3 Tips from FDA and EMA on QbD Risk Assessment for Regulatory Submissions

Posted On

Every day, I receive more than 25 emails just on Quality by Design. I’ll share these questions and answers in the future, but today, we cover the questions received by FDA and EMA.

FDA and EMA share 3 tips on QbD Risk Assessment and Design Space for Regulatory Submissions.

Based on FDA and EMA’s guideline (published on December 10 of 2014) this article highlights the questions regarding QbD Risk Assessment in regulatory submissions.

From my interactions with regulatory agencies and QbD scientists, I added “What This Means” and “What This Looks Like” as examples so you know exactly how to implement them in your own QbD projects.

Let’s dive right in.

Questions and answers on QbD Risk Assessment for regulatory submissions

 

1. What types of Risk Assessments (RAs) can be included in a regulatory submission?

It is up to the applicant to determine what Risk Assessments are included in the submission.

This information can be useful to aid the assessor/reviewer for determining how the applicant selected the specific formulation, manufacturing process and controls.

For example, the Agencies have seen risk assessments related to:

• selection of formulation variables, 

• delineation of impact of various process parameters and

• selection of in-process controls.

   

What this means:

You don’t have to use a FMEA format. In fact, FMEA alone, will not let you see the impact of process parameters to CQA’s and QTPP’s because its evaluation process is adhoc and subjective. Simply, it makes too big of a logical jump for the scientific minds.

As ICH Q8 (R2) and FDA suggests, QbD Risk Assessment’s main goal is linking process parameters (CPP) and material attributes (CMA) to patients’ safety and efficacy (QTPP).

That’s why it’s best to link these QbD elements in a sequential and logical order:

1. Link Patients (QTPP) to Drug Product & Substance Quality (CQA)
2. Link Drug Product & Substance Quality (CQA) to Process Parameters (CPP) & Material Attributes (CMA)
3. Calculate the correlations

Companies such as Merck, Medtronic, MedImmune and Dr. Reddy’s use Lean QbD for QbD Risk assessment to follow the 3 simple and effective steps above.

2. What level of detail should be considered for a Risk Assessment related to PROCESS design in a regulatory submission?

The level of detail should be commensurate with the significance of the outcome of the Risk Assessment to the commercial manufacturing process and to the control strategy.

What This Means:

 “high” critical process parameters should get priority when you plan for Design Space studies. Design space study will either help de-risk or mitigate the risk by changing parameter levels or improving the process. You also now have data to support your QbD Risk Assessment, and use as part of the CMC section New Drug Application (NDA) or Biologics License Applications (BLA).

For example, a risk assessment pertaining to identification of critical process parameters in a process or to the establishment of a design space for an important unit operation would normally be considered of high significance

What This Looks Like:

First, you build a process flow diagram (aka process map) and then list parameters underneath each unit operation. Below is a snapshot of the process map and process parameters string-ed together. (In “modify drug profile” of LeanQbD webapp version)

Zoom in and you can see it in further detail.

Second, after assigning correlations, you should be getting the Control Strategy Chart below. I circled and number-ranked the parameters in the order of criticality (1:highest to 3: lower)

Highest criticality process parameter is Load pH. Medium-critical items are bucketed into 2 and 3’s.

Lean QbD kindly lists the process parameters for you so you don’t have to guess.

This list should guide you on which process parameters to include in your Design Space first. Run some Design of Experiments (DOE’s) around these unit operations and process parameters and include the results in your technical reports.

The information that should be provided in such cases could include:

• A statement of which Risk Assessment tool was used; or if a novel Risk Assessment tool was used, a definition of the RA tool along with sample output from the tool.
• A comprehensive qualitative or quantitative summary of the outcome of the Risk Assessment (e.g. RPN (Risk Priority Number),
• Threshold value for RPN
• (and an explanation for why such a threshold was chosen) and final list of all RPNs) supporting and explaining the applicant’s rationale for the selection of variables/parameters that warranted further study and those which were not further studied.

What This Looks Like:

All of these information can be displayed easily with the graphs below.

