used in Stage 2 – anticipate 1 equivalent would react as reaction is high yielding. Hydroxylamine anticipated to be highly soluble in the isolation solvent (aqueous ethanol). Volatility during isolation and drying expected but not included within the prediction.
3.6.1.3 Alkyl Chloride 8
The alkyl chloride 8 is a monofunctional alkyl chloride8 and as such can be controlled to a class‐specific limit of <15 μg/day; refer Note 5 of ICH M7 [8]. This control limit approximates to 2.5% w/w, but as per ICH M7 section 7.2, the allowable intake of an MI should be capped at 0.5%; therefore, the impurity purge calculations for this material were set against this lower limit. This reagent is charged into the Stage 4 process at 1.15 equivalents, meaning that up to 0.15 equivalents could remain at the end of the reaction assuming a high yielding process, which would need to be reduced to <5000 μg/g of GW641597X, which gives a required purge factor of 30 (150 000/5000 = 30).
The alkyl chloride 8 reacts with the phenolate derived from 4 to provide the ethyl ester 9. An excess of reagent is used (1.15 eq.) to ensure reaction completion and maximize yield of product. It is assumed that 1 equivalent is required to form the high yield of ethyl ester 9, and therefore the residual 0.15 molar excess should be assessed.9 Reactivity for alkyl chloride 8 could be anticipated within Stage 4a through side reactions with the ethanol solvent and base to form the corresponding ethyl ether 10, but this was not scored. A moderate reactivity was predicted during Stage 4b where hydrolysis to alcohol 11 or ethanolysis to 10 could be reasonably predicted. Solubility was anticipated during product isolations in Stages 4b and 5. The resulting purge ratio (Table 3.9) was insufficient to justify an ICH M7 Option 4 control strategy as the purge ratio is less than 1000 and, without further data to support an Option 4 control strategy, an alternative ICH M7 control option would be recommended [19].
Table 3.9 Purge predictions for alkyl chloride 8.
| Stage | Reactivity | Solubility | Volatility | Total | Rationale |
|---|---|---|---|---|---|
| 4a | 1 | 1 | 1 | 1 | 1 Equivalent of 8 reacts with the phenolate from 4, and a high yield of the ethyl ester product 9 is formed. This has been reflected in starting equivalents. Anticipated reactivity to form ethyl ether 10 and hydrolysis product 11 have not been included in the prediction. |
| 4b | 10 | 10 | 1 | 100 | Excess alkyl chloride 8 anticipated to react to form ether 10, hydrolysis product 11, and potential homo coupled products. Anticipated to be soluble within the aqueous ethanol isolation solvents. |
| 5 | 1 | 10 | 1 | 10 | Solubility expected within the crystallization/isolation solvent (denatured ethanol). |
| Predicted purge factor (Required purge) | 1000 (30)Purge ratio = 33 (would inform ICH M7 Option 1, 2, or 3 control without additional data, e.g. experimentally measure PMI purging) | ||||
3.6.1.4 Additional Evidence for the Purging of Ethyl Bromoisobutyrate and Alkyl Chloride 8
In both the case of ethyl bromoisobutyrate 2 and the alkyl chloride 8, purge factors alone were insufficient to allow an Option 4 approach to be defined, and hence it was decided to generate further supporting data to justify the use of Option 4.
cssStyle="font-weight:bold;font-style:italic;" 3.6.1.4.1 Including “Measured” Purge into the Purge Rationale for Ethyl Bromoisobutyrate 2
Stage 1a and 1b products are generally telescoped through into the Stage 4 process; however, to understand likely levels of the alkyl bromide 2 that might be present, a sample was removed from Stage 1a and the reaction solvent removed under vacuum.10 Subsequent evaluation by 1H NMR (Figure 3.5) confirmed that the alkyl bromide 2 was present at <5% molar ratio (m/m) within this product, rather than the initial conservative estimate of 100% molar ratio based on the 1 molar equivalents excess used for this reagent. This observation aligns with the formation of polymer and allows a more realistic purge prediction to be estimated based on “point of last measurement.” Substituting a residual level of ethyl bromoisobutyrate of 5% (50 000 μg/g) into the Stage 4 process, we can now estimate that the predicted purge factor for Stages 4a, 4b, and 5 is 1 × 105 with a required purge factor of 20 (50 000/2500 = 20), which gives a purge ratio of 5000 and fully justifies an ICH M7 [8] Option 4 control. While not necessarily required, checking the 1H NMR of the crude GW641597X (Stage 4b product) confirmed that alkyl bromide 2 is not detected, helping to further validate the predicted purge factor.
Figure 3.5 1H NMR of Stage 1a product 3.
cssStyle="font-weight:bold;font-style:italic;" 3.6.1.4.2 Inclusion of “Measured” Purge