Randomized control trials (RCT) are complex and very expensive studies required by regulatory agencies, in this case the FDA, before a prescription drug is allowed to be touted for its benefits and marketed. The cost of a single trial, often with thousands of patients enrolled at as much as $5,000 per patient (payments are made to the institution or the physician and not the patient), can run into hundreds of millions of dollars. A study that shows no benefit for a drug can be a huge financial loss for the pharmaceutical company or government agency that sponsored the study. For most drugs, especially those for which a relatively small benefit over a generic or placebo is expected, the more patients enrolled in a trial, the more likely a small benefit can be obtained and therefore pass the qualifying test: a P-value of 0.05 or less. That is the ultimate goal of any study, and shows that the results of the trial are 95% likely to be true rather than a matter of mere chance. Big trials, however, come with big costs.
As the billion dollar drug, Plavix (clopidogrel), marketed by Bristol-Meyers Squibb, was approaching its patent expiration date, AstraZeneca released their results of the PLATO trial: a trial that compared their new drug, Brilinta (ticagrelor), with Plavix, in a classic, double-blind, randomized trial. With 18,624 patients and a substantial improvement in outcomes compared to patients who took Plavix, the drug won FDA approval. As Plavix finally became generic, and far less expensive, the leading societies, including the American Heart Association, the American College of Cardiology, and the European Society of Cardiology, gave Brilinta preferred use over Plavix.
But from the start, impartial physicians who reviewed the PLATO data raised alarming concerns with the results. The enormous benefit of Brilinta vs. Plavix found in two countries, Poland and Hungary, were responsible for much of the total favorable outcome. Yet strikingly, among the 1413 patients from the U.S. randomly assigned to Brilinta, there was a 27% HIGHER risk of vascular deaths, heart attacks and strokes compared to patients given Plavix. There were other flaws in the study as well. The study was designed as double blinded, yet anyone who cut into the capsule could clearly see if the patient was taking Plavix (there were two halves of Plavix pills within the capsule). Does that not leave an opening for fraud? And even more troubling, the original results of the trial appeared less promising until a post hoc, partial recount of the study added an extra 45 heart attacks to the Plavix group, which according to a physician who has sat on FDA panels, is quite unusual.
In spite of these alarming issues, and the great disparity in cost (a one month supply of Brilinta costs between $280.00 – $300.00 vs. a one month supply of Plavix easily purchased for under $10.00 dollars at Walmart and numerous other discount pharmacies), doctors have been prescribing Brilinta in increasing numbers. Some physicians seem unaware that the trail data from Eastern European countries may not be honest and that North Americans actually had a higher cardiac event rate when they took the more expensive Brilinta instead of Plavix. While some may enjoy the benefits of prescribing Brilinta and perhaps even becoming a paid speaker for AstraZenica, patients may be paying the price. In the end, writing prescriptions for Brilinta instead of Plavix actually has the stamp of approval from all the major cardiology societies who recently approved Brilinta over Plavix in patients undergoing coronary stenting procedures.
But how is it that Brilinta ultimately won out over Plavix? Is this because the drug is better, or perhaps because of poor data collection, or a convenient manipulation of study data from centers somewhat immune from close FDA and media scrutiny? Apparently the Department of Justice was interested in finding out; they announced, in October 2013, that they were investigating AstraZeneca’s PLATO trial.
In August of 2014, AstraZeneca announced that the U.S. Department of Justice had closed its investigation. Though the DOJ cleared the company of any wrongdoing in conducting the PLATO trial, they gave no explanation for the irregularities that were uncovered. The end-date of the trial was adjusted to exclude sixteen cardiac events (heart attacks, etc.) and there was an incomplete accounting of some “twenty-six out of 30 unreported events following early drug discontinuation.” Perhaps the DOJ just didn’t have enough proof to make such a damning charge, but it certainly leaves many still questioning the results, questioning why Brilinta became a preferred drug, compared to generic costing 5% of the price, and wondering why those enrolled in the United States were better off if they received the cheaper Plavix instead of Brilinta? Meanwhile, Brilinta sales have been rising dramatically from $24 million in the third quarter of 2012, to over $500 million dollars for 2014.
