Extracorporeal iron removal with the chelating polymer MEX-CD1 during continuous hemodialysis in a porcine model of sepsis-associated acute kidney injury: a proof-of-concept pilot study

Background Sepsis-associated acute kidney injury (SA-AKI) has no specific treatment. Labile iron and iron-dependent lipid peroxidation (ferroptosis) have been implicated in tubular injury in experimental models. MEX-CD1 is a high-molecular-weight chelating polymer designed to be added to dialysate to extract labile iron extracorporeally. This pilot study assessed the feasibility and tolerance of dialysis-based iron removal in a large-animal model of SA-AKI. Methods Three anesthetized pigs of both sexes (43–57 kg) received autologous fecal peritonitis. After an observation period (planned 12 h, shortened in two animals because of hemodynamic instability), resuscitation (antibiotics, fluids, norepinephrine targeting a mean arterial pressure > 60 mmHg) was started together with a 24-h continuous veno-venous hemodialysis session using a MEX-CD1-containing dialysate at a deliberately low effluent dose (≈4 mL/kg/h) chosen to maximize contact time between blood and chelator. Hemodynamics, urine output, serum creatinine and urea, iron parameters (transferrin saturation, hepcidin), dialysate metals (Fe, Cu, Zn, Mn, Al) and renal histology (PAS with Jablonski scoring, Perls, TUNEL, GPX4 immunostaining) were assessed. No control group was included; results are descriptive (median [range]). Results All animals developed vasopressor-dependent septic shock, transient oliguria (onset 4–7 h after peritonitis), KDIGO stage 1–2 AKI on the creatinine criterion at the start of resuscitation (T12), and acute tubular necrosis on histology (Jablonski score 2–3), and survived to T48. The planned 24-h session was delivered in all animals; circuit dysfunction occurred in the two animals that received the lowest initial heparin doses, and heparin requirements far exceeded the planned regimen. MEX-CD1 extracted 158, 236 and 712 µg of iron into the effluent, and effluent iron concentration remained detectable throughout the 24-h session. Transferrin saturation was very low at baseline (<12%) and followed no consistent pattern; hepcidin remained low (≤3.0 µg/L). Removal of copper, manganese and aluminum was small relative to daily intake; zinc removal reached 4.6 mg. Tubular GPX4 immunoreactivity was diffuse, although focally reduced in injured regions. TUNEL staining was largely negative, and Perls’ staining was negative in the cortex, with scattered medullary foci. Conclusions Extracorporeal iron removal with MEX-CD1 is feasible and was clinically tolerated in a porcine model of SA-AKI. In the absence of a control group, no conclusion can be drawn on efficacy. The histological pattern is compatible with a non-apoptotic form of tubular cell death but does not demonstrate ferroptosis. These data support a controlled study comparing dialysis with and without MEX-CD1.