The first graph and list shows a Pareto plot (ranked by RPN) of CQA’s and the threshold value for RPN’s.

The second graph and list shows a Pareto plot (ranked by RPN) of CPP’s and the threshold value for RPN’s.

Information on the way in which risk assessment activities were used to determine which process parameters and quality attributes are to be considered critical or non-critical.

What This Means:

As you conduct your QbD Risk Assessment, make the participants take notes on why the process parameters are evaluated as “high” (or medium or low) criticality. You can do this by right-clicking a box and just typing in the remarks.

Later, your team can see the summary of rationale here at the comments section. You can just use this table as the summary.

How did I do this?

Lean QbD Risk Assessment has a “Reports” feature where you can choose to include any graphs, tables or comments that your team generated during the process.

Below is just a reiteration of the information above. In other words, make sure you ask the risk assessment participants to comment on the rationale of their evaluations during evaluation.

Both agencies acknowledge that regulators should take the following into consideration when requesting additional details regarding an Risk Assessment in a submission: complexity of the dosage form, the commonality of the risks identified relative to products of the same type (i.e., available prior knowledge), and the amount of commercial scale manufacturing data available at the time of submission (i.e., state of knowledge).

The following question asks for the same information regarding the drug product.

3. What level of detail should be considered for a Risk Assessment related to PRODUCT design in a regulatory submission?

The level of detail should be commensurate with the level of risk presented by the formulation and dosage form, as it is intended to be used by the patient.

The information provided can include how the risk assessment approach was used to optimize and select the formulation, and/or to rank or prioritize formulation variables (e.g., raw material attributes) based on their potential effects on finished product Critical Quality Attributes (CQA). 

What This Means

Material attributes should be part of the QbD Risk Assessment at the level of process parameters.

The factors considered can also include how the drug product could be unintentionally or intentionally misused. For example, a risk assessment pertaining to a solid modified release oral formulation for an opioid would normally include an analysis of the potential for dose dumping. The factors discussed can also include considerations such as physical and chemical stability over shelf life.

What This Looks Like:

You should include misuse scenarios, and stability information in the risk assessment either in the form of parameters or comments. After experimenting, I recommend including them in the comments section in order to keep the CQA – CPP/CMA relationship simple and clean.

You can again use the comments feature in Lean QbD Risk Assessment.

Conclusion:

Let’s recap the 3 tips from FDA and EMA on QbD Risk Assessment and Design Space for Regulatory Submissions.

You can use other risk assessment tools. Go beyond FMEA. (Why QbD Risk Assess is not just risk assessment.)

Link QbD elements in a sequential and logical order:

1. Link Patients (QTPP) to Drug Product & Substance Quality (CQA)
2. Link Drug Product & Substance Quality (CQA) to Process Parameters (CPP) & Material Attributes (CMA)
3. Calculate the correlations

 

Include the information below for regulatory submission:

• Which Risk Assessment tool was used
• A summary of the outcome of the Risk Assessment (e.g. RPN)
• Threshold value for RPN (An explanation for why such a threshold was chosen)
• Final list of all RPNs
• Material attributes at the level of process parameters.
• Rationale for the selection of variables/parameters (in the comments section)
  

Here are some bonus tips:

Tie QbD Risk Assessment into Design Space by prioritizing “critical” process parameters.

Focus on the “high” risk parameters.

Be practical about what is “high” as this means more work and cost

Tie QbD Risk Assessment to your control strategy.

All of this can be facilitated simply on Lean QbD Risk Assessment.

Let’s remind ourselves why we set out on the journey of Pharmaceutical Quality by Design. Lawrence Yu of FDA said it well.

Committed to Helping Keep Patients Safe

At the end of the day, it’s the patients who rely on their medications and who suffer the consequences of poor quality; I find that deeply troubling. Providing patients with substandard or contaminated medications is simply unacceptable. As a regulator, I’m committed to helping to keep patients safe from unnecessary risk by fostering a quality-focused industry and regulations. Patients deserve it.