What CMC Questions (Would) They Answer?

In a recent draft guidance, FDA took a “Cliffs Notes” approach for cell and gene therapy (CGT) development, compiling Frequently Asked Questions into a one-stop resource. This begs the question: what if FDA did the same for small-molecule drug development? Small-molecule Chemistry, Manufacturing, and Controls (CMC) is governed by a patchwork of guidances (ICH Q-series, FDA guidances, etc.), and a consolidated FAQ could serve as a roadmap for sponsors. In this hypothetical Cliffs Notes-style guidance for small-molecule APIs, we outline key CMC questions – and answers – that such a document would likely cover. These questions address hot debate areas in development, providing technical insights for those looking for an edge.

What is the “Regulatory Starting Material” and Why Is It Important?

Q: Where does my GMP responsibility begin in the API synthesis? How do I designate the regulatory starting material?

A: The regulatory (API) starting material marks the point in your chemical synthesis where full Good Manufacturing Practice (GMP) controls must kick in. In simple terms, any steps before the designated starting material can be non-GMP (e.g. conducted by fine chemical suppliers), but every step after it must adhere to stringent GMP standards. Selecting this starting point is crucial – it’s literally the “line in the sand” dividing the non-regulated portion of synthesis from the regulated portion.

Sponsors have a powerful incentive to propose the starting material as late as possible in the synthetic route (sometimes called “late-stage designation”) to minimize GMP burden on early steps. Fewer GMP steps can mean lower costs and more sourcing flexibility for raw materials.

However, regulators push back if the proposed starting material is too far “downstream.” FDA and EMA have grown skeptical of synthetic schemes where a complex molecule seemingly “appears” as the starting material with very few subsequent steps. The concern is that with only 1–2 GMP steps after the starting material, there may be insufficient opportunity to purge impurities (especially mutagenic impurities) formed in the earlier non-GMP steps.

If you attempt to designate a late intermediate as the starting material without transparent lineage (i.e. without full knowledge of how it was made), reviewers may consider it an inadequate justification. The European Medicines Agency (EMA), for instance, has stated it will not accept opaque “third-party” information to justify the purity of an advanced intermediate, the sponsor must have full understanding of its synthesis.

Should regulators reject your proposed starting material as “too late” (i.e. reclassifying it as just another intermediate), the fallout is severe: you might be forced to shift the GMP boundary upstream, redo validations on earlier steps, audit new suppliers, and potentially face significant launch delays.

Takeaway: Define your API’s starting material carefully and justify it with sound science. It should be early enough to ensure impurity control and supply chain transparency, but not so early that you needlessly put simple raw materials under GMP. ICH guidance (e.g. ICH Q11) provides principles for selecting starting materials, emphasizing they should contribute a “significant structural fragment” of the API and be commercially available in a form with acceptable quality.

This remains a hotly debated area – sponsors often negotiate with FDA on how much synthetic detail is needed for a comfortable starting material designation. The bottom line is that the starting material choice can make or break your CMC strategy, affecting cost, timing, and regulatory risk in your development program.

When and How Should the Manufacturing Process Be Validated?

Q: At what stage do I need to perform full process validation for my API (and drug product), and how should I approach it?

A: Process validation – demonstrating that your manufacturing process consistently produces material meeting quality standards – is expected to be completed before you commercialize the product. FDA’s process validation guidance outlines Stage 1 (Process Design), Stage 2 (Process Qualification), and Stage 3 (Continued Process Verification). By the end of Stage 2 (often just before or as part of your NDA or marketing application), you should have executed Process Performance Qualification (PPQ) runs to confirm the process is reproducible at scale.

In fact, successful completion of Stage 2 is required before commercial distribution – meaning you generally cannot ship product for sale until PPQ is done and the data is reviewed (unless FDA specifically allows a limited concurrent validation approach in rare cases).

A common question is: How many PPQ batches are needed? In the past, “three large-scale batches” became an industry rule of thumb, but FDA avoids stating a rigid number. The hypothetical guidance would likely echo what regulators say elsewhere: there is no fixed number of lots required for PPQ.

Instead, the number of PPQ batches should be based on your process understanding and the complexity of your product. If you have a high degree of scientific knowledge and robust data on your process (e.g. through extensive development studies or continuous manufacturing controls), you might justify fewer PPQ lots.

Conversely, if there are lots of unknowns, more runs may be needed to demonstrate consistency. Ultimately, you must show with confidence that the commercial process performs as intended, including under worst-case or challenging conditions. Each PPQ batch should be manufactured under full GMP and tested comprehensively to ensure it meets all quality specs. FDA expects a detailed validation protocol and report as part of the filing.

