Oh, they kept taking rooms away from me, and help. But I—my best people stayed with me.
—Bernice Eddy, former NIH vaccine safety scientist, 19861
IN THE SPRING of 1960, at the National Institutes of Health in Bethesda, Maryland, just outside Washington, DC, a feisty fifty-seven-year-old PhD scientist with an open, square face and neatly coiffed dark brown hair was worrying about her alarming findings in a group of newborn hamsters.
Bernice Eddy had grown up in a family of physicians in Auburn, a town of 199 in rural West Virginia, and earned her PhD in microbiology from the University of Cincinnati. After a stint researching leprosy in Louisiana, she came to the NIH in 1937.2 By the late 1950s she had been working for more than twenty years in the NIH division that was then the gatekeeper of the U.S. vaccine market. The Division of Biologics Standards (DBS) filled the role played by a key branch of today’s Food and Drug Administration, assessing new vaccines and issuing licenses when it deemed a product ready for market. Eddy had done well in her work scrutinizing vaccine safety and effectiveness: in 1953 she won a “superior accomplishment” award and a pay raise from the NIH director.
The next year Eddy was put in charge of the unit’s polio vaccine safety tests. It was a high-stakes job. On the heels of a 1952 polio epidemic that had infected nearly 58,000 Americans and paralyzed more than 21,000 of them—the country’s worst polio epidemic ever—virologists were in the midst of an all-out sprint to get a polio vaccine licensed. That first vaccine, newly invented by Jonas Salk, was made from polio virus that multiplied in monkey kidney cells and was then killed with formaldehyde. The Salk vaccine was entering human trials in 1954 when Eddy isolated live polio virus from three lots of the vaccine made by the California-based Cutter company. She injected monkeys with the suspect vaccine and found that it paralyzed some of them. She reported the findings to her bosses and sent them photos of the afflicted animals.3 They were ignored, and in April 1955 the Salk vaccine was licensed.4 Mass vaccinations began, and the Cutter vaccine was among those distributed. It paralyzed 192 people, many of them children.5 Ten people died.6 The government was forced to temporarily recall all polio vaccine, sowing public panic. Manufacturing changes were mandated, and the vaccine returned to the market. But public confidence took months to rebuild, a fact that was responsible for many of the more than 28,000 cases of polio in the United States that year.7
Despite her earlier efforts to warn her bosses about the flawed vaccine, and over her protestations, Eddy was pulled off polio vaccine work as part of the Cutter-episode fallout.8 It was worse for the higher-ups: senior officials including the director of the NIH, William Sebrell, and the secretary of health, education and welfare, Oveta Culp Hobby, lost their jobs. So did William Workman, Eddy’s boss.
Eddy, who would show amazing staying power over the years, persisted at the DBS and in 1959 landed in Time magazine with her friend and NIH colleague physician/scientist Sarah Stewart. The pair had discovered a mouse virus that was now named “SE polyoma”—“SE” for Stewart and Eddy, and “polyoma” meaning “many tumors.” Stewart isolated the virus from tumors in three laboratory mice. Then the two scientists conducted experiments showing that fluid from the mouse tumors caused new malignant tumors not just when injected into other mice but also when injected into hamsters and rats.9 The notion that viruses might cause cancer was resurfacing and gaining traction, and their discovery got the attention of other scientists in a big way.
But beginning shortly before the Time article was published that July, Eddy launched another experiment—one born of a nagging worry that her polyoma studies had prompted and that now wouldn’t leave her alone.
Her concern was this: if virus-bearing fluid from a mouse tumor could so easily cross species lines to cause cancer in hamsters and rats, why couldn’t a virus from, say, a monkey—a species much closer to Homo sapiens—cause cancer in humans? Her question wasn’t academic. Salk’s polio vaccine had been injected into more than 69 million Americans since its launch in 1955. The vaccine was grown in monkey kidney cells. And Eddy well knew, as did any vaccinologist who worked with them, that those kidney cells harbored plenty of viruses—“simian” viruses, in scientific parlance—from simia, the Latin word for ape.
These were viruses that lurked in apparently healthy monkeys, and especially in their kidneys. However normal the animals seemed, the viruses regularly killed their kidney cells in the lab, forcing scientists to jettison spoiled cultures. The Wistar’s chief Koprowski, who had a keen interest in the matter because he was developing his own polio vaccine, noted that the viruses “are more often than not dormant in the intact [monkey], but go on a rampage when infected tissues are removed soon after the animal’s death.”10
A scientist at the drug company Eli Lilly, Robert Hull, had begun cataloging each new simian virus that was discovered, classifying it according to the cellular damage that it caused in monkey kidney cell cultures. In 1958 he reported that eighteen new simian viruses had been discovered in just the previous two years. “As long as primary monkey kidney cultures are used in the production and testing of virus vaccines,” his paper concluded, “the problem of simian virus contamination will remain.”fn1 11
The assumption about these simian viruses, made by everyone from Salk to the NIH’s top vaccine overseers, was that while they might be an annoyance in the lab or on the production line, they were not a danger to human beings, because they were killed by the same formaldehyde that killed the polio virus in the Salk vaccine. What was more, the reasoning went, even if simian viruses occasionally managed to survive the manufacturing process, they were clearly innocuous in humans: weren’t there tens of millions of healthy Salk vaccinees walking the streets to prove the point?
