2007
Rockville, Maryland
As the regulatory case against Ranbaxy moved slowly through different FDA divisions, it landed on the desk of Douglas A. Campbell, a thirty-seven-year-old officer with the agency’s international compliance team. Some of his superiors found little of interest. “We don’t really think there is anything there,” one of them advised. The case looked like a protracted fight over a single manufacturing plant. Ranbaxy’s lawyers were insisting that the FDA lift its temporary ban on new applications from the Paonta Sahib plant, claiming that the company had made all needed corrections. Regulators were demanding that Ranbaxy hand over full copies of the audits done by its consultant, Parexel. Stalemates like these littered the FDA’s offices.
But as Campbell read the paperwork for the case, he became intrigued. In its responses to the FDA’s inspection findings, the company blamed most of the problems on transcription errors, lost data, or mismatched internal systems. But in excerpts from the audits conducted by Parexel, which Ranbaxy’s lawyers had refused to release in full, the consultants had noted “inconsistent entries” in some of the data from stability tests. The company was now sending corrected data. But some of the discrepancies seemed awfully big to be mere mistakes, such as a forty-five-day discrepancy as to when a test was conducted. How could the company have been so confused? Or sloppy?
Companies had to continue testing their drugs at predetermined intervals long after they launched them on the market. Every year a company had to file the test results in annual reports to the FDA. Few at the agency actually read these reports, and they piled up in back offices. But the information in the reports still needed to be true.
On July 3, 2007, Campbell drove over to the Office of Generic Drugs, which was headquartered at the Metro Park North campus, and dug out the annual reports that Ranbaxy had filed for three anti-infective drugs: fluconazole, ciprofloxacin, and efavirenz. From the reports, the company appeared to be testing the drugs at appropriate intervals. But as Campbell compared that data to the corrected data on his desk, he was startled by the differences. The annual report for fluconazole stated that Ranbaxy had conducted the three-month stability test on September 26, 2004. But in its response to the warning letter, the company stated that it had performed that same test on August 17, 2005. There was almost a year’s difference between the two dates.
To Campbell, it looked like Ranbaxy had either lost control of its manufacturing process or lost track of its lies. “Once you matched up [the dates], it was blatant,” he recalled.
His discovery mattered—at least to him. The whole reason to test drugs so frequently was to ensure that unsafe products were detected quickly and didn’t stay on the market any longer than necessary. But Campbell’s superiors didn’t seem particularly excited by the mismatch of dates. Some seemed satisfied by the company’s claims that its problems stemmed largely from “transcription errors” or simple failures to update data. Campbell didn’t buy it. This was not just “.54 vs .45 type of stuff,” as he recalled. Ranbaxy’s data was so inaccurate that he couldn’t even “find a place where the data actually meant anything.” And if the data was meaningless, then there was no proof that Ranbaxy’s drugs were safe and effective.
Doug Campbell’s boss, Edwin Rivera-Martinez, the first FDA bureaucrat to have corresponded with Thakur, sided with Campbell. He was not mollified by Ranbaxy’s claims of correction. As he wrote to a colleague in March 2007, he was convinced that the agency should continue to withhold approval of new applications from Paonta Sahib until Ranbaxy had “completely addressed all issues” from the FDA’s warning letter. As Campbell and his colleagues continued to review the documentation from Ranbaxy, each thread they pulled, each strand of testing data they followed, seemed to lead to something bigger—some new revelation that the regulators had not anticipated.
By October 2007, the company reported that its gabapentin, a sensitive drug used to treat epileptic seizures, had shown a spike in an impurity called compound A. Under good manufacturing practices, abnormal spikes or troughs in data were known as out-of-specification (OOS). The company had been required to report the gabapentin finding to the FDA within three days of discovering it, but instead it had waited four months. The regulators soon discovered that the company had not just failed to report the gabapentin impurity spike. It had failed to report irregular test results to the FDA’s New Jersey district office for six years. For any vigilant drug company with a large volume of products, such reports should have been a regular occurrence.
