Chapter 22

The $600 Million Jacket

2011

Silver Spring, Maryland

Scattered across the desk of Karen Takahashi, a meticulous FDA compliance officer, lay Ranbaxy’s most valuable asset: its application to launch the first generic version of Lipitor. The controversial jacket was stamped ANDA 76-477, and it would fall to the soft-spoken Takahashi and her colleagues in the FDA’s international compliance branch to review it. The clock was ticking.

In the generic drug world, nothing was more lucrative than atorvastatin. Each year, U.S. government programs alone spent $2.5 billion on brand-name Lipitor. Any delay in launching a generic version could cost Americans up to $18 million a day, a group of U.S. senators reminded the FDA commissioner in a March 2011 letter. Ranbaxy was first in the FDA’s queue to make atorvastatin and, owing to a settlement with Pfizer, could legally start selling the drug by November 30, 2011. All the company needed was final approval from the FDA. But Ranbaxy’s mounting problems had turned the once seemingly surefire launch into a nail-biter. “For Ranbaxy, this is the fight of their lives,” a lawyer for a pharma company told Fortune magazine. “This is the biggest generic opportunity in history. None of us know where this is going to come out.”

Inside the FDA, despite murky and chaotic deliberations over Ranbaxy’s fate, it seemed clear to some that the agency would never allow a company so saturated in fraud to keep exclusive rights to launch the nation’s most important generic drug. Ranbaxy’s first-to-file application seemed “unsalvageable,” a word that had emerged from government huddles. “This application will never be approved,” Debbie Robertson had told a prosecutor in mid-2010. As the FDA’s deputy director of the Office of Pharmaceutical Science wrote in an internal memo, “To point out the obvious . . . an approval of Ranbaxy’s ANDA 76-477 may not seem in harmony with the agency’s regulatory actions against Ranbaxy.” One way or another, the company would be forced to forfeit its application, and some other company in the queue would step in. “Unfortunately for Ranbaxy, choices are very bleak,” Dinesh Thakur jotted down during a discussion with his lawyer, Andrew Beato, in early 2010. “Even if they go to court, they don’t have much choice. FDA is not going to sign off on Atorva.”

Publicly, the agency said nothing about its deliberations. But as Takahashi and her colleagues delved into the application, fully expecting to stop it, the agency found itself in a quagmire of evidentiary, regulatory, procedural, and bureaucratic problems, a number of the FDA’s own making. The battle with Ranbaxy had become so convoluted that many of the regulators waging it barely even understood it themselves. They joked bitterly about needing RANBAXY—BE STRONG bracelets. But they soon realized that they faced a serious checkmate: if the FDA did not approve Ranbaxy’s atorvastatin, it was not clear when—or even if—Americans would get access to a low-cost version.

Ranbaxy had filed its atorvastatin application in August 2002. Under rules in place then (which changed months later), unless the FDA revoked Ranbaxy’s application owing to a specific finding of fault, the company could simply park its exclusive rights. Competitors would pile up, unable to launch, and the public would get no price relief. As one FDA lawyer explained the rules to his colleagues, “As long as Ranbaxy is not able to sell generic Lipitor, no one can sell generic Lipitor.”

Ostensibly, the FDA had plenty of cause to shut down Ranbaxy’s bid for six months of exclusive atorvastatin sales. The company had committed extensive fraud and needed to be punished. But in the topsy-turvy world of the case, an argument had begun seeping into the government’s internal deliberations. Because the generic drug company operated on razor-thin margins, it needed atorvastatin profits in order to pay the record-breaking fine it deserved. Unless Takahashi found irrefutable proof of fraud in the application, Ranbaxy might just get away with a blockbuster launch of atorvastatin.

Takahashi systematically reviewed Ranbaxy’s application from the beginning. Since 2002, individual reviewers had flagged anomalies: missing electronic data, discrepancies in resubmitted data, impurity data that made no sense. As the application crept through review, Ranbaxy had blamed rounding errors, copying or calculation blunders, and even its own lax laboratory protocols for the glitches. Company executives had claimed that test dates were inconsistent because the company had been inconsistent in what it taught its analysts.

But agency reviewers had found numerous “unexplainable” discrepancies between the original application and amendments that Ranbaxy had filed in 2007. Some test results changed significantly. Impurities suspiciously decreased. Pills originally described as “white” became “off-white.” In several instances, tests originally reported as “out of specification” now fell “within specification.” The changes suggested that portions of data had been entirely false, in total disarray—or both.

