Attractive tools for systematic case accumulation

Excerpt

As Yoshizato et al. [1] mentioned in the recent International Journal of Hematology issue, to establish a causal relationship between a drug candidate and a suspected adverse reaction, it is insufficient only to assess individual adverse drug reaction (ADR) cases; evaluation of accumulated ADR cases is required. We point out that there are potentially effective database systems to support accumulated case review, such as the post-marketing safety database maintained by the market authorization holders (MAH) according to Japanese regulation. A subset of the data is open to the public through the Pharmaceutical and Medical Devices Agency (PMDA) (http://​www.​info.​pmda.​go.​jp/​fukusayou/​menu_​fukusayou_​attention.​html). Table 1 shows the number of reported cases for all ADR and for hypothyroidism. Supplemental data can be downloaded from the website, with more details for the reported ADR cases including year reported, age, sex, outcome, suspected drug, and ADR. The reported proportion of hypothyroidism among all ADRs in using sunitinib appeared to be more than for other tyrosine kinase inhibitors. This tendency appears to be compatible with what Yoshizato et al. mentioned in their article. Thus, reporting tendency in MAH’s safety databases may reflect facts to some extent. In fact, the PMDA’s data has been used to analyze other ADRs in limited cases [2, 3]. Even though the MAH’s safety databases have potentials for various applications, serious problems remain, as the MAH’s safety database includes spontaneously generated safety reports. First of all, as the MAH’s safety databases collect only ADR cases, they contain no information on others who are treated with a drug of interest, but do not experience any ADR. In other words, since the safety database consists nearly entirely of ADR reports, there are no denominators for calculating incidence. Second, the ADR term may vary between reporters. The MAH’s safety database usually registers ADR term based on the reporting physician’s diagnosis. Since these diagnoses are not based on identical criteria, the ADR terms registered in the MAH’s safety database may vary among reporting physicians, even though the ADR terms are coded according to the MedDRA (the Medical Dictionary for Regulatory Activities) terms, which is a standard term system designed to handle ADR in safety databases. Third, the MAH’s safety databases usually collect spontaneous reports. This means if physicians who experienced ADR do not report to the MAH, nothing is registered into the database. Even given these serious problems regarding spontaneous safety reports, however, seven of the 11 safety signals that led to medical product withdrawals from the US and UK markets between 1999 and 2000 were based on spontaneous reports [4], suggesting that such reports may nonetheless be useful for detecting safety signals. Moreover, since there are conditions for approval in the Japanese market for secondary tyrosine kinase inhibitors, including nilotinib and dasatinib, which require the registration of all patients to MAHs, MAHs can be expected to record the number of patients treated with such tyrosine kinase inhibitors, and to collect all adverse events (so long as the attending physicians report all adverse event occurring in the registered patients). Thus, the first and third problems should not affect the use of such databases in monitoring ADR associated with the use of secondary tyrosine kinase inhibitors. Here we propose a role for the society in maintaining the quality of the MAH safety databases in collaboration with MAHs. If the society reviews individual case reports registered in MAH databases and confirms their ADR terms or diagnosis of the event, and screens for causal relationships between candidate drugs and reported ADRs, the second problem associated with MAH safety database will be ameliorated. Thus, if we may be allowed to use such valuable tools, enabling us to study the safety of secondary tyrosine kinase inhibitors in an epidemiological manner, the society may not need to exert exhausting efforts in establishing and maintaining a separate database system. The topic mentioned in this letter may additionally be relevant to a wide range of medical societies other than the Japanese Society of Hematology. …