Philadelphia chromosome. In a bone marrow cell, defects on the second chromosome 22 cause chronic myelogenous leukemia. Credit: Addenbrookes Hospital / Science Source

US Food and Drug Administration (FDA)'s decision in late October 2013 to suspend marketing of Ariad's chronic myeloid leukemia (CML) drug Iclusig (ponatinib) sent shockwaves through the industry. Most clinicians weren't expecting a full-scale suspension, and neither apparently was the financial community, which sent the Cambridge, Massachusetts, biotech's shares down over 80%, forcing the company into a radical downsizing. But on December 20, the regulators gave Ariad the go-ahead to resume marketing, albeit to a narrower patient population, and with Iclusig's relaunch, Ariad's shares leapt up. What's already clear, though, is that as FDA continues to encourage early access to drugs covering diseases with few or no available treatments through its accelerated approval pathway, there will be more Iclusigs. And Ariad isn't the first—nor will it be the last—biotech forced to rapidly change gear in order to survive.

The FDA on October 30 requested Ariad to withdraw its drug due to an “increased frequency of blood clots and narrowing of blood vessels” since the drug's approval in December 2012. The agency declared that almost a quarter (24%) of patients in the phase 2 trial and almost half (48%) in the phase 1 trial had experienced “serious adverse vascular events,” including fatal and life-threatening heart attack, stroke and damaging loss of blood flow to the extremities, heart and brain. FDA also said it couldn't identify a safe dose level or exposure duration at that stage.

This was a damning indictment of a drug that had been accelerated through approval three months ahead of schedule and hailed as a major breakthrough in CML thanks in large part to its potent resistance profile. Iclusig is a multitargeted, small-molecule inhibitor. It is the most potent broadest range of mutations in the BCR-ABL kinase domain (the protein produced by the Philadelphia chromosome, a chromosomal abnormality associated with CML).

The news didn't come entirely out of the blue. Ariad's company release on October 9 said it was pausing patient enrollment in all Iclusig studies, and, nine days later, the biotech announced that owing to cardiovascular events in patients taking Iclusig, it was discontinuing a phase 3 trial pitting Iclusig against Novartis' well-established Gleevec (imatinib). And Iclusig has always carried a boxed warning about arterial thrombotic events.

But many treatments for a life-threatening condition carry boxed warnings, especially those approved under FDA's accelerated pathway. Cancer prescribers, and patients, are used to weighing up the possibility of side effects against a drug's benefit. And it wasn't as if Iclusig was a first-in-class treatment: there are four other tyrosine kinase inhibitors (TKIs) on the market for CML (Table 1). Each TKI has an individual side-effect profile, but none has—yet—thrown up any alarming safety signals.

FDA had initially determined that Iclusig's benefits outweighed its risks. Iclusig is the only tyrosine kinase inhibitor active against the T315I mutation, for instance (about 20% of all BCR-ABL mutations are T315I). Iclusig's label was broader than for just those patients with the T315I mutation, though: it was indicated for all adults with chronic, accelerated or blast-phase CML, resistant or intolerant to prior TKI therapy, and for those with Philadelphia chromosome–positive, acute lymphoblastic leukemia that is resistant or intolerant to prior TKI therapy. “Pulling ponatinib is most problematic for T315I patients. Virtually all others have other [treatment] options,” declares Peter Emanuel, director of the Winthrop P. Rockefeller Cancer Institute at the University of Arkansas for Medical Sciences in Little Rock. Ultimately, bone marrow transplant remains an option for some patients, too.

The fact that there are other treatments for CML likely influenced FDA's decision to pull Iclusig completely (the European Medicines Agency on November 22 only restricted the drug's use, though a further review announced December 6 may lead to additional changes). Indeed, Sprycel (dasatinib) and Tasigna (nilotinib) treat “most” of the mutations that develop as a result of Gleevec usage, according to Emanuel. But FDA may also have wanted to act firmly to avoid potential criticism of the accelerated pathway if further issues arose later on.

There are precedents for drug withdrawals, but not many. Even fewer return to market. The best-known come-back story is Cambridge, Massachusetts–based Biogen Idec's Tysabri (natalizumab) for multiple sclerosis. Tysabri was withdrawn in 2005 after reports connecting it to a rare viral disease that affects the brain and returned just over a year later with a boxed warning and risk management program. London-based AstraZeneca's lung cancer drug Iressa (gefitinib), restricted and ultimately withdrawn in the US, found a target subgroup among Asian patients with EGFR mutations.

