The Vaccine Race

by Meredith Wadman

  After all, the DBS-approved rubella vaccines, developed in the cells of African green monkeys and then grown in duck-embryo cells by Merck—the company branded its vaccine Meruvax—and in rabbit kidney cells by Smith, Kline & French, as Cendehill, had been on the U.S. market now for more than a year. Children were being routinely immunized by their pediatricians, and women who wanted the vaccine were getting it—usually the Cendehill vaccine in the case of grown women, because it seemed to cause fewer cases of arthritis than did Merck’s vaccine. (The vaccine made in puppy kidney cells by Philips Roxane of St. Joseph, Missouri, was by this point being rejected because of the bad side effects it caused.) The vaccines seemed to be doing their job: the feared rubella epidemic of 1970 had not materialized.

  But among the few Americans who cared about the fate of Plotkin’s RA 27/3 rubella vaccine was an exceedingly bright, determined woman named Dorothy Horstmann, who had been paying keen attention to the studies of the various rubella vaccines being published in medical journals—and who rarely, if ever, took no for an answer.

  In the autumn of 1970 Horstmann was a fifty-nine-year-old pediatrician and vaccine scientist at Yale Medical School in New Haven, Connecticut. She had watched closely the studies that Plotkin had published over the preceding five years. She had read, and continued to read, the papers on the other rubella vaccines. She had coauthored some of her own studies on them. And she was becoming convinced of the superiority of Plotkin’s RA 27/3 vaccine. Like Plotkin, Horstmann resembled a dog with a bone when she got onto a cause. Unlike him, she had a very high profile and a long and distinguished track record in vaccinology. She was also known as a woman of unimpeachable integrity.

  Horstmann was born in Spokane, Washington, in 1911 and grew up in San Francisco, where as a girl she shadowed a physician who was a family friend while he saw patients at the local hospital. By 1940 she had earned an undergraduate degree from the University of California at Berkeley and an MD from the University of California at San Francisco. Later, at Yale, she completed a specialization in pediatrics, and in 1961 she became the first woman professor at the medical school in New Haven. She had by then played a crucial role in the development of polio vaccines by discovering that polio invaded the central nervous system by way of the bloodstream.2

  In September 1970, the same month that Smith, Kline & French dropped Plotkin’s vaccine, Horstmann’s direct gaze—accented by a brunette bob and a toothy grin—stared out at readers from the pages of the New York Times under the headline NEW RESEARCH ON RUBELLA CHALLENGES EFFECTIVENESS OF VACCINATION PROGRAM. Recent studies, the article explained, indicated “that many vaccinated persons—perhaps more than half of them—can become reinfected with rubella virus and that the vaccine might not give life-long protection as originally thought.”3

  Horstmann’s was one of those recent studies, and it was days away from being published in the New England Journal of Medicine.4 Its findings were alarming. With colleagues from Yale and the University of Hawaii, Horstmann had set out to discover how well vaccination actually protected vaccinees when they were placed in the midst of a natural rubella epidemic. She knew where to find one: every year a group of military recruits known as the all-Hawaiian company arrived at the U.S. Army’s Fort Ord, California, for training. And every year at Fort Ord there was a rubella outbreak after these young Hawaiians—who were known to be particularly vulnerable to rubella because of their geographical isolation—arrived and began living in close quarters with others in a state where the virus circulated freely.5

  So in a study begun in the spring of 1969, the Horstmann team vaccinated the Hawaiian recruits with Smith, Kline & French’s Cendehill vaccine two to three months ahead of their departure for Fort Ord. During the outbreak that then ensued at the fort, 80 percent of the vaccinated recruits were reinfected with rubella. This is not to say that they broke out in rashes and nursed fevers, although on rare occasions, in other studies of reinfected vaccinees, this did occur.6 Rather, spikes in their antibody levels showed that the virus had reentered their bodies and multiplied silently.7 For the reinfected recruits this hardly mattered. But for pregnant women in contact with them, it might be positively dangerous.

