Preprocedural estimated glomerular filtration rate and all-cause mortality after TIPS in cirrhosis: a single-center retrospective cohort study

Technical Summary: Preprocedural eGFR and All-Cause Mortality After TIPS in Cirrhosis

Problem Statement
Transjugular intrahepatic portosystemic shunt (TIPS) is a standard therapy for portal hypertension complications, yet postoperative outcomes vary significantly, necessitating improved preprocedural risk stratification. While traditional scoring systems like Child–Pugh and Model for End-Stage Liver Disease (MELD) rely heavily on liver function, renal dysfunction is increasingly recognized as a critical prognostic determinant in cirrhosis. However, existing studies often utilize serum creatinine, a marker susceptible to confounding factors such as muscle mass and fluid status, rather than estimated glomerular filtration rate (eGFR). Furthermore, the specific prognostic value of preprocedural eGFR, particularly its independence from ascites status and its relationship with established liver scores, remains insufficiently defined in the context of covered-stent TIPS.

Methodology
This single-center retrospective cohort study analyzed 618 patients with cirrhosis who underwent first-time covered-stent TIPS between January 2015 and January 2024 at the Second Affiliated Hospital of Anhui Medical University and the First Affiliated Hospital of the University of Science and Technology of China.

• Data Collection: Preprocedural renal function was assessed using serum creatinine measured within one week of the procedure. To ensure stability, preference was given to patients with two consecutive measurements showing <20% change; others were included only if no acute kidney injury events were evident.
• eGFR Calculation: eGFR was calculated using the CKD-EPI 2021 creatinine-based equation. Patients were stratified into three groups: normal renal function (eGFR ≥ 90 mL/min/1.73 m², n=491), mild impairment (eGFR 60–89 mL/min/1.73 m², n=79), and significant impairment (eGFR < 60 mL/min/1.73 m², n=48).
• Statistical Analysis: Survival was evaluated using Kaplan–Meier analysis and Cox proportional hazards models. The primary multivariable model adjusted for age, serum sodium, and Child–Pugh score, deliberately excluding the MELD score initially to avoid structural overlap with creatinine-based variables. Sensitivity analyses were performed by adding the MELD score to the model and by modeling eGFR as a continuous variable. Restricted cubic spline analysis examined dose–response relationships, and subgroup analyses explored interactions across clinical strata.
• Outcomes: The primary outcome was all-cause mortality. Secondary outcomes included postprocedural hepatic encephalopathy, worsening ascites, and gastrointestinal rebleeding.

Key Results

• Mortality Association: During a median follow-up of 36.5 months, 146 patients (23.6%) died. In the primary multivariable model, preprocedural eGFR < 60 mL/min/1.73 m² was significantly associated with higher all-cause mortality compared to eGFR ≥ 90 (Hazard Ratio [HR] = 2.21, 95% CI 1.39–3.53, P < 0.001). The mild impairment group (60–89 mL/min/1.73 m²) did not show a statistically significant association after adjustment.
• Sensitivity Analysis: When the MELD score was added to the model, the association for eGFR < 60 was attenuated but remained directionally consistent (HR = 1.63, 95% CI 0.95–2.80, P = 0.078). Modeling eGFR as a continuous variable revealed that each 1 mL/min/1.73 m² increase in eGFR was associated with a 1.1% reduction in mortality risk (HR = 0.989, P = 0.004).
• Complications: Lower preprocedural eGFR was associated with significantly higher proportions of postprocedural hepatic encephalopathy (39.6% in the <60 group vs. 22.4% in the ≥90 group, P = 0.040) and worsening ascites (27.1% vs. 11.0%, P = 0.006). No significant difference was observed in gastrointestinal rebleeding.
• Subgroup Analysis: The association between low eGFR and mortality was generally consistent across subgroups defined by age, sex, Child–Pugh class, MELD score, and ascites severity, though interaction tests were not statistically significant, likely due to limited sample sizes in the impaired renal function strata.

Significance and Claims
The authors conclude that preprocedural eGFR < 60 mL/min/1.73 m² is a significant predictor of increased all-cause mortality following covered-stent TIPS in patients with cirrhosis. The study posits that this threshold (aligned with KDIGO guidelines) may serve as a practical reference marker for comprehensive pre-TIPS risk assessment.

However, the paper maintains a modest stance regarding the independence of this marker. The attenuation of the association upon adjustment for the MELD score suggests that the prognostic information provided by eGFR partially overlaps with overall liver disease severity and hepatorenal functional interactions captured by existing scoring systems. Consequently, the authors assert that preprocedural eGFR should be interpreted as a component of a broader clinical evaluation rather than as a standalone prognostic or decision-making criterion. The study highlights the need for multicenter prospective validation to confirm these findings and to better define the utility of eGFR in refining risk stratification strategies.