Around a third of us are infected with a brain parasite called Toxoplasma gondii. This single-celled creature spreads to humans from cats, and has a tendency to change the behaviour of its hosts. Now, a team of scientists led by Frederic Thomas and Kevin Lafferty have found that countries where more people are infected with the parasite have higher rates of brain cancer.
This does not mean that T.gondii causes brain cancer, or even that the two are actually linked. Patricia McKinney, who studies brain cancer and was not involved in the study, says, “This is a technically sound hypothesis-generating paper and, viewed as such, is interesting. It doesn’t tell us much, other than pointing towards some further investigation.”
What is Toxoplasma gondii?
Humans can become infected with T.gondii, either through contact with soil contaminated by cat faeces, or by eating infected meat. These infections are extremely common, and up a third of the world may carry the parasite. However, rates of infection vary greatly from country to country, from just 7 percent in the UK and 11 percent in the USA, to around 67 percent in Brazil.
In the first few weeks of an infection, the parasite causes mild flu-like symptoms, if any. Afterwards, it creates cysts within the blood cells and neurons of its host, and it can persist that way for an entire lifetime without any obvious ill effects. It only causes serious problems for people with weakened immune systems, who can become seriously ill.
But T.gondii can have subtler effects on its hosts. The parasite clearly changes the behaviour of mice and rats, fatally drawing them to the scent of cats. This increases the odds that the parasite will end up in a cat, the only animal in which they can mature and reproduce. There’s some evidence that it does similar mind-bending tricks in humans. All of these associations are disputed, but previous studies have linked the parasite to: subtle personality changes; mental illnesses like schizophrenia, anxiety and bipolar disorder; a higher risk of traffic accidents; and even global differences in human culture. Now, brain cancer joins this controversial list.
What does the new study show?
Thomas and Lafferty compared national figures from 37 countries and found that brain cancers are more common in countries where T.gondii infections are more common. Brain cancer was 1.8 times more common in countries where T.gondii was most common, than in those where it was virtually absent. To take the two extremes, in South Korea, where 4 percent of people are infected, 3.4 people out of every 100,000 develop brain cancer each year. In Brazil, 67 percent of people are infected, and 5.5 out of every 100,000 develop brain cancer each year.
However, these countries also differ in other important ways. Wealthier countries have more sophisticated technology that is better able to detect brain cancer, as well as more accurate cancer registries. To account for that, Thomas and Lafferty adjusted their results for national wealth, along with latitude (T.gondii is more common in the tropics), and cell phone use (which has been spuriously linked to brain cancer). None of these adjustments changed the link between T.gondii infections and brain cancer rates.
What does this mean?
This study is an example of an “ecological study”, one of the weakest designs in the hierarchy of medical research. Looking at differences between entire countries says nothing about the risks faced by the individual people living in those countries. At best, studies like this can suggest questions that are worth addressing in the future; at worst, they are shots in the dark that neither support not disprove a hypothesis.
These particular results simply point to a correlation between T.gondii and brain cancer, and do not imply that the parasite causes the cancer. The opposite may be true – it’s not implausible that brain tumours increase the risk of T.gondii, given that such tumours can weaken the immune system, which might make parasite infections more likely. Alternatively, something else could affect the risk of both brain cancer and T.gondii infection. Once again, weakened immune systems top the list.
McKinney has some other concerns about the study. “What are the toxoplasmosis data based on?” she asks. “Are they a true reflection of prevalence or a reflection of those who have a blood test?” She also says that it “weakens any argument about cause” to look at all brain cancers together, since there are many types that are all clinically distinct and probably have different causes.
Is the link plausible?
McKinney points out that there is no explanation for how T.gondii could trigger a tumour. But the idea that a parasite could increase the risk of brain cancer is not absurd. T.gondii lives for a long time in the brain, and brain cancers take several decades to develop. T.gondii stops cells from killing themselves, and it causes a mild degree of inflammation – two of the hallmarks of cancer.
And many infections are established causes of cancer. The virus HPV causes virtually all cervical cancers along with several other types, while hepatitis B and C cause liver cancer. The bacterium Helicobacter pylori can lead to stomach cancer. Schistosoma, a group of parasitic worms, can increase the risk of bladder cancer. Some scientists have suggested that a common virus called cytomegalovirus might be linked to brain cancer, but as with T.gondii, it’s too early to say if that’s a real association or a false start.
This is not the first time that T.gondii has been linked to brain cancer, although previous studies have been small and inconsistent. An Australian team found that people with meningioma – a type of brain cancer – were twice as likely to test positive for T.gondii as people without it. However, the parasite had no bearing on the odds of glioma, another type of brain cancer. Meanwhile, American researchers found that patients with astrocytoma – the most common glioma – are more likely to have T.gondii infections.
Thomas and Lafferty recognise the limitations of their study. Lafferty says, “These were the best data available and we felt they were sufficient to take the first step. Working with actual brain cancer patients is an obvious next step, but it would be an expensive proposition. It is a lot easier to justify the second, expensive step when you have some evidence for the hypothesis. We are hoping that our results motivate others in the field to do further studies.”
Brain cancer is such a rare disease that gathering enough people for a large and strong enough study is a challenge. But Lafferty thinks that there’s an easy solution at hand. “The National Cancer Institute currently has all the blood samples from brain cancer patients that one would need to do the definitive test at relatively little expense. The research hasn’t been done yet only because nobody has thought to do it.”
Even such samples might not be enough. Cancer-causing infections tend to be extremely common – take HPV and Epstein-Barr virus, for example – and they only cause diseases in a very small proportion of carriers. This means that it can be very hard to establish if an infection actually causes cancer, and we’re unlikely to get a solid coda to the the T.gondii-brain cancer story for a long while.
Update: I contacted Robert Newton, who studies infectious causes of cancer at the University of York, for a comment. He didn’t get back to me in time for publication, but I’m posting his thoughts below:
As you know, Toxoplasma gondii is a parasitic infection of cats. It can and does infect humans, but is usually most damaging in those who have underlying immunesuppression (who are not generally at increased risk of brain cancer, with the exception of lymphomas, which tend to be caused by infection with the virus EBV). It is a plausible cause of a proportion of brain cancers, since it can form cysts in the brain.
The paper by Thomas et al is very weak and, in my view, does not constitute real evidence of an association. Key issues include:
1. There are around 120 types of brain cancer that differ with respect to their epidemiology, prognosis, clinical presentation and probably aetiology. No data on subtype were analysed
2. Data sources – brain cancer is difficult to diagnose and data from GLOBOCAN may not be that reliable. I am unclear where the data for Toxoplasma came from although a paper is referenced
3. Geographical correlations of this sort are notoriously difficult to interpret and no data from patients with the disease (and controls without) are presented.
This is at best, a curiosity – and I would not personally have encouraged publication of the paper
Reference: Thomas, Lafferty, Brodeur, Elguero, Gauthier-Clerc & Misse. 2011. Incidence of adult brain cancers is higher in countries where the protozoan parasite Toxoplasma gondii is common. Biology Letters http://dx.doi.org/10.1098/rsbl.2011.0588
Image from Pulmonary Pathology
Disclosure: Until recently, I worked for Cancer Research UK, looking into aspects of cancer prevention, and I still do some freelance work for them. This post is not influenced by the charity, and does not represent their views.