Evaluation of potential complication of interstitial lung disease with abemaciclib and palbociclib treatments

Abstract Background Various cyclin‐dependent kinase 4/6 (CDK4/6) inhibitors have demonstrated promising anti‐tumor effects. The Japanese Ministry of Health, Labour and Welfare has issued a warning about interstitial lung diseases as an adverse effect of CDK4/6 inhibitors. However, a large‐scale evaluation of potential complications has not been conducted to date, and the occurrence of these adverse effects is unclear. Aim The aim of this study was to evaluate the clinical incidence of interstitial lung disease caused by two CDK4/6 inhibitors, abemaciclib and palbociclib, and assess the relationship between each drug and interstitial lung disease. Methods and results We evaluated the relationship between the CDK4/6 inhibitors (abemaciclib and palbociclib) and interstitial lung disease in clinical practice using data from the Japanese Adverse Drug Event Report (JADER) database and FDA Adverse Event Reporting System (FAERS) to detect adverse event signals with reported odds ratios (RORs). Furthermore, we performed an adverse event‐time analysis for each drug using data from the JADER database to examine the time of onset of the adverse events. The analysis of the reports in the JADER database showed that the lower limit of the 95% confidence interval (CI) of ROR for abemaciclib was >1 regardless of age, and a signal was detected. Interstitial lung disease associated with abemaciclib and palbociclib use has been reported, with an average onset period from treatment initiation [median (25th‐75th quartile)] of 65.1 [56.0 days (25.3‐98.3 days)] and 53.1 days [38.0 days (10.8‐76.0 days)], respectively. The analysis of the reports in the FAERS showed that the lower limit of the 95% CI of the ROR for the two drugs was >1, and a signal was detected. Conclusion Treatment with abemaciclib and palbociclib is associated with a potential complication of interstitial lung disease, regardless of age.

Methods and results: We evaluated the relationship between the CDK4/6 inhibitors (abemaciclib and palbociclib) and interstitial lung disease in clinical practice using data from the Japanese Adverse Drug Event Report (JADER) database and FDA Adverse Event Reporting System (FAERS) to detect adverse event signals with reported odds ratios (RORs). Furthermore, we performed an adverse event-time analysis for each drug using data from the JADER database to examine the time of onset of the adverse events.
The analysis of the reports in the JADER database showed that the lower limit of the 95% confidence interval (CI) of ROR for abemaciclib was >1 regardless of age, and a signal was detected. Interstitial lung disease associated with abemaciclib and palbociclib use has been reported, with an average onset period from treatment initia- showed that the lower limit of the 95% CI of the ROR for the two drugs was >1, and a signal was detected.
Conclusion: Treatment with abemaciclib and palbociclib is associated with a potential complication of interstitial lung disease, regardless of age. It has been demonstrated to be efficacious in combination with endocrine therapy (eg, anti-estrogen and aromatase inhibitors) against estrogen receptor (ER)-positive (ER+) and HER2-negative (HER2−) breast cancer. 4 Abemaciclib is another CDK4/6 inhibitor that is efficacious against ER+ and HER2− breast cancer when administered in combination with endocrine therapies, such as anti-estrogen and aromatase inhibitors. 5 It was approved in 2018. 4,6 The adverse effects of abemaciclib include myelosuppression, gastrointestinal symptoms, and skin reactions, which are clearly indicated on the package insert of abemaciclib. 7 In addition, the Japanese Ministry of Health, Labour and Welfare has issued an alert regarding interstitial lung disease caused by abemaciclib. 8 Similarly, for palbociclib, a directive has been provided to revise the package insert, and interstitial lung disease has been mentioned in the warning section. 9 Acute pulmonary fibrosis, an interstitial lung disease, is fatal. Therefore, great care must be taken to prevent the development of this disease. 10 However, considering that a large-scale evaluation of the potential complication of interstitial lung disease has not been performed to date, its incidence is unclear. Here, we evaluated the Using the data from these two databases, we assessed the association of palbociclib and abemaciclib with interstitial lung disease by detecting adverse event signals with ROR. In addition, we performed adverse event-time analysis using data from the JADER database and examined the adverse reaction-time profile of each drug.

