by Dan Fagin
What about young children with cancer? Could an environmentally induced mutation really be passed down from one generation to the next, just like hair or eye color? Hermann Joseph Muller thought so. Another in a long line of brilliant but abrasive biologists, Muller had been thinking about that question for practically his whole life. When he was a young boy in New York City, in 1899, his father had taken him to the American Museum of Natural History and explained, with the help of a display of horses’ hooves, how organisms evolved via natural selection acting on accidental variation. His father died soon afterward, but the lesson stayed with his eight-year-old son. If evolution could occur naturally, Muller thought, surely the deliberate actions of humans could affect the future evolution of his own species—for good or ill.6
Small and bellicose—he compensated for his size by standing on tiptoe while arguing—Muller was a social misfit who, like Boveri, was prone to depression.7 Despite his unhappiness, or perhaps driven by it, Muller did brilliant work as a young researcher at the University of Texas. He conducted a series of experiments in which he irradiated fruit flies with X-rays, which were already widely known to trigger cancerous tumors in animals and humans alike.8 In 1927, Muller was ready to divulge his results, which caused an immediate sensation. By exposing the sperm cells of fruit flies to X-rays, Muller reported, he was able to generate heritable mutations—white eyes in male flies instead of red, for example—almost at will. Many of those mutations were lethal, triggering a wide range of abnormalities, including sterility.9
Carcinogens could mutate genes; Muller had validated Boveri’s prediction. Not all mutagens triggered cancer, and not all carcinogens did their work via mutation, but there was a strong correlation between the two—as the people of Toms River would learn when they found out what was in their drinking water. More broadly, Muller helped to solidify the view that genes, with their unique ability to copy themselves and transmit information to daughter cells, must control cell development and therefore the growth of the entire organism. Genes were the wellspring of life—and also of death, depending on the particular the message they carried. Muller had answered the question that first occurred to him as a young boy at the museum: Yes, human beings could manipulate their genes in ways that would affect not only their survival as individuals but the long-term viability of their species. Muller was awarded a Nobel Prize for his genetics research in 1946, just after the atomic bombings of Hiroshima and Nagasaki. He spent the next twenty years studying fruit fly mutations and campaigning against the use of radiation in weaponry, industry, and medicine—“time bombing our descendants,” he called it.10 The sins of one generation, Muller suggested, would be visited on the next through genetic mutation and cancer.
One of the Toms River children whom Linda Boornazian met on the oncology ward at The Children’s Hospital of Philadelphia in the early 1990s was Michael Anderson. He was ten years old when he was first admitted to CHOP in September of 1991, the culmination of a summer of intensifying agonies that began with a sore thumb on May 8. Mike’s birthday was May 11, but by then he was already too sick to enjoy it—his ankles and hands ached, and he was running a fever. A month-long course of antibiotics did not help. By July, he was in a wheelchair, his feet and hands bent like an old man’s and too painful to move. On an initial visit to Philadelphia, the doctors diagnosed juvenile rheumatoid arthritis, but the medications they prescribed proved as useless as the antibiotics. His mother, Melanie, had to feed and wash her son as if he were a baby. When Mike went back to school, his mother went with him, carrying him up and down the stairs. His father, Bruce, who worked as a reactor operator at the Oyster Creek nuclear power plant, began to worry about the family’s health insurance. Mike’s three brothers—ages seven, thirteen, and sixteen—watched, terrified, as the unchecked and unidentified disease ravaged his small body. Circle-shaped bruises appeared all over him. They were not the impact bruises Mike used to get from playing soccer or riding his bike—the bruises all rambunctious young boys get—but something deeper, stranger.
