The FDA Oncologic Drugs Advisory Committee, in an 13-4 vote, recommended against approval of retifanlimab, a PD-1 inhibitor, for the treatment of squamous carcinoma of the anal canal.
The agent’s sponsor, Incyte Corp., was seeking an accelerated approval based on an ongoing, open-label, single-arm trial that accrued 94 patients with locally advanced or metastatic SCAC. The committee recommended deferring the decision until completion of a randomized trial, which is expected in 2025.
At the June 24 ODAC, the company was seeking an accelerated approval “for the treatment of patients with locally advanced or metastatic squamous carcinoma of the anal canal who have progressed on or are intolerant of platinum-based chemotherapy.”
If approved, Incyte’s retifanlimab would become the seventh PD-1/PD-L1 drug on the market.
FDA has been taking a hard look at “dangling” indications of this class of drugs. The term, coined by the agency’s cancer czar Richard Pazdur, refers to “accelerated approval indications where a required trial did not confirm benefit—hence, this indication is ‘dangling’ between an accelerated approval status and market withdrawal.”
In April, at a three-day meeting, ODAC reviewed five such indications (The Cancer Letter, April 30, 2021). Over preceding months, FDA convinced four drug companies in as many months to voluntarily withdraw four dangling indications. Press releases announcing the withdrawals said the indications were being pulled “in consultation with the agency” (The Cancer Letter, March 12, March 5, 2021).
The prevalence of dangling indications was thus reduced by 40%. The remaining 60%, presumably those that weren’t quite so open-and-shut, went to ODAC. (A date with ODAC had to be the or-else in the agency’s behind-the-scenes persuasion tour.)
By taking Incyte’s retifanlimab before ODAC, the agency was asking for a discussion of settings where single-arm trials are appropriate as well as about the value of demonstrating a small improvement in a response rate in such a trial.
“After three days of lengthy discussions, it appears that low response rate, even when some of these responses are durable, do not always translate into clinical benefit when larger numbers of patients are studied in clinical trials,” May Tun Saung, a clinical reviewer, said at the ODAC meeting June 24.
In Incyte’s trial, called POD1UM-202, 13 of the 94 patients in the trial had demonstrated objective response per Independent Central Review.
There is no reason why people cannot do randomized studies to get their drugs approved. And the single-arm trial is not the only way that a drug can be approved. We’ve advocated this multiple times to companies.Richard Pazdur
The FDA staff pointed out that it’s unclear whether this 14% response rate can be held as reasonably likely to predict clinical benefit. The ORR per ICR is 14% (95% confidence interval [CI]: 8, 22); median estimated duration of response (DoR) is 9.5 months (95% CI: 4.4, not estimable).
Making things worse, about half of the patients who demonstrated improved ORR in the study had limited follow-up for durability of response, meeting documents show. ORR hasn’t been shown to be a predictor of either overall survivors or progression-free survivors for immune checkpoint inhibitors.
The agency threw the application to ODAC to get advice on whether the request for accelerated approval should be deferred pending the results of a randomized, which is ongoing. The results are expected in about four years, the company said.
Also, the sponsor was presenting the results of a small study of patients who were not representative of the SCAC population, FDA said. Few of the POD1UM-202 patients were HIV-positive or were members of racial minority groups. About 8.1% of SCAC patients have HIV.
Incyte has filed an application despite words of caution it received at a pre-BLA meeting last September. The meeting, summarized by the agency, didn’t seem encouraging:
- ORR magnitude was modest in POD1UM-202 and less than the target ORR of 25% that was proposed by lncyte in POD1UM-202; thus, it was unclear if the ORR was reasonably likely to predict clinical benefit,
- DoR data was limited – only 7 of the 13 responders had DoR >6 months,
- The BLA would be a stronger application if it were supported by positive results from the POD1UM-303/lnterAACT 2 clinical trial, and
- If the BLA is submitted with results only from the POD1UM-202 clinical trial, FDA might elect to discuss the application in an ODAC.
“Although FDA cannot discuss the follow-up with respect to the [April] advisory committee meetings held in April, an important lesson is that when voting can maintain an indication, members of the advisory committee considered if alternative confirmatory trials were being conducted and the timing of when these trial results are expected,” Saung said at the ODAC meeting. “The FDA would like to highlight that in today’s meeting, we will discuss an application with a single confirmatory trial that has enrolled only 28 patients as of May 25, 2021, which is only 9% of the plan trial population.”
Though Incyte inferred delay in disease progression from the data, FDA said single-arm trials cannot produce such data.
“The problem with single-arm trials is you don’t get a great risk/benefit assessment, because you don’t have a control arm,” said Steven Lemery, acting director of the Division of Oncology 3 at the Office of Oncologic Diseases. “So, we have to carefully think about when single-arm trials shouldn’t use in, perhaps when they shouldn’t.”
Pazdur, director of the Oncology Center of Excellence and acting director of the Office of Oncologic Diseases, said the agency has “profound concerns” about continued reliance on single-arm studies as a basis for accelerated approval of PD-1/PD-L1 drugs.
There’s an adage: “Those who don’t learn from history are destined to repeat it.” And I think we had a very painful discourse over the past ODAC, with many trials that had relatively low response rates not demonstrating clinical benefit. And we really have to reassessed this.
And that’s why we’re bringing this to this ODAC meeting, and we’d like some discussion on this. There is no reason why people cannot do randomized studies to get their drugs approved. And the single-arm trial is not the only way that a drug can be approved.
We’ve advocated this multiple times to companies. This data was known many years ago, of the activity of this drug, and a randomized trial could have been initiated earlier, perhaps even in an earlier disease setting in anal canal cancer.
So, there are profound concerns here of whether continuing this practice for this class of drugs—and I want to make it quite clear—is a reasonable registration strategy. Here. again, there are areas where single-arm trials make sense.
These may include where there’s very high response rates for some of the targeted therapies, and we’ve given actually full approval on the basis of response rates, but there’s no reason why we only have to do single-arm trial for many of these diseases and then look at randomized trials.
One of the options that we would have had here is to do a randomized trial and take a look in interim analysis for response rates, and have a continuation of the trial to demonstrate clinical benefit. And we would actually have had a randomized trial going on here.
I’d also like to point out for the committee, since many of you may not be familiar. When we take a look at single-arm trials, we are only taking a look at response rates.
We cannot make any inferences regarding stable disease, because it may reflect the natural history of the patients that were enrolled in the studies, nor can we make any claims regarding time to progression, or overall survival.
So, although that was presented in the sponsor’s presentation, from a regulatory point of view, we would not be taking a look at these endpoints of disease stabilization, or time to progression, or overall survival.
These need to be demonstrated in a randomized setting.
Mark Cornfeld, Incyte’s vice president for immune-oncology drug development, disagreed with Pazdur:
We actually agree with FDA that that’s not all trials will be confirmed and should not be considered by way, but it’s the biology that’s the key here.
And if you take a look at those, the list of the trials, where there have been concerns, and if I did not participate in the April ODAC, but we certainly followed it with interest and all of these trials were in indications other than a squamous tumor, and specifically none of them were in HPV-driven disease.
And if you look at the trials specifically in HPV-driven malignancy, which is a very unique biology, and remember biology is key here, the results are consistently predictive of survival.
If we limit our discussion to the HPV-driven cancer biology, which is unique, and which is what we’re talking about here today, since all of anal cancer is an HPV-driven cancer, these very low response rates have consistently predicted for survival benefit.
There are no approved second-line treatments for SCAC, but other PD-1/PD-L1 agents are being used off-label for this rare cancer.