Gamma-glutamyl transferase as a redox-gatekeeper biomarker for pancreatic cystic neoplasms: A concise roadmap from epidemiology to bedside

Abstract

Serum gamma-glutamyl transferase (GGT), a marker of hepatobiliary and oxidative stress, emerged as a predictor of incident pancreatic cystic neoplasms (PCNs) in a 2.65-million nationwide cohort followed for > 10 years. The highest GGT quartile conferred 11% excess PCN risk, with hazard ratios rising across quartiles and persisting after 3- and 5-year lag exclusions. In redox-oncology, pancreatic GGT1 isoform upregulated by oncogenic KRAS cleaves extracellular glutathione, promoting reactive oxygen species-mediated DNA damage, Wnt/β-catenin, and interleukin-6/signal transducer and activator of transcription 3 inflammatory circuits cooperating with GNAS mutations to initiate intraductal papillary mucinous neoplasm. We propose: (1) Refining age-specific GGT thresholds via repeated measurements and restricted cubic splines; (2) Integrating GGT with carbohydrate antigen 19-9, carcinoembryonic antigen, and KRAS/GNAS circulating tumor DNA in machine-learning radiomic models for personalized 5-year malignancy risks per Fukuoka guidelines; and (3) Validating cost-effectiveness in multi-ethnic populations before screening. This repositions GGT as a globally available redox biosignature for PCN early detection, potentially reducing pancreatic cancer burden.

Key Words: Gamma-glutamyl transferase; Pancreatic cystic neoplasm; Biomarker; Cohort study; Early detection; Redox biology; KRAS; GNAS

Core Tip: This letter embeds the newly discovered gamma-glutamyl transferase-pancreatic cystic neoplasm association within a redox-oncology framework, supplies fresh single-cell evidence for pancreatic GGT1 induction by KRAS, and sketches a pragmatic three-tier roadmap (threshold refinement-multi-omic integration-cost-effectiveness validation) to expedite clinical translation.

TO THE EDITOR

We congratulate Lee et al[1] for their timely nationwide cohort analysis demonstrating that serum gamma-glutamyl transferase (GGT) anticipates incident pancreatic cystic neoplasms (PCNs) in a robust, dose-dependent manner. In 2655665 Koreans followed for a median 10.3 years, fully adjusted hazard ratios for PCN increased across ascending GGT quartiles: 1.043 (Q2), 1.075 (Q3), and 1.138 (Q4) vs Q1, with a significant P trend < 0.001. This gradient persisted after 3- and 5-year lag exclusions and remained consistent across sexes, metabolic phenotypes, and competing-risk models, conferring high epidemiological credibility to the observation.

GGT AND REDOX-DRIVEN PANCREATIC CARCINOGENESIS BEYOND THE LIVER

Although traditionally viewed as a hepatobiliary signal, membrane-bound GGT1 is highly expressed in pancreatic ductal epithelium, where it catalyzes extracellular glutathione cleavage. This process supplies cysteine for de novo glutathione synthesis but simultaneously generates reactive oxygen species (ROS) to activate KRAS-driven mitogen-activated protein kinase/phosphatidylinositol 3-kinase signaling. Persistent GGT upregulation, which is induced by oncogenic KRAS, amplifies Wnt/β-catenin and interleukin-6/signal transducer and activator of transcription 3 inflammatory circuits that promote intraductal papillary mucinous neoplasm (IPMN) progression. Using single-cell RNA-sequencing data from 26 human IPMN samples revealed a positive correlation between GGT1 transcript abundance and ROS-score (r = 0.42, P < 0.01), corroborating a direct redox link in human tissue[2]. Notably, this relationship may be bidirectional: KRAS activation drives GGT1 expression, while sustained GGT activity fosters a protumoral microenvironment rich in ROS and pro-inflammatory signals, potentially accelerating driver mutations (e.g., GNAS) and cystic transformation[3]. Thus, elevated serum GGT likely plays a more crucial role than previously known in facilitating GNAS-mutant clonal expansion and cystic transformation, thus might serving as a potential marker of a protumoral micro-environment.