So, was the agency too hasty in allowing Iclusig onto the market, and/or too generous with its label, given the limited data available at approval?

Yes, according to critics of the accelerated approval program, such as political scientist and Harvard University professor of government Daniel Carpenter. The current approval system is a “growing hodgepodge of exceptions to the rule of rigorous premarket review,” Carpenter told Reuters in October 2013. He pointed to a study published just before Iclusig's withdrawal showing that drugs approved in 2008 under FDA's accelerated program had been tested, on average, in less than one-fifth the number of patients that drugs approved normally are, with many safety questions remaining unanswered as a result (http://archinte.jamanetwork.com/article.aspx?articleID=1761917).

But some see Iclusig's withdrawal as proof that the system is working as it should. “What has occurred...is a clear demonstration that the accelerated approval pathway works,” noted Mikkael Sekeres, chair of FDA's Oncology Drugs Advisory Committee, on November 26 in Oncology Times. Emanuel—who follows hundreds of CML patients—also thinks FDA was right to withdraw Iclusig, given the data. Yet despite the withdrawal, “I don't think FDA needs to be more cautious when it comes to cancer,” he declares. “It's a balancing act. I don't think anyone could have foreseen that this TKI would have [such severe] cardiovascular side effects; nor should we jump to the conclusion and say all TKIs will have cardiovascular effects.”

The other problem is that cardiovascular side effects come in various shapes and sizes, and the methods used in cancer drug trials to screen for, and categorize, these events remain unstandardized. Indeed, Ariad claims that FDA, since approving the drug, changed its definition of a “serious adverse event.” At approval, it used “the standard US regulatory definition,” according to the company's chief medical officer and senior vice president clinical R&D Frank Haluska, talking on Ariad's third quarter results call on November 12. Yet the more recent FDA statistics used to back up the decision to withdraw were based on “a new definition of events...with more kinds of events clumped together, and more adverse events categorized as serious adverse events,” Haluska said. Had the same definitions been applied as at the time of approval, FDA's headline rates for serious adverse events would have been halved, from 48% to 22% in the phase 1 trial and from 24% to 12% in the phase 2 trial.

Despite the set-back, Ariad is still betting on Iclusig's potential to match the unmet clinical need. Fewer than half of CML patients resistant or intolerant to two or more TKIs have the mutation, according to 2011 data from UK-based research group Ipsos Healthcare. Ariad CEO Harvey Berger also points to FDA reports of over 200 US patients in the last month accessing Iclusig through single-patient Investigational New Drug applications, the only way to obtain Iclusig when it was withdrawn.

Dosing may also be tweaked to improve the risk-benefit profile. Data presented by Ariad at the American Society of Hematology (ASH) meeting in early December 2013 in New Orleans suggested that lower doses would reduce the risk of vaso-occlusive events, yet maintain efficacy in most patients. It also highlighted age and prior disease as influencing the rate of cardiovascular events—factors that may find themselves on revised clinical guidelines for the drug's use.

Ariad's drive to return Iclusig to as wide a portion of suitable US patients as possible is understandable: as the company's only marketed drug, it was propping up almost the entire valuation. The company slashed 40% of its workforce and pledged to reduce spending in 2014 by 35%. With cash supplies extended until mid-2015, Berger is confident the firm won't need to go back to investors before Iclusig returns. But to protect itself from an unwanted takeover, Ariad announced a shareholder rights plan—also known as a poison pill—the day after the withdrawal.

Citi Research analyst Jonathan Eckard in New York agreed back then that there would be a path back to market for Iclusig, even when sales might be limited. Eckard believes Ariad remains a “viable company that could turn profitable,” thanks also to phase 1/2 non-small cell lung cancer candidate AP26113. Indeed, “our basic strategy is unchanged. We're moving forward with filing [Iclusig] in Japan; [the plan in] Europe is in place, and we will solve these problems in the US,” declares Berger, reflecting back on a similar FDA setback experience while head of R&D at Philadelphia-based Centocor (now part of Johnson & Johnson) over 20 years ago.

Berger underlines two lessons from the recent experience. First, be aware that a safety-focused FDA could take a new line on how it accounts for adverse events. Second, consider studying multiple doses in pivotal trials, rather than focusing early on one optimal dose.

Iclusig's withdrawal and re-instatement won't likely affect FDA's accelerated approval pathway or its breakthrough therapy designation, says Citi's Eckard. There will be more Iclusig-like stories, where a first-round approval has to be adjusted. Each time this occurs, “both the Agency and the company will learn from it, and how to optimally approach it,” notes Eckard.