The company blamed the troubling lapse on a cascade of small internal problems. But when the FDA investigators went to Ranbaxy’s New Jersey headquarters, which should have been issuing the reports, they stumbled on an even bigger finding: for the 600-milligram gabapentin tablets, the company had conducted its three-, six-, and nine-month stability tests in a span of four days, and all on the same day for the 800-milligram tablets. But the test dates had been falsified and documented as if they were performed at the proper intervals.
Inside the FDA, this news was regarded as slam-dunk proof of fraud. Campbell emailed Rivera-Martinez, “Bullseye!!!” Rivera-Martinez reported the finding up the chain of command: “We hit a goldmine!” He laid out the details of the phony testing. Deb Autor, director of CDER’s Office of Compliance, responded with one word: “Wow.”
Suddenly, the lapses, irregularities, and omissions looked like something very different. Campbell scrawled in his government-issued notebook, “How Big Is the Problem?” And then, “What do they do to come into compliance?” Below this he wrote, “Present the chance to come clean or lie again. Will they continue to jerk us around?”
And below that: “Can we trust them?”
Campbell, a brawny former football player, spent eight years in the army, three on active duty. He had started his FDA career in 1998 at the Roanoke, Virginia, resident post, inspecting everything from infant formula to a tilapia aquaculture farm. By 2006, Campbell had moved to the Division of Manufacturing and Product Quality’s international compliance team, nested deep within CDER. His group performed about one hundred overseas inspections a year. Campbell himself had inspected cheese in Nicaragua and stuffed grape leaves in Greece.
Almost overnight, it seemed, the agency was hit by a wave of globalization. The number of pending inspections for Campbell’s division spiked. The foreign applications were “all stacked up in our offices,” Campbell recalled. From 2002 to 2009, the number of facilities overseas that required inspection by the FDA skyrocketed from around five hundred to over three thousand. As Campbell noted in a review, “The responsibilities related to our mission have exploded. The burden is immense, and the resources have not been readily allocated.”
As the need for inspections grew, FDA policy became an all-out scramble to keep up. At one point, Campbell wrote to a colleague to ask about the agency’s travel policy: “Do we send teams to India during Monsoon season? Do we send teams to India when it’s going to be 110+ degrees in the shade?” The answer came back: “In the past we would routinely defer travel to India during the monsoon, however, due to increase workload we no longer adhere to this.”
Before a company was cleared to manufacture a drug at a plant overseas, Campbell’s division was supposed to conduct a preapproval inspection to determine if the facility could safely and competently make the drug. But just keeping track of the facilities, and the respective codes that identified them, was a fearsome task. Were the manufacturing plants in the system those really making the drugs? Were the plants that were actually making the drugs the ones being inspected? Campbell understood that he and his colleagues held an enormous responsibility, which weighed on him as he sifted through the Ranbaxy paperwork: “[Should] we let these drugs come into the U.S. or not?”
In October, nine months after Jose Hernandez had inspected Paonta Sahib and come away with suspicions but little else, he received a remarkable email. It was from an employee, writing under the pseudonym “Sunny,” who said that he’d seen Hernandez at the plant in January and had finally summoned the courage to write. Sunny explained that Hernandez had been deceived, like so many FDA investigators before him. “At last my conscious did not allow me to keep quiet as its a matter of health of people. Ranbaxy has been hiding many facts till now.”
Sunny went on, “What you see at these locations is not the real stuff. It requires a minimum of one month to find out the real matter.” He described how the company’s top managers exerted relentless pressure on lower-level employees and forced them to clear crucial medicines—isotretinoin, gabapentin, flucanozole, metformin—for release. All had “issues but they are dispatched from the site by QA [quality assurance].” He named senior executives who had orchestrated the scheme: “Over the years they have all bluffed and fooled the FDA.”
Sunny explained that before Hernandez arrived at Paonta Sahib, a team of twenty people from research and development had descended on the plant to review and alter data. “This cleanup was done just before the FDA inspection,” he wrote. “Such thing is done routinely at all the plants at Ranbaxy.” Executives stage-managed the deceptions at the direction of top company leaders, he explained, with lower-level employees carrying out their orders under duress. To support the claim that the company had increased staffing at its stability lab, it had shifted employees from around the plant to that lab during Hernandez’s inspection.
