Crash Course

Neuraminidase crystals

Timing is critical. For the regular flu, initiation of treatment for the antivirals—whether the first swallowed capsule of Tamiflu or the first inhaled breath of Relenza—needs to happen within 36 or 48 hours of initial symptoms, such as a spiking fever, chills, cough, and body aches. The cytokine storm associated with H5N1 is so fast and furious, though, that experts advise initiating treatment as soon as possible, ideally within hours of onset.²⁵⁵⁸ Unlike antibiotics that can kill a bacterial invader, these antivirals don’t kill the virus, they just slow it down.

If the hemagglutinin (H) surface spike can be thought of as the key the virus uses to get into our cells, the neuraminidase (N) spike can be thought of as the virus’s key to get out. The H spike binds onto sialic acid receptors on the surfaces of our cells to gain entry and take over. The subsequent thousands of progeny viruses budding from the infected cell have a problem, though. As they bubble out of the cells, wrapping themselves in our cells’ own membrane, the exiting virus is snared on the dying cells’ remaining Velcro-like loops of sialic acid. The sticky ability of the H spikes to attach and infect new cells is the same property that keeps it anchored to the cell in which it was born. The virus solves this problem with its neuraminidase enzyme spike that lets it shave itself off from the cell’s sialic (also known as neuraminic) acid attachments, cutting the loops that bind it. The N spike that allows the virus to plow through our lung’s protective mucus is the same spike that can dissolve the chemical umbilical cord that keeps it bound to its cell of origin.²⁵⁵⁹

One of the ways our bodies deal with influenza is with targeted attacks against the virus, making antibodies that hone in on the N spikes and cripple their function. In this way, our immune system can limit the spread of the virus by preventing its release from infected cells. Under an electron microscope, scientists can visualize the antibodies in action, clumping all the budding viruses to the cell surface and preventing their escape. The problem is that it takes time to generate these antibodies, especially if the body has never seen the particular virus before. Scientists figured that if they could make a drug that mimics the body’s own strategy, a drug that plugs up the neuraminidase enzyme as effectively as the antibodies, they could stop the virus in its tracks.

To design “plug drugs,”²⁵⁶⁰ researchers first need to solve the three-dimensional structure of the enzyme to construct a molecule that can cork it up. Scientists can use X-ray crystallography to visualize the structure of proteins, but first they have to make a neuraminidase crystal. In one of the great epic stories of drug discovery, neuraminidase spikes from a virus swabbed from the feathered backside of a White-Capped Noddy Tern caught off Australia’s Great Barrier Reef were sent into outer space to crystallize in the weightless microgravity of the Soviet space station Mir.²⁵⁶¹ Other scientists scoffed at the notion that this team had actually succeeded in producing crystals from such a fastidious protein. “All you’ve got are salt crystals—don’t be such an idiot,” the scientists who made the breakthrough remember hearing.²⁵⁶² With the exact structure of the neuraminidase known, it was then possible to design neuraminidase inhibitors Relenza, and, subsequently, Tamiflu, to sheath the tips of the viral N spikes and prevent them from cutting the maternal ties that bind it.²⁵⁶³

This is why the drug has to be taken as soon as possible following infection. These antivirals don’t kill the virus; they just slow the spread, potentially giving the body the critical head start it needs to clear the virus before it can cause too much damage or, in the case of bird flu viruses like H5N1, trigger a cytokine storm. Taken too late, one of the investigators involved in the drugs’ discovery explains, “you may be able to stop the virus, but the immune response would likely kill you.”²⁵⁶⁴