These were only the small problems. There were more ominous signs that data had been compromised. The mysterious 4 degree Celsius refrigerators, which the FDA suspected were being used to artificially slow the degradation of test samples, had held bottles of atorvastatin. Takahashi learned that in October 2007 an informant notified the FDA that a Ranbaxy vice president had faked atorvastatin records just before the FDA inspected Paonta Sahib earlier that year. In the summer of 2008, while inspecting documents from Paonta Sahib, the FDA learned that Ranbaxy had been throwing out failed stability tests for atorvastatin and other drugs, then retesting the medicines until they passed and reporting only the successful tests. An inspection of Ranbaxy’s research and development laboratory in Gurgaon, India, in the spring of 2009 uncovered similar deceptions. Takahashi hadn’t found proof yet that the atorvastatin application itself had been compromised, but the chances of it being uncompromised were almost zero.

The agency had had a surefire way to shut down the application—and failed to. In February 2009, it had leveled a rare Application Integrity Policy against the company, which compelled Ranbaxy to prove its products weren’t fraudulent in order to get them approved. Takahashi and her colleague down the hall, Doug Campbell, had written memo after memo arguing that the punishment should be imposed on the entire company.

Remarkably, the FDA cracked down on Ranbaxy with one hand and opened an escape hatch with the other. That month, the agency announced that it would impose the AIP on only one plant, Paonta Sahib. Although this put the brakes on eighty-five drug applications, including atorvastatin, it also left Ranbaxy free to shift its most lucrative applications out of the embargoed plant to other manufacturing sites, so long as it submitted fresh data.

In December 2009, ten months after the FDA imposed the AIP on Paonta Sahib, Ranbaxy filed new amendments to its atorvastatin application. After settling its patent lawsuit and making a deal with Pfizer, it proposed using Pfizer’s active ingredient and moving production to the Ohm Laboratories plant in New Jersey, which had a comparatively clean regulatory record. Even high-level bureaucrats struggled to understand why the agency would suddenly permit the company to shift manufacturing sites. “From what I know Compliance did not want us to permit this because Ranbaxy could use this tactic as a simple way to circumvent AIP,” the chief of the regulatory support branch for the Office of Generic Drugs wrote to his colleagues. Yet somehow the FDA had authorized the shift. The move left Takahashi and her colleagues with yet more work: they had to sift through all of Ranbaxy’s amendments to ensure that fraudulent data from the embargoed plant wasn’t being reused.

In April 2010, Takahashi began to follow another important lead. The AIP required Ranbaxy to hire an outside auditor to verify that the data in its applications was accurate. Ranbaxy had chosen Quintiles Consulting, which was obligated to report its findings directly to the FDA. In November 2009, Quintiles auditors went to India to examine the atorvastatin application and ended up notifying the agency of a “very hot” finding, as Campbell relayed to his FDA colleagues. The original records for the dissolution data submitted in 2002 were nowhere to be found, and the raw data on hand didn’t match the data submitted to FDA. The situation threatened to upend Ranbaxy’s application. The FDA required that applications be “substantially complete” when submitted. This was to prevent companies from submitting sham applications simply to be first, before even finishing their studies or generating complete data. But the absence of original records meant that the data had either been lost or never existed in the first place. If the tests had been faked, it could mean that the original application had been incomplete, a disqualifying lapse. At the FDA’s Office of Criminal Investigations, Debbie Robertson flagged the missing data for prosecutors. “Sound familiar?” she quipped.

Quintiles’s report made clear that unless the original data could be located, Ranbaxy might have to repeat tests with new ingredients and a new clinical study. And this would certainly raise the question as to whether Ranbaxy really was the first generic company to have submitted a completed application. Should the company keep its place at the front of the line or move to the back?

In late May 2010, the FDA’s criminal investigators learned that Ranbaxy’s newest CEO, Atul Sobti, had summoned a top Quintiles official, Bob Rhoades, to India the previous November over the Thanksgiving weekend and demanded to know why he would report such egregious failings to the FDA without first consulting the company. “He was not happy,” Rhoades recalled of the meeting, where Sobti dressed him down. Sobti later expressed surprise “that an interaction that was just part of a long process is being given not only an exaggerated, but an inaccurate, view.”

But inside the FDA, officials learning about the meeting viewed it as an apparent effort at obstruction by Sobti. It “should be the absolute last straw,” Debbie Robertson’s supervisor declared to an FDA lawyer. “The CEO no less. Not only is the culture of fraud and deceit still alive and well at Ranbaxy—it exists all the way to the top. All bets should be off.”