Modern validation embraces a lifecycle approach. Stage 1 (development) is where you build understanding (often using design of experiments and identifying critical parameters). Stage 2 (PPQ) confirms the final process can scale and is locked in. Stage 3 (ongoing verification) means even after approval, you continuously monitor process performance (e.g. via annual review trending or continued process verification programs).

This ensures the process remains in a state of control. Notably, FDA allows innovative validation strategies – for example, continuous manufacturing or real-time monitoring might shrink the number of PPQ batches needed if they provide equivalent assurance of control. The key is that by the time you file for approval, you have done enough full-scale runs to feel confident in reproducibility and have a plan to monitor performance during commercial production.

Takeaway: Don’t wait until the last minute – engage with FDA early (e.g. at a pre-NDA meeting) about your PPQ plans. Ensure your process is well-characterized and your validation strategy is risk-based. FDA’s stance is clear that one size doesn’t fit all: “a greater understanding and knowledge of the product and process can reduce the number of PPQ lots” needed. But whichever approach you take, regulators will require proof that your manufacturing process is reliable, well-controlled, and capable of consistently yielding quality product before you hit the market.

At What Stage Do Analytical Methods Need to Be Fully Validated?

Q: When do my analytical methods (for potency, purity, etc.) need to be validated? Is it okay to use interim or qualified methods in early development?

A: Analytical methods are the backbone of CMC, ensuring you can reliably measure identity, strength, purity, and quality of your drug substance and product. In early development (preclinical, Phase 1 trials), it’s common to use methods that are qualified rather than fully validated.

A method qualification (or “phase-appropriate validation”) involves limited performance tests to show the method is suitable for its intended purpose (e.g. it can detect impurities above a certain level, it’s sufficiently precise for early formulation work). For example, in Phase 1 you might qualify an HPLC method by checking its precision, a few accuracy points, and that it can detect impurities above safety thresholds – but you might not do the exhaustive validation for every parameter yet.

This flexibility in early phases is generally acceptable because the primary goal is to ensure safety (e.g. no unknown toxic impurities) and basic quality for first-in-human trials.

By Phase 3 (pivotal trials and definitely for NDA submission), FDA expects your analytical methods to be fully validated per ICH guidelines (ICH Q2(R1) and the forthcoming Q2(R2)/Q14). Full validation means you have demonstrated all the required characteristics of the method – typically including accuracy, precision (repeatability and intermediate precision), specificity, sensitivity (LOD/LOQ), linearity, range, and robustness – with documented evidence under GMP conditions.

By Phase 3, the methods should be locked down and reliable enough for quality control of commercial-scale batches. In fact, FDA’s guidance “Analytical Procedures and Methods Validation for Drugs and Biologics” (2015) explicitly says that while less-rigorous method qualification can support Phase 1–2, full ICH Q2 validation is required for Phase 3 and beyond. Practically, this means before you manufacture your registration batches (the ones used in your NDA filing), all the assays used for release and stability testing of drug substance and drug product should be fully validated.

It’s worth noting that compendial methods (e.g. USP methods for things like residual solvents or dissolution) don’t need full re-validation, but they do require method verification to show they work as expected on your specific product. Also, as you make changes during development (new impurities, formulation changes, etc.), your methods may need updates – any significant method changes in late development should also undergo re-validation or at least partial validation to ensure continued suitability.

Takeaway: Use the phase-appropriate approach: early on, ensure methods are scientifically sound and stability-indicating (able to detect degradation) but avoid over-investing in formal validation too soon. As you approach Phase 3, tighten up – finalize your analytical procedures and execute full validation studies so that by the time of your NDA, each method has a complete validation report included in Module 3. This staged approach balances speed and rigor, and FDA is comfortable with it. Just remember: no shortcuts for Phase 3 – the pivotal trial material and anything beyond must be tested with fully validated methods to ensure the data supporting approval is rock solid.

How Do I Set Meaningful Specifications for My API (Drug Substance)?

Q: What specifications (tests and acceptance criteria) do I need to establish for my drug substance and drug product, and how should I justify them?

A: Specifications are the quality standards you propose to ensure every batch of API and final product is acceptable for use – basically, the list of tests (assays, impurity tests, dissolution, etc.) and the acceptance criteria each must meet. According to ICH guidance (notably ICH Q6A for small molecules), your specs should cover the critical quality attributes of the drug: identity, potency (assay), purity/impurities, and physical characteristics (like dissolution for tablets or particle size for certain APIs) that relate to safety and efficacy.

