Thank you very much to the pediatric section for this opportunity to share our work. My name is Andrew Hale, and I'm an MD-PhD student at Vanderbilt, currently applying to neurosurgery residency programs. We know that the burden of hydrocephalus is substantially higher in the developing world, and this is predominantly due to infection. However, the pathogens that cause hydrocephalus are largely unknown. Thus, identification of causal pathogens would inform development of targeted therapies aimed at hydrocephalus prevention. And one way to understand the causal relationship between infection and hydrocephalus is by using genetics. Thus, we first had to understand the genetic basis of infectious disease. And here, we conducted a comprehensive analysis of 35 infectious diseases in the electronic health record. And then, we analyzed adverse phenotypes associated with infection and causal phenotypic inference analysis by an approach called Mendelian randomization. And the purpose of this talk is to discuss one aspect of this study, causal inference by Mendelian randomization, and our efforts to understand the relationship between infection and hydrocephalus and shunt failure. But how does Mendelian randomization work? Well, it can be thought of like a randomized control trial. However, in the case of infection, it would be unethical to withhold treatment for infection, so a classic randomized control trial cannot be performed. And furthermore, we know that babies that develop infections are far more likely to develop other complications and may be a sicker population overall. So how do we know that the infection itself versus some other confounding factor is in fact causing hydrocephalus? So what Mendelian randomization does is assign groups according to genetic variation. And since genetic variants are inherited randomly according to Gregor Mendel's second law, we make the assumption that the genetic variant does not influence risk of these other confounding factors, such as prematurity and other complications. Thus, using the resource of infectious disease genetic associations we developed, we identified both risk-conferring and protective genetic variants associated with infectious diseases and compare hydrocephalus and shunt failure outcomes between those groups, in effect performing a randomized control trial using genetic variation to assign groups. Thus, of all 35 infectious diseases tested, only meningitis reached statistical significance after multiple testing correction for both hydrocephalus and CSF shunt failure. And this is perhaps not surprising. But what specific pathogens are causing hydrocephalus or shunt failure, and what about in cases where patients are infected with multiple pathogens? How do we disentangle the causative pathogenic organism? To begin to answer this question, we extracted all pathogen culture information from both blood and CSF, which identified 96 and 17 unique pathogens, respectively. And now we are in the process of performing Mendelian randomization for each individual pathogen to identify causative pathogens for both hydrocephalus and shunt failure. And these data will inform our understanding of the relationship between infection and hydrocephalus and shunt failure, and perhaps inform the development of novel targeted therapies. In conclusion, Mendelian randomization enables causal inference between risk factors and outcomes akin to a randomized control trial, which we apply to infectious disease and hydrocephalus and shunt failure. And lastly, identification of causal pathogens may lead to new antimicrobial strategies focused on hydrocephalus and CSF shunt failure prevention. This work would not have been possible without the extraordinary support of my co-authors and funding sources listed here. Thank you very much for your attention, and I'm happy to take any questions.

hydrocephalus

developing world

pathogens

infectious diseases

genetic basis