Biopharmaceutical Facility Cleaning Validation Using the Total Organic Carbon Test

ABSTRACT

The total organic carbon (TOC) test is a fast and effective analytical technique to evaluate the cleaning of biopharmaceutical manufacturing equipment. This technique can help ensure that the cleaning processes meet predetermined cleanability criteria for single and multiproduct production areas. This article presents a case study describing the use of the TOC test to validate the cleaning processes used for two types of biomanufacturing equipment.

Anumber of US and European documents describe the requirements and guidelines for biomanufacturing equipment cleaning processes.¹ Also, many journal articles have discussed strategies for performing cleaning validation.^2–5 However, cleaning validation problems, including a lack of documented procedures, inadequate training of operators, and insufficient validation of analytical or cleaning methods, still are among the four most commonly cited problems in Form 483s and warning letters issued by the US FDA.^6,7

Baxter Healthcare

This article discusses a strategy to validate biopharmaceutical facility cleaning processes using the total organic carbon method (TOC). An initial TOC measurement system qualification was performed, followed by an evaluation of the correlation between TOC and microorganism levels. Later, a swabbing recovery study was carried out with cells and proteins. Finally, the cleaning process for two types of biomanufacturing equipment was validated using the TOC test.

MATERIALS AND METHODS

MATERIALS Biological Samples

The cells and proteins used in the swabbing recovery study included Escherichia coli bacterial cells, Saccharomyces cerevisiae yeast cells, recombinant streptokinase (SK), recombinant epidermal growth factor (EGF), recombinant human alpha interferon (IFNα), and pegylated recombinant human alpha interferon (IFNα–PEG).

TOC Vials and Swabs

Glass 40-mL autosampler vials with caps were used for TOC measurements. For the recovery study a TX3340 TOC cleaning validation kit, containing 12 Eagle Picher 03464 40-mL clear vials, 24 Texwipe TX714K large SnapSwabs, and 12 blank vial labels was used. The 40-mL vials were certified as having TOC levels <10 ppb and the TOC levels of the swabs were certified at <50 ppb.

METHODS TOC Measurement Method Validation

The TOC test involves full oxidation of organic carbon and detection of the resulting CO₂. In this study, a TOC analyzer equipped with an auto-sampler was used. The analyzer measures TOC according to ASTM method D6317.⁸ It determines the amount of total carbon (TC), inorganic carbon (IC), and TOC in water in the range of 10 to 1,000 μg/L. The test method used persulfate and ultraviolet (UV) oxidation of organic carbon, coupled with a CO₂ selective membrane to recover the CO₂ in deionized water. The change in conductivity of the deionized water was measured and compared with the carbon concentration in the oxidized sample. IC was determined in a similar manner, but without the oxidation step. In both cases, the sample was acidified to facilitate CO₂ recovery through the membrane.

The relationship between the conductivity measurement and carbon concentration is described by a set of chemometric equations for the chemical equilibrium of CO₂, HCO₃–, and H+, and for the relationship between the ionic concentrations and conductivity. The chemometric model includes the temperature dependence of the equilibrium constants and the specific conductance, resulting in a linear response of the method over the stated range of TOC.

The TOC measurement method was validated according to the ICH Q2 (R1) guideline.⁹ Precision and accuracy were calculated using the TOC sucrose standard of TOC. For precision, the standard deviation (SD), and relative standard deviation (RSD) for the three TOC concentrations (250, 500, and 750 ppb; with three replicates for each concentration) readings were determined as follows:

in which Σ is the sum of each result and n is the number of measurements in a set (number of replicates – number of rejections). The RSD = (SD/measured TOC concentration) x 100.

The accuracy (percent difference between both the expected and measured TOC value) was calculated as follows:

% difference = (measured concentration – expected standard concentration/expected standard concentration) x 100.

Method linearity was verified using the sucrose standard at three concentrations, 250, 500, and 750 ppb of TOC (for each concentration with three replicates). The correlation coefficient (R²) was calculated as follows:

in which X is the certified values of TOC standards and Y is the measured values of TOC standards (blank–corrected).

The limits of detection and quantitation (LOD and LOQ, respectively) were calculated using three sucrose standards at 250, 500, and 750 ppb of TOC (for each concentration with three replicates). Using a least squares linear regression algorithm, the equation of the line relating the SDs of the standards to the actual measured concentrations of the standards was determined. The y-intercept represents the SD at zero ppb of standard sucrose concentration. The LOD was the yintercept value multiplied by 3, and the LOQ was the y-intercept value multiplied by 10.