  The problem was not that the vaccinations given in Hawaii had failed to stimulate the men’s immune systems to develop antibodies. In fact, the men had all developed decent levels of antirubella antibodies after being vaccinated and before shipping out to Fort Ord. Nor was it because the California microbe was some particularly aggressive strain of the virus—a group of men at Fort Ord who were naturally immune to rubella because they had been infected in childhood easily fought it off. Blood tests showed that just 3.4 percent of those men who were immune because of childhood bouts with the disease were reinfected.8

  The message was that the antibodies generated by the vaccine—as opposed to those generated by natural infection—were somehow failing to prevent reinfection by naturally occurring virus. They were in some way acting differently—they were weaker—than the antibodies generated by natural infection.

  Horstmann’s study joined others about rubella vaccine then being published in the New England Journal of Medicine, the American Journal of Diseases of Children, and the Journal of the American Medical Association.9 These studies found that people vaccinated with the Merck and Cendehill vaccines were being reinfected in the open community, in places like first-grade classrooms in Memphis; in other, nonmilitary institutions, like the residential Nazareth Childcare Center in Boston; and during clinical studies in which vaccinated people were deliberately exposed to the virus through nose drops.10The sum total of the work raised a question that Horstmann put bluntly in her New England Journal of Medicine paper: “If vaccinees can be readily reinfected a few months after successful immunization, what are the long-term prospects for durable protection of the young woman in the childbearing age who was successfully vaccinated at the age of six?”11

  Ten months before her paper was published, Horstmann had begun a quiet campaign to respond to that question—not with more studies on the vaccines already on the U.S. market but by presenting an alternative.

  In late 1969 or early 1970 Horstmann wrote to Plotkin asking for data on his RA 27/3 vaccine.12 Plotkin was only too happy to supply it and ended up sending Horstmann hundreds of ampules of his rubella vaccine, provided by the two companies that were in the process of winning market approvals for it abroad: the Institut Mérieux in Lyon, France, and Burroughs Wellcome in London.13

  Horstmann wanted to repeat her study of Hawaiian recruits when they landed at Fort Ord, this time using the RA 27/3 vaccine. She wanted to inject Plotkin’s vaccine in Hawaiian recruits who had never been exposed to rubella, watch them ship out to Fort Ord, and see if, once they were living in close quarters with other, stateside, recruits and were exposed to rubella for the first time, they got reinfected at the same high rates as had already been demonstrated in the recruits who had been vaccinated with the Cendehill vaccine. Her hunch was that they wouldn’t be. She suspected that Plotkin’s vaccine generated better immunity. But she needed to do the study to prove it.

  Then Horstmann hit a roadblock. She soon wrote to Plotkin that Colonel Edward Buescher, the deputy director, soon to be commandant, of the Walter Reed Army Institute of Research and as such the one in command of what studies could and could not be done on army recruits, was standing in the way of the new study of RA 27/3 vaccine in Hawaiian recruits “for reasons that do not make sense, either scientifically or administratively.”14 Horstmann couldn’t make out his “ulterior motives.”

  Not one to be turned aside, Horstmann proceeded to find another route to her goal. In December 1970 she launched a study of Plotkin’s vaccine among several hundred children in Danbury, Connecticut, a developing commuter town of fifty thousand outside New York City. She partnered with a local doctor, Martin Randolph, whom she described to Plotkin as the leading pediatrician in Danbury; with Warren Andiman, a young inf
ectious disease doctor at Yale; and with a Yale virologist and mother of five named Ann Schluederberg. The group vaccinated children in Randolph’s private practice, in a Head Start program, and at a day-care center in Danbury, comparing Plotkin’s vaccine, head to head, with the Merck and Cendehill vaccines. Then they watched the young vaccinees’ antibody levels over time. Their results would take fully eight years to generate and publish.15

  Horstmann also launched a smaller, quicker trial with Plotkin’s longtime collaborator, Ingalls, at the Boston School for the Deaf. The resulting paper, published early in 1972, showed that the antibody levels in children vaccinated with the Plotkin vaccine in 1968 hadn’t significantly declined two years after they were first vaccinated. What was more, the scientists had injected a “challenge” strain of disease-causing rubella into the noses of eighteen of the children who had been successfully vaccinated two years earlier. Just two of them—11 percent—showed a subsequent rise in antibody levels that indicated they had been reinfected.16 Eleven percent wasn’t trivial, but studies showed that even people who had natural rubella immunity from having had the disease in the past were reinfected at rates of 3 percent to 9 percent. And 11 percent certainly didn’t come close to the 40 percent to 100 percent reinfection rates that were being found with Merck’s duck-embryo vaccine and in Cendehill vaccinees.17

