Just before the November 4th AMDAC on solithromycin, we are being treated to a very fine piece of study write-up. Of course, we are speaking of the File paper on the SOLITAIRE-IV trial, and what a fine piece of crafty penmanship it is, if you haven’t noticed.
Just as a reminder: the SOLITAIRE-ORAL study was like a walk in the park: equal efficacy (make this: non-inferiority) and safety were shown between oral solithromycin and moxifloxacin in this CABP study of mostly ambulatory patients.
So far so good. In SOLITAIRE-IV, the File study, more sick CABP patients were enrolled in order to study IV/PO solithromycin and an IV to PO step-down concept. The study was conducted following the FDA CABP Guideline document to the letter, and the publication spares us none of the gruesome details. According to FDA wisdom, Early Clinical Response (ECR) around Day 3 is the primary efficacy parameter between test drug and control.
Solithromycin did not disappoint us: the IV/PO regimen showed very similar efficacy at that early time point to the comparator, moxifloxacin IV/PO. However, discordant results were seen at early f/u, the key time point for the European Agency which requires that clinical evaluation at short-term f/u (CE at SFU) be within a 10% NI delta.
The SOLITAIRE-IV authors carefully avoided providing this data, instead presenting only their post-hoc analysis of a “modified CE population” in Table 4 of the article. In this “modified CE population” analysis, a favorable outcome was seen in 87.6% for solithromycin vs 92.5% for moxifloxacin, CI [-9.4, -0.5].
One can bet the farm on this: without post-hoc data dredging, the CI would have been outside the -10% margin!
When a primary endpoint – prospectively specified in the protocol – is not detailed in the final article, when a post-hoc analysis is presented instead, one is dealing with obfuscation at the best, and with data manipulation at the worst. Tom File, the main author, should not have let this happen, and neither should the reviewers of this paper and editors of the Clinical Infectious Diseases Journal.
Such selective reporting is appalling, even during this election season when one is getting used to half-truths and spin-doctoring.
Of course, Tom File did not himself write this paper, ghost writers most likely wrote it who received their instructions, corrections and suggestions for ‘proper wording’ from company stakeholders. So much for the caliber of peer review and the publication process.
Looking at the data as presented, it seemed that SOLITAIRE-IV had met FDA criteria for efficacy but failed to meet EMA criteria. Now the FDA has a dilemma: prior FDA Guidelines relied on clinical assessment of efficacy after treatment, just as the current EMA approach. This was called the Test-of-Cure (TOC) visit, usually done some 14 days after EOT just like the SFU. Will the FDA turn a blind eye and ignore the post-treatment results?
Admittedly, making a decision on drug efficacy on Day 3 and not at the end-of-therapy is an abstruse concept hard to comprehend. It is a regulatory world, after all.
Will solithromycin become approved in the US but rejected in the EU? This is not likely but still possible. The drug passed the FDA litmus test of efficacy at the ECR timepoint; hence, we could predict that FDA approval is highly likely. EMA may have a tougher time with the data and not buy into Cempra’s post-hoc analysis.
Not always have US and EU regulators come to the same conclusions after looking at the same data. For instance, tigecycline/Tygacil was approved for CABP in the US, but not in Europe. Xigris labeling was quite different in the US than in the EU.
Now it appears that FDA’s new efficacy criteria have become less stringent than the older ones
Just think, what if SOLITAIRE-IV had failed to show NI at the Day 3 timepoint but, nevertheless, shown NI at early follow-up? What happens if a study meets EMA criteria for efficacy but not FDA? There is no need to speculate; FDA will likely tell us this was a review issue. Good answer!
There is another wrinkle in the SOLITAIRE-IV trial, and this one has to do with the inordinately high incidence of IV site reactions. A whopping 31.3% of patients on solithromycin IV had infusion site reactions! How fast did they infuse this otherwise well-tolerated 4th generation macrolide to create such a mess? Was it over 60 or 90 minutes? The article does not mention this. We have to dig deeper – a topic for another blog.
There was yet another peculiar aspect to this study, also related to dosing. As patients were switched from IV to PO solithromycin, the first oral dose was a loading dose of 800 mg. Folks, we need your help with this: Since when is a loading dose required in the middle of a treatment course? Is this also necessary with other antibiotics that have a long half-life? Have we missed that lecture – where have we been all this time? Please, enlighten us!
NOTE added in Proof (11/12/16):
Thankfully, the FDA AMDAC on 11/4/16 provided the missing data; the delta for the CE population at the SFU timepoint was 12.4%. This is not much off the mark and – as a single aberration – was not a big issue for FDA. We predict it will not be a big issue for EMA, either. Quite an unnecessary attempt to cover up in the File publication, we feel.
T File. SOLITAIRE-IV. IV Solithromycin vs Moxifloxacin in CABP. Clin Infect Dis 2016; 63: 1007
AMDAC Antimicrobial Drug Advisory Committee
ECR Early Clinical Response (Day 3 of Therapy)
SFU Short-term Follow-Up (Day 12-17 after EOT in this trial)
EOT End-of-therapy timepoint
TOC Test-of-Cure (usually 14 days after EOT)
CE population clinically evaluable for efficacy assessment