As the whistleblowers’ revelations spread among FDA officials, regulators were surprised by Ranbaxy’s duplicity, in part because it was difficult to fathom the extent of it. The FDA was confronting a system in which data had been so artfully altered that everything seemed perfect. Until then, despite the evidence that had emerged from the search warrant, they had thought of fraud as discrete acts committed by individuals. But what if the way the entire company operated was fraudulent? How could they pierce a scheme of lies that all the employees were in on?
As the FDA descended on Ranbaxy facilities, investigators suddenly noticed everywhere the clues they had previously walked by: the use of unapproved materials; the secret changes in formulation; the use of unregistered active ingredients; the plagiarism of already published data, down to copying chromatograms, or the graphs measuring impurities from brand-name drugs, and passing them off as the company’s own. The company seemed to have sprung a leak, with at least half a dozen whistleblowers now writing in to the agency with their own examples of fraud and misconduct.
Sunny continued to write, and Hernandez soon put him in touch with Debbie Robertson, who was leading the probe at the FDA’s Office of Criminal Investigations. Sunny revealed that Ranbaxy, unable to find a legitimate fix for its troubled Sotret drug, had secretly altered the formulation while it was on the market, adding oil to the wax base to try to improve the drug’s dissolution. None of this had been reported to the FDA, making it an egregious violation. Companies were strictly prohibited from making any changes to approved formulations without the FDA’s permission.
Another whistleblower also flagged the Sotret and urged the agency to check the difference in the formulations of the drug before and after December 2006, when the switch was made. “Some scientists wanted to address the issue earlier during 2005 and early 2006 but the commercial staff was ruthless with them,” the whistleblower wrote to CDER’s ombudsman. “They really have no appreciation for the honest people. I don’t know the harm that must have been caused to U.S. users but being part of the [manufacturing] team, I thought as a good world citizen, I must bring this to your notice.”
It seemed there was almost nothing the company wouldn’t do or say, no excuse it wouldn’t make, no claim too far-fetched, to trick the agency into approving its drugs. Instead of investigating its own out-of-specification results, as required, Ranbaxy claimed that its own lab had mishandled samples, which had erroneously led to poor test results. Sometimes it seemed that Ranbaxy was better at making excuses than it was at making drugs. As one FDA regulator urged her colleagues, the company’s “discrepancies, inconsistencies, mistakes, oversights and poorly executed investigations” were not to be taken at face value. The lies were so audacious that Debbie Robertson would later say that she had never encountered anything like it: “In all my years [investigating] drug dealers, I had never had that blatant disregard for the law. They lied right to your face. I was told, it’s a cultural thing. They understood, but they knew they could get away with it.”
The only answer to this global malfeasance seemed to be global punishment. And the agency had a rarely used remedy, one of the most onerous punishments it could inflict: an Application Integrity Policy (AIP), which it had imposed on only four drug companies since the policy was first established in 1991. The AIP allowed the FDA to halt all review of a company’s applications, to be resumed only after outside auditors—hired at the company’s expense—certified that the data was legitimate. The sanction would flip the regulatory dynamic. No longer would the FDA have the burden of proving fraud if it wanted to block a Ranbaxy product. Instead, in order to get its products approved, the company would have to prove that its products weren’t fraudulent.
The agency imposed an AIP only when it had found criminality or “untrue statements of material fact,” which it clearly had in Ranbaxy’s case, or so Doug Campbell believed. He had drafted a memorandum lining up the various falsehoods that regulators had been able to document. He proposed that CDER impose the AIP for “all approved and pending applications related to Ranbaxy Laboratories Ltd.” In short, bring the hammer down on the entire company.
But as drafts circulated and meetings gave way to more meetings, Campbell began to doubt his own ability to communicate. Nothing seemed to rise to the requisite level of proof. No one seemed exactly sure if the AIP was merited or even defensible. The FDA was crowded with lawyers who seemed more intent on heading off court challenges from companies than on protecting public health. The agency’s lawyers debated the FDA’s requirements. Did it say anywhere that companies could not keep drugs in unregistered refrigerators? If the company claimed that it had lost raw data, was there any written requirement that they keep it forever?
