Dr. Sperling
Welcome to Meet the Authors, a podcast brought to you by a collaboration of the Society for Clinical Neuropsychology and the journal Neuropsychology. My name is Dr. Scott Sperling, and I’m grateful to be your host. In this podcast, student leaders in Neuropsychology will discuss prominent recently published studies with the authors who undertook the research, thereby allowing for a behind the scenes look into the development, implementation, analysis and future implications of cutting edge neuropsychology research.

Today, our student leader, Mr. Humza Khan, will be discussing a very exciting paper entitled The Development and Application of the International Classification of Cognitive Disorders in Epilepsy: Initial results from a Multicentre study of adults with temporal lobe epilepsy, with two of the paper’s authors, doctors Carrie McDonald and Anny Reyes. I’d like to now introduce our student leader and our esteemed guests.

Mr. Humza Khan received his bachelors in Neuroscience at the University of Illinois at Chicago and is currently a fifth year clinical psychology doctoral student at the Illinois Institute of Technology. His clinical and research interests include health and numeracy and literacy and its impact on neuropsychological functioning within marginalized communities and advocacy in Pakistan and Guatemala. Dr. Carrie McDonald is a professor of radiation medicine and applied Sciences and psychiatry at UC San Diego and a board certified clinical neuropsychologist.

She’s a principal investigator at the UCSD Center for Multi-Modal Imaging and Genetics and director of the Lab of Cognitive Imaging. The primary research interests are in using advanced multimodal imaging in combination with patient specific risk and resilience factors to predict post-operative and post radiation cognitive outcomes in adults with epilepsy and brain tumors. Dr. McDonald is also the director of the Neuropsychology Assessment Clinic at the UCSD Moores Cancer Center.

She’s PI on three multicenter RO-1s from the NINDS and a co-investigator on multiple studies that use quantitative imaging to study cognitive networks in patients with neurological disorders. And Dr. Anny Reyes is a neuropsychology postdoctoral fellow in the Department of Radiation Medicine and Applied Sciences at UCSD. She completed her doctoral degree in clinical psychology, specializing in neuropsychology at the San Diego State University, UC San Diego Joint Doctoral program, and her Neuropsychology internship at Emory School of Medicine.

Dr. Reyes’s research includes the examination of cognitive diagnostic frameworks in epilepsy and brain tumors, validating the cultural applicability of neuropsychological tests and diagnostic approaches in understanding and reducing health disparities across neurological disorders. She’s worked tirelessly to increase diversity within neuropsychology, promote student mentorship and improve recruitment and retention of underrepresented minorities in neuropsychology and the neurosciences. So I’d like to now welcome Humza and welcome Drs McDonald and Reyes for being with us here today.

Dr. McDonald
Yeah, thank you.

Dr. Sperling
I’ll now turn over our discussion to Mr. Humza Khan.

Mr. Khan
Thank you very much. All righty. So in their paper, the authors describe the development and application of a consensus based, empirically driven approach, namely the international classification of cognitive disorders and epilepsy to classifying cognitive phenotypes. The authors assessed the ability of this classification system to produce definable and stable cognitive phenotypes in a multicenter study of over 2400 patients with temporal lobe epilepsy.

For those who may not have yet read your paper, can you briefly summarize the main findings?

Dr. McDonald
Sure. So thank you so much for inviting us here today to speak about this. This is a paper that we had envisioned for a long time as we were collaborating across many centers and discussing how our patients presented with many different neuropsychological profiles. We finally decided to try to combine datasets to see whether or not the same phenotypes or the same cognitive profiles would emerge across the various centers.

We did this by trying to take a very uniform and diagnostic approach that was harmonized based on a consensus from a number of experts, and then applied this across the cohorts. And what we found was the cognitive phenotypes were actually quite stable across all seven centers that we included in our study, and that the base rates of impairment were actually even quite similar.

And so what this told us was that the cognitive phenotypes that emerged, were largely reproducible and weren’t necessarily specific to any particular center, and that this might be a really valuable way to harmonize approaches and create a common language for how neuropsychologists can approach cognitive diagnostics in epilepsy. And then particularly on expanding this to a worldwide audience.

Mr. Khan
Thank you. So based on that worldwide audience question in your paper, you describe your work as an international effort and in significant detail, you very nicely discussed the importance of understanding how demographic and cultural factors impact neuropsychological testing, and thus this line of research. Though you had researchers from four continents in your working group, the study itself included data from sites solely within the US. Can you by chance discuss why international sites were not included and what additional efforts are needed in order to understand if the results might have varied –

If linguistic and cultural factors were considered in the analyses?

Dr. McDonald
Sure. So it’s a really good observation, and I think that we had initially considered using a cohort from London is our first international site. But the test battery itself was quite different from the US sites. And so after some consideration and discussion, we decided to keep our first set of analyses to seven sites that were all in the US and predominantly English speaking participants.

