Cleaning-in-place of ÄKTA process™ with sodium hydroxide

Microorganisms, such as, bacteria and yeasts are found in many laboratory and production environments and multiply rapidly after incubation. System cleaning is important since regulatory authorities have high standards for purity and low microbial presence in products used in clinical applications. The efficiency of your cleaning method can be evaluated using microbial challenge tests. Sodium hydroxide (NaOH) is frequently used to clean both chromatography equipment and resins since it is documented as an effective cleaning agent, is inexpensive, and readily available.

ÄKTA process™ automated liquid chromatography system is built for process scale-up and large-scale biopharmaceutical manufacturing. The system is available in three flow rate ranges that extend up to 2000 L/h for large volume manufacturing and can be constructed using either electropolished stainless steel or polypropylene, depending on process conditions and plant requirements.

The ability to clean the ÄKTA process™ system from yeast has previously been shown in several studies, therefore, the challenging organisms used in this study were two bacterial strains recommended by the United States Pharmacopoeia (USP 38), Escherichia coli (E. coli) and Pseudomonas aeruginosa (P. aeruginosa). The two organisms represent different types of bacteria that may be involved in contamination and biofilm formation.

Introduction

In large quantities, bacteria and yeasts can damage the function of chromatography columns and impair the performance of chromatography systems. Microorganisms adapt to a wide range of conditions. Many gram-negative bacteria produce endotoxins that can contaminate products. Certain groups, such as Pseudomonas, can proliferate in low nutrient, aqueous solutions and can be found in process chromatography buffers and process water. Consequently, it is important to follow cleaning routines throughout the whole production process.

We describe the cleaning procedure of ÄKTA process™ systems connected with inline dilution (ILD) units after challenging with a mixture of two microbial organisms. The two systems tested were constructed of 1/2 in. stainless steel (SS) and 1 in. polypropylene (PP) tubes, respectively. After challenging with the microbial organisms, systems were cleaned with a predefined cleaning method. The challenging solution was composed of a mix of two organisms recommended by USP 38, E. coli and P. aeruginosa.

Cleaning was performed by a two-step cleaning procedure at 75% of pump speed. First, the system was flushed with 0.22 µm filtered purified water followed by a flush with the 1 M NaOH cleaning agent. Second, the system was set to pause for 1 h followed by a wash out of the cleaning agent with 0.9% NaCl.

Outlets and waste are challenging to clean since they are connected to the outside of the system and extra consideration has to be taken to understand how to clean these areas as well as the flow-path. The waste valves have shown to be especially challenging since the most outlets can be cleaned by letting the CIP manifold create a water lock to contain the cleaning solution on the outside of the valve, but this is not possible for the waste valve. Inlets, although in contact with the outside of the system, are always in contact with the solutions used and thus do not render a problem.

The efficiency of NaOH as a cleaning agent was evaluated by sampling at predetermined points in the system flow path including points where the flow path via a valve met the outside of the system through waste pipes. Each of the two systems was examined for colony forming units (CFUs) of the two challenging organisms through collecting swab and liquid samples. Additional liquid samples were also collected for evaluation of total organic compound (TOC) and endotoxin (EU) contents. All inspected sampling sites in the flow path were visually clean and the results showed that 1 M NaOH effectively cleaned the ÄKTA process™ system flow path from E. coli and P. aeruginosa. The method showed excellent efficiency with a log reduction factor of ≥ 6 of CFUs of the two challenging organisms. The endotoxin level were below 0.5 EU/mL and the TOC level in the range of 0.4 to 1.1 mg C/L.

Extra caution was taken to clean the valve openings to the waste tube as they are self-drained after the valve is closed and therefore the contact time of the cleaning agent is limited. We treated the waste tubes connected to the top valves on the air trap (automatic drain valve) and filter units (manual drain valves) with a constant flow through overfilling both units using 75% of pump speed. The remaining waste tubes connected to the air trap and filter units were flushed with the volumes built up of the cleaning agent in the overfilled air trap and filter units through self-drainage. In general, the waste valves must always be considered as challenging and handled accordingly.

Materials and Methods

Preparation of the test organism

The organisms chosen for the microbial challenge tests were USP 38 recommended bacterial strains (Table 2). The challenging organisms were prepared and stored on Tryptic Soy Agar (TSA) plates at 37°C and 4°C, respectively. The day before each test, the organisms were inoculated in 200 mL of Tryptic Soy Broth (TSB) medium and incubated at 37°C for 16 to 18 h while shaking at 180 rpm. The bacterial cultures were suspended together in sterile 0.9% NaCl solution, and then diluted to an approximate concentration of 10⁷ viable organisms/mL for each organism.

