The Cancer Chronicles Read online

  As I learned about metastasis, I thought about the years before the cancer when Nancy and I worked so hard to turn a desiccated, junk-strewn weed patch—our backyard—into a xeriscape garden. Not a zeroscape—those gravel and cactus afterthoughts one sees in Phoenix or Las Vegas—but something akin to a dry highland meadow. We started with one small patch, clearing it of brush and scattering a packet of Beauty Beyond Belief wildflower seeds, a mix recommended for northern New Mexico. There were seeds for Colorado aster, goldfields, arroyo lupine, desert lupine, desert marigold, California poppy, alyssum, baby blue eyes, baby’s breath, bachelor button, black-eyed Susan, candytuft, catchfly, columbine, purple coneflower, yellow coneflower, coreopsis, cosmos, African daisy, Shasta daisy, blue flax, scarlet flax, mountain garland, gaillardia, larkspur, perennial lupine, Mexican hat, Rocky Mountain penstemon, corn poppy, sweet william pinks, and wallflower. We raked them into the dirt and let nature take its course.

  When the rains came it was clear that all we were going to get was blue flax, coneflower, and Mexican hat. They overflowed the garden and over the years found niches throughout our irregularly shaped quarter acre of land. The yellow coneflower and Mexican hat, both members of the genus Ratibida, mated to form hybrids that still appear each season. On Saturday mornings we would come home from the nursery with flats of new wildflowers to try. For all our efforts some would die not long after planting, but those that survived would set seed in the fall. The winds would come, then the rain, and we would find Rocky Mountain penstemon and red pineleaf penstemon in surprising new places. They would grow there and thrive in a way they never did when we were the ones to choose the locale.

  Some wildflowers native to the foothills where we lived flourished along the trails. Yet they were nearly impossible to cultivate: Hymenoxys argentea with its silvery leaves and yellow flowers, Phlox nana (locally called Santa Fe phlox), which bloomed little violet stars. A local nursery managed to grow only a few of the plants and there was a waiting list each spring. It took years of trial and error until the phlox finally found a spot, shaded by a pine tree, where it deigned to grow. Nancy had majored in biology and would show me how a leaf of a wildflower began to change at the tip, gradually in shape and color, until one day there was a blossom. It had never occurred to me that the same green cells that formed the leaf were differentiating into colorful petals—genes switched on and off, signaled by sunlight, temperature, moisture, whatever told the plant that it was time to bloom. Differentiation and development could occur at astonishing speeds.

  What adapted far more readily were the weeds. After our first summer rain in Santa Fe, a bluish green carpet that we welcomed as some unidentified native ground cover turned out to be seedlings of kochia, a member of the goosefoot family that originated in the harsh climate of the Russian steppes. For all its aridity, New Mexico must seem to this immigrant like a tropical paradise. The tiny plants rapidly shot up to form ugly, spindly weeds.

  Another hated intruder from Eurasia was western salsify, and we thought at first that it was no worse than a larger version of the American dandelion. We quickly learned better. One morning we were showing our fledgling gardens to our neighbor Vivian when she spotted one of these weeds, now more than a foot tall, with a podlike bud protruding outward that was about to open into a flower. Vivian shrieked melodramatically and pulled it up by its roots, advising us to kill every one we found. As we soon learned, the pretty yellow petals would turn, seemingly overnight, into a cloud of feathery white seeds, each so viable that western salsify would quickly spread throughout the yard outcompeting almost everything. It spread so viciously that we imagined it, in the dark of night, expectorating its deadly spores in one explosive burst. We thought of the pods in Invasion of the Body Snatchers, landing from some distant star to take over the earth. We nicknamed the weed “space plant,” and I learned to recognize and destroy the seedlings when they were barely half an inch high.

  That was a few years before Vivian died of ovarian cancer. The spreading of weeds became linked in my mind with metastasis. But maybe that was the wrong metaphor. Cancer, as Paget realized so long ago, is more discriminating in the way it propagates. Honed for life in a specific tissue, a metastasizing cancer cell had more in common with those delicate wildflowers—until it found its roost. Then it was more like the pods.

