QuarterWatch™ is an ISMP surveillance program that monitors serious, disabling, and fatal adverse drug events (ADEs) reported to the US Food and Drug Administration (FDA) by manufacturers, health professionals, and the public. The goal of QuarterWatch ™ is to improve patient safety through the identification of 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 of occurrence and establish a causal relationship to the suspect drug. The latest QuarterWatch ™ examines the leading potential drug safety problems reported to FDA during 2012.
In 2012, FDA received 210,648 domestic reports of serious, disabling, or fatal adverse events associated with therapeutic drugs, an increase of 16% from the previous year.
Anticoagulants and Hemorrhages
While 90% of adverse drug event reports are submitted to FDA by drug manufacturers, cases that health professionals and consumers report directly to FDA are an important indicator of drug safety issues. For the second straight year, these direct reports identified two anticoagulants, dabigatran (PRADAXA) and warfarin (COUMADIN, JANTOVEN), as the most frequent suspect drugs in serious or fatal adverse events in 2012. A third anticoagulant, rivaroxaban (XARELTO), ranked tenth. The predominant adverse event involved hemorrhage. In 2012, the three anticoagulants together accounted for 6,234 cases of serious injury, including 789 patient deaths when counting both manufacturer and direct reports (Table 1, in the PDF version). As in our previous QuarterWatch ™ analysis, reported adverse events with dabigatran were more likely to result in death (18%) when compared to warfarin (6.5%) and rivaroxaban (7.2%).
The same property that allows anticoagulants to reduce the risk of strokes and blood clots elsewhere in the body leads to a high risk for hemorrhage and other bleeding events. The result is tens of thousands of emergency department visits or hospitalizations each year. It is also of concern that 2 years after the beginning of a large volume of reports of hemorrhages with dabigatran and other anticoagulants, FDA has discounted the adverse event data and focused narrowly on the question of whether dabigatran has a higher risk of adverse events than warfarin.
Unlike warfarin, dabigatran and rivaroxaban do not have an antidote to rapidly reverse the anticoagulant effects of treatment, raising questions about how to manage hemorrhages that occur either as an adverse effect or as a result of an accident or traumatic injury. Because dabigatran-related hemorrhages have been reported primarily in older patients (median age 80) and are more likely to have a fatal outcome, reassessing the one-size-fits-all dose and enhancing the monitoring of kidney function in older patients could reduce the risk of hemorrhage. That FDA and the manufacturer have not recommended such changes remains unexplained.
Anti-TNF Agents and Infection, Skin Cancer, and Nerve Demyelination
When a drug manufacturer hears about a serious adverse event for which adequate warnings may not exist, the company is required to submit an expedited report to FDA within 15 days. (Other adverse event reports can be submitted quarterly or annually.) In 2012, three anti-tumor necrosis factor (TNF) agents—adalimumab (HUMIRA), etanercept (ENBREL), and inFLIXimab (REMICADE)—ranked first, second, and third in regards to manufacturers’ expedited reports to FDA with 2,962, 2,958, and 2,623 reports, respectively. To produce this volume of serious, expedited reports in 2012, unequalled by other drugs, requires a combination of three factors: extensive patient exposure, higher risks of serious side effects, and active manufacturer reporting.
These anti-TNF agents are approved for uses that include rheumatoid arthritis, ulcerative colitis, Crohn’s disease, and psoriasis. All together, and including other types of reports in addition to expedited cases, the three anti-TNF agents accounted for 11,215 serious adverse events, including 3,324 infections, 342 skin cancers, and 119 cases of nerve fiber demyelination. Because TNF assists in the immune response to infection, blocking TNF introduces an increased risk of viral, fungal, and bacterial infections, which accounted for 38.9% of the reports regarding adverse effects with these three drugs. TNF also has a role in cancer prevention, and blocking it raises fears of increased cancer risk. Although cancer rates were only modestly elevated in long-term clinical studies with these anti-TNF agents, the 2012 reports show unexpectedly large numbers of skin cancers reported for all three drugs.
FentaNYL Patches and Medication Errors, Product Problems
Safety issues with a drug may arise for two reasons unrelated to its potential adverse drug reactions profile: medication errors and product problems (e.g., contamination, manufacturing defect). In 2012, a signal for these types of safety issues was identified with fentaNYL patches. This product had a total of 1,890 reports of serious injury in 2012 of which 1,148 (60.7%) indicated a medication error and 802 (42.4%) described product problems. Reports often involved both categories.
