Biowaste Management During Biopharmaceutical Plant Start-Up: From Regulatory Guidance to Verified Inactivation Methods

Chinese hamster ovary (CHO) cells are one of the most commonly utilized expression systems for the production of biopharmaceutical products, principally because they have been well-characterized and there is a history of regulatory approval for recombinant proteins produced from these cells. Employing recombinant DNA technologies, CHO cells used in the biopharmaceutical industry are genetically manipulated enabling the production of proteins of interest. These cells are classed as genetically modified microorganisms (GMM). Current European Union (EU) legislation requires the biopharmaceutical industry to comply with regulations governing the contained use of GMM (Council Directive 98/81/EC, amending Direc-tive 90/219/EEC).¹ In Ireland, the Environmental Protection Agency (EPA) is the competent authority responsible for the implementation of Directive 98/81/EC, transposed into Irish law by the Genetically Modified Organisms (contained use) Regulations, S.I. No. 73 of 2001. The directive covers any activity involving the genetic manipulation, culturing, usage, storage, and destruction of GMM. Under the directive, there is a mandatory requirement for waste-containing Class 2 to 4 GMM to be inactivated prior to discharge. Inactivation of Class 1 GMM is optional but may be stipulated in certain circumstances by the competent authority. Inactivation refers to the destruction of GMM, ensuring that subsequent contact between the GMM and the general public or environment is limited, thereby providing an enhanced level of protection.

GMM inactivation may be achieved by thermal or chemical treatment. Heat treatment is the most common method of inactivating liquid effluent and may involve autoclave decontamination or the use of heat inactivation or "kill" systems.^2,3 Chemical treatment involves the addition of a bactericidal agent capable of inactivating the GMM and generally consists of the manual addition of chemicals such as chlorine, caustic solutions, or iodophor compounds.

In consultation with the EPA, a strategy was established for managing GMM waste during start-up and commercial operations at Wyeth BioPharma's new biopharmaceutical manufacturing facility in Ireland (Figure 1). The EPA required that all GMM waste be inactivated prior to discharge from the facility, including Class 1 organisms. This is a customary approach for most companies in the EU. Verification of cell-inactivation methodology prior to its site-wide implementation was also agreed upon in consultation with the EPA. A routine inactivation strategy based on autoclave inactivation and a heat "kill" system was devised for solid- and large-volume liquid waste, respectively. Cell inactivation based on sodium hypochlorite (NaOCl) addition was planned for routine use in the quality control and development laboratories for small culture volumes. Inactivation by sodium hydroxide (NaOH) addition to large-volume liquid waste was devised for use during the cell culture facility start-up, prior to commissioning and qualification of the "kill" system. It would also be available as a backup for the "kill" system during routine operation.

Figure 1. Wyeth Medica Ireland's New Biopharmaceutical Campus in Dublin, Ireland

Literature supporting the effectiveness of heat-inactivation systems for the destruction of mammalian cells is available.⁴ Although chemical treatment is a classical method of cell inactivation, there is a shortage of published information pertaining to its modes of use and efficacy for CHO cells and other GMM. Bench-scale trials were performed to characterize the chemical-inactivation schemes. This paper outlines the definition and verification of the NaOH- and NaOCl-based processes for inactivation of a CHO cell line producing a recombinant therapeutic protein.

CHEMICAL INACTIVATION STRATEGY

NaOCl is a broad-spectrum antimicrobial agent.⁵ However, its corrosive nature renders it unsuitable for in situ cell inactivation in stainless-steel stirred tank reactors. NaOH, as a disinfectant and cleaning agent, is compatible with stainless-steel tanks, is effective in dissolving proteins and denaturing nucleic acids, and is widely used in the pharmaceutical industry for cleaning, sanitizing, and system storage.⁵ As NaOCl and NaOH are sensitive to the chemical environments in which they are used and their activity may be modulated by the presence of large amounts of protein, validation of the inactivation procedures is required to confirm that inactivation is effective and reproducible within normal working conditions at the site.

