It should be recognized that the evaluation of mutagenic risk is an iterative process and needs to be updated in line with any process‐related changes and/or emerging information relating to impurities, and/or degradants, in drug substance or drug product. Other factors such as a change in the trial duration, trial population, specifically in terms of oncology where the treatment is extended from an initial S9 population to a non‐S9 population and/or dose may also require a review of the risk assessment.
3.4 Further Evaluation of Risk – Purge (Spiking) Studies
Alongside the theoretical evaluation of risk described above, there is often the need to examine this experimentally through conducting appropriate purging or spiking experiments. This is most likely required where a moderate to high risk of potential carryover into the API has been defined, i.e. where control Option 4 is not considered appropriate. Spiking refers to the practice of adding in a fixed quantity, or spike, of the material to be tracked, in order to confirm a quantifiable baseline. Purging refers to the extent to which the material in question is removed out of the downstream material or API as a consequence of the processing conditions to which it has been exposed.
3.5 Conclusion
The need to adequately assess the risk posed by MIs, and to limit the level present in API/DP, is clearly established in ICH M7. Within this chapter the authors have looked to define a risk‐based approach to such an assessment, one based on a combination of semiquantitative assessment, allied to analytical results and data from appropriate purging studies. Such an approach should ensure that any actual MI‐related risk is clearly identified and managed.
3.6 Case Studies
3.6.1 Case Study 1 – GW641597X
GW641597X was developed as a PPAR‐alpha agonist for the treatment of dyslipidemia. Described below is an overview of the MI risk assessment, and while this product predated ICH M7 [8], a useful retrospective assessment in alignment with ICH M7 was performed using current best practice to inform the reader of the specific steps required. The development of the process to GW641597X and a discussion of the MI control strategy has been published [21].
Applying mutagenic, or potentially mutagenic, impurity (PMI) controls in accordance with ICH M7 [8] for chronic dosing allows up to 1.5 μg/day for an individual MI specified in the drug substance or up to 5 μg/day for the total quantity of three or more PMIs that may be specified. A maximum dose of 600 μg/day was predicted for GW641597X, and therefore the “commercial” TTC‐based acceptable limit for GW641597X was determined as 2500 μg/g for individual MIs, and for three or more specified MIs, 8333 μg/g would be the maximum total amount.
The first stage for the assessment was to identify potential impurities, this was performed by assessing identified and reasonably predicted drug substance impurities (Figure 3.3), together with assessment of the synthetic process (Figure 3.4) for starting materials, intermediates, and reasonably predicted reaction by‐products from the synthetic process.
Figure 3.3 Identified (I) and reasonably predicted (RP) impurities within GW641597X drug substance.
Figure 3.4 Synthetic process to GW641597X.
The identified drug substance impurities and reasonably predicted impurities (derived from route reagents, intermediates, and potential by‐products/degradants) were all assessed for potential mutagenicity by (Q)SAR screening, which amounted to > 20 separate structures. In accordance with ICH M7, two methodologies were employed, one expert rule‐based software (Derek Nexus v6.0) [22] and one statistics‐based software (Leadscope v2.2.1) [23], and all output results were subject to expert assessment [2].
Out of all the materials assessed, only three structures of potential mutagenic or carcinogenic concern were identified, which were the reagents ethyl bromoisobutyrate 2, hydroxylamine hydrochloride, and the alkyl chloride 8 (Table 3.6).5
Table 3.6 Summary of mutagenicity assessment for synthetic process to GW641597X.
| Compound | Derek | Leadscope | Ames assay | ICH M7 impurity classification |
|---|---|---|---|---|
| Ethyl bromoisobutyrate 2 | Positive | Positive | Positive | 2 |
| Hydroxylamine hydrochloride | Negative | Negative | Negative | 5a |
| Chloromethyloxadiazole 8 | Positive | Positive | Not tested | 3b |
a Hydroxylamine is not mutagenic but is carcinogenic in rats and has a permitted daily exposure (PDE) of 23 μg/day. Hydroxylamine is non‐SAR alerting using the SAR tools for this case study. A published review of available data considers carcinogenesis to be via a non‐thresholded mechanism and as such hydroxylamine can therefore be considered ICH M7 Class 5, i.e. Ref. [19].
b Alkyl chloride 8 is a monofunctional alkyl chloride and should be controlled to a class‐specific limit <15 μg/day.
The remaining compounds were Ames negative, non‐SAR alerting, or the equivocal predictions could be refuted following expert review.
The next stage was to assess the probability for these impurities to be present within the drug substance at a level of concern. This was achieved using a paper‐based purge calculation using the principles established by Teasdale et al. [14, 15], and detailed knowledge of the processing as well as of the material attributes greatly facilitates this process. In each case, once the predicted purge was calculated, this purge was compared to the required purge (based on dose/duration factors) and a purge ratio calculated the magnitude that helps inform likely control strategy for that material [19]. For clarity, each of the potential mutagenic materials is discussed below together with a rationale for its proposed control.
3.6.1.1 Ethyl Bromoisobutyrate 2
Two equivalents of ethyl bromoisobutyrate 2 were charged into the Stage 1a process to achieve complete conversion to product at an acceptable rate of reaction, and as a consequence, there could theoretically be 1 equivalent of 2 remaining on completion of the reaction to prepare 3.6