Bernice Eddy couldn’t so casually accept those assumptions. In June of 1959, without her NIH boss’s knowledge, she launched a bold experiment. She took monkey kidney cultures prepared by the DBS—these were from rhesus monkeys, the species widely used in polio vaccine making, which will be important. She froze them, ground them up, put them through a very fine filter that strained out bacteria but not viruses, and injected a fraction of an ounce of the resulting fluid under the skin of newborn hamsters. Fully 70 percent of the 154 animals that she injected developed tumors, and every animal that did so died. What was worse, it wasn’t just one or two of the ground-up cultures that were at fault: she had prepared twelve lots of ground-up monkey kidney cells, each derived from between eight and thirty-two monkeys. Nine of the twelve turned out to be cancer causing. The virus appeared to be common in the monkeys.12
Eddy then went a step further, taking tumors from two of the sick hamsters, mincing them finely with scissors, and injecting tiny bits of tumor under the skin of forty more newborn hamsters. All but two of those hamsters got cancer and were dead in less than three months.13
It was by now the early summer of 1960, and Eddy was anxiously pondering her findings when she got word of a speech that had been given at a major polio vaccine conference in nearby Washington, DC. In the world of polio vaccinologists and public-health people, the talk, by Maurice Hilleman—a highly respected, tough-talking Montanan who headed vaccine research at Merck—landed with the explosive force of a hand grenade. It soon caused Eddy to gather her courage and approach her boss with her findings.
The leading polio vaccinologists who convened at the conference at Georgetown University that June were buzzing with optimism. The killed-virus Salk vaccine had already been on the market for five years, and the incidence of polio in the United States had fallen dramatically, from nearly 25 cases per 100,000 people in 1955 to fewer than 5 per 100,000 people in 1960.14 But pockets of polio stubbornly persisted, particularly in poor communities. Still more worrying was the fact that the worst cases—the ones that paralyzed people—had not declined at as steep a rate as cases that sickened people but from which they recovered fully. Alarmingly, cases of paralytic polio more than doubled between 1957 and 1959 to more than 6,000. It seemed clear that the Salk vaccine was going to curtail polio but not conquer it. And so the assembled virologists were eagerly awaiting the licensing of a second solution—the first live polio vaccine—a vaccine that, they anticipated, would be cheaper, easier to administer, and more effective than Salk’s. It would be swallowed rather than injected, and so would mimic the natural route of polio infection. It would generate robust levels of antibodies in the lining of the digestive tract, where the virus first encounters the human immune system, as well as in the blood—unlike Salk’s killed-virus vaccine, which was injected into muscle.
The scientists at the conference knew that they were witnessing the home stretch of an intense three-way race to win a U.S. license for a live vaccine. The high-powered, sharp-tongued Albert Sabin was clearly in the lead. But Herald Cox, Koprowski’s former boss at Lederle, had taken the vaccine that Koprowski developed while at the company and continued tweaking and testing it; he was fighting hard to stay in the race. And Koprowski himself had refused to give up on beating his archenemy, Sabin. Since leaving Lederle, the Wistar czar had adapted and renamed the vaccine he had brought with him from Lederle and continued to test it, vaccinating more than 300,000 people in central Africa, mainly in the Belgian Congo.
But what the audience heard from Hilleman at that June 1960 conference put a damper on the buzz about the hoped-for licensure of a live vaccine. Hilleman shocked them with the news of an unexpected finding by Ben Sweet, a scientist he supervised at Merck’s West Point, Pennsylvania, campus. Sweet had discovered a new, invisible simian virus contaminating Sabin’s live vaccine and almost certainly Cox’s and Koprowski’s live vaccines too. All three vaccines were made using monkey kidney cells. And unlike Salk’s killed vaccine, none of the live vaccines were treated with the formaldehyde that was presumed to kill simian viruses.
This new simian virus differed in an important way from the dozens of others that were now known to infect the monkey kidneys used by polio vaccine makers. The new virus didn’t declare itself by damaging cultures of the cells. It didn’t “go on a rampage,” as Koprowski put it. Instead it sat quietly, leaving the monkey cells looking and acting perfectly normally in their bottles. It was impossible to detect, making it impossible for vaccine makers to jettison infected cultures. But it was there nonetheless.