CLINICAL TRANSLATION TOWARD A MULTI-OMICS RISK STRATIFIER

The 11% excess PCN risk observed in the top GGT quartile remained significant after adjusting for metabolic syndrome, alcohol intake, and chronic kidney disease, underscoring its independent predictive value. However, PCN involves biologically heterogeneous lesions, spanning low-risk serous cystadenomas to high-grade IPMNs with malignant transformation rates exceeding 50%. We propose a tiered translational roadmap to address heterogeneity toward developing individualized strategies.

Tier 1: Refine cut-offs

For prospective cohorts, collect repeated GGT measurements and model intra-individual trajectories with restricted cubic splines to identify age-specific thresholds that maximize sensitivity for high-grade dysplasia. This approach must account for physiological fluctuations and comorbidities affecting GGT levels.

Tier 2: Integrated biosignature panels

Combine GGT with blood-based molecular markers (carbohydrate antigen 19-9, carcinoembryonic antigen, circulating KRAS/GNAS circulating tumor DNA) and radiomic features (e.g., cyst wall enhancement, main-duct diameter) to enhance discriminatory accuracy beyond individual biomarkers. Machine-learning algorithms trained on multi-modal data could generate personalized 5-year transformation probabilities, aligning with the revised Fukuoka guidelines[4,5]. However, challenges in integrating heterogeneous data sources include standardizing radiomic feature extraction, harmonizing circulating tumor DNA assays, and the need for large, annotated multi-institutional cohorts for robust model training.

Tier 3: Clinical implementation and population-level screening feasibility

Given Korea’s higher alcohol consumption (8.2 L per capita) compared to China (5.7 L) and Japan (6.7 L), external validation in Western cohorts with distinct metabolic and genetic backgrounds is imperative. Cost-effectiveness analyses should evaluate assay frequency, downstream imaging expenditure, and life-years gained from early resection of high-risk cysts, while considering the potential for overdiagnosis and the psychological impacts of screening.

LIMITATIONS AND FUTURE DIRECTIONS

The original cohort study by Lee et al[1] directly informs and justifies the proposed research agenda. However, we would like to address its limitations in differentiating between PCN subtypes (e.g., IPMN, mucinous cystic neoplasm, and serous cystic neoplasm), which is a critical distinction given their varying malignant potential. A primary constraint is the reliance on International Classification of Diseases-10 codes, which precludes differentiation of PCN subtypes. This limitation was compounded by the absence of PCN subtype adjudication in the underlying database, which would have prevented an evaluation of whether GGT has specificity for mucinous vs serous lesions[6,7]. Methodologically, the use of single-timepoint GGT measurements may not accurately capture dynamic risk over time. Furthermore, while lag analyses were employed, the possibility of reverse causation cannot be definitively excluded.

To address these gaps, future research should pursue several directions. Linkage to prospective imaging registries could determine if GGT preferentially predicts mucinous IPMNs harboring specific mutations like KRAS/GNAS. Future nested case-control studies with baseline magnetic resonance imaging could confirm cyst-free status to strengthen causal inference[8,9]. Future research strategies should incorporate longitudinal tracking of GGT levels correlated with cyst size and growth, validation in Western populations with different profiles of alcohol consumption, integrating emerging tools like liquid biopsies (e.g., circulating free DNA) to explore GGT as a marker for early malignancy detection, and employing Mendelian randomization using GGT1 polymorphisms to dissect causality from confounding[9].

As the therapeutic landscape for pancreatic cancer evolves, the potential role of GGT in predicting response to novel systemic therapies warrants further exploration. Recent advances in targeted therapy and immunotherapy for subsets of pancreatic adenocarcinoma highlight the need for biomarkers that guide precise treatment selection[10-14].

CONCLUSION

In conclusion, Lee et al[1] repositioned GGT from an overlooked liver enzyme to a credible, globally available biosignature for PCN risk. Their epidemiological breakthrough lays the foundation for redox-targeted prevention trials and multi-parametric early detection models that may reduce the burden of pancreatic cancer one cyst at a time. Our proposed roadmap offers a structured path forward, bridging population science with molecular mechanisms, urging concerted efforts to validate and integrate GGT into the evolving paradigm of precision pancreatic oncology. As emphasized in recent comprehensive reviews, the future of pancreatic cancer management will depend on targeting its complexity through personalized approaches that account for individual tumor biology and host factors[10,12,13,15].