Few at the FDA disagreed. Meanwhile, Takahashi returned to the question of the missing data. What did it mean? And what could the FDA do about it? “Is it written anywhere that the raw data for a drug application must be kept forever?” the director of the Office of Compliance’s Division of Manufacturing and Product Quality emailed his staff. Unfortunately, it wasn’t. As the regulatory counsel for the Office of Compliance noted, “We have not, as of yet, been able to point to a regulation that says they have to keep the data forever.”

On May 5, 2010, a group of almost a dozen FDA lawyers, regulators, and criminal investigators met at the agency’s headquarters with prosecutors and Justice Department officials to discuss the sprawling effort to settle the case against Ranbaxy. With every imaginable punishment on the table—from prosecuting individual executives to debarring the company to imposing a record-breaking fine—the conversation kept returning to atorvastatin.

“What’s in the best interest of the government?” Douglas Stearn, assistant director of the FDA’s Office of Compliance, asked those gathered. He noted that Ranbaxy had filed suspicious amendments for its atorvastatin ANDA, and that the FDA should not be rewarding the company with approval but imposing a heavy fine instead. “Big numbers talk.”

Stuart Berman, the assistant U.S. attorney for Maryland, pointed to the remarkable statement made to Quintiles auditors by a vice president at Ranbaxy’s research headquarters in Gurgaon: “Recording and verifying data contemporaneously with data generation is not a cultural trait in India. We need to learn this as a habit.” It seemed irrefutable, said Berman, that “every bit of data received [from Ranbaxy] is junk.”

“Why are we working so hard to keep Ranbaxy in business?” Steven Tave, the FDA’s associate chief counsel for enforcement, chimed in. “Why not take a broad sweep—take everything on the market from 2006 on and pull it off?”

Or, said Berman, “put the settlement aside and start indicting people.” The question of a settlement was simply “wagging the dog,” he said, unless they knew how much Ranbaxy could pay.

But as they debated their options, the discussion turned to the implications if the company was denied permission to make atorvastatin. Tave acknowledged that “if we take strong action, we may not have a huge settlement because the company can’t fund it.”

This concerned Stearn. “If we let them have the first-to-file and they get a huge settlement, what does that say?”

As they mulled how to stop Ranbaxy from making atorvastatin and still get a settlement that would send a strong message to the drug industry, Roann Nichols, an assistant U.S. attorney, asked, “Is there a way we can not approve the first-to-file and let them sell the rights?” That way, there would still be money for a settlement.

Berman pointed out that the Justice Department’s position should not be to fund the fine through “ill-gotten gains.”

Without atorvastatin, “we won’t get any money,” said Linda Marks, the senior litigation counsel for the Justice Department’s consumer protection branch, who had skirmished with Robertson over disclosing critical leads to Ranbaxy’s defense team.

“The company has to relinquish the exclusivity totally,” an FDA lawyer, Paige Taylor, interjected. “If the application is tainted, it’s tainted.”

There seemed to be no perfect path to justice. As Nichols concluded, “Someone will have to make a decision. Atorva will hit the fan either way.” That was almost certainly true. Officials at the meeting left with the view that the application was unsupportable, and therefore almost certainly doomed.

At the FDA, the essential compromises of regulating overseas plants were already whittling away the officials’ tough stance. Just three weeks after the May meeting, regulators got a request from Ranbaxy to inspect a plant in Mohali, India, that the agency had not yet approved. The compliance staff understood this to be part of Ranbaxy’s “creative tactics,” an effort to get more plants approved to circumvent the crackdown on Paonta Sahib. By now, Ranbaxy had been under criminal investigation for almost five years. But the agency responded to the company’s request using the same approach it took for all foreign facilities—by asking for permission to come inspect the plant, weeks in advance.

On August 19, 2010, an FDA investigator wrote to the firm requesting “your availability and the firm readiness for our next visit Oct. 4–8, 2010,” and asked for “your availability to assist in or information on round trip transportation airport/hotel/firm for the Investigators.” A Ranbaxy official wrote back providing more details about the plant, promising to work out hotel reservations, and saying that the company would “arrange the invitation letter as well.” In other words, the company would once again function as a host and travel agent and get six weeks of advance notice to transform the plant for the FDA’s arrival.

