RESUMO
BACKGROUND: Prescription drug labeling is an authoritative source of information that guides the safe and effective use of approved medications. In many instances, however, labeling may fail to be updated as new information about drug efficacy emerges in the postmarket setting. When labeling becomes outdated, it loses its value for prescribers and undermines a core part of the FDA's mission to communicate accurate and reliable information to patients and physicians. METHODS: We compared the number of drug uses indicated on product labels to the number of uses contained in a leading drug compendium for 43 cancer drugs approved between 1999 and 2011. We defined a "well-accepted off-label use" of a drug as one that was not approved by the FDA and received a category 1 or 2A evidence grade. RESULTS: Of the 43 drugs reviewed in this study, 34 (79%) had at least one well-accepted off-label use. In total, 253 off-label uses were identified; 91% were well accepted, and 65% were in cancer types not previously represented on labeling. Off-patent drugs had more well-accepted off-label uses than brand-name drugs, on average (mean 13.7 vs 3.8, P = .018). CONCLUSIONS: The labeling for many cancer drugs, particularly for older drugs, is outdated. Although FDA-approved labeling can never be fully aligned with real-world clinical practice, steps should be taken to better align the two when high-quality data exist. Such steps, if taken, will assist patients and prescribers in discerning which uses of drugs are supported by the highest quality evidence.