Ruggedness of Visible Residue Limits for Cleaning-Part III: Visible Residue Limits for Different Materials of Construction

Validation or re-validation of a cleaning program requires two criteria to establish equipment cleanliness. The first criterion is that there can be no carryover that will have an adverse impact on the patient (1). Setting a health-based limit to assure patient safety can begin with the allowable daily exposure (ADE) of the API approach (2, 3), or with the 1/1000th of the therapeutic dose approach (1, 4). The health-based limit is not only specific for the API but also for the manufacturing equipment in the facility.

The second criterion is that the equipment must be visually clean (5). Visually clean equipment can involve either a qualitative assessment or quantitative assessment using a visible residue limit (VRL), which is the level below which an API or formulation is not visible. VRLs have been established for use in clinical (6) and manufacturing (7) facilities. In Part I of this series, the ruggedness of the VRL approach was shown for hundreds of APIs, excipients, formulations, and detergents (8). In Part II, ruggedness was also demonstrated regarding the presentation of the coupons to the observers (9); however, to date, all VRL data has been generated using stainless-steel coupons. Because stainless steel comprises the vast majority of pharmaceutical manufacturing equipment product contact surfaces, it was the logical material of construction (MOC) on which to concentrate.

Determining VRLs for other materials of construction has been a question of considerable interest. Swab recovery studies have shown that a common recovery factor can be developed for the vast majority of different materials of construction (10); however, it is intuitive that the VRL for a given soil will differ from stainless steel to clear glass to the white background of polytetrafluorethylene (Teflon, DuPont). There are a variety of materials of construction to be considered for VRLs and depending on the equipment configuration, an alternate material of construction could comprise the majority of the product contact surface area. Once the VRL is established for a variety of MOC coupon types, which VRL to use going forward is the question. A distinct VRL for each material of construction is too cumbersome to develop, maintain, and assure compliance. Using the highest established VRL for the range of materials of construction is overly conservative unless the material of construction with the highest VRL is also the primary component of the manufacturing equipment. The current study was conducted to generate data to provide a basis for addressing these issues.

Coupons for VRL study

Represented coupons

Stainless steel

Hastalloy, aluminum

Glass

Acrylic, Lexan

Polytetrafluoroethylene (Teflon, DuPont)

Polypropylene, silicone, latex rubber, ethylene propylene diene monomer, neoprene

Methods
The VRL study was conducted to determine visual limits for a variety of soils on a range of MOC coupons using the following parameters. Table I lists multiple MOC coupon types, which are commonly used in pharmaceutical and biopharmaceutical manufacturing. A subset of coupon types was chosen as representative of the MOC coupon types. Stainless steel was chosen as representative of metal coupons; glass was chosen as representative of clear coupons; and polytetrafluorethylene (Teflon, DuPont) was chosen as the most difficult to see of the opaque coupons.

Concentration

Batch

Expiration date

Aspirin

81 mg/tablet

434549

Jan 2015

Caffeine

200 mg/tablet

10671868

May 2013

Calcium carbonate

600 mg/tablet

X27234

Aug 2013

Diphenhydramine

25 mg/tablet

P75456

Dec 2014

Ibuprofen

200 mg/tablet

2KE1259A

Jul 2014

Pharmaceutical soils included aspirin, caffeine, calcium carbonate, diphenhydramine, and ibuprofen tablets. The concentrations of the tablets are shown in Table II. The tablets were initially dispersed in several millimeters of distilled water. The soils were accurately diluted with isopropanol to the approximate concentrations listed in Table III. Isopropanol was chosen as the spotting solvent because it spreads easily due to its low surface tension, resulting in residues of approximately the same area, and it also evaporates quickly.

Spotting solution preparation

Residue spotted

Target concentration (20 cm²)

Solution A – 400 µg x 0.2 mL

80 µg

4 µg/cm²

Solution A – 400 µg x 0.1 mL

40 µg

2 µg/cm²

Solution B – 100 µg x 0.2 mL

20 µg

1 µg/cm²

Solution B – 100 µg x 0.1 mL

10 µg

0.5 µg/cm²

Solution C – 50 µg x 0.1 mL

5 µg

0.25 µg/cm²

Solvent

0 µg

0 µg/cm²

The soils were deposited onto individual coupons at the various concentrations in a serial order. Isopropanol was deposited on one of the coupons to serve as a negative control. Either 100 µL or 200 µL of the suspension were spread on the coupons. The final residue areas, concentrations, and observations of the five soils are shown in Table IV.