Meanwhile, at the international compliance branch, Takahashi continued reaching out to different scientific experts at the agency to see if, somewhere, ANDA 76-477 had crossed some obvious bright line. Each inquiry seemed to loop back to the same place: the application could not be verified as legitimate but couldn’t be proven obviously false. On the morning of March 16, 2011, however, Takahashi had an idea: to stop studying the application and to study the company’s modus operandi instead. She asked Debbie Robertson for a copy of the notes from her division’s interview with Raj Kumar back in 2007. As a witness with impeccable credentials and unshakable integrity, Kumar had been a vital source of information. Now, as Takahashi scoured the notes, one particular falsification that Kumar described grabbed her attention. Kumar told investigators that for two U.S. drug products, the company had taken the chemical profiles of brand-name drugs, and duplicated them to make them appear as though they were for Ranbaxy drug products.

This got Takahashi thinking. The dissolution data that Ranbaxy had submitted in its original 2002 application looked suspiciously similar to the dissolution data for brand-name Lipitor. The company also claimed that it couldn’t find the corresponding raw data for those tests. Perhaps Ranbaxy had cribbed all of Pfizer’s data and passed it off as its own. This possibility had to be pursued. Though something short of a “smoking gun,” it became known inside the agency as a “smoking gun theory.” If true, it would explain why some of the initial data was missing—perhaps because it had never existed in the first place.

The best way to figure out whether Ranbaxy had passed off the innovator data as its own would be to comb through Pfizer’s Lipitor application, patents, and other public sources in hopes of finding chromatograms or other results that perfectly matched the data in Ranbaxy’s application. It was perhaps a mark of Takahashi’s desperation that she thought this was even a doable idea. The task turned out to be monumental. Pfizer’s Lipitor application, which had been approved in 1996, was stored in 220 paper volumes in a government storehouse in West Virginia. Takahashi learned that it would take four to six weeks to get the documents shipped to CDER headquarters in Maryland.

Takahashi was forced to limit her search to public records. Thousands had been available when Ranbaxy filed its application. It was a true needle-in-a-haystack effort that turned up little of significance.

As her options dwindled, pressure was mounting on the FDA to make a decision. Not only had U.S. senators demanded quick approval, but the six other drug companies that had submitted atorvastatin applications were also demanding action. No one knew when—or even if—the FDA would approve Ranbaxy’s application, and competitors worried that the agency’s hesitation would delay their own approvals.

On March 18, Mylan filed a lawsuit against the FDA in federal district court in Washington, D.C., alleging that Ranbaxy should be forced to surrender its exclusivity, owing to its use of false data, and that the “FDA’s indecision is permitting Ranbaxy to maintain a benefit to which it otherwise is not entitled.” As news articles circulated about the suit, an FDA official based in India emailed his colleagues: “These stories are not good in this environment.”

The FDA responded on April 4 with a hard-hitting motion to dismiss Mylan’s suit, arguing that it was under no obligation to disclose confidential deliberations or to help Mylan with its business planning. On May 2, a judge agreed, ruling that one company couldn’t intervene in another company’s application. Investors interpreted the FDA’s stance in the Mylan case as a hint that the agency was planning to allow Ranbaxy to sell atorvastatin in November. Ranbaxy’s shares rose on the news. But no one, including Ranbaxy, knew where the agency’s deliberations would lead. By then, Wall Street analysts had mapped out complex flow charts of potential scenarios and their market impact.

But the compliance staff knew where it should come out. One month earlier, they circulated a memo with questions and possible answers. One was, “Should the government insist that Ranbaxy forfeit exclusivity for Atorvastatin, even if it means a substantial decrease in the settlement’s potentially large recovery?” The answer was, “Yes, it is in the best interest of public health.”

Yet the machinery of the agency kept moving toward approving the drug. In May 2011, Deb Autor’s division, the Office of Compliance, sent a memo to the director of the Center for Drug Evaluation and Research, with a fateful recommendation. The memo urged that the Office of Generic Drugs formally review Ranbaxy’s atorvastatin application and expedite the process under an exception to the Application Integrity Policy imposed on Ranbaxy. It argued that if there was fraud, a review could detect it, an argument that could be easily controverted, if anyone was inclined to argue. But with the U.S. government spending $2.5 billion on brand-name Lipitor each year, there were larger forces at work. The logic, as Doug Campbell later replayed it, was: “‘All we can do is block it, but we’re not going to block it, because we’re all going to save millions of dollars.’”

