The Cancer Chronicles Read online

  Chapter 13

  Beware the Echthroi

  On a clear winter day I drove the winding road up to the crest of Sandia Mountain, which looms 10,678 feet over Albuquerque, to spend time basking in the emanations of the Steel Forest, a thick stand of blinking broadcast and microwave antennas that serves as a communications hub for New Mexico and the Southwest. Microwaves are a weak form of electromagnetic radiation that sits in the lower half of the spectrum just above radio broadcast waves and below heat waves and the colors of light. Because of the compact size of the waves—half an inch to a foot across—they are easily focused into beams by dish antennas and used to relay television broadcasts, long-distance telephone calls, and other information from tower to tower and to satellites orbiting in the sky.

  Microwaves are also transmitted and received by cellular telephones and wireless Internet equipment, and Santa Fe had recently become a nexus for people who believe these emissions cause brain tumors and other sickness. They testified at hearings trying to keep wireless out of the public library and city hall. They opposed every new permit for a cell phone mast—even small ones in church steeples that no one could see. They would know they were there because of their emanations. Or so they believed. One Santa Fean sued his next- door neighbor for remotely poisoning him with her iPhone, and a Los Alamos physicist is sometimes seen in public wearing a chain mail hood for protection. Knowing I was skeptical that the small doses of microwaves the public receives could possibly be harmful, he laid down a challenge. Go to the mountain, he said, and spend an hour or two by the antennas. “See if aspirin cures the headache you’ll probably get, and see if you can sleep that night without medication.”

  After I reached the top, I walked around and admired the endless views, browsed through the gift shop, watched a small outdoor wedding ceremony. I sat for long stretches and read a book about mass hysteria and health scares. The cell phone fears seemed like a prime example, a case of metastatic memes—hard, impenetrable kernels of folk science passed from mind to mind with little deliberation. All the while I had in hand a microwave meter I’d purchased to make sure I was getting a dosage of at least 1 milliwatt per square centimeter. That is the threshold set by the Federal Communications Commission for what it considers safe exposure over a thirty-minute interval. (The sun shines upon us at about 100 milliwatts per square centimeter.) Anti-wireless advocates consider the FCC standard far too high, many times greater than what the brain can bear. After two hours, I drove home and woke up the next morning feeling fine. It might be decades of course before I would know if I had seeded a brain tumor.

  If so it would be through a means unknown to science. It is only when you reach the top of the spectrum—the highest frequencies of ultraviolet light, followed by x-rays and gamma rays—that radiation is proven to be carcinogenic. The higher the frequency, the higher the energy—and the smaller and more cutting the waves. Measured in billionths and trillionths of a meter, these are the rays that can zip through cells, tearing electrons from atoms and damaging DNA. Blunter radiation like microwaves can only cause harm by vibrating and heating tissue—that is how a microwave oven boils water and cooks meat. But cell phone and wireless Internet emissions are too feeble for even that. If they were causing cancer, it would have to be in more subtle ways. Electromagnetic fields, microwaves included, can influence the motion of charged particles. And in a living organism streams of charged ions—calcium, potassium, sodium, chloride—flow in and out of cells. So maybe rippling these currents at a particular rhythm somehow elicits malignant behavior, interfering with a crucial cellular pathway by amplifying or squelching it. The oscillations might conceivably suppress the immune system or have epigenetic influences—activating methylation or some other chemical reaction that can affect the output of genes without directly mutating the DNA.

  But all of that is speculation. There is no end to laboratory research investigating how the waves might affect mitosis, the expression of DNA, and other cellular functions or alter the efficiency of the blood-brain barrier or enhance known carcinogens. The results are contradictory and inconclusive. One study showed that glucose metabolism, the normal process by which cells turn sugar into energy, was higher in parts of the brain near where a cell phone antenna was held. Whatever clinical significance that might have is unknown, and the study was quickly contradicted by one finding that glucose activity was suppressed. A few studies—the outliers—have hinted that chronic microwave exposure might raise the risk of tumors in laboratory animals. But the experiments are far outnumbered by those finding no effect.

