Comments on FDA's Draft Guidance For Sterile Drug Products Produced by Aseptic Processing
November 4, 2003
Dockets Management Branch (HFA-305)
Food and Drug Administration
5630 Fishers Lane, Rm. 1061
Rockville, Maryland 20852
Re: Docket No. 2003D-0382, Federal Register: September 5, 2003 (Volume 68, Number 172, pp. 52782-52783)
The Biotechnology Industry Organization (BIO) appreciates the opportunity to comment on the Food and Drug Administration's Draft Guidance for Industry: "Sterile Drug Products Produced by Aseptic Processing." BIO represents more than 1,000 biotechnology companies, academic institutions, state biotechnology centers and related organizations in all 50 U.S. states and 33 other nations. BIO members are involved in the research and development of health-care, agricultural, industrial and environmental biotechnology products.
BIO believes that while the document provides a comprehensive and valuable discussion of aseptic processing, it at times appears overly prescriptive when viewed as guidance to industry and Food and Drug Administration (Agency) employees. While there is appropriate caution when discussing specific recommendations, proper training of inspectors and reviewers is mandatory to ensure that alternative procedures are allowed in practice.
BIO's specific recommendations are as follows:
Line 140 - 185: BIO suggests that the text be clarified to describe specifically the differential environmental expectations for stoppering and sealing filled vials.
Line 708 - 709: The apparent proscribed use of "microbial growth media" in aseptic simulations seems overly restrictive. We believe that flexibility should be allowed to enable use of simulations that more appropriately represent product characteristics (i.e., as regards flow for sterile powders) or are easier to clean - in these cases microbiological medium can be added to vials post-fill.
Lines 758 - 767: Confusion remains on the requirement for all employees entering the aseptic processing area to participate in an aseptic fill. BIO suggests that the blanket requirement for everyone to participate is burdensome and impractical.
Lines 869 - 872 and Lines 1304 - 1307: While the incubation section (869 - 872) provides a range (20 - 35 C) for incubation of media fill units and does not require two temperatures, the environmental monitoring section (lines 1304 - 1307) notes the two temperature ranges for aerobics and yeasts/molds, which accounts for the current direction to incubate media fills at two temperatures (this is assumed to capture both sets of contaminants). While not explicit, the two-temperature approach is implied by this guidance.
Line 1027: We suggest that the recommendation for adding a second filter be clarified to state circumstances when passage of a post-use integrity test by both of the redundant filters is necessary to provide evidence of asepsis.
Lines 1304 - 1307: Please see the comment on Lines 869 - 872 and Lines 1304 - 1307, above.
Lines 1509 - 1510: We believe the requirement that all in-process data must be included with the batch record documentations in accordance with section 2111.188 is misapplied. With (validated) Programmable Logic Controllers (PLCs) and Supervisory Control and Data Acquisition (SCADA) systems, it is not practical to include all data that a computer system might collect. For example, a printout of lyophilizer temperature with readings every minute for three days could run to over 100 pages. Simple reference to the archived location of these source data would seem adequate.
Lines 1554 - 1566: The text implies that all breaches to isolator integrity are equal in scope and consequence. However certain attributes have an impact on the consequence of a specific breach on product sterility, e.g., location, size, and material in which the breach occurs. This is especially important when the isolator is maintained under positive pressure relative to the surrounding environment. For example, evaluation of the consequence of potential breaches during media fill often shows no adverse effects.
Line 1577: While the term "turbulent flow" is common, the velocity of air within the isolator is usually not high enough to be turbulent. The terms "multi-directional flow" or "non-uni-directional flow" would be more accurate.
Lines 1604 - 1605: We are unaware of data that supports the induction argument put forward in the document. It is our understanding that induction will pull the outside air toward the exit hole but not into the isolator. BIO suggests that this statement regarding induction should therefore be deleted.
Lines 1661 - 1662: The text should be clarified to note that the stated "four to six log reduction" in binding inhibition (BI) titer includes the "margin of extra kill" (i.e., the intent is not to require a demonstrated overkill of > six logs lethality).
Lines 1681 - 1684: The text seemingly treats all breaches as equivalent events. However BIO believes that the requirement that detection of any breach would require a new decontamination is overly prescriptive and unnecessary. Rather, we propose that companies should develop a "decision tree" that describes how the consequence of various breaches will be evaluated, and the resulting actions that will be required.
Lines 1834 - 1835: We suggest that the Agency include a reference about both isolators and traditional laminar flow hoods/cabinets being acceptable. The statement as written does not allow for multi-directional flow isolators.
Thank you for your consideration of these comments. Please do not hesitate to contact me should you have any questions.
Gillian R. Woollett, MA, DPhil
Science and Regulatory Affairs