Last week, ProPublica reported
that the Food and Drug Administration allowed dozens of medications to stay on
the market, even though the research designed to prove their safety and
effectiveness was undermined by “egregious” violations at a major pharmaceutical
research laboratory in Houston. New information shows that even after the FDA
had cited the lab for falsifying data and other misconduct, the agency issued a
brand new approval to a drug tested there.

The FDA has refused to reveal the
names of any of the approximately 100 drugs affected by the fraud at the Houston
lab of the firm Cetero Research, saying that to do so
would reveal confidential commercial information. ProPublica was able to
identify five of those drugs, and now we’ve found a sixth.
This one was approved after the agency had already cited the Houston lab for
misconduct.

The drug is a generic version of Tussionex, which combines a long-acting narcotic cough
suppressant with an anti-allergy medication. Manufactured by TrisPharma, the drug has a
tongue-twisting chemical name: hydrocodone polistirex/chlorpheniraminepolistirex.

ProPublica discovered that both of
the clinical
trials used to show that the generic is equivalent to the name
brand — a key requirement for FDA approval — were analyzed in May
and June 2009 at the Houston lab of the firm Cetero Research.
The company, which conducted research for scores of pharmaceutical companies, has
acknowledged that chemists at its Houston lab committed research fraud, though it
says the misconduct was limited to a handful of employees and that none of
their tests have so far proven to be wrong or inaccurate.

Cetero
filed for Chapter 11 bankruptcy last year and emerged with a new name, PRACS
Institute. PRACS, in turn, filed for bankruptcy earlier this year.

The FDA got wind of problems at Cetero Houston in June 2009 when, prompted by a
whistleblower, the company alerted the agency of potential wrongdoing.

On May 3, 2010, three FDA agents
came to inspect the lab. The facility’s president turned over eight flatbed
carts double-stacked with file boxes and admitted that much of the lab’s work
was fraudulent, saying, “You got us,” lead agent Patrick Stone recalled. (The
lab’s president declined to comment.)

On May 7, the FDA issued an inspection
report that cited data falsification and other laboratory
violations at the facility during the time its chemists conducted the tests for
the TrisPharma drug.

But five months later, in October
2010, the FDA approved
Tris’ drug for sale in the United
States.

The FDA continued to inspect the
Houston lab, and in July 2011, the agency called
the lab’s misconduct so “egregious” and pervasive that more than five years of
tests conducted at the lab were potentially “unreliable.” The agency asked pharmaceutical
companies to repeat or reanalyze any tests conducted there during that time if
they had helped win drug approval. The FDA did not pull any drugs off pharmacy shelves, despite the fact that dozens were approved on data
the agency now said was questionable.

In April 2012 the FDA relaxed its
requirements. The agency announced that studies analyzed at Cetero Houston between March 2008 and August 2009 —
the time period in which tests for the TrisPharma drug were conducted — would not need to be
redone but instead would require only a “verification of data integrity by
an independent third-party audit.”

Tris said
it hired a consultant to perform the audit and it “came back 100 percent
clean,” Groner said. “Something could have
happened, but in our case, nothing did.” He said Tris
submitted the audit to the FDA in the second quarter of 2012.

The FDA has reviewed the audit and found
it “acceptable,” an agency spokesperson wrote in an email. “Our earlier
determination that this product meets FDA standards was upheld.” She declined
to add any information or answer more specific questions.

The main point of that audit was to
look for red flags that researchers might have cooked the books, said Scott Groner, TrisPharma’s
Director of Regulatory Affairs and Compliance. The Cetero
researchers ran blood samples through a machine known as a mass spectrometer. By
examining the timestamps of these runs, Groner said, the
auditors could tell whether or not a rogue researcher would have been able to
tamper with the results of the experiment. “If there was some
manipulation, there would have to be some down time” for the researcher to
fiddle with the samples or the machine in order to produce the desired result, he
explained. Auditors, he added, “were looking for many
other things. Were [experiments] done late at night, on weekends?”

While the FDA declined to comment on the specifics of the Tris
Pharma case, the agency has previously said that it
ran a risk assessment and concluded that the chances were low that drugs tested
at Cetero Houston were unsafe. The fact that the FDA
has since found no problems with drugs tested there “confirmed” that risk
assessment, a top FDA official said.

As of last week, the agency said it
had completed reviewing 21 out of 53 submissions it had received from
pharmaceutical companies that had medications tested at Cetero
Houston. It also said a few companies had not yet submitted new data or audits
on drugs tested there.

Rob Garver can be reached at [email protected], and Charles Seife can be reached at [email protected].