Samir Taneja, MD, discusses his research on MRI-guided prostate biopsies and other modalities in development.
Samir Taneja, MD
The use of MRI scans to detect and treat prostate cancer more effectively continues to be a growing area of investigation, according to Samir Taneja, MD.
Researchers are investigating how MRI scans can more precisely target and stage aggressive disease, as well as clinically insignificant cancers. Prior data published in the Journal of Urology showed that MRI-guided biopsies were 3 times as likely to identify cancer compared with traditional biopsies. MRI detected 38% of high-grade cancers that had been missed with a biopsy.
In an interview during the 2018 OncLive® State of the Science SummitTM on Genitourinary Cancers, Taneja, director, Urologic Oncology, Genito-Urologic program leader, NYU Langone’s Perlmutter Cancer Center, discussed his research on MRI-guided prostate biopsies and other modalities in development.Taneja: My research really has focused on the idea of using imaging to change the way we diagnose the disease, risk-assess prostate cancer, select treatments, and even the way we've started to treat it in order to reduce some of the side effects that men experience when they go through conventional therapy. We started thinking we could see the cancer before we biopsy it, very similar to the way a woman gets a mammogram before she gets a biopsy.
With the advent of the multiparametric MRI, which allows us to use different functional sequences to better assess the prostate and determine what’s cancer and what’s not, we're able to localize cancers accurately. NYU Langone is one of the early innovators in the area of prebiopsy MRI of the prostate. We think we can monitor the disease better than we used to be able to with fewer biopsies.
Finally, we have used it to change the way we think about treatment. Whereas previously, we used to treat patients uniformly by removing the whole prostate or radiating the whole prostate. Now, we are trying to transition to a paradigm similar to a lumpectomy where we find the tumor, localize it, and then destroy it using some energy source—heat, freezing, or the like. We have very early data on the effectiveness of that, but it is the imaging that has empowered us to do it.When we look at a prostate MRI, there are some basic functional sequences we look at. The first is what’s called the standard image, or the T2-weighted image. That's really an anatomic image that allows us to visualize the prostate at high resolution. We are looking for dark areas in the prostate that might represent cancer—essentially small, rounded areas or smudges through the cancer that look darker than the normal tissue.
The problem is that the sequence alone is very nonspecific; inflammation, scarring, and atrophy can cause a similar appearance to cancer on that T2-weighted image. We improve the accuracy and we reduce the false positive rate by adding functional sequences—one of which is called the diffusion-weighted image. Diffusion-weighted imaging is basically a method of measuring water movement in the tissue. Normal tissues allow passive movement of water molecules, but cancer, because it tends to be densely packed and higher pressure, restricts the movement of water. Restricted diffusion or restricted movement of water is another indicator of cancer. Now, if we see an area that’s dark on a T2-weighted image and there is a region of restricted diffusion, that increases our suspicion that there could be cancer.
Then, we look at perfusion images. We give gadolinium contrast to the patient and measure the rate at which the contrast accumulates in the area of suspicion and the rate at which it flows out. Cancers tend to accumulate contrast faster than normal tissues and lose contrast, or wash out, faster than normal tissues. By measuring that every 5 seconds and creating a curve, one can add additional information about suspicion.
Once we see that, the radiologist then assigns a score that assesses their suspicion level. We call that the PI-RADS score, which is now a nationally standardized 5-point scale created by the American College of Radiology that allows us to assess risk. A PI-RAD V classification would be almost definitely cancer. A PI-RAD I would be a normal MRI.
Staging is a little different. There, the radiologist is looking for whether or not the cancer extends outside of the prostate. MRI has some limitations in detecting small amounts of disease outside the prostate, but it can be used for that purpose, too. There are 3 kinds of men who come to us for diagnosis. One would be a man who has never had a biopsy but has an elevated prostate-specific antigen (PSA). Another would be a man who has had previous biopsies, but there is still concern because his PSA is going up. The third would be somebody who has been diagnosed with cancer on a standard biopsy, but it’s not clear whether he needs treatment or not.
