that an ovarian cancer would metastasize to an inguinal lymph node. He was puzzled by the order for a barium enema, which was scheduled for a few days later. “It’s a useless test,” he said. We explained about the long wait for a colonoscopy. He picked up the phone, called one of the physicians who owned the clinic, and we had an appointment two days later. “We are going to cure you,” he said. That at least is my memory.Oncologists are not supposed to say that. It was encouraging that he didn’t give a damn.
The results of the colonoscopy were negative, and the final stepwas aPET scan. Santa Fe had just gotten its own machine—it was no longer necessary to drive an hour south to Albuquerque—and Nancy was almost first in line. PET stands for positron emission tomography, a triumph of medical technology from the recondite world of particle physics. The patient fasts the night before so the body’s cells are starving. When radio-taggedglucose is injected it is eagerly consumed. Malignant, rapidly dividing cells are especially voracious, concentrating the radioactive molecules. As these decay they shoot out positrons, particles of antimatter that collide with electrons and produce bursts ofgamma rays. They strike a scintillator, which responds by emitting flashes of light. Nancy’s lower uterus was glowing from the feasting of hyperactive endometrial cells, descendants of a single cell gone mad, a cell that had forgotten it was part of a community, that began running its own show—an isolated act of betrayal that has been played out again and again since the first archaic cells grudgingly agreed to surrender their autonomy for the advantages of living in a collective.
In the days after the diagnosis, I beganreading about how this might have happened. To carry on in harmony, our cells constantly exchange chemical signals, conferring on when to start multiplying and creating new tissue. As each cell receives this information it responds by sending instructions to its nucleus, the central controller, for activating the appropriate combination of genes—pressing the right buttons, hitting an arpeggio of piano keys. A cancer cell is one that has cut itself out of the discussion, solipsistically deciding on its own. Random events—triggered by a cosmic ray, a carcinogenic chemical, or just plain dumb luck—must have altered the DNA inside one of Nancy’s cells, causing it to lose touch. The trouble might have begun with amutation to a gene that sends signals telling the cell that it is time to divide. Another mutation might have modified the molecular receptors that respond to the signals, causing them to become hypersensitive. Set on a hair trigger theyfire prematurely. Either way, the cell begins multiplying more rapidly than its neighbors.
In fact these kinds of errors happen all the time. We usually don’t get cancer because other genes react to sudden bursts of activity by reining in growth. But another mutation can cause that safeguard to fail. The nucleus of a cell is constantly receiving messages, weighing the evidence and deciding what to do next. The calculations depend on a tangle of molecular cascades—more things that can go wrong. And they do. All the time. The mistakes are caught and corrected. The DNA is repaired. If that fails, a cell can sense the inner turmoil and send itself suicide signals, killing itself for the common good. But another mutation can undermine that defense.
This is all usually described as though a single cell was sitting motionless and accumulating these defects over the years. I tried to imagine the process as it really is, unfolding dynamically. One hit causes a cell to start dividing repeatedly. Then one of its many progeny acquires another mutation, and its progeny acquire still more. The longer a lineage of cells has lived the more likely it is to have mutated to the brink. That still leaves another barrier against runaway growth: a counter that monitors and limits