His parents suspected leukemia, but the doctors said that was impossible; the blood tests were negative. Even so, Bruce and Melanie Anderson insisted on another examination in Philadelphia in September of 1991. This time, a new pediatric rheumatologist saw Mike. His name was Steven Goodman, and he was only a few years out of medical school. He looked at Mike’s case in a new way—asking more questions, conducting a more thorough physical exam. Examining Mike’s fingers, Goodman noticed that Mike’s bones, not arthritic joints, were the source of his pain. He ordered another set of blood tests, and this time the results were different. Mike’s platelets were low, and his white blood cell count was high—the same lethal combination Rudolf Virchow had dubbed leukemia after observing it under his microscope in 1845. The Andersons got the catastrophic news on Friday the Thirteenth: Mike had a rare form of leukemia that was essentially a combination of the two most common types, acute lymphocytic leukemia, known as ALL, and acute myelogenous leukemia, or AML. The prognosis was dire; the required double dose of extended chemotherapy would be vicious.
The next three years were “a roller coaster through hell,” as Bruce Anderson would remember them. After six weeks of intensive chemotherapy at the hospital, Mike went home to Toms River, returning at least once a month for another debilitating round of chemotherapy drugs, administered through a catheter surgically attached to his chest. There were many other trips to the hospital, too—harrowing, late-night runs across New Jersey whenever Mike had a high fever. The regimen of chemotherapy and steroids was gradually killing the cancer cells, but it was also wrecking Mike’s immune system, leaving him open to an array of opportunistic infections, any of which could have killed him. Melanie learned to keep an overnight bag packed and ready at all times in case they had to rush back to Philadelphia. She and Bruce also learned to be assertive with the nurses and doctors. They asked questions, checked medication dosages, and insisted that all of Mike’s visitors first wash their hands. The Andersons connected with Linda Gillick via Ocean of Love, a support group she had formed in 1988. Raising money in the community for families struck by childhood cancer, Gillick delivered Easter baskets, Thanksgiving turkeys, and Christmas gifts to the Andersons and other families. She organized summer picnics and trips to Broadway shows. And for families who needed it, she would provide direct financial support—including for funeral expenses. She was driven, and she was tireless. The Andersons were amazed by Gillick’s devotion to other sick children, even as her own son’s condition remained perilous.
Very gradually, as Mike Anderson’s health improved and the sense of perpetual crisis that had enveloped his family began to ease, his parents allowed themselves to ponder the deeper questions raised by their son’s illness. Bruce Anderson knew about Linda Gillick’s map of local childhood cancer cases, and he came to share her conviction that a chilling pattern was emerging—a cluster.
Robert Gialanella thought so too. He had met Linda Gillick in 1989 in a store checkout line at the Ocean County Mall, while she was waiting to pay for a huge basketful of holiday toys for the children. He was a local physician, and within a few weeks he was also a board member of Ocean of Love, where he would remain for thirteen years. Gialanella’s oldest son was born one day before Michael Gillick, and he came to think of the Ocean of Love children as extensions of his own family. “Linda was so committed to these kids,” he remembered. “I just felt I wanted to be involved.” Gialanella was a gastroenterologist, not an oncologist, and he treated adults, not children. But he knew a bit about epidemiology, and in 1991 he and Gillick became alarmed as the number of pushpins on her map proliferated across Ocean County and especially in Toms River. If something in the environment really was causing those cancer cases, it seemed to be gaining potency every year. Gialanella decided to try to find someone at the state health department who might be willing to investigate. It took a while to locate the right person, but in 1991 Gialanella spoke to Michael Berry, the same health d
epartment employee who five years earlier had responded to a very similar request from Chuck Kauffman of the Ocean County Health Department.
That earlier study had found no statistically significant elevation in the local childhood cancer rate, but Berry agreed to Gialanella’s request that he take another look because this time a few more years of registry data could be included in the analysis, increasing the overall number of cases and slightly reducing the effects of chance. In fact, there were enough cases by now that Berry could try to look at some specific categories of diseases instead of lumping all childhood cancers together. Also, there had been some improvements at the cancer registry in collecting data from out-of-state hospitals, so it was more likely that at least some of the Toms River children treated in Philadelphia and New York would be included this time. One problem, however, had gotten worse: The registry was now four years behind schedule; the most recent full year of data was 1987. The cases Chuck Kauffman had heard about in 1984 and 1985 would show up in Berry’s second study, but all of the cases that Linda Gillick had heard about in 1989, 1990, and 1991 would not. The state cancer registry was like a surveillance camera whose film took four years to develop. By the time the pictures were available, the world had moved on.
