Scholarlink

Background <p>Glioblastoma (GBM) has poor survival and lacks validated minimally invasive biomarkers for differential diagnosis, surveillance, and risk stratification. Proteomics can capture functional tumor biology and secreted/extracellular vesicle (EV)-associated proteins across tissue and accessible biofluids.</p> Methods <p>This PRISMA 2020-compliant systematic review was prospectively registered (PROSPERO: CRD420251111520). Databases were searched for 2015–2025 (initial: 31 May 2025; update: 09 Dec 2025). We included original human GBM studies in which candidates were nominated from high-throughput proteomics and evaluated for diagnostic performance (e.g., AUC/sensitivity/specificity) and/or prognostic time-to-event outcomes (Overall Survival/Progression-Free Survival, with HR/CI). Risk of bias was assessed using QUADAS-2 (diagnostic) and QUIPS (prognostic). Because the data were heterogeneous and frequently incompletely reported, the synthesis was narrative, organized by specimen type and clinical purpose.</p> Results <p>Of 836 screened records, 22 studies were included, comprising 9 diagnostic and 14 prognostic analyses across various specimen types (Serum/Plasma, Cerebrospinal Fluid (CSF), urine EVs, and tumour tissue), with one study contributing to both diagnostic and prognostic analyses. Diagnostic studies were largely retrospective case–control with limited external validation and frequent healthy-control comparators, yielding high but spectrum-dependent performance in some panels/markers. Prognostic evidence was mainly tissue-based, with several proteins/signatures associated with OS/PFS, including a small number with external validation. The overall risk of bias in both diagnostic and prognostic studies was high.</p> Conclusion <p>Proteomic biomarkers in GBM show recurrent biological themes but remain largely developmental. Translation requires spectrum-representative cohorts, standardized pre-analytics, prespecified thresholds, locked pipelines, transparent model reporting, and multicenter external validation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Proteomic biomarkers for diagnosis and prognosis in glioblastoma: A systematic review

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月4日

查看全文
  • Farnia Ghasemi,
  • Mohammad Amin Gholami,
  • Vahid Mohammadkarimi,
  • Erfan Sadeghi,
  • Hossein Molavi Vardanjani,
  • Ramin Ajami

摘要

Background

Glioblastoma (GBM) has poor survival and lacks validated minimally invasive biomarkers for differential diagnosis, surveillance, and risk stratification. Proteomics can capture functional tumor biology and secreted/extracellular vesicle (EV)-associated proteins across tissue and accessible biofluids.

Methods

This PRISMA 2020-compliant systematic review was prospectively registered (PROSPERO: CRD420251111520). Databases were searched for 2015–2025 (initial: 31 May 2025; update: 09 Dec 2025). We included original human GBM studies in which candidates were nominated from high-throughput proteomics and evaluated for diagnostic performance (e.g., AUC/sensitivity/specificity) and/or prognostic time-to-event outcomes (Overall Survival/Progression-Free Survival, with HR/CI). Risk of bias was assessed using QUADAS-2 (diagnostic) and QUIPS (prognostic). Because the data were heterogeneous and frequently incompletely reported, the synthesis was narrative, organized by specimen type and clinical purpose.

Results

Of 836 screened records, 22 studies were included, comprising 9 diagnostic and 14 prognostic analyses across various specimen types (Serum/Plasma, Cerebrospinal Fluid (CSF), urine EVs, and tumour tissue), with one study contributing to both diagnostic and prognostic analyses. Diagnostic studies were largely retrospective case–control with limited external validation and frequent healthy-control comparators, yielding high but spectrum-dependent performance in some panels/markers. Prognostic evidence was mainly tissue-based, with several proteins/signatures associated with OS/PFS, including a small number with external validation. The overall risk of bias in both diagnostic and prognostic studies was high.

Conclusion

Proteomic biomarkers in GBM show recurrent biological themes but remain largely developmental. Translation requires spectrum-representative cohorts, standardized pre-analytics, prespecified thresholds, locked pipelines, transparent model reporting, and multicenter external validation.