  A review by the World Health Organization of approximately 25,000 papers uncovered no convincing evidence that microwaves cause cancer. This is reflected in the epidemiology. For the last twenty years, while cell phone use has steadily increased, the annual age-adjusted incidence of malignant brain tumors has remained extremely low—6.1 cases per 100,000 people, or 0.006 percent—and for the last decade has been slightly but steadily decreasing. That has not kept epidemiologists from investigating whether cell phones might still be having a tiny impact. The most ambitious of these efforts, Interphone, gathered information from five thousand brain tumor patients in thirteen countries and compared it with a control group. No relationship was found between the amount of time talking on a cell phone and the incidence of gliomas, meningiomas, and acoustic neuromas—tumors that occur in areas of the head likely to get the most exposure from mobile phones. There actually was a slightly negative correlation: Regular users appeared to have a lower risk of getting brain tumors than people who didn’t use cell phones at all. Rejecting the likelihood of a protective effect, the authors of the final report interpreted the result as a fluke caused by unreliable data, sampling bias, random error—some flaw in methodology. The counterintuitive result also suggested that if there is some effect it is so minuscule that it is swamped by statistical noise.

  Interphone was a retrospective study, relying on memory, like the research that had led scientists to believe for a while that eating fruits and vegetables could drastically reduce the incidence of cancer. There was another reason, however, that has kept the results from being accepted as the final word. The study found no sign of a dose-response relationship, where cancer risk would rise steadily according to the number of hours spent on the phone. But for the 10 percent of people who reported the very highest use, the increased risk of glioma appeared to jump abruptly from 0 to 40 percent. A person’s odds of being diagnosed with the cancer, the most common of all brain tumors, is about 0.0057 percent. A 40 percent increase would make that 0.008 percent. There was a similar but smaller spike for the other tumors. These were also interpreted by the authors as the result of a methodological flaw. Some subjects reported outlandishly long times spent talking on cell phones—as much as twelve hours a day—and that may have skewed the results.

  Maybe people with brain cancer, desperate for an explanation, were overestimating the severity of their cell phone habit. Maybe their memory or their reason was impaired by the tumor. In any case, a later study by the National Cancer Institute looked at gliomas and found no sign that they have been increasing as cell phones have become a ubiquitous part of life. Many epidemiologists were surprised when the International Agency for Research on Cancer decided that there was still enough uncertainty to add microwaves to the long list of possible carcinogens—nowhere near being proven but worth keeping an eye on.

  More answers might come from a prospective study that is almost as ambitious as the EPIC project on nutrition and cancer. COSMOS (the Cohort Study of Mobile Phone Use and Health) is monitoring 250,000 volunteer cell phone users for twenty to thirty years, which surely is time enough to find delayed effects. But even when it is completed decades later not everyone will consider the matter resolved. It still can’t be said flat out that electrical power lines don’t slightly increase the risk of childhood leukemia—a hypothesis that was suggested to widespread disbelief more than three decades ago. The emanations from p
ower lines are many times weaker than even microwaves. Their wavelength is enormous. While the microwaves people have been worrying about are measured in inches and radio broadcast waves in feet—hundreds of feet for the lowest-frequency AM stations—60 hertz power line waves are more than 3,000 miles wide. As they gently roll through neighborhoods they can induce faint currents in whatever they cross, including human cells. No means have been discovered for how that might cause cancer. Over the years most epidemiological studies have turned up no evidence of a danger. But there are always a few anomalies suggesting otherwise.