How, then, did Sweet stumble on the new virus? A detailed rendering of his discovery comes from The Virus and the Vaccine by Debbie Bookchin and Jim Schumacher.15 The authors recount that Sweet happened to be working with a different species of monkey from the species whose kidneys were normally used to make polio vaccine. When he exposed kidney cells from this different species, the African green (or grivet) monkey, to fluids from rhesus monkey kidney cells—the cells used to make polio vaccine—the African green kidney cells sickened and died. It emerged that the rhesus cells had passed to the African green cells a virus—a virus that lay silent and invisible in rhesus monkey cells but caused major damage in the kidney cells of African greens, bloating the cells and riddling them with holes.16 (The virus was also discovered to live silently in the kidney cells of another monkey species also used in polio vaccine preparation: cynomolgus monkeys.)
If there was any good news in the Merck scientists’ findings, it was this: Sweet had found that the silent simian virus was inactivated by the same formaldehyde that was used to kill the polio virus in the Salk vaccine. So while the virus may have been injected into the arms of some of the seventy million Americans who had received the Salk vaccine, it had also, it seemed safe to say, been dead on arrival.
The remaining days of the conference buzzed with talk of the new virus, which Hilleman dubbed the “vacuolating virus” because of the holes, or vacuoles, it made in the cells, leaving them looking like Swiss cheese. Formally the new, silent virus was named SV40. News of it soon reached Eddy at the nearby NIH.
Eddy didn’t know what it was in the rhesus monkey kidney cells that had caused her hamsters’ tumors. She hadn’t had the tools on hand to identify whatever it was, so she had simply called it a “substance.” But when she heard of Hilleman’s vacuolating virus, she immediately suspected strongly that her “substance” and the new, silent virus, SV40, were one and the same thing.
On July 6, 1960, one month after Hilleman’s much-noted talk at the Washington conference, Eddy sent a memo to her boss, Joseph Smadel, who was newly in charge of vaccine safety testing at the DBS. She titled the memo “The presence of a [cancer-causing] substance or virus in monkey kidney cell cultures.”
Eddy wrote to Smadel that she had heard about Hilleman’s SV40 virus—and that she herself had inoculated newborn hamsters with “specially prepared monkey kidney cells.” Tumors occurred at the injection sites, she wrote. “Eventually, the animals die.” She hoped, she added, to do some follow-on experiments as quickly as possible, to see if the “substance” causing her hamsters’ tumors was in fact SV40.17
Smadel was a man’s man, a foul-mouthed, no-nonsense, dictatorial virologist in his early fifties who had staffed an advanced World War II field laboratory in France after D-Day and later, working in Malaysia, discovered that an antibiotic, chloramphenicol, could effectively treat typhus and typhoid fever. Smadel had come to the NIH in 1956 on the heels of the disastrous Cutter incident. He had been a key advocate of the Salk vaccine, and he was acutely aware of the damage that another round of Cutter-like bad press could do to public adoption of any polio vaccine. So as Smadel read Eddy’s note, he got angry. He summoned her and tore a strip off her in language he later acknowledged was “not even diplomatic.” He dismissed the tumors in Eddy’s hamsters as “lumps,” called her data “inadequate,” and shut down her “entirely unwarranted” suggestion that they might be related to SV40 or have implications for human cancer.18
Eddy wasn’t fazed. She went back to her lab determined to identify the “substance” that had caused the cancers in her hamsters. It was completely conceivable that her “substance” had been injected, alive, into the millions of people in other countries and the roughly 10,000 in the United States who had participated in Cox’s, Koprowski’s, and Sabin’s trials of live polio vaccine. And it was also possible, she was convinced, that it had been injected—and was still being injected, in the Salk vaccine—into the arms of tens of millions of U.S. school children.19
Hayflick recalls learning about the discovery of the SV40 contamination by Sweet at Merck even before Hilleman’s bombshell speech at the Washington conference in June of 1960. Hayflick was a friend of Sweet’s, and the circle of virologists in the Philadelphia area was a close one in which news was freely exchanged and traveled quickly. Similarly, Eddy’s findings soon made their way to Hayflick through the scientific grapevine, as bad news always does.