Sample Preparation for Determining the Correlation Between TOC and Microorganisms

Both bacterial and yeast broths were grown until an optical density of 0.6 was reached, at 500 nm and 600 nm, respectively. The culture media used to grow bacteria was LB (5 g/L yeast extract, 10 g/L triptone, 10 g/L NaCl). The culture media used for yeast was YPD (20 g/L glucose, 20 g/L peptone, and 10 g/L yeast extract). These cell suspensions were centrifuged in a TJ-6 centrifuge (Beckman Instrument, Fullerton, CA) for 10 min at 1,000g. The pellets were washed three times and suspended with low TOC purified water. The serial dilutions for wet cell concentrations shown in Figure 1 also were prepared using the same purified water.

Figure 1

Swab Recovery

The coupons used in this work were representative of the materials of construction of the manufacturing equipment.

Therefore, both 100 cm² stainless steel (316 L) and 100 cm² borosilicate coupons were used in the swab recovery study. One side of each coupon was spiked with a solution (or biologic sample) of cells or pure recombinant proteins, respectively. For bacterial and yeast cells, 200 μg, were applied. In the case of recombinant proteins, 1.2 mg of SK, 100 μg of EGF, 100 μg of IFNα, and 12.5 mg of IFNα–PEG, were applied. The coupons were allowed to dry completely for 4 h at 25 °C. A swab was moistened with low TOC water and the spiked coupon surfaces were swabbed both vertically and horizontally. The swab end was cut off and placed in a vial with 40-mL of TOC water. The vials were capped, shaked vigorously by inversion, and the TOC analysis was performed. The same quantity of each analyzed biologic sample was spiked directly in 40-mL of TOC water (reference control).

Manufacturing Equipment Cleaning

The cleaning process was performed using sodium hydroxide and orthophosphoric acid, as described by Verghese¹⁰ and Mollah,⁵ following the standard procedures previously developed by the manufacturing department and approved by quality assurance staff of the Center for Genetic Engineering and Biotechnology (CIGB).^11–13 The final rinses of the fermentation and chromatography equipment were performed with purified water and water for injection (WFI), respectively.

RESULTS AND DISCUSSION

TOC Measurement System Performance

According to the European good manufacturing practice (GMP) regulations, analytical methods used in biomanufacturing equipment cleaning should be able to detect the established acceptable level of the residues or contaminant.¹⁴ Also, according to US FDA cGMP regulations and the USP, the suitability of all compendial testing methods used must be verified under normal laboratory conditions.^15–16

Per the above regulations and considering that this is a quantitative test for impurities, the analytical parameters including accuracy, precision, linearity, detection limits, and quantiation limits, were evaluated with their corresponding assumption, per ICH Q2 (R1).⁹

The results obtained for each of the above parameters during the TOC analysis are summarized in Table 1. In the case of precision, 1.27% of RSD corresponds to 7.21 ppb of TOC. The linearity showed a fit correlation between ppb of TOC and the conductivity signal from its full oxidation. The lowest amount of TOC quantified in 21 ppb (21 ng/mL) demonstrated high sensitivity of this measurement system.

Table 1. Validated analytical parameters and the key results obtained

In addition, the suitability test was performed according to the procedure and specifications described in section 643 of the USP31–NF26 and 2.2.44 of the European Pharmacopeia (EP, data not shown).^16–17

Correlation Between TOC and Microorganisms

The bacterial and yeast cells commonly used in fermentation represent the "worst case scenario" in manufacturing system cleaning, because they lead to organic contaminants such as nucleic acids, lipids, carbohydrates, proteins, and endotoxins. The TOC test should provide a linear correlation between the measured molecules and the TOC response.

To analyze the relationship between cell concentration and TOC response, a correlation analysis was performed (Figure 1). In both cases, correlation coefficients were >0.983, showing linear relationship between wet cell concentration and the TOC values. Therefore, contamination with cells or any of those components can be quantitatively determined using the TOC measurement in a wide dilution range. These data can be used to extrapolate the potential wet cell concentration with its corresponding cell number estimate or relative organic carbon contribution in the final results of any cleaning process of a bioprocess equipment.¹⁸

Considering that both bacterial and yeast cultures were harvested at an exponential growth phase, we assumed that the majority of the cells were viable, although some lysed cells were not discarded.

It is significant to note that in first 27 ppb (27 ng/mL) of TOC, one could have a wet bacterial cell concentration of 1.29 mg/mL, which is equivalent to 10⁶ E. coli cells. Similarly, in as little as 16 ppb (16 ng/mL) of TOC, one could have a wet yeast cell concentration of 0.826 mg/mL, which is equivalent to around 10³ yeast cells.

However, according to the WFI specification (500 ppb of TOC) indicated in EP16 and the USP17, the TOC values (27 and 16 ppb) are more than one order below the specification. Thus, the amount of residue found in a final wash sample from manufacturing equipment cleaning could be considered as the "acceptable level" of undesirable residues. To use this value as acceptable level, the TOC value of the water used for cleaning the manufacturing system must be taken as the cleanliness reference.

