Clearing The Fog On New First Air Visualization Expectations

A conversation with Takeshi Ono, Daiichi Sankyo, and Masaru Fujita, Kyowa Kirin

A cleanroom is only as safe as the air moving within it. Regulators are increasingly scrutinizing first air, the undisturbed unidirectional flow exiting HEPA filters, yet execution remains inconsistent across the industry.

While smoke study visualization remains the definitive tool for verifying laminar flow, the methodology often suffers from high variability and subjective interpretation.

At this month's 2026 ISPE Aseptic Conference, Takeshi Ono, a manager at Daiichi Sankyo, and Masaru Fujita, an associate director at Kyowa Kirin, will discuss navigating this regulatory uncertainty. Before their talk, they offered to break down evolving standards, smoke study best practices, and how to avoid airflow-related 483s.

In the absence of a global standard, which existing regulatory annex does the SPP Community of Practice find most challenging or ambiguous to satisfy?

PIC/S describes the visualization study but does not specify the methods for the study. Visualization studies are qualitative in nature, making it difficult to establish clear criteria, and may incorporate subjective elements depending on the observers. In the past, Japan’s approach to visualization studies began with the assumption that merely confirming airflow was sufficient, which may have led to a lack of awareness regarding the importance of the visualization study. This background likely contributes to challenges with visualization studies, leading to Form 483s.

Furthermore, since there are no standards for smoke generator specifications, the required smoke volume and airflow characteristics vary between companies. While the explanation of first air is important, determining how to handle existing equipment poses significant challenges for manufacturers.

You note that visualization requirements are becoming more stringent. Can you explain how expectations have changed?

Historically, PIC/S guidance had been less explicit on  airflow or first air. While the U.S. has documented smoke in the past, it was difficult to envision how this study would translate into production. Currently, since a verification process has been established for how to generate first air (unidirectional air) from the time of equipment installation, requirements have become stricter compared to the past. This shift indicates that the focus has moved from merely verifying airflow to defining how operations should be performed. 

Furthermore, there is no doubt that the approach to contamination control strategy (CCS) has become more stringent, and one of the methods is the visualization study.

Can you describe the expectations and controls for maintaining uninterrupted flow amid dynamic process environments? Should smoke studies account for operator interventions, such as interrupting flow by reaching through an isolator?

It is an indisputable fact that airflow will be disrupted to some extent when equipment is operating and personnel access it. Generally, when personnel access the isolator, it is assumed that the air system is operating and filling equipment is stopped. The critical point when personnel access the isolator is where the air that contacts the surface of the gloves is directed. Verify the airflow when accessing the area using gloves. If airflow is confirmed in critical zones, significant improvements to the work process are required. If no airflow into critical areas is detected, acceptance may be possible.

This constitutes verification in operation, not dynamic testing. SPP distinguishes between "in operation" and "dynamic" testing. Dynamic testing is performed without personnel access, with only filling equipment operating.

Can you recommend best practices for producing smoke studies? What are some of the common mistakes related to camera placement and smoke opacity that could render results less useful?

For sterile products, smoke selection should prioritize substances with no potential for residual contamination in the isolator. Filming smoke requires multiple cameras because its appearance changes with light angles. Using only one camera risks discrepancies between the operator's perception and the recorded data. Therefore, at least two cameras are needed to capture data from different angles. For example, one camera can be fixed on the target area, while the other is operated to capture the smoke discharge point, how the operator interacts with the isolator, and the specific tasks being performed. This allows for three-dimensional spatial confirmation and facilitates easier verification of airflow conditions such as backflow and turbulence.

Performing these tasks will result in a correspondingly longer filming time. Care must be taken when introducing smoke so that airflow is not artificially accelerated.

Not every square inch of a cleanroom can be filmed. How do you justify the criticality of the areas chosen for visualization to an inspector?

It is impossible to film the entire cleanroom at once. This is considered impractical due to insufficient operating time from current smoke generators. Identifying critical areas through assessment and determining smoke measurement locations is considered effective.

Assessments must be fact-based and evaluated using appropriate criteria. For example, when selecting critical areas for sterile injectables, it is important to consider where open processes are located.

What are COP's next-step recommendations for when a smoke study reveals a dead zone or air reversal during the visualization phase?

To prevent such situations, it is crucial to thoroughly discuss your airflow requirements with the equipment manufacturer during installation, document appropriately in the URS, and perform qualification activities. If the situation is confirmed, methods to prevent inflow into critical areas must be considered. This requires collaboration with the equipment manufacturer and may involve implementing measures such as installing partition panels or adjusting the opening of air return sections.

Airflow ingress from outside the isolator is unlikely given the implementation of a leak test; however, if detected, equipment modification would be required. After corrective actions are completed, a visualization study should be conducted to verify the effectiveness of the improvements.

About The Experts:

Takeshi Ono is a manager in the quality assurance department at Daiichi Sankyo. He's spent more than two decades in pharmaceutical manufacturing, focusing on sterile products, deviations and change control, as well as tech transfer and CMO management. He's a member of the ISPE Sterile Products Processing Community of Practice Japan affiliate.

Masaru Fujita is an associate director in the drug product manufacturing department at Kyowa Kirin where his work includes pharmaceutical research, quality control, quality assurance, and manufacturing for commercial and clinical drug product. He is a licensed pharmacist with a master’s degree in pharmaceutical sciences.