For observation, the coupons were arranged first by soil and then sequentially by concentration on the three MOC coupon types. Observers independently viewed the coupons under well-defined, controlled conditions. The light level was between 350 lux to 400 lux. Data have shown that a light level of more than 200 lux lighting provides consistent results (6-8). Three observers viewed the coupons from a distance of 2 ft at a viewing angle of 40 º and determined the level of each soil that was visible on the MOC coupon. Changing viewing parameters did not significantly change the ability to see residue on stainless steel, but the ability to detect residue on both glass and polytetrafluorethylene (Teflon, DuPont) was greatly influenced by the viewing parameters. The optimal viewing parameters for glass were similar to stainless steel, but highly dependent on the background. Either dark or white backgrounds can make visual detection difficult. A medium, neutral color enhanced residue detection. Alternatively, looking through the glass with the light source behind and slightly offset as shown in Figure 1 also afforded equivalent viewing parameters.

Optimal viewing conditions for glass are fairly restricted and need to be established for each facility. Residues on polytetrafluorethylene (Teflon, DuPont) are difficult to detect but viewing from a shallow angle as shown in Figure 2 seemed to give the best chance of seeing the residue. The observers determined the VRL of each soil under optimal conditions on each material of construction as presented in Table V.

Optimal viewing conditions for glass are fairly restricted and need to be established for each facility. Residues on polytetrafluorethylene (Teflon, DuPont) are difficult to detect but viewing from a shallow angle as shown in Figure 2 seemed to give the best chance of seeing the residue. The observers determined the VRL of each soil under optimal conditions on each material of construction as presented in Table V.

Results and discussion
The MOC coupons chosen for the study were seen as representative of a wide variety of materials. A subset was also chosen, rather than to attempt a comprehensive survey of materials of construction because the combination of soils and materials of construction are typically restricted to a single site. Implementation of this approach at a site should include all soils on all materials of construction employed at the site unless an abbreviated approach can be justified.

The target residue concentrations as shown in Table III were based on past experience with a wide variety of pharmaceutical soils dissolved or suspended in methanol on stainless steel (8). The low surface tension of isopropanol allowed the residues to spread out on the materials of construction. The resulting concentrations of the dried residues are shown in Table IV. The combination of soils, MOC coupons, and solvents should be assessed for each facility to ensure a uniform distribution of residues.

Observations of the residues on the different MOC coupons varied based on the observation parameters. Residues on stainless-steel coupons were visible at a light level of  > 200 lux, a viewing angle of  > 30 º, and a distance of < 10 ft (3 m). Observations of residues on glass coupons were more challenging. Residues were visible when the viewing angle was at 30 º with the coupons laying flat on the viewing surface. Residues were not as visible at other angles. Residues were also highly visible when the glass coupons were viewed looking through the coupons with the light source at a slight offset behind the coupon. Under the optimal viewing conditions, the VRL on glass is comparable with the VRL on stainless steel. Residues on polytetrafluorethylene (Teflon, DuPont) coupons were not visible at the highest levels, > 5 µg/cm2. The observed VRLs for all of the residues tested on the three materials of construction are listed in Table V.

VRL (µg/cm2)

Stainless steel

Glass

Polytetrafluoroethylene (Teflon, DuPont)

Aspirin

< 0.5

< 0.4

> 5

Caffeine

< 0.4

< 0.4

> 5

Calcium

< 0.6

< 0.6

> 5

Diphenhydramine

< 1.6

< 1.6

> 5

Ibuprofen

< 1.0

< 1.6

> 5

The high VRLs of residues on polytetrafluorethylene (Teflon, DuPont) and similar materials of construction can be addressed using one of several options: use individual VRLs for each material of construction; use the highest VRL of the residue as the overall limit for all materials of construction; use the VRL for stainless steel and mitigate the risk for other materials of construction.