  Sometimes it feels like we’re chasing our tails, obsessed with finding causes where there may be none. Robert Weinberg once estimated that every second 4 million of the cells in our body are dividing, copying their DNA. With every division there are imperfections. That is the nature of living in a universe dominated by entropy—the natural tendency for order to give way to disorder. If we lived long enough, Weinberg observes, we all would eventually get cancer. That doesn’t mean that we can’t reduce the odds, even if only modestly, that we will get cancer before something else kills us. But genetic errors are inevitable and necessary for us to evolve. Evolution is by random variation and selection, and mutations are the grist for the mill. Along the way cells have evolved the ability to identify and repair broken DNA, but if the mechanism was foolproof evolution would stop. The trade-off is probably a threshold amount of cancer.

  Robert Austin, a biophysicist at Princeton University, goes so far as to argue that cancer is here “on purpose”—that it is a natural response by which organisms deal with stress. When bacteria are deprived of nutrients, they start replicating and mutating like mad—as though trying to evolve new survival skills. If the source of stress is an antibiotic, the winning adaptation might be one that produces an antidote to the poison—or quickens the pace at which the bacteria can flee. Maybe, Austin proposes, the cells in an organism do the same thing. Backed into a corner they try to mutate their way out of trouble, even if it endangers the rest of the body. The best response might not be to fight back with chemotherapy and radiation, increasing the stress, but to somehow maintain the exuberant cells—the tumor—in a quiescent state, something that can be lived with.

  Austin is one of dozens of scientists who have received money from the National Cancer Institute as part of an attempt to break the stalemate in the War on Cancer by importing ideas from outside the usual channels. Franziska Michor, the evolutionary biologist I met in Boston, is also part of the endeavor. In other laboratories, physicists and engineers are bringing their own perspective by studying the mechanical forces involved when cancer cells grow and divide and then migrate through the blood. Instead of speaking the language of biochemistry they use terms like “elasticity,” “translational and angular velocity,” “shear stress”—as if describing boats leaving dock to navigate down a river. Mathematicians are looking at cells at a different level of abstraction—as communications devices—and using the same ideas from information theory that might be applied in the analysis of radio signals or telephone transmission lines. Perhaps cells can be thought of as oscillators like tuning forks. Malignant ones might be identified by their discordant harmonics—their own special ring. If so there might be a way to retune them. A chemist at Rice University is trying to use radio frequency waves to kill cancer cells. First the cells would be injected with gold or carbon nanoparticles. Then the radio waves would cause them to vibrate, producing enough heat to destroy the cell from inside.

  The projects are done in collaboration with oncologists, and a lot of laboratory benchwork is involved. But there are also attempts to step back further and propose whole new theories of cancer. Cell biology is a science of details. There is a grand overarching framework—the modern theory of evolution—but you excel by digging down and mastering thick layers of knowledge about thousands of biochemical gears and the countless ways they can mesh or jam. There are models for how a neuron fires or how DNA is translated into protein. But the closer you look, the more elaborate these mechanisms appear. They are the outgrowth of a long chain of evolutionary accidents, a history that might have spun a different way.

  Theoretical physics rewards those who simplify—glossing over details and exceptions and explaining everything in terms of a few big ideas. The lumpers instead of the splitters. The last time I saw Paul Davies, a theoretical physicist and cosmologist, he was speculating on extraterrestrial biology. More recently he and an astrobiologist, Charles Lineweaver, have been playing with the notion that the human genome carries inside its coils an “ancient genetic toolkit”—long buried routines that primitive cells used to form colonies—early precursors to multicellular life. “If you travelled in a time machine back 1 billion years, you would see many clumps of cells resembling modern cancer tumours,” Davies ventured. As they join forces to become a malignancy, cancer cells are reenacting this legacy software, “marching to the beat of an ancient drum, recapitulating a billion-year-old lifestyle.” When earlier traits long dormant in the genome—hen’s teeth, three-toed horse hoofs, vestigial tails in humans—reemerge in later generations, biologists call them atavisms. Cancer, Davies speculates, is an atavistic phenomenon. Stretching out in another direction, he has suggested that the transition of a healthy cell to a cancerous one may have something to do with quantum physics.