Hayflick immediately grasped the headache that SV40 posed for polio vaccine makers and regulators. Admittedly, Hilleman and Sweet had argued that it wasn’t an insurmountable obstacle. They had tested the blood of people who had swallowed live polio vaccines in trials and found no antibodies to the simian virus, strongly suggesting that SV40 did not proliferate massively in the human intestinal tract and from there invade the body. Still, they conceded, the possibility could not definitively be ruled out. Nor could the long-term possibility that, having invaded the body, such a virus could eventually cause cancer, “especially when administered to babies” with their less-developed immune systems. Going forward, the pair wrote in the paper that summarized their SV40 findings, “the simple solution” would be to ensure that the live polio virus seed stocks—the stocks that were amplified by companies to make production-scale quantities of vaccine—were not contaminated with SV40, and to throw away vaccine lots that were already tainted.20
The problem, Hayflick was convinced, was not so simply solved. For one thing, the testing of the live vaccine to ensure that it was free of SV40—testing that would clearly be needed going forward—would be time-consuming and expensive. And even if no ill ever came to those already vaccinated, Hilleman and Sweet themselves had admitted that “other undetected [simian] viruses might also await demonstration.”21 Surely there was a better answer than passively waiting to stumble on the next unwelcome “passenger” virus in monkey kidney cells.
In the summer of 1960 Hayflick was still fully two years away from launching WI-38 from the lungs of Mrs. X’s fetus. As news of the discovery of SV40 landed, Hayflick was still working with the first twenty-five fetal cell lines, coming to realize that they aged and died in their dishes, with all that that portended. He was also finding them, to all appearances, free of lurking, unwanted viruses. At the same time, he was discovering that they could be infected with many other disease-causing viruses—including polio. And Moorhead, bent over his microscope, was reporting that their chromosomes were reassuringly normal. The obvious thing to do fairly shouted at Hayflick. He would make a polio vaccine using his normal human diploid cells. Then he would see if it worked.
Down the hall from Hayflick’s lab and across the atrium, in the other wing of the V-shaped Wistar Institute, another brain was buzzing with the vaccine-making implications of the new human cells. Koprowski had been front and center at the June conference in Washington, DC, where the sudden specter of SV40 took top billing. There he had given a speech downplaying SV40’s significance, and that of any other monkey virus that might be found contaminating live polio vaccine.
Koprowski pointed out that in the last several years of trials, millions of people around the world had already received various live poliovirus vaccines made using monkey kidney cells—his vaccine and Sabin’s and Cox’s—to no obvious detrimental effect. The discovery that SV40 had been lurking in those vaccines, he argued, “should hardly deter anybody from accepting the product.” Silent viruses inhabited all kinds of animals cells, he argued, and doubtless would be found, for example, in the calf lymph that had been successfully used to make smallpox vaccine for two hundred years. Should we therefore stop vaccinating against smallpox?
He conceded that it would be desirable, going forward, to try to rid monkey cells of extraneous viruses, and he mentioned several possible approaches, including using—he didn’t mention Hayflick or the Wistar by name—some newly available human cells.22
As he minimized the risks of SV40, Koprowski was still holding out the ambitious hope that the U.S. surgeon general—Leroy Burney, whom Koprowski had been sure to invite to the Wistar’s gala opening symposium the previous year—was going to pronounce his live polio vaccine the favorite child of the U.S. government; the vaccine that would be chosen to move through licensing, leaving its two competitors in the dust.
Burney was under pressure to make a choice, soon, among the three vaccine candidates, not least because the United States was still exclusively using Salk’s killed, injected vaccine while the country’s bitter cold war rival, the USSR, had already fed eighty million people Sabin’s purportedly superior, oral, live vaccine. Not only that, but there had been sporadic new outbreaks of polio in the United States in 1958 and 1959. These outbreaks had—fairly or unfairly, for most cases occurred in undervaccinated populations—undermined confidence in the Salk vaccine.
On August 24, 1960, Burney told a press conference that he had picked a winner: Sabin’s vaccine had bested its two competitors in safety tests in monkeys, and he had chosen it to proceed through licensing. Any live vaccine licensed in the future would have to be as good as Sabin’s, or better. Within hours, three major vaccine manufacturers announced that they would begin making the Sabin vaccine, and others appeared ready to follow suit. Koprowski’s vaccine was done, to all appearances.
But for the ever-enterprising Koprowski, Hayflick’s new human cells meant that the game was not necessarily over. In late October, Koprowski—joined by a twenty-eight-year-old Wistar physician/scientist, Stanley Plotkin, who had helped run trials of Koprowski’s monkey cell–based polio vaccine—sent off a long letter of advice to a World Health Organization committee. The group was soliciting input on the standards that live polio vaccines should be held to.
