Mapping the safety profile of biologicals: a disproportionality analysis using the WHO adverse drug reaction database, VigiBase.

BACKGROUND: Biologicals have specific characteristics, as compared with the small molecule drugs, and carry specific risks. Safety problems, for example infliximab and the risk for tuberculosis, have been identified via spontaneous reports of suspected adverse drug reactions (ADRs). However, in general there is limited data on the nature of spontaneously reported suspected ADRs for biologicals.

OBJECTIVE: To map the safety profile of biologicals as compared with all other drugs. In addition, mechanistic classes of biologicals will be compared.

METHODS: Data was obtained from the ADR database (VigiBase) maintained by the WHO Collaborating Centre for International Drug Monitoring. A disproportionality analysis was performed in which case reports for biologicals and all other drugs (the reference group), reported between January 1995 and December 2008, were selected. Vaccines were not included in the analysis. Suspected ADRs were classified according to Medical Dictionary for Regulatory Activities (MedDRA) version 12.0 at the System Organ Class (SOC) level. Biologicals were classified into mechanistic classes: antibodies, cytokines, enzymes, growth factors, hormones (reference group), interferons, receptors and others/various. The safety profile of the biologicals versus all other drugs in the database and of the various mechanistic classes of biologicals was compared using the proportional reporting ratio (PRR).

RESULTS: 191,004 case reports containing 546,474 suspected ADRs were reported for 62 different biologicals, and 2,556,209 case reports containing 8,761,522 suspected ADRs were reported for all other drugs (the reference group). It was found that two-thirds of all suspected ADRs reported for biologicals were reported for five active substances: etanercept (20.3%), interferon-beta-1a (15.6%), infliximab (11.6%), teriparatide (10.7%) and adalimumab (9.0%). Comparison of the safety profile of biologicals and the reference group showed that suspected ADRs for biologicals were more frequently reported in the SOCs 'Infections and infestations' (PRR 4.5), 'Surgical and medical procedures' (PRR 2.4) and 'Neoplasms benign, malignant and unspecified' (PRR 2.1), and less frequently reported in the SOCs 'Psychiatric disorders' (PRR 0.4), 'Vascular disorders' (PRR 0.4) and 'Pregnancy, puerperium and perinatal conditions' (PRR 0.4). Regarding the differences in safety profile between various mechanistic classes of biologicals, compared with hormones (reference group), 'Infections and infestations' were more frequently reported for receptors and antibodies. 'Neoplasms benign, malignant and unspecified' were more frequently reported for antibodies, cytokines, interferons and receptors, and less frequently for enzymes as compared with the reference group.

CONCLUSIONS: In VigiBase, five biologicals comprise two-thirds of the suspected ADRs reported for biologicals, which might distort the relation found between a specific biological and a specific adverse event in case of quantitative signal detection. Therefore the choice of reference group to be used in case of quantitative signal detection should be considered very carefully. This study confirmed that biologicals have a different safety profile compared with all other drugs in the database and, within the group of biologicals, differences exist between mechanistic classes. Infections are, for example, frequently reported for receptors and antibodies, which often have an immune compromising effect. Such predictable safety issues should be specifically studied by preregistration clinical trials and/or targeted pharmacovigilance. In addition, since not all adverse reactions can be predicted or detected during development, spontaneous reporting remains an important tool for the early detection of signals.