So, we actually have Dr. Mardor here, and she's going to tell us about treatment response assessment maps, increased-decreased sensitivity to tumor treatment effects. So, we are very intimate here now. I would like to thank the organizers for enabling me to share with you some new insights regarding our TRAMs. We heard a lot today about pseudoprogression, so the idea of the TRAMs is to differentiate tumor from treatment effects. These are my disclosures. I want to emphasize that we have licensed this technology to Brain Lab, and it is CE and FDA approved. So, how do we calculate the TRAMs? We scan the patient with standard T1-weighted MRI five minutes after contrast injection, and then we scan the patient again at least one hour after contrast injection, and simply subtract the two images in order to get what we call treatment response assessment maps. Those are the TRAMs. And in the TRAMs, there are two populations. Red is positive signal, blue is negative signal, and I'll explain. So, if you look at the blood vessels in the brain, the signal is maximal immediately after contrast injection, and then it decays with time. And therefore, if we subtract one hour minus five minutes, we get a negative signal showing blue in the maps. And therefore, all the blood vessels in the brain are blue. On the other hand, if you look at the cyst over here, gadolinium accumulates after one hour, and therefore, if we subtract, we get a positive signal showing red. And what we found by comparing the presurgical maps to the first, to the histology of the first 54 resected patients was that blue regions in the map represent morphologically active tumor, while red regions represent everything else, including treatment effects. So, we think the rationale for this stems in the morphology of the blood vessels in these regions. If you look at the blue slash tumor regions, you see tumor vasculature, and what's typical of these vessels is that they're viable. So, they're able to clear the gadolinium one hour after contrast injection, while when you look at regions of treatment effects, the red regions, you have vessels at different stages of vessel necrosis. And therefore, the whole idea of the TRAMs is that if you wait one hour, tumor will clear the gadolinium, treatment effects will accumulate gadolinium, and this is what gives us near complete separation between these two components. So, when we summarize the data of these 54 resected patients, we reach 93% specificity and 100% sensitivity to morphologically active tumor. And in a sub-cohort of patients that were resected in Blanc, where we could assess the tumor burden in histology, we could also compare the tumor burden in histology to the blue burden in the presurgical maps, and you can see the nice correlation. So, in Israel, the maps are being used routinely for clinical decision making with standard medical insurance, and we also recruited over 550 patients to our ongoing clinical studies. And I would like to show you two examples. This is classical pseudoprogression. A patient recruited to our study three weeks after the Stroop protocol with a small enhancing lesion. We scanned her every couple of months, and at four months post chemoradiation, you can see the enhancing lesion started to grow, but we saw in the trans that it was mostly the red component, the treatment effect component that was growing. So, assuming pseudoprogression, does she continue the adjuvant to mesolamide? The same thing at six months when the enhancing volume peaked in volume. Still a little bit of growth of the blue, but mostly a growth of the red. She continued adjuvant to mesolamide, and it shrank nicely later, classical pseudoprogression. The opposite example, progression, is this patient recruited one week after the Stroop protocol with this huge lesion growing into the surgery site, showing 40% blue, 60% red in the trans. We were not sure which one of these two components was growing rapidly and causing the clinical deterioration, so we scanned him again two and a half weeks later, and you can see in the trans that the tumor grew by nearly a factor of two, so it was resected a couple of days later, showing aggressive vascular tumor. In this study, what we wanted to do is test the sensitivity of the trans to the time when we acquired the first T1-weighted MRI. Remember, the nominal time is five minutes. So, here we scanned the patients with a rapid T1 sequence every two minutes, up to 30 minutes, and then the delayed MRI, and what I'm showing you here is the signal intensity of a single pixel within the enhancing lesion. You can see it goes up, and then it goes down with time, and then we calculated the trans signal for that single pixel. Remember, the trans is the late T1 minus the early T1. We normalize it to the signal decay in the blood vessels, and we get this intensity of the same pixel in the trans as a function of the early time point, and you can see that this specific pixel will not show tumor in the standard five-minute acquisition, but if you wait 15 minutes or 20 minutes, then it will become negative, showing blue in the maps, showing as a tumor. Our threshold for tumor is anything lower than minus 0.2, because there's some noise that goes between plus and minus 0.2. So, if we look now at multiple pixels from the same enhancing lesion, what you can see is that all pixels are below the threshold for tumor at 15 minutes, and then that means that at 15 minutes, we will reach a plateau. We will not get more tumor pixels. So, if we look really at the volume of the blue region in the trans as a function of the time when we acquire the first time point, you will see that at five minutes, it's a relatively small tumor. It's about 1.7 cc, but then if we waited 15 minutes, we would reach a three times larger tumor. So, this means that we can enhance the sensitivity to tumor by moving or by taking another trans with another early time point, which is later, and we think that the rationale for that is that the tumor vasculature is able to clear gadolinium also from further regions surrounding the tumor, which causes overestimation of the tumor in the trans, but again, if you want to increase your sensitivity to residual tumor post-surgery, for example, or to small pneumotestases that you may not depict in T1-weighted MRI, you can do that, and we see that for different patients, we have different times to plateau. Some of them are longer. Some of them are shorter. In the 15 patients that we used in this cohort, we reached an average of 25 minutes getting to a plateau, and this is why, really, for the trans, we need to wait at least one hour until we get clearance of the tumor. So to summarize these results, we said the nominal time that we use, which gives us only a little bit of overestimation on the tumor volume when we compare to histology is five minutes post-contrast injection. If you go lower than that, you'll get a smaller tumor, but you can go higher, and on average, if you will wait to 15 minutes, you'll get 60% more tumor than in the five minutes. If you go down, you'll get a factor of two below. You also have to note that since we are so sensitive, the tumor sensitivity to the time is so high when you follow the patient. You take the first time and it is five minutes on one follow-up. If you'll do it at six minutes at the next follow-up, you'll have a 20% error in the volume of the tumor, so you really have to repeat it to be always at five minutes. So to summarize, the trans enable near-complete separation with high resolution between tumor and treatment effects. We use T1-weighted MRIs so we don't have susceptibility artifacts, et cetera. It's model-independent. We simply subtract two images. Easy to interpret. Tumor is blue, treatment effects is red, and the sensitivity to tumor can be enhanced and decreased if we change the time of the acquisition of the early time point, but we should not do it for a specific patient for different follow-ups. Otherwise, we may be mistaken with progressional response. Thank you very much. Thank you very much. Thank you.

TRAMs

treatment effects

T1-weighted MRI

active tumor

timing of acquisition