• primary biliary cirrhosis
• rifampicin
• cholestatic pruritus

Prince et al described three patients with primary biliary cirrhosis who developed hepatotoxicity when given rifampicin to treat their cholestatic pruritus (Gut 2002;50:436–9). They describe the use of rifampicin as “secondline” treatment of cholestatic pruritus. Firstline therapy is generally considered to be cholestyramine, a bile acid sequestrant. Use of this agent is frequently unsatisfactory because of gastrointestinal side effects, especially constipation. I am writing to summarise new evidence that retention of endogenous bile acids causes cholestatic pruritus, and to call attention to a recent abstract indicating that colesevalem, a new bile acid sequestrant, appears to be more potent than cholestyramine and does not induce constipation.

The view that bile acid retention causes pruritus is a very old one. Varco¹ in 1947 noted that biliary drainage reduced pruritus in patients with extrahepatic biliary obstruction and that when bile was fed to patients, their pruritus returned. Huet and colleagues² reported that biliary drainage improved cholestatic pruritus in patients with intrahepatic cholestasis. Administration of cholestyramine, an anion exchange resin with a strong affinity for bile acids, improved pruritus³ as did passage of plasma over charcoal or anion exchange resins.⁴ More recently, extracorporeal albumin dialysis, a procedure that removes bile acids, has been shown to diminish cholestatic pruritus.⁵ In all of these procedures, retained substances in addition to bile acids could have been removed at the same time so that cause and effect relationships were uncertain.

Partial biliary diversion is effective in treating cholestatic pruritus. A likely explanation for the efficacy of this surgical procedure is that it reduces the load of bile acids to the ileal transport system and thence to the liver, resulting in less retention of bile acids. In an important study, Hollands et al reported that ileal bypass was effective in treating cholestatic pruritus.⁶ The ileal transport system is considered to solely transport bile acids. Thus this report in the surgical literature provided unequivocal evidence for bile acids being related directly or indirectly to cholestatic pruritus.

If bile acids are causal agents, and if bile acid depletion improves cholestatic pruritus, then bile acid administration to cholestatic patients should induce pruritus. Ricci et al noted that in two of four patients with primary biliary cirrhosis given cholylsarcosine, a non-metabolisable conjugated bile acid analogue, pruritus was induced.⁷

Despite these convincing lines of evidence, recent authors have been sceptical of the role of bile acids in cholestatic pruritus. This scepticism has arisen because of the lack of correlation between plasma bile acid levels and the magnitude of pruritus in some, but not all, studies. However, in my opinion, the lack of correlation between plasma bile acid levels and pruritus does not exclude a causal role for bile acids for three reasons. Firstly, the level of plasma bile acids fluctuates diurnally and the composition of plasma bile acids in cholestasis is quite complex. Secondly, bile acids might act slowly on peripheral or central receptors (or cause the release of substances that act on these receptors) so there could be a long delay between elevated plasma levels and pruritus. Thirdly, endogenous molecules such as opiates could act synergistically with bile acids to induce cholestatic pruritus.

Cholestyramine binds bile acids but is frequently ineffective. The efficiency of cholestyramine binding is quite low because the K_m for the ileal transport system is in the micromolar range, and at this concentration bile acid binding is relatively weak. Thus the ileal transport system for bile acids acts as a sink to strip the bound bile acids from the resin. A new bile acid binding polymer, colesevalem, was synthesised by GelTex Pharmaceuticals (now Genzyme General) with a much superior binding affinity for bile acids. Colesevalem is a hydrogel that was developed as therapy for hypercholesterolaemia. Berg, in a recent abstract, reported that in eight patients with cholestatic pruritus who had not responded to cholestyramine, colesevalem was effective in five.⁸ Colesevalem has an additional advantage over cholestyramine in that colesevalem does not have gastrointestinal side effects.

In their thoughtful case report, Prince et al did not note that rifampicin has a striking effect on bile acid metabolism, inducing 6-hydroxylation.⁹ It could well be that 6-hydroxy bile acids either do not induce pruritus or are not conserved by the ileal transport system, as compared with endogenous bile acids, or both. Rifampicin is presumed to induce 6-hydroxylation of bile acids via cyp3A, a microsomal enzyme, whose synthesis is induced by the interaction of rifampicin and the nuclear receptor PXR. PXR activation might also induce bile acid sulfation.

If the use of rifampicin is hazardous in cholestatic disease, as suggested by Prince et al, then the treatment of cholestatic pruritus should involve a bile acid sequestrant such as colesevalem and/or an opiate antagonist. It would seem desirable to initiate a prospective, placebo controlled, double blind study using state of the art methods for quantifying itching that would compare the efficacy, safety, and convenience of these two agents, alone or in combination, on cholestatic pruritus.