Guaranteeing the Quality, Efficacy and Safety of Pharmaceuticals

by Yoshihiko Izumoto, Chief Producer, Right_Seven Company, Ltd. Japan

Each year manufacturers contribute to the improvement of our health and welfare through the ongoing development and stable supply of pharmaceutical products.  And in so doing, they face a variety of challenges at each stage of development. Each of these new products must meet a consistent level of quality, efficacy and safety.

The principles of CMC (chemistry, manufacturing and control) guide the development of processes needed for approval by regulatory authorities, and define the information needed to manufacture pharmaceuticals. All stages of the development life cycle after drug discovery involve CMC.  For example, if variations occur in the compounds between lots, this may impact efficacy or safety. CMC is all-encompassing and may include:

  • Methods for evaluating manufactured compounds and the basis for these methods
  • The management of raw materials
  • Ingredient and manufacturing processes

Although many companies have specific organizations to ensure adherence to CMC principles, the areas that are under the jurisdiction of CMC vary from company to company.  Let’s take a look at the specific statistical analysis methods that can be used to ensure quality, efficacy and safety in non-clinical trial settings.

Guaranteeing Quality

Providing drugs of stable quality is extremely important for guaranteeing the efficacy of the product. For example, problems with storage conditions can lead to crystalline phase transition and toxicity. Unintended crystalline polymorphism and pseudo-crystalline polymorphism may accidentally occur in the manufacturing process.

Quality control in the pharmaceutical industry must be extremely stringent from the production process to packaging to storage. Design of experiments (formulation design, synthetic process research and analytical method development) and stability testing are representative statistical methods used to achieve quality control objectives.

Design of experiments” refers to a method for designing experiments to search for the factors (parameters) that affect quality attributes. This can be done with a few experiments, which means speed and reduced costs can be expected. This is impossible without statistical analysis. Based on the quality attributes and the factors, the program will provide a plan for accurately deducing the relevant parameters with few actual experiments.

Stability testing is a necessary process for meeting the ICH Q1E guideline. Statistical analysis allows the degradation of products and their active ingredients to be evaluated easily and the shelf life to be established accurately. For example, when analyzing the deterioration of a drug in a stability test, attributes such as content, time and lot information are entered into a statistical analysis tool. The appropriate model is then selected, and analysis of covariance is performed to calculate the shelf life of the product. A high-quality statistical analysis tool will present the results in the form of a visual graph.

Using a statistical analysis tool is also effective for creating control charts. The attributes you want to control, such as average values and failure rates, can be statistically visualized which can help distinguish accidental variations from abnormal causes. By setting an accidental variation range and narrowing down control limits, quality can be maintained at a high level.

Confirming Efficacy

Confirmation of a drug's efficacy is done in both non-clinical trials and clinical trials and involves a process that validates that the drug is new and innovative.

A variety of statistical models are created in non-clinical and clinical trials, in which statistical analysis is broadly used. Even in the prior stage of basic research, using statistical analysis in exploratory research of bioactivity allows for the modeling of information regarding the physical properties and bioactivity of candidate substances and various compounds by means of regression analysis, which can then be used to search for compounds that contribute to activity.

In pharmacological testing, statistical analysis is useful not only for dose response testing, but also for comparing efficacy between multiple groups. For example, a test formulation, subject formulation and placebo can be compared simultaneously to verify if the results are significant or not. An advanced statistical analysis tool outputs graphs and statistical results simultaneously as group comparison results, yielding statistical results that are easy for anyone to understand. Exploratory analyses, such as analysis by group and analysis of outliers, are also easy to conduct.

Verifying Safety

When developing a pharmaceutical, the goals are to achieve a product that effectively acts on the affected area with few side effects. It is the responsibility of the pharmaceutical manufacturer to guarantee the safety of its products.

Statistical analysis is widely used in pharmacokinetics testing to measure the absorption, distribution, metabolism and excretion (ADME) of a candidate substance to ensure safety. Statistical analysis tools that have a non-linear regression function apply pharmacokinetic models, compare applied models and estimate drug parameters.

There are also statistical analysis tools that handle various aspects of toxicity testing, such as genotoxicity, carcinogenicity and dependence. Such tools can be used to apply non-linear models, such as probit models and logistic models, to calculate Lethal Dose 50 (LD50) and easily create models for predicting toxicity from physical properties.

JMP Foreword

See the latest issue.