  Horstmann, as usual, was blunt in her conclusions. Yes, her numbers were small, and larger studies would be needed to confirm them. But the point was that to protect the fetuses of the future, it was crucial that any rubella vaccine deliver long-lasting immunity—and that that immunity be “qualitatively comparable” to the immunity that resulted from natural infection.18

  By “qualitatively comparable” Horstmann meant that a successful rubella vaccine needed to generate not only robust levels of antibodies to rubella virus but also the same types of antibodies that are made during natural infection. When the body’s immune system recognizes a foreign invader, it generates antibodies to numerous different pieces of that invader, called antigens. It also generates categories of antibodies that differ in how and where they attack the virus. Horstmann pointed to increasing evidence that showed that Plotkin’s vaccine mirrored the natural immune response in the kinds of antibodies it produced. This, she remarked, was in notable contrast to Merck’s HPV-77 duck vaccine and the Cendehill vaccine made by Smith, Kline & French using rabbit kidney cells.19

  Anyone reading the paper couldn’t miss Horstmann’s clear conviction that the RA 27/3 vaccine—Plotkin’s vaccine made in the WI-38 cells that Hayflick had launched a decade earlier—was simply better than anything then available in the United States.

  • • •

  In the early 1970s Horstmann began calling Maurice Hilleman, the vaccine chief at Merck. Regularly. Persistently. Always with the same message: Plotkin’s RA 27/3 vaccine was superior to Merck’s. The Plotkin vaccine was better because it mirrored natural rubella infection much more closely. It mirrored natural infection much more closely because it was far less weakened than the other rubella vaccines. It was less weakened because it had been passaged through cell cultures just twenty-five times. Not fifty-one times, like Cendehill. Not seventy-seven times, like the original Parkman-Meyer vaccine. Not eighty-two times, like Merck’s vaccine, which took Parkman-Meyer’s virus and sent it through five additional passages in duck-embryo cells.

  Hilleman took call after call from Horstmann. Finally he caved.

  “When she called me enough times, I said, ‘Dorothy, for Christ’s sake, you’re a pain in the ass,’” Hilleman told interviewer Paul Offit thirty years later. “I said, ‘I don’t know how much damage you’re doing in the scientific community, the medical community, with all of your promotion of RA 27/3. But I guess I have to think about it.’”20

  Hilleman may also have been influenced by a paper with Plotkin as the lead author that appeared in the Journal of the American Medical Association in August 1973. It began with Plotkin-esque self-deprecation.

  “Another [rubella] vaccine strain may seem like Marshal Ney at Waterloo, arriving after the battle is over,” he began. (Plotkin was soon informed that he had gotten his history wrong: it was the Marquis de Grouchy who arrived late at Waterloo.) He conceded that forty million American children had already been vaccinated with rubella vaccines that were, pointedly, not his. But then Plotkin laid out the facts that had accumulated about his vaccine and the others.

  People who received his vaccine generated numbers and kinds of antibodies that mimicked natural infection much more closely than the already-marketed vaccines. Similarly, people given his vaccine developed a resistance to reinfection that looked a lot like the strong resistance that occurred following natural infection with rubella. This was in stark contrast to people who received the Cendehill vaccine or Merck’s Meruvax, in whom reinfection occurred far more frequently than in RA 27/3 vaccinees. He offered one reason that this might be so: In some 40 percent of trial subjects who received jabs of his RA 27/3 rubella vaccine, the vaccine prompted antibodies to appear in the cells lining the inside of the nose and throat, where rubella first lands and multiplies before it invades the blood. Vaccines made using HPV-77 or the Cendehill virus did not produce any antibodies in the nose and throat.