At commercial production-scale (12,500 L) the recombinant CHO cells are grown in either protein-containing or protein-free media, in flasks or stirred-tank reactors. At various stages of the process the cells are present at densities ranging from 3.0x10⁵ to 4.5x10⁷ cells/ml. A strategy of "out-of-place" NaOCl inactivation of flask cultures and in situ NaOH inactivation of stirred-tank cultures was planned. Flask cultures signify cells growing in shake- or spinner-flasks, in protein-containing media, and at densities up to 3x10⁶ cells/ml. Inactivation was planned to involve the transfer of the cells to a reservoir and the addition of concentrated NaOCl stock solution to a pre-defined concentration of active chlorine and for a predetermined duration. GMM inactivation by NaOH addition to a predetermined concentration and for a defined time was planned for cells grown in stirred-tank reactors. As the cells are maintained in a variety of conditions in stirred-tank reactors, a single inactivation strategy that would encompass all conditions was essential. Although all stirred-tank reactors contain mixing and pH-monitoring capabilities, under specific circumstances inactivation of cultures at volumes lower than the level of the pH probe, impeller, or sample port could be required. To counteract the potential inability to monitor online pH or to easily recover a sample for off-line analysis, the inactivation method was designed to achieve a predetermined volumetric concentration of NaOH, rather than a specific pH.

MATERIALS AND METHODS

Recombinant CHO cells were grown as suspension culture in flasks (shake-flasks and spinner-flasks) and glass (Applikon) reactors, in either protein-containing or protein-free media. Flask cultures were grown without external pH control, while the pH of cultures grown in reactors was controlled at a set point of pH 7.2, by the addition of CO₂ and 1 M sodium carbonate.

Prior to cell inactivation, the density of the recombinant CHO culture was determined and cells were transferred to 60 ml sample tubes. Cell densities were adjusted by the addition of phosphate-buffered saline (PBS) or growth medium, if required. Concentrated NaOH (10 M) or NaOCl (13 percent active chlorine) was added to the cells to a predetermined final concentration. A constant volume and cell density was maintained in all sample tubes by the appropriate addition of PBS. Sample tubes were inverted to incorporate the inactivation solution into the cultures, which were then maintained at room temperature. Samples were withdrawn routinely, from 1 minute up to 90 minutes post-addition of the inactivating agent, and analyzed for viable cell numbers. Viable cell density was assessed using a Cedex automatic cell counter (from Innovatis AG), with viability determined by trypan blue exclusion. All test conditions were assessed in duplicate. The inactivation trial process is outlined in Figure 2. One hour after the addition of the inactivating agent, the pH of the NaOH-inactivated culture was measured. Statistical analysis was performed by analysis of variance (ANOVA), using Design-Expert software (from Stat-Ease, Inc.). The probability values (p-values) less than 0.05 indicated that test parameters were statistically significant.

Figure 2. Schematic of the Bench-scale Inactivation Experiments Small-scale experiments were conducted on recombinant CHO cells to establish inactivation conditions for site-wide implementation. Viable cell density (VCD) was measured using a Cedex automatic cell counter.

Cell inactivation in both protein-free and protein-containing media was analyzed in the NaOH inactivation study. Culture inactivation at high (5.0x10⁷ cells/ml; a cell density approximately 10 percent higher than the maximum density expected in the commercial process) and low (5.0x10⁶ cells/ml) densities was assessed. NaOCl inactivation was also assessed at high and low cell densities (3x10⁷ cells/ml and 3x10⁶ cells/ml) but only in protein-containing medium, reflecting the planned use of this inactivation process.

NaOCl INACTIVATION

NaOCl proved an effective inactivating agent for the GMM cells suspended in protein-containing media. NaOCl concentration, culture density, and length of contact between the inactivating agent and culture were significant factors in determining the efficacy of cell inactivation (Figures 3a and 3b). The level of inactivation was proportional to both the concentration of NaOCl (p-value < 0.0001) and duration of contact between the inactivating agent and culture (p-value = 0.0010). Cell inactivation was inversely proportional to the density of the culture (p-value = 0.0215).

Figure 3. NaOCl Inactivation of Recombinant CHO Cells CHO cells are suspended in protein-containing medium at high (3x107 cells/ml (a)) and low (3x106 cells/ml (b)) cell densities. Following exposure to 0%, 0.25%, and 0.5% (active chlorine) NaOCl, average cell kill (expressed relative to the VCD at time 0) is graphed against time.

At high densities (approximately 3x10⁷ cells/ml), increasing but incomplete cell inactivation was observed with increasing time and NaOCl concentrations, up to 0.5 percent active chlorine (Figure 3a). At lower densities (3x10⁶ cells/ml), complete cell inactivation was seen within one minute at NaOCl concentrations ≥0.25 percent active chlorine, i.e., any residual viable cells were below the level of detection of the system( Figure 3b).

