Background
Rheumatoid and osteoarthritis can often be very painful. The pain usually starts in the joints, where tiny nerves can sense that something is not right in their environment, for example because there is inflammation or bone damage. These tiny nerves then send electrical signals to the spinal cord and brain where we experience them as pain. Unfortunately, in people, we cannot really measure how these nerves are behaving, as they are nestled deep underneath our spinal cord vertebrae. Instead, scientists can nowadays use a new imaging technology where they can measure signals in joint nerves in anaesthetized mice. This paper, by George Goodwin and colleagues, set out to do this in mice with either inflammation in their joint (modelling rheumatoid arthritis) or in mice with joint damage due to surgery (modelling osteoarthritis). They also used microchips, like those used for pets, to track the movement of mice to determine whether they might be in pain.
Study findings
The research revealed that it is easy to show that there is a lot more signaling activity in nerves when mice have an inflamed joint – even at rest, when they are not doing anything at all. This is probably the equivalent of people feeling pain when they are just sitting on a sofa. The scientists were also able to show that the mice tend to move around less in their cages when they have inflammatory arthritis, presumably because moving is also painful – just like it would be in people. In contrast, when the mice had a surgically damaged joint, they seemingly moved about as normal, and it was not possible to detect more signaling activity in their nerves. This does not mean that they were not in pain: most likely their pain was simply less severe than that of mice with inflamed joints and therefore not detectable with the current study design.
Recommendations for Future Research
This study showed that joint inflammation can be used as a good model to study nerve dysfunction and pain in mice. On the other hand, damaging a joint surgically can be very useful for modelling joint abnormalities in osteoarthritis, but is probably not severe enough as an intervention to model osteoarthritis pain effectively with conventional sample sizes. This is especially true given that we want to keep the number of animals we need to use for research low. We recommend that, as always, scientists think carefully about whether their sample sizes are in keeping with expected effects in their future work. In the context of pain research, it may be useful to remember that, broadly, n=10 is enough to detect the acute effects of morphine (NNT 2), n=20 the effects of paracetamol (NNT 4) and n=50 the effect of anti-CGRP (NNT 6).