Unless you’ve been living under a (pharmaceutical) rock, you are aware of the struggle that both FDA and industry are having with Nitrosamines impurities in pharmaceutical products. As a reminder, APIs are at risk of forming nitrosamine drug substance-related impurities (NDSRIs) when they contain secondary amines or dimethyl tertiary amines. We have published many blogs on this topic many, many times- most recently: (here, here, here, here, and here).
Today seems to be a breakthrough day- FDA has issued a Guidance and a webpage to help industry through the NDSRI questions we all have. Let’s first discuss the Guidance for Industry (GFI), “Recommended Acceptable Intake Limits for Nitrosamine Drug Substance-Related Impurities (NDSRIs)”. As they state in the preface, this GFI is being implemented immediately without a comment period (so is FINAL not DRAFT). A high-level overview of what FDA is providing in this GFI follows:
- This GFI applies to APIs, OTC, and prescription drug products (including biological products regulated by CDER, either as standalone or biologic-led combination products), as well as drug products not marketed under a drug application. It includes drugs in clinical development, those that have pending applications, and those that are approved. However, it does not apply to NDSRIs that are detected in products indicated for use in patients with advanced cancers (see the Scope of the GFI for additional details).
- The GFI contains a recommended methodology for Acceptable Intake (AI) limit determination to generate predicted carcinogenic potency categorization and corresponding AI limits that can be applied by manufacturers and applicants when there are no FDA published recommended AI limits. The guidance complements the approach recommended in ICH M7(R2) that recommends the use of structure-activity relationship (SAR) concepts.
- They recognize that not all NDSRIs have the same carcinogenic potency, therefore “it is currently unknown if all or some NDSRIs are in fact high potency mutagenic carcinogens.” Therefore, section IV. The Predicted Carcinogenic Potency Categorization Approach, assigns a recommended AI limit to an NDSRI based on its activating and deactivating structural features alone.
- The predicted carcinogenic potency categorization and resulting recommended AI limit approach described in this guidance should not be applied to NDSRIs in circumstances in which FDA otherwise recommends an AI limit (e.g., based on compound-specific assessments or read-across analysis from a surrogate).
- Table 1 in the GFI provides five (5) Potency Categories and provided with associated recommended AI limits for NDSRIs that range from 26.5 ng/day to 1500 ng/day. This is followed by a flowchart (Figure 1) which can be used to predict the Carcinogenic Potency Category of an NDSRI. Appendix A further gives information on how to determine a potency score based on selected structural features present.
- Section V of the GFI discussed expectations from FDA for timing, implementations, etc., specifically when a product is marketed versus one in development or under review.
- FDA has also provided approaches on how to justify or qualify a proposed alternative AI.
In addition to this helpful new GFI, FDA has created an accompanying webpage (here): where they plan to provide more real-time updates to NDSRI-specific information as it arises. According to the webpage, FDA intends to provide updated information and recommendations on the following topics:
- AI limits for certain NDSRIs based on their predicted carcinogenic potency categorization listed by APIs that could hypothetically be at risk of forming NDSRIs.
- AI limits for certain NDSRIs based on compound-specific or read across analysis using a surrogate.
- Interim AI limits for certain NDSRIs.
- Testing methods to be used for confirmatory testing of certain NDSRIs.
- Safety testing methods for NDSRIs.
The FDA has started out strong, providing (in Table 1) Recommended AI limits for 247 Certain Hypothetical NDSRIs, including the API source and Potency Category. In Table 2, there are 4 NDSRI AI limits provided based on compound-specific or read-across analysis from a surrogate.
This offers some hope for those of us in the pharma industry who have been waiting for specific guidance from regulators so that the industry can continue to provide safe and effective drug products to the patients. Hopefully this is the light at the end of the tunnel we have been waiting for.