Response Surface Method
Liming Shi, Ph.D., senior staff scientist at Amylin Pharmaceuticals, shared his insights on successful bioassay development. “In order to develop a robust biological assay at low cost, the parameters and the process must be optimized to make it robust against noise factors. Design of experiment studies are used for assay development, especially for complicated multiple factor interaction assays,” he said.
“Response surface method (RSM) is a statistical technique for modeling responses via polynomial equations. The model becomes the basis for 2-D contour maps and 3-D surface plots for the purpose of optimization. Using statistical techniques, the data can be combined to produce a response surface that simultaneously optimizes all endpoints over the studied range for all factors,” Dr. Shi explained.
“During assay development, especially for cell-based potency assays, the major steps are getting the engineered cell line, establishing the biological activity detection system, screening and finding the significant parameters, optimizing the process settings, and eventually, performing the qualification to demonstrate assay accuracy/recovery, specificity, precision, linearity/range, and robustness.
“We screened multiple parameters and locked out three of the most significant parameters, which evoked the biggest changes in assay response. We used second-order polynomial functions to describe the response, which allowed us to determine the interactions between parameters.”
When using RSM for biological drug development, Dr. Shi explained that, “once the critical parameters have been determined, the next step is to obtain the optimum settings of the variables. For example, we are interested in determining the optimum levels of cell number and drug stimulation such that yield is maximized. For every combination there is a yield.”
Jill Crouse-Zeineddini, Ph.D., principal scientist at Amgen, summed up the future challenges of bioassay development during the next decade. “Potency assay development has come a long way. We have seen a progression from the use of animal models to the use of cell culture, and we are now seeing implementation of non-cell-based assays for potency testing.”
New technologies will continue to drive potency method selection, she insisted. “Furthermore, the use of automation, whether applied in a modular approach or in a fully automated capacity, will become more commonplace.” Lastly, she said that “it will be important for potency method development to keep pace with the vast number of therapeutics being developed. New classes of therapeutics, such as those designed to have dual functionality, are one example. Developers will need to ensure potency methods are relevant and fit for purpose to support these types of therapeutics.”