Why did Novartis wait about two months before launching an internal probe into the alleged data manipulation surrounding gene therapy Zolgensma? That's what officials at the FDA want to know.
After an on-site inspection at AveXis’ control testing lab in Irvine, California, which appears to have been prompted by the firm’s communication of the problem to the FDA, inspectors issued a Form 483 on Aug. 2. The report noticed a delay from when the initial Zolgensma data allegation was flagged by a scientist on Mar. 14 to when a formal internal probe was opened on May 15.
Two months seems like a lot, given that an FDA decision was looming, unnamed FDA officials told The Wall Street Journal.
What happened during those two months? By Novartis’ account, before it started the full-on investigation in May, it “conducted an independent investigation with the support of external counsel to determine the extent of data discrepancies and whether they could be explained,” a company spokesman said.
The reason why it took so long was because the preliminary probe “consisted of an extensive review of thousands of hand-written and electronic records concerning the life history of individual mice and a comparison of those records to entries on the hundreds of spreadsheets on which their data was recorded,” he said.
The FDA, citing ongoing investigation, declined a FiercePharma request for comment on whether it considers Novartis’ two-month delay in opening a Nonconformance Report to be problematic.
In a statement to WSJ, it said: “While in some instances it may be reasonable for a firm to make a determination of the merit of a complaint in order to decide whether or not to initiate an investigation, each decision must be considered on a case-by-case basis. We are not able to comment specifically on the circumstances of this case.”
The FDA first disclosed data manipulation happened in animal tests from Zolgensma’s application package on Aug. 6. Besides the very fact that falsified data were submitted, the initial public scrutiny harped on a delay of more than a month between when Novartis confirmed the allegation in an internal investigation and when it informed the FDA of the problem.
Novartis has said it was trying to understand the problem in detail during the period. “It was normal course of business and pretty typical for how we handle these items,” Novartis’ head of global regulatory affairs, Robert Kowalski, told analysts during a conference call after the revelation.
But the FDA officials WSJ interviewed said the company should have immediately notified the agency, even if the data breach was not confirmed.
The thing is, drugmakers don’t have to report suspected—not confirmed—data misconduct and won’t necessarily face FDA penalties. The FDA last October withdrew a proposed rule that would have required trial sponsors to report information suggesting any person “has, or may have, engaged in the falsification of data.”
A group of senators, led by Sen. Dick Durbin and including Sens. Elizabeth Warren and Bernie Sanders, recently requested an explanation from the agency as to why it had decided to scrap that proposal, and whether they would consider reinstating the regulation in the wake of the Zolgensma scandal.
As the FDA is still running its assessment, Sen. Chuck Grassley, chair of the Senate Finance Committee, has also started one, requesting all records related to the incident. He has set the deadline for Friday.
Editor's Note: The story has been updated with a statement from Novartis.