Interludes of classical music. Little kids talking in the background. Unmuted mics as panelists multitask. Gurgles of “underwater” sound. Periods of silence as speakers forget to unmute. And then the technical problems – lots of them.
Such are the challenges of addressing a virtual FDA advisory committee in the time of COVID-19.
“This is an important, critical day for the sponsor,” Steven Romano, executive vice president and chief scientific officer for Mallinckrodt plc, said of going before any adcom. Doing it virtually calls for a lot more sponsor preparation for what already can be a nerve-wracking experience.
Last week, Mallinckrodt and Glaxosmithkline plc were the first sponsors to have virtual adcoms in which the panelists voted on recommending FDA approval for a drug or device candidate. With COVID-19 cases surging, they likely will not be the last – even though the FDA has yet to add other product-oriented adcoms to its calendar. Both meetings could serve as learning experiences for other sponsors that may face a virtual adcom as the pandemic drags on.
The GSK meeting before the Oncologic Drugs Advisory Committee (ODAC) got off to a rocky start with technical difficulties shutting down the meeting less than an hour into it. Because of the nearly 1.5-hour delay in getting the virtual meeting back up, the committee decided to take an unscheduled lunch break, which resulted in the meeting wrapping up more than two hours after its planned 1:30 p.m. ET adjournment. Consequently, two panelists had to sign off before the 12-0 vote recommending approval.
That’s discouraging for a sponsor, Romano told BioWorld, as sponsors need to hear from all the expert panelists. It’s not so much the vote that’s important; it’s what comes after the vote when all the panelists are asked to explain their vote. That rationale often includes recommendations for labeling, risk mitigation strategies and further studies.
Since an advisory committee typically consists of academic and clinical experts from various fields, missing the voice of a particular expert can make a difference in the vote, as well as in the advice for going forward.
In the case of Mallinckrodt’s meeting July 15 with the Cardiovascular and Renal Drugs Advisory Committee, the vote was 8-7 in recommending approval for terlipressin, which has been in U.S. development for years as a treatment for hepatorenal syndrome type 1 (HRS-1). The two biostatisticians and three nephrologists on the panel voted against approval, while five of the committee’s six cardiologists voted for it. Had technical difficulties made it impossible for some of those experts to vote, the outcome could have been much different.
Anticipating what can go wrong
Recognizing that technical problems are to be expected with a virtual meeting that has so much riding on it, sponsors need to do a lot more preparation for an online adcom. In addition to putting together the presentation of clinical trial results and their significance, lining up presenters and key opinion leaders (KOL), practicing the presentation, and anticipating and assembling all the data that the committee may request, a sponsor also must think about the virtual technology and plan workarounds and ironclad backups should anything go wrong.
A virtual adcom “is not ideal,” Romano acknowledged. To prepare, Mallinckrodt tested the technology the FDA was using beforehand so its tech staff could iron out problems and become comfortable with it. The company also had an agreement with the FDA on how to handle the situation should it lose audio on its end. And the day of the adcom, Mallinckrodt had tech people across half a floor to support the company experts and staff participating in the meeting.
The challenges aren’t just technical, Romano said. There also are logistical issues imposed by the pandemic. Given COVID-19 restrictions, some of the company presenters and KOLs had to participate from other locations. Mallinckrodt had backup speakers online in case one of the scheduled speakers lost connection.
Another virtual “best practice” Mallinckrodt utilized was having a point person – Khurram Jamil, vice president for clinical research in the company’s hepatology critical care division – immediately respond to each panelist's questions and then ask a specific KOL to address the issues raised. That avoided awkward pauses or people speaking over each other.
Fortunately, the technical glitches that delayed the ODAC meeting the previous day didn’t recur Wednesday. But with panelists signing in from home or their individual offices, the Mallinckrodt adcom included the comic relief that’s becoming par for the course for virtual meetings – a small child speaking in the background, an unmuted mic picking up a conversation between a panelist and his colleague about the scheduling of a journal meeting and even a musical interlude. There also were moments of silence when panelists couldn’t unmute themselves.
Another glitch occurred during the public hearing session of both adcoms last week. Even though some of the people speaking had submitted slides, the slides couldn’t be shared – a fact that had some of the speakers scrambling to improvise their rehearsed comments.
But of all the challenges, Romano said what he missed most about an in-person adcom was the fluid exchange among the panelists. Since the members come together irregularly, the meetings offer an opportunity for discussing issues across specialties. Romano said he found the usual level of discussion was stymied somewhat in the virtual setting because of the technical complications and not having everyone in the same room.
The virtual nature also made it impossible for him and others speaking for Mallinckrodt to signal that they had something to add to the discussion. For instance, part of Wednesday’s discussion focused on serious adverse events (SAEs) in the CONFIRM trial involving respiratory failure. In looking at the data, Mallinckrodt raised the possibility that those SAEs, which hadn’t been noted in two previous trials conducted years ago, may be related to a more recent, dramatic increase in the use of albumin in treating patients with severe liver disease. Romano said he would have liked to have been able to engage with some of the panelists about that issue.
As it was, in explaining his yes vote, panelist David Assis, an assistant professor at the Yale University School of Medicine, agreed that the increased use of albumin may have played a role in the respiratory problems encountered in the CONFIRM trial. Hepatologists need to look at whether they are overusing albumin, he said, as it could affect other drug candidates in the hepatology space. He stressed that, if the drug is approved, any risk mitigation plan (RMP) should be used to educate hepatologists, as well as patients and their caregivers.
Romano said the RMP Mallinckrodt is proposing would serve as a learning tool for the use of albumin, as well as a risk management tool for terlipressin.
While virtual adcoms can inhibit some discussion and require a lot more contingency planning, there is one silver lining Romano noted. The virtual aspect enables a sponsor to do some backroom maneuvering during the meeting, if need be, to find requested data, present it in a requested form or change strategy.