2026 ECNP Congress

PS02-2050

A multi-axis blood-based biomarker framework for domain-specific clinical profiling in schizophrenia

Daniele Formisano — F. Iannotta, L. Vellucci, C. Ceparano, A. Nasti, C. Polito, D. Sorvillo, F. Iasevoli, D. Terracciano, A. de Bartolomeis · University Medical School of Naples “Federico II”, Section of Psychiatry- Laboratory of Translational and Molecular Psychiatry and Unit of Treatment-Resistant Psychosis- Department of Neuroscience- Reproductive Sciences and Dentistry, Naples, Italy

2026 ECNP Congress

Background

The identification of blood-based biomarkers in schizophrenia remains challenging, with findings often inconsistent and difficult to translate into clinically meaningful information. One possible explanation is that schizophrenia is clinically and biologically heterogeneous and circulating biomarkers may reflect distinct dimensions of disease rather than global psychopathological severity.

Aims

This study aimed to investigate whether plasma neurofilament light chain (NfL) and complete blood count-derived hematological parameters capture complementary biological domains across psychopathological, neurological, cognitive, and treatment-resistance dimensions in schizophrenia.

Methods

A total of 126 subjects were included: 63 patients with schizophrenia (31 treatment-responsive and 32 treatment-resistant) and 63 age- and sex-matched healthy controls. Psychopathology was assessed using the Positive and Negative Syndrome Scale (PANSS) and the PANSS Autism Severity Score (PAUSS). Neurological vulnerability was evaluated with the Neurological Evaluation Scale (NES). Cognitive functioning was assessed using the Brief Assessment of Cognition in Schizophrenia (BACS). Plasma NfL was measured using a fully automated chemiluminescent immunoassay. Blood cell count parameters were retrospectively extracted from routine clinical assessments. Group differences were assessed using the Shapiro–Wilk test for normality and Mann–Whitney U tests; associations were explored using Spearman correlations and regression models (all p<0.05).

Results

Psychopathological domain:blood count-derived haematological parameters showed selective associations with specific symptom dimensions: eosinophil count was inversely associated with overall psychopathological severity (p=0.038) and positive symptoms (p=0.006). Both eosinophil (p=0.029) and lymphocyte counts (p=0.033) were negatively associated with autistic traits. Neurological domain:higher plasma NfL levels were associated with greater neurological vulnerability, as reflected by positive correlations with NES total score (ρ=0.28, p=0.001), sensory integration abnormalities (ρ=0.497, p=0.002), and complex neurological functions (ρ=0.496, p=0.002). In regression analyses, NES total score was independently associated with large unstained cells (LUC; β≈−0.41, p=0.018), linking blood cell count parameters to neurological features. Cognitive domain: NfL was also negatively associated with problem-solving performance (ρ=−0.267, p=0.036), suggesting that neuroaxonal damage may contribute to selected cognitive dimensions in schizophrenia. Verbal cognitive functions showed distinct associations with blood cell count parameters: verbal fluency was independently associated with platelet count (β≈0.27, p=0.048), while platelet-related indices (PCT and PDW) correlated with verbal memory (all p<0.05). Treatment resistance:treatment-resistant schizophrenia was characterized by a specific haematological profile. Lower eosinophil (p=0.020) and lymphocyte counts (p=0.035), together with younger age (p=0.011), independently predicted treatment-resistant status in logistic regression analyses. These findings were consistent with group comparisons showing reduced eosinophil (p=0.022) and lymphocyte levels (p=0.018) compared with treatment-responsive patients.

Conclusions

Overall, these results support a multi-axis blood-based biomarker framework for schizophrenia. Plasma NfL appears to capture a neuroaxonal vulnerability axis mainly related to neurological abnormalities and selected cognitive functions, whereas complete blood count-derived haematological parameters reflect complementary psychopathological, cognitive, and treatment-resistance dimensions. Rather than identifying a single global biomarker of schizophrenia severity, combined blood-based profiling may provide a more informative strategy for capturing disease heterogeneity and supporting mechanistically informed patient stratification.

References

No references