Koprowski had minimized the SV40 monkey virus problem only four months earlier, when his own monkey kidney–based polio vaccine was still in the running for U.S. approval. Now, with Sabin’s vaccine rolling quickly toward being licensed, he sounded more alarmed. The letter noted that monkey kidneys are riddled with simian viruses. Eliminating SV40 from the live vaccine, the writers observed, “may present insurmountable obstacles.”23 And yet hundreds of thousands of fresh monkey kidneys would continue to be needed to make live polio vaccine, increasing the chances of a potentially cancer-causing virus finding its way into some lots of vaccine and “making the case even weaker for the use of such a tissue.”
By contrast, Koprowski and Plotkin wrote, cells were now available from normal human fetuses—cells that had normal numbers of chromosomes. They could be frozen and used for vaccine production when needed. And each line of such cells could be “scrupulously investigated” for hidden viruses. Koprowski and Plotkin laid out the pros and cons of monkey-kidney versus human cells in an accompanying table. Its columns had titles like “Possibility of freeing poliovirus from simian viruses” (with the human cells it was “possible”; with the monkey kidney cells it was “impossible”) and “Procurement of tissue” (with the monkey kidney cells it was “difficult”; with the human cells it was “easy”).24
Koprowski sent the letter off to Geneva. Then he waited for Hayflick to show just how a polio vaccine could be made—using the human fetal cells and not Sabin’s but Koprowski’s polio vaccine.
Hayflick first grew a small amount of Koprowski’s polio vaccine virus in petri-dish cultures of his normal WI-1 cells.fn2
Then he inoculated fluid from these petri-dish cultures into bigger quantities of WI-1 cells in quart-sized bottles. Koprowski’s polio vaccine virus destroyed the cells within two days, first causing the long, tapering fibroblasts to become round and then causing them to burst, or “lysing” them. Each lysed cell ejected up to ten thousand new virus particles into the nutrient fluid that bathed it.
Next Hayflick inoculated that fluid into still more WI-1 cells, this time in still bigger, gallon-plus bottles. These he left to incubate at body temperature. Five days later he filtered the resulting soup of culture medium, cellular debris, and live viruses to strain out bacteria. He froze it at –70 degrees Celsius. He had just produced the first vaccine grown on his human diploid cells.25 (It was not the first vaccine ever grown in human cells; Sven Gard’s group in Sweden had done that when they made that fleeting polio vaccine using fetal skin and muscle cells in the mid-1950s—the vaccine that was impracticable because they couldn’t make enough of it.)
As Hayflick prepared the vaccine, another stunning piece of news arrived about the silent simian virus, SV40. The previous summer, vaccine makers and regulators alike had taken heart from the finding by Sweet and Hilleman at Merck that SV40 was killed by the same formaldehyde used to kill the polio virus in the Salk vaccine.
But in March 1961 British researchers reported in the Lancet, a widely read medical journal, their discovery that, contrary to the Merck scientists’ report, SV40 was resistant to formaldehyde: it was not inactivated by the chemical as quickly as poliovirus was. That meant that the simian virus could survive sometimes, alive, in Salk’s vaccine. They argued for “an accumulating body of evidence that killed poliomyelitis vaccine in the past has contained [SV40], probably in the living state.”26
A few months later researchers from the Medical Research Council Laboratories in London reported in the British Medical Journal that they had found antibodies against the SV40 virus in eleven of twelve schoolboys who had received the full slate of three Salk injections.27 The presence of antibodies did not mean that the virus was alive and replicating in the boys—the whole principle of injecting a killed vaccine is that the body forms antibodies even to a dead virus. The boys might well have received dead SV40 with their Salk injections. But the presence of SV40 antibodies in eleven of the twelve boys was an alarming indication of just how widely SV40 might have infiltrated the supply of Salk vaccine. And the more widespread it was, the more likely that some vaccinees had received injections in which the virus had survived alive.
The same month, March 1961, that the Lancet article appeared, a House of Representatives subcommittee held hearings on production of the live vaccine. Koprowski had a conflict and couldn’t testify, but he wrote to the lawmakers, making the same argument that he had made to the World Health Organization a few months earlier. Given the ubiquity of the silent SV40 virus in the monkey cells used to make the live vaccine, it would be impossible to produce it in a cost-effective manner. It would be both cheaper and sounder scientifically to switch to human cell strains. There was, he added, “not a shred of evidence” that they caused cancer.28
This was the backdrop against which Hayflick pushed ahead with his human fetal-cell–produced polio vaccine. He finished making the vaccine in or about January 1961. Now, as the SV40 problem emerged as more widespread than anyone had first understood, Hayflick and Plotkin set about ensuring that the new vaccine would be safe to inject into the most fragile of human beings: newborn babies.
