QuarterWatch™ (2014 Annual Report): Signals for TNF Blocking Agents, Anticoagulants, Lipitor, Growth Hormone
A new edition of ISMP’s QuarterWatch™ provides an overview of drug safety issues reflected in 833,076 adverse drug events reported to the US Food and Drug Administration (FDA) during 2014. This total includes 293,810 fatal, disabling, and serious injury cases in the US. The report identifies:
Two tumor necrosis factor (TNF) blocking agents that accounted for the largest number of serious injury reports—more than 80,000 cases
Ongoing safety concerns about three newer anticoagulants, one of which has a better safety profile than the other two
A cholesterol-lowering medication that has generated the most legal claims, alleging it causes Type 2 diabetes in women
A growth hormone associated with the most frequent adverse events in children
Two anti-TNF products post most injury reports
The two most widely prescribed anti-TNF products accounted for the largest number of injury reports received overall by the FDA in 2014, mostly because of their potent effects on the immune system that increase the risk of opportunistic infections, reactivation of the hepatitis virus, and cancer. The two drugs, adalimumab (HUMIRA) and etanercept (ENBREL), are approved to treat several autoimmune disorders. Adalimumab ranked first with 46,937 new reports, and etanercept ranked second with 38,929 reports. Together, the two were the primary suspect drugs in 1,809 reports of patient deaths. To put it into perspective, in 2014 we identified 1,604 drugs with reports of injuries, with a median of 37 reports per drug. Only 168 drugs had more than 1,000 reports.
Infections, cancer, hypersensitivity reported. Accounting for nearly 1 million outpatient prescriptions in the fourth quarter of 2014, these drugs had moderate exposure in terms of the patient population taking these medications. Among the serious injuries reported, almost a quarter of the adalimumab cases and a third of the etanercept cases indicated an infection, and more than 10% of the adalimumab and etanercept cases indicated cancer, including 197 and 90 reported cancer deaths, respectively. A third large group of serious adverse events involved hypersensitivity. More than 10,000 injection site reactions, mostly non-serious, were also reported for each drug.
Conclusion. These unequalled totals of adverse event reports are a reminder that the prominent warnings about the risk of cancer, infection, hypersensitivity, and other harms are not boilerplate to satisfy legal departments and regulators. These two drugs account for thousands of serious and life-threatening injuries reported each year and many thousands of reports of less severe harm. Because of these significant risks, the drugs are intended to treat autoimmune disorders that are moderate to severe.
Safety contrasts for 3 newer anticoagulants
The adverse event reports involving oral anticoagulants in 2014 confirm that long-term use remains one of the highest-risk drug treatments in older patients, with injury rates of 15 to 20% per year. Rivaroxaban (XARELTO), dabigatran (PRADAXA), and apixaban (ELIQUIS) are newer oral anticoagulants approved between 2010-2012 and marketed as easier-to-use alternatives for warfarin (COUMADIN). As previously noted in QuarterWatch™, a new generation of oral anticoagulants entered the market claiming ease-of-use based on clinical trial designs that postulated the idea that periodic blood tests to establish an optimal dose were not required. Yet, two of the three newer anticoagulants have pharmacological profiles that raise questions about their suitability for simple, unmonitored, dosing regimens.
Rivaroxaban. Rivaroxaban is the only newer anticoagulant with once-a-day dosing. However, it has the shortest half-life—just 5 to 9 hours. Thus, rivaroxaban adverse events reported during 2014 were notable in one expected area: an excess of embolic-thrombotic events compared to the other two drugs. It also had the largest number and percentage of blood clot-related events, 33.9% compared to 20.1% for dabigatran and 22.1% for apixaban. The substantial drug peaks and troughs could be reduced with twice-a-day dosing, but it was not tested in the pivotal long-term use trial. The short half-life of once-a-day dosing also results in higher maximum drug concentrations during drug peaks, and thus, increases the risk of bleeding.
Dabigatran. Dabigatran had the highest overall total of domestic, serious adverse events reported among the three newer anticoagulants, the largest total of reported severe hemorrhages, and the most patient deaths in 2014. The differences persisted after adjusting for patient exposure and characteristics. Previously, we questioned whether a drug with a 5-fold variability of effect between patients getting the same dose was suitable in a single therapeutic dose without blood-level monitoring. The 2014 data further illustrate our concerns. Also, pharmacokinetic-pharmacodynamic data suggest that about 17% of patients taking dabigatran have subtherapeutic anticoagulation and therefore minimal protection against a stroke or heart attack. Nearly half receive more drug than needed, risking hemorrhage. The mortality rate for dabigatran events, at 20.9%, was about double that of the other two new oral anticoagulant drugs. The highest excess anticoagulation was seen in older patients. A drug-level blood test and a reduced dose (110 mg)—options available in Canada, Europe, and elsewhere outside the US—were not approved in the US; FDA was concerned that too many doctors would worry about bleeding and use the lower dose inappropriately.
