The title of our talk is The Incidents of Post-Traumatic Epilepsy After TBI, a TRACT-TBI Study. I'd like to thank the AANS, UCSF Neurosurgery, and also Drs. Jeffrey Manley, Ramon Diaz-Arrestia, Nancy Temkin, and all the TRACT-TBI investigators for this opportunity. So post-traumatic epilepsy, why is it important to diagnose and treat? So there's 2.8 million TBIs per year in the United States, and a subset of those will go on to develop seizures, and that's what's essentially referred to as PTE or post-traumatic epilepsy. Because of the large number of patients with TBI, PTE actually accounts for an estimated 20% of all symptomatic epilepsy in general. So it's an important issue to treat because of both the TBI, but also from a public health standpoint to treat epilepsy. The long-term effects of PTE on TBI are not fully known. On the left, I have a list of the study there, Bush et al. 2012, looking at a number of different factors, and quality of life, and also patient-reported cognitive factors. But there hasn't been an effect of PTE on GOST, overall outcome. So that's currently unknown. So the main research questions that we wanted to address, number one, how likely is, or what is the risk of developing PTE after TBI? And number two, what is the effect of PTE on TBI patients at 12-month follow-up? So very simple. So in order to do that, we're going to use the TRACT-TBI dataset. So TRACT-TBI stands for Transforming Research and Clinical Knowledge in Traumatic Brain Injury. It's a prospective longitudinal precision medicine study with 3,000 patients, including controls. And it's essentially people with head injuries are followed across the spectrum from concussion to coma for at least 12 months after the initial injury, in a prospective fashion. So there's 2,697 TRACT-TBI patients, and 299 orthopedic control patients, and 300 friend controls. So I'm going to briefly describe what that means. So the orthopedic control patients and friend control patients are followed in exactly the same way as TBI patients, and they're also identified prospectively. But orthopedic control patients have injuries due to orthopedic trauma, and essentially there's no head trauma. And friend control patients have no injuries at all, but they're friends and family of those who have TBI. So orthopedic control patients control for the fact of any trauma, and then friend control patients control for socioeconomic and environmental factors for any results that we find. So the inclusion criteria in this study is all adult patients, no pre-existing epilepsy condition, and all patients needed to have 12-month follow-up. When we applied those, our final data set consisted of 1,885 patients, 1,493 of those had TBI, 182 were orthopedic controls, and 210 were friend controls. And then patients were evaluated 12 months for having PTE or not. So how do we evaluate the patients for PTE? We essentially administered an NINDS epilepsy screening questionnaire, or NINDS ESQ. There are two levels of questioning with that. The first level are four questions that I've listed on the right, and those are supposed to be highly sensitive to picking up symptoms that may signify seizures. So are you having uncontrolled movements? Do you have any episodes of spacing out? So general questions that are designed to pick up if there's a risk factor for seizures. The level two questions are more specific. Specifically, we ask, have you been diagnosed with seizures? And second, did those seizures happen after seven days removed from your injury? If both of those are positive, then we consider that patient a self-reported diagnosis of PTE and a putative PTE patient, and they go in the PTE positive cohort. We'll get to the limitations when we talk about how this is a self-reported diagnosis. So the results, first of all, the overall rate of PTE in this study, 41 patients out of 1,493 patients with TBI screened positive for PTE, giving an overall risk and incidence of 2.7%, which was an agreement with historical PTE cohort from other studies. In addition, of the control patients, zero out of 182 of the orthopedic controls and zero out of 210 of the friend controls went on to develop PTE, which was a good validation of the screening questionnaire. On the left, I have a table of the percentages of people for each of those screening questions. You can see that, for example, everyone who is PTE positive, there was a high percentage of people who were told they actually have epilepsy. That's why this was caught. For the two cohorts, PTE positive and PTE negative, now that we have them, we can essentially compare clinical imaging and outcomes data in each of these cohorts. So the first set of results, we just looked at differences in gender and age. So overall, there was no difference in gender between people with TBI who went on to develop PTE and those who did not. But for age, there was a trend to have a significant finding, but it wasn't significant in the end. But at P-value 0.05, the younger patients had a higher risk factor, which is consistent with people who have found in the past. And one major risk factor was presenting GCS. So the gray plot shows, essentially, that