No more pilots, a phase III trial of fibrates in primary biliary cirrhosis is long overdue!


Dr John S Halliday, Department of Hepatology, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Email:

Primary biliary cirrhosis (PBC) is a chronic, cholestatic liver disease characterized by progressive destruction of the interlobular bile ducts that eventually leads to cirrhosis. Due to its hallmark serological signature, the antimitochondrial antibody (AMA) and similarly associated disease-specific T cell response, PBC is often considered a model autoimmune disease. Despite this, immunosuppressive therapies are ineffective for PBC and the only approved medical treatment is the hydrophilic bile acid, ursodeoxycholic acid (UDCA).1

A biochemical response to UDCA with normalization of alkaline phosphatase (ALP) is not achieved in some 30–40% of patients with PBC. Whilst it is established that responders to UDCA have a normal life expectancy, non-responders are at an increased risk of progression to liver transplantation or death.2,3 For this reason, the impetus for the discovery of adjuvant or alternative medical therapies for PBC persists. The potential efficacy of farnesoid X (FXR) receptor agonists such as obeticholic acid (OCA) are currently being evaluated in international multi centre trials with promising results in this group of refractory PBC patients.4 OCA is the first-in class agonist of the nuclear receptor farnesoid X, which controls bile acid synthesis and bile flow in the liver.

The potential role of fibrate therapy in PBC first became evident in the early 1990s when patients receiving fibrates for hypercholesterolemia were noted to have a reduction in their serum total ALP. In 1993, Day and colleagues demonstrated that this change resulted from reduced hepatic production of ALP.5 More recently, fibrates have been shown to activate peroxisome proliferator-activated receptor α (PPARα) and upregulate the expression of multiple drug resistance gene-3 (MDR3), both of which potentially ameliorate hepatic inflammation. However, whether the changes in cholestatic biochemistry observed with fibrate therapy will translate into an improved clinical outcome for PBC remains unclear.

The first studies to evaluate the use of fibrates for PBC appeared in the Japanese literature in the late 1990s and reports subsequently reached Western medical journals in 2000. There have now been approximately 20 small pilot studies/case series, 16 of which are from Japan, evaluating fibrate use either alone or in combination with UDCA for PBC.6–25

In the largest trial reported to date, Iwasaki and colleagues first compared fibrate monotherapy with UDCA; 45 patients were randomized to receive either therapy and evaluated at 52 weeks.18 They found bezafibrate (400 mg/day) to be as effective in reducing ALP, GGT, IgM and ALT levels as UDCA (600 mg/day). In a second study, they gave 21 patients with UDCA refractory PBC (defined by ALP > 1.5 normal) combined bezafibrate and UDCA therapy and importantly demonstrated a significant improvement in ALP levels.18

Overall, similar results to the work by Iwasaki have been reported in all fibrate studies in PBC. The great majority of these trials have used biochemical improvement alone as a measure of treatment success. In addition, no standardized criteria to define incomplete response to UDCA therapy have been applied, and all but a few studies have reported after a relatively short follow-up period of 3–12 months.16,26

Unfortunately, only two case series evaluating histological changes with fibrate therapy have been performed in a combined total of five patients; results have been mixed with histological improvement in some and worsening in others, irrespective of changes in liver biochemistry.12,26 Clearly, for an insidiously progressive disease like PBC, the conclusions that can be drawn from these small pilot trials are limited.

In this issue of JGH, Takeuchi and colleagues report yet another small pilot study of fibrate therapy in PBC. Over an 8-year period they consecutively enrolled 37 patients with PBC to receive 600 mg of UDCA. After 6 months treatment, those patients who failed to achieve a biochemical response to UDCA (defined by a fall in ALP > 40% or into the normal range), had bezafibrate therapy added. Fifteen (41%) of the 37 patients enrolled fell into this non-responder group and after one year of combined therapy, 12 of 15 (80%) had normalized their ALP and IgM levels with combination therapy. In an attempt to translate these biochemical improvements into a clinical outcome, Mayo risk scores were evaluated at enrollment and study conclusion at 2 years follow-up. No significant difference was noted between groups; this is not surprising, given the relatively short period of follow-up and small numbers.

The current study confirmed that at baseline, lower levels of ALP and early histological stage without PBC symptoms were both independent predictors of a “good response” to UDCA therapy. This has been previously demonstrated in a number of larger studies.2

Other than adding further weight to the effectiveness of fibrates in improving cholestatic biochemistry, what then does this new report tell us?

One apparent difference between bezafibrate and UDCA therapy are the changes in total serum IgM levels. In the “good responder” group who received only UDCA, IgM levels remained elevated in 8/22 patients at 24 months. In contrast, the addition of bezafibrate in the non-responder group resulted in normalization of IgM levels in all patients. An earlier study evaluating UDCA therapy in PBC demonstrated that improvements in total serum IgM may independently predict outcome.27 It is possible that the subgroup of biochemical responders identified in this study who have persistently elevated IgM levels may represent an additional subset for whom further therapy with newer agents might still be considered.

In conclusion, there is a now a substantial body of circumstantial evidence supporting the safety and possible efficacy of fibrates in PBC. Why have the plethora of positive successful pilot studies not led to a phase III study of fibrates for the treatment of PBC? Clearly, a controlled trial with sufficient numbers and follow-up to demonstrate changes in clinical outcome should be performed with the implication of the requirement of substantial financial resources. Fibrates are cheap, widely available and well tolerated. However, as they are now generic and off-label, pharma are unlikely to proceed with these studies. A concerted effort is required to achieve the required funding through non-pharma backed sources.