Sterility Assurance: What US FDA Is Emphasizing (and How Teams Can Prepare)

The “Knowns” That Matter Most

Sterile manufacturing inspections continue to focus on familiar fundamentals, and that’s good news, because expectations are well understood. What Dr. Arigo highlighted, though, is that a few recurring execution gaps persist across the industry. His talk offered a candid view of how the US FDA identifies those gaps during inspections, along with practical ways for quality and manufacturing teams to close them.

Using the US FDA’s internal large language model-based tool, Elsa, he reviewed more than 400 Form 483 inspection summaries from sterile facilities around the world. He came prepared with data, real examples, and reviewer-minded guidance. The key theme wasn’t “new rules,” it was alignment: making sure what you qualify, how you operate, and what you document all clearly reflect the reality of day-to-day production.

What the Inspection Data Is Showing

The inspection summaries analyzed by Arigo spanned the period of January 2024 through December 2025. The dataset included global facilities from the US, India, China, Japan, and Europe, across both commercial manufacturers and contract organizations.

A few themes showed up again and again; familiar topics, but still worth reinforcing given how often they appear: qualification studies that don’t fully reflect production reality, plus gaps in aseptic technique and personnel controls.

Qualification: Make Studies Look Like Real Production

A consistent theme underlying many qualification observations is straightforward and readily attainable: build studies that reflect how commercial production is actually intended to be run. When there’s a clear link between the qualified state and the day-to-day operating state, it’s easier to defend during inspection, and easier for reviewers to follow in a submission.

Unofficial Autoclave Loading Patterns

Arigo noted the importance of ensuring that the loads used in routine production are covered by the loads you qualify. Many facilities submit a set of fixed loads and perhaps a worst-case load, but day-to-day operations sometimes call for a variation outside the documented patterns. Designing (and clearly justifying) a true worst-case load, scientifically shown to be the hardest to heat, can provide more flexibility later, because small variations are easier to explain when the original study was built on sound worst-case logic.

Media Fills Disconnected from Commercial Production

Media fills are strongest when they mirror commercial scale and duration. For example, if routine production runs 10,000 vials over 15 hours, the simulation should be designed to represent that same scale and time meaningfully. When the simulation is much smaller (for instance, a few hundred units for a process that routinely runs in the tens of thousands), it can raise avoidable questions and lead to additional follow-up during review.

Equipment That Was Never on the Radar

Another helpful checkpoint is making sure all critical equipment is consistently represented across internal documentation and the application. If an inspection team encounters a critical autoclave (or another key system) that isn’t clearly described in the submission, aligning the inspection and review records becomes more complicated than it needs to be. A straightforward impact assessment and clear presentation of qualification status can prevent that disconnect.

Aseptic technique: Small habits, big impact

The second major theme is less about paperwork and more about day-to-day behaviors in the cleanroom. Arigo’s examples were a reminder that strong aseptic processing often comes down to consistent “back-to-basics” habits, especially around airflow, gowning discipline, and operator qualification.

He shared examples of the issues inspectors sometimes observe, such as operators inadvertently blocking first air over open containers, gowning practices that don’t match the operation’s cleanliness needs, or operator qualification approaches that don’t scale with commercial batch sizes.

One especially practical point was that when procedures exist, the biggest gains often come from making execution consistent, through targeted training, coaching, and routine reinforcement on the floor. In many cases, it’s less about creating more procedures and more about making the current ones easier to follow every time.

Sometimes the procedures are actually in place but they're not being followed. That's something that you would hope you wouldn't see. If the procedure is in place and you're just not following it, that's a costly mistake that could be avoided.

John Arigo, PhD, US FDA 

Submission Tips That Make Reviewers’ Lives Easier

Arigo offered several practical recommendations that can reduce back-and-forth questions and help keep timelines on track.

Organization and Readability

Reviewers often navigate large electronic submissions by clicking through folders and opening long PDFs, sometimes 600–700 pages. A clear opening summary can save real time. For example: “Autoclave X was qualified by performing three empty-chamber heat distribution runs and three worst-case heat penetration/BI runs in 2025. The worst-case load covers all proposed production loads. Results are provided on page 18.” Pair this with a simple summary table (minimum/maximum temperatures, values, and results) to make the reviewer’s job easier before they open a single attachment.

It also helps to write as if the reviewer has never seen your site, because they haven’t. Label equipment with IDs and room numbers and note whether it’s the only unit of its kind in the building. What’s obvious onsite isn’t always obvious in a PDF.

Drug Master Files (DMFs)

DMFs can be a powerful tool—especially when the letter of authorization (LOA) clearly directs reviewers to the exact validation information they need. A broad, well-structured DMF with qualification data covering equipment under worst-case or bracketing conditions can support multiple ANDAs and help compress review timelines. When the LOA is less specific (for example, referencing a document number without a roadmap), reviewers may need extra time to locate the right details.

