Dual-positive gastric cancer co-expressing AFP and CEA: an aggressive subtype defined by unique clinical and biological profiles.

While alpha-fetoprotein-producing gastric cancer (AFPGC) is a highly aggressive variant, a distinct subpopulation exhibits concurrent elevation of both AFP and carcinoembryonic antigen (CEA). In this review, patients with gastric cancer presenting with concurrent elevation of serum AFP (> 7 ng/mL) and CEA (> 5 ng/mL) are classified as dual-positive gastric cancer (DPGC). Compared to AFPGC and common gastric cancer, this phenotype exhibits profoundly more aggressive clinicopathological characteristics. Nearly all cases present at advanced clinical stage III-IV (98.25%) and exhibit extensive lymph node involvement (98.25%), alongside a remarkably high incidence of hepatic metastasis affecting 82.86% of the patients. Beyond its aggressiveness, DPGC frequently presents with severe systemic manifestations, including prominent immune-inflammation, malnutrition and cancer-associated thrombosis. While patients exhibit a median overall survival of merely 6 months under standard chemotherapy, they paradoxically achieve an extended survival of 18 months with combined immunochemotherapy, even though immune infiltration profiling reveals a highly immunosuppressive baseline microenvironment. Since the initial description of concurrent CEA and AFP elevation in gastric cancer in 1977, standardized diagnostic criteria for this extreme phenotype have remained undefined. To address this diagnostic gap, concurrent AFP and CEA evaluation is imperative for prognostic stratification and timely immunotherapeutic interventions. The clinical cohort data were derived from our previously published retrospective study enrolling 127 patients in total, including 57 dual-positive (DPGC) and 70 AFP single-positive (SPGC) individuals. Bioinformatic analyses were based on the TCGA stomach adenocarcinoma cohort. Normal samples and tumor samples without clinical data were excluded, leaving 370 gastric adenocarcinoma cases for analysis. AFP and CEACAM5 were used as transcriptomic surrogates for AFP and CEA, respectively. An AFP-CEA based risk score was calculated using Cox regression coefficients and corresponding gene expression values. Patients were divided into high- and low-risk groups according to the median score. All statistical analyses were performed using R software (version 4.5.1). Kaplan-Meier curves were used to estimate survival outcomes, and group differences were compared using the log-rank test. Cox proportional hazards regression was used for univariate and multivariate survival analyses. Immune infiltration analysis and gene expression comparisons were performed according to the risk groups. Between-group comparisons were performed using the Wilcoxon rank-sum test. All statistical tests were two-sided, and P < 0.05 was considered statistically significant.