Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive interstitial lung disease of unknown etiology (Spagnolo et al., 2021a). It primarily occurs in older men, and its prognosis is poor, with a median survival of only 2–5 years (Raghu et al., 2011; Jo et al., 2017). Globally, the incidence of IPF is increasing. A recent analysis of 22 studies covering 12 countries found that the adjusted incidence and prevalence of IPF range 0.09–1.30 and 0.33–4.51 per 10,000 persons, respectively (Maher et al., 2021). At present, lung transplantation is considered the only effective intervention to improve the life expectancy of patients with IPF, and it has the advantage of improving both symptoms and survival time (Glass et al., 2022). However, because of the limited number of organ donors and patients’ economic status, age, and comorbidities, only a few patients can undergo this treatment (Glass et al., 2020).

In 2015, an updated guideline of the 2011 version was released by the American Thoracic Society (ATS), European Respiratory Society (ERS), Japanese Respiratory Society (JRS), and Latin America Thoracic Association (ALAT) to focus on treatment interventions. Pirfenidone (PFD, Esbriet) and nintedanib (NDN, Ofev) were both conditionally recommended in this guideline, meaning that physicians should choose different treatment therapies consistent with patients’ values and preferences (Raghu et al., 2011). PFD (5-methyl-1-phenyl-2-[1H]-pyridone) is a novel anti-fibrotic and anti-inflammatory drug that can alleviate the deterioration of pulmonary function in patients with IPF (Noble et al., 2011). It was approved for the treatment of adults with mild-to-moderate IPF in the European Union in 2011 and for the treatment of IPF in the United States in 2014 (Spagnolo et al., 2021b). NDN is a multi-target tyrosine kinase inhibitor that can simultaneously block the fibroblast growth factor, vascular endothelial growth factor, and platelet-derived growth factor receptors (Lamb, 2021). The TOMORROW trial, a 52-week multi-country, double-blind, randomized, placebo-controlled phase II clinical trial, proved that NDN reduces the decline in forced vital capacity associated with IPF (Richeldi et al., 2011). The FDA and EMA approved NDN for the treatment of IPF in 2014 and 2015, respectively (Robalo-Cordeiro et al., 2017).

In November 2020, the UK Medicines and Health Products Regulatory Agency (MHRA, 2020) issued an information warning that PFD carries a risk of serious liver injury and updated advice on liver function testing. The Japanese Pharmaceuticals and Medical Devices Agency published pharmaceuticals safety information on NDN detailing cases of thrombocytopenia, including serious cases leading to bleeding and potentially leading nephrotic syndrome in August 2016 and April 2022 (PMDA, 2016; PMDA, 2022). On 14 December 2021, the Australian Government Department of Health and Age Care updated the product information of NDN to include new warnings on ischemic colitis and renal impairment (Medicines Safety Update, 2022).

Although physicians pay more attention to adverse drug reaction warnings regarding PFD and NDN, most warnings were reported in clinical trials or case reports. There is a lack of systematic and comprehensive pharmacovigilance studies regarding these two drugs. Therefore, we evaluated and compared PFD and NDN in a large real-world population by analyzing adverse events (AEs) reported to the FDA Adverse Event Reporting System (FAERS).

This study compared the AE reporting trends and characteristics of PFD and NDN. The number of reported AEs related to these two drugs increased significantly in 2015.

In that year, PFD and NDN were listed as novel agents conditionally recommended for the treatment of IPF by ATS/ERS/JRS/ALAT (Raghu et al., 2015). However, the total and annual numbers of AEs were higher for PFD than for NDN. Although both drugs were approved for the treatment of IPF by the FDA in 2014, PFD was approved for the treatment of mild-to-moderate IPF in adults in the European Union in 2011. At the same time, some studies found that PFD had significantly lower weighted annual mean anti-fibrotic drug costs than NDN (Corral et al., 2020). Based on this situation, PFD might be more widely used in more patients than NDN, thereby leading to a larger number of reported AEs. In terms of reporting regions, the number of reported AEs was significantly higher for the US than for other countries. This could be related to the higher incidence of IPF in North America and Europe than in Asia and South America (Hutchinson et al., 2015). Events were more common in males than in patients aged ≥65 years. This is explainable by the fact that male sex and older age have been identified as risk factors for IPF (Jee, et al., 2021).

Disproportionality analysis revealed that gastrointestinal disorders were the most commonly SOC for both drugs, which corresponded to previous findings. Regarding gastrointestinal disorders, the most frequently PTs were nausea (N = 3354, ROR = 4.437, 95% CI = 4.278–4.602) and diarrhea (N = 2933, ROR = 16.816, 95% CI = 16.102–17.562) for PFD and NDN, respectively. The clinical safety profile of PFD was demonstrated in the CAPACITY studies, which detected nausea in 36% of PFD-treated patients. However, the rate of treatment discontinuation related to nausea was only 1.4% (Noble et al., 2011). To prevent gastrointestinal AEs, it is recommended to take PFD during or after a meal (European Medicine Agency, 2023). A longer initial dosing titration scheme has been found to further improve the tolerability of PFD (Clinical trials.gov, 2016, NCT01933334). In the INPULSIS trials, diarrhea was the most commonly reported AE, being experienced by 62% of NDN-treated patients (Richeldi et al., 2014). However, only 4.4% of patients on NDN permanently discontinued treatment because of diarrhea (European Medicines Agency, 2015). Loperamide is currently recommended for the management of NDN-associated diarrhea (Corte et al., 2015).