In each of those applications, MRI can enhance our ability to detect the cancer and determine what the right course of action would be. The middle is where we can make the most impact because these are men who have been subjected to multiple procedures without a clear answer of what’s going on. We have learned that if we do an MRI, we can separate the men who need biopsies from the ones who don't. A normal MRI or low-suspicion MRI would predict a very low likelihood of a high-grade cancer, whereas a high-suspicion MRI allows us to target the right area and make the diagnosis.
A year and a half ago, my collaborator at NYU Langone, Dr Andrew B. Rosenkrantz, chaired a committee of the American Neurologic Association and Society of Abdominal Radiology. I chaired the urology side, and we created a consensus statement that has been adopted by both organizations, which advises urologists that they should consider MRI as standard in that setting as long as good quality MRI is available.
The application in the first scenario, in which a man who has never had a biopsy before just comes in with an elevated PSA, is a little more controversial. As a community, we have not adopted the use of MRI uniformly in those men. However, at NYU Langone, since 2012, nearly every man who comes to us receives an MRI. Our data clearly show that when we use MRI in those men, it allows us to find more high-grade cancers and to predict who is more likely to have low-grade disease, which perhaps we wouldn't want to find because we're not going to treat it anyway.
The example would be: someone comes to us at an older age and it’s not clear that treatment is going to benefit them. If they have a high PSA and the MRI shows no suspicion, we usually advise them it’s probably okay to [watch and wait]. In someone with a marginal PSA elevation but it’s not clear, we can use the MRI to help us. In most men, we still feel they need a biopsy; we do a better biopsy in that situation with the MRI.
In the coming years, there will be data from large global studies that will confirm that is a good paradigm for people to consider so that we are not missing cancers, and we can reduce biopsies.
In the third setting, MRI has a different purpose for the man who is on surveillance. It allows us to identify the men upfront who have high-grade disease so we can treat them as well as the ones who are truly at low risk, so we're able to do fewer biopsies and follow-up. By using MRI as a substitute, we are learning that we can minimize invasive procedures and still be able to carefully monitor when somebody needs treatment. We are still very interested in determining whether or not something in the MRI gives us a sense of which cancers are at risk for spreading and which are not. There can be quantitative metrics of diffusion you can measure to try and help with that. The future of that area will be combining the MRI signal with genetic findings in the tissue we sample, so that if that you can create somewhat of a radiomic signature, you might be able to predict who is at risk from the image.
Where we are making advances that are more tangible to the patient is in the area of molecular imaging—PET. When we fuse PET with MRI or fuse it with CT, we can localize PET signaling to areas that would help us to identify oligometastatic disease. That might help us to identify the extent of metastatic disease at much lower PSA levels. The typical setting in which we would use this is in somebody who has a rising PSA after they've been treated. We've removed or radiated the prostate, but now they come back with a rising PSA and we don't know where that disease is coming from.
There are a number of different PET ligands that have been investigated. Recently, the FDA approved FACBC (anti-1-amino-3-[18F]fluorocyclobutane-1- carboxylic acid; Axumin). We've been using that at NYU. Some of the newer ligands—choline, as an example, or something called prostate-specific membrane antigen (PSMA)—are PET scans which seem to have tremendous specificity for prostate. PSMA in particular is able to detect oligometastatic or widely metastatic disease at very low PSA levels.
However, it’s not FDA approved. We don't have it at NYU, so when we want to get PSMA-PET, we have to seek out protocols that are open for the patient. That is the future. At NYU, what we do have is one of the few PET/MRI facilities in the region. This is a scanner that simultaneously obtains MRI and PET. That increases the specificity because if we see an area in the bone that lights up on PET but also has a restricted diffusion signal on MRI, we can be pretty confident that that’s an area of prostate cancer.
PET/MRI may have wider applications than recurrent disease. We may use it in high-risk patients when we're deciding what treatment to offer them. We may even use it for better staging our patients with metastatic disease if we believe that there are new paradigms for oligometastatic disease.
Albertsen PC, Marks LS. MRI before prostate biopsy—yes or no? J of Urol. 2013; 190(6):1978-1980. doi: 10.1016/j.juro.2013.08.050.