| Targets of analysis
The adverse event identified for extraction was interstitial lung disease as described in the Medical Dictionary for Regulatory Activities (MedDRA) Ver. 22.1J by the International Council of Harmonization (ICH), using standardized MedDRA queries. The basic term "interstitial lung disease" was set as the preferred term to narrow the scope terms.
Abemaciclib and palbociclib were considered as the target drugs for analysis. In addition, sex and age were analyzed. Additional details on the reporting time of adverse reactions were also analyzed. Time-onset analysis of drug administration and interstitial lung disease was performed only in cases where both the date of drug administration and date of onset of interstitial lung disease were clear. We used the free statistical software EZR (Easy R) to perform the log-rank test. 14

| Analysis method
Access2016 (Microsoft) was used to create the JADER database.
NaviCat for SQLite (Premium Soft) was used to create the FAERS.
The safety evaluation of each anticancer drug was performed by signal detection. ROR was used as the index score 15 and was calculated using a 2 × 2 contingency table that was categorized based on the use or non-use of a drug and the presence or absence of specific adverse events. A signal was detected when the lower limit of a 95% confidence interval (CI) was greater than 1. In addition, signal detection for patients both over and under the age of 60 years was performed. [16][17][18] The latest cases reported in the FAERS were retained, whereas the duplicated cases were excluded. Regarding the cases in the JADER database, cases with missing values (unclear age or sex) were excluded.

| RESULTS
The total number of reports in the JADER database and that of reports on interstitial lung disease were 611 336 and 33 099, respectively. The total number of adverse events reported for abemaciclib and palbociclib was 151 and 1544, respectively, of which 38 and 50 were cases of interstitial lung disease. Abemaciclib or palbociclib were not administered to men. The RORs (95% CI) in the JADER for abemaciclib and palbociclib were 5.9 (95% CI: 4.1-8.5) and 0.5 (95% CI: 0.4-0.8), respectively. Among the two drugs, only abemaciclib had a lower limit of 95% CI of ROR greater than 1; thus, signals for only this drug were detected (Table 1) (Table 2). In the JADER database, regarding the association between the occurrence of interstitial lung disease and age, the ROR in patients under 60 years of age was 5.1 for abemaciclib (2.0-13.0) and 0.5 for palbociclib (0.2-1.3), and the ROR in patients over 60 years of age was 6.3 for abemaciclib (4.1-9.5) and 0.8 for palbociclib (0.6-1.2). The lower limit of the 95% CI of the ROR for abemaciclib was greater than 1, regardless of age, and a signal was detected ( Table 3).
The total number of cases reported in the FAERS was 11 448 913, and the total number of reported adverse reactions related to interstitial lung disease was 66 335 (0.58%). The RORs (95% CI) for abemaciclib and palbociclib in the FAERS were 4.8 (95% CI: 3.8-6.1) and 1.3 (95% CI: 1.1-1.4), respectively. The lower limits of the 95% CI of the ROR for both drugs were greater than 1, and a signal was detected (Table 4). Regarding the association between the occurrence of interstitial lung disease and age, the ROR in patients under 60 years of age was 5.1 for abemaciclib (2.8-9.2) and 1.4 for palbociclib (1.0-1.8), and the ROR in patients over 60 years of age was 3.9 for abemaciclib (2.7-5.4) and 0.6 for palbociclib (0.5-0.7). The lower limit of the 95% CI of the ROR for abemaciclib was greater than 1, regardless of age, and a signal was detected (Table 5).