In the late 1950s there had been a bump in polio cases in babies aged six to twelve months old—a vulnerable period when protective antibodies inherited from a baby’s mother have waned away but the baby’s immune system is still immature and can’t mount a full-fledged attack on foreign invaders. Koprowski and Plotkin argued in a 1959 paper that this meant that babies should be vaccinated as early in life as possible. The paper reported their success in a trial of Koprowski’s live polio vaccine, which was then still a contender for licensure; it had boosted antibody levels in babies as young as one day old.29
Importantly, babies also provided a source of unvaccinated trial subjects in which to test Hayflick’s new, live vaccine. Such subjects were otherwise tough to find: by 1961 an estimated 90 percent of the nation’s children and adolescents had received the Salk vaccine, as had 60 percent of adults younger than forty years old.30 Many among the other 40 percent likely had antipolio antibodies already, having been exposed to polio during decades of living. The presence of preexisting antibodies would make the results of a live vaccine test difficult to interpret.
Thanks to Koprowski’s connections, the Wistar researchers had a ready-made source of newborns to vaccinate, at an unusual women’s prison headed by an equally unusual woman.
Clinton State Farms was located in rural New Jersey, sixty miles northeast of Philadelphia. The campuslike institution, which housed women serving prison sentences of more than one year, was run by Warden Edna Mahan. She had been in charge of the prison for thirty-three years.
Mahan, then sixty, was a 1922 graduate of the University of California at Berkeley. She had a warm smile, an aquiline nose, and large, light, penetrating eyes. She was a 1960s-style liberal well before that decade arrived. An ardent advocate of rehabilitation, she banished handcuffs and allowed the several hundred inmates to earn trust by steps. They wore color-coded uniforms reflecting their degree of freedom, and the best-behaved were allowed to work in the surrounding community by day, at jobs as farmhands and domestics.
“The atmosphere at Clinton Farms is not that of a prison. No girl is locked in,” former First Lady Eleanor Roosevelt wrote after attending an eighth-grade graduation ceremony for inmates in 1956.31 Roosevelt remarked that the prisoners actually cheered when Mahan’s name was mentioned and noted that up to four hundred girls were allowed to picnic on a hillside with a single attendant accompanying them.
That same freedom made an impression on Koprowski, who reported in interviews late in his life that the Clinton inmates, being red-blooded women, took advantage of their freedom by hailing eighteen-wheelers on nearby Route 78 and enjoying moonlit trysts in the truck cabs—a claim that makes for a good story about how babies came to be born at the prison but for which there is no evidence but Koprowski’s word.32 What isn’t in dispute is that about sixty infants were born to inmates every year, either in the prison’s own Stevens Hospital or at nearby Hunterdon County Medical Center. In 1960 there were fifty-four births at the prison.33
At the time that Hayflick produced his polio vaccine, Koprowski had already been vaccinating newborns at Clinton Farms with experimental monkey cell–produced polio vaccines for five years. The door to the prison opened for him because a colleague, a University of Pennsylvania doctor named Joseph Stokes, socialized with a member of the Clinton Farms Board, and because Stokes’s brother, Emelen, also a physician at Penn, sat on New Jersey’s Board of Control of Institutions and Agencies. It didn’t hurt that the strong-willed Mahan herself was a big believer in medical research. In the first experiment she approved, in 1946, inmates were infected with body lice daily to gauge the effect of their nutritional status on the pace at which lice bred on their bodies.34
Plotkin, who did much of the actual vaccinating for Koprowski in the late 1950s, recalls making the drive to the prison as often as twice a week in that era. The prison babies had the great advantage that they didn’t disappear into the surrounding community within a week, as newborns did from city obstetrical wards. It was not unusual for babies to remain in the nursery at Clinton Farms for four to six months, allowing for follow-up testing to measure antibody levels in response to experimental vaccination.
Whether the inmates who were new mothers felt able to refuse to participate is an open question. “Dr. Agnes Flack, Medical Director, and Miss Edna Mahan, Superintendent, of Clinton State Farms, were extremely helpful in obtaining permission for vaccination of the infants,” Koprowski and Plotkin wrote in the acknowledgments of one 1959 paper.35
There may have been a strong incentive for the prisoners to volunteer their infants for the studies: the chance to spend more time with their babies. In his 1999 book The River, writer Edward Hooper examined Clinton Farms birth records from the mid-1950s and determined that babies enrolled in Koprowski’s polio vaccine trials stayed an average of four to six months at the prison, a figure also reported by Koprowski and his colleagues in the 1959 paper.36 Hooper found that those not in trials were placed with a social welfare agency or with relatives after about four to six weeks.37
Before taking the new live polio vaccine made in fetal cells to Clinton Farms, Plotkin and Hayflick ran what they would describe in the resulting paper as “exhaustive” safety tests on it. First they needed to ensure that the vaccine wasn’t contaminated with some virus or bacterium besides polio—a virus or bacterium that could theoretically be lurking in the WI-1 cells. So they injected the vaccine into dozens of mice, rabbits, guinea pigs, and hamsters, then watched them for symptoms of illnesses caused by microbes known to infect these species, like herpes simplex and the Bacillus that causes tuberculosis. None got ill. This test, they would later write, “presumably” ruled out the presence in the vaccine of such microbes.38
Next they neutralized the vaccine virus with antipolio antibodies and injected the resulting fluid into plates of monkey cells. If a hidden nonpolio virus was lurking in the vaccine, it would damage the kidney cells. Hayflick and Plotkin were thinking of SV40, and also of a lethal herpes virus known as B virus that occasionally killed researchers and animal handlers who were bitten by monkeys. But the plates of monkey cells remained healthy. (To be sure that SV40’s effects would be noticed, they used kidney cells from African green monkeys.)
