A retrospective descriptive study of the AOMs' adverse events from 2013 through 2020 was conducted using FAERS data in the large US pharmaceutical market. Due to clinical trial studies' inherent limitations, such as limited follow-up duration, stringent design, and relatively small sample size, the spontaneous reporting system (SRS) has been used in pharmacovigilance for safety assessment of a suspected drug’s AEs. Furthermore, SRS plays a integral part in signal identification, and the FDA has developed the FAERS database for postmarketing surveillance to monitor the safety of drugs and improve public health [19]. This study aimed to describe the most common AEs and investigate previously unreported potential AEs associated with using the AOMs.
To our knowledge, this is the first study that has systematically analyzed AEs reported for AOMs located in a nationally representative database for over a 7-year period. Over 73% of the reported AEs were for female adults, and that is consistent with previous studies that analyzed global AE reporting patterns of the AOMs [20], [21]. However, we cannot infer whether this can be explained by the increased risk of the incidence of AEs in this population. Drug ineffectiveness, dizziness, headache, nausea, vomiting, and diarrhea were among the most frequently reported AEs for all reports recorded in the FAERS database through 2002 [22], and that is consistent with our findings. The percentage of cardiovascular and AKF/ kidney injury AEs were sizable among phentermine, liraglutide, and phentermine/topiramate users (Table 2).
For liraglutide, which is the only injectable formulation of the AOMs, many AEs of the type gastrointestinal disorders, cancer, and CVD were found. A relatively high rate of death was associated with using liraglutide compared to the other studied medications. A previously published meta-analysis stated that the discontinuation rate due to side effects was the highest among liraglutide users compared to all FDA-approved AOMs [23]. Additionally, a relatively high percentage of disability reports and common gastrointestinal disorders were at the top of the AEs associated with naltrexone/bupropion medication.
Unlike the other medications, orlistat works by decreasing fat absorption, while the other five medications work mainly through CNS pathways that either enhance satiety or reduce appetite. This explains why a large number of diarrhea, renal complications, and drug ineffectiveness AEs were reported with orlistat. Death represents 4% of the reported AEs attributed to orlistat (Alli®), and since it is the only over- the-counter weight loss pill approved by the FDA [24], the marketing authorization holder should further investigate the fatal cases associated with using this medication.
Phentermine antiobesity monotherapy is approved for adults for 90-day use and contraindicated in cardiovascular disease patients [25]. It is still the most commonly prescribed AOM in the USA and worldwide, except in the European Union due to its potential adverse effects [26]. From the data shown in Table 2, it is apparent that CVD, AKF/kidney injury, paraesthesia, dry mouth, constipation, and insomnia are highly prevalent among phentermine and phentermine/topiramate users, which is consistent with previous studies [27–29]. Figure 3 reveals a considerably high frequency of the congenital anomaly type of hospitalization associated with phentermine/topiramate. A fetal safety issue and teratogenicity concerns related to phentermine/topiramate use was expressed by the FDA owing to the risk of increased oral clefts. Consequently, women of childbearing age are advised on contraceptive planning before using this medication [30–32].
Drug ineffectiveness, dizziness, headache, and dry mouth were among the most common AEs associated with lorcaserin. Compared to the other medications, the CVD-related AEs were relatively low (8%) as the medication showed a low rate of major cardiovascular events in previously published studies [25, 30, 33]. In February 2020, the drug was withdrawn from the market over the potential risk of an increased cancer occurrence. Our analysis revealed that cancer reports denote less than 0.9% of the overall lorcaserin submitted reports and represent 16% of the overall cancer reports associated with the studied medications, considering that not the only primary suspected medications were included. Rekha Kumar and Donna Ryan wondered whether lorcaserin withdrawal, owing to a numerical imbalance in the cancers occurring in a large, 12,000 participants clinical trial, has left more questions than answers. Ascertainment bias was among many concerns that led them to request a full disclosure of more information around the lorcaserin/cancers potential risk association [34].
The majority of serious AEs were found for phentermine, liraglutide, phentermine/topiramate and they were of the type 'cardiac' and 'renal' disorders. The proper use of medication through following the pharmacist's or the physician's instructions can significantly impact avoiding such undesirable effects. To ascertain whether these findings have contributed to detecting new AE signals or merely affirming existing knowledge, conducting comparative studies of the reported AE information with the AEs posted on the official product label are required.
These data must be interpreted with caution because there is no ultimate proof of the causal relationship between exposure to a drug and the reported event. Furthermore, because of the duplicate and incomplete reports (unknown denominator), the incidence of AEs cannot be estimated using FAERS data. Consequently, since we did not limit our analysis to the only primary suspected medications, the values only provide safety signals and do not denote a real risk. Additionally, there is a potential for bias based on physician preference when prescribing the appropriate choice of the AOM for a patient or when reporting the AE. Besides, comorbidities or concomitant drugs might confound the association between a drug and an AE [35]. Also, some medications have other uses that might confound the analysis, particularly liraglutide which is an increasingly common medication used for type 2 diabetes mellitus [36]. To overcome this limitation, we require either the indication to be for antiobesity or using the brand name of the medication. Ultimately, these limitations may lead to an inflation of the risks attributable to the AOMs.