there's a very significant association. But the more severe the head injury, the more likely that you're to develop seizures after that injury or essentially develop PTE. Just to make that very explicit, I plotted that out as the risk of developing PTE as a function of the presenting GCS. And you can really see GCS is on the x-axis. The risk is on the y-axis. That is, the GCS becomes more severe. You essentially have a 10 times greater risk. If you have a GCS less than 8 of developing seizures at 12 months, which is shown in the light gray line, and then 6 months is shown in the dark gray line. So a very significant association between GCS and developing PTE. Finally, for imaging, basically, this slide lists all the different hemorrhage compartments where the blood is. And the main takeaway is that any hemorrhage in the brain is a major risk factor for developing seizures, whether that's subdural blood, epidural blood, epidural only, subdural only, epidural and subdural subarachnoid or contusion. Any blood in the brain is a very, confers a lot of risk about developing seizures and PTE. So now that we have these two main risk factors, presenting GCS and imaging, and also a trend of age, we wanted to look at the outcomes data. So when we looked at GOSC, again, this was very surprising to us. The PTE positive patients had a very significant decrease in GOSC compared to the PTE negative patients. Almost a two-point drop in the GOSC. And that was extremely significant on the univariant. To take a little bit deeper dive into that is that you can see that for the PTE positive patients, the most likely GOSC outcome score is a three. PTE negative, it's an eight. That's a huge difference there. So this is highly significant. Of course, what we thought and the biggest confound here is that the PTE positive patients have a more severe injury and also more blood in the brain. So those two things are also associated with a worse outcome. So we wanted to create a multivariate analysis in order to control for that. And we were actually very surprised when we did that, when we tried to create a multivariate analysis for PTE positive, negative, controlling for GCS, which is injury severity, and head imaging findings, and corrected for multiple comparisons. We also used age as a covariate because it was trending to be significant. That there was still an independent risk of a poor GOSC score based upon the PTE. And it was very significant. So that was surprising. We also looked at two other outcome measures and only two to try to limit the multiple comparisons. We wanted an outcome measure on mood or emotional outcomes. So we did the brief symptom inventory and also a cognitive outcome. We used metric derived from the Rivermead cognitive test. And both of those also on multivariate analysis, the Rivermead cognitive effect more than the BSI. PTE positive group had essentially a worse outcome. So on all metrics tested, PTE was affecting patient outcome. The limitations. And quickly, I wanted to mention limitations. There's two very important ones that I want to highlight. The first is follow-up. So we have 12-month follow-up on a prospective data registry. That seems pretty good. But the reality, this is a paper that's probably the best cited paper about the incidence of PTE after TBI, Aniger et al., New England Journal, 1998. And they followed patients for 30 years and showed that there was a linear increase in risk throughout that 30-year interval. So the fact that we found 41 patients at 12 months, there's probably more than that that are out there who are going to develop seizures. And because we don't have the follow-up, we can't identify them to look at the risk factors they're in. So we need to increase the length of the follow-up. And also diagnosis. The screening questionnaire was great. It allowed us to, at 12 months, screen a lot of people. But the issue is these people are not being seen by an epileptologist. They're not having a scalp EEG done. And we need a more accurate way to do this. Briefly, we have a new study, TRACT-TBI-EPI, to address both of these. Five-year follow-up. Everyone who screens positive, we're going to bring back into the clinic to be seen by an epileptologist. And we're really excited that that may actually help address some of these limitations. So in review, we have shown that, number one, the estimated incidence of PTE is 2.7% and are all comers for all TBI patients in the TRACT-TBI data set. That the risk factors developing PTE or seizures after head injury are TBI severity. And also positive head CT. And then even after accounting for these risk factors, PTE has an independent effect on several metrics of outcome, including the GOSC. And the next step is the TRACT-TBI study to address those two limitations that we talked about. So I want to thank you very much, the AANS, and Dr. Jeff Manley, my mentor here in particular, the entire TRACT-TBI team, Dr. Roman Ramon-Diaz-Arrestia, Dr. David O'Connor, Nancy Temkin, Dan Lowenstein, and everyone at the SFGH basic team. Thank you very much. Thank you.

post-traumatic epilepsy

traumatic brain injury

risk factors

Glasgow Outcome Scale

TRACT-TBI-EPI study