Then they ran several tests to make sure that the genome of the vaccine virus hadn’t mutated to a different, dangerous form when it was grown in WI-1 cells. Their tests indicated that it hadn’t. This was decades before the advent of gene-sequencing techniques, so they had to infer the results from indirect laboratory tests. One test involved injecting the vaccine virus into the brains of five monkeys, then observing them for twenty-one days before euthanizing them and looking at slices of their brains and spinal cords under the microscope. The monkeys didn’t get polio, and their brains appeared normal.39
After weeks of testing the vaccine, Plotkin and Hayflick felt that they had done all that they could do. Hayflick recalled in a 2014 interview being so confident of the new vaccine’s safety that he at one point fed it to his own children, Joel, Deborah, and Susan, then aged four, three, and two years old respectively. He did not recall whether this occurred before or after the first trial of the vaccine at Clinton Farms.40
The first baby at Clinton Farms swallowed the human cell–based polio vaccine in the late spring or early summer of 1961. The vaccine was administered either by Plotkin or by Suzanne Richardson, a nurse assistant who often helped him.
The timing was apt. Bernice Eddy’s paper about the lethal hamster tumors caused by a “substance” in rhesus monkey kidney cells had finally been published in May, after being held up for months by her boss, Smadel. Then, in late June, Koprowski used the high-profile annual meeting of the American Medical Association in New York City to put the SV40 issue on the radar of practicing physicians for the first time. He warned them that the “obsolete” method of making polio vaccine with monkey kidney cells risked more “virus surprises.” He called Hayflick and Moorhead’s human cells “the obvious choice” for making polio vaccine going forward.41
Not long after, the lay press picked up on SV40. Late in July, the Associated Press, in a story that appeared on page thirty-three of the New York Times, reported that both Merck and Parke-Davis had stopped making the Salk vaccine because it had been found to contain a monkey virus “believed harmless” by the NIH. The story did not mention cancer.42 Nor did it report that Hilleman, at Merck, had insisted that the company drop its production of the Salk vaccine when the company’s tests found live SV40 in its vaccine.43
The National Enquirer wasn’t as sanguine. Despite its reputation for hyperbole, the tabloid that August ran an accurate, thorough story under the headline THE GREAT POLIO VACCINE COVER-UP. “The polio shots you have taken may KILL you,” it began, beside a subheadline that screamed 70% OF HAMSTERS DEVELOPED CANCER IN LAB TESTS. The paper reported clearly on Eddy’s findings and quoted Koprowski, again promoting Hayflick’s cells and suggesting that companies were clinging to monkey cells only due to fear of change.44
Days before the Enquirer story appeared, the Division of Biologics Standards at the NIH began for the first time to require vaccine makers to sample their Salk polio vaccine lots to ensure that they were free of live SV40 before sending them out. The change was made eighteen months after the letter to the editor in the Lancet first reported that live SV40 could survive in the Salk vaccine. The DBS did not recall any of the Salk vaccine that was already on the market. Nor did it at this point require companies to make the expensive switch to using African green monkey kidneys for making polio vaccine. These monkeys don’t naturally harbor SV40, so using kidneys from this species would have obviated the contamination problem.
By the end of the summer of 1961, six full-term Clinton Farms infants, aged between nine and fifty-seven days, had been vaccinated with Hayflick’s human-cell-propagated vaccine.45 That October Hayflick sent the resulting paper, with himself as first author—the others were Plotkin, Koprowski, and Koprowski’s administrative deputy and lab manager, Tom Norton—to the American Journal of Hygiene, the premier academic journal of the day for public-health research.
Their findings were encouraging, the authors reported in the paper, which was published in March 1962.46 The polio vaccine that was made using human fetal cells was free of extraneous, nonpolio viruses that critics might contend could have been living in the WI-1 cells in which it was produced. Five out of the six vaccinated babies excreted the virus in their feces for more than one week, a sign that the virus had established an active infection in their intestines. Such an infection was needed to provoke an antibody response. What was more, the vaccine appeared genetically stable—tests of virus from the feces of the five babies showed that it hadn’t morphed into something more toxic during its trip through the infants’ bowels.
