General discusson on use of analgesics

The bullet points on this page are an informal compilation of opinions of contributors to the online VOLE List. As such, they are not peer reviewed and may contain differences of opinion. Those wishing to contact the list may contact Adrian Smith.

We are interested in learning more about situations in which disagreements arise between vets/NVSs/techs/NACWOs/AWERBs etc and scientists (or any permutation thereof) over the use of analgesics in lab animals.

1) We would like to know how common these disagreements / disputes may be; a) in your experience; b) in your current workplace?

2) Are there "hot-spot" fields of study, e.g. inflammation, oncology, immunology, stroke?

• In my experience, in commercial drug development work, the main worries have been with the incidental use of NSAIDs for clinical purposes in long term studies with "immunomodulator" test materials. The point of friction is usually at the interface with the client rather than with internal stakeholders. In my experience again, this has always been managed properly, even at the expense of the client relationship. Obviously there are potential interferences with many other study types, but clearly defined humane end points seem to deal with most of these adequately.
• Yes we've experienced reluctance for researchers to give opioid pain relief to mice in an inflammation model "because it has immune-modulating effects and might alter the experiment" (no evidence provided) and also that "there's nothing in the literature about pain relief in this model". We pointed out that pain also has immune modulating effects, the issues of accuracy in reporting and the need to apply 3Rs but to no avail... The techs are currently scoring mice (with the Mouse Grimace Scale) through this recent experiment and we'll regroup with this information. At the risk of being controversial, it does seem that the onus is usually on the "welfare end" to provide evidence that a refinement won't mess up the experiment, whereas surely the opposite should be the case, that the burden should lie upon the PI to show the robust scientific evidence that applying such a humane refinement to a technique is not possible for whatever reason.
• Regarding pilot studies, my approach has always been that analgesia must be used by default, so if they don’t currently have evidence to show that they can’t study that question with analgesia, they need to set up analgesia vs no analgesia in order to provide an evidence base (unless it’s a pain study). I do find the severity limits unhelpful though. How a surgical study with no analgesia can be moderate just because it’s a rodent makes no sense to me (since when did we have evidence that their experience of pain was less detrimental than that of a monkey).  I have known researchers use the severity limit as evidence that analgesia isn’t needed (in good faith, i.e. they genuinely don’t believe rodents need analgesia and believe that’s supported by the view that the work is moderate).
• Bio-Serv has a range of bacon-flavoured analgesic tablets. See also: Use of Flavored Tablets of Gabapentin and Carprofen to Attenuate Postoperative Hypersensitivity in an Incisional Pain Model in Rats (Rattus norvegicus).
• The fact that pain itself has a myriad of uncontrolled side effects seems to be ignored.  Sometimes on the basis of the need to compare with historical results. With this argument.of course, nothing changes.  Alternatively, we can use that argument as a scientific rationale for trying both methods (with and without analgesia) to see which give the more consistent results.  If the results are different an explanation has to be found? The ethical argument is that of the precautionary principle in that if we err on the side of caution (try to promote better welfare) and are wrong LESS harm is caused than not going for better welfare.  If we assume that animals are not in pain and they are, not doing anythings will result in a greater harm. The North American argument has been going on for decades and revolves who has to prove what. You prove to me that the animal is in pain and I will use analgesics (etc). You prove to me the animal is not in pain and I will forego analgesia. The assumption that an animal is in pain and to provide analgesia is to err on the side of caution as less harm will result and so is the best ethics. (Assuming one agrees that pain is a harm).
• We have struggled with Osteoarthritis (OA)/Rheumatoid Arthritis (RA) and pain models; and of course, always wondering why the cancer model field never discusses the pain management in more detail (particularly orthotropic and/or metastasis models).I have had some frustrating discussions on pain management with groups running surgically induced OA models. Researchers strongly refute any use of analgesia (even opioids, not NSAIDs) as post-operative analgesics interfere with early nociceptive processing and chondrocyte homeostasis to prevent the transition from acute to chronic pain. "Silencing effects" of analgesics in the acute post-surgery/injury phase can prevent the development of nociceptor plasticity, which means the animals will not develop chronic pain. However, my argument was that animals recovering from isoflurane do not have any analgesic effects post-surgery. Surely, such dramatic recovery from surgery must have a considerable systemic effect –across all the endocrine & ANS pathways, giving maybe some non-controlled impact on experimental readouts (and of course all the perspectives on animal care & welfare!). I totally agree that an evidence –based approach in a pilot study to assess the effect of analgesia in study outcomes may be the best way forward. It is also important to assess that for many acute analgesics, there is also a reasonable short wash-out period (with very limited long lasting effects). This is possibly something that should be more systematically implemented by AWERBs. I always find this document from GV-SOLAS very informative and useful.
• There are some useful articles in Comparative Medicine - the paper by Peterson et al. (2017) To Treat or Not to Treat: The Effects of Pain on Experimental Parameters sets out a helpful approach to dealing with the problem. The 6th issue in 2019 has a series of papers, which review the literature in various fields of research. For example: DeMarco, G. J., & Nunamaker, E. A. (2019). A Review of the Effects of Pain and Analgesia on Immune System Function and Inflammation: Relevance for Preclinical Studies. Comparative Medicine, 69(6), 520-534. None of these reviews solves the problems of disagreement of interpretation of the literature, but approaching it as a scientific discussion is usually helpful, especially when the potential effect size of analgesic use is set against the effect size of surgery and pain. It has been quite some time since I have heard the “rodents don’t feel pain” argument, but it does still survive….. A recording of Professor Doug Taylor’s webinar on "Pain Management in Rodent Models of Cancer and Implications for Study Validity”, and the slides he used, are available. See also his paper on the Influence of Pain and Analgesia on Cancer Research Studies.
• I have similar experiences with strain specific differences in autotomy after sciatic nerve injury in mice. Across a couple of establishments they use buprenorphine and carprofen injection at time of surgery followed by buprenorphine in jelly, carprofen in water or injections for up to 3 days - a transgenic strain prone to self-mutilation has been dropped now - and they find this protocol "works". We haven't tried longer duration analgesia but there is still an obvious gait deficit at day 3, so worth looking at again.

