Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Bioinformatics, Computational and Systems Biology

Bioinformatics, Computational and Systems Biology - Poster Session E

Poster M1 - Predicting Functional Recovery Following Spinal Cord Injury Utilizing Proteomic Datasets

Saturday, October 26, 2024

10:00 AM – 11:00 AM EST

Location: Exhibit Hall E, F & G

Presenting Author(s)

Rebecca L. Riter, MS

Graduate Student
Northwestern University, United States

Co-Author(s)

AT

Ansen Tan

Graduate Student
Department of Biomedical Engineering, Northwestern University, United States
CA

Christopher S. Ahuja, MD, PhD

Clinician, Neurological Surgery
Department of Surgery, Northwestern University Feinberg School of Medicine, United States
NE

Najib El Tecle, MD

Assistant Professor
Department of Surgery, Northwestern University Feinberg School of Medicine, United States
AS

Adam M. Sonabend, MD

Associate Professor
Department of Surgery, Northwestern University Feinberg School of Medicine, United States

Primary Investigator(s)

GA

Guillermo Ameer

Daniel Hale Williams Professor of Biomedical Engineering, McCormick School of Engineering Professor
1 Department of Biomedical Engineering, Northwestern University, 2 Center for Advanced Regenerative Engineering (CARE), 3 Department of Surgery, Northwestern University Feinberg School of Medicine, United States

Introduction: Spinal cord injury (SCI) is a traumatic neurological injury that impacts hundreds of thousands of people a year and results in life-long debilitation and economic burden. The current gold standard of treatment involves immediate surgery to decompress the spinal cord followed by prolonged rehabilitation, however, there are no clinically available effective regenerative interventions despite increasing publications with promising animal study results [1]. Failure of translation is due in part to differences in response to spinal cord injury between animal models and humans and results in large losses of time and money [2]. To better predict clinical response, an approach called translatable components regression (TransCompR) developed by Brubaker, et. al, is used here to identify predictive proteomic components in a porcine proteomic dataset of human functional recovery [3]. The identified porcine proteomic components predictive of recovery outcome can then be used in future preclinical porcine studies to better predict translatability of potential neuroregenerative SCI therapeutics.

Materials and

Methods: Publicly available clinical and porcine proteomic datasets were obtained from Skinnider, et. al [4]. Cerebrospinal cord fluid (CSF) and serum samples were collected every twenty-four hours from one hundred eleven spinal cord injury patients for five days following injury. Neurological examination was used to determine injury severity and sensorimotor function at baseline (immediately following injury) and six-months post injury. Twenty-eight pigs received contusion spinal cord injuries with CSF and serum samples taken at five time points during the first five days following injury. Datasets were preprocessed before TransCompR method was applied at each time point. A porcine principal components analysis model was created that the human dataset was projected onto. A principal components regression against the human six-month recovery outcome (sensorimotor function improvement from baseline vs. no improvement from baseline) was then performed to identify translatable porcine principal components. The model was deemed significantly predictive of human recovery outcome if the overall p-value was significant (p < 0.05).

Results, Conclusions, and Discussions: Models utilizing the clinical and porcine CSF proteomic datasets were not significant (p>0.05) when compared across each time point directly. The TransCompR model was most predictive of human outcome when proteomic datasets at 48 hours were used (p=0.113) (Figure 1A). Further, it was not significantly predictive of human outcome at any combination of timepoints using the CSF data (Figure 1B). Clinical and porcine serum proteomic datasets were next compared and, interestingly, a similar pattern of model predictiveness was seen at the earlier time points. The model has a high p-value when using the 24-hour datasets (p=0.830) but becomes more significant at 48 hours before increasing again at 72 hours (p=0.167) (Figure 1C). However, with the serum proteomic datasets, the model was significantly predictive of human recovery outcomes at 48 hours (p=0.011) (Figure 1C). Further, when comparing across different time points, the model was significant when using human serum datasets from 96 hours post SCI and pig proteomic data from 72 hours post SCI (p=0.0093) (Figure 1D). Ongoing work involves identifying and analyzing pathways involved in the top principal components of these two significant models.

In summary, two models were found to be predictive of human functional recovery; serum proteomic datasets at 48 hours post SCI and serum proteomic datasets at 96 (human) and 72 (pig) hours post SCI. This work can improve translatability by providing a way to predict human functional recovery by using the most significant porcine principal components in these predictive models. Future work will involve integrating proteomic mouse data to potentially develop murine predictive models of human functional recovery following spinal cord injury. Significant principal components of predictive models will be used to identify animal biomarkers indicative of successful human recovery, providing a better metric of success for neuroregenerative technologies in animal studies before potentially investing large amounts of time and money in a clinical trial.

Acknowledgements (Optional): References
(1) Hejrati, N.; Fehlings, M. G. A Review of Emerging Neuroprotective and Neuroregenerative Therapies in Traumatic Spinal Cord Injury. Curr. Opin. Pharmacol. 2021, 60, 331–340. https://doi.org/10.1016/j.coph.2021.08.009.
(2) Donovan, J.; Kirshblum, S. Clinical Trials in Traumatic Spinal Cord Injury. Neurotherapeutics 2018, 15 (3), 654–668. https://doi.org/10.1007/s13311-018-0632-5.
(3) Brubaker, Douglas. K.; Kumar, Manu. P.; Chiswick, Evan. L.; Gregg, C.; Starchenko, A.; Vega, Paige. N.; Southard-Smith, Austin. N.; Simmons, Alan. J.; Scoville, Elizabeth. A.; Coburn, Lori. A.; Wilson, Keith. T.; Lau, Ken. S.; Lauffenburger, Douglas. A. An Interspecies Translation Model Implicates Integrin Signaling in Infliximab-Resistant Inflammatory Bowel Disease. Sci. Signal. 2020, 13 (643), eaay3258. https://doi.org/10.1126/scisignal.aay3258.
(4) Skinnider, M. A.; Rogalski, J.; Tigchelaar, S.; Manouchehri, N.; Prudova, A.; Jackson, A. M.; Nielsen, K.; Jeong, J.; Chaudhary, S.; Shortt, K.; Gallagher-Kurtzke, Y.; So, K.; Fong, A.; Gupta, R.; Okon, E. B.; Rizzuto, M. A.; Dong, K.; Streijger, F.; Belanger, L.; Ritchie, L.; Tsang, A.; Christie, S.; Mac-Thiong, J.-M.; Bailey, C.; Ailon, T.; Charest-Morin, R.; Dea, N.; Wilson, J. R.; Dhall, S.; Paquette, S.; Street, J.; Fisher, C. G.; Dvorak, M. F.; Shannon, C.; Borchers, C.; Balshaw, R.; Foster, L. J.; Kwon, B. K. Proteomic Portraits Reveal Evolutionarily Conserved and Divergent Responses to Spinal Cord Injury. Mol. Cell. Proteomics 2021, 20. https://doi.org/10.1016/j.mcpro.2021.100096.