As the Office of Generic Drugs started formally reviewing Ranbaxy’s revamped atorvastatin application, objections that once seemed insurmountable fell away. By mid-October, the necessary review divisions had signed off on the application. The FDA prepared to announce its approval. By early November, the agency finalized a press release and drafted responses to anticipated questions. It seemed that November 7, just three weeks before the final deadline for approval, would be the big day for the agency to greenlight Ranbaxy’s drug.

But that morning agency officials panicked. They realized—in a major oversight—that the FDA had not conducted a preapproval inspection of a plant that Ranbaxy had now listed as a main source for the active ingredients. Previously, Ranbaxy had told the agency that it would assemble and bottle its atorvastatin at its plant in New Jersey and use active ingredients from Pfizer’s plant in Cork, Ireland. But in late July 2011, Ranbaxy changed this and told the FDA that it would also use active ingredients made at its own plant Toansa, in Punjab, a decision that had bewildered the agency. “Seems ridiculous to me that Ranbaxy decided to use its own API,” the FDA lawyer Marci Norton wrote to Doug Campbell. “Given how high profile this drug is, you’d have thought the smarter course would’ve been to use Pfizer’s API.”

By its own standards, the agency was supposed to inspect every drug manufacturing facility once every two years. The Toansa plant was due for another visit. At an emergency meeting that afternoon, FDA officers decided to delay an announcement and rush an inspection there.

On November 21, nine days before Ranbaxy could potentially start shipping generic Lipitor, two FDA investigators arrived at Toansa to determine two things: whether the plant followed good manufacturing practices, and whether it was capable of safely making the active ingredient for atorvastatin. One of the investigators, Regina Brown, had originally found the unregistered refrigerators full of samples at Paonta Sahib.

As the inspection of Toansa progressed, the regulators back in Maryland waited anxiously for the outcome. On Thanksgiving, Brown wrote back to Campbell, summarizing problems she’d observed so far. She found workers poorly trained and openly using scrap paper while making a batch of atorvastatin, a breach that pointed to a lack of control and potential data manipulation. Fourteen hours later, when she sent him a draft of her findings, it was clear the inspection had deteriorated further. She’d found a shredder in the middle of the production floor, and evidence that records related to batches were being destroyed. “Each scrap paper incident led us into deeper doo-doo,” she wrote to Campbell.

On Friday, November 25, with five days to go until Ranbaxy could start manufacturing atorvastatin and with the FDA’s decision still unmade, Campbell was troubled not just by Brown’s findings but by what he felt to be the agency’s inexorable drift toward approval. He was also running out of time. The Office of Generic Drugs had been pressuring him to submit his team’s official recommendation either for or against approval by the end of the week—and it was now Friday morning. The lawyer Marci Norton, who’d been immersed in settlement negotiations with Ranbaxy’s lawyers, told Campbell to wait until Monday, when everyone would be back in the office. There was no room for mistakes. “This decision likely will undergo heavy scrutiny, perhaps in court and likely by Congress,” she wrote.

Still, Norton worried that Brown might classify the inspection as Official Action Indicated (OAI), the worst rating a drug facility can get. In the face of such a finding, it would make no sense for the FDA to approve Ranbaxy’s atorvastatin application. Campbell had reassured Norton: “I will touch bases with you before anything goes OAI.”

Norton knew that the agency was staring down the barrel of a no-win situation. If it didn’t approve Ranbaxy’s application, criticism would rain down from Congress and patient groups that were demanding an affordable Lipitor alternative. Also, Ranbaxy would likely refuse to settle the larger case if the company didn’t get its atorvastatin approved. But if the FDA did approve the application, it could knowingly be greenlighting a drug made in a subpar plant.

The white-knuckled regulators wanted a clear path to say yes. But less than an hour later, Regina Brown sent in a copy of the final inspection report. Norton’s biggest fear had come true. Brown recommended giving Toansa a flunking grade, or OAI, for failing to use good manufacturing practices. As she summed up in an email, “We saw too much scrap paper in use.” She noted: “It was suspect to even have a shredder there for general use.”

Presumably, since the plant couldn’t function safely, this should have been enough to disqualify its bid to make atorvastatin. But Brown essentially split her vote. In a remarkable decision, she deemed the plant “acceptable” to make the drug ingredients and recommended that the application be approved. She concluded that for the purposes of generic Lipitor, the deficiencies could be overlooked. “The firm has got a lot of experience now” making atorvastatin active ingredients, she wrote. It was, Brown admitted, “an unusual recommendation.” Under pressure, she had essentially rolled the dice: the plant was bad, but it could likely make a good drug.

