- 8:00 AM EST: Meeting Opens
- 8:20-9:50 – Hemispherx Presents
- 10:20-12:00 – FDA Presents
- 1-2:30 Public Presents
- 3-5:00 – The Committee answers the FDA’s questions and then votes.
The FDA released its background report and its list of committee members and, if you want to compare the two, the Agency was even-handed; the report was highly critical and the committee members they picked boded well for Ampligen chances..
Harshly Critical Analysis
We knew FDA Advisory Meetings can be intense, rough environments. With the FDA background report on Ampligen you can count that as confirmed. This is not going to be easy for Hemispherx..
The FDA analysis of the Ampligen trials, focusing on methodological issues, mistakes and what it believes are inadequate efficacy and safety data does not, on the face of it, appear to bode well for Ampligen and the market agreed with the stock shaving off half its value the day the report was released.
FDA reports do tend to highlight negatives, not positives which makes sense given the way the meetings are set up; the sponsoring drug company presents its data in its best light, then the FDA does its best to poke holes in it, and then the independent review committee decides.
In the meantime be prepared for a difficult hearing. These FDA Advisory hearings have been described as ‘blood baths’ from which few companies and drugs escape unscathed.
The long and rather tattered history of Ampligen at the FDA since 1988 is outlined. The first interagency transfer to a new department occurred in 1992. The long gap between the first proof of concept trial (1990-1991) and the confirmatory trial (1998-2004) probably demonstrated how difficult it was for Hemispherx to raise the money needed to carry out the second trial.
Ampligen’s New Drug Application (NDA) submitted on Oct, 2007 was refused filing because it was missing study reports, ECG and laboratory data, case report forms, ‘dose ranging’, ‘inconsistent statistical plans’, clinical pharmacology data, carcinogenicity data and
In Nov 2009, however, citing inadequate effectiveness and safety data, drug-drug interaction data, carcinogenicity data, anti-drug antibody information and inadequate analytical methods the FDA told Hemispherx, among other things, to conduct a new clinical trial; something which was beyond Hemispherx’s means.
“Your two main studies (AMP-502 and AMP-516) do not provide adequate evidence of effectiveness or safety.”
In 2012, after being moved to another branch of the FDA, Hemispherx requested to be allowed to re-analyze the data from the second, larger, confirmatory trial. The agency was leery about this approach stating “It would be unusual for this type of data to provide adequate evidence of efficacy.” but allowed the process to continue forward stating the efficacy issues would be addressed at the final review.
The FDA will be relying on evidence from two trials; AMP-502 and AMP-516. (The agency is, unfortunately, not using confirmatory data generated from patients who continued on the drug after the trial because that data was not placebo-controlled.)
First Trial (1990-1991) – The first study (502) involving 92 patients demonstrated significant increases in functioning (Karnofsky Performance Score (KPA)) and ETT (time on treadmill).
Problems had occurred. After finding that the original exercise protocol was too difficult, however, Hemispherx changed the protocol and eliminated the first seven patients from the analysis. ( If those patients are included in the analysis the significantly positive effects disappeared. )
The trial was originally supposed to last for a year but after an interim analysis it was stopped after 24 weeks. The FDA drew attention to the fact that the statistical analysis plan was written and dated 16 years after the completion of the study and later noted ‘multiple protocol violations’ including patients restarting their other medications during the trial had occurred.
Second Trial (1998-2004) – The second larger trial (516) designed to confirm the first trials results failed to show significant increases in self-reported functioning or ETT. Recently Hemispherx re-analyzed the data using patients with lower disease duration (1-10 years) and dichotomous branching techniques and received significantly positive ETT results but still no significant changes in self-reported functionality, daily living, SF-36 or symptom scores.
Hemispherx found that the percent of patients reaching 25% and 50% increases from baseline was significantly increased in patients receiving Ampligen and that patients with lower duration illness (<10 years) were more likely to benefit. Despite reports in the literature that lower duration patients (< 2 years) do tend to respond better the FDA found ‘no scientific rationale’ for the lower duration analysis and advised caution in interpreting them.
Several issues were reported including using median KPS values for some patients and minimum median values for others. Hemispherx also chose to use any response (ie the best response recorded) during the duration of the trial for its endpoint instead of set responses at the end of the trial.
The re-analysis was helpful but the FDA questioned whether these types of re-analyses are suitable for determining efficacy suggesting they are better suited for ‘hypothesis testing’.
Due to the nature of analyses conducted after data unblinding, the Agency generally considers post-hoc analyses to be hypothesis generating rather than forming the primary basis of efficacy for an application.
Hemispherx was caught in the first study with an exercise protocol that was apparently wiping the participants out. In the second , larger ‘confirmatory’ study Hemispherx had to analyze the results differently to get positive results in the exercise testing but not significantly positive results in self-report tests. Changes in protocols and analysis are often red flags to reviewers and questions were raised why a protocol and analysis would work in one trial and not the other.
It was rough stuff but a Hemispherx rep appeared unflustered and confident that the company was ready to respond to the FDA’s report.
