#71 – Katherine Eban: Widespread fraud in the generic drug industry

KEYPOINTS

• 01

  Patient complaints started the investigation

  Eban, an investigative journalist and author of Bottle of Lies, had seen generics as a public good: affordable, life-saving drugs that make programmes like Medicare and the VA workable. In 2008 Joe Graedon, host of NPR's The People's Pharmacy, told her he was deluged with complaints about generics not working. FDA officials dismissed them as psychosomatic, blaming switched pills that looked different. After about 20 years as a reporter, Eban says 'something tingles'.

• 02

  Generics are reverse-engineered, not copied

  A patent discloses the molecule, but the manufacturing process is often 'the secret sauce'. Attia compares it to tasting a finished dish that took 12 hours, seven steps and 13 ingredients, then trying to recreate it. Generic makers only have to show bioequivalence, meaning the drug behaves the same way in the body. They skip the phase 1–3 trials that test safety and efficacy, so the integrity of their data is everything.

• 03

  Hatch-Waxman created the modern generic industry

  The 1984 Hatch-Waxman Act, signed by Reagan, let generics be approved without new clinical trials. It gave the first generic filer six months of exclusivity, a valuable prize because prices collapse once about ten generics enter (to roughly 30 cents on the dollar). Attia wonders whether the trade-offs, including patent life, were set well. A bribery scandal at the FDA in the late 1980s showed that both industry and regulator could be compromised.

• 04

  Globalisation outran the FDA

  By 2005 the FDA had more plants to inspect overseas than in the US. Today about 80% of active ingredients in all US drugs, brand or generic, come from abroad. About 90% of prescriptions are generic, and about 40% of generics come from India. The US makes almost none of its own antibiotics. PEPFAR, the Bush administration's HIV programme in Africa, helped normalise cheap Indian generics, and Scott Gottlieb, then in the Bush White House, pushed questions about how to guarantee their quality.

• 05

  Ranbaxy fabricated data for more than 200 products in about 40 countries

  In 2003 engineer Dinesh Thakur left Bristol-Myers Squibb's spotless New Jersey campus for Ranbaxy in Gurgaon, India's would-be 'Pfizer of the 21st century'. Ranbaxy was nearing $1 billion in sales and held first-to-file rights on drugs including generic Lipitor. His new boss, R&D head Rajinder Kumar, who had come from GlaxoSmithKline, had Thakur's team audit its filings. They found the company had tested brand-name drugs and passed them off as its own to fake bioequivalence, then sold untested or substandard products. Eban says the 200 products were essentially the whole portfolio.

• 06

  The 'self-assessment report' went straight to the board

  Kumar presented the findings to Ranbaxy's board in a blandly titled 24-slide 'self-assessment report' (SAR), 'smoke and fire' on every slide. Attia compares it to finding bodies in every room of a house and showing the killer a PowerPoint. Rather than fixing the problem, the company moved to contain it. Thakur, anonymously through a Yahoo email account, went to the FDA.

• 07

  The investigation took years

  FDA agents raided Ranbaxy on February 14, 2007. A Ranbaxy enforcement team grew to about 30 people, unpicking each type of fraud, such as stability tests at 6, 9 and 18 months that were never actually done. The case settled in 2013. Eban's reporting drew on about 20,000 internal FDA documents and 16 Freedom of Information requests, and she even sued the FDA to obtain records.

• 08

  Ranbaxy was the tip of the iceberg

  FDA chemist Peter Baker, a tattooed, motorcycle-riding 32-year-old 'new sheriff in town', refused the usual courtesies. Overseas inspections were pre-announced, and companies arranged inspectors' travel, gave souvenirs and offered massages. Over five years Baker inspected 86 plants in India and China and found some form of data fraud or manipulation in about four-fifths of them. Some plants even had staff whose actual job was data fabrication.