  Chapter 5

  Information Sickness

  The first hint that cancer is a disease of information came in a laboratory at the University of Texas, where in the late 1920s Hermann J. Muller was experimenting with fruit flies. He was working in a long tradition that had begun with Mendel, who discovered in his monastery garden that certain traits like flower color are passed down among generations of pea plants according to predictable patterns. Purpleness is a dominant factor and whiteness is a recessive one. If a pea plant inherits the purple factor from each parent, its flowers will be purple. The same rule holds true if both inherited factors are white. But if one is white and the other is purple, they do not blend to make lavender. Purple trumps white so that is the color that appears in the progeny. The modern way of saying this is that there is a gene for flower color—a microscopic kernel of hereditary information—and that it comes in two forms. With fruit flies, which breed so rapidly, the shuffling of these tokens unfolds in fast-forward. Eyes red or white, bristles straight or forked—these genetic traits, as discrete as the ones and zeroes of binary code, can be followed and plotted as they travel down the family line.

  As a student, Muller had studied how the Mendelian process sometimes spits out a wild card. After many generations, purebred red-eyed flies would spontaneously produce a mutant with white eyes. Other kinds of mutations would also appear. This was long before DNA was identified as the stuff of genes, the helically shaped molecule that carries genetic information in a four-symbol alphabet—the nucleotides abbreviated G, C, A, and T. If a letter is changed, the meaning can be corrupted. The signal becomes noise or is silenced altogether. That kind of clarity would come decades later with the discoveries of Oswald Avery in 1944, Alfred Hershey and Martha Chase in 1952, and a year later when James Watson and Francis Crick cobbled together from cardboard, sheet metal, and wire their model of the double helix. For now Muller’s contribution was to show that whatever genes were made of and however they worked, you didn’t have to wait for mutations to occur. They could be produced at will by exposing the flies to x-rays.

  Most often the mutations sterilized the flies or killed them. That, he speculated, might explain why the rays were so effective at destroying rapidly dividing cancer cells—a therapy that had come into use almost as soon as x-rays were first produced in the laboratory of Wilhelm Röntgen in 1895. With each cellular division the genes had to be copied. The energy from a penetrating x-ray could damage the microscopic structure, inducing a lethal mutation and removing the cell from the game. Far more telling was that Muller’s x-rays could also create living mutants: albino fruit flies or fruit flies with forked bristles or shrunken wings. This ability to alter genetic material, he suggested, might explain a paradox: why the cancer-killing rays could also produce cancers, transforming normal cells into malignant ones. Cancer, this seemingly amorphous disease, this sprawl of hyped-up cells, might be the result of precise genetic mutations.

  The clues had been lingering barely visible since the early 1900s when a German biologist, Theodor Boveri, wondered why cancer cells had strange-looking chromosomes. Maybe, he speculated, they were damaged in a way that knocked out “factors,” whatever they might be, that would normally rein in growth, allowing the cells to “multiply without restraint.”

  Reverting to a more primitive state, a cancer cell abandoned its communal obligation to replicate only when “the needs of the whole organism require it.” What had been a responsible member of an organization became like a single-minded paramecium whose only aim, Boveri wrote, is to egotistically propagate itself. Half a century before DNA was decoded, he even ventured that a cancer cell goes native bec
ause “chemical and physical interventions” damage some of its internal workings without killing the cell outright. He was writing in 1914. Five years later, inspired by Boveri, the geneticists Thomas Hunt Morgan and Calvin B. Bridges found it “conceivable at least that mammalian cancer may be due to recurrent somatic mutation of some gene.” Another scientist spoke of cancer as “a new kind of cell” in which “an ever recurring process of mutation is taking place, with a tendency, however, to deviate more and more from the normal type.” It is as impressive as it is frustrating how close they came to the mark.