Tough Subject – With its subsets and its vague definition ME/CFS is a difficult field to test any drug in and drugs that practitioners say have positive results in some patients have failed to achieve them in placebo-controlled trials. If the disorder is larded with subsets then the results of any major trial of a drug that works probably should appear like Hemispherx’s did; that is, it should have weakly significant results because of the patients in the trial who have a different type of disorder altogether. We should also know of people who had excellent results (we do) and doctors who prescribe the medication to targeted patients should have good results (they do).
The FDA must realize that it’s facing an unusual challenge in a disorder that is so poorly defined. The background report did not take into account the possibility of subsets watering down the study results but the Committee is made of several members who are well aware of this problem. Dr. Unger of the CDC, for instance, has publicly stated she is sure this is true and is engaged in a study that should help ferret them out.
Indeed a recent analysis suggested the subset problem meant Hemispherx started its trial with one of its hands tied behind its back. Throw in the fact that it’s a small company with limited resources and you probably get what we see; midstream changes that attempt to account for new information and re-analyses to target possible subgroups.
Ampligen’s trials had no way to account for this meta-population problem beyond looking for interesting trends within the data set. And that is what they did… No one seems to know what an appropriate biomarker for the general CFS population is, but the FDA is holding Ampligen up to an unrealistic standard nonetheless. In the end, HEB did find statistical support that Ampligen provides a biologically meaningful impact on the lives of at least some patients with CFS.
The problems facing Ampligen-or any putative therapy for CFS for that matter, is that the underlying cause(s) are unknown. As a result, there are quite possibly distinct sub-groups within any potential CFS study population with markedly distinct pathophysiologies, and hence
Hemispherx has reported that tens of thousands of doses of Ampligen have been given without issue but the FDA will rely on the results from ‘only’ 567 patients of which ‘controlled data’ are available for only 162. The FDA stated that for drugs intended for long-term treatment they would like at least 1,500 people exposed to the drug for short term periods and 100 for over a year.
Here, too, the FDA cited numerous issues including serious adverse events not reported, marked laboratory abnormalities not reported, miscoding of adverse events and reasons for patient discontinuation of the trial and incomplete/misleading data presentation. Some serious adverse events that resulted in hospitalization were not reported. The discrepancies were enough for the FDA to call into question the data from all the trials.
On the other hand, the side effects that showed up more in patients receiving Ampligen (flu-like symptom, chills, vasodilatation, shortness of breath ) were not particularly significant. Other more serious side effects did occur but not more significantly in patients receiving Ampligen but the Agency expressed concerns about them.
Overall, the Agency has concerns regarding the reliability of the safety database for Ampligen, based upon lack of appropriate documentation and reporting. While review of the safety database for Ampligen suggests that the drug induces a systemic inflammatory reaction with a number of serious events of concern, the lack of reliable reporting prevents any accurate assessment of frequency. The key safety issue for discussion at the advisory committee meeting is whether conclusions can be drawn from the data as it stands …
Hemispherx will explain why the safety data is sufficient and why oversights, omissions, etc. occurred and why they shouldn’t bear negatively on the review.
COMMITTEE MAKEUP – the Good News
Taking the committee makeup into account one analyst predicted a close but positive vote. At least six people have history with chronic fatigue syndrome and I would be very surprised if Komaroff, Marshall weren’t strong yes votes and the others strongly leaning towards yes. If those six vote yes they’re two votes away from a tie vote and three from a majority vote. For me, I think Marshall can be very persuasive once he gets going J.
- Dr. Philip Bautista
- Dr. Robert Lahita – Rheumatologist – research focus: lupus; latest article – novel treatments for lupus
- Dr. Tuhina Neogi – Rheumatologist, Epidemiologist – research focus: osteoarthritis and gout
- Dr. Peter Peduzzi – Biostatistician
- Dr. Irwin Russell – Rheumatologist associated with Fibromyalgia Research and Consulting in San Antonio, Texas – research focus- fibromyalgia; one study on duloexetine
- Dr. Larry Borish – NIH grant recipient – allergic reactions in CFS
- Dr. Lenore Buckley
- Dr. Ralph D’Agostino (Ph.D)
- Dr. Jacqueline Gardner (Ph.D)
- Dr. Anthony Komaroff (MD) – ME/CFS expert
- Dr. Gailen Marshall (MD, Ph.D) – CFSAC representative
- Alaine Perry (MPH) – Patient representative – CFSAC representative
- Dr. Mathew Rudorfer (MD)
- Dr. Sean Hennessey (Ph.D)
- Dr. Elizabeth Unger (Ph.D) – Chief of CDC CFS research
- Dr. James Ware (Ph.D)
The harsh FDA report was produced by statisticians with little knowledge of the disorder but the FDA also produced a committee with a core of members with extensive knowledge of ME/CFS. The Committee vote is the key. Will Hemispherx and the patient advocates be able to convince the committee members to overlook a harsh FDA report? We’ll see.
The vote will come at the end of the day. The committee members will be asked to vote all at once so that the members are not influenced by the others votes.
The FDA will take everything into account and release its decision on Feb 2nd. They could approve the drug outright (not likely to happen), approve it with conditions, ask for more studies or just slam the door on the drug.