Importantly, each test and its acceptance limit must be scientifically justified with data. A “Cliffs Notes” guidance would remind sponsors that specs are not plucked from thin air or simply copied from early batch data – they should be based on what’s needed to ensure patient safety and drug effectiveness.

In fact, ICH Q6A/Q6B emphasize that specs should “focus on those … characteristics found to be useful in ensuring the safety and efficacy of the product.” In other words, every specification should have a rationale: why is this test needed and how does the limit ensure a quality product for the patient?

For example, you might set a potency specification of 98.0–102.0% for an API to ensure the correct dose, or an impurity limit at 0.10% for a particular impurity because higher levels were associated with toxicology findings or because that level is consistently achievable based on manufacturing data.

When establishing acceptance criteria, use your development data. Regulators expect that you look at results from clinical lots, consistency batches, and stability studies to decide on limits. A common practice is to compile assay values and impurity levels from your Phase 3/pivotal batches and earlier lots – your spec range should cover that variability with some allowance, without being so lax that quality could drift.

For impurities, ICH Q3A/Q3B provide thresholds for qualification (toxicology considerations) which often cap the limits. So your impurity specs might be the tighter of what your process normally produces or the safety threshold, whichever is lower. Specifications also must consider stability: for instance, if an impurity grows to 0.2% in 24-month stability, you might set the release limit lower (like 0.15%) to ensure it won’t exceed the ICH threshold before expiry.

Another current trend is pursuing “clinically relevant specifications.” This means linking certain spec limits to clinical outcomes. A classic case is dissolution specifications for oral drugs – rather than an arbitrary 80% in 30 minutes, companies may justify a dissolution spec that, say, ensures bioequivalence to clinical trial material, because perhaps a slightly lower dissolution rate had no impact on absorption.

If you can show that a wider spec does not affect safety/efficacy, regulators may accept it. Conversely, if a quality attribute is critical (e.g. particle size affecting inhaler deposition), the spec might be tightened to what was proven effective clinically. The overarching point is that specs are part of your control strategy – they work in concert with your manufacturing controls to assure quality.

They should be neither too lax (risking patient harm or product defects) nor too tight to achieve consistently (causing unnecessary batch failures). Each spec needs a justification in your filing, often summarizing how it was set based on manufacturing data, stability data, and clinical experience.

Takeaway: Approach spec-setting methodically. List the critical quality attributes, ensure you have analytical methods capable of measuring them, and gather data from throughout development. Set preliminary specs in early phases (often broad) and refine them as knowledge grows. By the NDA stage, your specifications should meaningfully guard safety and efficacy, and you should be able to defend each acceptance criterion with data and risk-based reasoning. Regulators will look for that justification.

As the ICH Q6 guidelines put it, every specification must be “scientifically sound” and each acceptance criterion “established and justified” with relevant data. In short, specs are your quality contract with the agency – make sure they reflect both patient needs and process capabilities.

How Much Stability Data Do I Need and What Should It Show?

Q: What stability studies are required to support my drug’s shelf life and expiration date?

A: Stability testing demonstrates how the quality of your drug substance and drug product changes over time under the influence of environmental factors like temperature, humidity, and light. The ultimate goal is to establish a re-test period for the API and an expiration (shelf life) for the drug product that ensure the medicine remains safe and effective until it’s used. In regulatory filings (IND, NDA), you must include stability data to justify that your proposed storage conditions and shelf life are appropriate.

For a new molecular entity, ICH guidelines (specifically the ICH Q1 series) define the standard stability package. Typically, by the time of NDA submission you need data from at least three primary batches of API and product on long-term storage (e.g. 12 months at 25°C/60% RH for most climates) and accelerated storage (6 months at 40°C/75% RH). These conditions may vary if your product is intended for refrigeration or other storage.

The three batches should ideally be the same ones used in your process validation (or represent the commercial process) – this is often called the “three-batch rule” to show consistency. Regulators will look to see that no significant degradation or variability has occurred in those batches over the testing period.

If trends are observed (e.g. a gradual increase in an impurity), you might need to do statistical analysis to extrapolate whether the product will remain within spec through the end of shelf life. ICH Q1E provides guidance on using statistics for shelf life estimation.