NaOH INACTIVATION

The addition of NaOH to a culture of GMM cells also resulted in cell inactivation. The final concentration of NaOH, the contact time between the cells and inactivation solution, and the protein content of the media were significant factors in determining the extent of cell kill (Figures 4a, 4b, 4c, and 4d). Increased cell kill was observed with increasing concentrations of NaOH (p-value < 0.0001). Recombinant CHO cells suspended in protein-free media appeared more susceptible to inactivation than cells in protein-containing media (p-value < 0.0001). However, NaOH-induced inactivation was independent of the initial culture density (p-value = 0.1236).

Figure 4. NaOH Inactivation of Recombinant CHO Cells. CHO cells are suspended in protein-containing (a, b) and protein-free medium (c, d), at high (5x107 cells/ml (a, c)) and low (5x106 cells/ml (b, d)) cell densities. The average cell kill (expressed relative to the VCD at time 0) is graphed against time for increasing concentrations of NaOH.

GMM inactivation was dependent on both NaOH concentration and media composition. In protein-containing media, NaOH addition to a final concentration ≤ 0.025 M resulted in negligible cell kill over a one-hour period. At concentrations ≥0.05 M, cell viability fell to below detectable levels within one minute of contact time( Figures 4a and 4b). However, in protein-free media, culture inactivation was achieved at lower NaOH concentrations; addition of 0.0125 M NaOH resulted in viable cell density (VCD) falling to below detectable levels within 30 minutes (Figures 4c and 4d).

Culture inactivation in both protein-containing and protein-free media was independent of cell density. In protein-containing media, inactivation of recombinant CHO cells by 0.05 M NaOH was comparable at both 5x10⁶ and 5x10⁷ cells/ml (Figures 4a and 4b). In protein-free media, cell inactivation was also comparable at both high and low cell densities but inactivation was achieved at a lower NaOH concentration (0.0125 M) (Figures 4c and 4d).

Following NaOH addition to the cultures, an increase in solution viscosity was noted. The increase in viscosity is a consequence of cell lysis, as the detection of both viable and dead cells decreased with increasing culture viscosity. From visual inspection, culture viscosity was broadly dependent on NaOH concentration, initial culture density, and the duration of exposure to NaOH. Increased NaOH concentrations, cell densities, and exposure time resulted in the generation of more viscous solutions. The viscosity of the solution prevented the accurate determination of cell densities. However, when the solution was examined microscopically, no cells (viable or dead) were observed.

The pH of the NaOH-inactivated culture was dependent on both the final concentration of NaOH (p-value = 0.0020) and the culture density (p-value = 0.0029), but was independent of the media composition (p-value = 0.9710) (Figure 5). The pH of the high-density control cultures (cultures unexposed to NaOH for 60 minutes) was significantly lower than the pH of low-density cultures. This may be accounted for by cellular metabolism during the study hold-time, such as lactate production by the viable culture. The finding was observed for cultures suspended in both protein-containing and protein-free media. Following inactivation, the extent of the pH increase in the culture was dependent on the amount of NaOH added. However, the effect of cell lysis and the associated release of intracellular components on the final pH was not accounted for in this study. In generating low-density cultures, high-density cultures were diluted with either PBS or culture medium. To minimize disparities in results due to the diluents, the pH of the PBS and media were identical and the same as the culture media in which the cells were grown, pH 7.2. PBS has an effective pH buffering range of 5.8 to 8.0. Both the protein-containing and protein-free culture media contain 15mM HEPES, which at room temperature has a pKa of 7.5 and an effective buffering range of 6.8 to 8.2. The similarities in solution pH and buffering range would indicate that cultures should behave similarly in response to NaOH addition, whether the cultures were diluted with PBS or media, or undiluted.

Figure 5. Increase in Culture pH Following NaOH Addition to Recombinant CHO Cells Suspended in Protein-containing and Protein-free Medium. The average pH of the inactivated cultured is graphed against final NaOH concentration.

A direct correlation between the pH of the inactivated GMM culture and the final NaOH concentration could not be established due to the influence of cell density and media composition. Addition of NaOH to GMM cultures suspended in protein-containing media to a final concentration of 0.025 M NaOH resulted in a final pH of 8.9 and 9.5, for high- and low-density cultures respectively. In protein-free media, addition of NaOH to a final concentration of 0.025 M resulted in a final pH of 7.5 and 11.9 for high- and low-density cultures respectively. However, 0.025 M NaOH resulted in low or negligible cell kill in the protein-containing culture, with 100 percent cell kill achieved in the protein-free culture.

