Study design, problematic statistical analyses, and risk management appeared to blunt the enthusiasm of FDA staff reviewing the investigational agent palovarotene for fibrodysplasia ossificans progressiva (FOP), an ultra-rare disease that causes bone to replace soft tissue.
In briefing documents released ahead of a Wednesday meeting of the Endocrinologic and Metabolic Drugs Advisory Committee, FDA staff expressed concerns not only about the statistical analyses drugmaker Ipsen employed, but the possibility that palovarotene, an oral selective retinoic acid receptor gamma agonist (retinoid), may actually increase the pathologic flare-ups that precipitate abnormal bone formation in people with FOP.
But despite those issues, the agency reviewers didn't quite shut the door on the tangled evidence for the treatment.
There's no treatment for FOP, which affects fewer than 1,000 people worldwide. It's caused by a mutation in ALK2, the gene that helps guide bone growth, including normal growth-related ossification of cartilage. Beginning in early childhood, patients experience flare-ups -- inflammatory swellings that precede abnormal bone growth. These can be triggered by any injury, even an injection, but also occur spontaneously. Eventually, ribbons, knobs, and sheets of bone replace sections of muscle and connective tissue, deforming and locking joints. Bone can't be surgically removed because that triggers the inflammatory response and more ossification. Patients are typically almost immobile by their mid-20s, have trouble breathing and eating, and usually die in their 50s.
According to the drugmaker, palovarotene works in two ways. It suppresses both the inflammation that causes soft tissue breakdown and the cellular signaling that sends chondrocytes to replace that soft tissue with mature bone.
The proposed indication is for the prevention of heterotopic ossification (new abnormal bone) in patients with FOP, both for adults and children (ages 8 years and older for girls; 10 and up for boys). The proposed dose is 5 mg daily for chronic treatment and 20 mg daily for 4 weeks during a flare-up, followed by 10 mg daily for another 8 weeks.
But the data package of one natural history study (NHS), a multi-armed phase II trial, and a single-arm phase III trial, contained hints that people taking the drug might be more prone to flare-ups than those taking placebo, especially with increasing dosages. Across the NHS and all phase II arms, new flare-ups occurred within 12 weeks in 23% of untreated flares, 30% of placebo-treated flares, 22% of flares treated with palovarotene 5/2.5 mg, 37% of those treated with 10/5 mg, and 36% of those treated with 20/10 mg.
However, FDA staff noted, "there is uncertainty as to whether palovarotene may in some cases trigger flare-ups, or symptoms that could mimic flare-ups, and if so, whether the development of new [bone] over longer periods of treatment ... could be worse than without palovarotene treatment."
The reviewers also questioned the efficacy data and analysis methods. Main support for the proposed indication derives from the single-arm pivotal 24-month phase III study, MOVE, which enrolled 107 patients. All received the 5 mg daily chronic dose with the episodic 20/10 mg dose during and after a flare-up. They were compared to 39 untreated patients from the 3-year long NHS. The primary endpoint was annualized new heterotopic ossification volume.
The study failed a futility analysis using a prespecified statistical method. After the data monitoring committee (DMC) stopped the study, the drugmaker unblinded the data and conducted a number of secondary post-hoc analyses, "which appeared to show evidence of benefit for palovarotene," FDA staff wrote. The DMC, however, determined that this new and "unexpectedly large" discrepancy muddied the picture even further.
"The DMC concluded that 'the dilemma created by these highly disparate results precludes a confident conclusion about futility,'" according to the briefing document.
While external controls are typically frowned upon in pivotal clinical trials, FOP is such a rare disease that it's very difficult to find and enroll patients, making the NHS an acceptable option. Ipsen conducted enough analyses to control for selection bias and confounding factors, the agency said.
According to the reviewers, the prespecified primary analysis "may not have been the appropriate method for analyzing the primary endpoint." Given these limitations, "we think it is reasonable to consider alternative more appropriate analyses to assess evidence of efficacy," wrote FDA staff.
It's palovarotene's fourth time at bat in the U.S. In 2021, the company received a priority review, but withdrew its new drug application after the FDA asked for additional analyses of the MOVE study. The drug was then up for committee review last October but that meeting along with several other advisory committee sessions, was rescheduled.
In December, however, the FDA rejected the drug for approval. At that time, the company didn't elaborate on why, saying only that the complete response letter didn't request additional efficacy or safety data. The drugmaker apparently addressed whatever deficiencies the letter noted.
Palovarotene fared better in Canada, where it was approved in January under the trade name Sohonos. But last week, the European Medicines Agency decided not to approve the drug.
On Wednesday, committee members will be asked to discuss the post-hoc data, the issue surrounding the flare-ups, and vote on whether the MOVE study supports palovarotene's efficacy, and whether the retinoid's benefits outweigh its risks.
The FDA is not required to follow its advisors' recommendations, but it usually does.