Stability indicating methods (validated per ICH Q2, as discussed earlier) are crucial – you must prove your analytical methods can detect changes (like potency loss or impurity formation). Agencies also expect you to conduct stress studies and forced degradation in development: these are studies where you deliberately expose the drug to harsh conditions (heat, light, acid/base, oxidation) to understand its degradation pathways and to prove the analytical methods can detect meaningful breakdown products. While these stress tests aren’t part of the formal shelf-life data, they guide you in formulating and packaging the product to enhance stability, and in setting alert levels for degradation.

A hot topic in stability is the concept of climatic zones. If you intend to distribute globally, you might need to consider more stringent conditions (like Zone IVb: 30°C/75% RH, representing very hot/humid climates). Recently, a new comprehensive draft stability guidance was issued that consolidates all Q1 guidelines. It underscores that a single global standard is emerging – meaning companies should be prepared to handle conditions as tough as 30°C/75% if they want a global shelf life.

This can be eye-opening for those used to only the traditional ICH Zone II (25°C) conditions. The draft also clarifies expectations like using bracketing and matrixing (where you test a subset of strengths or package configurations) – you need solid justification and perhaps additional data if you choose reduced testing plans.

Moreover, stability isn’t over at approval: ICH Q1 and Q12 tie in to require ongoing (post-approval) stability monitoring and a plan for stability data when changes are made (e.g. new formulation or site changes may trigger stability studies to confirm no impact).

Takeaway: Ensure you have a well-designed stability program from the start. For IND, you’ll need at least some data to show the investigational product will remain stable for the duration of the clinical trial (e.g. 3-6 months data may suffice early on). For NDA, plan for 12+ months on three batches at long-term conditions (and make sure those batches are adequately representative of commercial scale and packaging). Use interim data to set a provisional shelf life (e.g. you have 12 months data, you might request 24 months shelf life if trends are flat but be prepared to commit to follow-up data).

Always store retention samples and continue pulling data as you approach approval. Stability data should support that potency remains within spec, impurities remain below limits, and other quality attributes (like dissolution, appearance, etc.) stay acceptable over the shelf life. By thoroughly understanding your product’s stability profile – through both real-time studies and aggressive stress tests – you can confidently propose shelf lives and storage conditions.

And remember, a product’s shelf life is only as good as the data behind it. Regulators will scrutinize your justification: does the stability data truly show the product will be good through the last day of expiry? If you’ve done your homework per ICH Q1, the data will speak for itself in answering that question.

Conclusion: One Size Doesn’t Fit All – Engage Early for CMC Success

Small-molecule CMC development doesn’t have a single playbook, but these key questions highlight core principles that every sponsor should consider. A “Cliffs Notes” guidance for small molecules would likely echo a theme we see across FDA’s advice: there is no one-size-fits-all approach to CMC. Each drug has its quirks, and regulatory expectations can shift based on the specifics.

Sponsors should leverage existing guidance’s (ICH Q1–Q12, FDA’s process validation and analytical guidances, etc.) as pieces of the puzzle. Critically, take advantage of FDA meetings (Pre-IND, End-of-Phase 2, Pre-NDA) to ask product-specific CMC questions – you often only get one shot at certain meetings before key milestones. Getting clarity on issues like starting materials or validation plans in advance can save you from costly mistakes.

In the end, a hypothetical FAQ guidance would not necessarily reveal new rules but rather compile best practices and hard-earned lessons. The value is in distilling decades of requirements and lessons into an actionable roadmap. For companies looking for an edge, mastery of these CMC fundamentals – and understanding the latitude within them for innovative approaches – is vital. Early planning, risk-based thinking, and proactive communication with FDA are the formula for navigating CMC challenges.

Consider this “Cliffs Notes” a starting point in your CMC journey, but remember, as with any cheat-sheet, it’s no substitute for the real source material. Diligent study of guidances, and perhaps most importantly, learning from industry precedent and dialogue with regulators, will equip you to answer these questions for your own program and get your quality story right. In CMC, as in literature, those who do the reading (and ask the right questions) will ace the test.

Sources:

1. FDA Law Blog – FDA Issues “Cliffs Notes”-style Guidance on CGT; What CMC Questions Did They Answer? (Part 2)

2. DrugPatentWatch – Key Starting Material Designation and Regulatory Implications

3. FDA Guidance (2011) – Process Validation: General Principles and Practices

4. Rho Inc. – Analytical Method Lifecycle in Drug Development

5. ICH Q6A/B Guidelines – Setting Specifications for Drug Substances and Products

6. The FDA Group (2025) – Draft Stability Testing Guidance Breakdown