But the story had a final plot twist. As Li had predicted, with several additional doses of methotrexate, the hormone level that he had so compulsively trailed did finally vanish to zero. His patients finished their additional cycles of chemotherapy. Then, slowly, a pattern began to emerge. While the patients who had stopped the drug early inevitably relapsed with cancer, the patients treated on Li’s protocol remained free of disease—even months after the methotrexate had been stopped.
Li had stumbled on a deep and fundamental principle of oncology: cancer needed to be systemically treated long after every visible sign of it had vanished. The hcg level—the hormone secreted by choriocarcinoma—had turned out to be its real fingerprint, its marker. In the decades that followed, trial after trial would prove this principle. But in 1960, oncology was not yet ready for this proposal. Not until several years later did it strike the board that had fired Li so hastily that the patients he had treated with the prolonged maintenance strategy would never relapse. This strategy—which cost Min Chiu Li his job—resulted in the first chemotherapeutic cure of cancer in adults.
A model is a lie that helps334 you see the truth.
—Howard Skipper
Min Chiu Li’s experience with choriocarcinoma was a philosophical nudge for Frei and Freireich. “Clinical research is a matter of urgency,”335 Freireich argued. For a child with leukemia, even a week’s delay meant the difference between life and death. The academic stodginess of the leukemia consortium—its insistence on progressively and systematically testing one drug combination after another—was now driving Freireich progressively and systematically mad. To test three drugs, the group insisted336 on testing “all of the three possible combinations and then you’ve got to do all of the four combinations and with different doses and schedules for each.” At the rate that the leukemia consortium was moving, he argued, it would take dozens of years before any significant advance in leukemia was made. “The wards were filling up with these terribly sick children.337 A boy or girl might be brought in with a white cell count of three hundred and be dead overnight. I was the one sent the next morning to speak with the parents. Try explaining Zubrod’s strategy of sequential, systematic, and objective trials to a woman whose daughter has just slumped into a coma and died,” Freireich recalled.
The permutations of possible drugs and doses were further increased when yet another new anticancer agent was introduced at the Clinical Center in 1960. The newcomer, vincristine, was a poisonous plant-alkaloid that came from the Madagascar periwinkle, a small, weedlike creeper with violet flowers and an entwined, coiled stem. (The name vincristine comes from vinca, the Latin word for “bind.”) Vincristine had been discovered in 1958338 at the Eli Lilly company through a drug-discovery program that involved grinding up thousands of pounds of plant material and testing the extracts in various biological assays. Although originally intended as an antidiabetic, vincristine at small doses was found to kill leukemia cells. Rapidly growing cells, such as those of leukemia, typically create a skeletal scaffold of proteins (called microtubules) that allows two daughter cells to separate from each other and thereby complete cell division. Vincristine works by binding to the end of these microtubules and then paralyzing the cellular skeleton in its grip—thus, quite literally, evoking the Latin word after which it was originally named.
With vincristine added to the pharmacopoeia, leukemia researchers found themselves facing the paradox of excess: how might one take four independently active drugs—methotrexate, prednisone, 6-MP, and vincristine—and stitch them together into an effective regimen? And since each drug was potentially severely toxic, could one ever find a combination that would kill the leukemia but not kill a child?
Two drugs had already spawned dozens of possibilities; with four drugs, the leukemia consortium would take not fifty, but a hundred and fifty years to finish its trials. David Nathan, then a new recruit at the NCI, recalled the near standstill created by the avalanche of new medicines: “Frei and Freireich were simply taking drugs339 that were available and adding them together in combinations. . . . The possible combinations, doses, and schedules of four or five drugs were infinite. Researchers could work for years on finding the right combination of drugs and schedules.” Zubrod’s sequential, systematic, objective trials had reached an impasse. What was needed was quite the opposite of a systematic approach—an intuitive and inspired leap of faith into the deadly abyss of deadly drugs.
A scientist from Alabama, Howard Skipper340—a scholarly, soft-spoken man who liked to call himself a “mouse doctor”—provided Frei and Freireich a way out of the impasse. Skipper was an outsider to the NCI. If leukemia was a model form of cancer, then Skipper had been studying the disease by artificially inducing leukemias in animals—in effect, by building a model of a model. Skipper’s model used a mouse cell line called L-1210, a lymphoid leukemia that could be grown in a petri dish. When laboratory mice were injected with these cells, they would acquire the leukemia—a process known as engraftment because it was akin to transferring a piece of normal tissue (a graft) from one animal to another.
Skipper liked to think about cancer not as a disease but as an abstract mathematical entity. In a mouse transplanted with L-1210 cells, the cells divided with nearly obscene fecundity—often twice a day, a rate startling even for cancer cells. A single leukemia cell engrafted into the mouse could thus take off in a terrifying arc of numbers: 1, 4, 16, 64, 256, 1,024, 4,096, 16,384, 65,536, 262,144, 1,048,576 . . . and so forth, all the way to infinity. In sixteen or seventeen days, more than 2 billion daughter cells could grow out of that single cell—more than the entire number of blood cells in the mouse.
Skipper learned that he could halt this effusive cell division by administering chemotherapy to the leukemia-engrafted mouse. By charting the life and death of leukemia cells as they responded to drugs in these mice, Skipper emerged with two pivotal findings341. First, he found that chemotherapy typically killed a fixed percentage of cells at any given instance no matter what the total number of cancer cells was. This percentage was a unique, cardinal number particular to every drug. In other words, if you started off with 100,000 leukemia cells in a mouse and administered a drug that killed 99 percent of those cells in a single round, then every round would kill cells in a fractional manner, resulting in fewer and fewer cells after every round of chemotherapy: 100,000 . . . 1,000 . . . 10 . . . and so forth, until the number finally fell to zero after four rounds. Killing leukemia was an iterative process, like halving a monster’s body, then halving the half, and halving the remnant half.
Second, Skipper found that by adding drugs in combination, he could often get synergistic effects on killing. Since different drugs elicited different resistance mechanisms, and produced different toxicities in cancer cells, using drugs in concert dramatically lowered the chance of resistance and increased cell killing. Two drugs were therefore typically better than one, and three drugs better than two. With several drugs and several iterative rounds of chemotherapy in rapid-fire succession, Skipper cured leukemias in his mouse model.
For Frei and Freireich, Skipper’s observations had an inevitable, if frightening, conclusion. If human leukemias were like Skipper’s mouse leukemias, then children would need to be treated with a regimen containing not one or two, but multiple drugs. Furthermore, a single treatment would not suffice. “Maximal, intermittent, intensive, up-front”342