| DISCUSSION
Interstitial lung disease has been listed as a serious adverse effect in the package inserts of the CDK4/6 inhibitors abemaciclib and palbociclib. 7,9 Data of the analysis of reports from the JADER database showed that the lower limit of the 95% CI of the ROR for abemaciclib was greater than 1; however, the signal was not significant. In the T A B L E 1 Number of reports and the reporting odds ratio for each CDK4/6 inhibitor in the JADER database analysis of reports from the FAERS, signals were detected for both abemaciclib and palbociclib. Individually, abemaciclib showed a stronger signal than palbociclib. Abemaciclib was more selective for CDK4 than palbociclib, and this may have contributed to the difference in the occurrence of adverse events. 19 Nevertheless, the number of reports on palbociclib substantially in the FAERS exceeded that in the JADER database. Thus, it would be appropriate to conduct an analysis after an adequate number of cases have been reported in the JADER database. 11 Furthermore, both abemaciclib and palbociclib were associated with the onset of interstitial lung disease after a delay of 1-2 months from the start of treatment, indicating the need for monthly treatment monitoring in patients administered both drugs. In addition, the results showed that the common mechanism of action of these drugs may have contributed to the development of interstitial lung disease.
The results of this analysis showed that treatment with abemaciclib is associated with a possibility of development of interstitial lung disease, regardless of age. Although abemaciclib and palbociclib are commonly prescribed to patients older than 60 years, the association between age and the occurrence of CDK4/6 inhibitor-induced interstitial lung disease needs to be further clarified owing to a smaller number of cases in Japan.
As abemaciclib and palbociclib are relatively new drugs in Japan (launched in November 2018 and December 2017, respectively), information on their adverse effects is limited. While palbociclib is a specific inhibitor of CDK4/6, abemaciclib can inhibit various kinases at normal doses; it inhibits the binding of cyclin B and CDK1 and has effects on the G1 and G2 phases of the cell cycle. In other words, abemaciclib has a low specificity for CDK4/6, and therefore, it may cause more severe adverse effects, which could explain the difference in the potential complication of interstitial pneumonia between abemaciclib and palbociclib. 20 Hence, we were unable to perform an analysis based on disease severity. A sub-analysis to explore the association of age and sex in the reports from the FAERS showed results similar to those from the JADER database (Table 5). This was a limitation of our study. In addition, although anticancer drugs are generally used in combination, we did not consider the combination of CDK4/6 inhibitors with other anticancer drugs in this analysis. Recently, the adjustment of confounding factors using logistic regression and subset analyses has been reported as an attempt to prove this point. [20][21][22] The effects of concomitant medications and various confounding factors may be explored using such methods. 21 The FAERS is suitable for screening unknown adverse events and analysis of trends in adverse events owing to the reports of large number of cases in this database. In contrast, the JADER database is useful for analyzing the trend of reporting domestic adverse events and its time of onset. 21 Information on the onset time of adverse events is useful for alerting healthcare professionals and patients.
Information on new agents, such as CDK4/6 inhibitors, is useful for maintaining patient quality of life in cases of adverse events, especially interstitial lung disease, wherein the mechanism of event onset is typically unknown, and no preventive or therapeutic measures have been established. In the analysis of the reports from the JADER database, only abemaciclib was found to have a signal >1 for the lower limit of the 95% CI for ROR. However, in the analysis of the reports from the FAERS, both abemaciclib and palbociclib were found to have a signal, which might be due to the racial differences between patients from Japan and the United States. Nevertheless, the number of cases in the JADER database was less than that in the FAERS.
As the ROR only shows a high frequency of adverse drug reaction reports, the causal relationship between the drugs and the adverse drug reactions has to be evaluated in the future. Furthermore, the adverse effect profiles of CDK4/6 inhibitors have not been systematically evaluated due to the small number of available cases. Nonetheless, we believe that this study provides important suggestions for future research.
In summary, the results of this study show that treatment with abemaciclib and palbociclib require caution owing to the potential complication of interstitial lung disease. We believe that these results may be useful in preventing the occurrence of interstitial lung disease when using CDK4/6 inhibitors. In addition, this method may be used to obtain information useful for the safety management of various other drugs.

ACKNOWLEDGEMENT
We would like to thank Editage, Tokyo, Japan, for editing this manuscript.