Pulmonary Vascular Permeability and Extravascular Lung Water Index Vary with Infection Site in Patients with Liver...

Background: Cirrhotic patients with septic shock have variable responses to albumin infusion depending on the infection site. While albumin improves outcomes in spontaneous bacterial peritonitis (SBP), it offers no benefit in non-SBP infections and increases the risk of pulmonary edema, particularly in cirrhotic patients with pneumonia. We hypothesized cirrhotic patients with septic shock due to pneumonia would exhibit higher pulmonary vascular permeability index (PVPI) values. Using Pulse index Continuous Cardiac Output (PiCCO) monitoring, we examined the impact of infection site on PVPI and extravascular lung water index (EVLWI). Methods: We conducted a retrospective cohort study of 147 cirrhotic patients with septic shock. All patients underwent PiCCO monitoring. Measurements including PVPI, EVLWI, disease severity scores, and laboratory biomarkers on the first and second day of ICU admission were analyzed. Results: Pneumonia was the most common infection and was associated with significantly elevated PVPI and EVLWI values on both day 1 and day 2 compared to the non-pneumonia group. By contrast, patients with SBP had significantly lower PVPI and EVLWI values than those with non-SBP infections. Within the pneumonia subgroup, 28-day non-survivors had higher PVPI and EVLWI values than survivors. PVPI on day 2 was the strongest predictor of 28-day mortality (AUROC 0.83) among the PiCCO parameters. A PVPI on day 2 > 2.65 identified a high-risk phenotype, with a higher 28-day mortality (77.1% vs 18.9%, p < 0.001). PVPI on day 2 remained an independent predictor of 28-day mortality after adjustment for disease severity. Additionally, PaO₂/FiO₂ ratio on day 2 was an independent factor associated with day-2 PVPI ≤ 2.65. Conclusions: In cirrhotic patients with septic shock, non-SBP infections, mostly from pneumonia, were associated with elevated PVPI and EVLWI values. PVPI-based risk stratification may help optimize resuscitation strategies for patients at risk of pulmonary edema while reducing heterogeneity in future clinical trials.