Hemoperfusion with Hemodiafiltration Versus Standard Medical Therapy for Severe Acute Pancreatitis: A Retrospective Study.

BACKGROUND: Severe acute pancreatitis (SAP) triggers systemic inflammatory response syndrome (SIRS), often progressing to multiorgan dysfunction, with no targeted pharmacological interventions. This study evaluated hemoperfusion (HP) followed by sequential hemodiafiltration (HDF) as an adjunct to drug therapy for systemic inflammation. METHODS: This retrospective cohort study included 184 SAP patients; 104 received standard medical therapy alone (non-HP-HDF group), and 80 received HP sequential HDF (HP-HDF group) within 72 h. Propensity score matching (PSM) generated balanced cohorts (n = 41 each). Outcomes encompassed inflammatory markers, ICU/hospital stay, the rate of organ dysfunction, complications, and 28-day mortality. RESULTS: Inflammatory markers improved significantly in both groups after treatment (all P < 0.05). In a matched cohort, HP-HDF reduced SIRS rates (9.8% vs. 39.0%, P = 0.001), shortened ICU stay (median 0 vs. 1 day, P = 0.002), and hospital stay (median 8 vs. 10 days, P = 0.013) compared to standard medical therapy. 28-day mortality did not differ. The APACHE-II score (OR = 7.337, 95% CI: 3.377-19.102) was the predominant contributor to SIRS prediction in the non-HP-HDF group. The protective effect of HP-HDF against SIRS was consistent across three analytical approaches: PSM-adjusted analysis (aOR = 0.086, 95% CI: 0.011-0.642) in the matched cohort, as well as Inverse Probability of Treatment Weighting (OR = 0.127, 95% CI: 0.064-0.250) and Augmented Inverse Probability Weighting (ATE = -0.343, 95% CI: -0.471 to -0.214) in the full cohort (n = 184). The average treatment effect corresponded to an absolute risk reduction of approximately 34 percentage points. CONCLUSION: Early adjunctive HP-HDF was associated with reduced SIRS incidence and shortened hospitalization in SAP patients, without a significant effect on short-term survival. These findings warrant validation in prospective randomized controlled trials.