“Eye disorders” represented the most statistically significant SOC for PFD, but the total number of reports was limited. “Cardiac disorders” represented the most statistically significant SOC for NDN, and it was mainly represented by atrial fibrillation (N = 82, ROR = 2.129, 95% CI = 1.713–2.647), cardiac failure (N = 72, ROR = 2.136, 95% CI = 1.694–2.694), and pericardial effusion (N = 24, ROR = 2.654, 95% CI = 1.777–3.964). The phase II TOMORROW trial and the two phase III INPULSIS trials evaluated the most frequent AEs, defined as those with an incidence of more than 10% in the NDN and placebo groups. Cardiac disorders were not the most frequent AEs in all the aforementioned three trials (Richeldi et al., 2011; Richeldi et al., 2014). A subgroup analysis of data from the open-label extension study (INPULSIS^®-ON) of the INPULSIS^® trials focused on the long-term safety of NDN in Asian patients. In this trial, major adverse cardiovascular events were observed at event rates of 5.5 and 2.8 events per 100 exposure-years among patients who continued and initiated NDN, respectively (Song et al., 2020). From the safety-related labeling changes approved by the FDA Center for Drug Evaluation and Research, arterial thromboembolic events, especially myocardial infarction, were mentioned in the “warnings and precautions” section (https://www.accessdata.fda.gov/scripts/cder/safetylabelingchanges/index.cfm?event=searchdetail.page&DrugNameID=839#). A multi-center phase II study including 62 Chinese patients with advanced non-small-cell lung cancer who received NDN as second-line therapy recorded a heart failure rate of 69.35% (43/62) (Dai et al., 2015). Therefore, cardiac disorders should be closely monitored in clinical practice.

“Elevated liver enzymes and drug-induced liver injury” were listed in the “warnings and precautions” sections for both PFD and NDN. The prescribing information recommends that liver function should be monitored prior to initiating NDN and periodically during treatment. In 2020, the UK government published a drug safety update concerning risk of serious liver injury related to NDN, including two fatal outcomes (Verma et al., 2017; Benesic et al., 2019). The LiverTox^® database, which was produced by the National Institute of Diabetes and Digestive and Kidney Diseases, uses a five-point scale to estimate whether a medication is a cause of liver injury (Evans et al., 2001; LiverTox, 2012; Hutchinson et al., 2015; Corral et al., 2020; Neha et al., 2020; Jee, et al., 2021). In that database, NDN was assigned a likelihood score of E^* (unproven but suspected cause of clinically apparent liver injury), and PFD was assigned a likelihood score of D (possible rare cause of clinically apparent liver injury) (Nintedanib, 2017; LiverTox, 2020). In our study, liver injury associated with PFD was represented by hepatic failure (16, 2.778%), acute hepatic failure (10, 1.736%), and drug-induced liver injury (9, 1.563%). Simultaneously, NDN was associated with drug-induced liver injury (46, 6.461%), hepatic failure (27, 3.792%), acute hepatic failure (5, 0.702%), and hepatic necrosis (1, 0.140%). Therefore, NDN-associated liver injury should be further evaluated, and the database should be updated according to the evaluation results.

The disproportional analysis method used in this study has many advantages in pharmacovigilance studies, although it also has certain limitations. First of all, ADR signals detected by disproportional analysis indicate that there is a statistical association between adverse drug reaction events, which has a certain suggestive effect, however, the determination of causality needs further clinical studies to verify. Second, this approach is unable to include all factors in the analysis, due to the spontaneity of the reporting system and the absence of partial information. Third, although the OpenVigil 2.1 software provides an easy way to access, extract, and analyze the open FDA interface, it does not take dosage into account. Finally, reports of adverse reactions to both drugs came mainly from the United States, with limited reports from Asian populations. Further research is needed to determine whether there are differences in adverse effects between different ethnic groups.

The results of our study highlighted the adverse reactions and potential safety issues of PFD and NDN. The findings included associations of PFD with eye disorders and NDN with cardiovascular disorders, which were rarely mentioned or not reported in the medicine specification and previous reports. In addition, liver injury was more frequently associated with NDN than with PFD. Clinicians and regulators should pay more attention to the signals of related adverse reactions that occur frequently. These findings should be validated to guide clinicians, regulators, and the industry to focus on the most relevant signals beyond the information already contained in product characteristic summaries. Our research results could help clinicians identify the risks related to clinical drug use in the future and guide the rational and safe use of NDN and PFD. Further detailed analyses are required characterize the nature of the identified signals and investigate other correlations.