Apixaban. At this point, apixaban appears to have avoided the drawbacks as seen with the other two anticoagulants, and shows the strongest safety profile. After adjusting for exposure, 4.4 serious injury reports per 1,000 person-years were reported for apixaban, compared to 6.6 for rivaroxaban and 14.1 for dabigatran. Apixaban’s twice-a-day dosing regimen is consistent with its 12-hour half-life. In its long-term pivotal trials, apixaban use led to fewer cases of severe hemorrhage than warfarin (2.1% vs. 3.1%). In the 2014 data, the drug accounted for the fewest reports and the fewest patient deaths both before and after adjusting for patient exposure. Yet, the risk of bleeding is still high, and left unanswered is whether apixaban safety could be further improved by individualizing the dose for each patient, as with warfarin.
Concomitant use of platelet inhibitors. In the 2014 adverse event data, we found that concomitant therapy with platelet inhibitors while taking anticoagulants increased the odds of a hemorrhage event by threefold (OR 3.01 p < 0.01). The increased risk was found across all three of the newer anticoagulants and warfarin. However, patients on combined therapy had no greater risk of death (p = 0.861) and a reduced risk of a blood clot/treatment failure event (OR 0.64 p < 0.001)
Conclusion. FDA needs to take action to reduce the bleeding risks of dabigatran by approving a lower dose (110 mg) for older patients and requiring drug-level blood tests to identify patients with subtherapeutic or excessive blood levels. The safety and efficacy of once-a-day dosing of rivaroxaban compared to twice-a-day needs to be reassessed. Better guidance is needed regarding concomitant use of anti-platelet agents other than a simple warning that increased risk was observed. It is time to move towards individualizing the dose for all long-term anticoagulant therapy and drug-level blood monitoring when indicated.
Statins and diabetes
In 2014, the most frequent litigation target was atorvastatin (LIPITOR), and the issue was whether it causes Type 2 diabetes in women (98% of the reports involved women). Atorvastatin accounted for 4,727 reported legal cases, far more than any other therapeutic drug. Some event reports indicated known complications of diabetes, such as damage to the kidneys, vision, and nerves. However, these cases are allegations that have yet to be proven.
Benefit for low-risk women questioned. A recent reexamination of 13 large clinical trials of all statin drugs concluded that treatment might increase the risk of diabetes by around 9%. This was a concern but 9% seemed a small number compared to a 36% reduction in risk of cardiovascular events. But the statin trials had largely enrolled men. Observational studies in women showed that the risk of diabetes with statin treatment was much higher at 48% in the largest study. And women had a lower risk of cardiovascular disease compared to men. The studies suggested that all the statins shared roughly similar risks and benefits, although both might be higher for the most potent statins, rosuvastatin (CRESTOR) and atorvastatin.
Conclusion. To discover, after 20 years, that one of the most widely used drug treatments in medicine might do more harm than good in a huge subgroup–low-risk women–underscores the limited data that support long-term use in this population. When a drug has a relatively small chance of providing a future benefit, even a small risk of harm can alter the balance of risk and benefit. Women have a lower risk of cardiovascular disease than men, and if proven to have a 2 to 3 times higher risk of developing diabetes, guidelines for treating women with cholesterol-lowering drugs need to be reassessed.
Somatropin and children
Among children under 18, somatropin (recombinant human growth hormone) accounted for the most domestic, serious adverse events (n=232) in that age group. The median age was 13 years, with one-quarter 9 years old or younger and one-quarter 17 to 18 years. The reports included 103 cases requiring hospitalization, 72 cases of growth retardation, 44 cases of abnormal bone development, and 32 cases with an outcome of death. Somatropin was originally limited for use in children who were proven to be deficient in growth hormone, or had other rare disorders that resulted in short stature. But in 2003, FDA approved its use in children who are short in stature (for unknown reasons). At the time, it triggered a debate about whether somatropin should be used as a “lifestyle” drug because taller children might have higher self-esteem or increased social acceptance than shorter children.
Conclusion. Even though human growth hormone has been available for more than 25 years, the data about benefits and risks are limited. The benefit of accelerated growth is hard to measure, and although treatment typically lasts several years, late onset adverse events are particularly difficult to assess. These data illustrate the need for additional information about this hormone.
The full QuarterWatch™ report, with references, can be found here.
What is QuarterWatch™?
QuarterWatch™ is an independent ISMP surveillance program that monitors serious, disabling, and fatal adverse drug events reported to FDA by manufacturers, health professionals, and the public. The goal of QuarterWatch™ is to identify signals that may represent important drug safety issues. The term signal means evidence judged to be substantial enough to warrant publication but which requires further investigation to determine its frequency and establish a causal relationship to the suspect drug.