If your confidence is not high or very high, what study(ies) would you recommend to improve confidence?

The conclusions drawn from the studies already conducted seem very reasonable, I do not see any further need for additional study(ies) to improve confidence.

Test 1,3-D in glutathione over-expressing and under-expressing knockout/knockin mice.

1. If the dosing was not made each day (i.e., not repeated but rather intermittent dosing) (e.g., BID every 2 days, QID every 4 days), the GHS could be able to regenerate up to its control (basal) limit more rapidly. In this case, the elimination of 1,3-D would be higher under an intermittent dosing scenario compared to under a repeated dosing scenario. This means that the KMD or point of non-linearity may potentially significantly vary with the dosing scenario. Thus, performing intermittent dosing scenario versus repeated dosing would also be of great interest. Intermittent dosing/exposure scenario may represent better the real life conditions in humans. Accordingly, a more robust discussion on the time needed to regenerate the GSH after its depletion versus the dosing scenario is also necessary by the authors (or by the regulatory agency).

2. More discussions should be presented to explain why 30 ppm only has been chosen as the upper limit of the KMD versus potentially 15 ppm in the Figures 5 and 6. However, the reviewer agrees that this may not have a significant impact on the toxicological risk assessment conceding the already high MOE (Table 31).

3. It is generally assumed that only the free chemical moiety in the blood is active in the tissue to induce its toxicity or pharmacological action. If this is true, any significant difference in the binding of 1,3-D in plasma between mouse and human would affect the estimation of the KMD in human. In other words, if the unbound fraction in blood of 1,3-D is higher in human than in plasma, the KMD in mouse should be reduced by the same factor to estimate the KMD in human. It is know that small chemicals such as 1,3-D may bind to blood components such as the hemoglobin and potentially the lipoproteins, albumin and AAH from hydrophobic interactions. It could be of high interest to estimate the unbound fraction in the blood of mouse and human for 1,3-D. However, considering the MOE (margin of exposure) estimated in the whither paper, which are higher than 1000 (Table 31), this binding aspect would not affect the toxicological risk assessment. It could only affect the estimation of KMD and of the safe threshold.