Swabbing Recovery Study for Cells and Proteins

There are two types of sampling done following the cleaning process—the direct surface and rinse samples. The first one is considered "the most desirable" by the FDA.¹ The swab technique typically involves moistening a polyester swab with purified water to wipe a measured area in a systematic manner. Cleaning validation kits, specifically designed for TOC swabbing, are commercially available for this purpose. However, it is critical to validate the swabbing procedure in combination with the chosen analytical method for various combinations of contaminants and surfaces. Thus, TOC recovery study is mandatory before evaluating the efficiency of the cleaning process. In the absence of such a validation study, a manufacturer may erroneously assume that the equipment is clean, based on an apparent negative result.

To evaluate the efficiency of swabbing, the "worst case scenario" was evaluated using fermentation and purification facilities as examples. Moreover, three recombinant proteins were included in this study: SK and interferona 2b human recombinant (IFNα2bHu-r) were obtained from bacterial cells. The EGF was expressed in yeast cells. In all cases, the studies were performed considering the interaction of these cells with different manufacturing processing surfaces, which may in turn significantly affect the swabbing recovery during cleaning validation. The results obtained from microbial host and recombinant proteins on stainless steel and borosilicate glass are shown in Table 2. The mean of recovery factor ranged from 89 to 108%. For the cells, the results are closest to 100% recovery from both surfaces. These results are superior to those published in a similar study by using bacterial whole cell homogenate and stainless steel coupons.¹⁸ Those studies obtained recovery values of 67%, using 0.05 N phosphoric acid as the extraction solution. In a similar study, a recovery mean of 81.9% was obtained using 1 N sodium hydroxide as the solvent using E. coli and stainless steel coupons.²

Table 2. Values (%) of the cells and recombinant proteins recovered in the recovery study (SD: standard deviation; CV: coefficient of variation)

For the recombinant proteins, the lowest recovery value was obtained with SK on stainless steel coupons with a general recovery factor of 89.47%. The remaining recombinant protein–surfaces combinations showed recovery factor values ranging from 93.89 to 107.86%. These results match with those obtained by Lombardo, et al. in a similar study, but using purified recombinant human ciliary neurotrophic factor (rHCNTF) on a stainless steel surface and 0.05 N phosphoric acid as the extraction solution.¹⁸ However, Lombardo's results from borosilicate analysis (55% of recovery factor) were lower than those obtained in our study (from ~94 to 108%).

These results suggest that purified water with low levels of TOC can be used efficiently to recover all biological materials. In addition, the almost identical recovery factor values for the IFNα2bHu-r–PEG and IFNα2bHu-r show that the polyethylene glycol conjugation process does not affect the electrostatic interaction with the borosilicate surface and can be removed efficiently under the conditions described (see materials and methods above).

CLEANING VALIDATION OF TWO TYPES OF EQUIPMENT

Considering the results shown above, the TOC test was used to assess the cleaning of bioproduction equipment used in two types of manufacturing processes. The first set of equipment was used to obtain a recombinant protein from E. coli bacterial cells as host, and in the second scenario, S. cerevisiae yeast cells are used to express a heterologous protein. In both manufacturing systems, one fermentation vessel and one chromatography column were taken as examples.

Before the cleaning process was performed, the procedures used (who is responsible for performing and approving them, acceptance criteria, revalidation) were documented in detail.¹ Then, the cleaning process was performed, as described in preapproved protocols.^11–13

The results of three independent cleaning processes for both manufacturing systems are shown in Tables 3 and 4. For the fermenter vessel, cleaning validation results from rinse samples showed a reduction factor of three orders of magnitude, with final values ranging from 1.27 x 10³ to 2.25 x 10³. The values from the harvest port swabbing samples showed the same trend as for rinse samples, with a three orders of magnitude, and final values ranging from 1 x 10³ to 3.73 x 10³ .

Table 3. Cleaning validation results (TOC, in ppb) for fermenters used in E. coli and S. cerevisiæ production processes

On the other hand, the results of chromatography column cleaning showed, in general, lower reduction factor values compared with those for fermenter cleaning. Those results were expected, knowing that the chromatography system is in contact with cleaner material than that seen in fermentation steps.

The final samples (rinse and swabbing samples) of both manufacturing systems showed TOC values as low as 22 ppb (close to the LOQ, Table 1). The TOC value obtained for the water used to clean the biomanufacturing system was used as a reference value.

The results of the swabbing recovery study (Table 2) showed that the levels of organic charge reduction after cleaning reported in this work, showed a high level of cleaning efficiency. This proposed cleaning validation strategy can be applied to both single and multiproduct manufacturing facilities.