The test batteries were generally uniform and we thought that that was an advantage. There was some modest variability across tests and norms and demographics, but for the large part we thought this represented the first logical step so that we could ensure that if there was any lack of reproducibility across the cohorts, that it wasn’t directly related to major differences in the test batteries, languages in which the tests were administered or any major cultural differences.

But now that we have that and we’ve demonstrated the stability across our U.S. centers, we are now in a better position to extend this to international sites and test that diversity. And I think that Anny has done a wonderful job in already starting to accomplish this. And so maybe she could tell you more about what the next steps look like.

Dr. Reyes
Yeah, And really quickly, one also thing to mention is that although there were seven sites in the U.S. that were kind of spread across the U.S. so regionally, we know there are differences when it comes to factors that impact cognition. So the fact that we were able to show very similar rates northeast, you know, kind of the south, the West Coast, I think that also kind of strengthened our ability to show that the stability of the phenotypes.

Yeah. So regarding the international or just the cross-cultural application of this approach, we started with doing it in a Spanish speaking sample here in the US. This was data from NYU. We just recently published this paper in Epilepsia. And it was a relatively large sample of patients with temporal lobe epilepsy. So again, we’re trying to keep certain things kind of stable just to make sure that the phenotypes themselves is what we’re comparing.

So we showed that in the Spanish speaking sample with a very kind of structured battery that one of the rates of impairments were very similar to those for the English speaking sample. And the pattern itself was very similar. And then we’re also right now working with different sites internationally, including Dr. Rashi Shah from India. She actually presented this past INS her results where she shows that in a really large, multilingual sample, patients with temporal lobe epilepsy, that the rates of impairments, again, were very similar to the English speaking, as we call a proof of concept sample.

And the pattern itself was also very similar. One of the things that we’re noticing, though, across these different cultures or languages is that the single phenotype that we include looks slightly different. So for example, for the English speaking sample here in the U.S., these patients predominantly show impairments in language versus those in Spanish speaking samples. And then the sample from India, they show impairments in memory, right?

So we suspect that these subtle differences that we’re seeing across the cross-cultural application may be either due to underlying etiologies, the differences in that, whether the normative data that we’re using, the difference in test, however, it’s really going to take for us to examine that across other sites, international sites and then hopefully combine data and kind of do cross-culturally a comparison of these rates of impairment.

But so far, I think we feel pretty confident that these phenotypes are robust now across culture, across languages and across very different diverse samples.

Mr. Khan
Thank you for that. I’m really interested in that. I did go back to your point, Dr. McDonald, about the number of tests being administered within different populations. And so the number of test scores analyzed may have varied by site and perhaps even by patient. Given the number of scores analyzed having a significant relationship to the rates of impairment, do you think this affects the lens of which your results should be viewed and/or applied?

Dr. McDonald
So I think in terms of the number of tests given at each, whether there were two tests available or five available, Right. Yeah, you know, this is a really important issue and it’s one that has come up several times, not only from some of our collaborators and colleagues, but even from some journal reviews. And we’re in the process right now trying to determine what the optimal number of tests is to include per domain.

Given that the more tests you have, the higher likelihood that you might find an impaired score. Now, interestingly, we tested this in one of our larger samples and it wasn’t the case. We expected that the more tests we would include that, the more likelihood there would be of impairment. So we were surprised that that wasn’t the case. But we are actually planning on systematically testing that as we move forward in potentially even picking the most common tests and then capping the number of tests that we allow in order to be considered for impairment for each domain.

Mr. Khan
Sure. And then, Dr. Reyes, you mentioned the differences in phenotypes across cohorts, but some might argue that your core findings largely support that, what we know in general, that language and memory are most commonly impaired in this population. So can you explain how this work is indeed additive and to benefit future lines of research?

Dr. Reyes
Yeah. So we are seeing the kind of, as you mentioned, the expected isolated memory in language. However, the two phenotypes are very surprising is the generalized, so the global impairment. I mean, these patients have focal epilepsy, right? So based on the lesion model, we expect them to have impairments in memory or language. Right. And then there’s also a sizable group of patients are not demonstrating any deficits.

In some cohorts, it’s larger than 50%. Most of these patients are actually refractory patients who have been having chronic epilepsy for a very long time, which us neuropsychologists, you know, we think that these patients are going to demonstrate some level of impairment. And what we think is very interesting about this is that we can learn from these two unexpected groups, the groups with generalized impairments that are thought to have extra epilepsy pathology.

Right. We tend to focus mostly on the patients seizure control. However, these patients also demonstrate vascular risk factors. They demonstrate all the health co-morbidities that we tend to not really systematically examine in neuropsychological study. But seeing that they have this generalized impairment may suggest that there’s something else beyond the epilepsy going on. And then the intact phenotype is also another interesting group, because despite having very, you know, high epilepsy burden, very similar clinical profiles to other patients, they’re showing intact profiles.