  It was surprising to see Davies brainstorming about cancer. Even more unexpected was Daniel Hillis, a computer scientist and roboticist who is heading a team at the University of Southern California that is assembling detailed computer simulations of cancer—virtual tumors—that might be used to predict which drugs work best. I’d first heard of Hillis when as a student at MIT he helped build a Tinkertoy computer that played tic-tac-toe. He went on to start a company called Thinking Machines. He may be best known as the designer of a giant clock that is being assembled inside a mountain in West Texas, where it is supposed to keep running for ten thousand years, chiming through the millennia even if the human race is gone. At a session organized by the NCI he told an audience of oncologists that the way they were fighting cancer is all wrong—that we need to think of cancer as a process, not a thing. A body does not have cancer, it is “cancering.” Treatment should focus not on attacking a specific type of tumor in a specific organ but on looking at the patient as a complex system. Somewhere in the network of interlocking parts—the immune system, the endocrine system, the nervous system, the circulatory system—something has become unbalanced, and for every patient there may be a different way to set it right. That might have struck some listeners as so much holistic fuzz. But Hillis has been pursuing the idea by building another of his ambitious machines. Instead of the genome he was concentrating on the proteome—all of the proteins that are present in a cell at any one moment. Reading the genome gives you the instructions for making each of the cell’s working parts. Reading the proteome shows which parts are actually being made and in what abundance—a snapshot of the state of the system.

  Scientists have been working for years on mapping the proteome—a formidable task involving laboratory techniques like liquid chromatography and mass spectrometry. In collaboration with David Agus, an oncologist, Hillis started a company that is trying to automate the multiple steps with a robotic assembly line. Given a drop of blood, the machine extracts and sorts the proteins, arranging them in an image that looks like stars in a sky. Each kind of protein appears as an illuminated spot, and its brightness shows how much there is.

  Suppose you have two patients with the same kind of cancer. One responds to a drug and the other does not. Using a device like Hillis’s, you could take their proteomic snapshots and lay one on top of the other and look for something that is different. Even if you don’t know what the pattern means, it might be used as a marker to identify which patients will most likely benefit from the drug. I was reminded of Henrietta Leavitt, the astronomer who had died of stomach cancer but not before discovering Cepheid variables, the pulsating stars cosm
ologists use to measure the universe. She would start with two images of the same patch of sky—glass photographic plates taken a few weeks apart. One would be a negative with the stars glowing in black. She would place that plate on top of the other and hold the glass sandwich to the light. Stars that had grown brighter would appear as larger white spots with smaller black centers. On a plate taken weeks later the white spot would have shrunk to its previous size. No one yet knew the physics that caused the stars to blink, but she was able to correlate their rhythm with their distance from the earth. Sometimes our eyes can glimpse connections that our brains don’t understand.

  As the population ages, cancer is outrunning us. But placed under this stress we are like those madly replicating bacteria Austin talked about—spinning out combinations of memes instead of genes. New ideas. Maybe we really are getting smarter than cancer. Efforts like the Cancer Genome Atlas are continually announcing new discoveries—zeroing in on the genetic details of cancers and sorting them into subtypes, each one potentially vulnerable to a different treatment. As the information multiples, custom therapies will be further customized. Targeted drugs will become ever more precise. When a tumor finds a workaround, other drugs will be ready to go after the new mutation. Pursuing a different strategy, a new class of pharmaceuticals will switch back on apoptosis. Immune system boosters will learn to cleanly distinguish between what is a tumor and what is healthy flesh. A cocktail of these advanced treatments will stop cancer—even advanced metastatic cancer—in its tracks or manage it indefinitely as a chronic disease. Or maybe in ten years we will be reading how these approaches too are falling behind in the cellular arms race and we will be forced to look at cancer in an entirely different way.