Regulators could see the outlines of a disaster looming. The plant had failed a review of its good manufacturing practices. It didn’t meet the FDA’s standards. But there were only twenty-four hours left to either green-light Ranbaxy’s generic Lipitor or leave millions of Americans in the lurch.

In an emergency call with the FDA’s India office, regulators weighed the essential conundrum: The agency had found unacceptable procedures but still had faith in the active ingredients. What was the explanation for that? Several regulators wanted more time before they made their decision. Another asked a more fateful question: “Why should the American people want this [active ingredient] from this company in India?” The lawyer Norton chimed in, explaining that unless the agency approved the atorvastatin, Ranbaxy wouldn’t sign a consent decree, a settlement agreement that had been in the works for years. Ranbaxy, despite having committed fraud on a vast scale, seemed to have backed the agency into a corner. She concluded, “The commissioner may have to weigh in.”

November 30, 2011, dawned cold and dry in Washington, D.C. Doug Campbell awoke with a feeling of dread. He did not relish the day ahead. By the time he got to work, media calls about atorvastatin were pouring in. The Office of Compliance had yet to issue its final recommendation for Ranbaxy’s application. Though the inspection’s facts were the facts and presumably could not be changed, Regina Brown’s recommendation had left the FDA with a “narrative” problem: how could the agency justify approving Ranbaxy’s drug if the plant making the principal ingredient couldn’t meet basic manufacturing standards?

The facts could be altered. The Office of Compliance prepared a draft memo for approval in which it changed investigator Regina Brown’s dim assessment of Toansa’s good manufacturing practices from “Official Action Indicated” to “acceptable CGMP status.” It stated that after interviewing the inspection team “at length” and reviewing a draft report, “every subject matter expert consulted during the review process agreed that the facts surrounding the deficiencies do not warrant regulatory action.” To get from never to yes, the agency had changed an inconvenient fact and overlooked obvious warning signs.

By day’s end, the FDA informed Ranbaxy that its ANDA was approved. The scandal-ravaged company had gotten its prize: the lucrative rights to make generic Lipitor exclusively for six months, then alongside competitors after that. At 8:12 p.m., the agency issued a press release announcing the news to the public. In India, Ranbaxy’s latest CEO and managing director, Arun Sawhney, addressed his staff. The U.S. launch of generic Lipitor, he said, was a “historic moment for every Indian and the entire generic industry of the world.” He exulted, “It’s a spirit of never say die,” adding that the story of generic Lipitor had been “nothing short of a thriller” and that, in a competitive world, “Ranbaxy will turn out on top in the Atorvastatin market.” Ranbaxy started shipping the drug the next morning. Based on preorders alone, it generated $100 million of revenue in the first twenty-four hours, and nearly $600 million within six months.

At the FDA, the compliance staff congratulated one another and convinced themselves that, ultimately, they had done the right thing. But a deeper unease about the decision, and the process by which it was reached, settled over the group. And it didn’t take long for the regulators to realize just how badly they’d been manipulated. Ranbaxy had initially reassured the FDA that the drug would be made in the United States under the highest manufacturing quality it could muster. But on December 1, one day after the agency had given the company the green light, Ranbaxy filed a request with the FDA to use a different manufacturing site to make the finished doses of atorvastatin: its plant in Mohali, India. With its approval in place, the company planned to shift manufacturing back to a lower-cost plant with less oversight. It was the same plant where the FDA had asked for an “invitation” to inspect and gave the company six weeks advance notice of its arrival.

Even the regulators well versed in the company’s “creative tactics” were stunned. “What do we know about the Mohali site?” the deputy director of the pharmaceutical science office asked a group of colleagues.

“Well, it’s not in New Jersey!” a regulatory lawyer shot back.

From its seemingly unwinnable position, Ranbaxy had played its losing hand perfectly—and won. “They worked the Agency like a Hammond organ, a fine Stradivarius,” Campbell later said.

Ranbaxy continued to offer the FDA the veneer of compliance. In a March 2012 email, the head of Ranbaxy global quality wrote to Campbell: “In the spirit of transparency & cooperation with the FDA, I confirm that the shredder at Toansa is now confirmed as relocated out of the manufacturing area and that it now resides in the Toansa QA block with continuing controls for its use.” However, in the coming months, the concerns of Takahashi and her colleagues— that Ranbaxy was still concealing fraud and Americans would be getting an inferior drug—would prove dismayingly well founded.