SAFETY LABELING TIDBIT - No. 31a (Apr 19, 2021) : : : : Public pre-labeling risk communication by agencies. This TIDBIT is about US FDA’s practices (please read up yourself on what other agencies do). - - -  FDA has established a mechanism to alert the public about “important drug safety issues, including emerging drug safety information” without having to first negotiate labeling changes with companies (though FDA strives to notify them at least 24 hours before). FDA may do this if the issue is deemed know-worthy but assessment is still ongoing and no decision has been made that it is label-worthy. In fact, only a part of the issues later resulted in labeling changes (Tau et al, 2019).  - - - This lets us ask the following questions:  How do FDA’s thresholds for such alerts and labeling (for know-worthy vs label-worthy) differ, and why? What are companies to do as a reaction to such an alert? Propose a label change? What about CCDSs/CCSI and other countries? Will this lead to labeling becoming the repository for better established risks? Will it become more difficult for companies to label important potential risks at an appropriately low threshold?  - - - CONTINUED

SAFETY LABELING TIDBIT - No. 31b (Apr 19, 2021) : : : : Public pre-labeling risk communication by agencies. - - - Here are some references for the information in the previous TIDBIT:  Original FDA guidance (2007) “Drug Safety Information - FDA's Communication to the Public” - - - Draft revised guidance (2012) - - - FDA’s webpage: “Strategic Plan for Risk Communication” - - - Cited publication: Tau, N. et al. Association Between Data Sources and US Food and Drug Administration Drug Safety Communications. JAMA Intern Med. 2019 Nov; 179(11): 1-3  - - - See www.pharmiceutics.com for previous TIDBITS.

SAFETY LABELING TIDBIT - No. 32 (Apr 21, 2021) : : : : Drug interactions - Part 1. - - - In principle, labeling may inform about the existence of (1) beneficial interactions, (2) adverse interactions and (3), occasionally, about the absence of an otherwise expected drug interaction. - - - Including a beneficial interaction in labeling requires convincing evidence of its existence and clinical relevance. They are rather rarely found in labeling (think of some beneficial combination therapies). - - - Most interactions found in labeling are of the adverse (undesirable) kind. Including these in labeling follows rules that are, not surprisingly, similar to the rules for adding an adverse reaction to labeling. The required level of evidence for their existence generally depends on their clinical relevance. - - - Other than for “adverse reactions for the purposes of labeling” (see earlier TIDBITS), the criteria for accepting something as an “adverse drug interaction for the purposes of labeling” are not well documented in regulations and regulatory guidelines. - - - The following TIDBITS will address some of the factors to be considered in decision-making. - - - More information in my book. See www.pharmiceutics.com for previous TIDBITS.

SAFETY LABELING TIDBIT - No. 33 (Apr 22, 2021) : : : : Drug interactions - Part 2. - - - You may have noticed or read in publications about significant differences in which adverse interactions are listed in labeling and the various compendia/data bases that provide interaction information. These differences may be in part explained by differences in the amount and up-to-datedness of data utilized by the authors. Another explanation is the inherent subjectivity in weighing the evidence for and the clinical relevance of the potential interaction, i.e., in answering the following questions: (1) How confident do the data make me that the interaction does actually exist? (2) Given significant doubts about its existence, should I nonetheless include it in my compendium/database/label because of its potential clinical relevance? - - - In contrast to labeling, compendia and interaction databases are easily able categorize all listed interactions based on the level of confidence in their existence (e.g., as established, possible, unlikely) - - - On a side note: For label authors it is worthwhile to regularly review the various interaction databases/compendia. - - - CONTINUED - - - More information in my book. See www.pharmiceutics.com for previous TIDBITS.

SAFETY LABELING TIDBIT - No. 34 (Apr 23, 2021) : : : : Drug interactions - Part 3. - - - There can be well-established pharmacokinetic interactions (IAs) with an effect so small that they have no undesirable clinical effect or practical consequences and, therefore, are not “adverse”. Such IAs are generally not label-worthy. This is why regulations/guidelines ask us to label only “clinically significant” or “clinically relevant” IAs. (Don’t misunderstand them to ask us to label only “major” IAs. That’s not what clinical significance/relevance means in this context.) - - -  FDA’s guidance “Clinical Drug Interactions Studies - Cytochrome P450 Enzyme- and Transporter-Mediated Drug Interactions” addresses this topic in section V.B.1. “Approaches for determining No-Effect Boundaries”. When using “sharp” criteria for the existence of a clinical effect (e.g., 90% CI for exposure changes falling within established no-effect boundaries) for determining label-worthiness, we should avoid being too simplistic in our labeling decision-making and keep in mind the interactors’ safety margin, type and size of potential harm, and individual sensitivity. - - - CONTINUED - - - More information in my book. See www.pharmiceutics.com for previous TIDBITS.