CONCLUSIONS

To implement an effective inactivation system and to ensure inactivation in all circumstances, parameters were selected that significantly exceeded the conditions concluded from inactivation trials. GMM inactivation with NaOCl has been restricted to low-density cultures, where 0.25 percent (active chlorine) resulted in complete inactivation within one minute. To increase confidence in the inactivation process, the contact time between the CHO cells and NaOCl was extended. A NaOCl-inactivation protocol consisting of a room-temperature incubation of CHO cells, up to a maximum concentration of 3.0x10⁶ cells/ml with a volume of NaOCl ≥ a final concentration of 0.25 percent active chlorine and for a contact time equal to or greater than 30 minutes, has been implemented. When densities in excess of 3.0x10⁶ cells/ml were encountered, the culture was diluted to below this limit prior to inactivation.

In the inactivation study, 0.05 M NaOH was shown to effectively destroy the CHO cell line within one minute of contact with the culture, under all culture conditions employed. Again, to increase confidence in the inactivation procedure, the contact time between NaOH and cells was extended to 30 minutes. For routine NaOH inactivation, a strategy of a room-temperature incubation of a maximum concentration of 5.0x10⁷ cells/ml with a minimum concentration of 0.05M NaOH and for a minimum duration of 30 minutes was implemented. NaOH inactivation was based on attaining a predetermined final concentration of NaOH as opposed to attaining a target pH, as a direct correlation between the pH and the NaOH concentration of the inactivated GMM culture could not be established. The inability to easily assess the pH of inactivated culture in all circumstances, due to low culture volume in some instances and the positioning of probes and sample ports on the stirred-tank reactors, also contributed to the adoption of this strategy. The generation of highly viscous solutions during NaOH-induced cell inactivation should not be ignored when implementing inactivation strategies, as it may present a challenge when removing cultures from stirred-tank reactors.

SUMMARY

In formulating the inactivation strategy for GMM at our biopharmaceutical manufacturing facility, a combination of inactivation procedures was selected, reflecting the nature of the waste and phase of the commercial operation.

• Autoclave inactivation was employed for solid waste.

• A heat "kill" system was employed as the principal means of liquid-waste inactivation. Due to the availability of literature citing the use of "kill" systems, a scheme of commissioning following an annual verification of its effectiveness was adopted.

• Cell inactivation based on NaOCl addition was employed for small culture volumes and used routinely in the quality control and development laboratories. Inactivation consists of exposing cultures at a maximum density of 3x10⁶ cells/ml to a minimum concentration of 0.25 percent (active chlorine) NaOCl for at least 30 minutes.

• During start-up operations and prior to commissioning of the "kill" system, GMM inactivation by NaOH addition was implemented for large-volume liquid waste. NaOH inactivation of GMM would also be available as a backup for the "kill" system in instances of unforeseen equipment breakdown. Inactivation consists of exposing a maximum of 5x10⁷ cells/ml to a minimum concentration of 0.5 M NaOH for at least 30 minutes.

• Due to the dearth of published literature on chemical inactivation of GMM, the NaOH- and NaOCl-inactivation procedures were verified at bench-scale prior to their site-wide implementation. Following establishment of a chemical-inactivation regime, the procedure was implemented with a commitment to an annual verification of its efficacy.

Had pertinent data relating to chemical inactivation of CHO cells been available at the start-up of our facility, it would have been possible to move quickly and directly to an annual verification of the chosen methods. This paper should supply to other groups valuable in-formation that may be used in justifying the selection and use of similar inactivation procedures in their own facilities.

Mary Heenan, senior scientist, Process Development, Development & Technical Services, The Wyeth BioPharma Campus at Grange Castle, Wyeth Medica Ireland, Grange Castle International Business Park, Clondalkin Dublin 22, Ireland, 01.4696814, Fax: 01.4696896, heenanm@wyeth.com

REFERENCES

1. Fink R, Moran E. Biosafety for large-scale containment level 1 operations using recombinant DNA technology: No emerging hazards. Applied Biosafety. 2005; 10:30-39.

2. Ebling MJ, Miller S. Validation of a batch type biological waste treatment system. Pharmaceutical Engineering. 1991;11:42-48.

3. Carlson CJ. Biowaste Systems. Pharmaceutical Engineering. 2001;21:70-82.

4. Gregoriades N, Luzardo M, Lucquet B, Ryll T. Heat inactivation of mammalian cell cultures for biowaste kill system design. Biotechnol. Prog. 2003;19:14-20.

5. Block SS, ed. Disinfection, Sterilization, and Preservation. 5th ed. Philadelphia: Lippincott, Williams & Wilkins; 2001.