Plasma Protein Patterns Associated with Paliperidone Palmitate Maintenance Therapy in Schizophrenia: A Prospective Cohort Study

Background and Objective
Patients with schizophrenia exhibit significant interindividual variations in their response to pharmacotherapy, adverse effects, and clinical outcomes. While once-monthly paliperidone palmitate (PP1M) injections can improve treatment adherence and continuity compared with oral formulations, suboptimal therapeutic outcomes are still observed in some patients. Although the mechanisms underlying the variability in efficacy of long-acting injectables (LAIs) remain unclear, studies suggest an association with alterations in plasma protein expression. This study aims to investigate the correlation between interindividual differences in the response to PP1M treatment and changes in plasma protein abundance using proteomic analysis.

Methods
This prospective cohort study was conducted in Longgang District, Shenzhen, China. Utilizing Olink Proximity Extension Assay (PEA) proteomics technology, we analyzed plasma samples from 27 healthy controls and 28 patients with schizophrenia who were clinically indicated for and initiated on maintenance therapy with once-monthly paliperidone palmitate long-acting injection (PP1M), as assessed by their treating physicians. Plasma abundance levels of 92 proteins were measured in all patients with schizophrenia, both prior to their first PP1M administration and following at least 3 months of PP1M treatment.

Results
Our findings demonstrated that maintenance therapy with PP1M for at least 3 months effectively sustained and improved clinical symptoms in clinically stable patients with schizophrenia. However, it was associated with prolactin elevation, a known effect related to paliperidone. Utilizing Olink PEA analysis, we identified 25 plasma proteins, including SNAP23, ENO2, QDPR, ANXA11, and KYAT1, that distinguished patients with schizophrenia from healthy controls. Intriguingly, K-means clustering analysis of the differentially expressed proteins (DEPs) across the three groups (healthy controls, pretreatment, and posttreatment) revealed four distinct clusters characterized by specific expression patterns: class 1 (restoration to healthy control levels), class 2 (persistent elevation), class 3 (moderate recovery), and class 4 (further reduction). Analysis of plasma DEPs before and after at least 3 months of PP1M treatment identified significantly altered proteins (ENO2, QDPR, CRKL, ANXA11, KYAT1) exhibiting the class 1 expression pattern, characterized by a progressive restoration of plasma protein abundance to levels observed in healthy controls. Furthermore, analysis of DEPs associated with PP1M-induced hyperprolactinemia (a safety outcome) revealed that patients with schizophrenia demonstrated a higher probability of developing abnormally elevated prolactin levels after at least 3 months of PP1M treatment if they exhibited baseline elevations in ANXA11 and DIABLO, coupled with reduced CLEC5A expression.

Conclusions
This study reveals transition-associated proteomic signatures in schizophrenia, identifying protein markers that facilitate subgroup stratification and serve as treatment-emergent companion biomarkers to monitor pharmacodynamic shifts during therapeutic transition to PP1M injection therapy.