The text below is also taken from an informal compilation of opinions of contributors to the online VOLE List. As such, they are not peer reviewed and may contain differences of opinion. Those wishing to contact the list may contact Adrian Smith.

Does a 5mm skin incision result in the same degree of pain and distress in a horse, pig, dog, cat, monkey, rat and mouse? Or does a mouse feel it more?

In terms of nociceptor activity, I guess all are pretty similar as the density of nociceptors won’t be wildly different - will it be perceived as more pain because it’s a greater proportion of the animal’s body in the mouse compared to a horse (or if you just did the same thing in your kitchen – it’s only a scratch...?

I’ve always avoided the "procedure x hurts more - or less - in any particular species". I suspect the real answer lies in the degree of consciousness of the species and how it interprets the damage to its tissues and the threat this poses to its existence. Pain will (I think) be perceived differently in different species but it’s still pain, it’s aversive, and we should try to avoid causing it, and prevent it as effectively as we can.

Another extension to this debate is how long the pain lasts in different species, and how effective descending inhibition is in different species (for example, if you are really small with a high metabolic rate can you afford to lay around feeling sorry for yourself? No. I don’t think that means mice don’t feel pain or don’t feel it for as long or don’t feel it as intensely - we can’t easily address any of those questions, but what we can do is measure the relative pain intensity caused by different surgical procedures (probably) using validated pain assessment techniques based on behavioural changes (eg Mouse Grimace Scale (MGS), burrowing, grooming, wheel running, abnormal behaviours). In practical terms, the MGS is the best we have at present - if its elevated, the mice need more pain relief - well sometimes - I am not sure a very small (but statistically significant) elevation is an indication for more analgesics.