Even the FDA’s own role seemed unclear. Was it the agency’s job to help Ranbaxy comply? Or to cut off the company for failing to do so? Shocked by the internal dithering, Campbell jotted in his notebook, “Our goal cannot be to make it easy on Ranbaxy!” But in fact, that seemed to be the case. There was intense congressional and public pressure to find more cheap drugs. Ranbaxy also played a crucial role in providing AIDS drugs to Africa under the government’s PEPFAR program. Was Ranbaxy simply too big, or too important, to fail?
The FDA’s confusion over its own enforcement mission was not the only problem. “There was some force holding this case down,” Campbell would later conclude. Did it have to do with money? With connections? With political sway? He grew suspicious, particularly when some of the FDA’s Indian employees who’d never come to his office before seemed to find any excuse to visit him.
Then there was Deb Autor. As an attorney who headed CDER’s Office of Compliance, she sat a number of levels above him in the FDA bureaucracy and oversaw the work of some four thousand employees. Before she’d joined the federal government in 1995, she’d worked for three years at the law firm that became Buc & Beardsley, which was now representing Ranbaxy. As such, she was on a first-name basis with the partner who served as Ranbaxy’s outside counsel, Kate Beardsley.
Beardsley turned to Autor as her direct conduit to the workings of the case, calling and emailing to get insight into the government’s progress and trying, whenever possible, to move forward decisions that would be helpful to her client. Autor prodded the bureaucracy at Beardsley’s behest, an effort that she viewed as part of her FDA role: to respond to lawyers whose clients had matters pending before the agency. Autor had left the firm thirteen years before the FDA’s compliance office began considering the Ranbaxy case. But to Campbell, it seemed that sometimes Autor was more focused on assisting her former boss than on helping to advance the agency’s case against Ranbaxy.
Beardsley wrote to Autor in March 2007: “Deb, I left you a phone message, but thought it would be sensible to send you an email too asking if you could give me a call about Ranbaxy. We’re still trying to sort things out on the civil side, as they are also trying to cope with the criminal side.”
Autor responded, “Hi Kate. I’m happy to call you. But it seems like I know enough about your concerns to follow-up around here first and get back to you. Okay?”
In December, in another email, Beardsley contacted her again, to explain the delays on releasing the Parexel audits in their entirety and asking that Autor call her. Two hours later, Autor told her colleagues via email, “Ranbaxy needs to think through the implications for the criminal case of providing the audits. Please consider whether you can provide Ranbaxy with Doug’s 12/6/07 list of questions in order to give them more information about what GMP issues are still troubling us.” Given the overlapping connections and bureaucratic foot-dragging, Campbell found it hard to distinguish who was working for the FDA from who was working against it. As the case became increasingly fractious, Campbell hesitated to disclose things to Autor.
Even if agency regulators had not dragged their feet, they faced a problem without an obvious solution: how could they verify that a company was making the changes it claimed to be making in factories that were over seven thousand miles away?
In an email to Robertson, the whistleblower Sunny described how Ranbaxy used hidden areas of the plant to store and cover up testing machines that were not connected to the company’s main computer network. He was referring to the crucial high-performance liquid chromatography (HPLC) machines, the workhorses of any good testing laboratory. The bulky machines looked like a stack of computer printers. Once a drug sample is mixed with a solvent, injected into the machine, and pressed through a column filled with granular material, the machine separates out and measures the drug’s components, including impurities. It displays them as a series of peaks on a graph called a chromatogram.
In a compliant laboratory, HPLC machines would be networked with the main computer system, making all their data visible and preserved. During a recent inspection, Sunny wrote, the unauthorized HPLC machines were kept in two ancillary labs: “Ranbaxy creates small such hidden areas where these manipulations can be done.”
Sunny estimated that some thirty products on the U.S. market did not pass specifications and advised Robertson that the agency needed to raid Paonta Sahib and Dewas, just as it had done in New Jersey, to find the evidence. He warned, “The move has already started in Ranbaxy to share such details of problematic products personally and not on emails or letters.”
But because the U.S. Attorney had no jurisdiction in India, the FDA couldn’t execute a search warrant there. Robertson felt thwarted: “People said, ‘You need to go to India.’” But her response was, “What am I going to do [over there], knock on people’s doors and hope they talk to me? I don’t have authority over in India. It’s all a voluntary, good-faith system.” The case had crashed like a wrecking ball into the overtaxed agency, exposing the fact that the FDA had no effective way to police a foreign drug company.
