Mapping the preclinical therapeutic landscape of cirrhotic portal hypertension: a systematic evidence map.

Abstract

Introduction and objectives: Portal hypertension has a limited pharmacological armamentarium despite decades of preclinical therapeutic research. We mapped the experimental architecture of this literature, independently of reported treatment efficacy.

Materials and methods: PubMed/MEDLINE was searched from inception to March 31, 2026. Two independent title-and-abstract screening procedures identified 833 records retained by at least one screen; 540 formed a concordant core retained by both. Records were classified using a prespecified taxonomy of therapeutic strategy, intervention modality, animal model, species, sex, treatment schedule and reporting features. Global counts included all records; variable-specific distributions used available-case denominators. Treatment effects were not extracted.

Results: Among records with classifiable mechanisms (n=709), vasoactive/neurohumoral (29.1%), NO-cGMP/endothelial (23.8%) and antifibrotic/HSC-extracellular-matrix strategies (20.0%) represented 72.9%. Among records reporting the relevant variable, rats accounted for 82.7% (677/819), males only for 85.9% (548/638), and CCl₄, bile-duct ligation/cholestatic models and thioacetamide for 45.2%, 30.3% and 10.6% (n=568), respectively. Randomization and blinding were identifiable in indexed information for 5.3% and 0.1% of all records. Principal distributions were preserved in the concordant core.

Conclusions: Preclinical portal-hypertension research has generated mechanistic diversity on a narrow experimental foundation. Broader use of both sexes, etiologically diverse models, repeated dosing, rigorous design and complementary human-based platforms may strengthen translation.