Taken within 28 hours after experimental inoculation with flu virus, these drugs can, within a single day, effect a 100-fold reduction in the viral load—the number of viral mutants the original virus is able to spin off.²⁵⁶⁵ In a laboratory setting, every flu virus with pandemic potential so far tested is susceptible to this class of drugs,²⁵⁶⁶ including the genetically engineered 1918 resurrection²⁵⁶⁷ and H5N1. Clinical data on H5N1 are sparse since most of the human H5N1 infections have occurred in areas that lacked ready access to any antivirals. Tamiflu has also been successfully used on a long-term basis to prevent infection,²⁵⁶⁸ but given the drug’s scarcity, this has been considered wasteful of our limited global supply and may foster viral resistance.²⁵⁶⁹

The way the virus tries to outsmart the antiviral drugs is to weaken the binding power of its hemagglutinin spike so that it does not get as bound up trying to make its exit. If it can’t beat them, it tries to join them by attempting to rid itself of the need for the N spike altogether. The consequence for the virus, though, is reduced binding capacity, compromising its ability to infect new cells.²⁵⁷⁰ Even if the drug no longer works, the resulting resistant virus may be less of a threat, having hobbled itself to undermine the drug. “The indications from the lab data are that the virus is sort of a wimpier virus when it’s resistant to the drug,” says Earl Brown, a prominent influenza virologist at the University of Ottawa. “So if that’s always the case, that’s good. But I think given limited experience with the drug, you can’t be too categorical at this point.”²⁵⁷¹ Having a supply of both Tamiflu and Relenza might be a prudent strategy.

Late treatment is ineffective—the virus has already gotten out of hand.²⁵⁷² This is another reason why personal stockpiles have been considered by some to be so vital. Ideally, a personal supply wouldn’t be necessary, but studies show that only a fraction of patients are able to get to their doctors within 48 hours of initial flu symptoms, and that’s during regular flu season, not a crisis situation where medical facilities will be quickly overwhelmed.²⁵⁷³ With the Katrina debacle in mind, one worries whether the national U.S. stockpile will be efficiently distributed in time. “Let us say the pandemic came first to California,” one analyst asked. “Would supplies in Kansas be shipped out west? Would they ever come back? In any case we can expect time-consuming fights.”²⁵⁷⁴ After just a few days of infection, flu victims tend to have already started killing off the virus—or, the virus has already started killing off them.

The standard adult treatment course for Tamiflu is one capsule taken twice a day for five days, once in the morning and once in the evening without missing a single dose. Tamiflu tends to be well tolerated with minimal side effects (most often mild to moderate nausea in approximately one in ten people who take it).²⁵⁷⁵ Relenza is inhaled using the Rotadisk® inhaler device and instructions that come with each prescription—typically two inhalations twice a day for five days. In rare cases, Relenza may trigger asthma-like attacks and is therefore not recommended for people with underlying bronchial disease.

Recent data from experiments on mice suggest that the customary five-day course of treatment may not be sufficient, however. Webster’s group at St. Jude’s Hospital infected 20 mice with H5N1, which killed all of them in short order. The researchers took another 20 mice and put half on human-equivalent doses of Tamiflu for five days and the other half on Tamiflu for eight days. On the standard human-dosing regimen of five days, only half survived. Evidently the Tamiflu slowed H5N1 enough for the animals’ immune systems to beat back the virus, but once the drug was stopped at day five, the remaining virus resurged with a vengeance to kill off half of the survivors. Extending the treatment for three additional days allowed eight out of ten of the mice to successfully defeat the virus and survive.²⁵⁷⁶ While normally “unwise to extrapolate information concerning drugs from one species to another,”²⁵⁷⁷ lacking adequate human data, this experiment suggests that it may be prudent to take Tamiflu for longer than the usual five-day course due to the exceptional ferocity of this virus.

The experiment also suggests that even if one were to start one’s Tamiflu course early and extend it with probenicid or additional packs of pills, there remains a significant risk of dying. And the experiment was done under idealized conditions. The animals were started on Tamiflu not just early, but really early—hours before they were even infected. Given that under controlled laboratory conditions a significant percentage of the animals died—even after a prolonged eight-day course—the best course of action is not to get infected in the first place.