So we can learn a lot from this group. They may be demonstrating protective factors. We’re noticing that they have higher education for the most part, higher occupational attainment. Some of them are bilingual. So really identifying factors within this subgroup could actually help us come up with innovative ways to think about can we help these patients stop having greater cognitive decline or even progress later on in life?

Mr. Khan
And then to take that work and I guess widen the lens across other populations, given that the neuropsychological impairments in pediatric temporal lobe epilepsy are not as strongly lateralized as an adult, what are the implications of your research in pediatric populations, and do you have suggestions for future pediatric focused research in this area, if any?

Dr. Reyes
Yeah, I could take that question. So Dr. Kayla Hirano from Cleveland Clinic did a study where she applied the approach to frontal lobe epilepsy. And now we’re currently working in adults with genetic generalized epilepsy. So those are two adult different cohorts. And then when it comes to pediatric, Dr. Bruce Herman, Robin Bush and Dr. Janet Jones have initiated a study of over 400 youth with pharmaco-resistant epilepsy and are really applying those phenotypes.

We’re primarily interested in seeing how the phenotypes are stable over time, particularly as they intersect with the natural course of development. And as you mentioned, you know, there is not such a lateralized profile, so I think we can learn a lot from the pediatric cohorts and really follow these phenotypes to really see how they look in adulthood. So this is work that is currently going underway.

Mr. Khan
Okay. And I’d love to see that work as well. And then, as I understand it, other researchers have begun to apply your methodological approach to other kinds of populations beyond epilepsy. For those interested in doing so, how many patients or sites do you think are needed in order to produce reliable results? And are there any critical lessons you’ve learned?

Would you like to share?

Dr. McDonald
You’re absolutely right, and it’s very exciting to see it being picked up and applied in other populations. Anny has been doing this in some of our brain tumor work, especially so that we can track whether or not phenotypic membership changes following the course of radiation treatment. Doctors Bush and Herman have been on the forefront of applying it not only to COVID to look at cognitive profiles, but it’s been applied within an MS population.

And so I think it’s exciting to see the results extend way beyond epilepsy. I think it really could. And I think that does lead into one of the ways in which it could be impactful in the field within epilepsy or far beyond that. And that is that, you know, we still tend to lump patients, and this might be more so within epilepsy, according to syndrome.

So we’re putting them all patients with temporal lobe epilepsy, all patients with generalized epilepsy. And to really not consider the heterogeneity that we see within these disorders. In addition, we tend to study their cognitive deficits in isolation. Do they have a memory impairment or not? And that’s certainly how we look at clinical trials. Usually the outcome is a memory impairment or not.

And I think that concept is particularly important since a patient with a language and memory deficit is very different from a patient with an isolated language impairment. And the phenotyping approach really highlights that heterogeneity as well as having a more person centered approach. It really encourages us to consider the profile of impairments that a person has rather than looking at their deficits in isolation.

So we hope that the IC code and similar approaches will really encourage the use of profiling in research. It’s certainly what we do in clinical practice. So I’m seeing that we should treat our research the same way. We should study patients the same way that we contextualize them in clinical practice. And I think this is a step in that right direction.

Mr. Khan
Glad to hear that. Just a question about your phenotype and profiling. You guys used the one and a half standard deviation for impairment, is that correct? 1 to 1 and a half is that if I remember correctly, do you think that selective cutoff itself has unique relationship with neuropathology in general, but specifically within and across types of epilepsy?

Dr. Reyes
Yeah. So we tested negative one, -1.5. And then in the Spanish speaking sample, we also tested negative two. So we’re trying to figure out as well what is the optimal cutoff, where we are optimizing sensitivity as well as specificity. Right? And -1.5 is consistently coming across as that cutoff, where the rates are very similar to previous published findings and the rates are similar across the different cohorts that we have examined so far.

So it really is showing that sensitivity. However, we also are currently testing phenotypes in older adults, right where we might actually use a negative one cutoff. So one of the things that as initiative is to really think about the particular patient population that we’re seeing and what cutoff makes the most sense. But again, it’s going to take more studies and larger cohorts to really kind of test out the sensitivity of the cutoff, which for us neuropsychologists is quite important.

Mr. Khan
Definitely makes sense. All right. Thank you, Dr. Reyes and Dr. McDonald and I will turn it back to Dr. Sperling.

Dr. Sperling
Thank you. And that is all the time we have for today. I really appreciate, Humza, you leading this interview. And Drs McDonald and Reyes, for your valuable time. This is clearly very important work, very exciting work, not only for the implications and the results within the epilepsy populations as you demonstrated, but really in the fact that you’re able to put forth this framework, this sort of methodological framework for really investigating cognition, cognitive phenotypes, as we talked about, can really be applicable across just different disorders.

So again, thank you very much on behalf of the Society for Clinical Neuropsychology and the Journal of Neuropsychology. We appreciate you and your time and all the excellent research that you’ve put forth. Thank you. And take care.