Menu Back to Poster-Presentations-Details

P205: Comparing Safety Between Generic and Brand Drugs of Statins Marketed in Japan: A Cohort Study

Poster Presenter

Hotaka Maruyama

Division of PharmacoEpidemiology, Office of Pharmacovigilance ?
Pharmaceuticals and Medical Devices Agency (PMDA)
Japan

Objectives

Recent recalls of some generic drugs raise public concerns about drug use in Japan. To provide further drug safety information, safety profiles of statins, for which generic drugs were widely used, were compared on the occurrence of abnormal liver function between generic and brand drugs.

Method

New users with a statin from January 1, 2014 to March 31, 2022 were identified in the MID-NET®. Primary outcome was the first abnormal liver function (CTCAE grade 2). The adjusted hazard ratios (aHRs) for the outcome were estimated using high-dimensional propensity score (hdPS) weighted Cox models.

Results

50,379 new users with one of 6 statins were identified for analysis, including patients with rosuvastatin (generic:2,239, brand:13,715), atorvastatin (generic:12,821, brand:2,957), pitavastatin (generic:4,939, brand:6,493), pravastatin (generic:3,893, brand:2,267), simvastatin (generic:261, brand:363), and fluvastatin (generic:106, brand:325). After adjustment, Median duration of the patient follow-up in each cohort was as follows: rosuvastatin (generic:117 days, brand:126 days), atorvastatin (generic:145 days, brand:100 days), pitavastatin (generic:138 days, brand:119 days), pravastatin (generic:123 days, brand:98 days), simvastatin (generic:92 days, brand:99 days), and fluvastatin (hdPS was incalculable due to the limited population ). In the primary analysis, aHR (generic drug compared with the relevant brand drug) was 2.03 (95% confidence interval (CI): 1.16-3.56) for atorvastatin, and 0.47-1.02 (95% CI crossing 1.00) for the other five statins. When a follow-up period was censored on the date of 360, 180, 90, and 30 days from the first prescription date to examine impacts of events occurred at the longer period after the statin prescription, the aHRs for atorvastatin were 1.70 (95%CI: 0.93-3.13), 1.65 (95%CI: 0.85-3.20), 1.48 (95%CI: 0.73-3.02), and 1.33 (95%CI: 0.60-2.95), respectively.

Conclusion

The aHR for atorvastatin on the primary analysis was significantly greater than 1.0, although no increased trends on aHRs were observed for the other five statins. In the results with the censored follow-up periods, aHRs for atorvastatin approached 1.0 with the 95% CI crossing 1.0, as the follow-up period shortened. These results suggest that the observed increased risk by generic atorvastatin may be affected by other factors and do not necessarily show different safety profiles between generic and brand drugs. Results obtained in this study will be useful to further facilitate understanding about safety of generic drugs and may address the concerns in using generic drugs in clinical practice in Japan.