  Evidence had also been accumulating that radioactivity, like x-rays, was capable of causing mutations. Since ancient Rome, uranium had been mined and extracted from a rock called pitchblende for use as a yellow pigment in making glass and ceramics. No one knew of its more exotic qualities until 1896 when Henri Becquerel accidentally discovered that uranium salts wrapped in opaque paper or shielded with aluminum would fog photographic plates. He thought at first that the crystals were absorbing sunlight and then reemitting these piercing rays. What a chill he must have felt when he realized that the uranium was not sucking up the energy but producing it—this invisible and piercing light.

  The situation only grew stranger when Marie Curie noticed that pitchblende retained its power even after the uranium was removed—in fact, the leftover ore was far more radioactive than the purified uranium itself. There must be something else in the rock that was even hotter. She and her husband, Pierre, isolated and named a new radioactive element, polonium (after Poland, her native land), only to find that the rock remaining was still extremely radioactive. Something still hidden inside was shooting out these incredible rays.

  “Pierre, what if there is a kind of matter in the world that we’ve never even dreamed of.…What if there exists a matter that is not inert but alive?” That’s Greer Garson, playing Curie in the 1943 movie Madame Curie, in a scene as erudite as it is melodramatic. In a drafty shed at the University of Paris she sifts through piles of pitchblende and extracts the tiniest speck of what she names radium. In the best part of the movie she and Pierre come upon the shed at night and find it shining with an eerie glow. The real story, uncompressed and undramatized, is just as moving. Here is how Curie described it in her own writings: “One of our joys was to go into our workroom at night; we then perceived on all sides the feebly luminous silhouettes of the bottles or capsules containing our products. It was really a lovely sight and one always new to us. The glowing tubes looked like faint, fairy lights.” What the Curies were witnessing were contrails of light produced by charged particles shooting through the air, an optical analog of a sonic boom.

  Radium also glows when its rays strike a phosphorescent chemical like zinc sulfide, and before long the two substances were mixed to produce glow-in-the-dark watch dials. Painting the numbers was a painstaking task—the hook at the top of a 2 thinning just so to produce the narrow downstroke, thickening again to form the base line. The numerals 3, 6, and 8 were equally demanding. To clean the tips of the brushes and keep them pointed, workers were trained to wet and shape them with their lips and tongues. Assuming that the paint was harmless, some of the dial painters—they became known in news reports as the Radium Girls—used it to decorate their teeth, fingernails, and eyebrows. It must have been great for Halloween.

  Mistaken by the body for calcium, radium became incorporated into their bones, where it sat firing off high-speed electrons, alpha particles, and gamma rays, killing cells or transforming them and eventually giving some of the women cancer. Here was the paradox again: Curie herself had been promoting radium, like x-rays, as a therapy for shrinking cancerous tumors. But here it was producing tumors from healthy cells. In 1927 when the Radium Girls were making headlines, Muller’s paper appeared, speculating that the mutagenic power of x-rays might be responsible for their ability to cause cancer. If so, then the same was probably true for radium’s fairy light.

  Long before invisible rays became a suspect, doctors were seeing clues that cancer could also be caused by more tangible stuff. In 1775 a London surgeon realized that “soot warts,” sores appearing on the scrotums of chimney sweeps, were not venereal disease but a malignancy—apparently caused when skin came into contact with the black tars and dust left by burnt coal. The same cancer was later found in workers who manufactured paraffin and other coal tar distillates, and by the early twentieth century scientists were producing carcinomas by repeatedly applying coal tar to rabbits’ ears. Coal tar was found to consist of a witch’s brew of carbon-based compounds—benzene, aniline, naphthalene, phenols—and during the next few decades scientists discovered that many of them produced tumors in laboratory animals. It would have been unethical for them to expose human subjects to the carcinogens to see if they caused cancer. They didn’t have to. With the growth of the cigarette industry, people were performing the experiment on themselves.

