A New Risk Assessment Tool for Regulatory Starting Material Evaluation

• Genentech
• Roche
• Chugai Pharmaceutical

Abstract

Genentech/Roche/Chugai initiated a global working group regarding starting material selection due to the evolving regulatory environment. The group built a risk assessment tool for early development to highlight specific risks as well as overall risk for each potential starting material nomination. The risk assessment tool was based on the current regulatory climate, IQ Consortium papers and recent developments in the understanding of purging power of synthetic processes. The risk assessment tool has been well received by project teams globally at Genentech/Roche/Chugai as a harmonized way to compare projects, identify overall and specific risks for each starting material, develop strategies for risk mitigation and focus development resources. The new risk assessment tool is a valuable tool but does not replace good scientific assessment on a case-by-case basis.

Introduction

According to ICH Q7,¹ a starting material is defined as the point at which the sponsor commits to GMP manufacture of a drug substance. Drug substance processes are generally convergent and each branch of the synthesis begins with one (or more) starting materials. To protect patients from potential unknown impurities introduced prior to the GMP process, proposing very short synthetic routes with complex custom-made starting materials without an appropriate control strategy is not recommended; and, guidance is provided in ICHQ11.² While the ICHQ11 guidance has been in place for several years inconsistent application of ICHQ11 principles still lead to discussions between sponsors and regulators. Health authorities have challenged starting material designation for Phase III and market applications and are now starting to ask significant questions even for earlier phases of development.

Health authorities have presented their current thinking regarding starting materials in two recent regulatory documents to help minimize divergent interpretations of ICHQ11: an ICH Q11 Draft Q&A paper;³ and an EMA reflection paper.⁴ Current thinking focuses on concerns about introduction of impurities into the active substance from non-GMP manufacture, (e.g. from poor cleaning of vessels previously used for other purposes or inadequate control of processes), which would not necessarily be picked up by routine analytical testing. With fewer synthetic steps carried out under GMP, the risk to the quality of the active substance is perceived to be higher. Key considerations for starting material selection focus on the following:

• Proximity and Purging Power
• Stages and Steps
• “Telescoped” Process
• Complexity and Criticality
• Change Control and GMP

An IQ Consortium working group (work described in three papers)^5,6,7 had done extensive benchmarking in 2014 and proposed a risk based approach which considers:

• starting material molecular weight
• number of rings
• number of synthetic steps
• number of stereogenic centers.

Due to the evolving regulatory environment a Genentech/Roche/ Chugai global working group was formed to build an easy to use risk assessment tool for early development to highlight specific risks as well as overall risk for each potential starting material nomination.

Experimental

After careful consideration of the current regulatory climate, the IQ Consortium papers and recent developments in the understanding of purging power of synthetic processes (summarized in a paper by Teasdale et al)8 the Genentech/Roche/Chugai working group distilled starting material risks down to the following most essential criteria:

• Proximity (Propinquity)
• Sufficient Number of Stages
• Purging Power
• Impurity Purging Steps/Isolations
• Complexity
• % Wt of the API
• Number of Chiral Centers
• Number of Substitutions
• Number of Rings
• Impurity Carryover
• Stability

An internal benchmarking exercise was undertaken with careful consideration of recent health authority feedback. The benchmarking results were used to build and test a risk assessment tool. The specific criterial for the tool and the algorithm are contained in Table 1. From the synthetic scheme values for each of the criteria can be counted and the algorithm used to calculate a risk score.

Risk Score=4/A+8/B+(C ×0.06)+D+E+F+G+H

Risk scores less than 8 are considered low, scores from 8 to 10 medium and greater than 10 high.

Table 1. Risk Tool Criterion / Factor and Calculation

The overall risk score is built up from individual risks each contributing to the total. Proximity and purging power were captured as stages and steps respectively where a minimum of 4 bond-making stages and 8 impurity purging steps were considered low risk. Complexity was considered low if the %W/W of the starting material/API (exclusive of leaving groups not incorporated into the structure) was less than 1/3 and each other complexity factor remained low (e.g. chiral centers, substitutions, number of rings). Impurity carryover and stability were considered individually as binary risks (0 is low, 1 contributes to overall risk).

Results

The risk assessment tool was applied to starting materials for processes with recent submissions for which significant health authority queries had been received and to some development projects where the starting materials were identified as high risk during internal project review. These queries correlated well with high overall risk scores as illustrated by the following examples.

Demonstration of the Risk Assessment Tool: To demonstrate the risk assessment tool the synthesis scheme for Vemurafenib shown in Figure 1 (Patent submitted in 2009)⁹ was assessed. The results are shown in Table 2. In this demonstration no starting materials were identified as high risk by the risk assessment tool primarily and the starting materials were globally accepted health authorities.

