Friday, February 12, 2010

Pharmaceutical Companies are Chicken

Actually, this is not a blog about the use of antibiotics and hormones in poultry. It is about the pharmaceutical industry’s risk averse approach to developing antibiotics. In the interests of balance, I think we have to look critically at the industry as well as regulatory agencies.

Vancomycin has been a real workhorse antibiotic for the treatment of MRSA infections in our hospitals. But we now have three new antibiotics for the treatment of these infections, linezolid, daptomycin and, in 2009, telavancin. At the same time, there have been a number of scientific and clinical studies to suggest that vancomycin isn’t such a great antibiotic for the treatment of MRSA infections. Both linezolid and daptomycin were compared, in their late stage clinical trials, to vancomycin and both were shown to be “non-inferior.” Thus, they both were no worse than vancomycin within a reasonable statistical margin of error. Because antibiotics work so well in a normal clinical trial, it is usually impossible to actually show superiority of one antibiotic over another. In a few, more specialized trials, there is some evidence that linezolid might be superior to vancomycin, but these data were never approved by the FDA. Had the FDA actually accepted these data and approved a label suggesting that linezolid was superior to vancomycin, Pfizer would have been allowed to make that claim in their marketing efforts. At the time that daptomycin and linezolid were studied, there was no antibiotic other than vancomycin for them to use as a comparator. But for telavancin, which was just approved in 2009, both linezolid and daptomycin had been approved and on the market for several years before the pivotal telavancin trials were started. Cerexa/Forest is developing a cephalosporin antibiotic active against MRSA. For their recently completed Ph. III trials, they also chose to use vancomycin as their comparator.

So why did Theravance - Astellas and Cerexa/Forest choose vancomycin as the comparator? I can think of a number of reasons why they should not have done so. First, vancomycin sells at a lower price. In Europe, the price of your comparator comes into consideration when you get into the ultimate price negotiations that you have to go through with each individual country to market a new drug. So using a higher priced comparator would have made the argument that you too can charge a higher price more palatable to the national authorities in Europe. In addition, some influential clinicians believe strongly that vancomycin is an inferior drug compared to both linezolid and daptomycin. There is some evidence to support this view even though the data could not be used by regulatory authorities to approve such a claim. In that case, why would you not want to show non-inferiority compared to what might be an inferior antibiotic? It is of interest that Cerexa/Forest are running a Phase II trial comparing ceftaroline to linezolid in the treatment of skin infections. This suggests that they are aware that some clinicians consider linezolid a superior drug but made an active decision not to pursue linezolid as a comparator in their first pivotal trials.

I think the answer to this question goes back to risk. Companies are risk averse and they worry that they might be wrong and that they might not succeed against the newer antibiotics even though the risk, in my view, is a small one given everything we know about all these drugs. I’m sure other considerations come into the discussion as well. Both daptomycin and linezolid are expensive. The companies, in this example Theravance-Astellas and Cerexa/Forest, would have to pay the going rate for the comparator drug for hundreds of patients. This would significantly increase the cost of the trials. Also, they might have considered that because vancomycin had been used in so many trials as a comparator, it was much better characterized and provided a more reliable comparator. Finally, they might not have wanted to provide additional data for their future competitors, Cubist (daptomycin) and Pfizer (linezolid). Perhaps they were all dissuaded by Arpida who carried out two Ph. III trials of their antibiotic, iclaprim, compared to linezolid and failed to show non-inferiority.

Theravance-Astellas or Cerexa/Forest might argue that, based on volume of use, vancomycin is still the most widely used antibiotic for the treatment of serious Gram positive infections in hospitals where MRSA is common. That vancomycin is the most widely used of the anti-MRSA drugs remains true partly for cost reasons, where vancomycin is considerably less expensive than linezolid and daptomycin. The fact that there are no data approved by a regulatory authority recognizing the superiority of either daptomycin or linezolid over vancomycin in the treatment of MRSA infections probably also contributes to this situation.

I believe that companies should be providing the best data for physicians and patients. This should be a prime consideration in the choice of a comparator for pivotal clinical trials. The comparator should be the gold standard. The gold standard is not always the same as the so-called standard of care.

Monday, February 1, 2010

Generic Antibiotics: Effective and Safe?

Today I am going back to the FDA’s Advisory Committee (AIDAC) meeting on a new antibiotic, telithromycin (Ketek), in December, 2006. Telithromycin is an antibiotic distantly related to erythromycin and azithromycin (Zithromax, Z-Packs) that is active against strains resistant to the older antibiotics. During this meeting, the committee recommended that its marketing approval be withdrawn for the indications of sinusitis and bacterial exacerbations of chronic bronchitis. Telithromycin was contentious from the start. Bob Moellering and I wrote a Perspective on the entire process for Lancet Infectious Diseases in 2008 if you are interested in a summary of telethromycin’s history. But the argument at the AIDAC in 2006 essentially boiled down to the following logic:
1. The FDA has decided that superiority or placebo controlled trials would be required to show efficacy for self-limited infections like otitis media, sinusitis and bacterial exacerbations of chronic bronchitis.
2. Telithromycin is associated with rare but serious liver toxicity, aggravation of myasthenia gravis (a rare neurologic condition) and more common visual disturbances that are generally not very serious.
3. Because telithromycin was not subjected to superiority or placebo controlled trials to prove efficacy in sinusitis and bronchitis, the FDA asked, what is its benefit:risk ratio?
Predictably, the committee found that, since efficacy had not been proven, and telithromycin was associated with some toxicity, the benefit:risk ratio was poor (some might say 0). A few of us at the meeting, including the FDA itself, pointed out that generic antibiotics, like augmentin and others, had never been subjected to placebo controlled or superiority trials in the same indications and that they too had toxicities not unlike those associated with telithromycin. We asked what the FDA would do about that situation. The response from Dr. Jenkins of the FDA is quoted below from the meeting transcript:

DR. JENKINS: We definitely hear your concern about a fair level playing field for all the antimicrobials that have the indication and we will take that back and put on our thinking caps about what approaches there might be to get the level of evidence that we would like to have today for all of those products. But I can assure you it will be a very complex endeavor and maybe similar to me trying to climb Dinali.

Clearly, if one is to buy the arguments as posed by the FDA and its Advisory Committee in 2006, since in terms of numbers of prescriptions those for generic antibiotics far outweigh those for brand antibiotics, the potential risk for Americans is that much greater for generics. So, either the FDA does not really think that these drugs have no proven efficacy, or they do not believe their own data on the toxicity of generic antibiotics or they just have not yet been able to figure out what to do with their homemade conundrum.

To me, the solution is simpler. To require placebo-controlled trials dooms us to never having another new antibiotic for otitis, sinusitis or bronchitis. This is another example of the FDA requiring infeasible trials for approval. The FDA needs to put on their thinking caps about new trial designs that are feasible. Then, they need to ask whether the generics have come to an acceptable level of proof based on a more feasible approach. If the answer is “yes” then we’re done. If not, then marketing approval for the generics for these indications must be denied until they achieve the same level of proof the FDA is requiring for new antibiotics.

I have been trying to follow-up with the FDA since it has now been three years since they started their attempt on “Dinali.” So far, no one seems to be home.