  His rubella vaccine was now licensed, Plotkin wrote, in Israel, Ireland, New Zealand, and Yugoslavia, in addition to Great Britain, France, and a number of other countries. With trademark understatement, he noted, “There are . . . several reasons for dissatisfaction with the vaccines currently licensed in the United States.”21

  • • •

  For Plotkin 1972 was a year of big ups and downs. In May the patent on his rubella vaccine was granted, after four years and many rounds with the U.S. Patent and Trademark Office. Hilary Koprowski told him that, while the proceeds would flow to the Wistar, he was going, in turn, to direct 15 percent of them to Plotkin personally, as the inventor of the vaccine. Of course, that amounted to nothing in the United States, where Plotkin was doubtful that his vaccine would ever be licensed. But the Institut Mérieux in France and Burroughs Wellcome in the United Kingdom, having licensed the vaccine from the Wistar, were now selling it. There was going to be some significant money coming his way; that was clear.

  At the same time Plotkin’s marriage was falling apart. Three months after his fortieth birthday in August 1972, he decided to call it quits. He left Helen and with her his sons, Michael and Alec, aged ten and six. He moved into a high-rise near the Wistar and saw his sons on weekends. Leaving his boys was the hardest thing he had ever done. The experience left him feeling, he says, that between death and divorce there was not much to choose.

  During the coming half decade Plotkin’s personal life would improve dramatically. A weight of depression lifted from his shoulders. He began dating. A while after the separation, Koprowski, ever attuned to such details, noticed Plotkin’s new haircut and detected a livelier spirit emanating from his colleague. He declared him “a new man.”

  In 1976 Plotkin met Susan Lannon, a vivacious medical librarian who shared his love of classical music. The couple were married in 1979 in a ceremony in the anatomical museum on the first floor of the Wistar Institute, with a glass case of pelvic bones serving as the altar.

  A photo of Plotkin from the era seems to reveal the new man that Koprowski observed. He is standing in his office at the Children’s Hospital of Philadelphia, where he had become in 1969 the director of the Division of Infectious Diseases—an appointment he carried along with his Wistar position. He has one hand in his pocket, the other loosely on his desk. His posture is open to the camera. An unforced smile is on his face. He looks happy and relaxed.

  It was in this same office, back in the early autumn of 1973, soon after his “Waterloo” paper came out in the Journal of the American Medical Association, that Plotkin took a call out of the blue from Merck’s vaccine chief. After he hung up, it took quite a while for Hilleman’s word
s to sink in. A few days later Plotkin wrote a letter to Hayflick that began with the following words (he referred to Hilleman by the first name that those who knew the man tended to use: “Morris”):

  October 3, 1973

  Dear Len:

  I am writing to ask for several things but I would like to mention first of all that the other day I received a telephone call from Morris Hilleman asking me if the RA 27/3 strain is available to Merck. Needless to say I was left open-mouthed! . . . Please keep this under your hat for the moment.22

  Plotkin turned the request over to Koprowski, because the Wistar owned the patent. It was with Koprowski that Hilleman and a senior Merck licensing executive soon met.23 Koprowski gave them an estimate for the royalties the company would need to pay to the institute. After the requisite legal vetting, the deal was agreed, and RA 27/3 was licensed to Merck. Now all that Hilleman needed was WI-38 cells—the youngest ones, and plenty of them. So Hilleman did the obvious thing. He called Hayflick.

  • • •

  In the mid-1970s Hilleman, with the full resources of Merck behind him, ran Plotkin’s WI-38–propagated vaccine through the laboratory- and human-testing hoops needed for it to pass muster with the FDA’s new Bureau of Biologics and become a licensed vaccine. By early 1978 he and others, led by the University of Pennsylvania pediatrician Robert Weibel, had vaccinated nearly eight thousand antibody-lacking people with the Plotkin RA 27/3 vaccine in New York City, Philadelphia, and Costa Rica. They had also run head-to-head trials comparing Plotkin’s WI-38 cell–propagated vaccine against Merck’s HPV-77 duck-embryo vaccine in smaller groups of children and adult women. Hilleman would watch the vaccinees for two years before publishing the results, to make sure that the Plotkin vaccinees’ superior antibody levels persisted over time. They did.24

  The findings were unequivocal, Hilleman and his colleagues wrote in a paper that was published in 1980. The Plotkin vaccine “induced antibody in a larger proportion of individuals and at a substantially higher titer level” than did Merck’s HPV-77 duck vaccine. And in those people—98 percent of Plotkin vaccinees developed antibodies—the RA 27/3 vaccine induced the kinds of antibodies commonly found after natural infection.