  By the time the century was half through we knew that radiation caused both mutations and cancer. We knew that a host of different chemicals also caused cancer, and many of these were soon shown to be mutagens. They altered a cell’s genetic software by changing snippets of the DNA code. In the early 1970s Bruce Ames (the scientist best known for showing that ordinary fruits and vegetables contain carcinogens) came up with a striking demonstration. Instead of fruit flies, he worked with salmonella bacteria—strains that had lost the recipe for making histidine, an amino acid they needed in order to reproduce. If placed in a dish of nutrients with a dash of this vital ingredient, the bacteria would grow, but only until they had depleted the supply. Then the whole colony would die. Ames discovered that if carcinogens were added to the mix, some of the salmonella would keep on living, expanding and overtaking the dish. The chemicals were presumably producing mutations at random. But each bacterium’s genome carried so little information, and there were so many of the microbes—billions of them—that the mutations would include ones that happened to restore the ability to synthesize histidine.

  The procedure came to be called the Ames test—a fast and dirty way to see if a chemical might be mutagenic. In instance after instance, chemicals that passed the Ames test also produced tumors in laboratory animals. The case almost seemed clinched. What causes cancer, whether chemical or energetic, does so by altering genetic information. The pieces of a theory were falling into place, except for a stubborn exception—at least some cancers appeared to be caused neither by chemicals nor penetrating rays but by viruses.

  In retrospect that is not so surprising. Existing on the boundary between chemistry and life, viruses are packets of information—streamlined sequences of DNA or RNA wrapped in a protective sheath. They are wandering genomes so simple that some consist of only three genes. Like the handmade Internet viruses they later inspired, they infiltrate their hosts (the biological computers called cells) and commandeer the internal machinery. There the invader’s genes are dutifully duplicated and repackaged again and again, the viral copies spreading to other cells where they robotically carry out the same routine—life itself stripped of its capacity to do anything except reproduce.

  A few viruses operate in an even more convoluted way. They copy and splice their genes directly into a cell’s chromosomes. This infiltrating algorithm orders the host itself to replicate at an accelerated pace. It becomes a cancer cell. The earliest example was reported in 1910 by Peyton Rous, a scientist at the Rockefeller Institute for Medical Research who was studying chicken tumors. He began by extracting fluid from an irregularly shaped glob growing in the breast of a Plymouth Rock hen and then injecting it into another bird. Thirty-five days later the first chicken had died from the cancer, a sarcoma, and the second chicken had developed a tumor of the same kind. Material taken from the tumor could, in turn, be used to spread the cancer to another bird. And so it went from fowl to fowl. The transforming agent turned out to be a retrovirus—the kind that can smuggle cancer-causing genes into otherwise healthy cells.

  There was src, which was
part of the virus that caused sarcoma in chickens. Another gene, called ras, induced sarcoma in rats, while fes did the same in felines. Myc and myb induced blood cell cancers, myelocytomatosis and myeloblastosis, in poultry. If that is where the research had ended it would have made for a tidy picture. Cancer could be caused when chemicals or radiation mutated preexisting genes, or when viruses surreptitiously inserted entirely new ones—oncogenes, they were called—already capable of causing cancer. Two fundamental ways of modifying genetic information. But the real story turned out to be far more interesting.

  There was a problem reconciling Rous’s discovery with what appeared to be happening in the world. Cancer wasn’t acting like a contagious disease sweeping through populations like polio. It arose sporadically in various places. Even Rous’s chicken virus spread only when it was injected, and try as he might he couldn’t transfer it to other animals—pigeons, ducks, rats, mice, guinea pigs, rabbits. Only with great difficulty could it be induced in other chickens except closely related Plymouth hens. Even more suggestive, scientists were not finding the retroviruses inside human tumors. What they were discovering instead was that genomes of creatures throughout the animal kingdom contained what appeared to be naturally occurring versions of src, ras, fes, myb, myc—not ones that had been smuggled in. These were not broken, mutated genes like their viral counterparts. Their purpose was to govern how healthy cells divide, the process biologists call mitosis.