Bizarre Side Effects

Why would a drug designed to prevent and treat malaria, a parasitic infection of the blood and liver, also affect the central nervous system? The drug in question is mefloquine, marketed as Lariam, and I first learned about its bizarre side effects, including amnesia, psychosis, and hallucinations, while listening to “Contents Unknown,” an episode of the radio program This American Life. The episode intrigued me because it told the story of David MacLean, who was taking mefloquine while on a Fulbright scholarship in India and one day found himself in a train station in a different city from where he lived with no memory of who he was or how he got there. I went looking for primary scientific literature on how mefloquine affects both the malarial parasite and the human brain, and here is what I found. Before exploring the side effects of mefloquine, let’s tackle a more basic question: why do drugs have side effects? The answer lies in how drugs are discovered. In a perfect world, scientists would know so much about a disease that they would be able to design a precisely targeted drug, highly effective against the cause of the disease and benign for the patient. That is a major goal of biomedical research and highly desirable for malaria, which every year afflicts half a billion people (1) and kills one million children in Africa alone (2), but much more of this research is still needed. For many diseases, sadly including malaria, our knowledge is too limited to allow rational design of drugs, and most drugs are discovered either accidentally, like the first anti-malarial agent quinine (3), or by trial and error, which means taking some chemical compound, using it against the disease in an animal model, and seeing if the animal gets better. Therefore, a chemical compound often becomes a drug not because we understand how it works against the disease but rather because we have observed it to work. For a drug discovered in this way, we do not know whether its desired effect is its only effect until we try it out.

Mefloquine is an example of the trial-and-error approach. In the 1960s and 1970s, the Walter Reed Army Institute of Research tested over 300,000 chemical compounds for their ability to kill Plasmodium falciparum and Plasmodium vivax, the two most common malarial parasites, in owl monkeys, the best animal model at the time (4). Mefloquine showed the most promise and went on to clinical trials in humans that are meant, among other things, to test for side effects. In the case of mefloquine, these initial clinical trials showed no serious side effects, but they were conducted in vulnerable populations unable to give full consent, namely male prisoners, military personnel, and residents of developing countries, and may have been biased (5). More recent epidemiological research has shown that side effects in the central nervous system severe enough to require hospitalization occur in 1:10,000 patients taking mefloquine for malaria prevention and in 1:200 to 1:1200 patients using mefloquine for malaria treatment (6). Though mefloquine is still widely available, medical practitioners have an increased appreciation of its side effects, and it is now the drug of last resort, rather than of choice, for the U.S. military (5).

How does mefloquine act to kill Plasmodium parasites? This turned out to be a hard question to answer. Mefloquine is thought to inhibit growth of Plasmodium inside human red blood cells (7). The Plasmodium parasites have a complicated life cycle, proceeding from the saliva of mosquitoes to penetrate inside human red blood cells and liver cells. One possible target of mefloquine may be the food vacuole, a sort of microscopic stomach inside Plasmodium cells where they digest nutrients obtained from the cytoplasm of red blood cells, because the food vacuole changes shape in response to mefloquine treatment (8). I was unable to find any publications that identified the targets of mefloquine more specifically. This may be because laboratory experiments on Plasmodium species are difficult. My classmate Hao Li, a graduate student in the lab of Professor Matt Bogyo at Stanford, works with Plasmodium falciparum and has often described that the procedure of culturing it in blood cells is laborious, yielding precious little material for experimentation. And that’s the easiest parasite species to cultivate. The only way to obtain Plasmodium vivax for experiments is to let it infect and reproduce inside mice or monkeys (1).

Research on how mefloquine may be causing central nervous system side effects in humans was somewhat easier to find, though it is far from conclusive. Mefloquine doesn’t dissolve well in water but sticks quite well to the outside of blood cells and brain cells (5). Post-mortem examinations of both mice and humans exposed to mefloquine have found it to accumulate in the limbic system, a region of the brain responsible for emotions and memory (5) (for a bit more background on the limbic system, see my post “Linguistic Disconnect between the Brain and Emotions”). There it may be blocking connexins (5), proteins that are components of gap junctions. Gap junctions are channels in neuronal membranes that go from the cytoplasm of one neuron into the cytoplasm of the adjacent one. Thus, gap junctions are responsible for synchronizing the activity of neurons (5). Cap junctions are channels that form direct links between the cytoplasms of neighboring cells. Gap junctions are a critical pathway for direct cell-to-cell communication in both neurons (5) and glia (10). These channels pass both electrical current (in the form of charged ions) and intracellular signaling molecules, playing a role in synchronizing neuronal activity, as well as metabolic coupling and chemical signaling (11).

The blocking effect of mefloquine is strong enough and specific enough to have been used to study the signaling behavior of connexins (9). The blockage of connexins may impede the ability of the brain to control emotional impulses and may interfere with memory formation (5). However, the picture may be more complicated because I also found publications claiming blockage of a different class of proteins, called 5-HT3 receptors (7), and an effect on the ability of rat neurons to control their internal concentration of calcium ions, which is essential for neuronal signaling (6). Hopefully, future research will elucidate the relative importance of these various effects.

Current understanding of both the desired effect and the side effects of mefloquine is incomplete and more research is needed. But mefloquine also is a cautionary tale of pitfalls in the process of drug design that may misrepresent a compound with dangerous side effects as a perfectly safe one. It is an illustration of how much more we still need to understand about the human body and its parasites to be able to effectively treat malaria without driving anyone insane.



  1. Carlton JM et al. “Comparative genomics of the neglected human malaria parasite Plasmodium vivax.” Nature. 455:757, 2008. Paywall.
  2. Gardner MJ et al. “Genome sequence of the human malaria parasite Plasmodium falciparum.” Nature. 419:498, 2002. Paywall.
  4. Maugh TH. “Malaria drugs: new ones are available, but little used.” Science. 196:415, 1977. Paywall.
  5. Ritchie EC et al. “Psychiatric Side Effects of Mefloquine: Applications to Forensic Psychiatry.” The Journal of the American Academy of Psychiatry and the Law. 41:224, 2013. Paywall.
  6. Dow GS et al. “The acute neurotoxicity of mefloquine may be mediated through a disruption of calcium homeostasis and ER function in vitro.” Malaria Journal. 2:14, 2003. Open access.
  7. Thompson AJ et al. “The antimalarial drugs quinine, chloroquine and mefloquine are antagonists at 5-HT3 receptors.” British Journal of Pharmacology. 151:666, 2007. Paywall.
  8. Jacobs GH et al. “An ultrastructural study of the effects of mefloquine on malaria parasites.” Am J Trop Med Hyg. 36:9, 1987. Paywall.
  9. Cruikshank SJ et al. “Potent block of Cx36 and Cx50 gap junction channels by mefloquine.” PNAS. 101:12364, 2004.
  10. WIREs Membr Transp Signal 2013, 2:133–142. doi: 10.1002/wmts.87
  11. Bennett and Zukin. Electrical coupling and neuronal synchronization in the mammalian brain. Neuron 2004 Feb 19;41(4):495-511.