Continuous veno-venous haemodialysis (CVVHD) results in low extraction ratio and drug removal after massive sodium valproate overdose.

Abstract

Objective: Sodium valproate (VPA) has a small molecular weight and volume of distribution, with concentration-dependent protein binding. At supratherapeutic drug concentrations, free-drug fraction increases and contributes to its central nervous system toxicity. Intermittent haemodialysis (IHD) is recommended for rapid drug removal after overdose. Data on the effectiveness of CVVHD to enhance drug-elimination is limited. We report a case of VPA overdose, treated with CVVHD, with serial blood extraction ratios and dialysate drug recovery. Case report: A 26-year-old, 70 Kg male was found at home drowsy, 2 hours after suspected ingestion of 200 x 500mg (100 g, 1.4 g/Kg) enteric-coated VPA tablets. His Glasgow Coma Score (GCS) dropped to 3/15 during transit to hospital. He was emergently intubated in the emergency department and naso-gastric activated charcoal was administered. Three hours post-ingestion, serum [VPA] was 12,402 mumol/L (reference 350-700), ammonia 242 mumol/L (reference 16-50), and ethanol 27mmol/L. Cerebralcomputerised tomography (CT) scan was normal. IHD was recommended to increase VPA and ammonia elimination. Unfortunately, only CVVHD was available at the time. Serum [VPA] peaked 8-hours post-ingestion at 15,144 mumol/L. CVVHD was commenced 11 hours post-ingestion when serum [VPA] was 13,975 mumol/L and serum sodium was 161mmol/L. Multiple-dose activated charcoal (MDAC) was added to assist VPA elimination and intravenous carnitine was commenced for elevated ammonia. Afferent and efferent serum [VPA] was assayed across the dialysis filter. CVVHD blood flow was 200mL/min and dialysate flow 1.6 L/h. VPA recovery was measured in recovered dialysate fluid. Serum [VPA] fell to 932 mumol/L over 24 hours of CVVHD. Dialysis filter extraction-ratio averaged 23% when serum [VPA] was more than 8,000 mumol/L, however, this fell to less than 13% when serum [VPA] was below 3,500 mumol/L. Cumulative recovery of VPA in 30 litres of dialysate was 10.8 g. VPA elimination halflife was 26 and 5.4 hours, respectively, pre- and during CVVHD. Serum ammonia peaked at 616 mumol/L, 14 hours post-ingestion. The patient was extubated on day 3 post-ingestion, neurologically intact. Conclusion(s): IHD remains the treatment-of-choice to rapidly remove toxins. Sustained low-efficiency haemodialysis is also reported to be superior to CVVHD. Neither modality may be accessible in many hospital ICUs, and CVVHD is used as a replacement or bridging therapy. In this case, CVVHD achieved low extraction ratios and poor total recovery of VPA in dialysate fluid, despite potentially lethal serum VPA concentration, suggesting CVVHD played a moderate role in drug elimination in this patient. It is unclear how much MDAC contributed to VPA elimination in this case.