Monday, April 12, 2010

Rifaximin for Prevention of Hepatic Encephalopathy

Looking for a medication that can reduce a troubling symptom in an advanced disease and reduce hospitalization? Sound like the "perfect" palliative care drug? Read on and decide...

NEJM
published a randomized controlled trial comparing the antibiotic rifaximin at 550 mg po bid versus placebo for the prevention of recurrent hepatic encephalopathy. While other studies have suggested that rifaximin might be beneficial in treating hepatic encephalopathy, they have all been small studies of short duration and looked at the treatment of HE rather than prevention.

More than 90% of patients in both wings received concurrent lactulose therapy. Subjects were followed for six months with the primary endpoint being the time to the first breakthrough episode of hepatic encephalopathy. A major secondary endpoint was the time to rehospitalization. Patients with the most advanced disease were excluded (MELD score greater than 25 and the presence of several other complications such as creatinine greater than 2.0 g/dl). The study was funded by Salix Pharmaceuticals, the makers of rifaximin.

The results:

  • Breakthrough episodes of hepatic encephalopathy were reported in 31 of 140 patients in the rifaximin group (22.1%) and 73 of 159 patients in the placebo group (45.9%). The hazard ratio was significant at 0.42. The number needed to treat to prevent one episode of HE is 4.2.
  • Hospitalization involving hepatic encephalopathy was reported for 19 of 140 patients in the rifaximin group (13.6%) and 36 of 159 patients in the placebo group (22.6%). The hazard ratio is reported as 0.5 and the relative risk reduction was reported as 0.5 as well. However, based on my calculation, the relative risk reduction is closer to 0.4. They report a NNT of 9. I calculate a NNT of 11.
6.4% of patients in rifaximin group died in the six month period vs. 6.9% in placebo group. Rifaximin was generally well tolerated. Two patients in the study group developed Clostridium difficile versus none in the placebo group.

The authors point out potential downsides to alternative antibiotics, including neomycin. Oral neomycin can cause ototoxicity, nephrotoxicity, and peripheral neuropathy. (According to Lexicomp, this risk is less than 1% but there is a black box warning.) Metronidazole use may be limited by nausea and peripheral neuropathy.

While I don't prescribe neomycin for hepatic encephalopathy typically, there have been two previous clinical trials comparing rifaximin to neomycin. One showed no significant difference between the two over 21 days. Another showed a small advantage to rifaximin that was statistically insignificant. I've read several recommendations against neomycin because, in addition to the potential toxicity, there is clinical trial data suggesting that it is no better than placebo, but that's based on one clinical trial in patients who had acute encephalopathy where the neomycin group resolved within 40 hours compared to placebo 50 hours (not statistically significant).

I couldn't find any other trials head to head of rifaximin versus other therapies. Metronidazole and vancomycin don't appear to be well studied in general (although the former seems to be the most commonly used secondary agent behind lactulose).

After reviewing other studies, I wish there were a wing of this study for at least one of these agents. All are significantly less expensive and it would be nice to have a better understanding of the toxicity in a larger trial in this population. The "risk" for including one of these agents in this industry funded trial would be that the agent might work and not be as toxic as everyone thought it might be.

All of that aside, hepatic encephalopathy and the associated need for hospitalization can be a major burden for patients with advanced cirrhosis and their families. The study excluded the sickest patients (the ones most likely to be seen by palliative care or referred to hospice). But while it may be difficult to know how these results apply to a sicker population, it seems plausible that they can be extrapolated. Anything that reduces the incidence of encephalopathy or hospitalization is appreciated, but because of the cost, it makes sense that it should still be reserved for those with recurrent encephalopathy. And if started, one should monitor for the expected endpoints and discontinued if no improvement is seen. One could argue that other antibiotics should still be given a trial, as well- I'm not convinced that this trial knocks other antibiotics out of the "on-deck" circle behind lactulose. Look for a jump in the use of this medication because of this study and the lack of data behind other agents, though. Rifaximin now has FDA approval for this indication.

If rifaximin use does become more commonplace, it will be interesting to note how hospices handle the drug. While reducing hospitalization is always a good thing and may potentially save the overall health system money (maybe even a LOT of money, if you agree with this study), a hospice that is reimbursed on a per diem basis might struggle to cover this expensive medication (especially since the average patient with refractory encephalopathy may have more days to spend at home while taking this drug!) This type of therapy can be a real pickle for hospices- proven to benefit, palliative only, and expensive.

Even though it may be possible to extrapolate the results of this study to a population with more advanced liver disease, I don't think this study demonstrates any role for patients with a very limited prognosis (days to weeks).