Changes in Circulating Insulin and Glucose Homeostasis After Portosystemic Shunt Creation and Obliteration: A Systematic Review

BACKGROUND & AIMS: The liver removes about half of secreted insulin on first pass. Transjugular intrahepatic portosystemic shunt creation diverts portal blood past the liver; endovascular shunt obliteration restores it. We systematically reviewed glycaemic and insulin outcomes after both. METHODS: PubMed, Embase and Web of Science were searched through September 2026 for paired glucose or insulin measurements before and after either intervention. Ratios of post-intervention to pre-intervention means were pooled by random effects with Hartung-Knapp adjustment, excluding cohorts at critical risk of bias. RESULTS: Fifteen cohorts (17 records) were included. After shunt creation, fasting insulin rose 49% (95% confidence interval 24% to 80%; 6 cohorts) and the homeostasis model assessment of insulin resistance rose 55% (5% to 129%; 4 cohorts), whereas fasting glucose (2.5%; -0.3% to 5.4%; 5 cohorts), glycated haemoglobin, and fructosamine were unchanged. Glucose tolerance worsened in one of two cohorts tested (area under the curve +14%). Insulin rose more than C-peptide or proinsulin in four cohorts; clamp-measured insulin sensitivity was unchanged. After obliteration, fasting insulin fell 15% (one cohort) and insulin area under the curve 17% and 19% (two cohorts). Antihyperglycaemic requirements changed in both directions after shunt creation and fell after obliteration. CONCLUSIONS: Shunt creation raises insulin by nearly half without changing fasting glucose or glycation markers, and obliteration reverses the pattern. Secretion markers change less than insulin, indicating altered first-pass hepatic insulin extraction rather than secretion. Insulin-based resistance indices may mislead after shunting, and the consequences of chronic shunt-induced hyperinsulinemia need prospective study.