Bulk Bioburden Sampling

On bulk bioburden sampling, the US FDA’s point is straightforward: sample before any filtration. Even when organisms are removed by a filter, microbial byproducts (such as metabolites) and endotoxins can still be present, so pre-filtration sampling provides a more accurate picture of upstream control.

Biological Indicators: Seven Days, Not 24 Hours

US FDA continues to see questions around shorter incubation times (often 24 hours) for biological indicator studies. One nuance Arigo emphasized is that some supporting references come from a healthcare facility context rather than drug product manufacturing. USP <55> Biological Indicators, Resistance Performance Tests, and ISO 11138-1:2017 align on a seven-day incubation period for established sterilization processes such as moist heat. While a 24-hour read may detect many positives, teams should decide, based on risk assessment, whether that approach is sufficiently robust for pharmaceutical manufacturing.

Product Endotoxin Testing and Sample Pooling

When pooling endotoxin samples (which the US FDA may find acceptable for small volumes), remember that the maximum valid dilution (MVD) should be divided by the number of samples pooled. This adjustment is easy to overlook and can lead to avoidable follow-up. Also, be sure dosing calculations consider all relevant populations (including pediatric patients), since reviewers will typically verify the math independently.

Supplement Filing Tips

1. List all changes explicitly in the cover letter. Do not make a reviewer discover on page 400 that the rubber stopper, the filling duration, and the autoclave load were changed. Put every change up front, numbered and noted clearly.
2. Tell reviewers what you're changing from and to. Do not assume that the reviewer has the prior approval records. These records may be old and not easily accessible, such as those on paper or stored remotely. Clearly state in your writing what was previously approved and what is now being changed. Don't assume the agency knows.
3. Justify your filing category with specifics. If claiming Changes Being Effected Level 30 (CBE-30)* based on prior approval, provide the Abbreviated New Drug Application (ANDA) number, date, line, and machine identifiers, and a copy of the approval if possible. Arigo suggests using a statement such as "nothing else is changing" for clarification.
4. Specify the lyophilizer used in media fills. The lyophilizer used in simulations must be the one proposed for commercial production, and that should be clearly stated. This applies to all equipment in the filling process.

FDA PreCheck: An option for new domestic facilities

FDA’s PreCheck program is intended to encourage domestic pharmaceutical manufacturing by offering earlier regulatory engagement and more predictability during facility development. As of March 2026, the pilot was in the candidate selection phase.

The program has two core components. The first is a pre-operational review: inspection teams visit the facility during construction and setup to provide guidance on GMP systems, isolator configuration, and early qualification studies before commercial production begins.

The second involves assistance in setting up a Type V facility-based DMF, the idea being that a well-constructed facility DMF with all critical equipment qualified under worst-case or bracketing conditions could support multiple future ANDAs simultaneously, with much of the sterility assurance review effectively completed before the first application arrives.

Even for facilities not enrolled in PreCheck, the underlying logic applies. A well-structured worst-case qualification study, clearly documented in the general qualifications section of an ANDA, can accomplish much of the same efficiency for reviewers and future supplement filings.

Clear Expectations, Practical Opportunities

One reassuring theme from this talk is that sterility assurance expectations are not a moving target. US FDA guidance is established and accessible. The opportunities Arigo described are largely about consistent execution: qualification studies that reflect production reality, training programs that match commercial scale, and submissions that anticipate the questions a reviewer will naturally have.

In practice, the formula is straightforward: qualify what you truly do, train for the scale you truly run, and write submissions with enough context that someone who has never visited your facility can quickly understand the story.

Turning Guidance into Everyday Practice

Arigo’s presentation didn’t introduce new science or propose new rules. Instead, it delivered something many teams find even more helpful: a clear view of common inspection themes and a practical picture of what “good” looks like from the reviewer’s seat.

The themes he described aren’t about a lack of knowledge; they’re about making sure what we know is applied consistently, at scale, every time. That’s where strong quality systems, effective training, and clear documentation make the biggest difference.

For companies building new domestic facilities, the FDA PreCheck program may offer a proactive way to align early, before the first application is filed. For everyone else, the takeaway is still very actionable: qualify to what you actually do, train to the scale you actually operate, and write submissions as if your reviewer has never set foot in your building, because they likely haven’t.

The science is established. The guidance is clear. Success comes from disciplined implementation.

John Arigo, PhD, US FDA 

*Note: CBE-0 and CBE-30 submissions are used for minor changes that have minimal impact on the drug product. See FDA Guidance for Industry: Changes to an Approved NDA or ANDA: Specification- Use of Enforcement Discretion for Compendial Changes for more info.