FDA reviewers question efficacy, raise safety issues and cite a challenge on Clovis' rociletinib

The FDA today raised a host of thorny issues for an upcoming advisory committee meeting scheduled to review Clovis Oncology's ($CLVS) application for an accelerated approval of its non-small cell lung cancer drug rociletinib.

In an internal review released Friday morning, the agency noted that it wants its group of outside experts to consider whether rociletinib is better than existing drugs on the market, offered its opinion that serious safety issues associated with the drug will require a "black box" warning to patients and added that the agency challenged the company's use of unconfirmed responses in its data on the drug last fall.

The comparison the FDA is asking for will pit rociletinib directly against AstraZeneca's Tagrisso, which won an approval last year based on an overall response rate of 59%, far above the final 32% mark that the Boulder, CO-based company has posted for a 625-mg dose of its drug. Tagrisso and rociletinib both inhibit EGFR and are designed to target lung cancers with a T790M mutation.

The agency's review also notes serious adverse events in the rociletinib studies, singling out hyperglycemia and QTc prolongation leading to Torsades de pointes--a lethal irregular heart rhythm--which should require the black box warning if the drug is allowed on the market.

Significantly, the review includes a timeline on the interactions between the FDA and Clovis execs for this drug, which won the agency's breakthrough drug designation based on an overall tumor response rate that was reported early on at more than 50%. By last June, though, the response rate reported by the company for its 500-mg dose was 38%.

Then, on November 9, the FDA noted: "During the Mid-Cycle Communication with Clovis, FDA provided an update of the status of the review and communicated significant issues arising from the review. The FDA stated its disagreement with Clovis' reported efficacy results of the pooled analysis for Studies CO-1686-008 and CO-1686-019. The disagreement was based on Clovis' inclusion of patients with unconfirmed responses in the efficacy assessment."

Clovis admitted that it was using unconfirmed responses and submitted an update that required a three-month extension of the FDA review.

Several days later, the company updated its data, triggering a collapse of its stock price and sparking outrage among its investors over the company's timing.

At the beginning of this week, noted cancer drug development expert Kapil Dhingra stirred considerable discussion with a new analysis which questions Clovis's use of unconfirmed data long after most investigators would have nailed down a final confirmed response. In an interview with FierceBiotech, he charged the company with misrepresenting its data.

"I feel that the efficacy data have, consistently and repeatedly, over many years, been misrepresented," Dhingra told me earlier. "This is not simply a case of gray zones, this is black and white untrue presentation of the data. And it is not just a minor misrepresentation (such as photoshopping a western blot image etc that can get a basic scientist in trouble); the true efficacy is about half of what they represented."

Typically, the FDA's cancer advisory group would limit themselves to a conversation devoted to a drug's risk/benefit profile. Recently, practically any drug that offered a clinical benefit could expect an endorsement, even with serious safety issues. But this time around the FDA may be setting the stage for putting off an approval and waiting for confirmatory study data, or at the very least leaving the drug seriously hobbled in a head-to-head market fight with a competitor.

That's all something that the FDA advisers will get a chance to comment on next Tuesday. It's also a subject that FDA cancer czar Richard Pazdur could address. I asked him by email earlier this week if he plans to attend the AdComm next week, and in a one-word reply he responded "yep."

- here's the review