# T-cell biomarkers improve urinary tract infection risk stratification beyond clinical characteristics after acute traumatic spinal cord injury

## TL;DR

T-cell biomarkers can better predict urinary tract infection risk in spinal cord injury patients compared to clinical factors alone.

## Contribution

The study introduces T-cell biomarkers as novel predictors of UTI risk in traumatic SCI patients.

## Key findings

- Higher CD8+ T-cell counts and IFN-γ/IL-4 ratio are linked to lower UTI risk.
- Adding T-cell biomarkers improves model accuracy for predicting UTI.
- Systemic immune alterations contribute to UTI development after SCI.

## Abstract

Patients with traumatic spinal cord injury (SCI) frequently develop urinary tract infections (UTI) which may compromise neurological rehabilitation. We explored risk factors for UTI, including immunological biomarkers, in patients with acute traumatic SCI during the transition from acute to subacute phase. Two observational studies were analysed separately: a retrospective study (COaT-SCI, n = 296, follow-up 13 weeks), and a prospective immunological study (SCIentinel, n = 70, follow-up 10 weeks). Cox regression models were used to examine associations of first UTI with patient characteristics and longitudinal immunological parameters. Cumulative incidence of first UTI was approximately 50% at six weeks. Results indicated higher UTI risk in older patients and time-varying associations for SCI severity (ASIA impairment scale) and presence of accompanying injuries. Higher cellular immune competence, particularly CD8 + T-cell counts and higher IFN-γ/IL-4 ratio in functional T-cell assays, were associated with a lower UTI risk (HR [95% CI] per log-transformed unit: CD8 + T-cells 0.34 [0.18—0.65], IFN-γ/IL-4 ratio 0.53 [0.33–0.83]). Adding cellular immune markers improved models’ discriminative ability (weighted C-index at 21 days from 0.68 to 0.76 with CD8 + T-cells). These findings indicate UTI is facilitated by systemic immune alterations, with T-cell parameters representing promising predictive biomarkers requiring further validation before clinical implementation.

The online version contains supplementary material available at 10.1038/s41598-025-34852-0.