Low Pretreatment Albumin-to-Globulin Ratio Predicts Poor Prognosis in Gastric Cancer: Insight From a Meta-Analysis

Chengzhi Wei, Yu Zhu, Gonghe Wang, Yiming Zhou, Lei TianFrontiers in Oncology 2021 · DOI

Abstract

The albumin-to-globulin ratio (AGR) has emerged as a potential prognostic indicator in various malignancies. This meta-analysis investigates the relationship between pretreatment AGR and survival outcomes in gastric cancer (GC). Our findings suggest that AGR acts as an efficient prognostic indicator, with low pretreatment levels significantly predicting poor prognosis in GC patients.

Introduction

Gastric cancer (GC) remains a global health challenge, ranking as the third leading cause of cancer-related mortality. While the TNM (Tumor, Node, Metastasis) staging system is the primary tool for prognostic assessment, its predictive accuracy is limited by significant survival variability among patients within the same stage. Consequently, there is an urgent need for reliable, non-invasive biomarkers to refine prognostic stratification.

The albumin-to-globulin ratio (AGR) is a promising "liquid biopsy" marker that integrates two critical physiological parameters: nutritional status (represented by albumin) and systemic inflammation (represented by globulin). Low albumin levels often reflect malnutrition and a weakened host defense, while elevated globulin levels indicate a pro-tumor inflammatory environment. By combining these into a single ratio, AGR may offer a more robust prognostic signal than either marker alone, minimizing the influence of confounding factors like dehydration or fluid retention that can skew individual protein measurements.

Methods

A systematic review and meta-analysis were conducted following PRISMA guidelines. Researchers searched seven major databases, including PubMed, Embase, Cochrane, Web of Science, WanFang, CNKI, and VIP, for studies published through September 2020.

The analysis synthesized data from 12 cohort studies involving a total of 8,305 patients. Most studies were conducted in Asian populations (China, Japan) and Turkey. Eleven of the studies were retrospective, and one was prospective. Study quality was ensured using the Newcastle-Ottawa Scale (NOS).

Figure 1: Flow diagram of the study selection process.

Hazard Ratios (HRs) and 95% Confidence Intervals (CIs) were pooled using random-effects models due to significant heterogeneity ($I^2 = 76.8\%$). Subgroup analyses and meta-regression were performed to identify sources of bias, and the "trim and fill" method was used to address potential publication bias.

Results

The meta-analysis demonstrated a strong association between low pretreatment AGR and poor survival outcomes across various clinical contexts.

Overall Survival (OS)

Patients with low pretreatment AGR exhibited a significantly higher risk of death compared to those with high AGR. The pooled HR for OS was 1.531 (95% CI: 1.300–1.803, $P < 0.001$). This association remained statistically significant across various subgroups, including geographic location and tumor stage, indicating that low AGR is a consistent "red flag" for poor outcomes.

Figure 2: Forest plot showing the association between low pretreatment AGR and overall survival (OS).

Disease-Free and Progression-Free Survival

The impact of low AGR on disease recurrence was even more pronounced. The pooled HR for disease-free survival (DFS) or progression-free survival (PFS) was 2.008 (95% CI: 1.162–3.470, $P = 0.012$).

Sensitivity and Bias Analysis

Sensitivity analysis confirmed the stability of the results. Even after adjusting for potential publication bias using the "trim and fill" method, the association between low AGR and poor OS remained significant (adjusted HR = 1.204). Meta-regression identified the treatment method (surgery alone vs. multimodal therapy) as a primary source of heterogeneity between studies.

Discussion

This large-scale meta-analysis confirms that pretreatment AGR is a powerful, inexpensive, and readily available biomarker for predicting survival in gastric cancer. Its value lies in its ability to simultaneously capture the host's nutritional status and inflammatory state, both of which are critical drivers of cancer progression and cachexia.

While the findings are robust, several limitations must be addressed. The majority of data comes from Asian populations, and the lack of a standardized "cut-off" value for AGR (which ranged from 1.14 to 1.93 in the included studies) complicates its immediate clinical application. Future research should focus on large-scale, multi-center prospective trials to validate these findings in diverse populations and establish a universal threshold for AGR to facilitate its use in routine clinical practice.

Summary updated

Abstract

AGR could act as an efficient prognostic indicator for GC, and that low pretreatment AGR predicts poor prognosis in GC.
BibTeX
@inproceedings{wei2021low,
  author = {Chengzhi Wei and Yu Zhu and Gonghe Wang and Yiming Zhou and Lei Tian},
  title = {{Low Pretreatment Albumin-to-Globulin Ratio Predicts Poor Prognosis in Gastric Cancer: Insight From a Meta-Analysis}},
  booktitle = {Frontiers in Oncology},
  year = {2021},
  doi = {10.3389/fonc.2020.623046},
}
Figure