Predicting recovery from spinal cord injuries

Making predictions about the recovery success of spinal cord injuries is very difficult. Serological markers – derived from routine blood draws – are thought to correlate with the severity of spinal cord injury. Jutzeler and Bourguignon from ETH Zurich and an international team investigated the potential of such serological markers to improve recovery predictions.

The study investigated whether serological markers taken at the acute state (less than two weeks post-injury) of spinal cord injuries could predict motor performance of the chronic state (52 weeks post-injury). Researchers collected these markers within the first week post-injury from 363 patients, whose lower extremity motor scores (LEMS) were then evaluated a year later. LEMS assesses performance of several muscles in both limbs on a scale from 0 (total paralysis) to 5 (active movement).

Using different statistical methods and applying different regression models they tested the power of the serological markers in predicting the LEMS outcome after 52 weeks. This prediction was compared to a baseline prediction, which is simply based on the LEMS score in the acute state and the patient’s age.

No significant improvement was found in prediction accuracy using serological markers. However, the team found a way to enhance prediction accuracy: They were using different models for patients being completely paralyzed after the injury (LEMS score of 0) and patients that still showed some muscle movement (LEMS score above 0). This separation of patient groups made the statistical models more powerful for recovery predictions. This highlights the importance of incorporating clinical information into predictive models, rather than using serological markers.

Reference:
Matthias, J., Lukas, L.P., Brüningk, S.C. et al. Exploring the potential of routine serological markers in predicting neurological outcomes in spinal cord injury. Experimental Neurology, Vol. 380, October 2024. https://doi.org/10.1016/j.expneurol.2024.114918

Useful links:
About Catherine Jutzeler: https://hest.ethz.ch/forschung/Professuren/Professuren-nach-Forschungsbereich/persdetail.MTg3NjEz.U2VxdWVuY2UvMjQsLTE5NDc4MzQ2MTE=.html

Image:
Unsplash, National Cancer Institute