The Emperor of All Maladies: A Biography of Cancer (30 page)

In the early 1950s, at a cocktail party in California, Henry Kaplan, a professor of radiology at Stanford, overheard a conversation about
the plan to build a linear accelerator
for use by physicists at Stanford. A linear accelerator is an X-ray tube taken to an extreme form. Like a conventional X-ray tube, a linear accelerator also fires electrons onto a target to generate high-intensity X-rays. Unlike a conventional tube, however, the “linac” imbues massive amounts of energy into the electrons, pushing them to dizzying velocities before smashing them against the metal surface. The X-rays that emerge from this are deeply penetrating—powerful enough not only to pass through tissue, but to scald cells to death.

Kaplan had trained at the NCI, where he had learned to use X-rays to treat leukemia in animals, but his interest had gradually shifted to solid tumors in humans—lung cancer, breast cancer, lymphomas. Solid tumors could be treated with radiation, he knew, but the outer shell of the cancer, like its eponymous crab’s carapace, needed to be penetrated deeply to kill cancer cells. A linear accelerator with its sharp, dense, knifelike beam might allow him to reach tumor cells buried deep inside tissues.
In 1953, he persuaded a team
of physicists and engineers at Stanford to tailor-make
an accelerator exclusively for the hospital.
The accelerator was installed
in a vaultlike warehouse in San Francisco in 1956. Dodging traffic between Fillmore Street and Mission Hill, Kaplan personally wheeled in its colossal block of lead shielding on an automobile jack borrowed from a neighboring garage owner.

Through a minuscule pinhole in that lead block, he could now direct tiny, controlled doses of a furiously potent beam of X-rays—millions of electron volts of energy in concentrated bursts—to lancinate any cancer cell to death. But what form of cancer? If Kaplan had learned one lesson at the NCI, it was that by focusing microscopically on a single disease, one could extrapolate into the entire universe of diseases. The characteristics that Kaplan sought in his target were relatively well defined. Since the linac could only focus its killer beam on local sites, it would have to be a local, not a systemic, cancer. Leukemia was out of the question. Breast and lung cancer were important targets, but both were unpredictable, mercurial diseases, with propensities for occult and systemic spread. The powerful oculus of Kaplan’s intellect, swiveling about through the malignant world, ultimately landed on the most natural target for his investigation: Hodgkin’s disease.

“
Henry Kaplan
was
Hodgkin’s disease,” George Canellos, a former senior clinician at the NCI told me, leaning back in his chair. We were sitting in his office while he rummaged through piles of manuscripts, monographs, articles, books, catalogs, and papers, pulling out occasional pictures of Kaplan from his files. Here was Kaplan, dressed in a bow tie, looking at sheaves of papers at the NCI. Or Kaplan in a white coat standing next to the linac at Stanford, its 5-million-volt probe just inches from his nose.

Kaplan wasn’t the first doctor to treat Hodgkin’s with X-rays, but he was certainly the most dogged, the most methodical, and the most single-minded. In the mid-1930s, a Swiss radiologist named
Rene Gilbert had shown
that the swollen lymph nodes of Hodgkin’s disease could effectively and dramatically be reduced with radiation. But Gilbert’s patients had typically relapsed after treatment, often in the lymph nodes immediately contiguous to the original radiated area. At the Toronto General Hospital, a Canadian surgeon named Vera Peters had furthered Gilbert’s studies by broadening the radiation field even farther—delivering X-rays not to a single swollen node, but to an entire area of lymph nodes. Peters called
her strategy “extended field radiation.” In 1958, analyzing the cohort of patients that she had treated,
Peters observed that broad-field radiation could
significantly improve long-term survival for early-stage Hodgkin’s patients. But Peters’s data was retrospective—based on the historical analysis of prior-treated patients. What Peters needed was a more rigorous medical experiment, a randomized clinical trial. (Historical series can be biased by doctors’ highly selective choices of patients for therapy, or by their counting only the ones that do the best.)

Independently of Peters, Kaplan had also realized that extended field radiation could improve relapse-free survival, perhaps even cure early-stage Hodgkin’s disease. But he lacked formal proof. In 1962, challenged by one of his students, Henry Kaplan set out to prove the point.

The trials that Kaplan designed
still rank among the classics of study design. In the first set, called the L1 trials, he assigned equal numbers of patients to either extended field radiation or to limited “involved field” radiation and plotted relapse-free survival curves. The answer was definitive. Extended field radiation—“
meticulous radiotherapy
” as one doctor described it—drastically diminished the relapse rate of Hodgkin’s disease.

But Kaplan knew that a diminished relapse rate was not a cure
. So he delved further. Two years later, the Stanford team carved out a larger field of radiation, involving nodes around the aorta, the large arch-shaped blood vessel that leads out of the heart. Here they introduced an innovation that would prove pivotal to their success. Kaplan knew that only patients that had localized Hodgkin’s disease could possibly benefit from radiation therapy. To truly test the efficacy of radiation therapy, then, Kaplan realized that he would need a strictly limited cohort of patients whose Hodgkin’s disease involved just a few contiguous lymph nodes. To exclude patients with more disseminated forms of lymphoma, Kaplan devised an intense battery of tests to stage his patients. There were blood tests, a detailed clinical exam, a procedure called lymphangiography (a primitive ancestor of a CT scan for the lymph nodes), and a bone marrow biopsy. Even so, Kaplan was unsatisfied: doubly careful, he began to perform exploratory abdominal surgery and biopsy internal nodes to ensure that only patients with locally confined disease were entering his trials.