In November 2007, as the FDA prepared to inspect a factory at Ranbaxy’s Dewas site that made sterile injectable products to consider whether to approve its drugs for the U.S. market, Sunny emailed Robertson his most important tip yet. The FDA’s regulations required the facility to have one of the highest levels of sterility possible. But, Sunny warned Robertson, “the microbiological data is not actual and is manipulated to show less microbial counts,” adding that the plant had had several unreported sterility failures. He advised, “Be careful before approving this facility.”
A month before the FDA’s inspection, Sunny wrote again to Robertson to alert her to a cover-up in progress: “all the actual failure data of Environment monitoring and sterility failures” were being moved out of the Dewas plant to a warehouse fifteen miles away, in Raokheri. Sunny advised that those inside the plant “will try their best to confuse the auditors and they are trained for the same. People in QA have been told to keep their mouth shut about the deviations.”
The international compliance team members knew that if they wanted to get the evidence that microbial test results had been manipulated, they needed to go, unannounced, to the Raokheri warehouse. Rivera-Martinez submitted the team’s request to do that. But the answer he got back was not what he had expected. In a terse email marked “confidential,” Patricia Alcock, then deputy director in the Division of Field Investigations, wrote: “Please check your voicemail message. The warehouse is off the table.”
Alcock explained in her voicemail that an unannounced inspection could jeopardize ongoing diplomatic efforts between the FDA’s parent agency, the U.S. Department of Health and Human Services (HHS), and Indian health authorities. The two sides were negotiating a written statement of cooperation that might one day enhance the quality of FDA-regulated products made in India. To accomplish even a draft statement, HHS officials wanted to avoid angering the Indians.
Rivera-Martinez was outraged. He prepared an email to Alcock that he shared with colleagues, challenging the refusal. He stressed the seriousness of the allegations. If “true [they] raise significant questions regarding the corporation’s compliance attitude, adequacy of the firm’s quality management system, and manufacturing state of control.” He reminded Alcock that just last month, in a briefing to the staff of Senator Charles Grassley (R-IA) on what it was doing to strengthen foreign inspections, the FDA had even pledged to conduct them unannounced. In their congressional briefing, FDA officials had acknowledged that there was no legal requirement to notify foreign facilities in advance.
Rivera-Martinez’s office deemed an unannounced inspection of the warehouse to be “both warranted and necessary.” Armed with another email from Sunny about up-to-the-minute manipulations of data, Rivera-Martinez wrote to his superiors: “FYI, here’s another e-mail from the informant regarding Ranbaxy’s keen ability to manage our inspections and investigators. It’s quite clear to me that we have to consider different and bold inspection/investigative strategies and techniques in dealing with Ranbaxy to increase our chances of uncovering evidence of data falsification/manipulation. . . . That’s why I’m insisting on an unannounced inspection of the warehouse.”
Rivera-Martinez was insisting on common sense. But this time he heard back from the deputy associate commissioner for compliance policy in the Regulatory Affairs Office, who also nixed the warehouse inspection. The deputy explained the thinking of senior FDA officials: “We agreed that any unannounced inspections in a foreign country [need] to be well thought through and planned with all parties to ensure investigator safety and that we minimize the potential that any adverse international incident occurs. Since no such planning had taken place by the time the inspection team left, we agreed that no unannounced inspection would take place this trip.” So because of the risk of international conflict that might scuttle a diplomatic effort, the FDA would not fully and unreservedly investigate foreign facilities that presented potential threats to U.S. public health. The needs of patients came last.
Instead, while at Dewas, the investigators requested a visit to the Raokheri warehouse. They were taken there the following day and spent eight hours rummaging through drawers and boxes. As Alcock informed Rivera-Martinez, they “didn’t find anything pertinent or as described by the informant.” The whistleblower Sunny later told Robertson that “just before the inspection somebody inside alerted to move unaccounted material from the Raokheri warehouse.”
