Response to Roustan et al. ‘Evaluating methods of mouse euthanasia on the oocyte quality: cervical dislocation versus isoflurane inhalation’: animal welfare concerns regarding the aversiveness of isoflurane and its inability to cause rapid death

In a recent article in this journal, Roustan et al. ¹ examined the effect of euthanasia technique upon the quality of oocytes harvested from mice and concluded that cervical dislocation was preferable to an overdose of the volatile anaesthetic isoflurane in terms of oocyte quality. It is gratifying to see researchers examining their choice of euthanasia technique. However, the premise of this study appears to have been that, in terms of animal welfare, an overdose of isoflurane would have been preferable to cervical dislocation if it afforded an equal quality of oocyte; I suggest this presumption is not supported by the scientific evidence. In contrast to the assumption of Roustan et al., there is evidence that isoflurane may be aversive to rodents, whereas there is no current evidence that cervical dislocation, when performed correctly, causes suffering.

Several studies have demonstrated that rodents are aversive to isoflurane. Mice as well as rats will generally forego a highly desired food reward to avoid isoflurane exposure. ^2,3 When an aversive stimulus is inescapable and an animal cannot adapt to the stressor (as is the case during induction of isoflurane anaesthesia) it is likely that some degree of distress will result. ⁴ An ideal euthanasia agent would be one which would cause death without pain or distress; ^5–7 therefore, isoflurane does not appear to fulfil the criteria for such an ideal agent. Nonetheless, it has been demonstrated that isoflurane is less aversive than other gaseous or volatile euthanasia agents including carbon dioxide, ² the most commonly used agent, and for this reason its use has been suggested as a more humane inhalation technique. ^2,5 Isoflurane may be the ‘least worst’ option when an inhalation method of euthanasia must be used, but the presumption of Roustan and colleagues that euthanasia with isoflurane is more humane than a physical method such as cervical dislocation is questionable. To the best of my knowledge it has not been shown that euthanasia with isoflurane is more humane than physical methods. Cervical dislocation, when performed competently, renders mice insensible very rapidly ⁸ and therefore is likely to be relatively humane. Most animal welfare concerns regarding cervical dislocation centre around the technical competence of the operator ^7,9 or on issues of operator fatigue when multiple animals must be killed ⁵ rather than the humaneness of the technique per se. Cervical dislocation is also not favoured by some because of the unpleasantness of the technique for the persons performing it. ¹⁰ While such concerns are valid, in pure animal welfare terms it is likely that physical methods of euthanasia may offer a higher standard of welfare than inhalation techniques.

On a technical note, I was surprised by Roustan et al.'s statement that ‘…inhalation of a lethal dose of isoflurane (concentration of 5%) in a closed chamber…causes death in less than 60 s…’. ¹ In my experience, respiratory arrest does not occur until many minutes following loss of consciousness when rats or mice are exposed to 5% isoflurane. Cardiac arrest, which is often used to confirm death, ¹¹ likely occurs after respiratory arrest. Recent guidance from the Canadian Council on Animal Care recognizes that isoflurane is very slow to cause death and therefore recommends that animals are not killed with isoflurane, but only rendered unconscious before they are killed by another method such as exposure to CO₂. ⁵

To quantify the time taken to kill mice using isoflurane, I measured the time until cessation of respiration of four C57BL/6 mice (Charles River, Ramsgate, UK) scheduled for routine euthanasia (2 males and 2 females, aged 16 weeks). The study was conducted in accordance with the Animals (Scientific Procedures) Act (1986), under project licence 60/4126. The mice were killed in a commercially available 7 L induction chamber (VetTech Solutions Ltd, Congleton, UK) prefilled with 5% isoflurane delivered in oxygen from an Isotec 4 vaporizer (Datex Ohmeda, Steeton, UK) and maintained by a constant flow of 5% isoflurane. The isoflurane concentration in the chamber did not fall below 4.8% throughout the process as confirmed by continuous sampling of the chamber gas concentration using a calibrated agent analyser (Kolormon, Charter Kontron, Milton Keynes, UK). The mean time to the last visually detectable respiratory movement was 1605 ± 149 s (mean ± SD). Given that the shortest time to cessation of respiration was over 24 min, it is difficult to reconcile these data with the results of Roustan et al., even allowing for the use of a different age and strain of mouse. It may be that the ‘closed’ chamber employed by Roustan et al. ¹ caused death more rapidly, in which case it would be useful for the authors to clarify how their chamber differs from the standard vaporizer and induction chamber apparatus as their system may be better suited for euthanasia procedures. In the meantime, I believe that it is important to clarify that most commercially available anaesthetic chamber and vaporizer combinations are not capable of killing animals within 60 s. This issue is important because it is crucial that it is not assumed that animals are dead without confirmation of death, ^7,11 as they may survive for many minutes under isoflurane anaesthesia. Such animals likely remain capable of regaining consciousness if removed from the chamber, exposing them to the possibility of severe suffering if invasive procedures are then performed upon them. Many guidelines require confirmation of the cessation of circulation, cervical dislocation or destruction of the brain to ensure death following euthanasia with inhaled agents to prevent the possibility of animals regaining consciousness. ^6,7,10

In summary, there is no evidence that euthanasia with isoflurane is more humane than cervical dislocation. If euthanasia is performed solely with isoflurane, death is likely to be prolonged and should be confirmed before procedures are performed on the animal or the carcass disposed of.