Police profiling based on race, sex or gender is a highly controversial practice. FDA doing the same, not so much.
Profiling based on similarities to other drugs in the same category is nothing new. Class labeling can be expected for the next new B-lactam or quinolone regardless of whether during the drug’s development an anaphylactic reaction or tendinitis occurs.
AMDAC and the FDA meeting documents led us to think about the value and consequences of compound-class related profiling. In the case of solithromycin it is hard to overcome this stigma of chemical heritage or similarities in MoA. The chemical structures of solithromycin and telithromycin are almost identical.
Solithromycin’s efficacy in CABP was beyond question even before the AMDAC; indeed, few had issues with the pivotal trials or the study results for PO and IV/PO sequential therapy. It is a truism that efficacy can be established quite reliably in Phase 2 and 3 trials with very few exceptions. However, safety is an altogether different issue.
Everyone has his or her favorite story of a surprise safety signal emerging only after a drug has been marketed. We like to remind newcomers to drug safety of the temafloxacin/Omniflox disaster (HUS-like syndrome [3]) or the less dramatic gatifloxacin / Tequin AEs (sporadic hypo- and hyperglycemia). For FDA reviewers, one of the most traumatic memories in recent history is telithromycin / Ketek and hepatotoxicity.
Telithromycin was the 800-pound gorilla in the room during the AMDAC meeting on Nov. 4th. The FDA reviewers and expert hepatologists homed in on solithromycin’s hepatic safety record and provided insights hard to find in the public domain. They stated unequivocally that there was a strong hepatotoxicity signal in the solithromycin study program. They pointed out that solithromycin was already worse than telithromycin at the same point in development.
With little enthusiasm for the drug in the room and FDA telling panelists that solithromycin looks like telithromycin, just worse, we were treated to a hefty dose of profiling. Ketek still looms large; its shadow was felt everywhere.
The data analysis, esp. the expert evaluation provided by Dr. Mark Avigan, a co-author of the FDA Ketek publication in 2009 [5], was thorough and authoritative; it provided insights into the drug’s hepatotoxic profile, and put it into perspective, as only the FDA can do. It really helps to be in possession of all the data.
Hepatic toxicity was not much of an issue in any particular solithromycin study when looked at in isolation. Both the PO solithromycin CABP [1] and the IV/PO solithromycin CABP study publications [2] mention liver function abnormalities in detail. In all fairness, there were no cases of Hy’s law, no irreversible cases of liver toxicity requiring prolonged hospitalization or cases of liver failure requiring transplant and/or resulting in death. In response to this, FDA would say this: “If it has not happened yet, it will! Don’t you see the writing on the wall?”
Indeed, data from preclinical toxicology, Phase 1 in volunteers and in Phase 2/3 patient trials, all point to the liver as the target organ for injury. The small NASH [3] study with longer term exposure to solithromycin added oil to the fire. From FDA documents we learned that the dose in this study had to be down-titrated repeatedly because of hepatic AEs.[4]
There are lots of lessons to be learned here, as always. Firstly, individual study publications really do not provide the most reliable information on new drugs, neither for efficacy nor for safety. All too often, information provided is subtly biased. Secondly, safety assessment is a much more difficult task than people generally admit; it requires diligence, perspective, knowledge and a good look at the entirety of data.
We can thank the FDA reviewers for doing a really thorough job, much better than anyone else
Still, there is some unease in our minds with this approach to new drugs: How often are we wrong when we apply a profile and do we care to find out about it? Of course, one can always stay on the safe side and never approve a drug that is not super-clean. Once profiled, how do we get out from under? If a drug after years on the market does not follow the class profile, is FDA equally active then in removing an inappropriate Warning / Precaution or restriction from the label? We know of no example in which this happened.
The risk associated with solithromycin, small as it currently appears, can be reduced further. The PO form of solithromycin is clearly safer, especially with short-treatment courses. Hence, duration of dosing in general, and that of IV dosing in particular, could be kept as short as necessary. In addition, FDA agrees with us that doubling of the oral dose when transitioning from IV to PO is unnecessary.
One could have profiled the drug’s efficacy easily just by using microbiological data. Still the Agency would not have approved the drug based on surrogate efficacy data alone; they do not trust the data derived from animal models for efficacy or the consistent bacteriological response seen with earlier generations of erythromycin congeners. FDA and EMA insist on proving the obvious in lengthy, complicated, and costly Phase 3 efficacy studies. They do so because they are concerned about a possible exception to the rule, the rare outlier that does not work.
Cempra has characterized the safety profile of solithromycin and provided all necessary information for the safe use of this macrolide. We believe that monitoring of LFTs, already routine for most in-patients, should be added to the labeling of IV solithromycin. Post-marketing surveillance would further provide information on frequency and severity of liver toxicity, the impact of renal impairment or co-medications. In contrast, a post-marketing study would take years to finish.
With prudent restrictive labeling and post-marketing surveillance in place, we have enough monitoring already in place for safe use and will learn enough from reports to the AERS system.
To put matters in perspective, we have lived with the QT prolonging effects of IV erythromycin for a very, very long time. There is no worse offender than this drug in this regard, a drug which can prolong the QT interval up to 40 msec! Nevertheless, it is still on the market and we still use it extensively. Talking about safety, the playing field is anything but level.
References:
[1] D Oldach. AAC 2013; 57: 2526
[2] T File. CID 2016; 63:1007
[3] NASH = non-alcoholic steatohepatitis HUS = hemolytic-uremic syndrome
[4] clinicaltrials.gov mentions only that the solithromycin dose was reduced from 400 mg PO QD for 13 weeks to 200 mg PO QD for 1 week, followed by 200 mg PO TIW for 12; this change was entered into the database on Sept. 28, 2016.
[5] A Brinker. Telithromycin-Associated Hepatotoxicity: Clinical Spectrum and Causality Assessment of 42 Cases. Hepatology 49; 250, 2009