The doses of radiation were now daringly high. But gratifyingly, the responses soared as well. Kaplan documented even greater relapse-free
intervals, now stretching out into dozens of months—then years. When the first batch of patients had survived five years without relapses, he began to speculate that some may have been cured by extended field X-rays. Kaplan’s experimental idea had finally made its way out of a San Francisco warehouse into the mainstream clinical world.

But hadn’t Halsted wagered on the same horse and lost? Hadn’t radical surgery become entangled in the same logic—carving out larger and larger areas for treatment—and then spiraled downward? Why did Kaplan succeed where others had failed?

First, because Kaplan meticulously restricted radiotherapy to patients with early-stage disease. He went to exhaustive lengths to stage patients before unleashing radiation on them. By strictly narrowing the group of patients treated, Kaplan markedly increased the likelihood of his success.

And second, he succeeded because he had picked the right disease. Hodgkin’s was, for the most part, a regional illness. “
Fundamental to all attempts at curative treatment
of Hodgkin’s disease,” one reviewer commented memorably in the
New England Journal of Medicine
in 1968, “is the assumption that in the significant fraction of cases, [the disease] is localized.” Kaplan treated the intrinsic biology of Hodgkin’s disease with utmost seriousness. If Hodgkin’s lymphoma had been more capricious in its movement through the body (and occult areas of spread more common, as in some forms of breast cancer), then Kaplan’s staging strategy, for all his excruciatingly detailed workups, would inherently have been doomed to fail. Instead of trying to tailor the disease to fit his medicine, Kaplan learned to tailor his medicine to fit the right disease.

This simple principle—the meticulous matching of a particular therapy to a particular form and stage of cancer—would eventually be given its due merit in cancer therapy. Early-stage, local cancers, Kaplan realized, were often inherently different from widely spread, metastatic cancers—even within the same form of cancer. A hundred instances of Hodgkin’s disease, even though pathologically classified as the same entity, were a hundred variants around a common theme. Cancers possessed temperaments, personalities—behaviors. And biological heterogeneity demanded therapeutic heterogeneity; the same treatment could not indiscriminately be applied to all. But even if Kaplan understood it fully in 1963 and made an example of it in treating Hodgkin’s disease, it would take decades for a generation of oncologists to come to the same realization.

Now we are an army on the march
.

—Sidney Farber in 1963

The next step—the complete cure
—is almost sure to follow.

—Kenneth Endicott,
NCI director, 1963

The role of aggressive multiple drug therapy
in the quest for long-term survival [in cancer] is far from clear.

—R. Stein, a scientist in 1969

One afternoon in the late summer of 1963,
George Canellos, then a senior fellow at the NCI
, walked into the Clinical Center to find Tom Frei scribbling furiously on one of the institute’s blackboards. Frei, in his long white coat, was making lists of chemicals and drawing arrows. On one side of the board was a list of cytotoxic drugs—Cytoxan, vincristine, procarbazine, methotrexate. On the other side was a list of new cancers that Zubrod and Frei wanted to target: breast, ovarian, lung cancers, lymphomas. Connecting the two halves of the blackboard were chalky lines matching combinations of cytotoxic drugs to cancers. For a moment, it almost looked as if Frei had been deriving mathematical equations: A+B kills C; E+F eliminates G.

The drugs on Frei’s list came largely from three sources. Some, such as aminopterin or methotrexate, were the products of inspired guesswork by scientists (Farber had discovered aminopterin by guessing that an antifolate might block the growth of leukemia cells). Others, such as nitrogen mustard or actinomycin D, came from serendipitous sources, such
as mustard gas or soil bacteria, found accidentally to kill cancer cells. Yet others, such as 6-MP, came from drug-screening efforts in which thousands of molecules were tested to find the handful that possessed cancer-killing activity.

The notable common feature that linked all these drugs was that they were all rather indiscriminate inhibitors of cellular growth. Nitrogen mustard, for instance, damages DNA and kills nearly all dividing cells; it kills cancer cells somewhat preferentially because cancer cells divide most actively. To design an ideal anticancer drug, one would need to identify a specific molecular target in a cancer cell and create a chemical to attack that target. But the fundamental biology of cancer was so poorly understood that defining such molecular targets was virtually inconceivable in the 1960s. Yet, even lacking such targets, Frei and Freireich had cured leukemia in some children. Even generic cellular poisons, dosed with adequate brio, could thus eventually obliterate cancer.

The bravado of that logic was certainly hypnotic. Vincent DeVita, another fellow at the institute during that time, wrote, “
A new breed of cancer investigators in the 1960s
had been addressing the generic question of whether or not cytotoxic chemotherapy was ever capable of curing patients with any type of advanced malignancies.” For Frei and Zubrod, the only way to answer that “generic question” was to direct the growing armamentarium of combination chemotherapy against another cancer—a solid tumor this time—which would retrace their steps with leukemia. If yet another kind of cancer responded to this strategy, then there could be little doubt that oncology had stumbled upon a generic solution to the generic problem. A cure would then be within reach for all cancers.

But which cancer would be used to test the principle? Like Kaplan, Zubrod, DeVita, and Canellos also focused on Hodgkin’s disease—a cancer that lived on the ill-defined cusp between solid and liquid, a stepping-stone between leukemia and, say, lung cancer or breast cancer. At Stanford, Kaplan had already demonstrated that Hodgkin’s lymphoma could be staged with exquisite precision and that local disease could be cured with high-dose extended field radiation. Kaplan had solved half the equation: he had used local therapy with radiation to cure localized forms of Hodgkin’s disease. If metastatic Hodgkin’s disease could be cured by systemic and aggressive combination chemotherapy, then Zubrod’s “generic solution” would begin to sound plausible. The equation would be fully solved.