Medication errors with fentaNYL patches included mostly prescribing errors (48 cases) and patient self-administration errors, including omitting a dose (130 cases), putting the patch on an inappropriate site (115 cases, absorption varies at different body sites), leaving the patch on too long (40 cases), and accidental exposure to the patch by children, caregivers, or others (33 cases), which has led to fatalities. In September 2013, FDA reported it had learned of two more accidental exposure deaths in children during the past 18 months, both cases originally reported to the ISMP National Medication Errors Reporting Program. The product-related quality problems with the patches were vaguely characterized in the reports as a “product adhesion issue,” a “product quality issue,” or “poor quality drug.” Another group of fentaNYL adverse reactions were those expected of a potent opioid: withdrawal (158 cases), overdose (83 cases), cardiopulmonary arrest (32 cases), as well as classic opioid side effects, nausea (85 cases) and vomiting (68 cases).
FDA has taken action to address the safety risks of potent and long-acting opioid drugs such as fentaNYL patches. The agency mandated a Risk Evaluation and Mitigation Strategies (REMS) plan, launched in July 2012, that targets misprescribing, misuse, and abuse. The REMS require the 20 pharmaceutical companies that market long-acting opioids to jointly educate prescribers on the safe use of these drugs, with a goal of training 25% of the 320,000 opioid prescribers in the US in the first year and 60% within 4 years. Prescriber education programs and other tools can be found at: www.er-la-opioidrems.com. However, health professional participation is voluntary because FDA concluded that monitoring compliance would be unduly burdensome. In addition, in July 2013 FDA announced it would be changing the prescribing information for opioids to provide improved warnings for this class of drugs and said it would mandate postmarket studies. In September 2013, the agency sought to reduce the risk of accidental exposure to fentaNYL patches by requesting manufacturers to change the color of the text on patch labels so they can be seen more clearly. This action will also help caregivers more easily find fentaNYL patches on the skin or patches that have fallen off.
Alendronate and Bone Injuries
The use of statistical tools to search for links between drugs and specific side effects resulted in a signal strongly associating the osteoporosis drug alendronate (FOSAMAX) with 1,327 reports of bone and joint injuries in 2012, notably 1,201 cases of femur fractures and an additional 155 reports of osteonecrosis of the jaw. Using a measure called the proportional reporting ratio (PRR), which can identify cases in which a specific drug is disproportionately associated with a particular adverse effect, alendronate was 186 times more likely to be associated with bone and joint injuries (PRR=186, p<0.01) compared to all other drugs after adjusting for the volume of reports. Results were similar for risedronate (ACTONEL), another osteoporosis drug, but the bone injury cases were fewer (n=68) yet still disproportionately more than expected (PRR=27.9, p<0.01).
Alendronate is the most frequently prescribed bisphosphonate. Early studies showed that the drug increases bone density over time and reduces the risk of mostly asymptomatic vertebral fractures. After a decade, however, questions have arisen about long-term bone health: Do these drugs over the long term render the femur and other bones more vulnerable to atypical fractures that may heal slowly? FDA considered duration of treatment in 2010 and concluded that the benefits versus risks after 3 to 5 years of treatment were unknown. FDA required a warning in the prescribing information stating, “Optimal duration of use has not been determined. For patients at low-risk for fracture, consider drug discontinuation after 3 to 5 years of use.” While the 1,327 cases of bone injury reported in 2012 cannot determine whether long-term use of alendronate is beneficial or harmful, they do highlight a weakness in the system that allows millions of people to be exposed to long-term drug treatment where long-term risks and benefits have not been determined.
The 2012 adverse event report data show, in our judgment, that major priorities in drug safety should include action to reduce high rates of hemorrhages for all the anticoagulants rather than the current FDA focus on differences between the various drugs on the market. On the other hand, FDA has launched an ambitious drug safety initiative to mandate risk management programs to educate physicians about safe use of long-acting opioid drugs, including fentaNYL patches highlighted in this report. The data continue to emphasize that most drugs that suppress the immune system have substantial risks, and anti-TNF agents rank high in that group and require careful patient selection and monitoring. The signal linking alendronate to femur fractures highlights a fundamental flaw in the overall system in which ongoing monitoring of the long-term effects of drugs is not included or required in the initial approval process. To obtain new drugs more quickly and at lower cost, long-term benefits and risks are hard to determine before approval. But in this case, 18 years after marketing approval, the optimum duration of treatment with alendronate is still unknown, and in some large patient subgroups, a question remains whether the benefits outweigh the risks.
See the full QuarterWatch ™ report, including references, at: www.ismp.org/QuarterWatch/default.aspx.