Berry ran the numbers, and this time the results were ambiguous—the number of cases was higher than expected, but not by much. Gialanella was disappointed, but not surprised. “We knew there were more kids than were documented in the state reports,” Gialanella remembered. “We were looking at real-time data, all of the kids we knew about. I think we kind of said to ourselves that if the state had more current data it might show something more significant.” Berry’s inability to confirm the existence of a cluster was extremely frustrating to Linda Gillick and Bruce Anderson, who by now were convinced the cluster was real and getting worse. Ocean of Love stopped asking the state to investigate—that route now seemed hopeless—but Gillick kept adding pins to her map, and Anderson started reading everything he could find about environmental causes of cancer.
Anderson’s personality was as forceful and stubborn as Gillick’s, but his intensity smoldered, while hers often burst into flames. He had an engineer’s mind; he liked to pull things apart and see how they worked. He also liked to build with his own hands, including the house on Malcolm Street where the family had lived since 1984. Now he applied that same meticulousness to the process of learning everything he could about the causes of childhood leukemia, including the possible role of environmental pollutants in combination with inherited vulnerabilities. One of his son’s doctors had mentioned something called the “two hit” theory of carcinogenesis, developed by a researcher who was now in Philadelphia. The idea intrigued Anderson; he wanted to know more.
More than forty years earlier, in another hospital ward in another city, the father of the two-hit hypothesis, Alfred Knudson, first encountered the mystery and pathos of childhood cancer. He would go on to become one of the world’s most influential cancer geneticists, but in March of 1949 he was a twenty-six-year-old medical resident on a one-month rotation in the pediatric oncology ward at Memorial Hospital in New York City. Most of the patients in the twenty-bed ward had acute lymphocytic leukemia, and for the first time there was some hope of curing them. A new class of chemotherapy drugs seemed to slow tumor growth by impeding the function of folic acid. “Before the antifolates, there was nothing you could do except just watch the leukemia kids die. It would usually take three or four months,” Knudson remembered many years later. “I was with those kids all of the time for that one month on the ward. It was total immersion, and it was exciting because there were remissions for the first time in leukemia.”
Knudson’s rotation in the ward soon ended, and so did the remissions. Most of the children with leukemia relapsed within a few months, with little to show from the experimental treatment except a few extra months of life and an awful array of debilitating side effects. Antifolate drugs would later become a useful part of the standard chemotherapy regimen, but their early failure was a wrenching experience for the young doctor-in-training. “When you see twenty kids all together on the ward,” he said, “it just suddenly hits you: Why, and how, do these kids get cancer?”
His memories of those helpless young patients, many of whom would never go home, stuck with Knudson—so much so that he later decided to give up clinical medicine for a research career focused on trying to understand the causes of childhood cancer. In 1964, he was at the City of Hope Medical Center in California when Hermann Muller came through as a visiting researcher. The diminutive leftist was now seventy-five years old and was still working with fruit flies; briefly hospitalized for a heart ailment while in California, he insisted that an assistant bring his flies to his bedside so that he could inspect them. He was also still buzzing energetically over the unsolved problems of mutation and cancer. Why did it often take many years—or several generations, in the case of fruit flies—for radiation exposure to trigger cancerous tumors? The answer, Muller believed, was that several mutation “events” were required to transform a healthy cell into a malignant one.11 Cancer, in other words, was more like a long-distance relay than a solo sprint.
