As a mouse vet, my job is to suggest refinements to make cancer studies as easy as possible on the mice. 

Cancer researchers grow tumor cells for study. They buy them from a catalog, or they get them fresh from the hospital, where patients have donated their tumors to science. They grow them in tissue culture; they grow them under the skin on a mouse’s back. And they grow them right inside the mouse brain, when they want a tumor that’s most like the complicated many-armed invasive brain tumors people get. They grow the tumors inside mouse brains when they need to see how cancer drugs work, in an organ notoriously difficult to get good drug penetration. 

The route to the mouse brain is surgical. You make a skin incision on the scalp, drill a hole in the top of the skull, lower a needle into the brain and inject the tumor cells. Close the skull hole and the skin and wait, days to months, for the tumors to develop. 

As a scientist, you have a legal obligation to consult the literature on how these surgeries are done, and consult your vet for recommendations on anesthesia and pain management. Vet school had no class on anesthesia for mouse brain surgery, so I too am looking at that same literature and there we meet my pet peeve: People who do these studies and don’t publish all the details of how they care for their animals.

I went to Pubmed last week with the search terms “glioblastoma” and “mouse.” Plenty of research available there, some with cells in vitro, some with tumors on mouse flank, and some with tumors in mouse brains. Quickly scrutinizing the 15 most recent intracerebral tumor ‘models’ I found ---- 6 of the 15 mentioned that the mouse was actually anesthetized for his or her brain surgery (4 with a ketamine-xylazine mix that’s questionable by current standards). Zero of fifteen reported any kind of painkillers. Ouch!

Trust me, I’m a veterinarian. It is impossible to do mouse brain tumor surgery without anesthesia. Pain concerns aside, there is no way an awake mouse will stay still for this delicate procedure, so I know that at least part of the problem is in the reporting. 9 of 15 scientists failed to describe their mouse anesthesia, but they have to have used anesthesia. Maybe all 15 used analgesics too, but simply failed to report that. Maybe. You CAN accomplish major surgery without post-op painkillers --- you shouldn’t (in most cases), but you can, and the mice will survive, and the pain will likely pass, and they will grow their tumors and you can do your science. 

Here’s what’s wrong with publishing animal studies without pain management details.

1) People will assume you don’t care enough about the animals to care for their pain. Is this REALLY what a scientist wants animal protection watchdog groups saying about their research?  

2) One of the reasons scientists publish their Materials and Methods instead of just their findings is so their peers can more fully evaluate their work and make a decision how much to trust it. Use or non-use of analgesics, the choice of a specific anesthetic --- these are potentially powerful sources of uncontrolled variability in an experiment and really need to be disclosed. Scientists often have reason to fear how an analgesic will affect the progress of a study. Might morphine or NSAIDs alter the course of how the glioma cells grow in the mouse brain? You bet, but so might untreated pain and inflammation. Pain and painkillers, inflammation and anti-inflammatories ALL can affect the outcome of an experiment – to leave them unreported in publishing research studies strikes me as a really bad idea.

3) Finally, a scientist looks to the literature for info on how to do a kind of experiment in a way that will get a publishable result. If all s/he sees are glioblastoma experiments published without analgesia, s/he’ll easily assume that’s how it’s supposed to be done. And s/he may not be receptive to meddlesome vets and IACUCs insisting on more aggressive animal pain management. 

Is it bad science or bad animal welfare to leave animal care details out of scientific articles? That’s a false choice: those two can co-exist.