More Troubling Results Involving Trial Data from Eastern European Countries
For years physicians have searched for an effective treatment for patients with what is known as HFPEF – heart failure with a preserved ejection fraction. Mostly elderly, they are often found to behave like classic heart failure patients, even when their hearts pump out a normal amount of blood with each beat. One trial, known under the acronym TOPCAT (Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist), and funded by the National Heart, Lung, and Blood Institute (NHLBI), aimed to show whether or not a generic drug, spironolactone, was effective in treating this disorder.
The results of this trial, published in 2013, did not show any significant reduction in the primary outcome, a composite of cardiovascular death, aborted cardiac arrest, or hospitalization for heart failure, for the drug spironolactone vs. placebo. It appeared from these results that spironolactone had no benefit in treating these HFPEF patients.
But late in 2014, a new post hoc analysis of the trial showed something quite startling. The NHBLI had funded clinics in Russia and the Republic of Georgia to also enroll patients for TOPCAT. Results there markedly conflicted with results from patients enrolled in North America. The event rate for the patients treated with placebo in Russia and the Republic of Georgia, in spite of their less sophisticated healthcare system, were so low (one quarter of the event rate in controlled subjects compared to North American patients), that some questioned if some of the patients enrolled had even had heart failure. The expected adverse effects seen with spironolactone: high potassium and renal insufficiency, were mostly seen in the group treated in the Americas, yet hardly noted in patients treated in Russia and Georgia, even as the Russian and Georgian investigators claimed their patients had a much better compliance with medications than those in America. Could it be that doctors, starving for income, enrolled as many patients as they could in order to collect the payments from NHBLI, even if those patients hadn’t had heart failure, and possibly never even bothered to administer the drugs?
In late 2014 these post hoc results were reported at the American Heart Association meetings. TOPCAT patients, enrolled from the Americas, had a 26% reduction in cardiovascular mortality. Like PLATO, the patient studies in North America had a completely different result than those studied in Eastern Europe. But unlike Brilinta, the generic, spironolactone, didn’t have a huge corporation pushing the results or physicians paid by that corporation, arguing its merits.
One of the lead authors of the TOPCAT Trial, Dr. Marc Pfeffer, in a recent interview conducted by the American College of Cardiology (ACCEL June, 2015) was questioned about the results from these eastern European countries “is it fraud?” the interviewer asked him. Dr. Pfeffer responded, “I don’t think we can say what happened. It may be some investigators may have been inexperienced… we can’t say it’s fraud, we don’t know. But it is striking that when they got spironolactone in Russia and Georgia we don’t see the response we expected so what happened there we honestly don’t know. We’re perplexed.” He goes on to say that “rather than use this F-word [fraud], which doesn’t help anybody and certainly doesn’t help our patients, we should focus on the results of the Americas.” When asked by the interviewer if he knew who the CRO was (the CRO is the Contract Research Organization that provide all the support for the clinical study) for Russia or the Republic of Georgia, Dr. Pfeffer responded, “I don’t know.”
Two different trials, one by Big Pharma and one by a United States agency, allowed enrollment systems that may be very difficult to vet or trust, yet were willing to provide enough patients to ensure that the study population met the goals of the study designers. Talking to a researcher familiar with the process I learned that it is indeed much easier to find study patients in poorer countries and physicians who are willing to accept less funding to enroll them.
These two large trials also create disturbing concerns: might physicians in less stable countries, who are desperate for income, lie about their patients and their results? Has the NHBLI, in an effort to save dollars, published a study that is corrupted and untrustworthy? Can Big Pharma not only reduce their cost to fund large trials but also covertly manipulate the results by allowing enrollment of patients in foreign countries that are not accessible for an adequate vetting process or the eye of the media? Could it be that some of our leading researchers are just too hesitant to call out what appears to be fraud, even as they seemingly have no reasonable explanation for some of their results, except for duplicitous activity by researchers?
I’m asking the FDA and other regulators as well as the media to interview the nurses, doctors, and patients that were involved in these studies and find out if what we have here is fraudulent data that could affect millions of lives and cost billions in healthcare dollars that we cannot afford to waste.
I will continue to prescribe generic Plavix to my patients and I would not hesitate to accept Plavix as a treatment if I suffered a heart attack.
About the author: Evan S. Levine, MD FACC, is Director of the Cardiovascular Center at Saint Joseph’s Hospital and a Clinical Assistant Professor of Medicine at Montefiore Medical Center – Albert Einstein College of Medicine. He is also the author of the book “What Your Doctor Won’t (or can’t) Tell You”. He lives in Connecticut with his wife and children.