Table 2. Organisms chosen for the microbial challenge tests

Preparation and infection of the system

The system was prepared for cleaning by connecting manifolds at the system inlets, outlets, and column in/out. The inlet manifolds were then connected by tubing to the CIP inlet valves and tanks filled with cleaning and rinsing solutions were connected to the CIP inlet valves (Fig 1). Prior to cleaning, the filter insert in the filter holder was removed. Note that if biomass film is found at the bottom of the filter housing, we recommend that the surface is manually cleaned with a detergent and sprayed with 70% ethanol or spore cleanse and wiped clean. This also applies to sensors that are calibrated externally.

The detachable parts of the system were disassembled, cleaned manually with a detergent, and sprayed with 70% ethanol before reassembling, and the pH electrode was replaced with a plug to avoid damage at extreme pH.

The challenge organism’s suspension was pumped into the system at a pump speed between 33% and 50% of maximal pump speed. The air trap was over-filled, and the filter housing was inline (but without filter) and over filled (manual valve opened). The systems were incubated without flow for 16 to 20 h at room temperature (RT).

Fig 1. Connection of CIP manifolds to inlet A and B.

Cleaning procedure

After incubation, the system was rinsed with purified water then filled with 1 M NaOH at 75% of maximal pump speed making sure all valves were opened and closed at least 2 times. After filling with NaOH, the method was paused for 1 h. The procedure was finished by rinsing with 0.9% NaCl at 75% of maximal pump speed to achieve neutral pH before sampling. The cleaning method is described in Table 3.

Table 3. The defined cleaning method

Microbial sampling

Microbial samples were taken at predetermined swab sampling points (Table 4). Liquid samples were taken at infection, post-infection after incubation for 16 to 18 h, and after ending the cleaning procedure. EU samples and TOC samples were also taken.

Table 4. Number of sampling sites with challenging organisms remaining after cleaning with 1 M NaOH of the ÄKTA process™ chromatography systems including the ILD carts

¹ Numbers within brackets are referred to the ½ in. SS system. The reason for the difference in the number of sampling sites was due to differences in component configurations between the two systems and/or due to changes in sampling strategies when going from the cleaning studies on the ½ in. SS system to the 1 in. PP system.

² Numbers within parentheses refer to the number of sampling sites with growth of another organism than the challenging ones.

The following methods were used to estimate the numbers of challenging organisms:

Sampling method 1

Liquid samples (~ 50 mL) were collected in sterile tubes and then filtered through a 0.45 µm polyvinylidene difluoride membrane filter. They were then incubated on TSA plates at 32°C for 5 d before the number of CFUs were counted and the number of viable microorganisms in the samples were calculated.

Sampling method 2

Surface samples were taken with nylon flocked swabs. The swab was inserted into the tube containing the isotonic swab rinse solution. After dissolution, the whole solution was mixed into molten TSA and allowed to solidify in Petri dishes. The plates were incubated at 32°C for 5 d before the number of CFUs were counted and the number of viable microorganisms in the samples were calculated.

Sampling method 3

Viable count samples of the organisms’ solutions before and after applications, were diluted and incubated on TSA plates at 32°C for 1 d before the number of CFUs were counted and the number of viable microorganisms in the samples were calculated.

Results

We evaluated the concentrations of the challenge organisms four phases during the cleaning study: in the inoculation solution, immediately after infection, prior to cleaning, and after cleaning. The results in Table 5 and 6 show that the two systems had infection and post-infection solutions with approximately 1 × 10⁷ viable organisms/mL of the two organisms. After treatment the concentration of the challenging organisms in the flowthrough solutions was zero. It is common practice to define an efficient cleaning procedure with a log reduction factor of ≥ 6 of CFUs.

Table 5. Number of CFUs of challenging organisms at different phases of the studies

Table 6. Endotoxin samples and TOC samples

¹ Sampling of EU and TOC were not performed in this study

² Below detection level

Conclusions

In conclusion we have shown that:

• The two sizes of ÄKTA process™ chromatography systems with two different contact materials (SS and PP) were efficiently cleaned with 1 M NaOH after contamination with high concentrations of a mix of E. coli and P. aeruginosa in suspension.
• Manifold design is important for efficient cleaning of inlets and outlets including the column valves. On the outlet side, the manifold should be attached in such a way that the common outlet gives rise to a liquid lock. By doing so, the whole valve gets cleaned completely, even if it is closed.
• Extra care also needs to be taken to address cleaning the waste tubes with prolonged contact time and with a larger flowthrough volume of cleaning agent filling the entire tube.
• The cleaning method alone cannot guarantee good hygienic status of a chromatographic process. It should always be applied in conjunction with other well-designed and carefully controlled routines, as well as rigorous control of buffers, water, and other input material.

CY37746-14-Nov-2023-AN