66 - Adverse Events Associated with Analgesics Used for Osteoarthritis Pain: Analysis of Post-Marketing Data
Raveendhara Bannuru1, Mo Dimbil2, Wenhui Wei3, Susan Colilla4, Clotilde Huyghues-Despointes3, Joanne Nettleship4, Ravi Iyer4
1Tufts Medical Center, Boston, MA, USA. 2Advera Health Analytics, Santa Rosa, CA, USA. 3Regeneron Pharmaceuticals, Tarrytown, NY, USA. 4Teva Pharmaceutical Industries, Frazer, PA, USA
Purpose Several pharmacological interventions are available for treating patients with osteoarthritis (OA) pain. Whilst each of these therapies is associated with potential adverse events (AEs), those risks have not been comprehensively quantified in the real-world setting. Real-world data for AEs associated with OA analgesics may be useful to patients and healthcare providers to better assess treatment risk when considered within the context of an individual’s medical history and existing comorbidities. The aim of this study was to analyze post-marketing data for AEs associated with analgesics most commonly used for OA pain. Methods Data were obtained from the US Food and Drug Administration Adverse Event Reporting System (FAERS) database between January 1, 2001, and June 30, 2019. We identified the known labeled risks associated with nonsteroidal anti-inflammatory drugs (NSAIDs), Drug Enforcement Administration (DEA) class II/III opioids, tramadol, and acetaminophen from the boxed warning, warnings, precautions, and adverse reactions sections of the drug labels for these agents. Based on this research, lists of custom Medical Dictionary for Regulatory Activities query searches were developed to identify AE cases. Cumulative case counts on primary suspect outcomes (i.e., cases in which the reporter listed the specified drug as the primary suspect cause associated with the AE) were calculated, and all cases were required to have OA as a reported condition. The reporting odds ratio (ROR) approach was used to compute a measure akin to the odds ratio to quantify the strength of the association between the specific aforementioned analgesic classes and AEs reported in a post-marketing safety database such as FAERS. RORs were calculated by dividing the odds of a specific AE occurring with the specific analgesic used for OA pain by the odds of that AE occurring with all drug classes (non-stratified) or only drug classes with OA as a reported condition (stratified). RORs greater than 1 indicated an elevated association between the specific drug class and the AE compared with the other drug classes, whereas RORs below 1 indicated lack of such an association. The limits of the 95% confidence intervals of ROR were obtained via an approximation of the normal distribution. The lower limit of this distribution (ROR05) was calculated and reported for both the non-stratified and stratified RORs, providing 95% certainty that the true mean of the population was at or above the number reported. Results During the study period in the FAERS database, the cumulative primary suspect case counts, with OA being a reported condition, were as follows: 7,128 for NSAIDs, 938 for DEA class II/III opioids, 296 for tramadol, and 149 for acetaminophen. For primary suspect AE cases overall, non-stratified ROR05s were greater than 1 (indicative of higher odds of AEs compared with the other drug classes) for 13 of the 16 expected safety issues for NSAIDs, 7 of the 18 for DEA class II/III opioids, 4 of the 18 for tramadol, and 1 of the 16 for acetaminophen. ROR05s were the highest for gastrointestinal (GI) ulceration/perforation (9.67), GI bleeding (8.24), and myocardial infarction (3.09) for NSAIDs; withdrawal symptoms (9.43), sedation (4.58), and drug abuse/dependence (2.13) for DEA class II/III opioids; withdrawal symptoms (2.92), GI obstruction (1.73), and drug overdose (1.22) for tramadol; and hepatotoxicity (7.37) for acetaminophen. Stratified ROR05 results showed similar trends but with different magnitudes. For primary suspect AE cases, when comparing to other drugs with OA as a reported condition, stratified ROR05s for NSAIDs were 2.72 for GI ulceration/perforation and 2.28 for GI bleeding. For DEA class II/III opioids, stratified ROR05s were 10.79 for withdrawal symptoms, 6.38 for sedation, and 2.56 for drug abuse/dependence. For tramadol, an association was demonstrated for seizures (1.42) when comparing to other drug classes with OA as a reported condition, which was not apparent in the non-stratified results. For acetaminophen, the stratified ROR05 was 6.58 for hepatotoxicity. Of all studied primary suspect cases with OA as a reported condition, serious AE cases (defined as death, life-threatening AEs, hospitalization, disability/permanent damage, congenital anomalies/birth defects, AEs requiring intervention to prevent permanent impairment/damage, or other serious AEs) comprised 55.7% (83/149) of all studied AEs with acetaminophen, 63.8% (598/938) with DEA class II/III opioids, 75.4%, (5,371/7,128) with NSAIDs, and 90.5% (268/296) with tramadol. Conclusions This study provides a comprehensive and quantitative evaluation of the safety issues with drugs used to treat patients with OA pain in the real-world setting. The results of our study demonstrate that the most commonly used OA analgesics are associated with elevated risks for certain AEs compared with the other drug classes, underscoring the need for new and safer therapies for OA pain. These data will help patients and healthcare providers to better assess the benefit-risk profiles of these analgesics before incorporating them into treatment regimens. This study was limited by the lack of control for other confounding factors and the potential for use of multiple medications in individual patients. Further research is required to evaluate the downstream effects of the studied AEs (e.g., medical costs) and to benchmark these AEs against the safety profiles for the evaluated analgesics reported in randomized clinical trials.