SAFETY LABELING TIDBIT - No. 35 (Apr 26, 2021) : : : : Drug interactions - Part 4. - - - Regulatory definitions of “clinically significant interaction” - USA: In the preamble to FDA’s 1979 publication of the then new US PI labeling regulation, the following was provided: “If information on an interaction would not contribute to the safe and effective use of the drug, the interaction would not be clinically significant and, accordingly, is unnecessary in prescription drug labeling.” In 2020, in its Guidance “Clinical Drug Interaction Studies — Cytochrome P450 Enzyme- and Transporter-Mediated Drug Interactions”, FDA gives us a slightly different but equivalent definition: “An interaction is clinically significant if concomitant use of the drugs leads to safety, efficacy, or tolerability concerns greater than those present when the drugs are administered alone.” - - - The EU agencies, however, in training materials originally created for regulatory reviewers and posted on EMA’s website, provide a significantly different, somewhat problematic definition, which I will discuss in the next TIDBIT. - - - More information in my book. See www.pharmiceutics.com for previous TIDBITS.

SAFETY LABELING TIDBIT - No. 36 (Jun 8, 2021) : : : : Drug interactions - Part 5 - - - Regulatory definitions of “clinically significant interaction” - EU: The SmPC guideline does not provide a definition. A definition is, however, found in a deck posted on EMA’s website. There, clinically significant IAs are defined as “those resulting in [a] recommendation on the use of this medicine or other medicines” (deck on section 4.5, slide 3). Slide 4 then lists what such recommendations “might be: Contraindications of concomitant use. Concomitant use not recommended. Precautions including dose adjustment …” - strongly implying that IAs not accompanied by such a recommendation would not be clinically significant and, thereby, not label-worthy. In my view, this definition is too narrow and not supported by the guideline itself. It does not allow for IAs that don’t require such a “bold” recommendation but are still know-worthy, e.g., for interpreting a clinical or lab observation. The original version of the SmPC guideline actually had acknowledged the existence of these by offering, as an alternative recommendation, a “to be taken into account” category. More information in my book. See www.pharmiceutics.com for previous TIDBITS

SAFETY LABELING TIDBIT - No. 37 (Jun 11, 2021) : : : : Drug interactions - Part 6 - - - Pharmacodynamic interactions (PD IAs) - For the purpose of these TIDBITs, I define PD IAs as those that are not caused by a pharmacokinetic interference but by interference of the biologic effects (desirable or undesirable; synergistic, antagonistic; etc.). On one hand, compared with PK IA’s, when it comes to assessing clinical relevance, labeling decision making on PD IAs is simpler and similar to decision making for adverse reactions. On the other hand, it is complicated by the fact that SOME PD IAs are so obviously expectable to the reader of labeling that one can wonder if they need to be mentioned at all (think, e.g., about simply additive blood-pressure effects of two drugs with hypotensive side effects). So, in contrast to adverse reactions and PK IAs, labeling of PD IAs does not only require (1) consideration of the level of confidence in that the interaction can happen and (2) consideration of its clinical relevance but also (3) consideration of its “tell-worthiness” (or “triviality”). - - CONTINUED in the next TIDBIT. - - More information in my book. See www.pharmiceutics.com for previous TIDBITS.

SAFETY LABELING TIDBIT - No. 38 (Jun 22, 2021) : : : : Drug interactions - Part 7 - - - What to do with pharmacodynamic interactions (PD IAs) that appear too “trivial” to be label-worthy? (Read TIDBIT 37 first.) - - - Before deciding to omit from labeling such a PD IA, ask yourself: (1) Do other labels mention this or a comparable PD IA, especially the labels of drugs in the same class or therapeutic area? If yes, this may mean that the label readers have an established expectation that such a PD IA be included and - although unlikely - may assume that its omission from one drug’s label may be sign that this drug is positioned as safer. (2) Can the interaction be covered by means other than including a dedicated statement in the Interactions section? Such PD IA’s can often be packaged together with other risk factors for the undesirable PD effect and presented in a statement under, e.g., Warnings and Precautions. An example to illustrate the principle: “… the risk of symptomatic hypotension is increased in dehydrated patients and patients on antihypertensive medication …”. This may allow us to omit a dedicated statement under Interactions.  - - - More information in my book “Core Labeling of Medicines for Pharmacovigilance Professionals”. See www.pharmiceutics.com for previous TIDBITS.