radioligand therapy. The likelihood of this patient responding is high. And then they stop the radioligand therapy. They give another molecule. It's in today's practice. Lutetium-177 PSMA, that's the isotope. They infuse that drug in the patient, and that starts to kill the tumor cells. Luckily, the Lutetium-177 emits two things: One is a beta particle that kills the cell, and the other one is a gamma ray. That gamma ray we can take a picture of using another molecular imaging scan, the SPECT scanner.
So we take a picture of the actual therapeutic drug where it is, and the doctor can confirm it's going to the right places. It's not going to dangerous places, but not too much in the kidneys, for example, that could hurt the kidneys. So you can see if the therapy is working or not. And there are many cycles of this therapy. When the doctors are looking at it, they want to see if there are new lesions showing up in the middle of the therapy cycles. If there are new lesions going up and growing, oh, that's a red flag. It means it is not working.
They also measure the patient's PSA level for prostate cancer, and they see if the lesions are shrinking or not between one cycle and another. If the patient is not responding at all, they will stop and try something else. If the patient is responding enough, they will continue. And there is the other extreme. They can do up to six cycles of this type of drug. If the patient responds completely, wipes out the disease, there are cases where they may decide, "OK, let's pause." So that's the personalization aspect of it, using imaging, you get really smart about what to do next.
MHN: Are there certain stages where theranostics doesn't work anymore?
Calvo: Certain patients, certain types of tumors, if the tumor is already very advanced, highly differentiated, maybe the patient will not respond. And there are some unknowns there. Certain patients, they do not express the target. This type of cancer biology turns out not to express this target, so the therapies may not work.
So the personalization is something that exists today to a certain degree. You select patients, you make certain decisions, but this personalization is going to increase over time. It's going to get more sophisticated. Why? For prostate cancer, for example, there will be other options. There are many other drugs being developed, and at some point, you're going to have to decide, well, what's the best one for this case? What's the best dose today? We do not adjust how much activity you inject. It's a standard based on the clinical trial. Everybody anticipates that, at some point, we will be able to modulate that according to the patient. We don't know how to do that. We anticipate, but we don't know how. So there's an opportunity to develop digital and AI solutions for that, and we are investing in it ourselves.
