September 1, 2014 (Vol. 34, No. 15)

Lena Lee marketing manager Beckman Coulter Life Sciences

Goal Is to Optimize Growth Conditions for Sustaining High Productivity in Culture

Maintaining cell lines in culture, under optimal growth conditions, is essential for production of many biological drugs. The overall health of the culture is generally assessed by the determination of both cell concentration and percentage of viable cells.

Many cell culture facilities, growing cells in bioreactors or flask cultures, use the standard manual trypan blue vital dye-exclusion cell-viability assay. Viable cells, which have an intact plasma membrane, exclude the trypan blue stain, whereas nonviable cells have a permeable cell membrane and stain dark blue.

The manual method requires that an operator be present, using a hemacytometer and light microscope to identify and count both the stained and unstained cells. The total cell count and percentage viability is then calculated.

This manual technique has significant limitations. It is both time-consuming and labor-intensive. Variation in results can also be significant among users analyzing the same cell sample due to the subjective nature of the determination, or to pipetting or mixing errors.

In addition, only several hundred cells are generally counted, which means statistical confidence is low. This variability may contribute to an unstable manufacturing process.

An automated cell counting process removes the variability between operators and frees them for other tasks. This tutorial will describe the key steps of a validation study conducted by a large biopharmaceutical company changing from a manual counting process to an automated counting process. Beckman Coulter Life Sciences’ (BEC) Vi-CELL XR was the automated cell viability analyzer (automated cell counter) used in the study.

Evaluation of Measurement Accuracy

The first phase of the validation study was to compare the accuracy of the automated counting method with the manual method. This was determined by calculating inter and intra operator % CV (Vi-CELL XR vs. hemacytometer). Accuracy was tested using control beads of known concentration (1 × 106 beads/mL).

Using control beads is a good practice for assessing accuracy as they eliminate sample variability and allow for true comparison of the underlying methods. In the example, bead concentration was determined by two operators, three replicates under the same operating conditions. The final concentrations of control beads determined by manual and automated counting methods were comparable and in the correct range of 0.9 × 106 count/mL–1.1 × 106 count/mL (1 × 106 count/mL ± 10%) as specified in the concentration control assay sheet (Tables 1 & 2).


Table 1

While the coefficient of variation between each count and also between operators did not exceed 5% for both counting methods, it is important to note that the operator-to-operator variability for the automated method was six times less than the manual method (2.81 vs. 0.46). This amount of difference in variability has major implications for global biopharmaceutical manufacturers who have to contend with globally distributed teams, several different shifts, and different levels of training among operators.


Table 2

Optimization of Automated Cell-Type Settings

An automated cell counter, such as the Vi-CELL XR, will typically provide image analysis parameters that can be optimized for a particular cell line. It is important to optimize the analysis settings to achieve accuracy and/or correlation with historical results. As a starting point, it is recommended to use the default cell type and then adjust the cell-type settings to optimize performance for cells of interest.

In this validation example, the following parameters were adjusted for the particular cell line:

  • Minimum and maximum size range. This parameter helped to exclude larger or smaller objects that are not representative of the cell-line population (Figure 1).
  • Minimum circularity for nonviable cells. This parameter helped to exclude debris that may be counted as a nonviable cell.

After the initial cell-type optimization was performed, additional lots of the cell line were analyzed and the cell type was further optimized to ensure repeatability between lots. 


Figure. Plot of size distribution of cells in the Vi-CELL Software

Validation of Automated Cell Counter

Before beginning validation an acceptance criteria was established. The acceptance limits took into consideration the tolerance of all the variables involved in the measurement. Variables such as sample handling, timing of running the sample, instrument variability, and people variability will have an impact on the result.

There is also inherent variance between the automated and manual method. The automated method, such as that performed on the Vi-CELL, will count 10 times more cells than the manual method. Having a higher sample population will increase the accuracy of the result.

Two or three key values were compared to validate the automated method against the manual method. Total cell density (TCD) and viable cell density results were compared. Several lots and sampling intervals were used to complete the validation. Results were compared statistically using the paired t-test.

Other important considerations include: (1) Mix, mix, and mix—thorough mixing of the cell culture or concentration control is critical to obtaining a representative sample. (2) Side by side—analyze samples using the manual and automated methods as close in time together as you can. The viability will change over time and affect the results. (3) Instrument calibration—have the service engineer for the automated counter ensure that the instrument is properly calibrated before beginning the evaluation process.

Conclusion

The Vi-CELL automated cell count and viability analyzer was validated as an accurate and suitable method for cell counting and is comparable to the currently used quality control method. While no significant difference was observed between cell counts obtained using the Vi-CELL or by the manual method, automated counting resulted in significantly less variability.

Moving away from an established technique takes time, patience, and trust in the new method. Collecting sufficient data and applying good statistical analyses will support and justify the changes to the process.

An automated process will eliminate the operator-to-operator count variation inherent with the use of a hemacytometer. An automated cell viability analyzer such as the Vi-CELL counts 10 times more cells than a hemacytometer, thus providing much greater statistical confidence. Images from the automated counter can also be archived for future analyses.

For regulated facilities, and unlike a hemacytometer, Vi-CELL software is designed to comply with 21 CFR Part 11 regulations for electronic records. As the example shows, the major benefit for adapting the new methods is a dramatic reduction in operator-to-operator variability and a corresponding increase in the confidence of results. 

Lena Lee ([email protected]) is a marketing manager with Beckman Coulter Life Sciences. Note that this tutorial is also applicable for the replacement of an obsolete cell viability analyzer.

Previous articleDeveloping New Therapies to Address the Reemergence of Bubonic Plague
Next articleDepth Filtration vs. Centrifugation