But the company could not conceal a more startling problem at the supposedly sterile facility. As Alcock conveyed to her colleagues, “The building . . . is surrounded by pig farms. [The investigator also observed] a lack of instructions/procedures advising personnel to wash their hands/feet before entering the sterile core. (Many of the workers wear sandals. . . . There are also a tremendous amount of pigs scattered on/near the facility???).”
The FDA did not approve the sterile facility. But as agency regulators continued to approve other Ranbaxy applications, their permissive approach seemed increasingly unsustainable. Toward the end of 2007, Deb Autor learned that a different federal agency, the U.S. Agency for International Development (USAID), was considering terminating Ranbaxy as a provider of low-cost drugs to Africa. Instead of applauding the aggressive action, Autor expressed concern that such a clear rebuke of the company would make the FDA’s regulators look bad. It would raise “questions about why FDA has not shut down Ranbaxy,” she warned her bosses. She offered to take up this question in person or by phone. One of those copied on the email advised colleagues, “Keep this information to yourselves.”
On December 12, 2007, USAID sent a stern letter to Ranbaxy. It accused the company of delaying reporting negative test results to the FDA. “I am very troubled by this apparent lack of business integrity and honesty by a company performing on a subcontract funded by the U.S. Government,” the letter stated. It was signed by the acting director of the Office of Acquisition and Assistance and stated that USAID was considering suspending or debarring Ranbaxy from the program. By contrast, the FDA seemed more comfortable to continue its accustomed approach: “Regulators were perfectly happy to stall and stall,” Campbell recalled.
By early 2008, Ranbaxy had submitted an application for approval of a plant called Batamandi, in Himachal Pradesh. It proposed to make sensitive drugs there, including tacrolimus, an immunosuppressant used by transplant patients to prevent organ rejection. The application caused immediate suspicion among Campbell and his colleagues. Batamandi was close enough to Paonta Sahib to possibly be a part of it. Was Ranbaxy trying to pass off a part of Paonta Sahib as a new plant as a way to evade the agency’s restrictions there?
As was customary, the agency ordered a preapproval inspection for the manufacture of tacrolimus at the Batamandi facility. But there was nothing ordinary about the assignment. Jose Hernandez was tapped to go. In early March, he arrived at Batamandi, armed with his “broad scope of thinking,” as he liked to put it, and a determination not to be fooled again. As was typical, Hernandez began his inspection outside, taking in the grounds. By standing at the edge of the property, he could see the Paonta Sahib plant, about two and a half miles away. He noted an eight-foot-high fence that surrounded almost all of the Batamandi plant. There was only one entryway, secured by a well-staffed guard post. The guards there, former military policemen, seemed proud to confirm that they logged all employees and visitors in and out through the gate. No one got past them.
This gave Hernandez the opportunity he needed. By examining checkpoint records at the gate, he learned that the supervisors who had signed off as present for the manufacturing of key tacrolimus batches had not actually been at the plant on those days. They had not signed in to the security gate logs. The dates, times, and signatures of the batch records were fake and had been filled in after the fact. One evening at his hotel, where other Ranbaxy executives who had flown in for his inspection were also staying, Hernandez cornered one of them and said, “When Catholics do something bad, they go to the priest and confess. Now, think that I am the priest and take this opportunity to tell me who was involved in creating the fraudulent data.”
They did not confess then and there. But by the end of his inspection, while at the closeout meeting at the plant, the managers essentially admitted that the company was trying to rush Batamandi into operation to circumvent the FDA’s restrictions on Paonta Sahib.
Hernandez explained to Ranbaxy’s senior executives that he was planning to recommend against approving the plant. The company’s senior vice president for global manufacturing, who’d arrived to oversee the inspection, took him aside. He became “agitated and desperate,” Hernandez documented. The vice president admitted that the plant’s construction had been rushed and mistakes had been made. Over and over, the man pledged to correct anything Hernandez wanted, but begged the investigator not to use the word “falsified” in his report.
The FDA refused to certify Batamandi as a separate plant, and Ranbaxy withdrew its tacrolimus application. It was yet another round in a global game of whack-a-mole. Each time the FDA found fraud inside Ranbaxy, it responded with a small regulatory restriction, only to find fraud pop up somewhere else. The agency had done nothing yet to clamp down on Ranbaxy’s entire method of operating. But the game was about to change.