An earlier draft of the paper, written before most of the babies had been vaccinated, conceded that all of the testing in the lab and in animals wasn’t proof positive of the vaccine’s safety. “It is possible,” the draft read, “that the growth of a [weakened] poliovirus in human cells may increase its [ability to cause illness in humans] or change other of its characteristics. Only a large scale field trial, which we are presently organizing, can determine whether or not pathogenicity for man is increased.”47
By the time the paper was published, however, the wording had been changed. “Our results with a [weakened] poliovirus vaccine grown in human cells indicates [sic] that the virus does not increase its pathogenicity for the human, or change any other of its characteristics,” the paper stated. “We are presently organizing a large-scale field trial to further demonstrate the safety and efficacy of this material.”48
The authors ended the published paper by singing the virtues of all of Hayflick’s human diploid cell strains for making antiviral vaccines. These were superior, they wrote, in a dozen ways to the monkey kidney cells still being used to make both the Salk and the Sabin vaccines, and there was theoretically no end to the vaccines that could be made with them, against diseases as diverse as rabies, measles, chicken pox, and even the common cold.
And what became of SV40, and Eddy? Her dressing-down from her boss, Smadel, in the summer of 1960 didn’t cow her. In fact, within weeks she used the occasion of a meeting of the New York Cancer Society to present her findings that a “substance” in rhesus monkey kidneys caused hamster-killing tumors. She hadn’t cleared the content of her presentation with Smadel.
“I knew when I was doing it I’d be in trouble,” Eddy recalled late in life. “And I didn’t care much.”49
Soon thereafter, Roderick Murray, the chief of the NIH’s Division of Biologics Standards, where Eddy worked, told her in a memo that she was being freed of her “irksome” responsibilities trying to ensure vaccine safety.50 Instead she would begin conducting her own independent research. In a separate memo Smadel let her know that she wouldn’t speak at any more meetings without his reviewing and approving her remarks first.51
In the summer of 1961, shortly after her paper documenting the hamster tumors appeared, Eddy was downsized to two assistants and assigned to room 207 in building 29 at the NIH, which had until then been a supply room.52 It measured about sixteen by fourteen feet. Ruth Kirschstein, who was a young DBS scientist during Eddy’s demotion and who later rose to become the NIH’s deputy director, would, late in her own life, describe the DBS’s failure—including her own—to take Eddy’s findings seriously as “not a terribly pretty story.”53
Hilleman, whose Merck lab identified SV40 and who ended the company’s production of the Salk vaccine, recalled years later that Eddy’s NIH bosses “tore the hell out of her” for her findings because her studies failed to use rigorous controls. “But,” he added, “she was right.”54
Eddy’s next paper, on which she labored for a year despite the punishing atmosphere, would demonstrate that SV40 and the monkey kidney “substance” that caused tumors in her hamsters were one and the same thing. It was published in May 1962 after Smadel again held it up for several months before allowing Eddy to submit it.55
The following month Koprowski rushed into print a paper by Wistar researchers and the surgeon Robert Ravdin showing for the first time that the silent monkey virus affected human cells—cells scraped from people’s skin and the insides of their cheeks. The cells became abnormally shaped, divided faster, piled up on top of one another, and outlived their uninfected counterparts. The paper featured photos of the cells’ disrupted, abnormal chromosomes at 2,500 magnifications.56 A short time later the Nobelist John Enders and his colleagues Harvey Shein and Jeana Levinthal published similar findings in the Proceedings of the National Academy of Sciences. The trio had found that SV40 caused malignant changes in kidney cells from human fetuses, newborns, and three-month-olds.57
Nine months later, in March 1963, the NIH’s Division of Biologics Standards began requiring that all polio vaccine be free of SV40 before its inactivation by formaldehyde, rather than allowing companies to sample vaccine once it was inactivated and ready for market. This in effect forced manufacturers to switch to using the kidneys of African green monkeys, which don’t naturally harbor SV40, rather than continuing to use kidneys from SV40’s natural hosts: rhesus and cynomolgus monkeys.
By this time 98 million Americans had received the Salk vaccine since its launch in 1955. Another 10,000 had potentially been exposed to SV40 when they participated in trials of live polio vaccines between 1959 and 1961. In addition 100,000 members of the military were potentially exposed to the silent monkey virus when they were injected with adenovirus vaccine between 1955 and 1961 to protect them from the respiratory infection, which travels easily in the close confines of military barracks.58