Using a separate VRL for each material of construction is a viable approach as long as the highest VRL is lower than the acceptable cleaning limit (ARL) but it would be logistically challenging. Maintaining accurate records for each residue and material of construction as well as training site personnel would require significant site resource expenditures. The larger issue would be in those instances where the VRLs for some materials of construction are below the ARL, but the VRLs are higher than the ARL for other materials of construction. In those instances, a secondary confirmation would be necessary for those materials of construction with the high VRL. This scenario would be likely considering the high VRLs on polytetrafluorethylene (Teflon, DuPont) compared to those on stainless steel and glass.

Use of the highest VRL as the overall limit is logistically more manageable and is a viable approach as long as it is lower than the ARL. The risk of a cleaning failure, however, becomes much greater the closer the VRL is to the ARL. This approach could severely limit the application of VRLs for cleaning determinations.

Continued use of the established VRL on stainless steel would need to address the higher visible limits on other materials of construction or accept the risk that these materials are also clean. Acceptance of the risk that the materials with high VRLs are clean might be reasonable if the material of construction in question comprised a small percentage (i.e., < 5%) of the total product contact surface area of the manufacturing equipment, which would be the case for gasket material or sight glass material on a tank.

However, it would be preferable to either mitigate or address the risk of higher VRLs. One approach to address higher VRLs would be to determine the relative cleanability of residues on the site materials of construction (11). If the residues rinse from the MOC coupons at the same rate as stainless steel or the cleaning cycle is longer than all of the cleanability times, it could then be concluded that if the stainless-steel product contact surfaces are clean, the other product contact surfaces are also clean. Although cleanability experiments are additional work, on top of strengthening a site’s visual-inspection program, they provide objective data to establish the hardest-to-clean residue at the site. For a site with a large number of residues, a paper exercise (12) could group and/or eliminate a majority of the residues prior to performing the laboratory cleanability studies.

Conclusions
VRL have been established for different materials of construction. VRLs were higher on some materials of construction than on stainless steel. The relationship of the coupon, observer, and light source was also different for other materials of construction, making optimal viewing conditions material of construction dependent. Risk of a cleaning failure due to visual failure for different materials of construction can be mitigated or eliminated using complementary developmental cleaning process and validation studies.
Part I of this series of articles examined the ruggedness of VRL viewing conditions and defined optimal viewing conditions (8). In Part II, current detection methodologies were challenged and the author discussed questions about the constant parameters of VRL studies that remained to be answered (9).

References

1. FDA, “Guide to Inspection of Validation of Cleaning Processes,” (Division of Field Investigations, Office of Regional Operations, Office of Regulatory Affairs, July 1993).
2. D. G. Dolan et al., Regul. Toxicol. Pharmacol. 43, 1-9 (2005).
3. R. J. Forsyth and D. V. Haynes, Pharm. Technol. 22 (9) 90-112 (1998).
4. G. L. Fourman and M. V. Mullen, Pharm. Technol. 17 (4) 54- 60 (1993).
5. Code of Federal Regulations, Title 21, Food and Drugs (Government Printing Office, Washington, DC) Part 211, Section 67.b.6.
6. R. J. Forsyth, V. Van Nostrand and G. Martin, Pharm. Technol. 28 (10) 58-72 (2004).
7. R. J. Forsyth and V. Van Nostrand, Pharm. Technol. 29 (10) 152-161 (2005).
8. R. J. Forsyth, Pharm. Technol. 33 (3) 102-500 (2009).
9. R. J. Forsyth, Pharm. Technol 35 (3) 122-128 (2011).
10. R. J. Forsyth, J. C. O’Neill, and J. L. Hartman, Pharm. Techno.l 31 (10)
    102-116 (2007).
11. R. J. Forsyth, “Cleanability of Pharmaceutical Residues from Different Materials of Construction,” in press.
12. A. J. Canhoto, et al., J Valid. Technol. 11 (1) 6-15 (2004)

Richard Forsyth* is a senior consultant at Forsyth Pharmaceutical Company, 907 Shamrock Ct. Royersford, PA 19468, Keith Bader is senior director of technology, and Kelly Jordan is an engineer, both at Hyde Engineering + Consulting, 6260 Lookout Rd, Suite 120 Boulder, Colorado 80301, forsythpharmaconsulting@gmail.com.

*To whom correspondence should be addressed.

Submitted: Feb. 1, 2013. Accepted: Mar. 3, 2013.