 Figure 1. Synthetic Scheme for Vemurafenib

Table 2. Demonstration of the Risk Assessment Tool Using Vemurafenib

Example 1: In the example shown in Table 3 Starting Material 3 was identified as high risk by health authorities. The high risk score of Starting Material 3 correlates with the health authority finding and is largely related to individual risks of complexity (high %w/w and number of rings).

Table 3. First Example of Risk Assessment Tool

Example 2: In example shown in Table 4 both starting materials were identified as high risk by health authorities. The high risk score of both starting materials correlates with the health authority finding and is largely related to individual risks of proximity (1 bond forming stage), purging power (3 isolation steps) and complexity (high %w/w, number of stereogenic centers and number of rings).

Table 4. Second Example of Risk Assessment Tool

Example 3: In the example shown in Table 5 Starting Material 3 was identified as high risk during internal project review. The high risk score of Starting Material 3 correlates with the internal finding and is largely related to individual risks of proximity (1 bond making stage), and complexity (high %w/w and number of rings).

Table 5. Third Example of Risk Assessment Tool

Example 4: In the example shown in Table 6 Starting Material 1 and 3 were identified as high risk during internal project review. The high risk scores of Starting Material 1 and 3 correlate with internal finding and are largely related to individual risks of proximity (low number of bond making stages for SM 3), impurity purging (2 steps for SM 3), and complexity (high %w/w for SM 1).

Table 6. Fourth Example of Risk Assessment Tool

Discussion

While final starting material strategies in the Market Application are built up and justified based on the principles outlined in ICH Q11, the tool demonstrated to be valuable for teams to recognize risks early, focus process development resources and develop appropriate risk mitigation strategies. The output of the tool is never considered in isolation. It is merely a starting point to level-set expectations which is particularly helpful in large organizations to ensure consistent assumptions.

The new risk assessment tool is a simple algorithm which can help identify potential health authority concerns. The risk assessment tool has been well received by technical R&D project teams globally at Genentech/Roche/Chugai as a harmonized way to identify overall and specific risks for each starting material early in the development process. To produce high quality API in acceptable quantities risk mitigation strategies focus on ensuring patient safety. Strategies to mitigate risks include: understanding the purging power of the API Schema; defining the supply chain and synthesis by using reliable starting material sources with quality agreements/change control; and, establishing a thorough quality control plan/control strategy.

References

1. ICH Q7A, Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients, Current Step 4 version, dated 10 November 2014
2. ICH Q11, DEVELOPMENT AND MANUFACTURE OF DRUG SUBSTANCES (CHEMICAL ENTITIES AND BIOTECHNOLOGICAL/BIOLOGICAL ENTITIES), Current Step 4, version dated 1 May 2012
3. ICH Q11 Draft Question and Answers document regarding the selection and justification of starting materials (Step 2b Version of 13 October 2016)
4. EMA Reflection Paper on ICH Q11, Reflection paper on the requirements for selection and justification of starting materials for the manufacture of chemical active substances, 16 September 2014, EMA/448443/2014
5. Part 1: A Review and Perspective of the Regulatory Guidance to Support Designation and Justification of API Starting Material, Margaret M. Faul, William F. Kiesman*, Maciej Smulkowski, Steven Pfeiffer, Carl A. Busacca, Magnus C. Eriksson, Fred Hicks, and John D. Orr, Org. Process Res. Dev., 2014, 18 (5), pp 587–593, DOI: 10.1021/op500059k
6. Part 2: Designation and Justification of API Starting Materials: Current Practices across Member Companies of the IQ Consortium, Margaret M. Faul, Carl A. Busacca, Magnus C. Eriksson, Fred Hicks, William F. Kiesman, Maciej Smulkowski, John D. Orr, and Steven Pfeiffer, Org. Process Res. Dev., 2014, 18 (5), pp 594–600, DOI: 10.1021/op5000607
7. Part 3: Designation and Justification of API Starting Materials: Proposed Framework for Alignment from an Industry Perspective, Margaret M. Faul, Mark D. Argentine, Marjorie Egan, Magnus C. Eriksson, Zhihong Ge, Fred Hicks, William F. Kiesman, Ingrid Mergelsberg, John D. Orr, Maciej Smulkowski, and Gerald A. Wächter, Org. Process Res. Dev., 2015, 19 (8), pp 915–924, DOI: 10.1021/acs.oprd.5b00079
8. Risk Assessment of Genotoxic Impurities in New Chemical Entities: Strategies To Demonstrate Control, Andrew Teasdale, David Elder, Sou-Jen Chang, Sophie Wang, Richard Thompson, Nancy Benz, and Ignacio H. Sanchez Flores, dx.doi.org/10.1021/op300268u, Org. Process Res. Dev. 2013, 17, 221−230
9. S. Hildbrand, H.-J. Mair, R. Radinov, Y. Ren, J. A. Wright, WO 2011/015522, 2009 (F. Hoffmann-La Roche)