Table 4. Cleaning validation results (TOC, in ppb) for chromatography columns used in E. coli and S. cerevisiæ production processes

CONCLUSIONS

A general strategy for assessing cleaning validation using the total organic carbon (TOC) test was applied. The qualification of the TOC measurement system showed a precision of 1.27%, 1.3% of accuracy, and a detection limit of 6.30 ppb of TOC.

A linear correlation between cells and TOC values was demonstrated with this analytical system. This study showed that in both 27 ppb and 16 ppb of TOC, ~10⁶E. coli cells and 10³S. cerevisiae cells were present, respectively. The swab recovery assessment showed recovery factor values ranging from 90 to 110% using stainless steel and borosilicate surfaces and SK, EGF, IFNα2bHu-r, as well as both E. coli cells and S. cerevisiae as biological matrices, and purified water as a recovery solvent. It also showed that applying the polyethylene glycol conjugation process to IFNα2bHu-r does not affect its interaction with borosilicate surfaces. Given the cleaning reduction factors values, the final TOC values for the studied systems, the agreement between them, and the TOC values from final samples and for the water used to clean each system, the strategy described here can be a useful tool in similar cleanliness validation strategies.

ACKNOWLEDGMENTS

The authors want to thank Magaly García Blanco for proofreading this manuscript and the epidermal growth factor and interferon manufacturing personnel from the CIGB for supplying the cleaning validation samples used in this work.

Julio César Sánchez is the head, and Raudel Sosa and Meily Sánchez are technicians in the process control department, Rebeca Bouyon Albarran is a research scientist in the epidermal growth factor department, Luciano Hernández and Marbel Ramos Alfonso are research scientists in the interferon manufacturing department, Alexis Musacchio Lasa is a research scientist in the bioinformatics department, and Leopoldo Núñez is a technical service assistant, all at the Center for Genetic Engineering and Biotechnology, Havana, Cuba, +537.271.6022 (ext. 7101), julio.sanchez@cigb.edu.cu

REFERENCES

1. The US Food and Drug Administration. Mid-Atlantic region inspection guide: cleaning validation. Rockville, MD; 1993 Jul. Available from: http://www.fda.gov/ICECI/Inspections/InspectionGuides/ucm074922.htm.

2. Mark AS, Terry LS, Brett TF, Avinash LL. Total organic carbon analysis of swab samples for the cleaning validation of bioprocess fermentation equipment. BioPharm Intl. 1996;9(4):42–5.

3. Clark K. How to develop and validate a total organic carbon method for cleaning applications. J Pharm Sci Technol. 2001;55(5):290–94.

4. Glover C. Validation of the total organic carbon (TOC) swab sampling and test method. J Pharm Sci Technol. 2006;60(5):284–90.

5. Mollah AH. Cleaning validation for pharmaceutical manufacturing at Genentech—Part 2. BioPharm Intl. 2008;22(3):68–75.

6. McCormick D. FDA's Evans reviews causes of warning and recall. Pharm Technol. 2005 Oct. Available from: www.pharmtech.com/pharmatch/content/printContentPopus.jsp?id=190230and?id=383878.x

7. McCormick D. Poor OOS Review leads causes of FDA citations. Pharm Technol. 2006 Nov. Available from: www.pharmtech.com/pharmatch/content/printContentPopus.jsp??id=383878.

8. ASTM international standard. Available from: http://www.astm.org/Standards/D6317.htm.

9. ICH Q2 (R1) Validation of analytical procedures: Text and methodology. Available from: www.ich.org/LOB/media/MEDIA417.pdf.

10. George Verghese. Cleaning agents for biopharma manufacturing. Gen Eng News. 2003;23(6):46–52.

11. IT 60.602 Preparation and application of API to chromatographic systems for cleaning validation in campaign changes.

12. IT 60.201 Preparation and application to Fermentors of biomass used in the cleaning validation in campaign changes.

13. PVL-005 Procedure of parenteral Epidermal Growth Factor manufacturing equipment cleaning validation.

14. Working Party on Control of medicine and inspections. Qualification and validation. Annex 15 to the EU guide to good manufacturing practice. European commission: Brussels, Belgium, July 2001. Available from:

.

15. US Food and Drug Administration. Current good manufacturing practice regulations; 21 CFR 211.194 (a) (2), 2005.

16. United States Pharmacopeia. USP 31–NF 26, USP <643> Total Organic Carbon. 2007 Nov.

17. European Pharmacopeia Method 2.2.44 – total organic carbon in water for pharmaceutical use, European Pharmacopoeia Commission; 6th ed, 2009.

18. Lombard S, Inampudi P, Scotton A, Ruezinsky G, Rupp R, Nigam S. Development of surface swabbing procedures for a cleaning validation program in a biopharmaceutical manufacturing facility. Biotechnol Bioeng. 1995;48:513–519.