Alfred Knudson agreed with his eminent visitor. To Knudson, a multistage sequence of carcinogenesis fit the evidence. He and Muller were not the only scientists who thought so. The idea had been embraced by cancer researchers as diverse as Katsusaburo Yamagiwa, Ernest Kennaway, and Wilhelm Hueper, each of whom knew from firsthand experience that inducing tumors in rabbits, mice, dogs, or other experimental animals required many toxic doses over long periods of time. They also knew that some individuals were especially vulnerable to carcinogenic exposures and that most victims were elderly. Those observations all suggested that there were multiple steps to malignancy and that the journey could take many years. Richard Doll theorized that carcinogenesis required six or seven mutations.12 But that was just an educated guess, an estimation based on Doll’s observation that the cancer death rate for the elderly was more than seven times higher than for young adults.
Knudson wondered whether he could push the multistage hypothesis beyond guesswork. He had read Theodor Boveri’s groundbreaking 1914 book on mutation and cancer and was excited by a burst of recent findings that seemed to confirm many of Boveri’s ideas. Genetics was in the midst of a revolution in the 1960s. James Watson and Francis Crick’s success in elucidating the double helix structure of DNA in 1953 at last revealed exactly how hereditary information is stored in the chromosomes of living organisms, as Boveri had predicted sixty years earlier. Now researchers could peer inside the twenty-three pairs of chromosomes in every human cell and identify DNA segments, or genes, associated with specific traits and bodily functions—and malfunctions, too.
Many of the breakthroughs that followed concerned cancer. The most famous was the 1960 identification of a mutation known as the Philadelphia chromosome, named for the city where it was discovered. The bone marrow cells of 95 percent of adults with chronic myelogenous leukemia carried the telltale genetic defect. It was the first direct evidence that chromosome alterations preceded cancer and thus was another confirmation of Boveri’s ideas. The Philadelphia mutation was a scramble: A chunk of DNA from Chromosome 9 swapped places with a chunk from Chromosome 22. What was puzzling was that not everyone whose cells had this chromosomal translocation got leukemia; many did not. To Knudson, that was additional evidence that more than one mutational event was required. The identification of the Philadelphia chromosome and other translocations associated with cancers would eventually lead to the discovery of oncogenes, the rogue mutations predicted by Boveri that promote the rapid cell division of cancer. But even the presence of an oncogene in a cell was not a surefire indicator that malignancy would result—one mutation was often not enough.
The multiple-hit theory of carcinogenesis made sense to Knudson, but his experience treating childhood cancer patients led him to doubt Richard Doll’s suggestion that six or seven successive
mutations were needed to trigger a malignancy. After all, some of Knudson’s young patients were born with neuroblastoma; others developed leukemia as infants. Mutations were rare events; surely a baby had not lived long enough to take a half-dozen genetic hits. On the other hand, it did not seem to make sense that only one mutation was necessary. The human body produced one hundred billion white blood cells daily, yet leukemia was still a relatively rare condition. If just one mutation was required, why didn’t everyone have it? With those questions in his head, Knudson began looking for a childhood cancer he could study to test his ideas about multiple hits.
He found retinoblastoma. Like Michael Gillick’s neuroblastoma and Randy Lynnworth’s medulloblastoma, retinoblastoma begins with the malignant transformation of precursor stem cells, in this case, in the eye. These retinoblasts produce the specialized light-sensitive cells that line the retina, making vision possible. A very rare disease, afflicting one in fifteen thousand children, retinoblastoma comes in two varieties: hereditary and sporadic. Hereditary retinoblastomas, about 40 percent of all cases, afflict children who have a family history of the disease, while sporadic cases do not. What caught Knudson’s attention were three quirks that apply only to hereditary retinoblastomas. First, they sometimes skip a generation: A grandparent is afflicted, a parent is not, and then the disease returns in a grandchild. Second, kids with a family history of retinoblastoma usually develop the disease very early in childhood—even during infancy. And finally, children with hereditary retinoblastoma very often develop more than one tumor in one or both eyes.
