Introduction About 5% of patients with severe cholestatic liver disease have intractable pruritus. Albumin dialysis is an effective treatment for these patients but the mechanism through which it reduces the severity of pruritus is not clear. Autotaxin (ATX) is a 125 kD protein which cleaves a choline group of lysophosphatidylcholine (LPC) thereby forming the biological active lysophosphatidic acid (LPA). Increased serum ATX activity has been described in patients with cholestatic pruritus (Gastroenterology, 2010; in press).

Aim The aim of the study was to determine the effect of albumin dialysis using molecular adsorbents recirculating system (MARS) on autotaxin activity and bile salts and their relationship to the severity of pruritus.

Method 15 patients (11F/4M, PBC 10, PSC 2, other 3) with severe pruritus that was resistant to medical therapy were treated in 31 sessions (each consisting of 3, 8-h treatments; median follow-up of 12.8 months) with MARS. The intensity and severity of itch was quantified using an itch severity scale (ISS) and the visual analogue scale (VAS) pre and post treatment and then weekly for up to 12 weeks. ATX activity was measured in diluted serum samples and albumin dialysate before and after MARS treatments using a fluorescence assay. ATX protein levels were determined by Western blotting. Bile salts were measured by LCMS.

Results MARS treatment was associated with immediate, significant and complete response (R) in 11 patients, two patients had a partial response (PR) and two patients had no response (NR). Median ATX activities were not different between responders and non-responders (R=216.5; NR=306.6). A mean reduction of ATX activity of 27.6±4.13% was seen in R, 10.7±10.3% in PR and 1.5±1.48% in NR. This change in ATX activity was directly correlated with the reduction in ISS (r2=0.59; p<0.005) and VAS (r²=0.47; p<0.02). The change in serum ATX activity correlated closely with the change in serum ATX protein level (r²=0.5; p<0.01). Expectedly, no ATX activity was measureable in the albumin dialysate. ATX levels returned to pre-treatment values with relapse of itching which occurred in all responders between 6 weeks and 4 months. No significant changes in serum bile salts were observed.

Conclusion Our study suggests an important role for ATX activity in modulating the severity of pruritus in cholestatic patients providing novel insights into the pathogenesis of itch in cholestatic disease. The reduction is not due to removal of circulating ATX into the dialysate suggesting that MARS treatment either reduces the production of ATX or enhances its hepatic clearance. Alternatively, or in addition, either a substrate or co-factor of ATX is extracted by the MARS system.